list:\r\n commands: list = []\r\n\r\n arguments = CommandArguments()\r\n\r\n compiler_path: str = self.options.compiler_path\r\n flags_path: str = self.options.flags_path\r\n output_path: str = self.options.output_path\r\n\r\n if self.options.no_incremental_build:\r\n psc_paths: dict = self.psc_paths\r\n else:\r\n psc_paths = self._try_exclude_unmodified_scripts()\r\n\r\n # add .psc scripts whose .pex counterparts do not exist\r\n for object_name, script_path in self.missing_scripts.items():\r\n if object_name not in psc_paths.keys():\r\n psc_paths[object_name] = script_path\r\n\r\n source_import_paths = deepcopy(self.import_paths)\r\n\r\n # TODO: depth sorting solution is not foolproof! parse psc files for imports to determine command order\r\n for object_name, script_path in psc_paths.items():\r\n import_paths: list = self.import_paths\r\n\r\n if self.options.game_type != GameType.FO4:\r\n object_name = script_path\r\n\r\n # remove unnecessary import paths for script\r\n if self.options.game_type == GameType.FO4:\r\n for import_path in reversed(self.import_paths):\r\n if self._can_remove_folder(import_path, object_name, script_path):\r\n import_paths.remove(import_path)\r\n\r\n arguments.clear()\r\n arguments.append(compiler_path, enquote_value=True)\r\n arguments.append(object_name, enquote_value=True)\r\n arguments.append(flags_path, key='f', enquote_value=True)\r\n arguments.append(';'.join(import_paths), key='i', enquote_value=True)\r\n arguments.append(output_path, key='o', enquote_value=True)\r\n\r\n if self.options.game_type == GameType.FO4:\r\n # noinspection PyUnboundLocalVariable\r\n if self.release:\r\n arguments.append('-release')\r\n\r\n # noinspection PyUnboundLocalVariable\r\n if self.final:\r\n arguments.append('-final')\r\n\r\n if self.optimize:\r\n arguments.append('-op')\r\n\r\n arg_s = arguments.join()\r\n commands.append(arg_s)\r\n\r\n self.import_paths = source_import_paths\r\n\r\n return commands","def build(c):","def executable(output=None, sources=None, rule=None, slibs=[], libs='', **kwargs):\n objs = compile(\n sources = sources,\n bldprefix = output.name + '_',\n rule = rule,\n **kwargs)\n objs += [str(x) for x in slibs]\n w.build(str(output), 'executable',\n inputs = objs,\n variables = {'libs': libs})\n return str(output) # convenience","def get_compilation_args(\n filename: str, compilation_database: Optional[str] = None\n) -> List[str]:\n if not compilation_database:\n return []\n directory = os.path.dirname(compilation_database)\n comp_db = cindex.CompilationDatabase.fromDirectory(directory)\n commands = comp_db.getCompileCommands(filename)\n\n if not commands:\n return []\n\n # For now only handling the first file instance seen in the database.\n # First argument is compiler path, last is the file to compile\n args = list(commands[0].arguments)[1:-1]\n\n # Since things like includes and defines could be relative we force the working\n # directory.\n working_dir = commands[0].directory\n args.append(f\"-working-directory={working_dir}\")\n\n return args","def build(config_file, install=False):\n return exec_fn(lambda: _build(config_file, install))","def compile(self, **options):\n pass","def build(self):\n if os.path.exists(self.build_path) and not os.path.isdir(self.build_path):\n self._logger.error('Output path must be a folder if it already exists')\n return\n Compiler._mkdir(self.build_path)\n self._execute_backend_on_spec()","def compile(self):\n return None # pragma: no cover","def parsed(self):\n if not self._parsed:\n\n self._parsed = compile(self.content, self.path, \"exec\")\n\n return self._parsed","def get_buildall_target(self):\n raise NotImplementedError","def compilation(self, config: Config, files: List[str]) -> CallbackResult:\n return [], files","def compile_package(self):\n build_package = [\n self.mock,\n '--root=%s' % self.root,\n '--arch=%s' % self.arch,\n '--shell',\n '/build_package.sh',\n \n ]\n output, errors = self._run_command(build_package)","def compile_target(target_obj, search_paths, compile_path, ref_controller, globals_cached=None, **kwargs):\n start = time.time()\n compile_objs = target_obj[\"compile\"]\n ext_vars = target_obj[\"vars\"]\n target_name = ext_vars[\"target\"]\n\n if globals_cached:\n cached.from_dict(globals_cached)\n\n use_go_jsonnet = kwargs.get(\"use_go_jsonnet\", False)\n if use_go_jsonnet:\n logger.debug(\"Using go-jsonnet over jsonnet\")\n\n for comp_obj in compile_objs:\n input_type = comp_obj[\"input_type\"]\n output_path = comp_obj[\"output_path\"]\n input_params = comp_obj.setdefault(\"input_params\", {})\n\n if input_type == \"jinja2\":\n input_compiler = Jinja2(compile_path, search_paths, ref_controller, comp_obj)\n elif input_type == \"jsonnet\":\n input_compiler = Jsonnet(compile_path, search_paths, ref_controller, use_go=use_go_jsonnet)\n elif input_type == \"kadet\":\n input_compiler = Kadet(compile_path, search_paths, ref_controller, input_params=input_params)\n elif input_type == \"helm\":\n input_compiler = Helm(compile_path, search_paths, ref_controller, comp_obj)\n elif input_type == \"copy\":\n ignore_missing = comp_obj.get(\"ignore_missing\", False)\n input_compiler = Copy(compile_path, search_paths, ref_controller, ignore_missing)\n elif input_type == \"remove\":\n input_compiler = Remove(compile_path, search_paths, ref_controller)\n elif input_type == \"external\":\n input_compiler = External(compile_path, search_paths, ref_controller)\n if \"args\" in comp_obj:\n input_compiler.set_args(comp_obj[\"args\"])\n if \"env_vars\" in comp_obj:\n input_compiler.set_env_vars(comp_obj[\"env_vars\"])\n else:\n err_msg = 'Invalid input_type: \"{}\". Supported input_types: jsonnet, jinja2, kadet, helm, copy, remove, external'\n raise CompileError(err_msg.format(input_type))\n\n input_compiler.make_compile_dirs(target_name, output_path, **kwargs)\n input_compiler.compile_obj(comp_obj, ext_vars, **kwargs)\n\n logger.info(\"Compiled %s (%.2fs)\", target_obj[\"target_full_path\"], time.time() - start)","def _compile(self, tocompile, parameters):\n compiler = self.dialect.statement_compiler(self.dialect, tocompile, parameters)\n compiler.compile()\n return compiler","def build(self, *args, **kwargs):\n self._layer_counter = 0\n r = self._build_impl(*args, **kwargs)\n \n # Call the init functions \n if self._build_counter == 0:\n for initlayer in self._layers_to_init:\n if initlayer['initfnkwargs']:\n initlayer['initfn'](initlayer['layer'], **initlayer['initfnkwargs'])\n else:\n initlayer['initfn'](initlayer['layer'])\n \n self._build_counter += 1\n return r","def calculate_compile_sources(targets, is_thrift_target):\r\n\r\n basedirs = set()\r\n sources = set()\r\n def collect_sources(target):\r\n basedirs.add(target.target_base)\r\n sources.update(target.sources_relative_to_buildroot())\r\n for target in targets:\r\n target.walk(collect_sources, predicate=is_thrift_target)\r\n return basedirs, sources","def make(*args, **kwargs):\n return _uhd_swig.amsg_source_make(*args, **kwargs)","def _compile(self, source: str, filename: str) -> CodeType:\n return compile(source, filename, \"exec\") # type: ignore","def build_all(self):\n self.android_build()\n self.generate_patch_build('')\n self.generate_specs_build()\n self.generate_interfaces()","def CompileWithClosure(js_files, definitions, entry_points, output_file):\n\n cmd = [\n 'java', '-jar',\n './node_modules/google-closure-compiler-java/compiler.jar',\n '--language_out=ES5_STRICT', '--dependency_mode=PRUNE',\n '--js_output_file=%s' % output_file\n ]\n cmd += ['--entry_point=%s' % e for e in entry_points]\n cmd += ['--output_manifest=%s' % ('%s.manifest' % output_file)]\n cmd += [\n '../node_modules/google-closure-library/closure/**.js',\n '!../node_modules/google-closure-library/closure/**_test.js',\n '../node_modules/google-closure-library/third_party/closure/**.js',\n '!../node_modules/google-closure-library/third_party/closure/**_test.js'\n ]\n cmd += js_files\n cmd += definitions\n subprocess.check_call(cmd)","def cpp_compile(threads=2):\n cmake_ret = subprocess.run([\"cmake\", \"../\"], encoding='utf-8', stdout=subprocess.PIPE, stderr=subprocess.PIPE)\n output = cmake_ret.stdout\n errors = cmake_ret.stderr\n\n output = \"\"\n errors = \"\"\n\n make_ret = subprocess.run([\"make\", f\"-j{threads}\"], encoding='utf-8', stdout=subprocess.PIPE, stderr=subprocess.PIPE)\n output = make_ret.stdout\n errors = make_ret.stderr\n return (make_ret.returncode == 0, errors if make_ret != 0 else None)","def build(texFileLists):\n LaTeX_CMD = [\"latex -shell-escape -halt-on-error {texfile}.tex\",\n \"bibtex {texfile}.aux\",\n \"latex -shell-escape -halt-on-error {texfile}.tex\",\n \"pdflatex -shell-escape -halt-on-error {texfile}.tex\"]\n\n for texFile in texFileLists:\n filename, file_extension = os.path.splitext(texFile)\n for lcmd in LaTeX_CMD:\n try:\n call(lcmd.format(texfile=filename), shell=True)\n except:\n print \"Error occurs while building \", filename\n exit(1)","def compile(self, exclude=None, recurse=True, references=None, verbose=False):\n from clay.shell.core import lsgrep\n\n _os.chdir(self.directory)\n sources = self.sources\n\n if sources is None:\n sources = [_os.path.splitext(x)[0] for x in lsgrep(self.src_ext, self.directory, recurse=recurse)]\n if exclude is not None and len(exclude) > 0:\n sources = list(filter(lambda x: all(not(y in x) for y in exclude), sources))\n # if any flags, include them\n if len(self.flags) > 0:\n opt_str = '-' + ' -'.join(self.flags)\n else:\n opt_str = ''\n\n statechanged = False\n for src in sources:\n src_name = src + self.src_ext\n dst_name = src + self.dst_ext\n\n if not _os.path.exists(src_name):\n print(src, 'doesn\\'t exist, skipping...')\n continue\n\n src_mtime = _os.stat(src_name).st_mtime\n if _os.path.exists(dst_name):\n dst_mtime = _os.stat(dst_name).st_mtime\n else:\n dst_mtime = 0 # file doesn't exist\n\n # if edited more than five seconds ago, compile it\n if src_mtime - dst_mtime >= 5:\n print('Compiling ({}):'.format(self.compiler_name), src)\n cmd = self.compiler_name\n if self.compiler_name == 'csc': # C Sharp specific handling\n if references is not None:\n cmd += ' /r:' + ','.join(references)\n cmd += ' /out:{} '.format(dst_name)\n cmd += '{} \"{}\"'.format(opt_str, src_name)\n if verbose:\n print('cmd:', cmd)\n _os.system(cmd)\n statechanged = True\n\n if not statechanged:\n print('Nothing new to compile in \"{}\" when recurse={}'.format(self.directory, recurse))","def build(setup_kwargs: MutableMapping[Text, Any]) -> None:\n setup_kwargs.update({\n 'ext_modules': extensions,\n 'cmdclass': {\n 'build_ext': CMakeBuild\n }\n })","def build(self) -> ast.Module:\n self.build_missing_config()\n self.build_output()\n self.build_input()\n self.build_missing_imports()\n\n ret = ast.parse(\"\")\n ret.body = self.before_tree.body + self.tree.body + self.after_tree.body\n # Add fake line numbers to the nodes, so we can generate a traceback on error\n i = 0\n for node in dfs_walk(ret):\n if isinstance(node, ast.stmt):\n i += 1\n node.lineno = i\n node.end_lineno = i\n\n return ast.fix_missing_locations(ret)","def make(files: Dict[str, List[File]] = files) -> List[File]:\n progs = [\n make_prog(prog_path, c_files)\n for prog_path, c_files in files.items()\n ]\n return progs","def pyo():\n local('python -O -m compileall .')","def pyo():\n local('python -O -m compileall .')","def compile(self):\n raise NotImplementedError()","def _GetFlags(self, lines, build_dir):\n is_win = sys.platform == 'win32'\n flags_by_output = {}\n for line in lines:\n command_line = shlex.split(line.strip(), posix=not is_win)[1:]\n\n output_name = _FindAndRemoveArgWithValue(command_line, '-o')\n dep_name = _FindAndRemoveArgWithValue(command_line, '-MF')\n\n command_line = _MergeSpacedArgs(command_line, '-Xclang')\n\n cc_file = [x for x in command_line if x.endswith('.cc') or\n x.endswith('.c') or\n x.endswith('.cpp') or\n x.endswith('.mm') or\n x.endswith('.m')]\n if len(cc_file) != 1:\n self._skipped.append(command_line)\n continue\n assert len(cc_file) == 1\n\n if is_win:\n rsp_file = [x for x in command_line if x.endswith('.rsp')]\n assert len(rsp_file) <= 1\n if rsp_file:\n rsp_file = os.path.join(build_dir, rsp_file[0][1:])\n with open(rsp_file, \"r\") as open_rsp_file:\n command_line = shlex.split(open_rsp_file, posix=False)\n\n defines = [x for x in command_line if x.startswith('-D')]\n include_dirs = [x for x in command_line if x.startswith('-I')]\n dash_f = [x for x in command_line if x.startswith('-f')]\n warnings = \\\n [x for x in command_line if x.startswith('/wd' if is_win else '-W')]\n others = [x for x in command_line if x not in defines and \\\n x not in include_dirs and \\\n x not in dash_f and \\\n x not in warnings and \\\n x not in cc_file]\n\n for index, value in enumerate(include_dirs):\n if value == '-Igen':\n continue\n path = value[2:]\n if not os.path.isabs(path):\n path = os.path.join(build_dir, path)\n include_dirs[index] = '-I' + os.path.normpath(path)\n\n # GYP supports paths above the source root like <(DEPTH)/../foo while such\n # paths are unsupported by gn. But gn allows to use system-absolute paths\n # instead (paths that start with single '/'). Normalize all paths.\n cc_file = [os.path.normpath(os.path.join(build_dir, cc_file[0]))]\n\n # Filter for libFindBadConstructs.so having a relative path in one and\n # absolute path in the other.\n others_filtered = []\n for x in others:\n if x.startswith('-Xclang ') and \\\n (x.endswith('libFindBadConstructs.so') or \\\n x.endswith('libFindBadConstructs.dylib')):\n others_filtered.append(\n '-Xclang ' +\n os.path.join(os.getcwd(), os.path.normpath(\n os.path.join('out/gn_flags', x.split(' ', 1)[1]))))\n elif x.startswith('-B'):\n others_filtered.append(\n '-B' +\n os.path.join(os.getcwd(), os.path.normpath(\n os.path.join('out/gn_flags', x[2:]))))\n else:\n others_filtered.append(x)\n others = others_filtered\n\n flags_by_output[cc_file[0]] = {\n 'output': output_name,\n 'depname': dep_name,\n 'defines': sorted(defines),\n 'include_dirs': sorted(include_dirs), # TODO(scottmg): This is wrong.\n 'dash_f': sorted(dash_f),\n 'warnings': sorted(warnings),\n 'other': sorted(others),\n }\n return flags_by_output","def build(*onts, fail=False, n_jobs=9, write=True):\n from joblib import Parallel, delayed # importing in here saves 150ms\n tail = lambda:tuple()\n lonts = len(onts)\n if lonts > 1:\n for i, ont in enumerate(onts):\n if ont.__name__ == 'parcBridge':\n onts = onts[:-1]\n def tail(o=ont):\n return o.setup(),\n if i != lonts - 1:\n raise ValueError('parcBridge should be built last to avoid weird errors!')\n # ont_setup must be run first on all ontologies\n # or we will get weird import errors\n if n_jobs == 1 or True:\n return tuple(ont.make(fail=fail, write=write) for ont in\n tuple(ont.setup() for ont in onts) + tail())\n\n # have to use a listcomp so that all calls to setup()\n # finish before parallel goes to work\n return Parallel(n_jobs=n_jobs)(delayed(o.make)(fail=fail, write=write)\n for o in\n #[ont_setup(ont) for ont in onts])\n (tuple(Async()(deferred(ont.setup)()\n for ont in onts)) + tail()\n if n_jobs > 1\n else [ont.setup()\n for ont in onts]))","def build_step(self):\n run_cmd('./compile.sh', log_all=True, simple=True, log_ok=True)","def build(self, depVersions, output):\n raise NotImplementedError()","def here(*allowed_target_types):\r\n SourceRoot.register(buildfile_dir, *allowed_target_types)","def resolve(self, targets):\n context = self.context(target_roots=targets)\n self.execute(context)\n return context.products.get_data('compile_classpath')","def build(self, backend: Optional[str] = 'thread', copy_flow: bool = False, *args, **kwargs) -> 'Flow':\n\n op_flow = self._build_graph(copy_flow)\n\n if not backend:\n op_flow.logger.warning('no specified backend, build_level stays at %s, '\n 'and you can not run this flow.' % op_flow._build_level)\n elif backend in {'thread', 'process'}:\n op_flow._service_contexts.clear()\n for v in op_flow._service_nodes.values():\n p_args = v['parsed_args']\n p_args.parallel_backend = backend\n # for thread and process backend which runs locally, host_in and host_out should not be set\n p_args.host_in = BaseService.default_host\n p_args.host_out = BaseService.default_host\n op_flow._service_contexts.append((Flow._service2builder[v['service']], p_args))\n op_flow._build_level = Flow.BuildLevel.RUNTIME\n else:\n raise NotImplementedError('backend=%s is not supported yet' % backend)\n\n return op_flow","def assemble(compilers, paths, output, promoted):\n\n\t# assemble C compile command\n\tcommand = []\n\tcommand.append(compilers['cc'])\n\n\tcommand.extend(['-I', paths['include']])\n\n\tif output:\n\t\tcommand.extend(['-o', output])\n\n\tcommand.extend(promoted)\n\n\t# execute compiler\n\treturn subprocess.call(command)","def BuildFileTargets(target_list, build_file):\n return [p for p in target_list if BuildFile(p) == build_file]","def _compile_sources(sources: Union[Sequence[_SourceTypes], _SourceTypes]) -> Dict[str, Any]:\n if not isinstance(sources, Sequence):\n sources = [sources]\n\n compiled: Dict[str, Any] = {}\n for source in sources:\n if callable(source):\n source = source(compiled) # Pass what we have parsed so far to the function\n compiled = _merge_sources(dest=compiled, source=source)\n return compiled","def partial_build(src, lang, bsp_c3, opt_level, march, reporter):\n if lang == 'c3':\n srcs = [\n relpath('..', 'librt', 'io.c3'),\n bsp_c3,\n io.StringIO(src)]\n o2 = c3c(\n srcs, [], march, opt_level=opt_level,\n reporter=reporter, debug=True)\n objs = [o2]\n elif lang == 'bf':\n o3 = bfcompile(src, march, reporter=reporter)\n o2 = c3c(\n [bsp_c3], [], march, reporter=reporter)\n objs = [o2, o3]\n elif lang == 'c':\n o2 = c3c(\n [bsp_c3], [], march, reporter=reporter)\n coptions = COptions()\n include_path1 = relpath('..', 'librt', 'libc')\n coptions.add_include_path(include_path1)\n with open(relpath('..', 'librt', 'libc', 'lib.c'), 'r') as f:\n o3 = cc(\n f, march, coptions=coptions, debug=True,\n reporter=reporter)\n o4 = cc(\n io.StringIO(src), march, coptions=coptions, debug=True,\n reporter=reporter)\n objs = [o2, o3, o4]\n else:\n raise NotImplementedError('language not implemented')\n obj = link(\n objs, partial_link=True,\n use_runtime=True, reporter=reporter, debug=True)\n return obj","def get_compiled(self, flags=0):\n return re.compile(self.get_regex(), flags)","def getBuildSets():","def build_run_bowtie(bowtie_fp, input_contigs_fasta, output_index_fp, pe1_fastq, pe2_fastq, u_fastq, output_sam_fp):\n\n logging.debug('preparing bowtie build and align function calls')\n\n # bowtie build executable should be in the same directory as bowtie2 commands\n bowtie_build_fp = bowtie_fp + '-build'\n\n # build index files\n buildout, builderr = run_bowtie_build(bowtie_build_fp, input_contigs_fasta, output_index_fp)\n\n # once index files are built, run aligner using said index files\n stdout, stderr = run_bowtie(bowtie_fp, output_index_fp, pe1_fastq, pe2_fastq, u_fastq, output_sam_fp)\n\n return buildout, builderr, stdout, stderr","def run_compiler(text_or_file: str, compiler_suite: str, fail_when: ErrorCode = ERROR) -> Any:\n preprocessor, parser, ast, compiler = load_compiler_suite(compiler_suite)\n return compileDSL(text_or_file, preprocessor(), parser(), ast(), compiler(), fail_when)","def build(name, file, config, backend, debug):\n setup_lithops_logger(logging.DEBUG)\n\n verify_runtime_name(name)\n\n if config:\n config = load_yaml_config(config)\n\n config_ow = set_config_ow(backend, runtime_name=name)\n config = default_config(config, config_ow, load_storage_config=False)\n\n if config['lithops']['mode'] != SERVERLESS:\n raise Exception('\"lithops build\" command is only valid for serverless backends')\n\n compute_config = extract_serverless_config(config)\n compute_handler = ServerlessHandler(compute_config, None)\n compute_handler.build_runtime(name, file)","def builds(self):\n builds = [b for b in self.statuses() if b[\"type\"] == \"build\"]\n for build in builds:\n yield Build(build, **self._new_session_args)\n\n return","def __call__(self, *args, **kwargs):\n\n self._check_compiled()\n return self._compiled(*args, **kwargs)","def resolve(self,**bindings):\n for solution in self.resolve_all(**bindings):\n return solution","def build(parameters):\n\n\n print(\"In Build module\")","def build(self, progname):\n self.run_programm(self.COMPILED[self.progtype][0], \"%s %s %s\" %\\\n (progname, self.COMPILED[self.progtype][1], COMPILED_FILENAME ))\n\n compiled_progname=COMPILED_FILENAME\n return compiled_progname","def abstract_builder(pkg_name, name_list, return_list = False):\n # some handlers needs dicts (commands, beacons), while some need lists (encoders,decoders, etc)\n if return_list:\n ret_val = []\n else:\n ret_val = {}\n \n # Go through the string names and get the appropriate Class from the appropriate module.\n # Once you have that, do a dynamic import so we can use it, then map that class type\n # so we can instantiate the appropriate instance when going through a beaconing interation.\n for module_name in name_list:\n\n module_class = Controller.easy_import(pkg_name, module_name) # imports the class\n if return_list:\n ret_val.append(module_class) # adds the Class object to a list\n else:\n ret_val[module_name] = module_class # maps the Class object to the appropriate module name\n \n return ret_val","def process_input_files(inputs):\n for ifile in inputs:\n with open(ifile) as fin:\n exec(compile(fin.read(), ifile, 'exec'))","def build(self, **kwargs):\n if not self.built:\n with tf.name_scope(self.name):\n self._build(**kwargs)\n self.built = True\n return self","def actionBuild():\n\n #Init builder logger\n Builder.init()\n\n for target in Settings.targets:\n targetsToBuild, combineLibs, copyToOutput = Builder.getTargetGnPath(target)\n for platform in Settings.targetPlatforms:\n for cpu in Settings.targetCPUs:\n if System.checkIfCPUIsSupportedForPlatform(cpu,platform):\n for configuration in Settings.targetConfigurations:\n if not Summary.checkIfActionFailed(ACTION_PREPARE, target, platform, cpu, configuration):\n Logger.printStartActionMessage('Build ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)\n result = Builder.run(target, targetsToBuild, platform, cpu, configuration, combineLibs, copyToOutput)\n Summary.addSummary(ACTION_BUILD, target, platform, cpu, configuration, result, Builder.executionTime)\n if result != NO_ERROR:\n Logger.printEndActionMessage('Failed building ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.RED)\n #Terminate script execution if stopExecutionOnError is set to True in userdef\n shouldEndOnError(result)\n else:\n Logger.printEndActionMessage('Build ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration)\n else:\n Logger.printColorMessage('Build cannot run because preparation has failed for ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)\n Logger.printEndActionMessage('Build not run for ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)","def Run(self):\n self.__compile = False\n return self._Run()","def compile(self):\n if self._outputs is None or len(self._outputs) == 0:\n print \"Error! No templates to compile! Did you specify a valid configuration file?\"\n else:\n # for each output file in output list \n for output in self._outputs:\n # Generate context\n context = output[\"context\"]\n # Generate template stream for that output\n tmpl = self._env.get_template(output[\"source_filename\"]).stream(context)\n # Verify output directory\n output_filename = output[\"output_filename\"]\n output_path = abspath(dirname(output_filename))\n if not exists(output_path):\n makedirs(output_path)\n # Render it to the output file\n tmpl.dump(output_filename)","def compile(self, relations):\n # Compile each relation one-by-one, accumulating them into a list\n boxDeclarations = []\n for boxSyntax in self.boxes:\n box = boxSyntax.compile(boxDeclarations, relations)\n boxDeclarations.append(box)\n\n # With all the boxes compiled, compile all of the connections and use\n # both to construct and return a new relation declaration\n return RelationDeclaration(\n self.name,\n boxDeclarations,\n [c.compile(boxDeclarations) for c in self.connections])","def build_runtime(self, runtime_name, file):\n self.compute_handler.build_runtime(runtime_name, file)","def cl_program_from_files(context, filenames):\n source = ''\n for filename in filenames:\n source += open(os.path.join(CL_PATH, filename)).read()\n return cl.Program(context, source)","def run_thrifts(self):\r\n\r\n def is_py_thrift(target):\r\n return isinstance(target, PythonThriftLibrary)\r\n\r\n all_thrifts = set()\r\n\r\n def collect_sources(target):\r\n for source in target.sources:\r\n all_thrifts.add((target.target_base, source))\r\n\r\n self.target.walk(collect_sources, predicate=is_py_thrift)\r\n\r\n copied_sources = set()\r\n for base, relative_source in all_thrifts:\r\n abs_source = os.path.join(base, relative_source)\r\n copied_source = os.path.join(self._workdir, relative_source)\r\n\r\n safe_mkdir(os.path.dirname(copied_source))\r\n shutil.copyfile(abs_source, copied_source)\r\n copied_sources.add(self._modify_thrift(copied_source))\r\n\r\n for src in copied_sources:\r\n if not self._run_thrift(src):\r\n raise PythonThriftBuilder.CodeGenerationException(\"Could not generate .py from %s!\" % src)","def build(self):\n if not self.built:\n with tf.name_scope(self.name):\n self._build()\n self.built = True\n return self","def _do_build(self) -> List[types.Action]:\n return [\n docker_command.DockerRun(\n command=[\"/entrypoint.sh\", self.tag],\n builder=builder.GO_BUILDER,\n run_config=docker_command.default_run_config(\n constants.STORAGE_OPERATOR_ROOT / \"entrypoint.sh\"\n ),\n mounts=[\n utils.bind_mount(\n target=Path(\"/storage-operator\"),\n source=constants.STORAGE_OPERATOR_ROOT,\n ),\n # This container (through operator-sdk) will call `docker\n # build`, so we need to expose our Docker socket.\n utils.bind_mount(\n target=Path(\"/var/run/docker.sock\"),\n source=Path(\"/var/run/docker.sock\"),\n ),\n ],\n )\n ]"],"string":"[\n \"def compile_modules(base, output, source, bind=True):\\n return compile_files(base, output, source, bind, amd=True)\",\n \"def compile(base, output, source, bind=True, skip=False):\\n\\n # Paths\\n path = lambda *x: r.path(*[base] + list(x))\\n tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')\\n output_path = path(*output)\\n source_path = path(*source)\\n source_folder = path(*source[:-1])\\n\\n # Build\\n build = r.build()\\n build.notice('Typescript compile')\\n build.chdir(source_folder)\\n build.run('node', tsc_path, source_path, '--declaration', '--out', output_path)\\n\\n # Target\\n target = r.target(timeout=10)\\n target.pattern('.*\\\\.ts$', dirname(source_path), recurse=True)\\n\\n # Bind if required\\n if bind:\\n r.bind(target, build)\\n\\n # Run?\\n if ruffx.BUILD and not skip:\\n build.execute()\\n\\n return build\",\n \"def build(ctx: typer.Context):\\n from .tasks import build, main\\n\\n sys.argv = sys.argv[:1] + (ctx.args or [\\\"list\\\"])\\n main(vars(build))\",\n \"def cross_compile(*args, **kwargs):\\n return compile(*args, **kwargs)\",\n \"def parse(self, **globalargs):\\r\\n if self.buildfile not in ParseContext._parsed:\\r\\n buildfile_family = tuple(self.buildfile.family())\\r\\n\\r\\n pants_context = self.default_globals(Config.load())\\r\\n\\r\\n with ParseContext.activate(self):\\r\\n for buildfile in buildfile_family:\\r\\n self._active_buildfile = buildfile\\r\\n # We may have traversed a sibling already, guard against re-parsing it.\\r\\n if buildfile not in ParseContext._parsed:\\r\\n ParseContext._parsed.add(buildfile)\\r\\n\\r\\n buildfile_dir = os.path.dirname(buildfile.full_path)\\r\\n\\r\\n # TODO(John Sirois): XXX imports are done here to prevent a cycles\\r\\n from twitter.pants.targets.jvm_binary import Bundle\\r\\n from twitter.pants.targets.sources import SourceRoot\\r\\n\\r\\n class RelativeBundle(Bundle):\\r\\n def __init__(self, mapper=None, relative_to=None):\\r\\n super(RelativeBundle, self).__init__(\\r\\n base=buildfile_dir,\\r\\n mapper=mapper,\\r\\n relative_to=relative_to)\\r\\n\\r\\n # TODO(John Sirois): This is not build-dictionary friendly - rework SourceRoot to allow\\r\\n # allow for doc of both register (as source_root) and source_root.here(*types).\\r\\n class RelativeSourceRoot(object):\\r\\n @staticmethod\\r\\n def here(*allowed_target_types):\\r\\n \\\"\\\"\\\"Registers the cwd as a source root for the given target types.\\\"\\\"\\\"\\r\\n SourceRoot.register(buildfile_dir, *allowed_target_types)\\r\\n\\r\\n def __init__(self, basedir, *allowed_target_types):\\r\\n SourceRoot.register(os.path.join(buildfile_dir, basedir), *allowed_target_types)\\r\\n\\r\\n eval_globals = copy.copy(pants_context)\\r\\n eval_globals.update({\\r\\n 'ROOT_DIR': buildfile.root_dir,\\r\\n '__file__': buildfile.full_path,\\r\\n 'globs': partial(Fileset.globs, root=buildfile_dir),\\r\\n 'rglobs': partial(Fileset.rglobs, root=buildfile_dir),\\r\\n 'zglobs': partial(Fileset.zglobs, root=buildfile_dir),\\r\\n 'source_root': RelativeSourceRoot,\\r\\n 'bundle': RelativeBundle\\r\\n })\\r\\n eval_globals.update(globalargs)\\r\\n Compatibility.exec_function(buildfile.code(), eval_globals)\",\n \"def cli_build(\\n family: str,\\n versions: Union[str, List[str]],\\n ports: Union[str, List[str]],\\n boards: Union[str, List[str]],\\n clean: bool,\\n force: bool,\\n # stub_type: str,\\n):\\n\\n # lists please\\n versions = list(versions)\\n ports = list(ports)\\n boards = list(boards)\\n\\n # db = get_database(publish_path=CONFIG.publish_path, production=production)\\n log.info(f\\\"Build {family} {versions} {ports} {boards}\\\")\\n\\n results = build_multiple(\\n family=family,\\n versions=versions,\\n ports=ports,\\n boards=boards,\\n production=True, # use production database during build\\n force=force,\\n clean=clean,\\n )\\n # log the number of results with no error\\n log.info(f\\\"Built {len([r for r in results if not r['error']])} stubs\\\")\\n print(tabulate(results, headers=\\\"keys\\\"))\",\n \"def test(base, output, source, bind=True):\\n\\n # Paths\\n path = lambda *x: r.path(*[base] + list(x))\\n runner = 'node'\\n output_path = path(*output)\\n source_path = path(*source)\\n output_folder = path(*output[:-1])\\n\\n # Build\\n build = compile(base, output, source, bind=False, skip=True)\\n build.notice('Typescript tests')\\n build.chdir(output_folder)\\n build.run(runner, output_path)\\n\\n # Target\\n target = r.target(timeout=10)\\n target.pattern('.*\\\\.ts$', dirname(source_path), recurse=True)\\n\\n # Bind if required\\n if bind:\\n r.bind(target, build)\\n\\n # Run?\\n if ruffx.BUILD:\\n build.execute()\\n\\n return build\",\n \"def compile(self, targets=None, name=None, out_dir=None,\\n target=None, use_goma_module=False, **kwargs):\\n\\n targets = targets or self.c.compile_py.default_targets.as_jsonish()\\n assert isinstance(targets, (list, tuple))\\n\\n if self.c.gyp_env.GYP_DEFINES.get('clang', 0) == 1:\\n # Get the Clang revision before compiling.\\n self._clang_version = self.get_clang_version()\\n\\n goma_env = self.get_env()\\n goma_env.update(self.m.context.env)\\n ninja_env = goma_env.copy()\\n\\n goma_env['GOMA_CACHE_DIR'] = self.m.goma.default_cache_path\\n\\n # Enable goma DepsCache\\n goma_env['GOMA_DEPS_CACHE_FILE'] = \\\"goma_deps_cache\\\"\\n\\n if self.c.compile_py.mode:\\n if (self.c.compile_py.mode == 'google_chrome' or\\n self.c.compile_py.mode == 'official'):\\n ninja_env['CHROMIUM_BUILD'] = '_google_chrome'\\n\\n if self.c.compile_py.mode == 'official':\\n # Official builds are always Google Chrome.\\n ninja_env['CHROME_BUILD_TYPE'] = '_official'\\n\\n if self.c.compile_py.goma_hermetic:\\n goma_env['GOMA_HERMETIC'] = self.c.compile_py.goma_hermetic\\n if self.c.compile_py.goma_enable_remote_link:\\n goma_env['GOMA_ENABLE_REMOTE_LINK'] = 'true'\\n if self.c.compile_py.goma_enable_localoutputcache:\\n # Use per-slave cache. LocalOutputCache could use a lot of disks.\\n # To run GC for older caches, we should share the same build\\n # among builders.\\n goma_env['GOMA_LOCAL_OUTPUT_CACHE_DIR'] = (\\n self.m.path.join(self.m.goma.default_cache_path_per_slave,\\n \\\"localoutputcache\\\"))\\n if self.c.compile_py.goma_store_local_run_output:\\n goma_env['GOMA_STORE_LOCAL_RUN_OUTPUT'] = 'true'\\n if self.c.compile_py.goma_max_active_fail_fallback_tasks:\\n goma_env['GOMA_MAX_ACTIVE_FAIL_FALLBACK_TASKS'] = (\\n self.c.compile_py.goma_max_active_fail_fallback_tasks)\\n if (self.m.tryserver.is_tryserver or\\n self.c.compile_py.goma_failfast):\\n # We rely on goma to meet cycle time goals on the tryserver. It's better\\n # to fail early.\\n goma_env['GOMA_FAIL_FAST'] = 'true'\\n else:\\n goma_env['GOMA_ALLOWED_NETWORK_ERROR_DURATION'] = '1800'\\n\\n if self.c.TARGET_CROS_BOARD:\\n # Wrap 'compile' through 'cros chrome-sdk'\\n kwargs['wrapper'] = self.get_cros_chrome_sdk_wrapper()\\n\\n if self.m.platform.is_linux and self.c.TARGET_CROS_BOARD:\\n out_dir = 'out_%s' % self.c.TARGET_CROS_BOARD\\n elif out_dir is None:\\n out_dir = 'out'\\n\\n target_output_dir = self.m.path.abspath(\\n self.m.path.join(self.m.path['checkout'], out_dir,\\n target or self.c.build_config_fs))\\n\\n command = [str(self.m.depot_tools.ninja_path), '-w', 'dupbuild=err',\\n '-C', target_output_dir]\\n\\n if self.c.compile_py.show_ninja_stats:\\n command.extend(['-d', 'stats'])\\n\\n if self.c.compile_py.build_args:\\n command.extend(self.c.compile_py.build_args)\\n\\n # TODO(tikuta): Remove this and let goma module set '-j'\\n # inside build_with_goma.\\n if use_goma_module:\\n # Set -j just before 'with self.m.goma.build_with_goma('\\n # for ninja_log_command being set correctly if starting goma\\n # fails.\\n if self.c.compile_py.goma_high_parallel:\\n # This flag is set for experiment.\\n command += ['-j', 3 * self.m.goma.recommended_goma_jobs]\\n else:\\n command += ['-j', self.m.goma.recommended_goma_jobs]\\n\\n if targets is not None:\\n # Add build targets to command ('All', 'chrome' etc).\\n command += targets\\n\\n assert 'env' not in kwargs\\n\\n assert 'cwd' not in kwargs\\n\\n\\n if not use_goma_module:\\n compile_exit_status = 1\\n try:\\n with self.m.context(cwd=self.m.context.cwd or self.m.path['checkout']):\\n self._run_ninja(ninja_command=command,\\n name=name or 'compile',\\n ninja_env=ninja_env,\\n ninja_confirm_noop=self.c.compile_py.ninja_confirm_noop,\\n **kwargs)\\n compile_exit_status = 0\\n except self.m.step.StepFailure as e:\\n compile_exit_status = e.retcode\\n raise e\\n finally:\\n upload_ninja_log_args = [\\n '--gsutil-py-path', self.m.depot_tools.gsutil_py_path,\\n '--skip-sendgomatsmon',\\n '--ninja-log-outdir', target_output_dir,\\n '--ninja-log-command', str(command),\\n '--ninja-log-exit-status', compile_exit_status,\\n '--ninja-log-compiler', self.c.compile_py.compiler or 'unknown'\\n ]\\n self.m.python(\\n name='upload_ninja_log',\\n script=self.package_repo_resource(\\n 'scripts', 'slave', 'upload_goma_logs.py'),\\n args=upload_ninja_log_args)\\n\\n return\\n\\n try:\\n with self.m.context(cwd=self.m.context.cwd or self.m.path['checkout']):\\n self._run_ninja_with_goma(\\n name=name or 'compile',\\n ninja_command=command,\\n ninja_env=ninja_env,\\n goma_env=goma_env,\\n ninja_log_outdir=target_output_dir,\\n ninja_log_compiler=self.c.compile_py.compiler or 'goma',\\n ninja_confirm_noop=self.c.compile_py.ninja_confirm_noop,\\n **kwargs)\\n except self.m.step.StepFailure as e:\\n # Handle failures caused by goma.\\n step_result = self.m.step.active_result\\n failure_result_code = ''\\n\\n json_status = self.m.goma.jsonstatus['notice'][0]\\n\\n if (not json_status.get('infra_status')):\\n failure_result_code = 'GOMA_SETUP_FAILURE'\\n elif json_status['infra_status']['ping_status_code'] != 200:\\n failure_result_code = 'GOMA_PING_FAILURE'\\n elif json_status['infra_status'].get('num_user_error', 0) > 0:\\n failure_result_code = 'GOMA_BUILD_ERROR'\\n\\n if failure_result_code:\\n assert len(failure_result_code) <= 20\\n properties = self.m.step.active_result.presentation.properties\\n if not properties.get('extra_result_code'):\\n properties['extra_result_code'] = []\\n properties['extra_result_code'].append(failure_result_code)\\n raise self.m.step.InfraFailure('Infra compile failure: %s' % e)\\n\\n raise e\",\n \"def build_drivers(compilation_commands, linker_commands, kernel_src_dir,\\n target_arch, clang_path, llvm_link_path, llvm_bit_code_out, is_clang_build):\\n output_llvm_sh_file = os.path.join(llvm_bit_code_out, 'llvm_build.sh')\\n fp_out = open(output_llvm_sh_file, 'w')\\n fp_out.write(\\\"#!/bin/bash\\\\n\\\")\\n log_info(\\\"Writing all compilation commands to\\\", output_llvm_sh_file)\\n all_compilation_commands = []\\n obj_bc_map = {}\\n for curr_compilation_command in compilation_commands:\\n if is_clang_build:\\n wd, obj_file, bc_file, build_str = _get_llvm_build_str_from_llvm(clang_path, curr_compilation_command.curr_args,\\n kernel_src_dir, target_arch,\\n curr_compilation_command.work_dir,\\n curr_compilation_command.src_file,\\n curr_compilation_command.output_file,\\n llvm_bit_code_out)\\n else:\\n wd, obj_file, bc_file, build_str = _get_llvm_build_str(clang_path, curr_compilation_command.curr_args,\\n kernel_src_dir, target_arch,\\n curr_compilation_command.work_dir,\\n curr_compilation_command.src_file,\\n curr_compilation_command.output_file, llvm_bit_code_out)\\n all_compilation_commands.append((wd, build_str))\\n obj_bc_map[obj_file] = bc_file\\n fp_out.write(\\\"cd \\\" + wd + \\\";\\\" + build_str + \\\"\\\\n\\\")\\n fp_out.close()\\n\\n log_info(\\\"Got\\\", len(all_compilation_commands), \\\"compilation commands.\\\")\\n log_info(\\\"Running compilation commands in multiprocessing modea.\\\")\\n p = Pool(cpu_count())\\n return_vals = p.map(run_program_with_wd, all_compilation_commands)\\n log_success(\\\"Finished running compilation commands.\\\")\\n\\n output_llvm_sh_file = os.path.join(llvm_bit_code_out, 'llvm_link_cmds.sh')\\n fp_out = open(output_llvm_sh_file, 'w')\\n fp_out.write(\\\"#!/bin/bash\\\\n\\\")\\n log_info(\\\"Writing all linker commands to\\\", output_llvm_sh_file)\\n all_linker_commands = []\\n recursive_linker_commands = []\\n for curr_linked_command in linker_commands:\\n curr_ret_val = _get_llvm_link_str(llvm_link_path, kernel_src_dir,\\n curr_linked_command.input_files, obj_bc_map,\\n curr_linked_command.output_file,\\n curr_linked_command.work_dir, llvm_bit_code_out)\\n if curr_ret_val is not None:\\n wd, obj_file, bc_file, build_str = curr_ret_val\\n all_linker_commands.append((wd, build_str))\\n obj_bc_map[obj_file] = bc_file\\n fp_out.write(\\\"cd \\\" + wd + \\\";\\\" + build_str + \\\"\\\\n\\\")\\n else:\\n # these are recursive linker commands.\\n recursive_linker_commands.append(curr_linked_command)\\n\\n log_info(\\\"Got\\\", len(all_linker_commands), \\\"regular linker commands.\\\")\\n log_info(\\\"Running linker commands in multiprocessing mode.\\\")\\n p = Pool(cpu_count())\\n return_vals = p.map(run_program_with_wd, all_linker_commands)\\n log_success(\\\"Finished running linker commands.\\\")\\n\\n if len(recursive_linker_commands) > 0:\\n log_info(\\\"Got\\\", len(recursive_linker_commands), \\\" recursive linker commands.\\\")\\n _process_recursive_linker_commands(recursive_linker_commands, kernel_src_dir, llvm_link_path,\\n llvm_bit_code_out, obj_bc_map, fp_out)\\n fp_out.close()\\n\\n return True\",\n \"def build_command(self, args):\\n self._validate_common(args)\\n self._set_manifests(args)\\n\\n manifest = self._manager._recipes\\n\\n records = self._get_matching_records(args, manifest)\\n\\n if not len(records):\\n sys.exit(\\\"No matching items found.\\\")\\n\\n for record in records:\\n result = self._manager.build(record['_type'], **record)\\n print('The requested asset has been built: {}'.format(result['_path']))\\n\\n if len(records) > 1:\\n print('All files have been successfully built. Thank you.')\",\n \"def build():\\n local('wintersmith build')\",\n \"def build(working_directory=None, args=None):\\n from .buildme import main\\n if args is None:\\n args = []\\n return main(working_directory, args)\",\n \"def build(self, gyp_file, target=None, **kw):\\n raise NotImplementedError\",\n \"def build(_):\",\n \"def form_bowtie_build_cmd_list(bowtie_build_fp, input_contigs_fasta, output_index_fp):\\n if bowtie_build_fp is '':\\n raise ValueError('bowtie2_build_path name is empty')\\n if output_index_fp is '':\\n raise ValueError('output file name invalid. index_output_fp is None')\\n if input_contigs_fasta is '' or input_contigs_fasta is None:\\n raise ValueError('no fasta file found')\\n\\n # required arguments\\n calls_args_list = [bowtie_build_fp, input_contigs_fasta, output_index_fp]\\n\\n return calls_args_list\",\n \"def build(which):\\n return subprocess.Popen([p['paths']['w2l'],'obj',os.path.abspath(which)]).wait()\",\n \"def _sphinx_build(self, kind: str):\\n if kind not in (\\\"html\\\", \\\"latex\\\"):\\n raise ValueError(f\\\"kind must be html or latex, not {kind}\\\")\\n\\n cmd = [\\\"sphinx-build\\\", \\\"-b\\\", kind]\\n if self.num_jobs:\\n cmd += [\\\"-j\\\", self.num_jobs]\\n if self.warnings_are_errors:\\n cmd += [\\\"-W\\\", \\\"--keep-going\\\"]\\n if self.verbosity:\\n cmd.append(f\\\"-{'v' * self.verbosity}\\\")\\n cmd += [\\n \\\"-d\\\",\\n os.path.join(BUILD_PATH, \\\"doctrees\\\"),\\n SOURCE_PATH,\\n os.path.join(BUILD_PATH, kind),\\n ]\\n return subprocess.call(cmd)\",\n \"def _sphinx_build(self, kind: str):\\n if kind not in (\\\"html\\\", \\\"latex\\\"):\\n raise ValueError(f\\\"kind must be html or latex, not {kind}\\\")\\n\\n cmd = [\\\"sphinx-build\\\", \\\"-b\\\", kind]\\n if self.num_jobs:\\n cmd += [\\\"-j\\\", self.num_jobs]\\n if self.warnings_are_errors:\\n cmd += [\\\"-W\\\", \\\"--keep-going\\\"]\\n if self.verbosity:\\n cmd.append(f\\\"-{'v' * self.verbosity}\\\")\\n cmd += [\\n \\\"-d\\\",\\n os.path.join(BUILD_PATH, \\\"doctrees\\\"),\\n SOURCE_PATH,\\n os.path.join(BUILD_PATH, kind),\\n ]\\n return subprocess.call(cmd)\",\n \"def build_all():\\n client = TestClient()\\n client.save({\\\"conanfile.py\\\": GenConanfile().with_setting(\\\"build_type\\\")})\\n client.run(\\\"export . --name=foo --version=1.0 --user=user --channel=testing\\\")\\n client.save({\\\"conanfile.py\\\": GenConanfile().with_require(\\\"foo/1.0@user/testing\\\")\\n .with_setting(\\\"build_type\\\")})\\n client.run(\\\"export . --name=bar --version=1.0 --user=user --channel=testing\\\")\\n client.save({\\\"conanfile.py\\\": GenConanfile().with_require(\\\"foo/1.0@user/testing\\\")\\n .with_require(\\\"bar/1.0@user/testing\\\")\\n .with_setting(\\\"build_type\\\")})\\n client.run(\\\"export . --name=foobar --version=1.0 --user=user --channel=testing\\\")\\n client.run(\\\"install --requires=foobar/1.0@user/testing --build='*'\\\")\\n return client\",\n \"def compile(self, args):\\n if args not in self._compileinfos:\\n cres = compile_with_dppl(self.py_func, None, args, debug=self.debug)\\n func = cres.library.get_function(cres.fndesc.llvm_func_name)\\n cres.target_context.mark_ocl_device(func)\\n first_definition = not self._compileinfos\\n self._compileinfos[args] = cres\\n libs = [cres.library]\\n\\n if first_definition:\\n # First definition\\n cres.target_context.insert_user_function(self, cres.fndesc,\\n libs)\\n else:\\n cres.target_context.add_user_function(self, cres.fndesc, libs)\\n\\n else:\\n cres = self._compileinfos[args]\\n\\n return cres.signature\",\n \"def generate_cython_transpile(self, target: build.BuildTarget) -> \\\\\\n T.Tuple[T.MutableMapping[str, File], T.MutableMapping[str, File], T.List[str]]:\\n static_sources: T.MutableMapping[str, File] = OrderedDict()\\n generated_sources: T.MutableMapping[str, File] = OrderedDict()\\n cython_sources: T.List[str] = []\\n\\n cython = target.compilers['cython']\\n\\n args: T.List[str] = []\\n args += cython.get_always_args()\\n args += cython.get_buildtype_args(target.get_option(OptionKey('buildtype')))\\n args += cython.get_debug_args(target.get_option(OptionKey('debug')))\\n args += cython.get_optimization_args(target.get_option(OptionKey('optimization')))\\n args += cython.get_option_compile_args(target.get_options())\\n args += self.build.get_global_args(cython, target.for_machine)\\n args += self.build.get_project_args(cython, target.subproject, target.for_machine)\\n args += target.get_extra_args('cython')\\n\\n ext = target.get_option(OptionKey('language', machine=target.for_machine, lang='cython'))\\n\\n pyx_sources = [] # Keep track of sources we're adding to build\\n\\n for src in target.get_sources():\\n if src.endswith('.pyx'):\\n output = os.path.join(self.get_target_private_dir(target), f'{src}.{ext}')\\n element = NinjaBuildElement(\\n self.all_outputs, [output],\\n self.compiler_to_rule_name(cython),\\n [src.absolute_path(self.environment.get_source_dir(), self.environment.get_build_dir())])\\n element.add_item('ARGS', args)\\n self.add_build(element)\\n # TODO: introspection?\\n cython_sources.append(output)\\n pyx_sources.append(element)\\n else:\\n static_sources[src.rel_to_builddir(self.build_to_src)] = src\\n\\n header_deps = [] # Keep track of generated headers for those sources\\n for gen in target.get_generated_sources():\\n for ssrc in gen.get_outputs():\\n if isinstance(gen, GeneratedList):\\n ssrc = os.path.join(self.get_target_private_dir(target), ssrc)\\n else:\\n ssrc = os.path.join(gen.get_subdir(), ssrc)\\n if ssrc.endswith('.pyx'):\\n output = os.path.join(self.get_target_private_dir(target), f'{ssrc}.{ext}')\\n element = NinjaBuildElement(\\n self.all_outputs, [output],\\n self.compiler_to_rule_name(cython),\\n [ssrc])\\n element.add_item('ARGS', args)\\n self.add_build(element)\\n pyx_sources.append(element)\\n # TODO: introspection?\\n cython_sources.append(output)\\n else:\\n generated_sources[ssrc] = mesonlib.File.from_built_file(gen.get_subdir(), ssrc)\\n # Following logic in L883-900 where we determine whether to add generated source\\n # as a header(order-only) dep to the .so compilation rule\\n if not self.environment.is_source(ssrc) and \\\\\\n not self.environment.is_object(ssrc) and \\\\\\n not self.environment.is_library(ssrc) and \\\\\\n not modules.is_module_library(ssrc):\\n header_deps.append(ssrc)\\n for source in pyx_sources:\\n source.add_orderdep(header_deps)\\n\\n return static_sources, generated_sources, cython_sources\",\n \"def compile_regex(self, paths):\\r\\n if isinstance(paths, list):\\r\\n ret = []\\r\\n for regex in paths:\\r\\n ret.append(re.compile(regex, re.I))\\r\\n return ret\\r\\n else:\\r\\n return re.compile(paths, re.I)\",\n \"def get_artifacts(self):\\n\\n self.clml_modules = list(\\n filter(lambda mod: mod.type_key == \\\"clml\\\", self.libm.get_lib().imported_modules)\\n )\\n self.clml_builds[\\\"file_header\\\"] = [self.MakeFileHeader.substitute()]\\n\\n for cmod in self.clml_modules:\\n (sub_module_name, clml_code) = CLMLGetSubModuleSrc(cmod).get_src()\\n self.clml_builds[sub_module_name] = clml_code\\n\\n main_code = []\\n main_code.append(\\n \\\"\\\"\\\"\\n std::vector<CLMLRunner> BuildModules(ToolArgs& args,\\n cl_platform_id arg_platform,\\n cl_context arg_context,\\n cl_device_id arg_device_id,\\n cl_command_queue arg_queue) {\\n std::vector<CLMLRunner> runners;\\\"\\\"\\\"\\n )\\n for key, val in self.clml_builds.items():\\n if key != \\\"file_header\\\":\\n main_code.append(\\n \\\"runners.push_back(\\\"\\n + key\\n + '(\\\"'\\n + key\\n + '\\\", args, arg_platform, arg_context, arg_device_id, arg_queue));'\\n )\\n main_code.append(\\\"return runners;}\\\")\\n self.clml_builds[\\\"MainBuild\\\"] = main_code\\n\\n for key, val in self.clml_builds.items():\\n self.gen_src.extend(val)\\n\\n return (self.get_clml_params(), self.gen_src)\",\n \"def compile(self):\\n\\n for v in list(self.graph.values()):\\n if isinstance(v, Graph):\\n v.compile()\\n self._compiled = compile_graph(self.graph)\\n return self._compiled\",\n \"def build(*args, **kwargs):\\n\\n\\treturn Spider(*args, **kwargs)\",\n \"def build(path=None, output=None):\\n conf.load(path)\\n cache = Cache()\\n if cache.processing_errors():\\n for file_name, error in cache.processing_errors():\\n message = \\\"error processing source file '%s' - %s\\\"\\n logger.error(message % (file_name, error))\\n if output:\\n conf.set('build_path', output)\\n logger.info('build directory: ' + conf.get('build_path'))\\n for builder in builders.order():\\n builder(cache)\",\n \"def build(c, force=None):\\n for sp_ns in ns_foreach_task_subdir(c):\\n print(\\\"-- running build in \\\", os.getcwd())\\n\\n # sp_ns.tasks.build(c, force)\\n c.run('invoke build')\",\n \"def build():\",\n \"def build(self) -> None:\\n def do_process(fname) -> bool:\\n for sfx in skip_suffixes:\\n if fname.endswith(sfx):\\n return False\\n return True\\n\\n for dirpath, _, fnames in os.walk(self.template_dir):\\n for fname in fnames:\\n if do_process(fname):\\n self.process(dirpath, fname)\",\n \"def build_pipeline(roots):\\n if not isinstance(roots, list):\\n roots = [roots]\\n\\n build_graph = BuildGraph()\\n build_graph.add_targets(roots)\\n\\n return build_graph\",\n \"async def builds(self, *args, **kwargs):\\n\\n return await self._makeApiCall(self.funcinfo[\\\"builds\\\"], *args, **kwargs)\",\n \"def compile_build_files(ctx):\\n project_dir = Path(__file__).parent\\n\\n build_dir = project_dir / \\\"build\\\"\\n ninja_dir = project_dir / \\\"build/ninja\\\"\\n artifacts_dir = project_dir / \\\"build/artifacts\\\"\\n\\n if artifacts_dir.exists():\\n shutil.rmtree(artifacts_dir)\\n if ninja_dir.exists():\\n shutil.rmtree(ninja_dir)\\n\\n os.makedirs(artifacts_dir)\\n os.makedirs(ninja_dir)\\n\\n call_cmake = (\\n f\\\"cmake \\\"\\n f\\\"-DCMAKE_BUILD_TYPE=Release \\\"\\n f'-G Ninja \\\"{build_dir}\\\" '\\n f\\\"-DPYTHON_EXECUTABLE={sys.executable} \\\"\\n )\\n call_ninja = \\\"ninja -j 8\\\"\\n call_install = \\\"ninja install\\\"\\n\\n with ctx.cd(str(project_dir / \\\"build/ninja\\\")):\\n if sys.platform == \\\"win32\\\":\\n paths = (\\n os.path.expandvars(\\n r\\\"${PROGRAMFILES(X86)}\\\\Microsoft Visual Studio\\\\2017\\\\Community\\\\VC\\\\Auxiliary\\\\Build\\\\vcvarsall.bat\\\"\\n ),\\n os.path.expandvars(\\n r\\\"${PROGRAMFILES(X86)}\\\\Microsoft Visual Studio\\\\2017\\\\BuildTools\\\\VC\\\\Auxiliary\\\\Build\\\\vcvarsall.bat\\\"\\n ),\\n os.path.expandvars(\\n r\\\"${PROGRAMFILES(X86)}\\\\Microsoft Visual Studio\\\\2017\\\\Professional\\\\VC\\\\Auxiliary\\\\Build\\\\vcvarsall.bat\\\"\\n ),\\n os.path.expandvars(\\n r\\\"${PROGRAMFILES(X86)}\\\\Microsoft Visual Studio\\\\2017\\\\WDExpress\\\\VC\\\\Auxiliary\\\\Build\\\\vcvarsall.bat\\\"\\n ),\\n # Path for vcvars on GithubAction\\n r\\\"C:\\\\Program Files (x86)\\\\Microsoft Visual Studio\\\\2019\\\\Enterprise\\\\VC\\\\Auxiliary\\\\Build\\\\vcvars64.bat\\\",\\n )\\n for msvc_path in paths:\\n if os.path.isfile(msvc_path):\\n break\\n else:\\n raise RuntimeError(\\n \\\"Couldn't find MSVC compiler in any of:\\\\n{}\\\".format(\\\"- \\\" + \\\"\\\\n- \\\".join(paths))\\n )\\n\\n call_cmd = f'call \\\"{msvc_path}\\\" amd64'\\n ctx.run(command=call_cmd + \\\"&\\\" + call_cmake + \\\"&&\\\" + call_ninja + \\\"&&\\\" + call_install)\\n\\n else:\\n ctx.run(command=call_cmake + \\\"&&\\\" + call_ninja + \\\"&&\\\" + call_install)\",\n \"def try_build_target(self, makefile_path, files=None, c_compiler=\\\"cc\\\", cxx_compiler=\\\"g++\\\"):\\n if len(self.makefile_am_info.get(makefile_path, dict())) == 0:\\n return\\n make_info = self.makefile_am_info.get(makefile_path, dict())\\n path = os.path.dirname(makefile_path)\\n if \\\"target\\\" not in make_info:\\n return\\n\\n targets = make_info.get(\\\"target\\\", dict())\\n for target_key, target in targets.items():\\n # Build files\\n if len(target.get(\\\"files\\\", list())) != 0:\\n files = target.get(\\\"files\\\", list())\\n # Avoid unknown list contains list problem\\n move_to_top = lambda x: (z for y in x for z in (isinstance(y, list) and move_to_top(y) or [y]))\\n files = list(move_to_top(files))\\n files_str = \\\" \\\".join(files)\\n files = files_str.split(\\\" \\\\t\\\\n\\\")\\n for file_line in files:\\n sub_files = re.split(r\\\"\\\\s+\\\", file_line)\\n c_files = filter(lambda file_name: True if file_name.split(\\\".\\\")[-1] == \\\"c\\\" else False,\\n sub_files)\\n cpp_files = filter(lambda file_name: True if file_name.split(\\\".\\\")[-1] in \\\\\\n [\\\"cxx\\\", \\\"cpp\\\", \\\"cc\\\"] else False, sub_files)\\n target[\\\"c_files\\\"] = list(map(lambda file: os.path.join(path, file), c_files))\\n target[\\\"cxx_files\\\"] = list(map(lambda file: os.path.join(path, file), cpp_files))\\n # Get test case\\n logger.info(\\\"## Get case from files.\\\")\\n c_case = None\\n cxx_case = None\\n if \\\"c_files\\\" in target and len(target.get(\\\"c_files\\\", list())) != 0:\\n c_files = target.get(\\\"c_files\\\", list())\\n idx = random.randint(0, len(c_files) - 1)\\n c_case = os.path.join(path, c_files[idx])\\n if \\\"cxx_files\\\" in target and len(target.get(\\\"cxx_files\\\", list())) != 0:\\n cxx_files = target.get(\\\"cxx_files\\\", list())\\n idx = random.randint(0, len(cxx_files) - 1)\\n cxx_case = os.path.join(path, cxx_files[idx])\\n\\n # Try build\\n if c_case is None and cxx_case is None:\\n return\\n else:\\n if \\\"flags\\\" not in target:\\n target[\\\"c_flags\\\"] = {}\\n target[\\\"cpp_flags\\\"] = {}\\n continue\\n\\n if \\\"CPPFLAGS\\\" in target.get(\\\"flags\\\", dict()):\\n cppsorted_flags = target[\\\"flags\\\"].get(\\\"CPPFLAGS\\\", list())\\n else:\\n cppsorted_flags = [\\\"\\\"]\\n\\n c_compiler_status = False\\n cxx_compiler_status = False\\n sorted_cppsorted_flags = sort_flags_line(cppsorted_flags)\\n\\n flags_dict = target.get(\\\"flags\\\", dict())\\n c_flags_line = flags_dict.get(\\\"CFLAGS\\\", [\\\"\\\"])\\n sorted_c_flags = sort_flags_line(c_flags_line)\\n\\n cxx_flags_line = flags_dict.get(\\\"CXXFLAGS\\\", [\\\"\\\"])\\n sorted_cxx_flags = sort_flags_line(cxx_flags_line)\\n\\n final_c_flags_lines = []\\n final_cxx_flags_lines = []\\n for lines in sorted_cppsorted_flags:\\n default_includes, default_macros, default_flags = format_flags(lines, path)\\n default_macros.append(\\\"HAVE_CONFIG_H\\\")\\n\\n # Sorted by macros count\\n if len(sorted_c_flags) == 0:\\n final_c_flags_lines.append(([], [], []))\\n for c_lines in sorted_c_flags:\\n includes, macros, flags = format_flags(c_lines, path)\\n includes.extend(default_includes)\\n macros.extend(default_macros)\\n flags.extend(default_flags)\\n final_flags_line = (includes, macros, flags)\\n final_c_flags_lines.append(final_flags_line)\\n\\n if len(sorted_cxx_flags) == 0:\\n final_cxx_flags_lines.append(([], [], []))\\n for cxx_lines in sorted_cxx_flags:\\n includes, macros, flags = format_flags(cxx_lines, path)\\n includes.extend(default_includes)\\n macros.extend(default_macros)\\n flags.extend(default_flags)\\n final_flags_line = (includes, macros, flags)\\n final_cxx_flags_lines.append(final_flags_line)\\n\\n final_c_flags_lines.sort(key=lambda x: len(x[1]))\\n final_cxx_flags_lines.sort(key=lambda x: len(x[1]))\\n default_cxx_flags = {\\n \\\"macros\\\": final_cxx_flags_lines[0][1],\\n \\\"includes\\\": final_cxx_flags_lines[0][0],\\n \\\"flags\\\": final_cxx_flags_lines[0][2]\\n }\\n default_c_flags = {\\n \\\"macros\\\": final_c_flags_lines[0][1],\\n \\\"includes\\\": final_c_flags_lines[0][0],\\n \\\"flags\\\": final_c_flags_lines[0][2]\\n }\\n\\n for flags_type in (\\\"C\\\", \\\"CXX\\\"):\\n flags_lines = final_c_flags_lines if flags_type == \\\"C\\\" else final_cxx_flags_lines\\n case = c_case if flags_type == \\\"C\\\" else cxx_case\\n if case is None and flags_type == \\\"C\\\":\\n c_compiler_status = True\\n continue\\n if case is None and flags_type == \\\"CXX\\\":\\n cxx_compiler_status = True\\n continue\\n compiler = c_compiler if flags_type == \\\"C\\\" else cxx_compiler\\n flags_type_name = \\\"c_flags\\\" if flags_type == \\\"C\\\" else \\\"cxx_flags\\\"\\n\\n for (includes, macros, flags) in flags_lines:\\n include_line = \\\" \\\".join(map(\\\"-I{}\\\".format, includes))\\n macros_line = \\\" \\\".join(map(\\\"-D{}\\\".format, macros))\\n flags_line = \\\" \\\".join(flags)\\n cmd = \\\"{} -c {} -o {} {} {} {}\\\".format(\\n compiler, case, os.path.join(self._build_path, case + \\\".o\\\"),\\n include_line, macros_line, flags_line\\n )\\n logger.debug(cmd)\\n (returncode, out, err) = capture_util.subproces_calling(cmd, path)\\n if returncode == 0:\\n logger.info(\\\"Try compile for target: %s success.\\\" % target_key)\\n if flags_type == \\\"C\\\":\\n save_status = False if c_compiler_status else True\\n c_compiler_status = True\\n else:\\n save_status = False if cxx_compiler_status else True\\n cxx_compiler_status = True\\n\\n if save_status:\\n target[flags_type_name] = {\\n \\\"definitions\\\": macros,\\n \\\"includes\\\": includes,\\n \\\"flags\\\": flags,\\n }\\n\\n if len(target.get(\\\"c_flags\\\", dict())) == 0:\\n target[\\\"c_flags\\\"] = default_c_flags\\n if len(target.get(\\\"cxx_flags\\\", dict())) == 0:\\n target[\\\"cxx_flags\\\"] = default_cxx_flags\\n return\",\n \"def __init__(self, paths):\\n self._makefile = pymake.data.Makefile(env={}, justprint=True)\\n self._dependencies = {}\\n\\n for f in paths:\\n for stmt in pymake.parser.parsefile(f):\\n self._ExecStatement(stmt)\",\n \"def build(ctx):\\n generate_build_files(ctx)\\n compile_build_files(ctx)\\n _package_plugins(ctx)\",\n \"def run(self, argv, binds=None, readonly_binds=None):\\n\\n nspawn_ro_binds = []\\n\\n # make osbuild API-calls accessible to the container\\n nspawn_ro_binds.append(f\\\"{self.api}:/run/osbuild/api\\\")\\n\\n # We want to execute our stages and other scripts in the container. So\\n # far, we do not install osbuild as a package in the container, but\\n # provide it from the outside. Therefore, we need to provide `libdir`\\n # via bind-mount. Furthermore, a system-installed `libdir` has the\\n # python packages separate in `site-packages`, so we need to bind-mount\\n # them as well.\\n # In the future, we want to work towards mandating an osbuild package to\\n # be installed in the container, so the build is self-contained and does\\n # not take scripts from the host. However, this requires osbuild\\n # packaged for those containers. Furthermore, we want to keep supporting\\n # the current import-model for testing and development.\\n if self.libdir is not None:\\n # caller-specified `libdir` must be self-contained\\n nspawn_ro_binds.append(f\\\"{self.libdir}:/run/osbuild/lib\\\")\\n else:\\n # system `libdir` requires importing the python module\\n nspawn_ro_binds.append(\\\"/usr/lib/osbuild:/run/osbuild/lib\\\")\\n modorigin = importlib.util.find_spec(\\\"osbuild\\\").origin\\n modpath = os.path.dirname(modorigin)\\n nspawn_ro_binds.append(f\\\"{modpath}:/run/osbuild/lib/osbuild\\\")\\n\\n return subprocess.run([\\n \\\"systemd-nspawn\\\",\\n \\\"--quiet\\\",\\n \\\"--register=no\\\",\\n \\\"--keep-unit\\\",\\n \\\"--as-pid2\\\",\\n \\\"--link-journal=no\\\",\\n \\\"--capability=CAP_MAC_ADMIN\\\", # for SELinux labeling\\n f\\\"--directory={self.root}\\\",\\n \\\"--setenv=PYTHONPATH=/run/osbuild/lib\\\",\\n *[f\\\"--bind-ro={b}\\\" for b in nspawn_ro_binds],\\n *[f\\\"--bind={b}\\\" for b in (binds or [])],\\n *[f\\\"--bind-ro={b}\\\" for b in (readonly_binds or [])],\\n f\\\"/run/osbuild/lib/runners/{self.runner}\\\"\\n ] + argv, check=False, stdin=subprocess.DEVNULL)\",\n \"def build_from_c_and_cpp_files(extensions):\\n for extension in extensions:\\n sources = []\\n for sfile in extension.sources:\\n path, ext = os.path.splitext(sfile)\\n if ext in ('.pyx', '.py'):\\n if extension.language == 'c++':\\n ext = '.cpp'\\n else:\\n ext = '.c'\\n sfile = path + ext\\n sources.append(sfile)\\n extension.sources = sources\",\n \"def buildAll(self, packages):\\r\\n return [self.build(package) for package in packages]\",\n \"def build_commands(self) -> list:\\r\\n commands: list = []\\r\\n\\r\\n arguments = CommandArguments()\\r\\n\\r\\n compiler_path: str = self.options.compiler_path\\r\\n flags_path: str = self.options.flags_path\\r\\n output_path: str = self.options.output_path\\r\\n\\r\\n if self.options.no_incremental_build:\\r\\n psc_paths: dict = self.psc_paths\\r\\n else:\\r\\n psc_paths = self._try_exclude_unmodified_scripts()\\r\\n\\r\\n # add .psc scripts whose .pex counterparts do not exist\\r\\n for object_name, script_path in self.missing_scripts.items():\\r\\n if object_name not in psc_paths.keys():\\r\\n psc_paths[object_name] = script_path\\r\\n\\r\\n source_import_paths = deepcopy(self.import_paths)\\r\\n\\r\\n # TODO: depth sorting solution is not foolproof! parse psc files for imports to determine command order\\r\\n for object_name, script_path in psc_paths.items():\\r\\n import_paths: list = self.import_paths\\r\\n\\r\\n if self.options.game_type != GameType.FO4:\\r\\n object_name = script_path\\r\\n\\r\\n # remove unnecessary import paths for script\\r\\n if self.options.game_type == GameType.FO4:\\r\\n for import_path in reversed(self.import_paths):\\r\\n if self._can_remove_folder(import_path, object_name, script_path):\\r\\n import_paths.remove(import_path)\\r\\n\\r\\n arguments.clear()\\r\\n arguments.append(compiler_path, enquote_value=True)\\r\\n arguments.append(object_name, enquote_value=True)\\r\\n arguments.append(flags_path, key='f', enquote_value=True)\\r\\n arguments.append(';'.join(import_paths), key='i', enquote_value=True)\\r\\n arguments.append(output_path, key='o', enquote_value=True)\\r\\n\\r\\n if self.options.game_type == GameType.FO4:\\r\\n # noinspection PyUnboundLocalVariable\\r\\n if self.release:\\r\\n arguments.append('-release')\\r\\n\\r\\n # noinspection PyUnboundLocalVariable\\r\\n if self.final:\\r\\n arguments.append('-final')\\r\\n\\r\\n if self.optimize:\\r\\n arguments.append('-op')\\r\\n\\r\\n arg_s = arguments.join()\\r\\n commands.append(arg_s)\\r\\n\\r\\n self.import_paths = source_import_paths\\r\\n\\r\\n return commands\",\n \"def build(c):\",\n \"def executable(output=None, sources=None, rule=None, slibs=[], libs='', **kwargs):\\n objs = compile(\\n sources = sources,\\n bldprefix = output.name + '_',\\n rule = rule,\\n **kwargs)\\n objs += [str(x) for x in slibs]\\n w.build(str(output), 'executable',\\n inputs = objs,\\n variables = {'libs': libs})\\n return str(output) # convenience\",\n \"def get_compilation_args(\\n filename: str, compilation_database: Optional[str] = None\\n) -> List[str]:\\n if not compilation_database:\\n return []\\n directory = os.path.dirname(compilation_database)\\n comp_db = cindex.CompilationDatabase.fromDirectory(directory)\\n commands = comp_db.getCompileCommands(filename)\\n\\n if not commands:\\n return []\\n\\n # For now only handling the first file instance seen in the database.\\n # First argument is compiler path, last is the file to compile\\n args = list(commands[0].arguments)[1:-1]\\n\\n # Since things like includes and defines could be relative we force the working\\n # directory.\\n working_dir = commands[0].directory\\n args.append(f\\\"-working-directory={working_dir}\\\")\\n\\n return args\",\n \"def build(config_file, install=False):\\n return exec_fn(lambda: _build(config_file, install))\",\n \"def compile(self, **options):\\n pass\",\n \"def build(self):\\n if os.path.exists(self.build_path) and not os.path.isdir(self.build_path):\\n self._logger.error('Output path must be a folder if it already exists')\\n return\\n Compiler._mkdir(self.build_path)\\n self._execute_backend_on_spec()\",\n \"def compile(self):\\n return None # pragma: no cover\",\n \"def parsed(self):\\n if not self._parsed:\\n\\n self._parsed = compile(self.content, self.path, \\\"exec\\\")\\n\\n return self._parsed\",\n \"def get_buildall_target(self):\\n raise NotImplementedError\",\n \"def compilation(self, config: Config, files: List[str]) -> CallbackResult:\\n return [], files\",\n \"def compile_package(self):\\n build_package = [\\n self.mock,\\n '--root=%s' % self.root,\\n '--arch=%s' % self.arch,\\n '--shell',\\n '/build_package.sh',\\n \\n ]\\n output, errors = self._run_command(build_package)\",\n \"def compile_target(target_obj, search_paths, compile_path, ref_controller, globals_cached=None, **kwargs):\\n start = time.time()\\n compile_objs = target_obj[\\\"compile\\\"]\\n ext_vars = target_obj[\\\"vars\\\"]\\n target_name = ext_vars[\\\"target\\\"]\\n\\n if globals_cached:\\n cached.from_dict(globals_cached)\\n\\n use_go_jsonnet = kwargs.get(\\\"use_go_jsonnet\\\", False)\\n if use_go_jsonnet:\\n logger.debug(\\\"Using go-jsonnet over jsonnet\\\")\\n\\n for comp_obj in compile_objs:\\n input_type = comp_obj[\\\"input_type\\\"]\\n output_path = comp_obj[\\\"output_path\\\"]\\n input_params = comp_obj.setdefault(\\\"input_params\\\", {})\\n\\n if input_type == \\\"jinja2\\\":\\n input_compiler = Jinja2(compile_path, search_paths, ref_controller, comp_obj)\\n elif input_type == \\\"jsonnet\\\":\\n input_compiler = Jsonnet(compile_path, search_paths, ref_controller, use_go=use_go_jsonnet)\\n elif input_type == \\\"kadet\\\":\\n input_compiler = Kadet(compile_path, search_paths, ref_controller, input_params=input_params)\\n elif input_type == \\\"helm\\\":\\n input_compiler = Helm(compile_path, search_paths, ref_controller, comp_obj)\\n elif input_type == \\\"copy\\\":\\n ignore_missing = comp_obj.get(\\\"ignore_missing\\\", False)\\n input_compiler = Copy(compile_path, search_paths, ref_controller, ignore_missing)\\n elif input_type == \\\"remove\\\":\\n input_compiler = Remove(compile_path, search_paths, ref_controller)\\n elif input_type == \\\"external\\\":\\n input_compiler = External(compile_path, search_paths, ref_controller)\\n if \\\"args\\\" in comp_obj:\\n input_compiler.set_args(comp_obj[\\\"args\\\"])\\n if \\\"env_vars\\\" in comp_obj:\\n input_compiler.set_env_vars(comp_obj[\\\"env_vars\\\"])\\n else:\\n err_msg = 'Invalid input_type: \\\"{}\\\". Supported input_types: jsonnet, jinja2, kadet, helm, copy, remove, external'\\n raise CompileError(err_msg.format(input_type))\\n\\n input_compiler.make_compile_dirs(target_name, output_path, **kwargs)\\n input_compiler.compile_obj(comp_obj, ext_vars, **kwargs)\\n\\n logger.info(\\\"Compiled %s (%.2fs)\\\", target_obj[\\\"target_full_path\\\"], time.time() - start)\",\n \"def _compile(self, tocompile, parameters):\\n compiler = self.dialect.statement_compiler(self.dialect, tocompile, parameters)\\n compiler.compile()\\n return compiler\",\n \"def build(self, *args, **kwargs):\\n self._layer_counter = 0\\n r = self._build_impl(*args, **kwargs)\\n \\n # Call the init functions \\n if self._build_counter == 0:\\n for initlayer in self._layers_to_init:\\n if initlayer['initfnkwargs']:\\n initlayer['initfn'](initlayer['layer'], **initlayer['initfnkwargs'])\\n else:\\n initlayer['initfn'](initlayer['layer'])\\n \\n self._build_counter += 1\\n return r\",\n \"def calculate_compile_sources(targets, is_thrift_target):\\r\\n\\r\\n basedirs = set()\\r\\n sources = set()\\r\\n def collect_sources(target):\\r\\n basedirs.add(target.target_base)\\r\\n sources.update(target.sources_relative_to_buildroot())\\r\\n for target in targets:\\r\\n target.walk(collect_sources, predicate=is_thrift_target)\\r\\n return basedirs, sources\",\n \"def make(*args, **kwargs):\\n return _uhd_swig.amsg_source_make(*args, **kwargs)\",\n \"def _compile(self, source: str, filename: str) -> CodeType:\\n return compile(source, filename, \\\"exec\\\") # type: ignore\",\n \"def build_all(self):\\n self.android_build()\\n self.generate_patch_build('')\\n self.generate_specs_build()\\n self.generate_interfaces()\",\n \"def CompileWithClosure(js_files, definitions, entry_points, output_file):\\n\\n cmd = [\\n 'java', '-jar',\\n './node_modules/google-closure-compiler-java/compiler.jar',\\n '--language_out=ES5_STRICT', '--dependency_mode=PRUNE',\\n '--js_output_file=%s' % output_file\\n ]\\n cmd += ['--entry_point=%s' % e for e in entry_points]\\n cmd += ['--output_manifest=%s' % ('%s.manifest' % output_file)]\\n cmd += [\\n '../node_modules/google-closure-library/closure/**.js',\\n '!../node_modules/google-closure-library/closure/**_test.js',\\n '../node_modules/google-closure-library/third_party/closure/**.js',\\n '!../node_modules/google-closure-library/third_party/closure/**_test.js'\\n ]\\n cmd += js_files\\n cmd += definitions\\n subprocess.check_call(cmd)\",\n \"def cpp_compile(threads=2):\\n cmake_ret = subprocess.run([\\\"cmake\\\", \\\"../\\\"], encoding='utf-8', stdout=subprocess.PIPE, stderr=subprocess.PIPE)\\n output = cmake_ret.stdout\\n errors = cmake_ret.stderr\\n\\n output = \\\"\\\"\\n errors = \\\"\\\"\\n\\n make_ret = subprocess.run([\\\"make\\\", f\\\"-j{threads}\\\"], encoding='utf-8', stdout=subprocess.PIPE, stderr=subprocess.PIPE)\\n output = make_ret.stdout\\n errors = make_ret.stderr\\n return (make_ret.returncode == 0, errors if make_ret != 0 else None)\",\n \"def build(texFileLists):\\n LaTeX_CMD = [\\\"latex -shell-escape -halt-on-error {texfile}.tex\\\",\\n \\\"bibtex {texfile}.aux\\\",\\n \\\"latex -shell-escape -halt-on-error {texfile}.tex\\\",\\n \\\"pdflatex -shell-escape -halt-on-error {texfile}.tex\\\"]\\n\\n for texFile in texFileLists:\\n filename, file_extension = os.path.splitext(texFile)\\n for lcmd in LaTeX_CMD:\\n try:\\n call(lcmd.format(texfile=filename), shell=True)\\n except:\\n print \\\"Error occurs while building \\\", filename\\n exit(1)\",\n \"def compile(self, exclude=None, recurse=True, references=None, verbose=False):\\n from clay.shell.core import lsgrep\\n\\n _os.chdir(self.directory)\\n sources = self.sources\\n\\n if sources is None:\\n sources = [_os.path.splitext(x)[0] for x in lsgrep(self.src_ext, self.directory, recurse=recurse)]\\n if exclude is not None and len(exclude) > 0:\\n sources = list(filter(lambda x: all(not(y in x) for y in exclude), sources))\\n # if any flags, include them\\n if len(self.flags) > 0:\\n opt_str = '-' + ' -'.join(self.flags)\\n else:\\n opt_str = ''\\n\\n statechanged = False\\n for src in sources:\\n src_name = src + self.src_ext\\n dst_name = src + self.dst_ext\\n\\n if not _os.path.exists(src_name):\\n print(src, 'doesn\\\\'t exist, skipping...')\\n continue\\n\\n src_mtime = _os.stat(src_name).st_mtime\\n if _os.path.exists(dst_name):\\n dst_mtime = _os.stat(dst_name).st_mtime\\n else:\\n dst_mtime = 0 # file doesn't exist\\n\\n # if edited more than five seconds ago, compile it\\n if src_mtime - dst_mtime >= 5:\\n print('Compiling ({}):'.format(self.compiler_name), src)\\n cmd = self.compiler_name\\n if self.compiler_name == 'csc': # C Sharp specific handling\\n if references is not None:\\n cmd += ' /r:' + ','.join(references)\\n cmd += ' /out:{} '.format(dst_name)\\n cmd += '{} \\\"{}\\\"'.format(opt_str, src_name)\\n if verbose:\\n print('cmd:', cmd)\\n _os.system(cmd)\\n statechanged = True\\n\\n if not statechanged:\\n print('Nothing new to compile in \\\"{}\\\" when recurse={}'.format(self.directory, recurse))\",\n \"def build(setup_kwargs: MutableMapping[Text, Any]) -> None:\\n setup_kwargs.update({\\n 'ext_modules': extensions,\\n 'cmdclass': {\\n 'build_ext': CMakeBuild\\n }\\n })\",\n \"def build(self) -> ast.Module:\\n self.build_missing_config()\\n self.build_output()\\n self.build_input()\\n self.build_missing_imports()\\n\\n ret = ast.parse(\\\"\\\")\\n ret.body = self.before_tree.body + self.tree.body + self.after_tree.body\\n # Add fake line numbers to the nodes, so we can generate a traceback on error\\n i = 0\\n for node in dfs_walk(ret):\\n if isinstance(node, ast.stmt):\\n i += 1\\n node.lineno = i\\n node.end_lineno = i\\n\\n return ast.fix_missing_locations(ret)\",\n \"def make(files: Dict[str, List[File]] = files) -> List[File]:\\n progs = [\\n make_prog(prog_path, c_files)\\n for prog_path, c_files in files.items()\\n ]\\n return progs\",\n \"def pyo():\\n local('python -O -m compileall .')\",\n \"def pyo():\\n local('python -O -m compileall .')\",\n \"def compile(self):\\n raise NotImplementedError()\",\n \"def _GetFlags(self, lines, build_dir):\\n is_win = sys.platform == 'win32'\\n flags_by_output = {}\\n for line in lines:\\n command_line = shlex.split(line.strip(), posix=not is_win)[1:]\\n\\n output_name = _FindAndRemoveArgWithValue(command_line, '-o')\\n dep_name = _FindAndRemoveArgWithValue(command_line, '-MF')\\n\\n command_line = _MergeSpacedArgs(command_line, '-Xclang')\\n\\n cc_file = [x for x in command_line if x.endswith('.cc') or\\n x.endswith('.c') or\\n x.endswith('.cpp') or\\n x.endswith('.mm') or\\n x.endswith('.m')]\\n if len(cc_file) != 1:\\n self._skipped.append(command_line)\\n continue\\n assert len(cc_file) == 1\\n\\n if is_win:\\n rsp_file = [x for x in command_line if x.endswith('.rsp')]\\n assert len(rsp_file) <= 1\\n if rsp_file:\\n rsp_file = os.path.join(build_dir, rsp_file[0][1:])\\n with open(rsp_file, \\\"r\\\") as open_rsp_file:\\n command_line = shlex.split(open_rsp_file, posix=False)\\n\\n defines = [x for x in command_line if x.startswith('-D')]\\n include_dirs = [x for x in command_line if x.startswith('-I')]\\n dash_f = [x for x in command_line if x.startswith('-f')]\\n warnings = \\\\\\n [x for x in command_line if x.startswith('/wd' if is_win else '-W')]\\n others = [x for x in command_line if x not in defines and \\\\\\n x not in include_dirs and \\\\\\n x not in dash_f and \\\\\\n x not in warnings and \\\\\\n x not in cc_file]\\n\\n for index, value in enumerate(include_dirs):\\n if value == '-Igen':\\n continue\\n path = value[2:]\\n if not os.path.isabs(path):\\n path = os.path.join(build_dir, path)\\n include_dirs[index] = '-I' + os.path.normpath(path)\\n\\n # GYP supports paths above the source root like <(DEPTH)/../foo while such\\n # paths are unsupported by gn. But gn allows to use system-absolute paths\\n # instead (paths that start with single '/'). Normalize all paths.\\n cc_file = [os.path.normpath(os.path.join(build_dir, cc_file[0]))]\\n\\n # Filter for libFindBadConstructs.so having a relative path in one and\\n # absolute path in the other.\\n others_filtered = []\\n for x in others:\\n if x.startswith('-Xclang ') and \\\\\\n (x.endswith('libFindBadConstructs.so') or \\\\\\n x.endswith('libFindBadConstructs.dylib')):\\n others_filtered.append(\\n '-Xclang ' +\\n os.path.join(os.getcwd(), os.path.normpath(\\n os.path.join('out/gn_flags', x.split(' ', 1)[1]))))\\n elif x.startswith('-B'):\\n others_filtered.append(\\n '-B' +\\n os.path.join(os.getcwd(), os.path.normpath(\\n os.path.join('out/gn_flags', x[2:]))))\\n else:\\n others_filtered.append(x)\\n others = others_filtered\\n\\n flags_by_output[cc_file[0]] = {\\n 'output': output_name,\\n 'depname': dep_name,\\n 'defines': sorted(defines),\\n 'include_dirs': sorted(include_dirs), # TODO(scottmg): This is wrong.\\n 'dash_f': sorted(dash_f),\\n 'warnings': sorted(warnings),\\n 'other': sorted(others),\\n }\\n return flags_by_output\",\n \"def build(*onts, fail=False, n_jobs=9, write=True):\\n from joblib import Parallel, delayed # importing in here saves 150ms\\n tail = lambda:tuple()\\n lonts = len(onts)\\n if lonts > 1:\\n for i, ont in enumerate(onts):\\n if ont.__name__ == 'parcBridge':\\n onts = onts[:-1]\\n def tail(o=ont):\\n return o.setup(),\\n if i != lonts - 1:\\n raise ValueError('parcBridge should be built last to avoid weird errors!')\\n # ont_setup must be run first on all ontologies\\n # or we will get weird import errors\\n if n_jobs == 1 or True:\\n return tuple(ont.make(fail=fail, write=write) for ont in\\n tuple(ont.setup() for ont in onts) + tail())\\n\\n # have to use a listcomp so that all calls to setup()\\n # finish before parallel goes to work\\n return Parallel(n_jobs=n_jobs)(delayed(o.make)(fail=fail, write=write)\\n for o in\\n #[ont_setup(ont) for ont in onts])\\n (tuple(Async()(deferred(ont.setup)()\\n for ont in onts)) + tail()\\n if n_jobs > 1\\n else [ont.setup()\\n for ont in onts]))\",\n \"def build_step(self):\\n run_cmd('./compile.sh', log_all=True, simple=True, log_ok=True)\",\n \"def build(self, depVersions, output):\\n raise NotImplementedError()\",\n \"def here(*allowed_target_types):\\r\\n SourceRoot.register(buildfile_dir, *allowed_target_types)\",\n \"def resolve(self, targets):\\n context = self.context(target_roots=targets)\\n self.execute(context)\\n return context.products.get_data('compile_classpath')\",\n \"def build(self, backend: Optional[str] = 'thread', copy_flow: bool = False, *args, **kwargs) -> 'Flow':\\n\\n op_flow = self._build_graph(copy_flow)\\n\\n if not backend:\\n op_flow.logger.warning('no specified backend, build_level stays at %s, '\\n 'and you can not run this flow.' % op_flow._build_level)\\n elif backend in {'thread', 'process'}:\\n op_flow._service_contexts.clear()\\n for v in op_flow._service_nodes.values():\\n p_args = v['parsed_args']\\n p_args.parallel_backend = backend\\n # for thread and process backend which runs locally, host_in and host_out should not be set\\n p_args.host_in = BaseService.default_host\\n p_args.host_out = BaseService.default_host\\n op_flow._service_contexts.append((Flow._service2builder[v['service']], p_args))\\n op_flow._build_level = Flow.BuildLevel.RUNTIME\\n else:\\n raise NotImplementedError('backend=%s is not supported yet' % backend)\\n\\n return op_flow\",\n \"def assemble(compilers, paths, output, promoted):\\n\\n\\t# assemble C compile command\\n\\tcommand = []\\n\\tcommand.append(compilers['cc'])\\n\\n\\tcommand.extend(['-I', paths['include']])\\n\\n\\tif output:\\n\\t\\tcommand.extend(['-o', output])\\n\\n\\tcommand.extend(promoted)\\n\\n\\t# execute compiler\\n\\treturn subprocess.call(command)\",\n \"def BuildFileTargets(target_list, build_file):\\n return [p for p in target_list if BuildFile(p) == build_file]\",\n \"def _compile_sources(sources: Union[Sequence[_SourceTypes], _SourceTypes]) -> Dict[str, Any]:\\n if not isinstance(sources, Sequence):\\n sources = [sources]\\n\\n compiled: Dict[str, Any] = {}\\n for source in sources:\\n if callable(source):\\n source = source(compiled) # Pass what we have parsed so far to the function\\n compiled = _merge_sources(dest=compiled, source=source)\\n return compiled\",\n \"def partial_build(src, lang, bsp_c3, opt_level, march, reporter):\\n if lang == 'c3':\\n srcs = [\\n relpath('..', 'librt', 'io.c3'),\\n bsp_c3,\\n io.StringIO(src)]\\n o2 = c3c(\\n srcs, [], march, opt_level=opt_level,\\n reporter=reporter, debug=True)\\n objs = [o2]\\n elif lang == 'bf':\\n o3 = bfcompile(src, march, reporter=reporter)\\n o2 = c3c(\\n [bsp_c3], [], march, reporter=reporter)\\n objs = [o2, o3]\\n elif lang == 'c':\\n o2 = c3c(\\n [bsp_c3], [], march, reporter=reporter)\\n coptions = COptions()\\n include_path1 = relpath('..', 'librt', 'libc')\\n coptions.add_include_path(include_path1)\\n with open(relpath('..', 'librt', 'libc', 'lib.c'), 'r') as f:\\n o3 = cc(\\n f, march, coptions=coptions, debug=True,\\n reporter=reporter)\\n o4 = cc(\\n io.StringIO(src), march, coptions=coptions, debug=True,\\n reporter=reporter)\\n objs = [o2, o3, o4]\\n else:\\n raise NotImplementedError('language not implemented')\\n obj = link(\\n objs, partial_link=True,\\n use_runtime=True, reporter=reporter, debug=True)\\n return obj\",\n \"def get_compiled(self, flags=0):\\n return re.compile(self.get_regex(), flags)\",\n \"def getBuildSets():\",\n \"def build_run_bowtie(bowtie_fp, input_contigs_fasta, output_index_fp, pe1_fastq, pe2_fastq, u_fastq, output_sam_fp):\\n\\n logging.debug('preparing bowtie build and align function calls')\\n\\n # bowtie build executable should be in the same directory as bowtie2 commands\\n bowtie_build_fp = bowtie_fp + '-build'\\n\\n # build index files\\n buildout, builderr = run_bowtie_build(bowtie_build_fp, input_contigs_fasta, output_index_fp)\\n\\n # once index files are built, run aligner using said index files\\n stdout, stderr = run_bowtie(bowtie_fp, output_index_fp, pe1_fastq, pe2_fastq, u_fastq, output_sam_fp)\\n\\n return buildout, builderr, stdout, stderr\",\n \"def run_compiler(text_or_file: str, compiler_suite: str, fail_when: ErrorCode = ERROR) -> Any:\\n preprocessor, parser, ast, compiler = load_compiler_suite(compiler_suite)\\n return compileDSL(text_or_file, preprocessor(), parser(), ast(), compiler(), fail_when)\",\n \"def build(name, file, config, backend, debug):\\n setup_lithops_logger(logging.DEBUG)\\n\\n verify_runtime_name(name)\\n\\n if config:\\n config = load_yaml_config(config)\\n\\n config_ow = set_config_ow(backend, runtime_name=name)\\n config = default_config(config, config_ow, load_storage_config=False)\\n\\n if config['lithops']['mode'] != SERVERLESS:\\n raise Exception('\\\"lithops build\\\" command is only valid for serverless backends')\\n\\n compute_config = extract_serverless_config(config)\\n compute_handler = ServerlessHandler(compute_config, None)\\n compute_handler.build_runtime(name, file)\",\n \"def builds(self):\\n builds = [b for b in self.statuses() if b[\\\"type\\\"] == \\\"build\\\"]\\n for build in builds:\\n yield Build(build, **self._new_session_args)\\n\\n return\",\n \"def __call__(self, *args, **kwargs):\\n\\n self._check_compiled()\\n return self._compiled(*args, **kwargs)\",\n \"def resolve(self,**bindings):\\n for solution in self.resolve_all(**bindings):\\n return solution\",\n \"def build(parameters):\\n\\n\\n print(\\\"In Build module\\\")\",\n \"def build(self, progname):\\n self.run_programm(self.COMPILED[self.progtype][0], \\\"%s %s %s\\\" %\\\\\\n (progname, self.COMPILED[self.progtype][1], COMPILED_FILENAME ))\\n\\n compiled_progname=COMPILED_FILENAME\\n return compiled_progname\",\n \"def abstract_builder(pkg_name, name_list, return_list = False):\\n # some handlers needs dicts (commands, beacons), while some need lists (encoders,decoders, etc)\\n if return_list:\\n ret_val = []\\n else:\\n ret_val = {}\\n \\n # Go through the string names and get the appropriate Class from the appropriate module.\\n # Once you have that, do a dynamic import so we can use it, then map that class type\\n # so we can instantiate the appropriate instance when going through a beaconing interation.\\n for module_name in name_list:\\n\\n module_class = Controller.easy_import(pkg_name, module_name) # imports the class\\n if return_list:\\n ret_val.append(module_class) # adds the Class object to a list\\n else:\\n ret_val[module_name] = module_class # maps the Class object to the appropriate module name\\n \\n return ret_val\",\n \"def process_input_files(inputs):\\n for ifile in inputs:\\n with open(ifile) as fin:\\n exec(compile(fin.read(), ifile, 'exec'))\",\n \"def build(self, **kwargs):\\n if not self.built:\\n with tf.name_scope(self.name):\\n self._build(**kwargs)\\n self.built = True\\n return self\",\n \"def actionBuild():\\n\\n #Init builder logger\\n Builder.init()\\n\\n for target in Settings.targets:\\n targetsToBuild, combineLibs, copyToOutput = Builder.getTargetGnPath(target)\\n for platform in Settings.targetPlatforms:\\n for cpu in Settings.targetCPUs:\\n if System.checkIfCPUIsSupportedForPlatform(cpu,platform):\\n for configuration in Settings.targetConfigurations:\\n if not Summary.checkIfActionFailed(ACTION_PREPARE, target, platform, cpu, configuration):\\n Logger.printStartActionMessage('Build ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)\\n result = Builder.run(target, targetsToBuild, platform, cpu, configuration, combineLibs, copyToOutput)\\n Summary.addSummary(ACTION_BUILD, target, platform, cpu, configuration, result, Builder.executionTime)\\n if result != NO_ERROR:\\n Logger.printEndActionMessage('Failed building ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.RED)\\n #Terminate script execution if stopExecutionOnError is set to True in userdef\\n shouldEndOnError(result)\\n else:\\n Logger.printEndActionMessage('Build ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration)\\n else:\\n Logger.printColorMessage('Build cannot run because preparation has failed for ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)\\n Logger.printEndActionMessage('Build not run for ' + target + ' ' + platform + ' ' + cpu + ' ' + configuration,ColoredFormatter.YELLOW)\",\n \"def Run(self):\\n self.__compile = False\\n return self._Run()\",\n \"def compile(self):\\n if self._outputs is None or len(self._outputs) == 0:\\n print \\\"Error! No templates to compile! Did you specify a valid configuration file?\\\"\\n else:\\n # for each output file in output list \\n for output in self._outputs:\\n # Generate context\\n context = output[\\\"context\\\"]\\n # Generate template stream for that output\\n tmpl = self._env.get_template(output[\\\"source_filename\\\"]).stream(context)\\n # Verify output directory\\n output_filename = output[\\\"output_filename\\\"]\\n output_path = abspath(dirname(output_filename))\\n if not exists(output_path):\\n makedirs(output_path)\\n # Render it to the output file\\n tmpl.dump(output_filename)\",\n \"def compile(self, relations):\\n # Compile each relation one-by-one, accumulating them into a list\\n boxDeclarations = []\\n for boxSyntax in self.boxes:\\n box = boxSyntax.compile(boxDeclarations, relations)\\n boxDeclarations.append(box)\\n\\n # With all the boxes compiled, compile all of the connections and use\\n # both to construct and return a new relation declaration\\n return RelationDeclaration(\\n self.name,\\n boxDeclarations,\\n [c.compile(boxDeclarations) for c in self.connections])\",\n \"def build_runtime(self, runtime_name, file):\\n self.compute_handler.build_runtime(runtime_name, file)\",\n \"def cl_program_from_files(context, filenames):\\n source = ''\\n for filename in filenames:\\n source += open(os.path.join(CL_PATH, filename)).read()\\n return cl.Program(context, source)\",\n \"def run_thrifts(self):\\r\\n\\r\\n def is_py_thrift(target):\\r\\n return isinstance(target, PythonThriftLibrary)\\r\\n\\r\\n all_thrifts = set()\\r\\n\\r\\n def collect_sources(target):\\r\\n for source in target.sources:\\r\\n all_thrifts.add((target.target_base, source))\\r\\n\\r\\n self.target.walk(collect_sources, predicate=is_py_thrift)\\r\\n\\r\\n copied_sources = set()\\r\\n for base, relative_source in all_thrifts:\\r\\n abs_source = os.path.join(base, relative_source)\\r\\n copied_source = os.path.join(self._workdir, relative_source)\\r\\n\\r\\n safe_mkdir(os.path.dirname(copied_source))\\r\\n shutil.copyfile(abs_source, copied_source)\\r\\n copied_sources.add(self._modify_thrift(copied_source))\\r\\n\\r\\n for src in copied_sources:\\r\\n if not self._run_thrift(src):\\r\\n raise PythonThriftBuilder.CodeGenerationException(\\\"Could not generate .py from %s!\\\" % src)\",\n \"def build(self):\\n if not self.built:\\n with tf.name_scope(self.name):\\n self._build()\\n self.built = True\\n return self\",\n \"def _do_build(self) -> List[types.Action]:\\n return [\\n docker_command.DockerRun(\\n command=[\\\"/entrypoint.sh\\\", self.tag],\\n builder=builder.GO_BUILDER,\\n run_config=docker_command.default_run_config(\\n constants.STORAGE_OPERATOR_ROOT / \\\"entrypoint.sh\\\"\\n ),\\n mounts=[\\n utils.bind_mount(\\n target=Path(\\\"/storage-operator\\\"),\\n source=constants.STORAGE_OPERATOR_ROOT,\\n ),\\n # This container (through operator-sdk) will call `docker\\n # build`, so we need to expose our Docker socket.\\n utils.bind_mount(\\n target=Path(\\\"/var/run/docker.sock\\\"),\\n source=Path(\\\"/var/run/docker.sock\\\"),\\n ),\\n ],\\n )\\n ]\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.6088479","0.56501824","0.5504514","0.5331496","0.5290386","0.52033865","0.5071304","0.50649625","0.5030966","0.49910313","0.49463353","0.49420282","0.49213454","0.49104938","0.4905638","0.49018425","0.48995948","0.48995948","0.48929653","0.48861158","0.48784515","0.48736057","0.48582754","0.48312917","0.48227343","0.48219872","0.4821308","0.4815596","0.48114997","0.47803763","0.47801414","0.47726846","0.47643018","0.47526765","0.47498006","0.47362244","0.47349203","0.47291514","0.47268617","0.4726009","0.47066423","0.47006202","0.47005144","0.46971312","0.46951458","0.469201","0.46908435","0.46809298","0.46808437","0.46759716","0.46743503","0.46539944","0.4651915","0.46494246","0.4630476","0.46190098","0.46058053","0.45954254","0.45872706","0.45810467","0.4562527","0.45617914","0.45531097","0.45524555","0.4546404","0.4546404","0.45357874","0.4535124","0.45333183","0.4532277","0.4527178","0.4524876","0.45168024","0.45093477","0.45045444","0.44961533","0.4490744","0.44893262","0.44880116","0.44865856","0.44761476","0.44727","0.44714952","0.4469824","0.44628558","0.4462184","0.44560772","0.4446326","0.4444588","0.44420773","0.44418117","0.44390905","0.44355115","0.44351262","0.44292054","0.4427343","0.44212064","0.44107816","0.44022456","0.44015896"],"string":"[\n \"0.6088479\",\n \"0.56501824\",\n \"0.5504514\",\n \"0.5331496\",\n \"0.5290386\",\n \"0.52033865\",\n \"0.5071304\",\n \"0.50649625\",\n \"0.5030966\",\n \"0.49910313\",\n \"0.49463353\",\n \"0.49420282\",\n \"0.49213454\",\n \"0.49104938\",\n \"0.4905638\",\n \"0.49018425\",\n \"0.48995948\",\n \"0.48995948\",\n \"0.48929653\",\n \"0.48861158\",\n \"0.48784515\",\n \"0.48736057\",\n \"0.48582754\",\n \"0.48312917\",\n \"0.48227343\",\n \"0.48219872\",\n \"0.4821308\",\n \"0.4815596\",\n \"0.48114997\",\n \"0.47803763\",\n \"0.47801414\",\n \"0.47726846\",\n \"0.47643018\",\n \"0.47526765\",\n \"0.47498006\",\n \"0.47362244\",\n \"0.47349203\",\n \"0.47291514\",\n \"0.47268617\",\n \"0.4726009\",\n \"0.47066423\",\n \"0.47006202\",\n \"0.47005144\",\n \"0.46971312\",\n \"0.46951458\",\n \"0.469201\",\n \"0.46908435\",\n \"0.46809298\",\n \"0.46808437\",\n \"0.46759716\",\n \"0.46743503\",\n \"0.46539944\",\n \"0.4651915\",\n \"0.46494246\",\n \"0.4630476\",\n \"0.46190098\",\n \"0.46058053\",\n \"0.45954254\",\n \"0.45872706\",\n \"0.45810467\",\n \"0.4562527\",\n \"0.45617914\",\n \"0.45531097\",\n \"0.45524555\",\n \"0.4546404\",\n \"0.4546404\",\n \"0.45357874\",\n \"0.4535124\",\n \"0.45333183\",\n \"0.4532277\",\n \"0.4527178\",\n \"0.4524876\",\n \"0.45168024\",\n \"0.45093477\",\n \"0.45045444\",\n \"0.44961533\",\n \"0.4490744\",\n \"0.44893262\",\n \"0.44880116\",\n \"0.44865856\",\n \"0.44761476\",\n \"0.44727\",\n \"0.44714952\",\n \"0.4469824\",\n \"0.44628558\",\n \"0.4462184\",\n \"0.44560772\",\n \"0.4446326\",\n \"0.4444588\",\n \"0.44420773\",\n \"0.44418117\",\n \"0.44390905\",\n \"0.44355115\",\n \"0.44351262\",\n \"0.44292054\",\n \"0.4427343\",\n \"0.44212064\",\n \"0.44107816\",\n \"0.44022456\",\n \"0.44015896\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6061279"},"document_rank":{"kind":"string","value":"1"}}},{"rowIdx":90866,"cells":{"query":{"kind":"string","value":"sends an email when lead saved successfully"},"document":{"kind":"string","value":"def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())","def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has been submitted by %s.<br><br>\n\nGo here to view or edit the request: <br>\n<a href=\"%s\">%s</a>\n<br><br>\nSincerely,<br><br>\nThe Janelia Parking Permit Program\n \"\"\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\n subject = 'A new parking permit request has been entered'\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\n text_content = re.sub(r'<[^>]+>','',message)\n html_content = message\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def save(self, commit=False):\n mail_result = self.send_email()\n if mail_result:\n self.instance.is_admin_notified = True\n\n contact = super().save(commit=commit)\n\n return contact","def save(self, **kwargs):\n\t\tif (self.answer != \"\"):\n\t\t\tself.answered = True\n\t\t\tsuper().save(**kwargs)\n\t\t\tmail_subject = \"Your Question has been answered.\"\n\t\t\t# Do not break the following string or the email will get cut off\n\t\t\tmessage = f\"Hi {self.user.first_name},\\n\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\n\\n\"\n\n\t\t\temail = EmailMessage(\n\t\t\t\tmail_subject, message, to=[self.user.email]\n\t\t\t)\n\t\t\temail.send()\n\t\telse:\n\t\t\tsuper().save(**kwargs)","def save(self): # noqa\n self.bound_admin.message_success(self.bound_request, _('Done.'))","def action_my_payslip_sent(self):\n self.ensure_one()\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\n if template:\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\n self.flag = True","def test_email_sent_on_failure(self):\n self._authorize()\n data = {\n 'Subject_Number': '000-1111',\n 'Pin_Code': '1234',\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\n 'Mobile_Number': '2223334444',\n }\n patient = self.create_xml_patient(data)\n payload = self.create_xml_payload([patient])\n response = self._post(payload)\n self.assertEqual(response.status_code, 500)\n self.assertEqual(len(mail.outbox), 1)","def save_model(self, request, obj, form, change):\n if change and obj.status == 'V':\n if Organization.objects.get(pk=obj.pk).status == 'P':\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\n try:\n sender = Plugin_Contact.objects.all()[0].email\n except IndexError:\n sender = None\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\n send_mail(u'Validation de votre fiche sur la PASR',\n u'Bonjour,\\n\\nVotre fiche vient d\\'être validée sur la PASR.',\n sender, dests)\n super(OrganizationAdmin, self).save_model(request, obj, form, change)","def send_referral_saved(cls, referral, created_by):\n\n template_id = settings.SENDINBLUE[\"REFERRAL_SAVED_TEMPLATE_ID\"]\n\n data = {\n \"params\": {\"case_number\": referral.id},\n \"replyTo\": cls.reply_to,\n \"templateId\": template_id,\n \"to\": [{\"email\": created_by.email}],\n }\n\n cls.send(data)","def callback_success_message(request):\n msg = 'Sucessfully recorded form :)'\n logger.info(msg)\n messages.info(request._request, msg)","def save(self):\n super(Notification, self).save()\n # get user recipent\n us = self.notified_user\n # check that user has a valid email address\n if us.email.find('@') > 0 and us.email.find('.') > 0:\n # mandatory fields\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\n to = us.email\n from_email = settings.DEFAULT_FROM_EMAIL\n # get text version of the message\n text_content = self.get_email_content_from_type(\n self.notification_type\n )\n # FIXME: HTML version implementation pending\n html_content = self.get_email_content_from_type(\n self.notification_type\n )\n msg = EmailMultiAlternatives(\n subject, \n text_content,\n from_email,\n [to]\n )\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)","def mail(request):\n email_admin.delay('testinggg')\n return JsonResponse({\"details\":\"working\"})","def on_post(self):\n return \"Ok, the stuff is being saved\"","def send_email(request):\n # send emails and return some manner of success response\n send(**request.params)\n return {'success': 'mail sent!'}","def unsuccessful(self, **post):\n cr, uid, context = request.cr, request.uid, request.context\n attchment_list = []\n mail_obj = request.env['mail.mail'].sudo()\n\n sale_order_id = request.session.get('sale_order_id')\n\n print(\"\\n\\n\\n\\n\\n=========sale order id===========\", sale_order_id)\n if sale_order_id:\n order = request.env['sale.order'].sudo().browse(sale_order_id)\n print(\"\\n\\n\\n\\n\\n=========sale order order===========\", order)\n else:\n return request.redirect('/enrolment_book')\n request.website.sale_reset()\n\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\n raise_if_not_found=False)\n if template_id:\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\n order.id)\n enroll_file_name = \"Pro-Forma \" + order.name\n if pdf_data_enroll:\n pdfvals = {'name': enroll_file_name,\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas_fname': enroll_file_name + '.pdf',\n 'res_model': 'sale.order',\n 'type': 'binary'}\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\n attchment_list.append(pdf_create)\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\n if agreement_id:\n attchment_list.append(agreement_id)\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\n if banking_detail_id:\n attchment_list.append(banking_detail_id)\n\n # mail_values = {\n # 'email_from': template_id.email_from,\n # 'reply_to': template_id.reply_to,\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # 'subject': \"Charterquest FreeQuote/Enrolment \" + order.name,\n # 'notification': True,\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\n # 'auto_delete': False,\n # }\n # msg_id = mail_obj.create(mail_values)\n # msg_id.send()\n\n template_id.sudo().write(\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\n template_id.sudo().with_context(\n # email_to=each_request.get('email'),\n # event_list=event_list,\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # prof_body=invoice_id.prof_body.name,\n ).send_mail(order.id, force_send=True)\n return request.render(\"cfo_snr_jnr.event_unsuccessful\", {'order': order})","def claim_email(request):\n email = request.POST.get('email', '')\n email_user = User.objects.filter(email=email)\n payload = {\n 'res': 'failed'\n }\n if email_user.exists() and \\\n not email_user[0].profile.send_mail:\n request.user.profile.add_email(email)\n payload['res'] = 'success'\n\n return payload","def test_admin_approval_complete_email(self):\n new_user = UserModel().objects.create_user(**self.user_info)\n profile = self.registration_profile.objects.create_profile(new_user)\n profile.send_admin_approve_complete_email(Site.objects.get_current())\n self.assertEqual(len(mail.outbox), 1)\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])","def add_manualpost_email(request, submission_id=None, access_token=None):\n\n if request.method == 'POST':\n try:\n button_text = request.POST.get('submit', '')\n if button_text == 'Cancel':\n return redirect(\"submit/manual_post.html\")\n \n form = SubmissionEmailForm(request.POST)\n if form.is_valid():\n submission_pk = form.cleaned_data['submission_pk']\n message = form.cleaned_data['message']\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n \n if form.cleaned_data['direction'] == 'incoming':\n msgtype = 'msgin'\n else:\n msgtype = 'msgout'\n \n submission, submission_email_event = (\n add_submission_email(request=request,\n remote_ip=request.META.get('REMOTE_ADDR', None),\n name = form.draft_name,\n rev=form.revision,\n submission_pk = submission_pk,\n message = message,\n by = request.user.person,\n msgtype = msgtype) )\n \n messages.success(request, 'Email added.')\n \n try:\n draft = Document.objects.get(name=submission.name)\n except Document.DoesNotExist:\n # Assume this is revision 00 - we'll do this later\n draft = None\n \n if (draft != None):\n e = AddedMessageEvent(type=\"added_message\", doc=draft)\n e.message = submission_email_event.submissionemailevent.message\n e.msgtype = submission_email_event.submissionemailevent.msgtype\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\n e.by = request.user.person\n e.desc = submission_email_event.desc\n e.time = submission_email_event.time\n e.save()\n \n return redirect(\"ietf.submit.views.manualpost\")\n except ValidationError as e:\n form = SubmissionEmailForm(request.POST)\n form._errors = {}\n form._errors[\"__all__\"] = form.error_class([\"There was a failure uploading your message. (%s)\" % e.message])\n else:\n initial = {\n }\n\n if (submission_id != None):\n submission = get_submission_or_404(submission_id, access_token)\n initial['name'] = \"{}-{}\".format(submission.name, submission.rev)\n initial['direction'] = 'incoming'\n initial['submission_pk'] = submission.pk\n else:\n initial['direction'] = 'incoming'\n \n form = SubmissionEmailForm(initial=initial)\n\n return render(request, 'submit/add_submit_email.html',dict(form=form))","def receive_email_view(request):\n save_inbound_email(request.POST, request.FILES)\n return HttpResponse(200)","def validation_email_sent(request):\n assert(settings.EMAIL_VALIDATION == True)\n logging.debug('')\n data = {\n 'email': request.user.email,\n 'change_email_url': reverse('user_changeemail'),\n 'action_type': 'validate'\n }\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))","def postprocess():\n if ERRORS:\n address = 'tamakoshihiroki@gmail.com'\n body = '\\n\\n'.join( ERRORS )\n msg = create_message( body, address )\n send_mail( msg, address )","def send_mail_when_failed(self, body):\r\n pass","def test_send_email_on_invite(self):\n\n league = self.create_league()\n\n season = self.create_season(league)\n team = self.create_team(season)\n\n player = self.create_player()\n\n send_user_email_on_join(player, team.id)\n\n self.assertEqual(len(mail.outbox), 1)\n\n # if testing manually:\n # import pathlib\n # pathlib.Path(\"test_email.html\").write_text(last_sent.body)","def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\n lead = Lead.objects.get(id=lead_id)\n created_by = lead.created_by\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\n for user in recipients:\n recipients_list = []\n user = User.objects.filter(id=user, is_active=True).first()\n if user:\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\n recipients_list.append(user.email)\n context = {}\n context[\"url\"] = protocol + '://' + domain + \\\n reverse('leads:view_lead', args=(lead.id,))\n context[\"user\"] = user\n context[\"lead\"] = lead\n context[\"created_by\"] = created_by\n context[\"source\"] = source\n subject = 'Assigned a lead for you. '\n html_content = render_to_string(\n 'assigned_to/leads_assigned.html', context=context)\n msg = EmailMessage(\n subject,\n html_content,\n to=recipients_list\n )\n msg.content_subtype = \"html\"\n msg.send()","def form_valid(self, form):\n self.object = form.save()\n self.send_verify_email()\n return super().form_valid(form)","def save_response(self):\r\n self.q(css='input.save-button').first.click()\r\n EmptyPromise(\r\n lambda: 'save' in self.alert_message.lower(),\r\n \"Status message saved\"\r\n ).fulfill()","def form_valid(self, form):\n\n ModelAccountVerification.objects.send_verification_email(form.user,\n verification_type=\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\n\n messages.info(self.request, \"Email sent successfully to the {0}\".format(form.user.email))\n\n return HttpResponseRedirect(self.success_url)","def execute(self):\n return LOGGER.info(f\"{datetime.datetime.now()} - Sending EMail to the configured email list\")","def test_post_external_attendance_successfull(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-03T19:00',\n 'attendee_email': 'student@external.mx',\n 'exit_datetime': '2021-08-03T20:00'}\n response = self.client.post(\n reverse('tracking:external-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def save_model(self, request, obj, form, change):\n notify.send(\n request.user,\n recipient=obj.reported_by,\n verb='updated your report',\n action_object=obj,\n description=obj.notes,\n target=obj\n )\n obj.save()","def post(self):\n return send_email(request.args)","def send_warning(self):\n\n # Check whether all the necessary parameters for SMS are present\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\n client = Client(self.account_sid, self.auth_token)\n\n try:\n sms = client.messages.create(\n body=\"\"\"Last will: It was at least 30 days since your last check in. \n This is a reminder to check in in the next 24 hours.\"\"\",\n from_=self.twilio_phone,\n to=self.your_phone)\n sms\n print(\"\\nSMS sent\")\n except Exception as e:\n print(f\"An error occurred while trying to send the SMS. Error: {e}\")\n\n else:\n print(\"\\nMissing SMS parameters. SMS not sent\")\n\n # Check whether all the necessary parameters for email are present\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\n message = f\"\"\"It has been at least 30 days since you last checked in. \nYou need to check in in the next 24 hours.\\n\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\n\nIn order to reset simply go to the working directory and run python3 last_will.py\"\"\"\n\n # send_email will return 0 if everything went ok, otherwise it will return an error message\n status = send_email(self.sender_name, self.your_email,\n self.email, self.email_pwd,\n subject='Last will: Reminder to check in', unencrypted_message=message)\n\n if status != 0:\n print(status)\n exit(1)\n else:\n print(\"Email sent\\n\")\n\n print(f\"You have until {self.deadline} to check in. \"\n f\"In order to do that simply go to the working directory and run ./last_will.sh\\n\")\n else:\n print(\"Missing email parameters. Email not sent.\\n\")\n exit(1)","def send_reminder(self):\n pass","def complete_allotment(request):\n reg_no = request.user.username\n user = get_object_or_404(User, username=reg_no)\n sec_email = user.get_profile().secondary_email\n options_chosen = get_chosen_options(user)\n context = {'username': reg_no, 'email': sec_email, \n 'options_chosen': options_chosen}\n ##Sending mail with allotment details \n admin = User.objects.get(pk=1) \n from_email = admin.email\n subject = \"JAM 2012 admissions\"\n content = \"The following options were chosen by you \\n \\n\"\n if options_chosen:\n counter = 1\n for option in options_chosen:\n content += \"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\n\" %(counter, option.opt_code, option.opt_name, option.opt_location)\n counter += 1 \n \n content += \"\\n \\n \\nPlease do not delete this email and keep it for reference purposes. \\n \\n \\n \\n Regards, \\n JAM Office, IIT Bombay\"\n send_mail(subject, content, from_email, [sec_email])\n admin_content = content\n admin_content +=\"\\n\\n\\n#%s:\" % (reg_no)\n counter = 1\n for option in options_chosen:\n admin_content += \"%s,%s:\" %(counter, option.opt_code) \n counter += 1\n admin_content +=\"#\"\n admin_content += time.ctime() \n mail_admins(subject, admin_content) \n return render(request, 'allotter/complete.html', context)","def create_response_and_send_mail(token, partner_id, email):\n \t #set url\n\t url = survey.public_url\n \t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\n\t\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\n\t\t template = self.pool.get('email.template').browse(cr,uid,template_id)\n\t\t body = template.body_html\n\t if token:\n \t url = url + '/' + token\n\n\t # post the message\n \t values = {\n \t'model': None,\n\t 'res_id': None,\n \t 'subject': template.subject,\n \t'body': body.replace(\"__URL__\", url),\n\t 'body_html': body.replace(\"__URL__\", url),\n \t 'parent_id': None,\n \t'partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'email_from': 'info@sba.org' or None,\n \t 'email_to': partner.email,\n\t\t }\n\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\n \t mail_mail_obj.send(cr, uid, [mail_id], context=context)","def send_reminder(self):\n message_contents = \"This is a reminder that your event: \" + self.event_title + \" takes place on \" + self.event_date + \" in \" + self.event_location\n subject = \"Event Reminder\"\n attendees = self.gameplanuser_set.all()\n for attendee in attendees:\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\n remindermessage.save()","def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\n\n location = reverse(\"KursAnmeldung-verify_email\", kwargs={\"hash\":rnd_hash})\n verify_link = request.build_absolute_uri(location)\n\n # FIXME: convert to users local time.\n now = datetime.datetime.utcnow()\n\n email_context = {\n \"verify_link\": verify_link,\n \"db_entry\": db_entry,\n \"now\": now,\n }\n\n # Render the internal page\n emailtext = render_to_string(\"kurs_anmeldung/verify_mailtext.txt\", email_context)\n\n # Get the preferences from the database:\n raw_notify_list = preferences[\"notify\"]\n notify_list = raw_notify_list.splitlines()\n notify_list = [i.strip() for i in notify_list if i]\n\n email_kwargs = {\n \"from_email\": preferences[\"from_email\"],\n \"subject\": preferences[\"email_subject\"],\n \"body\": emailtext,\n \"to\": [db_entry.email],\n \"bcc\": notify_list,\n }\n\n if MAIL_DEBUG == True:\n msg = u\"MAIL_DEBUG is on: No Email was sended!\"\n request.page_msg(msg)\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n\n request.page_msg(\"django.core.mail.EmailMessage kwargs:\")\n request.page_msg(email_kwargs)\n\n request.page_msg(\"debug mail text:\")\n request.page_msg(mark_safe(\"<pre>%s</pre>\" % emailtext))\n return\n\n # We can't use django.core.mail.send_mail, because all members\n # of the recipient list will see the others in the 'To' field.\n # But we would like to notify the admins via 'Bcc' field.\n\n connection = SMTPConnection(fail_silently=False)\n email = EmailMessage(**email_kwargs)\n\n try:\n sended = email.send(fail_silently=False)\n except Exception, err:\n msg = \"Error sending mail: %s\" % err\n LogEntry.objects.log_action(app_label=\"kurs_anmeldung\", action=\"error\",\n message=msg\n )\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n if settings.DEBUG or request.user.is_staff:\n db_entry.save()\n raise\n else:\n db_entry.mail_sended = sended\n db_entry.log(request, \"mail sended: %s\" % sended)","def confirm_email(self):\n self.active = True\n self.save()","def save(self, register=False):\n sha = sha1(self.email).hexdigest()\n infos = self.to_dict()\n infos[\"plan\"] = infos[\"plan\"][\"id\"] if infos[\"plan\"] else None\n\n if not redis.hmset(\"sl:account:{}\".format(sha), infos):\n raise SleekException(\"Could not save current user.\", 401)\n \n if register:\n try:\n send_email.delay(\n \"florent.esp@gmail.com\", \"Welcome to sleekstatus !\",\n \"Welcome message\", [self.email]\n )\n except:\n pass # Cannot send email","def task_send_reminder_email():\n send_reminder_email()\n logger.info(\"Sent reminder email\")","def account_activation_sent(request):\r\n\treturn render(request, 'account_activation_sent.html')","def save(self, *args, **kwargs):\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\n allUsers = Bruker.get_all_dict(Bruker)\n mottaker = allUsers[int(data['mottaker'])]\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\n melding.save()","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n reminders.Patient.objects.filter(\n pk__in=[self.test_patient.pk, self.other_patient.pk]\n ).update(next_visit=appt_date)\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n self.startRouter()\n self.router.logger.setLevel(logging.DEBUG)\n\n # run email job\n from afrims.apps.reminders.app import daily_email_callback\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)\n self.stopRouter()","def save_appointment_details(request, calendar_id):\n def schedule_mail(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n def schedule_sms(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n start_time = request.GET['start_time'][:19]\n end_time = request.GET['end_time'][:19]\n \n start_time = datetime.strptime(start_time, \"%Y-%m-%dT%H:%M:%S\")\n end_time=datetime.strptime(end_time, \"%Y-%m-%dT%H:%M:%S\")\n \n calendar_obj = Calendar.objects.get(pk=calendar_id)\n # if this is a POST request we need to process the form data\n if request.method == 'POST':\n\n # create a form instance and populate it with data from the request:\n form = AppointmentForm(request.POST)\n\n # check whether it's valid and save it\n if form.is_valid():\n # Save appointment details\n \n mobilephone = form.data['mobilephone']\n email = form.data['email']\n first_name = form.data['first_name']\n last_name = form.data['last_name']\n notes = form.data['notes']\n\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\n \n appointment.calendar = calendar_obj\n appointment.save()\n\n try:\n send_appointment_mail(appointment) # send appointment details email\n except Exception as exp:\n print(exp)\n \n try:\n send_appointment_sms(appointment) # send appointment details sms\n except Exception as exp:\n print(exp)\n \n # Calculate reminder schedule dates\n reminder1 = start_time - timedelta(hours=2)\n reminder2 = start_time - timedelta(hours=24)\n reminder3 = start_time - timedelta(days=7)\n\n # Schedule mails\n schedule_mail(reminder1, appointment)\n schedule_mail(reminder2, appointment)\n schedule_mail(reminder3, appointment)\n \n # Schedule sms\n schedule_sms(reminder1, appointment)\n schedule_sms(reminder2, appointment)\n schedule_sms(reminder3, appointment)\n \n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\n \n # if a GET (or any other method) we'll create a blank form\n else:\n form = AppointmentForm()\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\n 'office_location': calendar_obj.office_location})","def post(\n self,\n email,\n company_name,\n location,\n job_profile,\n salary,\n username,\n password,\n security_question,\n security_answer,\n notes,\n date_applied,\n status,\n):","def post_save_order_receiver(sender, instance, created, *args, **kwargs):\n sender_email = instance.cargo.sender.email\n recepient_email = instance.cargo.recepient.email\n\n booking_agent = instance.cargo.booking_station.branch_agent.email\n\n if created:\n instance._set_order_price()\n instance._set_time_approximations()\n price = instance.price\n\n subject = \"Order Finalized and ready to go.\"\n message = f\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\"\n\n send_async_email(\n subject=subject,\n message=message,\n sender=booking_agent,\n recepients=[sender_email, recepient_email],\n )","def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\n\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n parts = pyzmail.parse.get_mail_parts(message)\n body=''\n for part in parts:\n if part.is_body == 'text/plain' and part.disposition == None:\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\n body = body + payload + '\\n'\n\n msg = submit_message_from_message(message, body, by)\n\n if (submission_pk != None):\n # Must exist - we're adding a message to an existing submission\n submission = Submission.objects.get(pk=submission_pk)\n else:\n # Must not exist\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\n if submissions.count() > 0:\n raise ValidationError(\"Submission {} already exists\".format(name))\n \n # create Submission using the name\n try:\n submission = Submission.objects.create(\n state_id=\"waiting-for-draft\",\n remote_ip=remote_ip,\n name=name,\n rev=rev,\n title=name,\n note=\"\",\n submission_date=datetime.date.today(),\n replaces=\"\",\n )\n from ietf.submit.utils import create_submission_event, docevent_from_submission\n desc = \"Submission created for rev {} in response to email\".format(rev)\n create_submission_event(request, \n submission,\n desc)\n docevent_from_submission(request,\n submission,\n desc)\n except Exception as e:\n log(\"Exception: %s\\n\" % e)\n raise\n\n if msgtype == 'msgin':\n rs = \"Received\"\n else:\n rs = \"Sent\"\n\n desc = \"{} message - manual post - {}-{}\".format(rs, name, rev)\n submission_email_event = SubmissionEmailEvent.objects.create(\n desc = desc,\n submission = submission,\n msgtype = msgtype,\n by = by,\n message = msg)\n #in_reply_to = in_reply_to\n\n save_submission_email_attachments(submission_email_event, parts)\n return submission, submission_email_event","def handle_sent(self, instance):\n if not instance.created_by:\n return\n\n activity = Activity(\n actor=instance.created_by,\n verb=RestrictedMailSent,\n object=instance,\n time=instance.used,\n extra_context={},\n )\n self.manager.add_activity(\n activity, [instance.created_by.pk], [NotificationFeed]\n )\n\n # Send notification\n notification = RestrictedMailSentNotification(instance.created_by)\n notification.notify()","def handle_failure(self, instance, sender=None, reason=None):\n send_email.delay(\n to_email=sender,\n context={\"reason\": reason},\n subject=\"Kunne ikke sende ut begrenset epost\",\n plain_template=\"restricted/email/process_failure.txt\",\n html_template=\"restricted/email/process_failure.html\",\n )","def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):\n print(\"member email\",mem_mail)\n logger.info(\"in sending existing customer mail task\")\n return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website)","def handle(self, *args, **options):\n\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\n if candidate.contact_address and candidate.participating]\n\n\n print 'sending e-mails'\n conn = get_connection()\n for c in candidates_with_email:\n if c.should_send_reminder():\n\n print 'emailing', c\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\n c.last_reminder_sent = timezone.now()\n c.save()\n msg = make_email(c)\n conn.send_messages([msg])\n conn.close()","def send(self, request, id, tribe_slug):\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\n document = self.get_document(id, tribe_slug)\n \n form = self._set_user_email_address(request)\n email = self._get_user_email_address(request)\n if form or not email:\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\n \n # NB: Temporarily disabling actual email sending for development\n #email_document(document, to=[email], subject='Document: %s' % document.title)\n print \"Sending email to %s\" % email \n \n # Send a signal to let everyone know about this document interaction\n document_interaction.send(sender=self, document=document, mode=\"sent\", request=request, recipient=email)\n \n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))","def submit(self):\r\n try:\r\n response = self.send_request()\r\n if response.ok:\r\n self.submitted_at = datetime.now(pytz.UTC)\r\n self.status = \"submitted\"\r\n self.save()\r\n else:\r\n self.status = \"must_retry\"\r\n self.error_msg = response.text\r\n self.save()\r\n except Exception as error:\r\n log.exception(error)\r\n self.status = \"must_retry\"\r\n self.save()","def _send(self, email_message): \n\t\tif not email_message.recipients(): \n\t\t\treturn False \n\t\trecipients = map(self._sanitize, email_message.recipients()) \n\t\tMsg = o.CreateItem(0)\n \t\tMsg.To = recipients\n\t\tMsg.Subject = 'subject'\n\t\tMsg.Body = 'text'\n\t\tself.connection.SaveChanges(0)\n\n\t\tMsg.Send()\n\t\treturn True","def set_email():\n bid_fields = {'assertion':request.form['bid_assertion'],\n 'audience':settings.DOMAIN}\n headers = {'Content-type':'application/x-www-form-urlencoded'}\n h.disable_ssl_certificate_validation=True\n resp, content = h.request('https://browserid.org/verify',\n 'POST',\n body=urlencode(bid_fields),\n headers=headers)\n bid_data = json.loads(content)\n if bid_data['status'] == 'okay' and bid_data['email']:\n user = f.userFromBidOrNew(bid_data['email'])\n session['user_id'] = user.id\n session['user_email'] = user.email\n\n return redirect(url_for('main'))","def wasSent(self):\n\t\tself.sent = True\n\t\tself.save()","def action_send_email(self):\n self.ensure_one()\n if self.communication_channel != \"email\":\n return False # Maybe we should raise an error here.\n\n if self.state in (\"sent\", \"done\"):\n raise ValidationError(_(\"This communication is already sent.\"))\n\n lines_2be_processed = self.credit_control_line_ids.filtered(\n lambda line: line.state != \"sent\"\n )\n\n if not lines_2be_processed:\n raise ValidationError(_(\"There is no draft lines to send.\"))\n\n self = self.with_context(lang=self.partner_id.lang)\n\n partner = self.partner_id\n mail_template = self.policy_level_id.email_template_id\n # Send the email\n partner.with_context(credit_control_mail=True).message_post_with_template(\n template_id=mail_template.id,\n model=self._name,\n res_id=self.id,\n )\n # Set the state of the credit control lines to \"queued\"\n lines_2be_processed.write({\"state\": \"sent\"})\n self.state = \"sent\"","def test_activate_form(self, mock_sendmail):\r\n res = self.testapp.post('/api/v1/suspend',\r\n params={'email': u'testing@dummy.com'},\r\n status=200)\r\n\r\n success = json.loads(res.body)\r\n self.assertTrue(\r\n 'message' in success,\r\n \"Should be successful with admin email address: \" + str(res))\r\n self.assertTrue(mock_sendmail.called)","def test_post_internal_attendance_success(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-04T10:00',\n 'attendee_email': 'student@anahuac.mx',\n 'exit_datetime': '2021-08-04T11:30'}\n response = self.client.post(\n reverse('tracking:internal-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def save(self, *args, **kwargs):\n self.expire_date = self.sent_date + timedelta(days=self.period)\n super(Referral, self).save(*args, **kwargs)","def post(self, request):\n\n try:\n eventoid = request.POST.get('id', '')\n correo = request.POST.get('correo', '')\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\n print(\"Exito en la asignación de staff\")\n except:\n print(\"Error en la asignacion de staff\")\n\n \n return render(request, self.template, self.context)\n #return render(request, self.template, self.context)","def save(self, *args, **kwargs):\n create = self.id is None\n if create:\n self.username = u'lc_' + self.city.lower()\n # Give the user a randomly generated password\n password = generate_random_string()\n self.password = password\n # Compose the email to send to the LC\n context = {\n 'lc': self,\n 'password': password,\n }\n subject = render_to_string(\n 'registration/lc_register_email_subject.txt')\n subject = subject.strip()\n message = render_to_string(\n 'registration/lc_register_email_text.txt',\n context)\n send_mail(subject,\n message,\n settings.DEFAULT_FROM_EMAIL,\n [self.email])\n # Delegate upward to the parent classes\n super(LocalCommittee, self).save(*args, **kwargs)","def contact():\n if request.method == 'POST':\n send_email()\n return \"\"","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n # run email job\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)","def save_model(self, request, obj, form, change):\n admin.ModelAdmin.save_model(self, request, obj, form, change)\n\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\n post_title = obj.title\n post_url = 'http://www.aptitudeworld.com.au/blog'\n\n # send an email if we are adding a new post\n if not change:\n # send an email to all users in the system to notify them that a new post is available\n email_title = 'Aptitude World ANZ - A new post is available'\n email_from = 'info@aptitudeworld.com.au'\n\n try:\n # get all user emails in the database\n users = apps.users.models.User.objects.all()\n confirmation_email_txt_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\n confirmation_email_html_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\n for u in users:\n email_to = u.email_address\n full_name = u.first_name + \" \" + u.last_name\n\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\n\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\n email_message.attach_alternative(email_html, 'text/html')\n email_message.send()\n except BadHeaderError:\n pass","def send_email():\n send_mail(\"You've got some problem.\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\n ['dimazarj2009@rambler.ru'], fail_silently=False,)","def create_sent_email(self, *args, **kwargs):\n receiver = kwargs['receiver']\n sender = kwargs['sender']\n user = kwargs['user']\n body = kwargs['body']\n subject = kwargs['subject']\n tracker_id = kwargs['tracker_id']\n if receiver and sender and subject and body:\n sent_email = SentEmail()\n if tracker_id:\n sent_email.tracker_id = tracker_id\n sent_email.read_status = 'Tracking'\n else:\n sent_email.read_status = 'Not Tracking'\n sent_email.receiver = receiver\n sent_email.subject = subject\n sent_email.sender = sender\n sent_email.status = 'sent'\n sent_email.user = user\n sent_email.body = body\n sent_email.save()\n return sent_email\n else:\n return False","def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\n\n if is_new:\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\n\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))","def sendEmail(householdID):\n contactID = mdb.getContact(householdID)\n sqlq = \"\"\"\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\n FROM Contact\n WHERE idContact = '{}';\n \"\"\".format(contactID)\n result = mdb.getSQL(sqlq)[0]\n\n # thisName = (\"%s %s\" % (result['Name'], result['Surname']))\n thisName = (\"%s\" % (result['Name']))\n thisAddress = (\"%s</br>%s</br>%s %s\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\n thisAddress = thisAddress.replace(\"None </br>\", \"\")\n thisEmail = (\"%s\" % (result['email']))\n thisStatus = (\"%s\" % (result['status']))\n thisAddress = thisAddress.replace(\"None</br>\", \"\")\n participantCount = (\"%s\" % mdb.getParticipantCount(str(householdID)))\n # prepare the custom email\n\n thisPath = os.path.dirname(os.path.abspath(__file__))\n if (thisStatus == 'de'):\n # DOESN'T happen yet - de is excluded from query for now\n # emailPath = os.path.join(thisPath, \"emails/email_confirm_de.html\")\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\n else:\n emailPath = os.path.join(thisPath, \"emails/email_automated_date.html\")\n\n templateFile = open(emailPath, \"r\")\n templateText = templateFile.read()\n templateFile.close()\n templateText = templateText.replace(\"[householdID]\", householdID)\n templateText = templateText.replace(\"[contactID]\", contactID)\n templateText = templateText.replace(\"[name]\", thisName)\n templateText = templateText.replace(\"[address]\", thisAddress)\n templateText = templateText.replace(\"[securityCode]\", mdb.getSecurityCode(householdID))\n templateText = templateText.replace(\"[participantCount]\", participantCount)\n\n # Subject\n subjectLine = templateText.splitlines()[0]\n templateText = templateText[templateText.find('\\n') + 1:] # find line break and return all from there - i.e. remove first line\n \n # email file\n emailFilePath = os.path.join(thisPath, \"tempEmail.htmail\")\n\n emailFile = open(emailFilePath, \"w+\")\n emailFile.write(templateText)\n emailFile.close()\n\n # call('mutt -e \"set content_type=text/html\" -s \"[TESTING]' + subjectLine + '\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\n call('mutt -e \"set content_type=text/html\" -s \"' + subjectLine + '\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)","def save(self,owner=None):\n\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\n\n t = Ticket(\n title = self.cleaned_data['title'],\n owner=owner,\n submitter_email = self.cleaned_data['submitter_email'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n account = owner.account\n )\n\n t.save()\n\n f = FollowUp(\n ticket = t,\n title = _('Ticket Opened Via Web'),\n date = datetime.now(),\n public = True,\n comment = self.cleaned_data['body'],\n account = owner.account\n )\n\n f.save()\n\n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n }\n\n messages_sent_to = []\n\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self, user):\n\n q = self.cleaned_data['queue']\n\n t = Ticket( title = self.cleaned_data['title'],\n submitter_email = self.cleaned_data['submitter_email'],\n account = self.cleaned_data['account'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n owner = self.cleaned_data['owner']\n )\n\n if HAS_TAG_SUPPORT:\n t.tags = self.cleaned_data['tags']\n\n if self.cleaned_data['assigned_to']:\n try:\n t.assigned_to = self.cleaned_data['assigned_to']\n except User.DoesNotExist:\n t.assigned_to = None\n t.save()\n\n f = FollowUp( ticket = t,\n title = _('Ticket Opened'),\n date = datetime.now(),\n public = False,\n comment = self.cleaned_data['body'],\n systemuser = user.account,\n )\n if self.cleaned_data['assigned_to']:\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\n 'name': t.get_assigned_to\n }\n\n f.save()\n \n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n 'comment': f.comment,\n }\n \n messages_sent_to = []\n\n if t.submitter_email:\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n #FIX USERSETTINGS\n #=======================================================================\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\n # send_templated_mail(\n # 'assigned_to',\n # context,\n # recipients=t.assigned_to.email,\n # sender=q.from_address,\n # fail_silently=True,\n # files=files,\n # )\n # messages_sent_to.append(t.assigned_to.email)\n #=======================================================================\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self):\n email = self.validated_data['email']\n name = self.validated_data['name']\n password = self.validated_data['password']\n\n email_query = models.Email.objects.filter(address=email)\n if email_query.exists():\n email_instance = email_query.get()\n\n # If the email is already verified, we send a duplicate\n # notification and exit.\n if email_instance.is_verified:\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"is already verified.\",\n email_instance,\n )\n email_instance.send_duplicate_notification()\n\n return\n\n # If the email is not verified, we send a new verification\n # token to the address.\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"already exists. Sending a new verification token instead.\"\n )\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()\n\n return\n\n # The email doesn't exist, so we create a new user and email,\n # then send a verification token to the email.\n user = models.User.objects.create_user(name, password)\n email_instance = models.Email.objects.create(address=email, user=user)\n\n # The user's primary email is their only email. This is the only\n # time the primary email can be unverified.\n user.primary_email = email_instance\n user.save()\n\n logger.info(\n \"Registered new user %r with email address %r\",\n user,\n email_instance,\n )\n\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()","def send_email(self, message):\n pass","def send_mail(self, subject):\r\n pass","def post():\n contactus_json = request.get_json()\n\n try:\n dict_data = ContactUsSchema().load(contactus_json)\n dict_data['description'] = escape(dict_data['description'])\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\n response, status = 'Received successfully', HTTPStatus.OK\n except ValidationError as project_err:\n response, status = {'systemErrors': project_err.messages}, \\\n HTTPStatus.BAD_REQUEST\n return response, status","async def handle_email(self, email):\n\t\tif 'logdir' in self.log_settings:\n\t\t\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\n\t\t\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\n\t\t\t\tf.write(email.email.as_bytes())\n\n\t\tawait self.log('You got mail!')","def set_email_sent(self, instance_id, creator, type):\n table = self.type_map[type]\n item = {\n \"InstanceID\": instance_id,\n \"Creator\": creator,\n \"Scheduled For Deletion\": True,\n \"EmailSent\": True\n }\n return table.put_item(item)","def order_success(request, job_id, token):\n if request.user.is_active:\n job_obj = Job.objects.get(job_id= job_id)\n if job_obj.employeer == request.user:\n if verifiers.payment_token_is_valid(token, job_obj):\n if job_obj.accepted == False:\n job_obj.status = Status.objects.create(status_name=\"PN\", comment=\"Awaiting payment confirmation from Stripe, this might take a few minutes.\")\n job_obj.accepted = True\n job_obj.save()\n \n messages.add_message(request, messages.INFO, 'Your order has been created.')\n return redirect('/') \n else:\n #someone tries to manually enter the URL with wrong payment_token\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n #the person making the request is not the same as the person who created the job\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n return redirect(\"/login\")","def order_created(order_id):\n order = Order.objects.get(id=order_id)\n subject = f'Order nr. {order.id}'\n message = f'Dear {order.first_name},\\n\\n' \\\n f'You have successfully placed an order.' \\\n f'Your order ID is {order.id}.'\n mail_sent = send_mail(subject,\n message,\n 'info@tishman.com.ng',\n [order.email])\n return mail_sent","def send_email_alert_slack(email_pk: str):\n logging.info(\"Sending alert for Email<%s>\", email_pk)\n email = Email.objects.get(pk=email_pk)\n issue = email.issue\n assert issue, f\"Email<{email_pk}> does not have an associated Issue\"\n paralegal = issue.paralegal\n alert_sent = False\n case_email_url = settings.CLERK_BASE_URL + reverse(\n \"case-email-list\", args=(str(issue.pk),)\n )\n case_url = settings.CLERK_BASE_URL + reverse(\n \"case-detail-view\", args=(str(issue.pk),)\n )\n msg = (\n \"*New Email Notification*\\n\"\n f\"A new email has been received for case <{case_url}|{issue.fileref}> .\\n\"\n f\"You can view this case's emails here: <{case_email_url}|here>.\\n\"\n )\n if paralegal and issue.is_open:\n # Send alert directly to paralegal for open issues.\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\n logging.info(\"Looking up %s in Slack\", alert_email)\n slack_user = get_slack_user_by_email(alert_email)\n if slack_user:\n logging.info(\"Notifying %s of Email<%s> via Slack\", alert_email, email_pk)\n send_slack_direct_message(msg, slack_user[\"id\"])\n alert_sent = True\n\n if not alert_sent:\n # In all other cases send to alerts channel.\n logging.info(\n \"Could not find someone to DM in Slack, sending generic alert for Email<%s>\",\n email_pk,\n )\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\n\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\n logging.info(\"Alert sent sucessfully for Email<%s>\", email_pk)","def send_sbemail(email_addy, city):\n print(\"Single email send attempt: %s, %s\" % (email_addy, city))\n given_pair, weather = subject_phrase_picker(city)\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\n print(\"Going to try a mailgun post to: %s with subject %s\" % (email_addy, given_pair['subject']))\n response = requests.post(\n MG_API_URL,\n auth=(\"api\", MG_API_KEY),\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\n 'to': [email_addy],\n 'subject': given_pair['subject'],\n 'text': mailtext\n })\n print(\"status: %s | %s and response: %s\" % (response.status_code, response.reason, response.text))\n print(\"Attempted send %s to: %s!\" % (mailtext, email_addy))\n return \"Sent %s to: %s!\" % (mailtext, email_addy)","def on_update(self):\n\t\tfor email_account in frappe.get_all(\"Email Account\", filters={\"domain\": self.name}):\n\t\t\ttry:\n\t\t\t\temail_account = frappe.get_doc(\"Email Account\", email_account.name)\n\t\t\t\tfor attr in [\"email_server\", \"use_imap\", \"use_ssl\", \"use_tls\", \"attachment_limit\", \"smtp_server\", \"smtp_port\", \"use_ssl_for_outgoing\", \"append_emails_to_sent_folder\", \"incoming_port\"]:\n\t\t\t\t\temail_account.set(attr, self.get(attr, default=0))\n\t\t\t\temail_account.save()\n\n\t\t\texcept Exception as e:\n\t\t\t\tfrappe.msgprint(_(\"Error has occurred in {0}\").format(email_account.name), raise_exception=e.__class__)","def put(self, campaign_id):\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\n if campaign.candidates.count() > 0:\n campaign.launch_task('generate_mailer_file')\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\n else:\n api.abort(404, message='Campaign does not exist', success=False)","def post_save_moderation(self, sender, comment, request, **kwargs):\n model = comment.content_type.model_class()\n if model not in self._registry:\n return\n self._registry[model].email(comment, comment.content_object, request)","def delegate_last_day():\n\n regs = Registration.objects.all()\n\n template = 'notifications/last_day_mail.html'\n\n for reg in regs:\n subject = 'SciPy.in 2011: Schedule and other details'\n message = loader.render_to_string(\n template, dictionary={'name': reg.registrant.username})\n\n reg.registrant.email_user(subject=subject, message=message,\n from_email='madhusudancs@gmail.com')","def send_email(email_dict, appointment_id):\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)","def post(self):\n # se captura y se parsea a json el body del request recibido por el\n # webhook\n request_body = json.loads(self.request.body)\n\n for body in request_body:\n \"\"\" Evaluar el tipo de evento ya que trae campos diferentes \"\"\"\n logging.info(request_body)\n\n event = str(body['event'])\n correo = str(body['email'])\n numero_folio = str(body['numero_folio'])\n tipo_dte = str(body['tipo_dte'])\n\n logging.info(event)\n\n if event and correo and numero_folio and tipo_dte:\n\n if event == 'processed':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.processed_event = event\n email_model.processed_sg_event_id = body['sg_event_id']\n email_model.processed_sg_message_id = body['sg_message_id']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.processed_event = event\n e.processed_sg_event_id = body['sg_event_id']\n e.processed_sg_message_id = body['sg_message_id']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'delivered':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.delivered_event = event\n email_model.delivered_sg_event_id = body['sg_event_id']\n email_model.delivered_sg_message_id = body['sg_message_id']\n email_model.delivered_response = body['response']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.delivered_event = event\n e.delivered_sg_event_id = body['sg_event_id']\n e.delivered_sg_message_id = body['sg_message_id']\n e.delivered_response = body['response']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'open':\n model = EmailModel()\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n if email_model.opened_first_date == None:\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_event = event\n email_model.opened_ip = body['ip']\n email_model.opened_user_agent = body['useragent']\n email_model.opened_sg_event_id = body['sg_event_id']\n email_model.opened_sg_message_id = body['sg_message_id']\n model.email_add_count(email_model)\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n if e.opened_first_date == None:\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_event = event\n e.opened_ip = body['ip']\n e.opened_user_agent = body['useragent']\n e.opened_sg_event_id = body['sg_event_id']\n e.opened_sg_message_id = body['sg_message_id']\n e.email_add_count(e)\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'dropped':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.dropped_sg_event_id = body['sg_event_id']\n email_model.dropped_sg_message_id = body['sg_message_id']\n email_model.dropped_reason = body['reason']\n email_model.dropped_event = event\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.dropped_sg_event_id = body['sg_event_id']\n e.dropped_sg_message_id = body['sg_message_id']\n e.dropped_reason = body['reason']\n e.dropped_event = event\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'bounce':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.bounce_event = event\n email_model.bounce_sg_event_id = body['sg_event_id']\n email_model.bounce_sg_message_id = body['sg_message_id']\n email_model.bounce_reason = body['reason']\n email_model.bounce_status = body['status']\n email_model.bounce_type = body['type']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.bounce_event = event\n e.bounce_sg_event_id = body['sg_event_id']\n e.bounce_sg_message_id = body['sg_message_id']\n e.bounce_reason = str(body['reason']).decode(\"utf-8\")\n e.bounce_status = body['status']\n e.bounce_type = body['type']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'unsubscribe':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.unsubscribe_uid = body['uid']\n email_model.unsubscribe_purchase = body['purchase']\n email_model.unsubscribe_id = body['id']\n email_model.unsubscribe_event = body['event']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.unsubscribe_uid = body['uid']\n e.unsubscribe_purchase = body['purchase']\n e.unsubscribe_id = body['id']\n e.unsubscribe_event = body['event']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n else:\n logging.info('body con campos vacios')","def send_mail(self, address, title, message):\n pass","def mail_sent():\n\n url = settings.SITE_URL + '\\charts'\n subject = 'Анализ запрошенного ресурса'\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\n mail_sent = send_mail(subject,\n message,\n 'admin@myshop.com',\n ['user@mail.ru,'])\n print(message)\n return mail_sent","def emailNote(self, authenticationToken, parameters):\r\n pass","def send_new_email(user):\n token = user.get_token()\n message = Message(\n 'Verify Your New Email',\n sender='storcwebsite@gmail.com',\n recipients=[user.temp_email])\n message.body = f\"The email address associated with your Storc \" \\\n f\"account has changed.\\n\\nTo verify your new email address, \" \\\n f\"please click the link below:\\n\\n\" \\\n f\"{url_for('users.new_email', token=token, _external=True)}\"\n mail.send(message)","def report_save(message, bot):\n text = message.text\n reports = parse_report(text)\n author_id = message.from_user.id\n author_username = message.from_user.username\n to_save = []\n for report in reports:\n obj_to_save = {\n 'author': author_id,\n 'author_name': author_username,\n 'text': report,\n }\n to_save.append(obj_to_save)\n try:\n request = requests.post(\n url='http://127.0.0.1:8000/api/v1/report/',\n headers={\n 'Authorization': f'Bearer {jwt_token}',\n },\n json=to_save,\n )\n bot.send_message(\n message.from_user.id,\n request.status_code,\n )\n except requests.exceptions.ConnectionError:\n bot.send_message(\n message.from_user.id,\n 'Ошибка соединения с сервером',\n )","def perform_create(self, serializer):\n user = serializer.save()\n signals.user_registered.send(\n sender=self.__class__, user=user, request=self.request\n )\n\n context = get_email_context(user)\n to = [get_user_email(user)]\n if djconf.SEND_ACTIVATION_EMAIL:\n djconf.EMAIL.activation(self.request, context).send(to)\n elif djconf.SEND_CONFIRMATION_EMAIL:\n djconf.EMAIL.confirmation(self.request, context).send(to)","def send(self):\n ReferralActivity.objects.create(\n actor=self.user,\n verb=ReferralActivityVerb.CREATED,\n referral=self,\n )\n # Confirm the referral has been sent to the requester by email\n Mailer.send_referral_saved(self)\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\n # email notifications)\n for unit in self.units.all():\n contacts = unit.members.filter(\n unitmembership__role=UnitMembershipRole.OWNER\n )\n for contact in contacts:\n Mailer.send_referral_received(self, contact=contact, unit=unit)","def execute_automatic_email(self, request, pk=None):\n try:\n retreat = Retreat.objects.get(pk=pk)\n except Exception:\n response_data = {\n 'detail': \"Retreat not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n try:\n email = AutomaticEmail.objects.get(\n id=int(request.GET.get('email'))\n )\n except Exception:\n response_data = {\n 'detail': \"AutomaticEmail not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n # Notify a user for every reserved seat\n emails = []\n for reservation in retreat.reservations.filter(is_active=True):\n if reservation.automatic_email_logs.filter(email=email):\n pass\n else:\n send_automatic_email(reservation.user, retreat, email)\n AutomaticEmailLog.objects.create(\n reservation=reservation,\n email=email\n )\n emails.append(reservation.user.email)\n\n response_data = {\n 'stop': True,\n 'emails': emails\n }\n return Response(response_data, status=status.HTTP_200_OK)","def save(self):\n owner = self.context[\"request\"].user\n recipient = self._recipient_email_inst.user\n\n logger.info(\n \"Transferring subscription %r from user %r to user %r\",\n owner.know_me_subscription,\n owner.pk,\n recipient.pk,\n )\n\n with transaction.atomic():\n models.Subscription.objects.filter(user=recipient).delete()\n models.Subscription.objects.filter(user=owner).update(\n user=recipient\n )","def send_message(self):\r\n return \"success\"","def alert_for_pending_mails_1(request):\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tThread(target=alert_for_pending_mails_1_worker).start()\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tresponse = {}\n\n\tresponse[\"info_to_contact\"] = \"Ok\"\n\n\treturn response","def save(self, request, item, linked_item, linked_item_form):\n\t\titem.description = linked_item_form.cleaned_data['subject']\n\t\tlinked_item.text = linked_item_form.cleaned_data['text']\n\t\tlinked_item.delivery_notes = linked_item_form.cleaned_data['delivery_notes']\n\t\titem.priority = Priority.objects.get(id = linked_item_form.cleaned_data['priority'])\n\t\t#\n\t\t# Look for actions other than save\n\t\t#\n\t\tif request.POST.get('update', '') == 'Completed':\n\t\t\titem.fixed = True\n\t\tif request.POST.get('update', '') == 'Failed':\n\t\t\titem.fixed = False\n\t\t\titem.validated = True\n\t\tif request.POST.get('update', '') == 'Verified':\n\t\t\titem.fixed = True\n\t\t\titem.validated = True\n\n\t\titem.save()\n\t\tlinked_item.save()\n\n\t\t#\n\t\t# Check for comments and add\n\t\t#\n\t\tcomment_text = linked_item_form.cleaned_data['comments'].strip()\n\t\tif comment_text != '':\n\t\t\tpost_comment(request, item, comment_text)"],"string":"[\n \"def save(self, fail_silently=False):\\n send_mail(fail_silently=fail_silently, **self.get_message_dict())\",\n \"def send_created_email(self):\\n if settings.NOTIFY_NEW_REG:\\n to = settings.NOTIFY_NEW_REG\\n message = \\\"\\\"\\\"\\\\\\nGreetings,<br><br>\\n\\nA new vehicle registration has been submitted by %s.<br><br>\\n\\nGo here to view or edit the request: <br>\\n<a href=\\\"%s\\\">%s</a>\\n<br><br>\\nSincerely,<br><br>\\nThe Janelia Parking Permit Program\\n \\\"\\\"\\\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\\n subject = 'A new parking permit request has been entered'\\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\\n text_content = re.sub(r'<[^>]+>','',message)\\n html_content = message\\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def save(self, commit=False):\\n mail_result = self.send_email()\\n if mail_result:\\n self.instance.is_admin_notified = True\\n\\n contact = super().save(commit=commit)\\n\\n return contact\",\n \"def save(self, **kwargs):\\n\\t\\tif (self.answer != \\\"\\\"):\\n\\t\\t\\tself.answered = True\\n\\t\\t\\tsuper().save(**kwargs)\\n\\t\\t\\tmail_subject = \\\"Your Question has been answered.\\\"\\n\\t\\t\\t# Do not break the following string or the email will get cut off\\n\\t\\t\\tmessage = f\\\"Hi {self.user.first_name},\\\\n\\\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\\\n\\\\n\\\"\\n\\n\\t\\t\\temail = EmailMessage(\\n\\t\\t\\t\\tmail_subject, message, to=[self.user.email]\\n\\t\\t\\t)\\n\\t\\t\\temail.send()\\n\\t\\telse:\\n\\t\\t\\tsuper().save(**kwargs)\",\n \"def save(self): # noqa\\n self.bound_admin.message_success(self.bound_request, _('Done.'))\",\n \"def action_my_payslip_sent(self):\\n self.ensure_one()\\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\\n if template:\\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\\n self.flag = True\",\n \"def test_email_sent_on_failure(self):\\n self._authorize()\\n data = {\\n 'Subject_Number': '000-1111',\\n 'Pin_Code': '1234',\\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\\n 'Mobile_Number': '2223334444',\\n }\\n patient = self.create_xml_patient(data)\\n payload = self.create_xml_payload([patient])\\n response = self._post(payload)\\n self.assertEqual(response.status_code, 500)\\n self.assertEqual(len(mail.outbox), 1)\",\n \"def save_model(self, request, obj, form, change):\\n if change and obj.status == 'V':\\n if Organization.objects.get(pk=obj.pk).status == 'P':\\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\\n try:\\n sender = Plugin_Contact.objects.all()[0].email\\n except IndexError:\\n sender = None\\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\\n send_mail(u'Validation de votre fiche sur la PASR',\\n u'Bonjour,\\\\n\\\\nVotre fiche vient d\\\\'être validée sur la PASR.',\\n sender, dests)\\n super(OrganizationAdmin, self).save_model(request, obj, form, change)\",\n \"def send_referral_saved(cls, referral, created_by):\\n\\n template_id = settings.SENDINBLUE[\\\"REFERRAL_SAVED_TEMPLATE_ID\\\"]\\n\\n data = {\\n \\\"params\\\": {\\\"case_number\\\": referral.id},\\n \\\"replyTo\\\": cls.reply_to,\\n \\\"templateId\\\": template_id,\\n \\\"to\\\": [{\\\"email\\\": created_by.email}],\\n }\\n\\n cls.send(data)\",\n \"def callback_success_message(request):\\n msg = 'Sucessfully recorded form :)'\\n logger.info(msg)\\n messages.info(request._request, msg)\",\n \"def save(self):\\n super(Notification, self).save()\\n # get user recipent\\n us = self.notified_user\\n # check that user has a valid email address\\n if us.email.find('@') > 0 and us.email.find('.') > 0:\\n # mandatory fields\\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\\n to = us.email\\n from_email = settings.DEFAULT_FROM_EMAIL\\n # get text version of the message\\n text_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n # FIXME: HTML version implementation pending\\n html_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n msg = EmailMultiAlternatives(\\n subject, \\n text_content,\\n from_email,\\n [to]\\n )\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\\n\\n logger.info(\\\"in send lead mail task\\\")\\n return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)\",\n \"def mail(request):\\n email_admin.delay('testinggg')\\n return JsonResponse({\\\"details\\\":\\\"working\\\"})\",\n \"def on_post(self):\\n return \\\"Ok, the stuff is being saved\\\"\",\n \"def send_email(request):\\n # send emails and return some manner of success response\\n send(**request.params)\\n return {'success': 'mail sent!'}\",\n \"def unsuccessful(self, **post):\\n cr, uid, context = request.cr, request.uid, request.context\\n attchment_list = []\\n mail_obj = request.env['mail.mail'].sudo()\\n\\n sale_order_id = request.session.get('sale_order_id')\\n\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order id===========\\\", sale_order_id)\\n if sale_order_id:\\n order = request.env['sale.order'].sudo().browse(sale_order_id)\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order order===========\\\", order)\\n else:\\n return request.redirect('/enrolment_book')\\n request.website.sale_reset()\\n\\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\\n raise_if_not_found=False)\\n if template_id:\\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\\n order.id)\\n enroll_file_name = \\\"Pro-Forma \\\" + order.name\\n if pdf_data_enroll:\\n pdfvals = {'name': enroll_file_name,\\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas_fname': enroll_file_name + '.pdf',\\n 'res_model': 'sale.order',\\n 'type': 'binary'}\\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\\n attchment_list.append(pdf_create)\\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\\n if agreement_id:\\n attchment_list.append(agreement_id)\\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\\n if banking_detail_id:\\n attchment_list.append(banking_detail_id)\\n\\n # mail_values = {\\n # 'email_from': template_id.email_from,\\n # 'reply_to': template_id.reply_to,\\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # 'subject': \\\"Charterquest FreeQuote/Enrolment \\\" + order.name,\\n # 'notification': True,\\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\\n # 'auto_delete': False,\\n # }\\n # msg_id = mail_obj.create(mail_values)\\n # msg_id.send()\\n\\n template_id.sudo().write(\\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\\n template_id.sudo().with_context(\\n # email_to=each_request.get('email'),\\n # event_list=event_list,\\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # prof_body=invoice_id.prof_body.name,\\n ).send_mail(order.id, force_send=True)\\n return request.render(\\\"cfo_snr_jnr.event_unsuccessful\\\", {'order': order})\",\n \"def claim_email(request):\\n email = request.POST.get('email', '')\\n email_user = User.objects.filter(email=email)\\n payload = {\\n 'res': 'failed'\\n }\\n if email_user.exists() and \\\\\\n not email_user[0].profile.send_mail:\\n request.user.profile.add_email(email)\\n payload['res'] = 'success'\\n\\n return payload\",\n \"def test_admin_approval_complete_email(self):\\n new_user = UserModel().objects.create_user(**self.user_info)\\n profile = self.registration_profile.objects.create_profile(new_user)\\n profile.send_admin_approve_complete_email(Site.objects.get_current())\\n self.assertEqual(len(mail.outbox), 1)\\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])\",\n \"def add_manualpost_email(request, submission_id=None, access_token=None):\\n\\n if request.method == 'POST':\\n try:\\n button_text = request.POST.get('submit', '')\\n if button_text == 'Cancel':\\n return redirect(\\\"submit/manual_post.html\\\")\\n \\n form = SubmissionEmailForm(request.POST)\\n if form.is_valid():\\n submission_pk = form.cleaned_data['submission_pk']\\n message = form.cleaned_data['message']\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n \\n if form.cleaned_data['direction'] == 'incoming':\\n msgtype = 'msgin'\\n else:\\n msgtype = 'msgout'\\n \\n submission, submission_email_event = (\\n add_submission_email(request=request,\\n remote_ip=request.META.get('REMOTE_ADDR', None),\\n name = form.draft_name,\\n rev=form.revision,\\n submission_pk = submission_pk,\\n message = message,\\n by = request.user.person,\\n msgtype = msgtype) )\\n \\n messages.success(request, 'Email added.')\\n \\n try:\\n draft = Document.objects.get(name=submission.name)\\n except Document.DoesNotExist:\\n # Assume this is revision 00 - we'll do this later\\n draft = None\\n \\n if (draft != None):\\n e = AddedMessageEvent(type=\\\"added_message\\\", doc=draft)\\n e.message = submission_email_event.submissionemailevent.message\\n e.msgtype = submission_email_event.submissionemailevent.msgtype\\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\\n e.by = request.user.person\\n e.desc = submission_email_event.desc\\n e.time = submission_email_event.time\\n e.save()\\n \\n return redirect(\\\"ietf.submit.views.manualpost\\\")\\n except ValidationError as e:\\n form = SubmissionEmailForm(request.POST)\\n form._errors = {}\\n form._errors[\\\"__all__\\\"] = form.error_class([\\\"There was a failure uploading your message. (%s)\\\" % e.message])\\n else:\\n initial = {\\n }\\n\\n if (submission_id != None):\\n submission = get_submission_or_404(submission_id, access_token)\\n initial['name'] = \\\"{}-{}\\\".format(submission.name, submission.rev)\\n initial['direction'] = 'incoming'\\n initial['submission_pk'] = submission.pk\\n else:\\n initial['direction'] = 'incoming'\\n \\n form = SubmissionEmailForm(initial=initial)\\n\\n return render(request, 'submit/add_submit_email.html',dict(form=form))\",\n \"def receive_email_view(request):\\n save_inbound_email(request.POST, request.FILES)\\n return HttpResponse(200)\",\n \"def validation_email_sent(request):\\n assert(settings.EMAIL_VALIDATION == True)\\n logging.debug('')\\n data = {\\n 'email': request.user.email,\\n 'change_email_url': reverse('user_changeemail'),\\n 'action_type': 'validate'\\n }\\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))\",\n \"def postprocess():\\n if ERRORS:\\n address = 'tamakoshihiroki@gmail.com'\\n body = '\\\\n\\\\n'.join( ERRORS )\\n msg = create_message( body, address )\\n send_mail( msg, address )\",\n \"def send_mail_when_failed(self, body):\\r\\n pass\",\n \"def test_send_email_on_invite(self):\\n\\n league = self.create_league()\\n\\n season = self.create_season(league)\\n team = self.create_team(season)\\n\\n player = self.create_player()\\n\\n send_user_email_on_join(player, team.id)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n\\n # if testing manually:\\n # import pathlib\\n # pathlib.Path(\\\"test_email.html\\\").write_text(last_sent.body)\",\n \"def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\\n lead = Lead.objects.get(id=lead_id)\\n created_by = lead.created_by\\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\\n for user in recipients:\\n recipients_list = []\\n user = User.objects.filter(id=user, is_active=True).first()\\n if user:\\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\\n recipients_list.append(user.email)\\n context = {}\\n context[\\\"url\\\"] = protocol + '://' + domain + \\\\\\n reverse('leads:view_lead', args=(lead.id,))\\n context[\\\"user\\\"] = user\\n context[\\\"lead\\\"] = lead\\n context[\\\"created_by\\\"] = created_by\\n context[\\\"source\\\"] = source\\n subject = 'Assigned a lead for you. '\\n html_content = render_to_string(\\n 'assigned_to/leads_assigned.html', context=context)\\n msg = EmailMessage(\\n subject,\\n html_content,\\n to=recipients_list\\n )\\n msg.content_subtype = \\\"html\\\"\\n msg.send()\",\n \"def form_valid(self, form):\\n self.object = form.save()\\n self.send_verify_email()\\n return super().form_valid(form)\",\n \"def save_response(self):\\r\\n self.q(css='input.save-button').first.click()\\r\\n EmptyPromise(\\r\\n lambda: 'save' in self.alert_message.lower(),\\r\\n \\\"Status message saved\\\"\\r\\n ).fulfill()\",\n \"def form_valid(self, form):\\n\\n ModelAccountVerification.objects.send_verification_email(form.user,\\n verification_type=\\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\\n\\n messages.info(self.request, \\\"Email sent successfully to the {0}\\\".format(form.user.email))\\n\\n return HttpResponseRedirect(self.success_url)\",\n \"def execute(self):\\n return LOGGER.info(f\\\"{datetime.datetime.now()} - Sending EMail to the configured email list\\\")\",\n \"def test_post_external_attendance_successfull(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-03T19:00',\\n 'attendee_email': 'student@external.mx',\\n 'exit_datetime': '2021-08-03T20:00'}\\n response = self.client.post(\\n reverse('tracking:external-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def save_model(self, request, obj, form, change):\\n notify.send(\\n request.user,\\n recipient=obj.reported_by,\\n verb='updated your report',\\n action_object=obj,\\n description=obj.notes,\\n target=obj\\n )\\n obj.save()\",\n \"def post(self):\\n return send_email(request.args)\",\n \"def send_warning(self):\\n\\n # Check whether all the necessary parameters for SMS are present\\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\\n client = Client(self.account_sid, self.auth_token)\\n\\n try:\\n sms = client.messages.create(\\n body=\\\"\\\"\\\"Last will: It was at least 30 days since your last check in. \\n This is a reminder to check in in the next 24 hours.\\\"\\\"\\\",\\n from_=self.twilio_phone,\\n to=self.your_phone)\\n sms\\n print(\\\"\\\\nSMS sent\\\")\\n except Exception as e:\\n print(f\\\"An error occurred while trying to send the SMS. Error: {e}\\\")\\n\\n else:\\n print(\\\"\\\\nMissing SMS parameters. SMS not sent\\\")\\n\\n # Check whether all the necessary parameters for email are present\\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\\n message = f\\\"\\\"\\\"It has been at least 30 days since you last checked in. \\nYou need to check in in the next 24 hours.\\\\n\\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\\\n\\nIn order to reset simply go to the working directory and run python3 last_will.py\\\"\\\"\\\"\\n\\n # send_email will return 0 if everything went ok, otherwise it will return an error message\\n status = send_email(self.sender_name, self.your_email,\\n self.email, self.email_pwd,\\n subject='Last will: Reminder to check in', unencrypted_message=message)\\n\\n if status != 0:\\n print(status)\\n exit(1)\\n else:\\n print(\\\"Email sent\\\\n\\\")\\n\\n print(f\\\"You have until {self.deadline} to check in. \\\"\\n f\\\"In order to do that simply go to the working directory and run ./last_will.sh\\\\n\\\")\\n else:\\n print(\\\"Missing email parameters. Email not sent.\\\\n\\\")\\n exit(1)\",\n \"def send_reminder(self):\\n pass\",\n \"def complete_allotment(request):\\n reg_no = request.user.username\\n user = get_object_or_404(User, username=reg_no)\\n sec_email = user.get_profile().secondary_email\\n options_chosen = get_chosen_options(user)\\n context = {'username': reg_no, 'email': sec_email, \\n 'options_chosen': options_chosen}\\n ##Sending mail with allotment details \\n admin = User.objects.get(pk=1) \\n from_email = admin.email\\n subject = \\\"JAM 2012 admissions\\\"\\n content = \\\"The following options were chosen by you \\\\n \\\\n\\\"\\n if options_chosen:\\n counter = 1\\n for option in options_chosen:\\n content += \\\"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\\\n\\\" %(counter, option.opt_code, option.opt_name, option.opt_location)\\n counter += 1 \\n \\n content += \\\"\\\\n \\\\n \\\\nPlease do not delete this email and keep it for reference purposes. \\\\n \\\\n \\\\n \\\\n Regards, \\\\n JAM Office, IIT Bombay\\\"\\n send_mail(subject, content, from_email, [sec_email])\\n admin_content = content\\n admin_content +=\\\"\\\\n\\\\n\\\\n#%s:\\\" % (reg_no)\\n counter = 1\\n for option in options_chosen:\\n admin_content += \\\"%s,%s:\\\" %(counter, option.opt_code) \\n counter += 1\\n admin_content +=\\\"#\\\"\\n admin_content += time.ctime() \\n mail_admins(subject, admin_content) \\n return render(request, 'allotter/complete.html', context)\",\n \"def create_response_and_send_mail(token, partner_id, email):\\n \\t #set url\\n\\t url = survey.public_url\\n \\t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\\n\\t\\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\\n\\t\\t template = self.pool.get('email.template').browse(cr,uid,template_id)\\n\\t\\t body = template.body_html\\n\\t if token:\\n \\t url = url + '/' + token\\n\\n\\t # post the message\\n \\t values = {\\n \\t'model': None,\\n\\t 'res_id': None,\\n \\t 'subject': template.subject,\\n \\t'body': body.replace(\\\"__URL__\\\", url),\\n\\t 'body_html': body.replace(\\\"__URL__\\\", url),\\n \\t 'parent_id': None,\\n \\t'partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'email_from': 'info@sba.org' or None,\\n \\t 'email_to': partner.email,\\n\\t\\t }\\n\\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\\n \\t mail_mail_obj.send(cr, uid, [mail_id], context=context)\",\n \"def send_reminder(self):\\n message_contents = \\\"This is a reminder that your event: \\\" + self.event_title + \\\" takes place on \\\" + self.event_date + \\\" in \\\" + self.event_location\\n subject = \\\"Event Reminder\\\"\\n attendees = self.gameplanuser_set.all()\\n for attendee in attendees:\\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\\n remindermessage.save()\",\n \"def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\\n\\n location = reverse(\\\"KursAnmeldung-verify_email\\\", kwargs={\\\"hash\\\":rnd_hash})\\n verify_link = request.build_absolute_uri(location)\\n\\n # FIXME: convert to users local time.\\n now = datetime.datetime.utcnow()\\n\\n email_context = {\\n \\\"verify_link\\\": verify_link,\\n \\\"db_entry\\\": db_entry,\\n \\\"now\\\": now,\\n }\\n\\n # Render the internal page\\n emailtext = render_to_string(\\\"kurs_anmeldung/verify_mailtext.txt\\\", email_context)\\n\\n # Get the preferences from the database:\\n raw_notify_list = preferences[\\\"notify\\\"]\\n notify_list = raw_notify_list.splitlines()\\n notify_list = [i.strip() for i in notify_list if i]\\n\\n email_kwargs = {\\n \\\"from_email\\\": preferences[\\\"from_email\\\"],\\n \\\"subject\\\": preferences[\\\"email_subject\\\"],\\n \\\"body\\\": emailtext,\\n \\\"to\\\": [db_entry.email],\\n \\\"bcc\\\": notify_list,\\n }\\n\\n if MAIL_DEBUG == True:\\n msg = u\\\"MAIL_DEBUG is on: No Email was sended!\\\"\\n request.page_msg(msg)\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n\\n request.page_msg(\\\"django.core.mail.EmailMessage kwargs:\\\")\\n request.page_msg(email_kwargs)\\n\\n request.page_msg(\\\"debug mail text:\\\")\\n request.page_msg(mark_safe(\\\"<pre>%s</pre>\\\" % emailtext))\\n return\\n\\n # We can't use django.core.mail.send_mail, because all members\\n # of the recipient list will see the others in the 'To' field.\\n # But we would like to notify the admins via 'Bcc' field.\\n\\n connection = SMTPConnection(fail_silently=False)\\n email = EmailMessage(**email_kwargs)\\n\\n try:\\n sended = email.send(fail_silently=False)\\n except Exception, err:\\n msg = \\\"Error sending mail: %s\\\" % err\\n LogEntry.objects.log_action(app_label=\\\"kurs_anmeldung\\\", action=\\\"error\\\",\\n message=msg\\n )\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n if settings.DEBUG or request.user.is_staff:\\n db_entry.save()\\n raise\\n else:\\n db_entry.mail_sended = sended\\n db_entry.log(request, \\\"mail sended: %s\\\" % sended)\",\n \"def confirm_email(self):\\n self.active = True\\n self.save()\",\n \"def save(self, register=False):\\n sha = sha1(self.email).hexdigest()\\n infos = self.to_dict()\\n infos[\\\"plan\\\"] = infos[\\\"plan\\\"][\\\"id\\\"] if infos[\\\"plan\\\"] else None\\n\\n if not redis.hmset(\\\"sl:account:{}\\\".format(sha), infos):\\n raise SleekException(\\\"Could not save current user.\\\", 401)\\n \\n if register:\\n try:\\n send_email.delay(\\n \\\"florent.esp@gmail.com\\\", \\\"Welcome to sleekstatus !\\\",\\n \\\"Welcome message\\\", [self.email]\\n )\\n except:\\n pass # Cannot send email\",\n \"def task_send_reminder_email():\\n send_reminder_email()\\n logger.info(\\\"Sent reminder email\\\")\",\n \"def account_activation_sent(request):\\r\\n\\treturn render(request, 'account_activation_sent.html')\",\n \"def save(self, *args, **kwargs):\\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\\n allUsers = Bruker.get_all_dict(Bruker)\\n mottaker = allUsers[int(data['mottaker'])]\\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\\n melding.save()\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n reminders.Patient.objects.filter(\\n pk__in=[self.test_patient.pk, self.other_patient.pk]\\n ).update(next_visit=appt_date)\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n self.startRouter()\\n self.router.logger.setLevel(logging.DEBUG)\\n\\n # run email job\\n from afrims.apps.reminders.app import daily_email_callback\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\\n self.stopRouter()\",\n \"def save_appointment_details(request, calendar_id):\\n def schedule_mail(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n def schedule_sms(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n start_time = request.GET['start_time'][:19]\\n end_time = request.GET['end_time'][:19]\\n \\n start_time = datetime.strptime(start_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n end_time=datetime.strptime(end_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n \\n calendar_obj = Calendar.objects.get(pk=calendar_id)\\n # if this is a POST request we need to process the form data\\n if request.method == 'POST':\\n\\n # create a form instance and populate it with data from the request:\\n form = AppointmentForm(request.POST)\\n\\n # check whether it's valid and save it\\n if form.is_valid():\\n # Save appointment details\\n \\n mobilephone = form.data['mobilephone']\\n email = form.data['email']\\n first_name = form.data['first_name']\\n last_name = form.data['last_name']\\n notes = form.data['notes']\\n\\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \\n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\\n \\n appointment.calendar = calendar_obj\\n appointment.save()\\n\\n try:\\n send_appointment_mail(appointment) # send appointment details email\\n except Exception as exp:\\n print(exp)\\n \\n try:\\n send_appointment_sms(appointment) # send appointment details sms\\n except Exception as exp:\\n print(exp)\\n \\n # Calculate reminder schedule dates\\n reminder1 = start_time - timedelta(hours=2)\\n reminder2 = start_time - timedelta(hours=24)\\n reminder3 = start_time - timedelta(days=7)\\n\\n # Schedule mails\\n schedule_mail(reminder1, appointment)\\n schedule_mail(reminder2, appointment)\\n schedule_mail(reminder3, appointment)\\n \\n # Schedule sms\\n schedule_sms(reminder1, appointment)\\n schedule_sms(reminder2, appointment)\\n schedule_sms(reminder3, appointment)\\n \\n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\\n \\n # if a GET (or any other method) we'll create a blank form\\n else:\\n form = AppointmentForm()\\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\\n 'office_location': calendar_obj.office_location})\",\n \"def post(\\n self,\\n email,\\n company_name,\\n location,\\n job_profile,\\n salary,\\n username,\\n password,\\n security_question,\\n security_answer,\\n notes,\\n date_applied,\\n status,\\n):\",\n \"def post_save_order_receiver(sender, instance, created, *args, **kwargs):\\n sender_email = instance.cargo.sender.email\\n recepient_email = instance.cargo.recepient.email\\n\\n booking_agent = instance.cargo.booking_station.branch_agent.email\\n\\n if created:\\n instance._set_order_price()\\n instance._set_time_approximations()\\n price = instance.price\\n\\n subject = \\\"Order Finalized and ready to go.\\\"\\n message = f\\\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\\\"\\n\\n send_async_email(\\n subject=subject,\\n message=message,\\n sender=booking_agent,\\n recepients=[sender_email, recepient_email],\\n )\",\n \"def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\\n\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n parts = pyzmail.parse.get_mail_parts(message)\\n body=''\\n for part in parts:\\n if part.is_body == 'text/plain' and part.disposition == None:\\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\\n body = body + payload + '\\\\n'\\n\\n msg = submit_message_from_message(message, body, by)\\n\\n if (submission_pk != None):\\n # Must exist - we're adding a message to an existing submission\\n submission = Submission.objects.get(pk=submission_pk)\\n else:\\n # Must not exist\\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\\n if submissions.count() > 0:\\n raise ValidationError(\\\"Submission {} already exists\\\".format(name))\\n \\n # create Submission using the name\\n try:\\n submission = Submission.objects.create(\\n state_id=\\\"waiting-for-draft\\\",\\n remote_ip=remote_ip,\\n name=name,\\n rev=rev,\\n title=name,\\n note=\\\"\\\",\\n submission_date=datetime.date.today(),\\n replaces=\\\"\\\",\\n )\\n from ietf.submit.utils import create_submission_event, docevent_from_submission\\n desc = \\\"Submission created for rev {} in response to email\\\".format(rev)\\n create_submission_event(request, \\n submission,\\n desc)\\n docevent_from_submission(request,\\n submission,\\n desc)\\n except Exception as e:\\n log(\\\"Exception: %s\\\\n\\\" % e)\\n raise\\n\\n if msgtype == 'msgin':\\n rs = \\\"Received\\\"\\n else:\\n rs = \\\"Sent\\\"\\n\\n desc = \\\"{} message - manual post - {}-{}\\\".format(rs, name, rev)\\n submission_email_event = SubmissionEmailEvent.objects.create(\\n desc = desc,\\n submission = submission,\\n msgtype = msgtype,\\n by = by,\\n message = msg)\\n #in_reply_to = in_reply_to\\n\\n save_submission_email_attachments(submission_email_event, parts)\\n return submission, submission_email_event\",\n \"def handle_sent(self, instance):\\n if not instance.created_by:\\n return\\n\\n activity = Activity(\\n actor=instance.created_by,\\n verb=RestrictedMailSent,\\n object=instance,\\n time=instance.used,\\n extra_context={},\\n )\\n self.manager.add_activity(\\n activity, [instance.created_by.pk], [NotificationFeed]\\n )\\n\\n # Send notification\\n notification = RestrictedMailSentNotification(instance.created_by)\\n notification.notify()\",\n \"def handle_failure(self, instance, sender=None, reason=None):\\n send_email.delay(\\n to_email=sender,\\n context={\\\"reason\\\": reason},\\n subject=\\\"Kunne ikke sende ut begrenset epost\\\",\\n plain_template=\\\"restricted/email/process_failure.txt\\\",\\n html_template=\\\"restricted/email/process_failure.html\\\",\\n )\",\n \"def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):\\n print(\\\"member email\\\",mem_mail)\\n logger.info(\\\"in sending existing customer mail task\\\")\\n return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website)\",\n \"def handle(self, *args, **options):\\n\\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\\n if candidate.contact_address and candidate.participating]\\n\\n\\n print 'sending e-mails'\\n conn = get_connection()\\n for c in candidates_with_email:\\n if c.should_send_reminder():\\n\\n print 'emailing', c\\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\\n c.last_reminder_sent = timezone.now()\\n c.save()\\n msg = make_email(c)\\n conn.send_messages([msg])\\n conn.close()\",\n \"def send(self, request, id, tribe_slug):\\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\\n document = self.get_document(id, tribe_slug)\\n \\n form = self._set_user_email_address(request)\\n email = self._get_user_email_address(request)\\n if form or not email:\\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\\n \\n # NB: Temporarily disabling actual email sending for development\\n #email_document(document, to=[email], subject='Document: %s' % document.title)\\n print \\\"Sending email to %s\\\" % email \\n \\n # Send a signal to let everyone know about this document interaction\\n document_interaction.send(sender=self, document=document, mode=\\\"sent\\\", request=request, recipient=email)\\n \\n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))\",\n \"def submit(self):\\r\\n try:\\r\\n response = self.send_request()\\r\\n if response.ok:\\r\\n self.submitted_at = datetime.now(pytz.UTC)\\r\\n self.status = \\\"submitted\\\"\\r\\n self.save()\\r\\n else:\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.error_msg = response.text\\r\\n self.save()\\r\\n except Exception as error:\\r\\n log.exception(error)\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.save()\",\n \"def _send(self, email_message): \\n\\t\\tif not email_message.recipients(): \\n\\t\\t\\treturn False \\n\\t\\trecipients = map(self._sanitize, email_message.recipients()) \\n\\t\\tMsg = o.CreateItem(0)\\n \\t\\tMsg.To = recipients\\n\\t\\tMsg.Subject = 'subject'\\n\\t\\tMsg.Body = 'text'\\n\\t\\tself.connection.SaveChanges(0)\\n\\n\\t\\tMsg.Send()\\n\\t\\treturn True\",\n \"def set_email():\\n bid_fields = {'assertion':request.form['bid_assertion'],\\n 'audience':settings.DOMAIN}\\n headers = {'Content-type':'application/x-www-form-urlencoded'}\\n h.disable_ssl_certificate_validation=True\\n resp, content = h.request('https://browserid.org/verify',\\n 'POST',\\n body=urlencode(bid_fields),\\n headers=headers)\\n bid_data = json.loads(content)\\n if bid_data['status'] == 'okay' and bid_data['email']:\\n user = f.userFromBidOrNew(bid_data['email'])\\n session['user_id'] = user.id\\n session['user_email'] = user.email\\n\\n return redirect(url_for('main'))\",\n \"def wasSent(self):\\n\\t\\tself.sent = True\\n\\t\\tself.save()\",\n \"def action_send_email(self):\\n self.ensure_one()\\n if self.communication_channel != \\\"email\\\":\\n return False # Maybe we should raise an error here.\\n\\n if self.state in (\\\"sent\\\", \\\"done\\\"):\\n raise ValidationError(_(\\\"This communication is already sent.\\\"))\\n\\n lines_2be_processed = self.credit_control_line_ids.filtered(\\n lambda line: line.state != \\\"sent\\\"\\n )\\n\\n if not lines_2be_processed:\\n raise ValidationError(_(\\\"There is no draft lines to send.\\\"))\\n\\n self = self.with_context(lang=self.partner_id.lang)\\n\\n partner = self.partner_id\\n mail_template = self.policy_level_id.email_template_id\\n # Send the email\\n partner.with_context(credit_control_mail=True).message_post_with_template(\\n template_id=mail_template.id,\\n model=self._name,\\n res_id=self.id,\\n )\\n # Set the state of the credit control lines to \\\"queued\\\"\\n lines_2be_processed.write({\\\"state\\\": \\\"sent\\\"})\\n self.state = \\\"sent\\\"\",\n \"def test_activate_form(self, mock_sendmail):\\r\\n res = self.testapp.post('/api/v1/suspend',\\r\\n params={'email': u'testing@dummy.com'},\\r\\n status=200)\\r\\n\\r\\n success = json.loads(res.body)\\r\\n self.assertTrue(\\r\\n 'message' in success,\\r\\n \\\"Should be successful with admin email address: \\\" + str(res))\\r\\n self.assertTrue(mock_sendmail.called)\",\n \"def test_post_internal_attendance_success(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-04T10:00',\\n 'attendee_email': 'student@anahuac.mx',\\n 'exit_datetime': '2021-08-04T11:30'}\\n response = self.client.post(\\n reverse('tracking:internal-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def save(self, *args, **kwargs):\\n self.expire_date = self.sent_date + timedelta(days=self.period)\\n super(Referral, self).save(*args, **kwargs)\",\n \"def post(self, request):\\n\\n try:\\n eventoid = request.POST.get('id', '')\\n correo = request.POST.get('correo', '')\\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\\n print(\\\"Exito en la asignación de staff\\\")\\n except:\\n print(\\\"Error en la asignacion de staff\\\")\\n\\n \\n return render(request, self.template, self.context)\\n #return render(request, self.template, self.context)\",\n \"def save(self, *args, **kwargs):\\n create = self.id is None\\n if create:\\n self.username = u'lc_' + self.city.lower()\\n # Give the user a randomly generated password\\n password = generate_random_string()\\n self.password = password\\n # Compose the email to send to the LC\\n context = {\\n 'lc': self,\\n 'password': password,\\n }\\n subject = render_to_string(\\n 'registration/lc_register_email_subject.txt')\\n subject = subject.strip()\\n message = render_to_string(\\n 'registration/lc_register_email_text.txt',\\n context)\\n send_mail(subject,\\n message,\\n settings.DEFAULT_FROM_EMAIL,\\n [self.email])\\n # Delegate upward to the parent classes\\n super(LocalCommittee, self).save(*args, **kwargs)\",\n \"def contact():\\n if request.method == 'POST':\\n send_email()\\n return \\\"\\\"\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n # run email job\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\",\n \"def save_model(self, request, obj, form, change):\\n admin.ModelAdmin.save_model(self, request, obj, form, change)\\n\\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\\n post_title = obj.title\\n post_url = 'http://www.aptitudeworld.com.au/blog'\\n\\n # send an email if we are adding a new post\\n if not change:\\n # send an email to all users in the system to notify them that a new post is available\\n email_title = 'Aptitude World ANZ - A new post is available'\\n email_from = 'info@aptitudeworld.com.au'\\n\\n try:\\n # get all user emails in the database\\n users = apps.users.models.User.objects.all()\\n confirmation_email_txt_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\\n confirmation_email_html_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\\n for u in users:\\n email_to = u.email_address\\n full_name = u.first_name + \\\" \\\" + u.last_name\\n\\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\\n\\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\\n email_message.attach_alternative(email_html, 'text/html')\\n email_message.send()\\n except BadHeaderError:\\n pass\",\n \"def send_email():\\n send_mail(\\\"You've got some problem.\\\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\\n ['dimazarj2009@rambler.ru'], fail_silently=False,)\",\n \"def create_sent_email(self, *args, **kwargs):\\n receiver = kwargs['receiver']\\n sender = kwargs['sender']\\n user = kwargs['user']\\n body = kwargs['body']\\n subject = kwargs['subject']\\n tracker_id = kwargs['tracker_id']\\n if receiver and sender and subject and body:\\n sent_email = SentEmail()\\n if tracker_id:\\n sent_email.tracker_id = tracker_id\\n sent_email.read_status = 'Tracking'\\n else:\\n sent_email.read_status = 'Not Tracking'\\n sent_email.receiver = receiver\\n sent_email.subject = subject\\n sent_email.sender = sender\\n sent_email.status = 'sent'\\n sent_email.user = user\\n sent_email.body = body\\n sent_email.save()\\n return sent_email\\n else:\\n return False\",\n \"def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\\n\\n if is_new:\\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\\n\\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))\",\n \"def sendEmail(householdID):\\n contactID = mdb.getContact(householdID)\\n sqlq = \\\"\\\"\\\"\\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\\n FROM Contact\\n WHERE idContact = '{}';\\n \\\"\\\"\\\".format(contactID)\\n result = mdb.getSQL(sqlq)[0]\\n\\n # thisName = (\\\"%s %s\\\" % (result['Name'], result['Surname']))\\n thisName = (\\\"%s\\\" % (result['Name']))\\n thisAddress = (\\\"%s</br>%s</br>%s %s\\\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\\n thisAddress = thisAddress.replace(\\\"None </br>\\\", \\\"\\\")\\n thisEmail = (\\\"%s\\\" % (result['email']))\\n thisStatus = (\\\"%s\\\" % (result['status']))\\n thisAddress = thisAddress.replace(\\\"None</br>\\\", \\\"\\\")\\n participantCount = (\\\"%s\\\" % mdb.getParticipantCount(str(householdID)))\\n # prepare the custom email\\n\\n thisPath = os.path.dirname(os.path.abspath(__file__))\\n if (thisStatus == 'de'):\\n # DOESN'T happen yet - de is excluded from query for now\\n # emailPath = os.path.join(thisPath, \\\"emails/email_confirm_de.html\\\")\\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\\n else:\\n emailPath = os.path.join(thisPath, \\\"emails/email_automated_date.html\\\")\\n\\n templateFile = open(emailPath, \\\"r\\\")\\n templateText = templateFile.read()\\n templateFile.close()\\n templateText = templateText.replace(\\\"[householdID]\\\", householdID)\\n templateText = templateText.replace(\\\"[contactID]\\\", contactID)\\n templateText = templateText.replace(\\\"[name]\\\", thisName)\\n templateText = templateText.replace(\\\"[address]\\\", thisAddress)\\n templateText = templateText.replace(\\\"[securityCode]\\\", mdb.getSecurityCode(householdID))\\n templateText = templateText.replace(\\\"[participantCount]\\\", participantCount)\\n\\n # Subject\\n subjectLine = templateText.splitlines()[0]\\n templateText = templateText[templateText.find('\\\\n') + 1:] # find line break and return all from there - i.e. remove first line\\n \\n # email file\\n emailFilePath = os.path.join(thisPath, \\\"tempEmail.htmail\\\")\\n\\n emailFile = open(emailFilePath, \\\"w+\\\")\\n emailFile.write(templateText)\\n emailFile.close()\\n\\n # call('mutt -e \\\"set content_type=text/html\\\" -s \\\"[TESTING]' + subjectLine + '\\\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\\n call('mutt -e \\\"set content_type=text/html\\\" -s \\\"' + subjectLine + '\\\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)\",\n \"def save(self,owner=None):\\n\\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\\n\\n t = Ticket(\\n title = self.cleaned_data['title'],\\n owner=owner,\\n submitter_email = self.cleaned_data['submitter_email'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n account = owner.account\\n )\\n\\n t.save()\\n\\n f = FollowUp(\\n ticket = t,\\n title = _('Ticket Opened Via Web'),\\n date = datetime.now(),\\n public = True,\\n comment = self.cleaned_data['body'],\\n account = owner.account\\n )\\n\\n f.save()\\n\\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n }\\n\\n messages_sent_to = []\\n\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self, user):\\n\\n q = self.cleaned_data['queue']\\n\\n t = Ticket( title = self.cleaned_data['title'],\\n submitter_email = self.cleaned_data['submitter_email'],\\n account = self.cleaned_data['account'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n owner = self.cleaned_data['owner']\\n )\\n\\n if HAS_TAG_SUPPORT:\\n t.tags = self.cleaned_data['tags']\\n\\n if self.cleaned_data['assigned_to']:\\n try:\\n t.assigned_to = self.cleaned_data['assigned_to']\\n except User.DoesNotExist:\\n t.assigned_to = None\\n t.save()\\n\\n f = FollowUp( ticket = t,\\n title = _('Ticket Opened'),\\n date = datetime.now(),\\n public = False,\\n comment = self.cleaned_data['body'],\\n systemuser = user.account,\\n )\\n if self.cleaned_data['assigned_to']:\\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\\n 'name': t.get_assigned_to\\n }\\n\\n f.save()\\n \\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n 'comment': f.comment,\\n }\\n \\n messages_sent_to = []\\n\\n if t.submitter_email:\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n #FIX USERSETTINGS\\n #=======================================================================\\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\\n # send_templated_mail(\\n # 'assigned_to',\\n # context,\\n # recipients=t.assigned_to.email,\\n # sender=q.from_address,\\n # fail_silently=True,\\n # files=files,\\n # )\\n # messages_sent_to.append(t.assigned_to.email)\\n #=======================================================================\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self):\\n email = self.validated_data['email']\\n name = self.validated_data['name']\\n password = self.validated_data['password']\\n\\n email_query = models.Email.objects.filter(address=email)\\n if email_query.exists():\\n email_instance = email_query.get()\\n\\n # If the email is already verified, we send a duplicate\\n # notification and exit.\\n if email_instance.is_verified:\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"is already verified.\\\",\\n email_instance,\\n )\\n email_instance.send_duplicate_notification()\\n\\n return\\n\\n # If the email is not verified, we send a new verification\\n # token to the address.\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"already exists. Sending a new verification token instead.\\\"\\n )\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\\n\\n return\\n\\n # The email doesn't exist, so we create a new user and email,\\n # then send a verification token to the email.\\n user = models.User.objects.create_user(name, password)\\n email_instance = models.Email.objects.create(address=email, user=user)\\n\\n # The user's primary email is their only email. This is the only\\n # time the primary email can be unverified.\\n user.primary_email = email_instance\\n user.save()\\n\\n logger.info(\\n \\\"Registered new user %r with email address %r\\\",\\n user,\\n email_instance,\\n )\\n\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\",\n \"def send_email(self, message):\\n pass\",\n \"def send_mail(self, subject):\\r\\n pass\",\n \"def post():\\n contactus_json = request.get_json()\\n\\n try:\\n dict_data = ContactUsSchema().load(contactus_json)\\n dict_data['description'] = escape(dict_data['description'])\\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\\n response, status = 'Received successfully', HTTPStatus.OK\\n except ValidationError as project_err:\\n response, status = {'systemErrors': project_err.messages}, \\\\\\n HTTPStatus.BAD_REQUEST\\n return response, status\",\n \"async def handle_email(self, email):\\n\\t\\tif 'logdir' in self.log_settings:\\n\\t\\t\\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\\n\\t\\t\\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\\n\\t\\t\\t\\tf.write(email.email.as_bytes())\\n\\n\\t\\tawait self.log('You got mail!')\",\n \"def set_email_sent(self, instance_id, creator, type):\\n table = self.type_map[type]\\n item = {\\n \\\"InstanceID\\\": instance_id,\\n \\\"Creator\\\": creator,\\n \\\"Scheduled For Deletion\\\": True,\\n \\\"EmailSent\\\": True\\n }\\n return table.put_item(item)\",\n \"def order_success(request, job_id, token):\\n if request.user.is_active:\\n job_obj = Job.objects.get(job_id= job_id)\\n if job_obj.employeer == request.user:\\n if verifiers.payment_token_is_valid(token, job_obj):\\n if job_obj.accepted == False:\\n job_obj.status = Status.objects.create(status_name=\\\"PN\\\", comment=\\\"Awaiting payment confirmation from Stripe, this might take a few minutes.\\\")\\n job_obj.accepted = True\\n job_obj.save()\\n \\n messages.add_message(request, messages.INFO, 'Your order has been created.')\\n return redirect('/') \\n else:\\n #someone tries to manually enter the URL with wrong payment_token\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n #the person making the request is not the same as the person who created the job\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n return redirect(\\\"/login\\\")\",\n \"def order_created(order_id):\\n order = Order.objects.get(id=order_id)\\n subject = f'Order nr. {order.id}'\\n message = f'Dear {order.first_name},\\\\n\\\\n' \\\\\\n f'You have successfully placed an order.' \\\\\\n f'Your order ID is {order.id}.'\\n mail_sent = send_mail(subject,\\n message,\\n 'info@tishman.com.ng',\\n [order.email])\\n return mail_sent\",\n \"def send_email_alert_slack(email_pk: str):\\n logging.info(\\\"Sending alert for Email<%s>\\\", email_pk)\\n email = Email.objects.get(pk=email_pk)\\n issue = email.issue\\n assert issue, f\\\"Email<{email_pk}> does not have an associated Issue\\\"\\n paralegal = issue.paralegal\\n alert_sent = False\\n case_email_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-email-list\\\", args=(str(issue.pk),)\\n )\\n case_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-detail-view\\\", args=(str(issue.pk),)\\n )\\n msg = (\\n \\\"*New Email Notification*\\\\n\\\"\\n f\\\"A new email has been received for case <{case_url}|{issue.fileref}> .\\\\n\\\"\\n f\\\"You can view this case's emails here: <{case_email_url}|here>.\\\\n\\\"\\n )\\n if paralegal and issue.is_open:\\n # Send alert directly to paralegal for open issues.\\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\\n logging.info(\\\"Looking up %s in Slack\\\", alert_email)\\n slack_user = get_slack_user_by_email(alert_email)\\n if slack_user:\\n logging.info(\\\"Notifying %s of Email<%s> via Slack\\\", alert_email, email_pk)\\n send_slack_direct_message(msg, slack_user[\\\"id\\\"])\\n alert_sent = True\\n\\n if not alert_sent:\\n # In all other cases send to alerts channel.\\n logging.info(\\n \\\"Could not find someone to DM in Slack, sending generic alert for Email<%s>\\\",\\n email_pk,\\n )\\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\\n\\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\\n logging.info(\\\"Alert sent sucessfully for Email<%s>\\\", email_pk)\",\n \"def send_sbemail(email_addy, city):\\n print(\\\"Single email send attempt: %s, %s\\\" % (email_addy, city))\\n given_pair, weather = subject_phrase_picker(city)\\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\\n print(\\\"Going to try a mailgun post to: %s with subject %s\\\" % (email_addy, given_pair['subject']))\\n response = requests.post(\\n MG_API_URL,\\n auth=(\\\"api\\\", MG_API_KEY),\\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\\n 'to': [email_addy],\\n 'subject': given_pair['subject'],\\n 'text': mailtext\\n })\\n print(\\\"status: %s | %s and response: %s\\\" % (response.status_code, response.reason, response.text))\\n print(\\\"Attempted send %s to: %s!\\\" % (mailtext, email_addy))\\n return \\\"Sent %s to: %s!\\\" % (mailtext, email_addy)\",\n \"def on_update(self):\\n\\t\\tfor email_account in frappe.get_all(\\\"Email Account\\\", filters={\\\"domain\\\": self.name}):\\n\\t\\t\\ttry:\\n\\t\\t\\t\\temail_account = frappe.get_doc(\\\"Email Account\\\", email_account.name)\\n\\t\\t\\t\\tfor attr in [\\\"email_server\\\", \\\"use_imap\\\", \\\"use_ssl\\\", \\\"use_tls\\\", \\\"attachment_limit\\\", \\\"smtp_server\\\", \\\"smtp_port\\\", \\\"use_ssl_for_outgoing\\\", \\\"append_emails_to_sent_folder\\\", \\\"incoming_port\\\"]:\\n\\t\\t\\t\\t\\temail_account.set(attr, self.get(attr, default=0))\\n\\t\\t\\t\\temail_account.save()\\n\\n\\t\\t\\texcept Exception as e:\\n\\t\\t\\t\\tfrappe.msgprint(_(\\\"Error has occurred in {0}\\\").format(email_account.name), raise_exception=e.__class__)\",\n \"def put(self, campaign_id):\\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\\n if campaign.candidates.count() > 0:\\n campaign.launch_task('generate_mailer_file')\\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\\n else:\\n api.abort(404, message='Campaign does not exist', success=False)\",\n \"def post_save_moderation(self, sender, comment, request, **kwargs):\\n model = comment.content_type.model_class()\\n if model not in self._registry:\\n return\\n self._registry[model].email(comment, comment.content_object, request)\",\n \"def delegate_last_day():\\n\\n regs = Registration.objects.all()\\n\\n template = 'notifications/last_day_mail.html'\\n\\n for reg in regs:\\n subject = 'SciPy.in 2011: Schedule and other details'\\n message = loader.render_to_string(\\n template, dictionary={'name': reg.registrant.username})\\n\\n reg.registrant.email_user(subject=subject, message=message,\\n from_email='madhusudancs@gmail.com')\",\n \"def send_email(email_dict, appointment_id):\\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)\",\n \"def post(self):\\n # se captura y se parsea a json el body del request recibido por el\\n # webhook\\n request_body = json.loads(self.request.body)\\n\\n for body in request_body:\\n \\\"\\\"\\\" Evaluar el tipo de evento ya que trae campos diferentes \\\"\\\"\\\"\\n logging.info(request_body)\\n\\n event = str(body['event'])\\n correo = str(body['email'])\\n numero_folio = str(body['numero_folio'])\\n tipo_dte = str(body['tipo_dte'])\\n\\n logging.info(event)\\n\\n if event and correo and numero_folio and tipo_dte:\\n\\n if event == 'processed':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.processed_event = event\\n email_model.processed_sg_event_id = body['sg_event_id']\\n email_model.processed_sg_message_id = body['sg_message_id']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.processed_event = event\\n e.processed_sg_event_id = body['sg_event_id']\\n e.processed_sg_message_id = body['sg_message_id']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'delivered':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.delivered_event = event\\n email_model.delivered_sg_event_id = body['sg_event_id']\\n email_model.delivered_sg_message_id = body['sg_message_id']\\n email_model.delivered_response = body['response']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.delivered_event = event\\n e.delivered_sg_event_id = body['sg_event_id']\\n e.delivered_sg_message_id = body['sg_message_id']\\n e.delivered_response = body['response']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'open':\\n model = EmailModel()\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n if email_model.opened_first_date == None:\\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_event = event\\n email_model.opened_ip = body['ip']\\n email_model.opened_user_agent = body['useragent']\\n email_model.opened_sg_event_id = body['sg_event_id']\\n email_model.opened_sg_message_id = body['sg_message_id']\\n model.email_add_count(email_model)\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n if e.opened_first_date == None:\\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_event = event\\n e.opened_ip = body['ip']\\n e.opened_user_agent = body['useragent']\\n e.opened_sg_event_id = body['sg_event_id']\\n e.opened_sg_message_id = body['sg_message_id']\\n e.email_add_count(e)\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'dropped':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.dropped_sg_event_id = body['sg_event_id']\\n email_model.dropped_sg_message_id = body['sg_message_id']\\n email_model.dropped_reason = body['reason']\\n email_model.dropped_event = event\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.dropped_sg_event_id = body['sg_event_id']\\n e.dropped_sg_message_id = body['sg_message_id']\\n e.dropped_reason = body['reason']\\n e.dropped_event = event\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'bounce':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.bounce_event = event\\n email_model.bounce_sg_event_id = body['sg_event_id']\\n email_model.bounce_sg_message_id = body['sg_message_id']\\n email_model.bounce_reason = body['reason']\\n email_model.bounce_status = body['status']\\n email_model.bounce_type = body['type']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.bounce_event = event\\n e.bounce_sg_event_id = body['sg_event_id']\\n e.bounce_sg_message_id = body['sg_message_id']\\n e.bounce_reason = str(body['reason']).decode(\\\"utf-8\\\")\\n e.bounce_status = body['status']\\n e.bounce_type = body['type']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'unsubscribe':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.unsubscribe_uid = body['uid']\\n email_model.unsubscribe_purchase = body['purchase']\\n email_model.unsubscribe_id = body['id']\\n email_model.unsubscribe_event = body['event']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.unsubscribe_uid = body['uid']\\n e.unsubscribe_purchase = body['purchase']\\n e.unsubscribe_id = body['id']\\n e.unsubscribe_event = body['event']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n else:\\n logging.info('body con campos vacios')\",\n \"def send_mail(self, address, title, message):\\n pass\",\n \"def mail_sent():\\n\\n url = settings.SITE_URL + '\\\\charts'\\n subject = 'Анализ запрошенного ресурса'\\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\\n mail_sent = send_mail(subject,\\n message,\\n 'admin@myshop.com',\\n ['user@mail.ru,'])\\n print(message)\\n return mail_sent\",\n \"def emailNote(self, authenticationToken, parameters):\\r\\n pass\",\n \"def send_new_email(user):\\n token = user.get_token()\\n message = Message(\\n 'Verify Your New Email',\\n sender='storcwebsite@gmail.com',\\n recipients=[user.temp_email])\\n message.body = f\\\"The email address associated with your Storc \\\" \\\\\\n f\\\"account has changed.\\\\n\\\\nTo verify your new email address, \\\" \\\\\\n f\\\"please click the link below:\\\\n\\\\n\\\" \\\\\\n f\\\"{url_for('users.new_email', token=token, _external=True)}\\\"\\n mail.send(message)\",\n \"def report_save(message, bot):\\n text = message.text\\n reports = parse_report(text)\\n author_id = message.from_user.id\\n author_username = message.from_user.username\\n to_save = []\\n for report in reports:\\n obj_to_save = {\\n 'author': author_id,\\n 'author_name': author_username,\\n 'text': report,\\n }\\n to_save.append(obj_to_save)\\n try:\\n request = requests.post(\\n url='http://127.0.0.1:8000/api/v1/report/',\\n headers={\\n 'Authorization': f'Bearer {jwt_token}',\\n },\\n json=to_save,\\n )\\n bot.send_message(\\n message.from_user.id,\\n request.status_code,\\n )\\n except requests.exceptions.ConnectionError:\\n bot.send_message(\\n message.from_user.id,\\n 'Ошибка соединения с сервером',\\n )\",\n \"def perform_create(self, serializer):\\n user = serializer.save()\\n signals.user_registered.send(\\n sender=self.__class__, user=user, request=self.request\\n )\\n\\n context = get_email_context(user)\\n to = [get_user_email(user)]\\n if djconf.SEND_ACTIVATION_EMAIL:\\n djconf.EMAIL.activation(self.request, context).send(to)\\n elif djconf.SEND_CONFIRMATION_EMAIL:\\n djconf.EMAIL.confirmation(self.request, context).send(to)\",\n \"def send(self):\\n ReferralActivity.objects.create(\\n actor=self.user,\\n verb=ReferralActivityVerb.CREATED,\\n referral=self,\\n )\\n # Confirm the referral has been sent to the requester by email\\n Mailer.send_referral_saved(self)\\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\\n # email notifications)\\n for unit in self.units.all():\\n contacts = unit.members.filter(\\n unitmembership__role=UnitMembershipRole.OWNER\\n )\\n for contact in contacts:\\n Mailer.send_referral_received(self, contact=contact, unit=unit)\",\n \"def execute_automatic_email(self, request, pk=None):\\n try:\\n retreat = Retreat.objects.get(pk=pk)\\n except Exception:\\n response_data = {\\n 'detail': \\\"Retreat not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n try:\\n email = AutomaticEmail.objects.get(\\n id=int(request.GET.get('email'))\\n )\\n except Exception:\\n response_data = {\\n 'detail': \\\"AutomaticEmail not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n # Notify a user for every reserved seat\\n emails = []\\n for reservation in retreat.reservations.filter(is_active=True):\\n if reservation.automatic_email_logs.filter(email=email):\\n pass\\n else:\\n send_automatic_email(reservation.user, retreat, email)\\n AutomaticEmailLog.objects.create(\\n reservation=reservation,\\n email=email\\n )\\n emails.append(reservation.user.email)\\n\\n response_data = {\\n 'stop': True,\\n 'emails': emails\\n }\\n return Response(response_data, status=status.HTTP_200_OK)\",\n \"def save(self):\\n owner = self.context[\\\"request\\\"].user\\n recipient = self._recipient_email_inst.user\\n\\n logger.info(\\n \\\"Transferring subscription %r from user %r to user %r\\\",\\n owner.know_me_subscription,\\n owner.pk,\\n recipient.pk,\\n )\\n\\n with transaction.atomic():\\n models.Subscription.objects.filter(user=recipient).delete()\\n models.Subscription.objects.filter(user=owner).update(\\n user=recipient\\n )\",\n \"def send_message(self):\\r\\n return \\\"success\\\"\",\n \"def alert_for_pending_mails_1(request):\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tThread(target=alert_for_pending_mails_1_worker).start()\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tresponse = {}\\n\\n\\tresponse[\\\"info_to_contact\\\"] = \\\"Ok\\\"\\n\\n\\treturn response\",\n \"def save(self, request, item, linked_item, linked_item_form):\\n\\t\\titem.description = linked_item_form.cleaned_data['subject']\\n\\t\\tlinked_item.text = linked_item_form.cleaned_data['text']\\n\\t\\tlinked_item.delivery_notes = linked_item_form.cleaned_data['delivery_notes']\\n\\t\\titem.priority = Priority.objects.get(id = linked_item_form.cleaned_data['priority'])\\n\\t\\t#\\n\\t\\t# Look for actions other than save\\n\\t\\t#\\n\\t\\tif request.POST.get('update', '') == 'Completed':\\n\\t\\t\\titem.fixed = True\\n\\t\\tif request.POST.get('update', '') == 'Failed':\\n\\t\\t\\titem.fixed = False\\n\\t\\t\\titem.validated = True\\n\\t\\tif request.POST.get('update', '') == 'Verified':\\n\\t\\t\\titem.fixed = True\\n\\t\\t\\titem.validated = True\\n\\n\\t\\titem.save()\\n\\t\\tlinked_item.save()\\n\\n\\t\\t#\\n\\t\\t# Check for comments and add\\n\\t\\t#\\n\\t\\tcomment_text = linked_item_form.cleaned_data['comments'].strip()\\n\\t\\tif comment_text != '':\\n\\t\\t\\tpost_comment(request, item, comment_text)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.67478585","0.64101136","0.61686337","0.61469436","0.61077154","0.61048436","0.60561985","0.6044016","0.60433817","0.5962845","0.59153545","0.5879812","0.5870707","0.5869888","0.58417296","0.5837118","0.5825478","0.58196616","0.57915473","0.5787865","0.57669544","0.574968","0.57293534","0.5725928","0.57227284","0.5714649","0.5712621","0.56969035","0.5689443","0.5681858","0.56777817","0.5677773","0.5672467","0.5663979","0.566298","0.5644856","0.5630909","0.56081456","0.56014544","0.5594918","0.5574689","0.55703187","0.5560041","0.5559436","0.55421066","0.55217606","0.55094653","0.55061436","0.55015546","0.5498323","0.54872906","0.5486167","0.54831564","0.5477915","0.5476625","0.54750633","0.54570115","0.5435575","0.5427781","0.5424537","0.5414667","0.5401194","0.5398877","0.53959084","0.53936905","0.53778183","0.5365919","0.53633624","0.5341283","0.5337631","0.532041","0.5315671","0.53116167","0.5303751","0.5300678","0.52967966","0.5288687","0.5277436","0.52724516","0.52620757","0.5260432","0.5252903","0.52519655","0.52476853","0.5245848","0.5243782","0.5242832","0.5242335","0.52420616","0.52394414","0.52360696","0.523167","0.5231379","0.52313536","0.52290714","0.5227354","0.5226612","0.52210313","0.521271","0.5211467"],"string":"[\n \"0.67478585\",\n \"0.64101136\",\n \"0.61686337\",\n \"0.61469436\",\n \"0.61077154\",\n \"0.61048436\",\n \"0.60561985\",\n \"0.6044016\",\n \"0.60433817\",\n \"0.5962845\",\n \"0.59153545\",\n \"0.5879812\",\n \"0.5870707\",\n \"0.5869888\",\n \"0.58417296\",\n \"0.5837118\",\n \"0.5825478\",\n \"0.58196616\",\n \"0.57915473\",\n \"0.5787865\",\n \"0.57669544\",\n \"0.574968\",\n \"0.57293534\",\n \"0.5725928\",\n \"0.57227284\",\n \"0.5714649\",\n \"0.5712621\",\n \"0.56969035\",\n \"0.5689443\",\n \"0.5681858\",\n \"0.56777817\",\n \"0.5677773\",\n \"0.5672467\",\n \"0.5663979\",\n \"0.566298\",\n \"0.5644856\",\n \"0.5630909\",\n \"0.56081456\",\n \"0.56014544\",\n \"0.5594918\",\n \"0.5574689\",\n \"0.55703187\",\n \"0.5560041\",\n \"0.5559436\",\n \"0.55421066\",\n \"0.55217606\",\n \"0.55094653\",\n \"0.55061436\",\n \"0.55015546\",\n \"0.5498323\",\n \"0.54872906\",\n \"0.5486167\",\n \"0.54831564\",\n \"0.5477915\",\n \"0.5476625\",\n \"0.54750633\",\n \"0.54570115\",\n \"0.5435575\",\n \"0.5427781\",\n \"0.5424537\",\n \"0.5414667\",\n \"0.5401194\",\n \"0.5398877\",\n \"0.53959084\",\n \"0.53936905\",\n \"0.53778183\",\n \"0.5365919\",\n \"0.53633624\",\n \"0.5341283\",\n \"0.5337631\",\n \"0.532041\",\n \"0.5315671\",\n \"0.53116167\",\n \"0.5303751\",\n \"0.5300678\",\n \"0.52967966\",\n \"0.5288687\",\n \"0.5277436\",\n \"0.52724516\",\n \"0.52620757\",\n \"0.5260432\",\n \"0.5252903\",\n \"0.52519655\",\n \"0.52476853\",\n \"0.5245848\",\n \"0.5243782\",\n \"0.5242832\",\n \"0.5242335\",\n \"0.52420616\",\n \"0.52394414\",\n \"0.52360696\",\n \"0.523167\",\n \"0.5231379\",\n \"0.52313536\",\n \"0.52290714\",\n \"0.5227354\",\n \"0.5226612\",\n \"0.52210313\",\n \"0.521271\",\n \"0.5211467\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6403527"},"document_rank":{"kind":"string","value":"2"}}},{"rowIdx":90867,"cells":{"query":{"kind":"string","value":"sends an email when lead saved successfully"},"document":{"kind":"string","value":"def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):\n print(\"member email\",mem_mail)\n logger.info(\"in sending existing customer mail task\")\n return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())","def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has been submitted by %s.<br><br>\n\nGo here to view or edit the request: <br>\n<a href=\"%s\">%s</a>\n<br><br>\nSincerely,<br><br>\nThe Janelia Parking Permit Program\n \"\"\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\n subject = 'A new parking permit request has been entered'\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\n text_content = re.sub(r'<[^>]+>','',message)\n html_content = message\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)","def save(self, commit=False):\n mail_result = self.send_email()\n if mail_result:\n self.instance.is_admin_notified = True\n\n contact = super().save(commit=commit)\n\n return contact","def save(self, **kwargs):\n\t\tif (self.answer != \"\"):\n\t\t\tself.answered = True\n\t\t\tsuper().save(**kwargs)\n\t\t\tmail_subject = \"Your Question has been answered.\"\n\t\t\t# Do not break the following string or the email will get cut off\n\t\t\tmessage = f\"Hi {self.user.first_name},\\n\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\n\\n\"\n\n\t\t\temail = EmailMessage(\n\t\t\t\tmail_subject, message, to=[self.user.email]\n\t\t\t)\n\t\t\temail.send()\n\t\telse:\n\t\t\tsuper().save(**kwargs)","def save(self): # noqa\n self.bound_admin.message_success(self.bound_request, _('Done.'))","def action_my_payslip_sent(self):\n self.ensure_one()\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\n if template:\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\n self.flag = True","def test_email_sent_on_failure(self):\n self._authorize()\n data = {\n 'Subject_Number': '000-1111',\n 'Pin_Code': '1234',\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\n 'Mobile_Number': '2223334444',\n }\n patient = self.create_xml_patient(data)\n payload = self.create_xml_payload([patient])\n response = self._post(payload)\n self.assertEqual(response.status_code, 500)\n self.assertEqual(len(mail.outbox), 1)","def save_model(self, request, obj, form, change):\n if change and obj.status == 'V':\n if Organization.objects.get(pk=obj.pk).status == 'P':\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\n try:\n sender = Plugin_Contact.objects.all()[0].email\n except IndexError:\n sender = None\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\n send_mail(u'Validation de votre fiche sur la PASR',\n u'Bonjour,\\n\\nVotre fiche vient d\\'être validée sur la PASR.',\n sender, dests)\n super(OrganizationAdmin, self).save_model(request, obj, form, change)","def send_referral_saved(cls, referral, created_by):\n\n template_id = settings.SENDINBLUE[\"REFERRAL_SAVED_TEMPLATE_ID\"]\n\n data = {\n \"params\": {\"case_number\": referral.id},\n \"replyTo\": cls.reply_to,\n \"templateId\": template_id,\n \"to\": [{\"email\": created_by.email}],\n }\n\n cls.send(data)","def callback_success_message(request):\n msg = 'Sucessfully recorded form :)'\n logger.info(msg)\n messages.info(request._request, msg)","def save(self):\n super(Notification, self).save()\n # get user recipent\n us = self.notified_user\n # check that user has a valid email address\n if us.email.find('@') > 0 and us.email.find('.') > 0:\n # mandatory fields\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\n to = us.email\n from_email = settings.DEFAULT_FROM_EMAIL\n # get text version of the message\n text_content = self.get_email_content_from_type(\n self.notification_type\n )\n # FIXME: HTML version implementation pending\n html_content = self.get_email_content_from_type(\n self.notification_type\n )\n msg = EmailMultiAlternatives(\n subject, \n text_content,\n from_email,\n [to]\n )\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)","def mail(request):\n email_admin.delay('testinggg')\n return JsonResponse({\"details\":\"working\"})","def on_post(self):\n return \"Ok, the stuff is being saved\"","def send_email(request):\n # send emails and return some manner of success response\n send(**request.params)\n return {'success': 'mail sent!'}","def unsuccessful(self, **post):\n cr, uid, context = request.cr, request.uid, request.context\n attchment_list = []\n mail_obj = request.env['mail.mail'].sudo()\n\n sale_order_id = request.session.get('sale_order_id')\n\n print(\"\\n\\n\\n\\n\\n=========sale order id===========\", sale_order_id)\n if sale_order_id:\n order = request.env['sale.order'].sudo().browse(sale_order_id)\n print(\"\\n\\n\\n\\n\\n=========sale order order===========\", order)\n else:\n return request.redirect('/enrolment_book')\n request.website.sale_reset()\n\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\n raise_if_not_found=False)\n if template_id:\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\n order.id)\n enroll_file_name = \"Pro-Forma \" + order.name\n if pdf_data_enroll:\n pdfvals = {'name': enroll_file_name,\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas_fname': enroll_file_name + '.pdf',\n 'res_model': 'sale.order',\n 'type': 'binary'}\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\n attchment_list.append(pdf_create)\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\n if agreement_id:\n attchment_list.append(agreement_id)\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\n if banking_detail_id:\n attchment_list.append(banking_detail_id)\n\n # mail_values = {\n # 'email_from': template_id.email_from,\n # 'reply_to': template_id.reply_to,\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # 'subject': \"Charterquest FreeQuote/Enrolment \" + order.name,\n # 'notification': True,\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\n # 'auto_delete': False,\n # }\n # msg_id = mail_obj.create(mail_values)\n # msg_id.send()\n\n template_id.sudo().write(\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\n template_id.sudo().with_context(\n # email_to=each_request.get('email'),\n # event_list=event_list,\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # prof_body=invoice_id.prof_body.name,\n ).send_mail(order.id, force_send=True)\n return request.render(\"cfo_snr_jnr.event_unsuccessful\", {'order': order})","def claim_email(request):\n email = request.POST.get('email', '')\n email_user = User.objects.filter(email=email)\n payload = {\n 'res': 'failed'\n }\n if email_user.exists() and \\\n not email_user[0].profile.send_mail:\n request.user.profile.add_email(email)\n payload['res'] = 'success'\n\n return payload","def test_admin_approval_complete_email(self):\n new_user = UserModel().objects.create_user(**self.user_info)\n profile = self.registration_profile.objects.create_profile(new_user)\n profile.send_admin_approve_complete_email(Site.objects.get_current())\n self.assertEqual(len(mail.outbox), 1)\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])","def add_manualpost_email(request, submission_id=None, access_token=None):\n\n if request.method == 'POST':\n try:\n button_text = request.POST.get('submit', '')\n if button_text == 'Cancel':\n return redirect(\"submit/manual_post.html\")\n \n form = SubmissionEmailForm(request.POST)\n if form.is_valid():\n submission_pk = form.cleaned_data['submission_pk']\n message = form.cleaned_data['message']\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n \n if form.cleaned_data['direction'] == 'incoming':\n msgtype = 'msgin'\n else:\n msgtype = 'msgout'\n \n submission, submission_email_event = (\n add_submission_email(request=request,\n remote_ip=request.META.get('REMOTE_ADDR', None),\n name = form.draft_name,\n rev=form.revision,\n submission_pk = submission_pk,\n message = message,\n by = request.user.person,\n msgtype = msgtype) )\n \n messages.success(request, 'Email added.')\n \n try:\n draft = Document.objects.get(name=submission.name)\n except Document.DoesNotExist:\n # Assume this is revision 00 - we'll do this later\n draft = None\n \n if (draft != None):\n e = AddedMessageEvent(type=\"added_message\", doc=draft)\n e.message = submission_email_event.submissionemailevent.message\n e.msgtype = submission_email_event.submissionemailevent.msgtype\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\n e.by = request.user.person\n e.desc = submission_email_event.desc\n e.time = submission_email_event.time\n e.save()\n \n return redirect(\"ietf.submit.views.manualpost\")\n except ValidationError as e:\n form = SubmissionEmailForm(request.POST)\n form._errors = {}\n form._errors[\"__all__\"] = form.error_class([\"There was a failure uploading your message. (%s)\" % e.message])\n else:\n initial = {\n }\n\n if (submission_id != None):\n submission = get_submission_or_404(submission_id, access_token)\n initial['name'] = \"{}-{}\".format(submission.name, submission.rev)\n initial['direction'] = 'incoming'\n initial['submission_pk'] = submission.pk\n else:\n initial['direction'] = 'incoming'\n \n form = SubmissionEmailForm(initial=initial)\n\n return render(request, 'submit/add_submit_email.html',dict(form=form))","def receive_email_view(request):\n save_inbound_email(request.POST, request.FILES)\n return HttpResponse(200)","def validation_email_sent(request):\n assert(settings.EMAIL_VALIDATION == True)\n logging.debug('')\n data = {\n 'email': request.user.email,\n 'change_email_url': reverse('user_changeemail'),\n 'action_type': 'validate'\n }\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))","def postprocess():\n if ERRORS:\n address = 'tamakoshihiroki@gmail.com'\n body = '\\n\\n'.join( ERRORS )\n msg = create_message( body, address )\n send_mail( msg, address )","def send_mail_when_failed(self, body):\r\n pass","def test_send_email_on_invite(self):\n\n league = self.create_league()\n\n season = self.create_season(league)\n team = self.create_team(season)\n\n player = self.create_player()\n\n send_user_email_on_join(player, team.id)\n\n self.assertEqual(len(mail.outbox), 1)\n\n # if testing manually:\n # import pathlib\n # pathlib.Path(\"test_email.html\").write_text(last_sent.body)","def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\n lead = Lead.objects.get(id=lead_id)\n created_by = lead.created_by\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\n for user in recipients:\n recipients_list = []\n user = User.objects.filter(id=user, is_active=True).first()\n if user:\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\n recipients_list.append(user.email)\n context = {}\n context[\"url\"] = protocol + '://' + domain + \\\n reverse('leads:view_lead', args=(lead.id,))\n context[\"user\"] = user\n context[\"lead\"] = lead\n context[\"created_by\"] = created_by\n context[\"source\"] = source\n subject = 'Assigned a lead for you. '\n html_content = render_to_string(\n 'assigned_to/leads_assigned.html', context=context)\n msg = EmailMessage(\n subject,\n html_content,\n to=recipients_list\n )\n msg.content_subtype = \"html\"\n msg.send()","def form_valid(self, form):\n self.object = form.save()\n self.send_verify_email()\n return super().form_valid(form)","def save_response(self):\r\n self.q(css='input.save-button').first.click()\r\n EmptyPromise(\r\n lambda: 'save' in self.alert_message.lower(),\r\n \"Status message saved\"\r\n ).fulfill()","def form_valid(self, form):\n\n ModelAccountVerification.objects.send_verification_email(form.user,\n verification_type=\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\n\n messages.info(self.request, \"Email sent successfully to the {0}\".format(form.user.email))\n\n return HttpResponseRedirect(self.success_url)","def execute(self):\n return LOGGER.info(f\"{datetime.datetime.now()} - Sending EMail to the configured email list\")","def test_post_external_attendance_successfull(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-03T19:00',\n 'attendee_email': 'student@external.mx',\n 'exit_datetime': '2021-08-03T20:00'}\n response = self.client.post(\n reverse('tracking:external-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def save_model(self, request, obj, form, change):\n notify.send(\n request.user,\n recipient=obj.reported_by,\n verb='updated your report',\n action_object=obj,\n description=obj.notes,\n target=obj\n )\n obj.save()","def post(self):\n return send_email(request.args)","def send_warning(self):\n\n # Check whether all the necessary parameters for SMS are present\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\n client = Client(self.account_sid, self.auth_token)\n\n try:\n sms = client.messages.create(\n body=\"\"\"Last will: It was at least 30 days since your last check in. \n This is a reminder to check in in the next 24 hours.\"\"\",\n from_=self.twilio_phone,\n to=self.your_phone)\n sms\n print(\"\\nSMS sent\")\n except Exception as e:\n print(f\"An error occurred while trying to send the SMS. Error: {e}\")\n\n else:\n print(\"\\nMissing SMS parameters. SMS not sent\")\n\n # Check whether all the necessary parameters for email are present\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\n message = f\"\"\"It has been at least 30 days since you last checked in. \nYou need to check in in the next 24 hours.\\n\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\n\nIn order to reset simply go to the working directory and run python3 last_will.py\"\"\"\n\n # send_email will return 0 if everything went ok, otherwise it will return an error message\n status = send_email(self.sender_name, self.your_email,\n self.email, self.email_pwd,\n subject='Last will: Reminder to check in', unencrypted_message=message)\n\n if status != 0:\n print(status)\n exit(1)\n else:\n print(\"Email sent\\n\")\n\n print(f\"You have until {self.deadline} to check in. \"\n f\"In order to do that simply go to the working directory and run ./last_will.sh\\n\")\n else:\n print(\"Missing email parameters. Email not sent.\\n\")\n exit(1)","def send_reminder(self):\n pass","def complete_allotment(request):\n reg_no = request.user.username\n user = get_object_or_404(User, username=reg_no)\n sec_email = user.get_profile().secondary_email\n options_chosen = get_chosen_options(user)\n context = {'username': reg_no, 'email': sec_email, \n 'options_chosen': options_chosen}\n ##Sending mail with allotment details \n admin = User.objects.get(pk=1) \n from_email = admin.email\n subject = \"JAM 2012 admissions\"\n content = \"The following options were chosen by you \\n \\n\"\n if options_chosen:\n counter = 1\n for option in options_chosen:\n content += \"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\n\" %(counter, option.opt_code, option.opt_name, option.opt_location)\n counter += 1 \n \n content += \"\\n \\n \\nPlease do not delete this email and keep it for reference purposes. \\n \\n \\n \\n Regards, \\n JAM Office, IIT Bombay\"\n send_mail(subject, content, from_email, [sec_email])\n admin_content = content\n admin_content +=\"\\n\\n\\n#%s:\" % (reg_no)\n counter = 1\n for option in options_chosen:\n admin_content += \"%s,%s:\" %(counter, option.opt_code) \n counter += 1\n admin_content +=\"#\"\n admin_content += time.ctime() \n mail_admins(subject, admin_content) \n return render(request, 'allotter/complete.html', context)","def create_response_and_send_mail(token, partner_id, email):\n \t #set url\n\t url = survey.public_url\n \t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\n\t\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\n\t\t template = self.pool.get('email.template').browse(cr,uid,template_id)\n\t\t body = template.body_html\n\t if token:\n \t url = url + '/' + token\n\n\t # post the message\n \t values = {\n \t'model': None,\n\t 'res_id': None,\n \t 'subject': template.subject,\n \t'body': body.replace(\"__URL__\", url),\n\t 'body_html': body.replace(\"__URL__\", url),\n \t 'parent_id': None,\n \t'partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'email_from': 'info@sba.org' or None,\n \t 'email_to': partner.email,\n\t\t }\n\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\n \t mail_mail_obj.send(cr, uid, [mail_id], context=context)","def send_reminder(self):\n message_contents = \"This is a reminder that your event: \" + self.event_title + \" takes place on \" + self.event_date + \" in \" + self.event_location\n subject = \"Event Reminder\"\n attendees = self.gameplanuser_set.all()\n for attendee in attendees:\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\n remindermessage.save()","def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\n\n location = reverse(\"KursAnmeldung-verify_email\", kwargs={\"hash\":rnd_hash})\n verify_link = request.build_absolute_uri(location)\n\n # FIXME: convert to users local time.\n now = datetime.datetime.utcnow()\n\n email_context = {\n \"verify_link\": verify_link,\n \"db_entry\": db_entry,\n \"now\": now,\n }\n\n # Render the internal page\n emailtext = render_to_string(\"kurs_anmeldung/verify_mailtext.txt\", email_context)\n\n # Get the preferences from the database:\n raw_notify_list = preferences[\"notify\"]\n notify_list = raw_notify_list.splitlines()\n notify_list = [i.strip() for i in notify_list if i]\n\n email_kwargs = {\n \"from_email\": preferences[\"from_email\"],\n \"subject\": preferences[\"email_subject\"],\n \"body\": emailtext,\n \"to\": [db_entry.email],\n \"bcc\": notify_list,\n }\n\n if MAIL_DEBUG == True:\n msg = u\"MAIL_DEBUG is on: No Email was sended!\"\n request.page_msg(msg)\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n\n request.page_msg(\"django.core.mail.EmailMessage kwargs:\")\n request.page_msg(email_kwargs)\n\n request.page_msg(\"debug mail text:\")\n request.page_msg(mark_safe(\"<pre>%s</pre>\" % emailtext))\n return\n\n # We can't use django.core.mail.send_mail, because all members\n # of the recipient list will see the others in the 'To' field.\n # But we would like to notify the admins via 'Bcc' field.\n\n connection = SMTPConnection(fail_silently=False)\n email = EmailMessage(**email_kwargs)\n\n try:\n sended = email.send(fail_silently=False)\n except Exception, err:\n msg = \"Error sending mail: %s\" % err\n LogEntry.objects.log_action(app_label=\"kurs_anmeldung\", action=\"error\",\n message=msg\n )\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n if settings.DEBUG or request.user.is_staff:\n db_entry.save()\n raise\n else:\n db_entry.mail_sended = sended\n db_entry.log(request, \"mail sended: %s\" % sended)","def confirm_email(self):\n self.active = True\n self.save()","def save(self, register=False):\n sha = sha1(self.email).hexdigest()\n infos = self.to_dict()\n infos[\"plan\"] = infos[\"plan\"][\"id\"] if infos[\"plan\"] else None\n\n if not redis.hmset(\"sl:account:{}\".format(sha), infos):\n raise SleekException(\"Could not save current user.\", 401)\n \n if register:\n try:\n send_email.delay(\n \"florent.esp@gmail.com\", \"Welcome to sleekstatus !\",\n \"Welcome message\", [self.email]\n )\n except:\n pass # Cannot send email","def task_send_reminder_email():\n send_reminder_email()\n logger.info(\"Sent reminder email\")","def account_activation_sent(request):\r\n\treturn render(request, 'account_activation_sent.html')","def save(self, *args, **kwargs):\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\n allUsers = Bruker.get_all_dict(Bruker)\n mottaker = allUsers[int(data['mottaker'])]\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\n melding.save()","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n reminders.Patient.objects.filter(\n pk__in=[self.test_patient.pk, self.other_patient.pk]\n ).update(next_visit=appt_date)\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n self.startRouter()\n self.router.logger.setLevel(logging.DEBUG)\n\n # run email job\n from afrims.apps.reminders.app import daily_email_callback\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)\n self.stopRouter()","def save_appointment_details(request, calendar_id):\n def schedule_mail(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n def schedule_sms(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n start_time = request.GET['start_time'][:19]\n end_time = request.GET['end_time'][:19]\n \n start_time = datetime.strptime(start_time, \"%Y-%m-%dT%H:%M:%S\")\n end_time=datetime.strptime(end_time, \"%Y-%m-%dT%H:%M:%S\")\n \n calendar_obj = Calendar.objects.get(pk=calendar_id)\n # if this is a POST request we need to process the form data\n if request.method == 'POST':\n\n # create a form instance and populate it with data from the request:\n form = AppointmentForm(request.POST)\n\n # check whether it's valid and save it\n if form.is_valid():\n # Save appointment details\n \n mobilephone = form.data['mobilephone']\n email = form.data['email']\n first_name = form.data['first_name']\n last_name = form.data['last_name']\n notes = form.data['notes']\n\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\n \n appointment.calendar = calendar_obj\n appointment.save()\n\n try:\n send_appointment_mail(appointment) # send appointment details email\n except Exception as exp:\n print(exp)\n \n try:\n send_appointment_sms(appointment) # send appointment details sms\n except Exception as exp:\n print(exp)\n \n # Calculate reminder schedule dates\n reminder1 = start_time - timedelta(hours=2)\n reminder2 = start_time - timedelta(hours=24)\n reminder3 = start_time - timedelta(days=7)\n\n # Schedule mails\n schedule_mail(reminder1, appointment)\n schedule_mail(reminder2, appointment)\n schedule_mail(reminder3, appointment)\n \n # Schedule sms\n schedule_sms(reminder1, appointment)\n schedule_sms(reminder2, appointment)\n schedule_sms(reminder3, appointment)\n \n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\n \n # if a GET (or any other method) we'll create a blank form\n else:\n form = AppointmentForm()\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\n 'office_location': calendar_obj.office_location})","def post(\n self,\n email,\n company_name,\n location,\n job_profile,\n salary,\n username,\n password,\n security_question,\n security_answer,\n notes,\n date_applied,\n status,\n):","def post_save_order_receiver(sender, instance, created, *args, **kwargs):\n sender_email = instance.cargo.sender.email\n recepient_email = instance.cargo.recepient.email\n\n booking_agent = instance.cargo.booking_station.branch_agent.email\n\n if created:\n instance._set_order_price()\n instance._set_time_approximations()\n price = instance.price\n\n subject = \"Order Finalized and ready to go.\"\n message = f\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\"\n\n send_async_email(\n subject=subject,\n message=message,\n sender=booking_agent,\n recepients=[sender_email, recepient_email],\n )","def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\n\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n parts = pyzmail.parse.get_mail_parts(message)\n body=''\n for part in parts:\n if part.is_body == 'text/plain' and part.disposition == None:\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\n body = body + payload + '\\n'\n\n msg = submit_message_from_message(message, body, by)\n\n if (submission_pk != None):\n # Must exist - we're adding a message to an existing submission\n submission = Submission.objects.get(pk=submission_pk)\n else:\n # Must not exist\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\n if submissions.count() > 0:\n raise ValidationError(\"Submission {} already exists\".format(name))\n \n # create Submission using the name\n try:\n submission = Submission.objects.create(\n state_id=\"waiting-for-draft\",\n remote_ip=remote_ip,\n name=name,\n rev=rev,\n title=name,\n note=\"\",\n submission_date=datetime.date.today(),\n replaces=\"\",\n )\n from ietf.submit.utils import create_submission_event, docevent_from_submission\n desc = \"Submission created for rev {} in response to email\".format(rev)\n create_submission_event(request, \n submission,\n desc)\n docevent_from_submission(request,\n submission,\n desc)\n except Exception as e:\n log(\"Exception: %s\\n\" % e)\n raise\n\n if msgtype == 'msgin':\n rs = \"Received\"\n else:\n rs = \"Sent\"\n\n desc = \"{} message - manual post - {}-{}\".format(rs, name, rev)\n submission_email_event = SubmissionEmailEvent.objects.create(\n desc = desc,\n submission = submission,\n msgtype = msgtype,\n by = by,\n message = msg)\n #in_reply_to = in_reply_to\n\n save_submission_email_attachments(submission_email_event, parts)\n return submission, submission_email_event","def handle_sent(self, instance):\n if not instance.created_by:\n return\n\n activity = Activity(\n actor=instance.created_by,\n verb=RestrictedMailSent,\n object=instance,\n time=instance.used,\n extra_context={},\n )\n self.manager.add_activity(\n activity, [instance.created_by.pk], [NotificationFeed]\n )\n\n # Send notification\n notification = RestrictedMailSentNotification(instance.created_by)\n notification.notify()","def handle_failure(self, instance, sender=None, reason=None):\n send_email.delay(\n to_email=sender,\n context={\"reason\": reason},\n subject=\"Kunne ikke sende ut begrenset epost\",\n plain_template=\"restricted/email/process_failure.txt\",\n html_template=\"restricted/email/process_failure.html\",\n )","def handle(self, *args, **options):\n\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\n if candidate.contact_address and candidate.participating]\n\n\n print 'sending e-mails'\n conn = get_connection()\n for c in candidates_with_email:\n if c.should_send_reminder():\n\n print 'emailing', c\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\n c.last_reminder_sent = timezone.now()\n c.save()\n msg = make_email(c)\n conn.send_messages([msg])\n conn.close()","def send(self, request, id, tribe_slug):\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\n document = self.get_document(id, tribe_slug)\n \n form = self._set_user_email_address(request)\n email = self._get_user_email_address(request)\n if form or not email:\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\n \n # NB: Temporarily disabling actual email sending for development\n #email_document(document, to=[email], subject='Document: %s' % document.title)\n print \"Sending email to %s\" % email \n \n # Send a signal to let everyone know about this document interaction\n document_interaction.send(sender=self, document=document, mode=\"sent\", request=request, recipient=email)\n \n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))","def submit(self):\r\n try:\r\n response = self.send_request()\r\n if response.ok:\r\n self.submitted_at = datetime.now(pytz.UTC)\r\n self.status = \"submitted\"\r\n self.save()\r\n else:\r\n self.status = \"must_retry\"\r\n self.error_msg = response.text\r\n self.save()\r\n except Exception as error:\r\n log.exception(error)\r\n self.status = \"must_retry\"\r\n self.save()","def _send(self, email_message): \n\t\tif not email_message.recipients(): \n\t\t\treturn False \n\t\trecipients = map(self._sanitize, email_message.recipients()) \n\t\tMsg = o.CreateItem(0)\n \t\tMsg.To = recipients\n\t\tMsg.Subject = 'subject'\n\t\tMsg.Body = 'text'\n\t\tself.connection.SaveChanges(0)\n\n\t\tMsg.Send()\n\t\treturn True","def set_email():\n bid_fields = {'assertion':request.form['bid_assertion'],\n 'audience':settings.DOMAIN}\n headers = {'Content-type':'application/x-www-form-urlencoded'}\n h.disable_ssl_certificate_validation=True\n resp, content = h.request('https://browserid.org/verify',\n 'POST',\n body=urlencode(bid_fields),\n headers=headers)\n bid_data = json.loads(content)\n if bid_data['status'] == 'okay' and bid_data['email']:\n user = f.userFromBidOrNew(bid_data['email'])\n session['user_id'] = user.id\n session['user_email'] = user.email\n\n return redirect(url_for('main'))","def wasSent(self):\n\t\tself.sent = True\n\t\tself.save()","def action_send_email(self):\n self.ensure_one()\n if self.communication_channel != \"email\":\n return False # Maybe we should raise an error here.\n\n if self.state in (\"sent\", \"done\"):\n raise ValidationError(_(\"This communication is already sent.\"))\n\n lines_2be_processed = self.credit_control_line_ids.filtered(\n lambda line: line.state != \"sent\"\n )\n\n if not lines_2be_processed:\n raise ValidationError(_(\"There is no draft lines to send.\"))\n\n self = self.with_context(lang=self.partner_id.lang)\n\n partner = self.partner_id\n mail_template = self.policy_level_id.email_template_id\n # Send the email\n partner.with_context(credit_control_mail=True).message_post_with_template(\n template_id=mail_template.id,\n model=self._name,\n res_id=self.id,\n )\n # Set the state of the credit control lines to \"queued\"\n lines_2be_processed.write({\"state\": \"sent\"})\n self.state = \"sent\"","def test_activate_form(self, mock_sendmail):\r\n res = self.testapp.post('/api/v1/suspend',\r\n params={'email': u'testing@dummy.com'},\r\n status=200)\r\n\r\n success = json.loads(res.body)\r\n self.assertTrue(\r\n 'message' in success,\r\n \"Should be successful with admin email address: \" + str(res))\r\n self.assertTrue(mock_sendmail.called)","def test_post_internal_attendance_success(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-04T10:00',\n 'attendee_email': 'student@anahuac.mx',\n 'exit_datetime': '2021-08-04T11:30'}\n response = self.client.post(\n reverse('tracking:internal-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def save(self, *args, **kwargs):\n self.expire_date = self.sent_date + timedelta(days=self.period)\n super(Referral, self).save(*args, **kwargs)","def post(self, request):\n\n try:\n eventoid = request.POST.get('id', '')\n correo = request.POST.get('correo', '')\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\n print(\"Exito en la asignación de staff\")\n except:\n print(\"Error en la asignacion de staff\")\n\n \n return render(request, self.template, self.context)\n #return render(request, self.template, self.context)","def save(self, *args, **kwargs):\n create = self.id is None\n if create:\n self.username = u'lc_' + self.city.lower()\n # Give the user a randomly generated password\n password = generate_random_string()\n self.password = password\n # Compose the email to send to the LC\n context = {\n 'lc': self,\n 'password': password,\n }\n subject = render_to_string(\n 'registration/lc_register_email_subject.txt')\n subject = subject.strip()\n message = render_to_string(\n 'registration/lc_register_email_text.txt',\n context)\n send_mail(subject,\n message,\n settings.DEFAULT_FROM_EMAIL,\n [self.email])\n # Delegate upward to the parent classes\n super(LocalCommittee, self).save(*args, **kwargs)","def contact():\n if request.method == 'POST':\n send_email()\n return \"\"","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n # run email job\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)","def save_model(self, request, obj, form, change):\n admin.ModelAdmin.save_model(self, request, obj, form, change)\n\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\n post_title = obj.title\n post_url = 'http://www.aptitudeworld.com.au/blog'\n\n # send an email if we are adding a new post\n if not change:\n # send an email to all users in the system to notify them that a new post is available\n email_title = 'Aptitude World ANZ - A new post is available'\n email_from = 'info@aptitudeworld.com.au'\n\n try:\n # get all user emails in the database\n users = apps.users.models.User.objects.all()\n confirmation_email_txt_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\n confirmation_email_html_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\n for u in users:\n email_to = u.email_address\n full_name = u.first_name + \" \" + u.last_name\n\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\n\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\n email_message.attach_alternative(email_html, 'text/html')\n email_message.send()\n except BadHeaderError:\n pass","def send_email():\n send_mail(\"You've got some problem.\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\n ['dimazarj2009@rambler.ru'], fail_silently=False,)","def create_sent_email(self, *args, **kwargs):\n receiver = kwargs['receiver']\n sender = kwargs['sender']\n user = kwargs['user']\n body = kwargs['body']\n subject = kwargs['subject']\n tracker_id = kwargs['tracker_id']\n if receiver and sender and subject and body:\n sent_email = SentEmail()\n if tracker_id:\n sent_email.tracker_id = tracker_id\n sent_email.read_status = 'Tracking'\n else:\n sent_email.read_status = 'Not Tracking'\n sent_email.receiver = receiver\n sent_email.subject = subject\n sent_email.sender = sender\n sent_email.status = 'sent'\n sent_email.user = user\n sent_email.body = body\n sent_email.save()\n return sent_email\n else:\n return False","def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\n\n if is_new:\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\n\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))","def sendEmail(householdID):\n contactID = mdb.getContact(householdID)\n sqlq = \"\"\"\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\n FROM Contact\n WHERE idContact = '{}';\n \"\"\".format(contactID)\n result = mdb.getSQL(sqlq)[0]\n\n # thisName = (\"%s %s\" % (result['Name'], result['Surname']))\n thisName = (\"%s\" % (result['Name']))\n thisAddress = (\"%s</br>%s</br>%s %s\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\n thisAddress = thisAddress.replace(\"None </br>\", \"\")\n thisEmail = (\"%s\" % (result['email']))\n thisStatus = (\"%s\" % (result['status']))\n thisAddress = thisAddress.replace(\"None</br>\", \"\")\n participantCount = (\"%s\" % mdb.getParticipantCount(str(householdID)))\n # prepare the custom email\n\n thisPath = os.path.dirname(os.path.abspath(__file__))\n if (thisStatus == 'de'):\n # DOESN'T happen yet - de is excluded from query for now\n # emailPath = os.path.join(thisPath, \"emails/email_confirm_de.html\")\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\n else:\n emailPath = os.path.join(thisPath, \"emails/email_automated_date.html\")\n\n templateFile = open(emailPath, \"r\")\n templateText = templateFile.read()\n templateFile.close()\n templateText = templateText.replace(\"[householdID]\", householdID)\n templateText = templateText.replace(\"[contactID]\", contactID)\n templateText = templateText.replace(\"[name]\", thisName)\n templateText = templateText.replace(\"[address]\", thisAddress)\n templateText = templateText.replace(\"[securityCode]\", mdb.getSecurityCode(householdID))\n templateText = templateText.replace(\"[participantCount]\", participantCount)\n\n # Subject\n subjectLine = templateText.splitlines()[0]\n templateText = templateText[templateText.find('\\n') + 1:] # find line break and return all from there - i.e. remove first line\n \n # email file\n emailFilePath = os.path.join(thisPath, \"tempEmail.htmail\")\n\n emailFile = open(emailFilePath, \"w+\")\n emailFile.write(templateText)\n emailFile.close()\n\n # call('mutt -e \"set content_type=text/html\" -s \"[TESTING]' + subjectLine + '\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\n call('mutt -e \"set content_type=text/html\" -s \"' + subjectLine + '\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)","def save(self,owner=None):\n\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\n\n t = Ticket(\n title = self.cleaned_data['title'],\n owner=owner,\n submitter_email = self.cleaned_data['submitter_email'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n account = owner.account\n )\n\n t.save()\n\n f = FollowUp(\n ticket = t,\n title = _('Ticket Opened Via Web'),\n date = datetime.now(),\n public = True,\n comment = self.cleaned_data['body'],\n account = owner.account\n )\n\n f.save()\n\n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n }\n\n messages_sent_to = []\n\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self, user):\n\n q = self.cleaned_data['queue']\n\n t = Ticket( title = self.cleaned_data['title'],\n submitter_email = self.cleaned_data['submitter_email'],\n account = self.cleaned_data['account'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n owner = self.cleaned_data['owner']\n )\n\n if HAS_TAG_SUPPORT:\n t.tags = self.cleaned_data['tags']\n\n if self.cleaned_data['assigned_to']:\n try:\n t.assigned_to = self.cleaned_data['assigned_to']\n except User.DoesNotExist:\n t.assigned_to = None\n t.save()\n\n f = FollowUp( ticket = t,\n title = _('Ticket Opened'),\n date = datetime.now(),\n public = False,\n comment = self.cleaned_data['body'],\n systemuser = user.account,\n )\n if self.cleaned_data['assigned_to']:\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\n 'name': t.get_assigned_to\n }\n\n f.save()\n \n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n 'comment': f.comment,\n }\n \n messages_sent_to = []\n\n if t.submitter_email:\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n #FIX USERSETTINGS\n #=======================================================================\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\n # send_templated_mail(\n # 'assigned_to',\n # context,\n # recipients=t.assigned_to.email,\n # sender=q.from_address,\n # fail_silently=True,\n # files=files,\n # )\n # messages_sent_to.append(t.assigned_to.email)\n #=======================================================================\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self):\n email = self.validated_data['email']\n name = self.validated_data['name']\n password = self.validated_data['password']\n\n email_query = models.Email.objects.filter(address=email)\n if email_query.exists():\n email_instance = email_query.get()\n\n # If the email is already verified, we send a duplicate\n # notification and exit.\n if email_instance.is_verified:\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"is already verified.\",\n email_instance,\n )\n email_instance.send_duplicate_notification()\n\n return\n\n # If the email is not verified, we send a new verification\n # token to the address.\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"already exists. Sending a new verification token instead.\"\n )\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()\n\n return\n\n # The email doesn't exist, so we create a new user and email,\n # then send a verification token to the email.\n user = models.User.objects.create_user(name, password)\n email_instance = models.Email.objects.create(address=email, user=user)\n\n # The user's primary email is their only email. This is the only\n # time the primary email can be unverified.\n user.primary_email = email_instance\n user.save()\n\n logger.info(\n \"Registered new user %r with email address %r\",\n user,\n email_instance,\n )\n\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()","def send_email(self, message):\n pass","def send_mail(self, subject):\r\n pass","def post():\n contactus_json = request.get_json()\n\n try:\n dict_data = ContactUsSchema().load(contactus_json)\n dict_data['description'] = escape(dict_data['description'])\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\n response, status = 'Received successfully', HTTPStatus.OK\n except ValidationError as project_err:\n response, status = {'systemErrors': project_err.messages}, \\\n HTTPStatus.BAD_REQUEST\n return response, status","async def handle_email(self, email):\n\t\tif 'logdir' in self.log_settings:\n\t\t\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\n\t\t\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\n\t\t\t\tf.write(email.email.as_bytes())\n\n\t\tawait self.log('You got mail!')","def set_email_sent(self, instance_id, creator, type):\n table = self.type_map[type]\n item = {\n \"InstanceID\": instance_id,\n \"Creator\": creator,\n \"Scheduled For Deletion\": True,\n \"EmailSent\": True\n }\n return table.put_item(item)","def order_success(request, job_id, token):\n if request.user.is_active:\n job_obj = Job.objects.get(job_id= job_id)\n if job_obj.employeer == request.user:\n if verifiers.payment_token_is_valid(token, job_obj):\n if job_obj.accepted == False:\n job_obj.status = Status.objects.create(status_name=\"PN\", comment=\"Awaiting payment confirmation from Stripe, this might take a few minutes.\")\n job_obj.accepted = True\n job_obj.save()\n \n messages.add_message(request, messages.INFO, 'Your order has been created.')\n return redirect('/') \n else:\n #someone tries to manually enter the URL with wrong payment_token\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n #the person making the request is not the same as the person who created the job\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n return redirect(\"/login\")","def order_created(order_id):\n order = Order.objects.get(id=order_id)\n subject = f'Order nr. {order.id}'\n message = f'Dear {order.first_name},\\n\\n' \\\n f'You have successfully placed an order.' \\\n f'Your order ID is {order.id}.'\n mail_sent = send_mail(subject,\n message,\n 'info@tishman.com.ng',\n [order.email])\n return mail_sent","def send_email_alert_slack(email_pk: str):\n logging.info(\"Sending alert for Email<%s>\", email_pk)\n email = Email.objects.get(pk=email_pk)\n issue = email.issue\n assert issue, f\"Email<{email_pk}> does not have an associated Issue\"\n paralegal = issue.paralegal\n alert_sent = False\n case_email_url = settings.CLERK_BASE_URL + reverse(\n \"case-email-list\", args=(str(issue.pk),)\n )\n case_url = settings.CLERK_BASE_URL + reverse(\n \"case-detail-view\", args=(str(issue.pk),)\n )\n msg = (\n \"*New Email Notification*\\n\"\n f\"A new email has been received for case <{case_url}|{issue.fileref}> .\\n\"\n f\"You can view this case's emails here: <{case_email_url}|here>.\\n\"\n )\n if paralegal and issue.is_open:\n # Send alert directly to paralegal for open issues.\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\n logging.info(\"Looking up %s in Slack\", alert_email)\n slack_user = get_slack_user_by_email(alert_email)\n if slack_user:\n logging.info(\"Notifying %s of Email<%s> via Slack\", alert_email, email_pk)\n send_slack_direct_message(msg, slack_user[\"id\"])\n alert_sent = True\n\n if not alert_sent:\n # In all other cases send to alerts channel.\n logging.info(\n \"Could not find someone to DM in Slack, sending generic alert for Email<%s>\",\n email_pk,\n )\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\n\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\n logging.info(\"Alert sent sucessfully for Email<%s>\", email_pk)","def send_sbemail(email_addy, city):\n print(\"Single email send attempt: %s, %s\" % (email_addy, city))\n given_pair, weather = subject_phrase_picker(city)\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\n print(\"Going to try a mailgun post to: %s with subject %s\" % (email_addy, given_pair['subject']))\n response = requests.post(\n MG_API_URL,\n auth=(\"api\", MG_API_KEY),\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\n 'to': [email_addy],\n 'subject': given_pair['subject'],\n 'text': mailtext\n })\n print(\"status: %s | %s and response: %s\" % (response.status_code, response.reason, response.text))\n print(\"Attempted send %s to: %s!\" % (mailtext, email_addy))\n return \"Sent %s to: %s!\" % (mailtext, email_addy)","def on_update(self):\n\t\tfor email_account in frappe.get_all(\"Email Account\", filters={\"domain\": self.name}):\n\t\t\ttry:\n\t\t\t\temail_account = frappe.get_doc(\"Email Account\", email_account.name)\n\t\t\t\tfor attr in [\"email_server\", \"use_imap\", \"use_ssl\", \"use_tls\", \"attachment_limit\", \"smtp_server\", \"smtp_port\", \"use_ssl_for_outgoing\", \"append_emails_to_sent_folder\", \"incoming_port\"]:\n\t\t\t\t\temail_account.set(attr, self.get(attr, default=0))\n\t\t\t\temail_account.save()\n\n\t\t\texcept Exception as e:\n\t\t\t\tfrappe.msgprint(_(\"Error has occurred in {0}\").format(email_account.name), raise_exception=e.__class__)","def put(self, campaign_id):\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\n if campaign.candidates.count() > 0:\n campaign.launch_task('generate_mailer_file')\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\n else:\n api.abort(404, message='Campaign does not exist', success=False)","def post_save_moderation(self, sender, comment, request, **kwargs):\n model = comment.content_type.model_class()\n if model not in self._registry:\n return\n self._registry[model].email(comment, comment.content_object, request)","def delegate_last_day():\n\n regs = Registration.objects.all()\n\n template = 'notifications/last_day_mail.html'\n\n for reg in regs:\n subject = 'SciPy.in 2011: Schedule and other details'\n message = loader.render_to_string(\n template, dictionary={'name': reg.registrant.username})\n\n reg.registrant.email_user(subject=subject, message=message,\n from_email='madhusudancs@gmail.com')","def send_email(email_dict, appointment_id):\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)","def post(self):\n # se captura y se parsea a json el body del request recibido por el\n # webhook\n request_body = json.loads(self.request.body)\n\n for body in request_body:\n \"\"\" Evaluar el tipo de evento ya que trae campos diferentes \"\"\"\n logging.info(request_body)\n\n event = str(body['event'])\n correo = str(body['email'])\n numero_folio = str(body['numero_folio'])\n tipo_dte = str(body['tipo_dte'])\n\n logging.info(event)\n\n if event and correo and numero_folio and tipo_dte:\n\n if event == 'processed':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.processed_event = event\n email_model.processed_sg_event_id = body['sg_event_id']\n email_model.processed_sg_message_id = body['sg_message_id']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.processed_event = event\n e.processed_sg_event_id = body['sg_event_id']\n e.processed_sg_message_id = body['sg_message_id']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'delivered':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.delivered_event = event\n email_model.delivered_sg_event_id = body['sg_event_id']\n email_model.delivered_sg_message_id = body['sg_message_id']\n email_model.delivered_response = body['response']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.delivered_event = event\n e.delivered_sg_event_id = body['sg_event_id']\n e.delivered_sg_message_id = body['sg_message_id']\n e.delivered_response = body['response']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'open':\n model = EmailModel()\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n if email_model.opened_first_date == None:\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_event = event\n email_model.opened_ip = body['ip']\n email_model.opened_user_agent = body['useragent']\n email_model.opened_sg_event_id = body['sg_event_id']\n email_model.opened_sg_message_id = body['sg_message_id']\n model.email_add_count(email_model)\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n if e.opened_first_date == None:\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_event = event\n e.opened_ip = body['ip']\n e.opened_user_agent = body['useragent']\n e.opened_sg_event_id = body['sg_event_id']\n e.opened_sg_message_id = body['sg_message_id']\n e.email_add_count(e)\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'dropped':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.dropped_sg_event_id = body['sg_event_id']\n email_model.dropped_sg_message_id = body['sg_message_id']\n email_model.dropped_reason = body['reason']\n email_model.dropped_event = event\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.dropped_sg_event_id = body['sg_event_id']\n e.dropped_sg_message_id = body['sg_message_id']\n e.dropped_reason = body['reason']\n e.dropped_event = event\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'bounce':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.bounce_event = event\n email_model.bounce_sg_event_id = body['sg_event_id']\n email_model.bounce_sg_message_id = body['sg_message_id']\n email_model.bounce_reason = body['reason']\n email_model.bounce_status = body['status']\n email_model.bounce_type = body['type']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.bounce_event = event\n e.bounce_sg_event_id = body['sg_event_id']\n e.bounce_sg_message_id = body['sg_message_id']\n e.bounce_reason = str(body['reason']).decode(\"utf-8\")\n e.bounce_status = body['status']\n e.bounce_type = body['type']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'unsubscribe':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.unsubscribe_uid = body['uid']\n email_model.unsubscribe_purchase = body['purchase']\n email_model.unsubscribe_id = body['id']\n email_model.unsubscribe_event = body['event']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.unsubscribe_uid = body['uid']\n e.unsubscribe_purchase = body['purchase']\n e.unsubscribe_id = body['id']\n e.unsubscribe_event = body['event']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n else:\n logging.info('body con campos vacios')","def send_mail(self, address, title, message):\n pass","def mail_sent():\n\n url = settings.SITE_URL + '\\charts'\n subject = 'Анализ запрошенного ресурса'\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\n mail_sent = send_mail(subject,\n message,\n 'admin@myshop.com',\n ['user@mail.ru,'])\n print(message)\n return mail_sent","def emailNote(self, authenticationToken, parameters):\r\n pass","def send_new_email(user):\n token = user.get_token()\n message = Message(\n 'Verify Your New Email',\n sender='storcwebsite@gmail.com',\n recipients=[user.temp_email])\n message.body = f\"The email address associated with your Storc \" \\\n f\"account has changed.\\n\\nTo verify your new email address, \" \\\n f\"please click the link below:\\n\\n\" \\\n f\"{url_for('users.new_email', token=token, _external=True)}\"\n mail.send(message)","def report_save(message, bot):\n text = message.text\n reports = parse_report(text)\n author_id = message.from_user.id\n author_username = message.from_user.username\n to_save = []\n for report in reports:\n obj_to_save = {\n 'author': author_id,\n 'author_name': author_username,\n 'text': report,\n }\n to_save.append(obj_to_save)\n try:\n request = requests.post(\n url='http://127.0.0.1:8000/api/v1/report/',\n headers={\n 'Authorization': f'Bearer {jwt_token}',\n },\n json=to_save,\n )\n bot.send_message(\n message.from_user.id,\n request.status_code,\n )\n except requests.exceptions.ConnectionError:\n bot.send_message(\n message.from_user.id,\n 'Ошибка соединения с сервером',\n )","def perform_create(self, serializer):\n user = serializer.save()\n signals.user_registered.send(\n sender=self.__class__, user=user, request=self.request\n )\n\n context = get_email_context(user)\n to = [get_user_email(user)]\n if djconf.SEND_ACTIVATION_EMAIL:\n djconf.EMAIL.activation(self.request, context).send(to)\n elif djconf.SEND_CONFIRMATION_EMAIL:\n djconf.EMAIL.confirmation(self.request, context).send(to)","def send(self):\n ReferralActivity.objects.create(\n actor=self.user,\n verb=ReferralActivityVerb.CREATED,\n referral=self,\n )\n # Confirm the referral has been sent to the requester by email\n Mailer.send_referral_saved(self)\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\n # email notifications)\n for unit in self.units.all():\n contacts = unit.members.filter(\n unitmembership__role=UnitMembershipRole.OWNER\n )\n for contact in contacts:\n Mailer.send_referral_received(self, contact=contact, unit=unit)","def execute_automatic_email(self, request, pk=None):\n try:\n retreat = Retreat.objects.get(pk=pk)\n except Exception:\n response_data = {\n 'detail': \"Retreat not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n try:\n email = AutomaticEmail.objects.get(\n id=int(request.GET.get('email'))\n )\n except Exception:\n response_data = {\n 'detail': \"AutomaticEmail not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n # Notify a user for every reserved seat\n emails = []\n for reservation in retreat.reservations.filter(is_active=True):\n if reservation.automatic_email_logs.filter(email=email):\n pass\n else:\n send_automatic_email(reservation.user, retreat, email)\n AutomaticEmailLog.objects.create(\n reservation=reservation,\n email=email\n )\n emails.append(reservation.user.email)\n\n response_data = {\n 'stop': True,\n 'emails': emails\n }\n return Response(response_data, status=status.HTTP_200_OK)","def save(self):\n owner = self.context[\"request\"].user\n recipient = self._recipient_email_inst.user\n\n logger.info(\n \"Transferring subscription %r from user %r to user %r\",\n owner.know_me_subscription,\n owner.pk,\n recipient.pk,\n )\n\n with transaction.atomic():\n models.Subscription.objects.filter(user=recipient).delete()\n models.Subscription.objects.filter(user=owner).update(\n user=recipient\n )","def send_message(self):\r\n return \"success\"","def alert_for_pending_mails_1(request):\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tThread(target=alert_for_pending_mails_1_worker).start()\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tresponse = {}\n\n\tresponse[\"info_to_contact\"] = \"Ok\"\n\n\treturn response","def save(self, request, item, linked_item, linked_item_form):\n\t\titem.description = linked_item_form.cleaned_data['subject']\n\t\tlinked_item.text = linked_item_form.cleaned_data['text']\n\t\tlinked_item.delivery_notes = linked_item_form.cleaned_data['delivery_notes']\n\t\titem.priority = Priority.objects.get(id = linked_item_form.cleaned_data['priority'])\n\t\t#\n\t\t# Look for actions other than save\n\t\t#\n\t\tif request.POST.get('update', '') == 'Completed':\n\t\t\titem.fixed = True\n\t\tif request.POST.get('update', '') == 'Failed':\n\t\t\titem.fixed = False\n\t\t\titem.validated = True\n\t\tif request.POST.get('update', '') == 'Verified':\n\t\t\titem.fixed = True\n\t\t\titem.validated = True\n\n\t\titem.save()\n\t\tlinked_item.save()\n\n\t\t#\n\t\t# Check for comments and add\n\t\t#\n\t\tcomment_text = linked_item_form.cleaned_data['comments'].strip()\n\t\tif comment_text != '':\n\t\t\tpost_comment(request, item, comment_text)"],"string":"[\n \"def save(self, fail_silently=False):\\n send_mail(fail_silently=fail_silently, **self.get_message_dict())\",\n \"def send_created_email(self):\\n if settings.NOTIFY_NEW_REG:\\n to = settings.NOTIFY_NEW_REG\\n message = \\\"\\\"\\\"\\\\\\nGreetings,<br><br>\\n\\nA new vehicle registration has been submitted by %s.<br><br>\\n\\nGo here to view or edit the request: <br>\\n<a href=\\\"%s\\\">%s</a>\\n<br><br>\\nSincerely,<br><br>\\nThe Janelia Parking Permit Program\\n \\\"\\\"\\\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\\n subject = 'A new parking permit request has been entered'\\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\\n text_content = re.sub(r'<[^>]+>','',message)\\n html_content = message\\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\\n\\n logger.info(\\\"in send lead mail task\\\")\\n return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)\",\n \"def save(self, commit=False):\\n mail_result = self.send_email()\\n if mail_result:\\n self.instance.is_admin_notified = True\\n\\n contact = super().save(commit=commit)\\n\\n return contact\",\n \"def save(self, **kwargs):\\n\\t\\tif (self.answer != \\\"\\\"):\\n\\t\\t\\tself.answered = True\\n\\t\\t\\tsuper().save(**kwargs)\\n\\t\\t\\tmail_subject = \\\"Your Question has been answered.\\\"\\n\\t\\t\\t# Do not break the following string or the email will get cut off\\n\\t\\t\\tmessage = f\\\"Hi {self.user.first_name},\\\\n\\\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\\\n\\\\n\\\"\\n\\n\\t\\t\\temail = EmailMessage(\\n\\t\\t\\t\\tmail_subject, message, to=[self.user.email]\\n\\t\\t\\t)\\n\\t\\t\\temail.send()\\n\\t\\telse:\\n\\t\\t\\tsuper().save(**kwargs)\",\n \"def save(self): # noqa\\n self.bound_admin.message_success(self.bound_request, _('Done.'))\",\n \"def action_my_payslip_sent(self):\\n self.ensure_one()\\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\\n if template:\\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\\n self.flag = True\",\n \"def test_email_sent_on_failure(self):\\n self._authorize()\\n data = {\\n 'Subject_Number': '000-1111',\\n 'Pin_Code': '1234',\\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\\n 'Mobile_Number': '2223334444',\\n }\\n patient = self.create_xml_patient(data)\\n payload = self.create_xml_payload([patient])\\n response = self._post(payload)\\n self.assertEqual(response.status_code, 500)\\n self.assertEqual(len(mail.outbox), 1)\",\n \"def save_model(self, request, obj, form, change):\\n if change and obj.status == 'V':\\n if Organization.objects.get(pk=obj.pk).status == 'P':\\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\\n try:\\n sender = Plugin_Contact.objects.all()[0].email\\n except IndexError:\\n sender = None\\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\\n send_mail(u'Validation de votre fiche sur la PASR',\\n u'Bonjour,\\\\n\\\\nVotre fiche vient d\\\\'être validée sur la PASR.',\\n sender, dests)\\n super(OrganizationAdmin, self).save_model(request, obj, form, change)\",\n \"def send_referral_saved(cls, referral, created_by):\\n\\n template_id = settings.SENDINBLUE[\\\"REFERRAL_SAVED_TEMPLATE_ID\\\"]\\n\\n data = {\\n \\\"params\\\": {\\\"case_number\\\": referral.id},\\n \\\"replyTo\\\": cls.reply_to,\\n \\\"templateId\\\": template_id,\\n \\\"to\\\": [{\\\"email\\\": created_by.email}],\\n }\\n\\n cls.send(data)\",\n \"def callback_success_message(request):\\n msg = 'Sucessfully recorded form :)'\\n logger.info(msg)\\n messages.info(request._request, msg)\",\n \"def save(self):\\n super(Notification, self).save()\\n # get user recipent\\n us = self.notified_user\\n # check that user has a valid email address\\n if us.email.find('@') > 0 and us.email.find('.') > 0:\\n # mandatory fields\\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\\n to = us.email\\n from_email = settings.DEFAULT_FROM_EMAIL\\n # get text version of the message\\n text_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n # FIXME: HTML version implementation pending\\n html_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n msg = EmailMultiAlternatives(\\n subject, \\n text_content,\\n from_email,\\n [to]\\n )\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\\n\\n logger.info(\\\"in send lead mail task\\\")\\n return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)\",\n \"def mail(request):\\n email_admin.delay('testinggg')\\n return JsonResponse({\\\"details\\\":\\\"working\\\"})\",\n \"def on_post(self):\\n return \\\"Ok, the stuff is being saved\\\"\",\n \"def send_email(request):\\n # send emails and return some manner of success response\\n send(**request.params)\\n return {'success': 'mail sent!'}\",\n \"def unsuccessful(self, **post):\\n cr, uid, context = request.cr, request.uid, request.context\\n attchment_list = []\\n mail_obj = request.env['mail.mail'].sudo()\\n\\n sale_order_id = request.session.get('sale_order_id')\\n\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order id===========\\\", sale_order_id)\\n if sale_order_id:\\n order = request.env['sale.order'].sudo().browse(sale_order_id)\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order order===========\\\", order)\\n else:\\n return request.redirect('/enrolment_book')\\n request.website.sale_reset()\\n\\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\\n raise_if_not_found=False)\\n if template_id:\\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\\n order.id)\\n enroll_file_name = \\\"Pro-Forma \\\" + order.name\\n if pdf_data_enroll:\\n pdfvals = {'name': enroll_file_name,\\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas_fname': enroll_file_name + '.pdf',\\n 'res_model': 'sale.order',\\n 'type': 'binary'}\\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\\n attchment_list.append(pdf_create)\\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\\n if agreement_id:\\n attchment_list.append(agreement_id)\\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\\n if banking_detail_id:\\n attchment_list.append(banking_detail_id)\\n\\n # mail_values = {\\n # 'email_from': template_id.email_from,\\n # 'reply_to': template_id.reply_to,\\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # 'subject': \\\"Charterquest FreeQuote/Enrolment \\\" + order.name,\\n # 'notification': True,\\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\\n # 'auto_delete': False,\\n # }\\n # msg_id = mail_obj.create(mail_values)\\n # msg_id.send()\\n\\n template_id.sudo().write(\\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\\n template_id.sudo().with_context(\\n # email_to=each_request.get('email'),\\n # event_list=event_list,\\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # prof_body=invoice_id.prof_body.name,\\n ).send_mail(order.id, force_send=True)\\n return request.render(\\\"cfo_snr_jnr.event_unsuccessful\\\", {'order': order})\",\n \"def claim_email(request):\\n email = request.POST.get('email', '')\\n email_user = User.objects.filter(email=email)\\n payload = {\\n 'res': 'failed'\\n }\\n if email_user.exists() and \\\\\\n not email_user[0].profile.send_mail:\\n request.user.profile.add_email(email)\\n payload['res'] = 'success'\\n\\n return payload\",\n \"def test_admin_approval_complete_email(self):\\n new_user = UserModel().objects.create_user(**self.user_info)\\n profile = self.registration_profile.objects.create_profile(new_user)\\n profile.send_admin_approve_complete_email(Site.objects.get_current())\\n self.assertEqual(len(mail.outbox), 1)\\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])\",\n \"def add_manualpost_email(request, submission_id=None, access_token=None):\\n\\n if request.method == 'POST':\\n try:\\n button_text = request.POST.get('submit', '')\\n if button_text == 'Cancel':\\n return redirect(\\\"submit/manual_post.html\\\")\\n \\n form = SubmissionEmailForm(request.POST)\\n if form.is_valid():\\n submission_pk = form.cleaned_data['submission_pk']\\n message = form.cleaned_data['message']\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n \\n if form.cleaned_data['direction'] == 'incoming':\\n msgtype = 'msgin'\\n else:\\n msgtype = 'msgout'\\n \\n submission, submission_email_event = (\\n add_submission_email(request=request,\\n remote_ip=request.META.get('REMOTE_ADDR', None),\\n name = form.draft_name,\\n rev=form.revision,\\n submission_pk = submission_pk,\\n message = message,\\n by = request.user.person,\\n msgtype = msgtype) )\\n \\n messages.success(request, 'Email added.')\\n \\n try:\\n draft = Document.objects.get(name=submission.name)\\n except Document.DoesNotExist:\\n # Assume this is revision 00 - we'll do this later\\n draft = None\\n \\n if (draft != None):\\n e = AddedMessageEvent(type=\\\"added_message\\\", doc=draft)\\n e.message = submission_email_event.submissionemailevent.message\\n e.msgtype = submission_email_event.submissionemailevent.msgtype\\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\\n e.by = request.user.person\\n e.desc = submission_email_event.desc\\n e.time = submission_email_event.time\\n e.save()\\n \\n return redirect(\\\"ietf.submit.views.manualpost\\\")\\n except ValidationError as e:\\n form = SubmissionEmailForm(request.POST)\\n form._errors = {}\\n form._errors[\\\"__all__\\\"] = form.error_class([\\\"There was a failure uploading your message. (%s)\\\" % e.message])\\n else:\\n initial = {\\n }\\n\\n if (submission_id != None):\\n submission = get_submission_or_404(submission_id, access_token)\\n initial['name'] = \\\"{}-{}\\\".format(submission.name, submission.rev)\\n initial['direction'] = 'incoming'\\n initial['submission_pk'] = submission.pk\\n else:\\n initial['direction'] = 'incoming'\\n \\n form = SubmissionEmailForm(initial=initial)\\n\\n return render(request, 'submit/add_submit_email.html',dict(form=form))\",\n \"def receive_email_view(request):\\n save_inbound_email(request.POST, request.FILES)\\n return HttpResponse(200)\",\n \"def validation_email_sent(request):\\n assert(settings.EMAIL_VALIDATION == True)\\n logging.debug('')\\n data = {\\n 'email': request.user.email,\\n 'change_email_url': reverse('user_changeemail'),\\n 'action_type': 'validate'\\n }\\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))\",\n \"def postprocess():\\n if ERRORS:\\n address = 'tamakoshihiroki@gmail.com'\\n body = '\\\\n\\\\n'.join( ERRORS )\\n msg = create_message( body, address )\\n send_mail( msg, address )\",\n \"def send_mail_when_failed(self, body):\\r\\n pass\",\n \"def test_send_email_on_invite(self):\\n\\n league = self.create_league()\\n\\n season = self.create_season(league)\\n team = self.create_team(season)\\n\\n player = self.create_player()\\n\\n send_user_email_on_join(player, team.id)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n\\n # if testing manually:\\n # import pathlib\\n # pathlib.Path(\\\"test_email.html\\\").write_text(last_sent.body)\",\n \"def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\\n lead = Lead.objects.get(id=lead_id)\\n created_by = lead.created_by\\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\\n for user in recipients:\\n recipients_list = []\\n user = User.objects.filter(id=user, is_active=True).first()\\n if user:\\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\\n recipients_list.append(user.email)\\n context = {}\\n context[\\\"url\\\"] = protocol + '://' + domain + \\\\\\n reverse('leads:view_lead', args=(lead.id,))\\n context[\\\"user\\\"] = user\\n context[\\\"lead\\\"] = lead\\n context[\\\"created_by\\\"] = created_by\\n context[\\\"source\\\"] = source\\n subject = 'Assigned a lead for you. '\\n html_content = render_to_string(\\n 'assigned_to/leads_assigned.html', context=context)\\n msg = EmailMessage(\\n subject,\\n html_content,\\n to=recipients_list\\n )\\n msg.content_subtype = \\\"html\\\"\\n msg.send()\",\n \"def form_valid(self, form):\\n self.object = form.save()\\n self.send_verify_email()\\n return super().form_valid(form)\",\n \"def save_response(self):\\r\\n self.q(css='input.save-button').first.click()\\r\\n EmptyPromise(\\r\\n lambda: 'save' in self.alert_message.lower(),\\r\\n \\\"Status message saved\\\"\\r\\n ).fulfill()\",\n \"def form_valid(self, form):\\n\\n ModelAccountVerification.objects.send_verification_email(form.user,\\n verification_type=\\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\\n\\n messages.info(self.request, \\\"Email sent successfully to the {0}\\\".format(form.user.email))\\n\\n return HttpResponseRedirect(self.success_url)\",\n \"def execute(self):\\n return LOGGER.info(f\\\"{datetime.datetime.now()} - Sending EMail to the configured email list\\\")\",\n \"def test_post_external_attendance_successfull(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-03T19:00',\\n 'attendee_email': 'student@external.mx',\\n 'exit_datetime': '2021-08-03T20:00'}\\n response = self.client.post(\\n reverse('tracking:external-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def save_model(self, request, obj, form, change):\\n notify.send(\\n request.user,\\n recipient=obj.reported_by,\\n verb='updated your report',\\n action_object=obj,\\n description=obj.notes,\\n target=obj\\n )\\n obj.save()\",\n \"def post(self):\\n return send_email(request.args)\",\n \"def send_warning(self):\\n\\n # Check whether all the necessary parameters for SMS are present\\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\\n client = Client(self.account_sid, self.auth_token)\\n\\n try:\\n sms = client.messages.create(\\n body=\\\"\\\"\\\"Last will: It was at least 30 days since your last check in. \\n This is a reminder to check in in the next 24 hours.\\\"\\\"\\\",\\n from_=self.twilio_phone,\\n to=self.your_phone)\\n sms\\n print(\\\"\\\\nSMS sent\\\")\\n except Exception as e:\\n print(f\\\"An error occurred while trying to send the SMS. Error: {e}\\\")\\n\\n else:\\n print(\\\"\\\\nMissing SMS parameters. SMS not sent\\\")\\n\\n # Check whether all the necessary parameters for email are present\\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\\n message = f\\\"\\\"\\\"It has been at least 30 days since you last checked in. \\nYou need to check in in the next 24 hours.\\\\n\\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\\\n\\nIn order to reset simply go to the working directory and run python3 last_will.py\\\"\\\"\\\"\\n\\n # send_email will return 0 if everything went ok, otherwise it will return an error message\\n status = send_email(self.sender_name, self.your_email,\\n self.email, self.email_pwd,\\n subject='Last will: Reminder to check in', unencrypted_message=message)\\n\\n if status != 0:\\n print(status)\\n exit(1)\\n else:\\n print(\\\"Email sent\\\\n\\\")\\n\\n print(f\\\"You have until {self.deadline} to check in. \\\"\\n f\\\"In order to do that simply go to the working directory and run ./last_will.sh\\\\n\\\")\\n else:\\n print(\\\"Missing email parameters. Email not sent.\\\\n\\\")\\n exit(1)\",\n \"def send_reminder(self):\\n pass\",\n \"def complete_allotment(request):\\n reg_no = request.user.username\\n user = get_object_or_404(User, username=reg_no)\\n sec_email = user.get_profile().secondary_email\\n options_chosen = get_chosen_options(user)\\n context = {'username': reg_no, 'email': sec_email, \\n 'options_chosen': options_chosen}\\n ##Sending mail with allotment details \\n admin = User.objects.get(pk=1) \\n from_email = admin.email\\n subject = \\\"JAM 2012 admissions\\\"\\n content = \\\"The following options were chosen by you \\\\n \\\\n\\\"\\n if options_chosen:\\n counter = 1\\n for option in options_chosen:\\n content += \\\"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\\\n\\\" %(counter, option.opt_code, option.opt_name, option.opt_location)\\n counter += 1 \\n \\n content += \\\"\\\\n \\\\n \\\\nPlease do not delete this email and keep it for reference purposes. \\\\n \\\\n \\\\n \\\\n Regards, \\\\n JAM Office, IIT Bombay\\\"\\n send_mail(subject, content, from_email, [sec_email])\\n admin_content = content\\n admin_content +=\\\"\\\\n\\\\n\\\\n#%s:\\\" % (reg_no)\\n counter = 1\\n for option in options_chosen:\\n admin_content += \\\"%s,%s:\\\" %(counter, option.opt_code) \\n counter += 1\\n admin_content +=\\\"#\\\"\\n admin_content += time.ctime() \\n mail_admins(subject, admin_content) \\n return render(request, 'allotter/complete.html', context)\",\n \"def create_response_and_send_mail(token, partner_id, email):\\n \\t #set url\\n\\t url = survey.public_url\\n \\t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\\n\\t\\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\\n\\t\\t template = self.pool.get('email.template').browse(cr,uid,template_id)\\n\\t\\t body = template.body_html\\n\\t if token:\\n \\t url = url + '/' + token\\n\\n\\t # post the message\\n \\t values = {\\n \\t'model': None,\\n\\t 'res_id': None,\\n \\t 'subject': template.subject,\\n \\t'body': body.replace(\\\"__URL__\\\", url),\\n\\t 'body_html': body.replace(\\\"__URL__\\\", url),\\n \\t 'parent_id': None,\\n \\t'partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'email_from': 'info@sba.org' or None,\\n \\t 'email_to': partner.email,\\n\\t\\t }\\n\\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\\n \\t mail_mail_obj.send(cr, uid, [mail_id], context=context)\",\n \"def send_reminder(self):\\n message_contents = \\\"This is a reminder that your event: \\\" + self.event_title + \\\" takes place on \\\" + self.event_date + \\\" in \\\" + self.event_location\\n subject = \\\"Event Reminder\\\"\\n attendees = self.gameplanuser_set.all()\\n for attendee in attendees:\\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\\n remindermessage.save()\",\n \"def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\\n\\n location = reverse(\\\"KursAnmeldung-verify_email\\\", kwargs={\\\"hash\\\":rnd_hash})\\n verify_link = request.build_absolute_uri(location)\\n\\n # FIXME: convert to users local time.\\n now = datetime.datetime.utcnow()\\n\\n email_context = {\\n \\\"verify_link\\\": verify_link,\\n \\\"db_entry\\\": db_entry,\\n \\\"now\\\": now,\\n }\\n\\n # Render the internal page\\n emailtext = render_to_string(\\\"kurs_anmeldung/verify_mailtext.txt\\\", email_context)\\n\\n # Get the preferences from the database:\\n raw_notify_list = preferences[\\\"notify\\\"]\\n notify_list = raw_notify_list.splitlines()\\n notify_list = [i.strip() for i in notify_list if i]\\n\\n email_kwargs = {\\n \\\"from_email\\\": preferences[\\\"from_email\\\"],\\n \\\"subject\\\": preferences[\\\"email_subject\\\"],\\n \\\"body\\\": emailtext,\\n \\\"to\\\": [db_entry.email],\\n \\\"bcc\\\": notify_list,\\n }\\n\\n if MAIL_DEBUG == True:\\n msg = u\\\"MAIL_DEBUG is on: No Email was sended!\\\"\\n request.page_msg(msg)\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n\\n request.page_msg(\\\"django.core.mail.EmailMessage kwargs:\\\")\\n request.page_msg(email_kwargs)\\n\\n request.page_msg(\\\"debug mail text:\\\")\\n request.page_msg(mark_safe(\\\"<pre>%s</pre>\\\" % emailtext))\\n return\\n\\n # We can't use django.core.mail.send_mail, because all members\\n # of the recipient list will see the others in the 'To' field.\\n # But we would like to notify the admins via 'Bcc' field.\\n\\n connection = SMTPConnection(fail_silently=False)\\n email = EmailMessage(**email_kwargs)\\n\\n try:\\n sended = email.send(fail_silently=False)\\n except Exception, err:\\n msg = \\\"Error sending mail: %s\\\" % err\\n LogEntry.objects.log_action(app_label=\\\"kurs_anmeldung\\\", action=\\\"error\\\",\\n message=msg\\n )\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n if settings.DEBUG or request.user.is_staff:\\n db_entry.save()\\n raise\\n else:\\n db_entry.mail_sended = sended\\n db_entry.log(request, \\\"mail sended: %s\\\" % sended)\",\n \"def confirm_email(self):\\n self.active = True\\n self.save()\",\n \"def save(self, register=False):\\n sha = sha1(self.email).hexdigest()\\n infos = self.to_dict()\\n infos[\\\"plan\\\"] = infos[\\\"plan\\\"][\\\"id\\\"] if infos[\\\"plan\\\"] else None\\n\\n if not redis.hmset(\\\"sl:account:{}\\\".format(sha), infos):\\n raise SleekException(\\\"Could not save current user.\\\", 401)\\n \\n if register:\\n try:\\n send_email.delay(\\n \\\"florent.esp@gmail.com\\\", \\\"Welcome to sleekstatus !\\\",\\n \\\"Welcome message\\\", [self.email]\\n )\\n except:\\n pass # Cannot send email\",\n \"def task_send_reminder_email():\\n send_reminder_email()\\n logger.info(\\\"Sent reminder email\\\")\",\n \"def account_activation_sent(request):\\r\\n\\treturn render(request, 'account_activation_sent.html')\",\n \"def save(self, *args, **kwargs):\\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\\n allUsers = Bruker.get_all_dict(Bruker)\\n mottaker = allUsers[int(data['mottaker'])]\\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\\n melding.save()\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n reminders.Patient.objects.filter(\\n pk__in=[self.test_patient.pk, self.other_patient.pk]\\n ).update(next_visit=appt_date)\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n self.startRouter()\\n self.router.logger.setLevel(logging.DEBUG)\\n\\n # run email job\\n from afrims.apps.reminders.app import daily_email_callback\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\\n self.stopRouter()\",\n \"def save_appointment_details(request, calendar_id):\\n def schedule_mail(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n def schedule_sms(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n start_time = request.GET['start_time'][:19]\\n end_time = request.GET['end_time'][:19]\\n \\n start_time = datetime.strptime(start_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n end_time=datetime.strptime(end_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n \\n calendar_obj = Calendar.objects.get(pk=calendar_id)\\n # if this is a POST request we need to process the form data\\n if request.method == 'POST':\\n\\n # create a form instance and populate it with data from the request:\\n form = AppointmentForm(request.POST)\\n\\n # check whether it's valid and save it\\n if form.is_valid():\\n # Save appointment details\\n \\n mobilephone = form.data['mobilephone']\\n email = form.data['email']\\n first_name = form.data['first_name']\\n last_name = form.data['last_name']\\n notes = form.data['notes']\\n\\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \\n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\\n \\n appointment.calendar = calendar_obj\\n appointment.save()\\n\\n try:\\n send_appointment_mail(appointment) # send appointment details email\\n except Exception as exp:\\n print(exp)\\n \\n try:\\n send_appointment_sms(appointment) # send appointment details sms\\n except Exception as exp:\\n print(exp)\\n \\n # Calculate reminder schedule dates\\n reminder1 = start_time - timedelta(hours=2)\\n reminder2 = start_time - timedelta(hours=24)\\n reminder3 = start_time - timedelta(days=7)\\n\\n # Schedule mails\\n schedule_mail(reminder1, appointment)\\n schedule_mail(reminder2, appointment)\\n schedule_mail(reminder3, appointment)\\n \\n # Schedule sms\\n schedule_sms(reminder1, appointment)\\n schedule_sms(reminder2, appointment)\\n schedule_sms(reminder3, appointment)\\n \\n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\\n \\n # if a GET (or any other method) we'll create a blank form\\n else:\\n form = AppointmentForm()\\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\\n 'office_location': calendar_obj.office_location})\",\n \"def post(\\n self,\\n email,\\n company_name,\\n location,\\n job_profile,\\n salary,\\n username,\\n password,\\n security_question,\\n security_answer,\\n notes,\\n date_applied,\\n status,\\n):\",\n \"def post_save_order_receiver(sender, instance, created, *args, **kwargs):\\n sender_email = instance.cargo.sender.email\\n recepient_email = instance.cargo.recepient.email\\n\\n booking_agent = instance.cargo.booking_station.branch_agent.email\\n\\n if created:\\n instance._set_order_price()\\n instance._set_time_approximations()\\n price = instance.price\\n\\n subject = \\\"Order Finalized and ready to go.\\\"\\n message = f\\\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\\\"\\n\\n send_async_email(\\n subject=subject,\\n message=message,\\n sender=booking_agent,\\n recepients=[sender_email, recepient_email],\\n )\",\n \"def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\\n\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n parts = pyzmail.parse.get_mail_parts(message)\\n body=''\\n for part in parts:\\n if part.is_body == 'text/plain' and part.disposition == None:\\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\\n body = body + payload + '\\\\n'\\n\\n msg = submit_message_from_message(message, body, by)\\n\\n if (submission_pk != None):\\n # Must exist - we're adding a message to an existing submission\\n submission = Submission.objects.get(pk=submission_pk)\\n else:\\n # Must not exist\\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\\n if submissions.count() > 0:\\n raise ValidationError(\\\"Submission {} already exists\\\".format(name))\\n \\n # create Submission using the name\\n try:\\n submission = Submission.objects.create(\\n state_id=\\\"waiting-for-draft\\\",\\n remote_ip=remote_ip,\\n name=name,\\n rev=rev,\\n title=name,\\n note=\\\"\\\",\\n submission_date=datetime.date.today(),\\n replaces=\\\"\\\",\\n )\\n from ietf.submit.utils import create_submission_event, docevent_from_submission\\n desc = \\\"Submission created for rev {} in response to email\\\".format(rev)\\n create_submission_event(request, \\n submission,\\n desc)\\n docevent_from_submission(request,\\n submission,\\n desc)\\n except Exception as e:\\n log(\\\"Exception: %s\\\\n\\\" % e)\\n raise\\n\\n if msgtype == 'msgin':\\n rs = \\\"Received\\\"\\n else:\\n rs = \\\"Sent\\\"\\n\\n desc = \\\"{} message - manual post - {}-{}\\\".format(rs, name, rev)\\n submission_email_event = SubmissionEmailEvent.objects.create(\\n desc = desc,\\n submission = submission,\\n msgtype = msgtype,\\n by = by,\\n message = msg)\\n #in_reply_to = in_reply_to\\n\\n save_submission_email_attachments(submission_email_event, parts)\\n return submission, submission_email_event\",\n \"def handle_sent(self, instance):\\n if not instance.created_by:\\n return\\n\\n activity = Activity(\\n actor=instance.created_by,\\n verb=RestrictedMailSent,\\n object=instance,\\n time=instance.used,\\n extra_context={},\\n )\\n self.manager.add_activity(\\n activity, [instance.created_by.pk], [NotificationFeed]\\n )\\n\\n # Send notification\\n notification = RestrictedMailSentNotification(instance.created_by)\\n notification.notify()\",\n \"def handle_failure(self, instance, sender=None, reason=None):\\n send_email.delay(\\n to_email=sender,\\n context={\\\"reason\\\": reason},\\n subject=\\\"Kunne ikke sende ut begrenset epost\\\",\\n plain_template=\\\"restricted/email/process_failure.txt\\\",\\n html_template=\\\"restricted/email/process_failure.html\\\",\\n )\",\n \"def handle(self, *args, **options):\\n\\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\\n if candidate.contact_address and candidate.participating]\\n\\n\\n print 'sending e-mails'\\n conn = get_connection()\\n for c in candidates_with_email:\\n if c.should_send_reminder():\\n\\n print 'emailing', c\\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\\n c.last_reminder_sent = timezone.now()\\n c.save()\\n msg = make_email(c)\\n conn.send_messages([msg])\\n conn.close()\",\n \"def send(self, request, id, tribe_slug):\\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\\n document = self.get_document(id, tribe_slug)\\n \\n form = self._set_user_email_address(request)\\n email = self._get_user_email_address(request)\\n if form or not email:\\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\\n \\n # NB: Temporarily disabling actual email sending for development\\n #email_document(document, to=[email], subject='Document: %s' % document.title)\\n print \\\"Sending email to %s\\\" % email \\n \\n # Send a signal to let everyone know about this document interaction\\n document_interaction.send(sender=self, document=document, mode=\\\"sent\\\", request=request, recipient=email)\\n \\n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))\",\n \"def submit(self):\\r\\n try:\\r\\n response = self.send_request()\\r\\n if response.ok:\\r\\n self.submitted_at = datetime.now(pytz.UTC)\\r\\n self.status = \\\"submitted\\\"\\r\\n self.save()\\r\\n else:\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.error_msg = response.text\\r\\n self.save()\\r\\n except Exception as error:\\r\\n log.exception(error)\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.save()\",\n \"def _send(self, email_message): \\n\\t\\tif not email_message.recipients(): \\n\\t\\t\\treturn False \\n\\t\\trecipients = map(self._sanitize, email_message.recipients()) \\n\\t\\tMsg = o.CreateItem(0)\\n \\t\\tMsg.To = recipients\\n\\t\\tMsg.Subject = 'subject'\\n\\t\\tMsg.Body = 'text'\\n\\t\\tself.connection.SaveChanges(0)\\n\\n\\t\\tMsg.Send()\\n\\t\\treturn True\",\n \"def set_email():\\n bid_fields = {'assertion':request.form['bid_assertion'],\\n 'audience':settings.DOMAIN}\\n headers = {'Content-type':'application/x-www-form-urlencoded'}\\n h.disable_ssl_certificate_validation=True\\n resp, content = h.request('https://browserid.org/verify',\\n 'POST',\\n body=urlencode(bid_fields),\\n headers=headers)\\n bid_data = json.loads(content)\\n if bid_data['status'] == 'okay' and bid_data['email']:\\n user = f.userFromBidOrNew(bid_data['email'])\\n session['user_id'] = user.id\\n session['user_email'] = user.email\\n\\n return redirect(url_for('main'))\",\n \"def wasSent(self):\\n\\t\\tself.sent = True\\n\\t\\tself.save()\",\n \"def action_send_email(self):\\n self.ensure_one()\\n if self.communication_channel != \\\"email\\\":\\n return False # Maybe we should raise an error here.\\n\\n if self.state in (\\\"sent\\\", \\\"done\\\"):\\n raise ValidationError(_(\\\"This communication is already sent.\\\"))\\n\\n lines_2be_processed = self.credit_control_line_ids.filtered(\\n lambda line: line.state != \\\"sent\\\"\\n )\\n\\n if not lines_2be_processed:\\n raise ValidationError(_(\\\"There is no draft lines to send.\\\"))\\n\\n self = self.with_context(lang=self.partner_id.lang)\\n\\n partner = self.partner_id\\n mail_template = self.policy_level_id.email_template_id\\n # Send the email\\n partner.with_context(credit_control_mail=True).message_post_with_template(\\n template_id=mail_template.id,\\n model=self._name,\\n res_id=self.id,\\n )\\n # Set the state of the credit control lines to \\\"queued\\\"\\n lines_2be_processed.write({\\\"state\\\": \\\"sent\\\"})\\n self.state = \\\"sent\\\"\",\n \"def test_activate_form(self, mock_sendmail):\\r\\n res = self.testapp.post('/api/v1/suspend',\\r\\n params={'email': u'testing@dummy.com'},\\r\\n status=200)\\r\\n\\r\\n success = json.loads(res.body)\\r\\n self.assertTrue(\\r\\n 'message' in success,\\r\\n \\\"Should be successful with admin email address: \\\" + str(res))\\r\\n self.assertTrue(mock_sendmail.called)\",\n \"def test_post_internal_attendance_success(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-04T10:00',\\n 'attendee_email': 'student@anahuac.mx',\\n 'exit_datetime': '2021-08-04T11:30'}\\n response = self.client.post(\\n reverse('tracking:internal-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def save(self, *args, **kwargs):\\n self.expire_date = self.sent_date + timedelta(days=self.period)\\n super(Referral, self).save(*args, **kwargs)\",\n \"def post(self, request):\\n\\n try:\\n eventoid = request.POST.get('id', '')\\n correo = request.POST.get('correo', '')\\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\\n print(\\\"Exito en la asignación de staff\\\")\\n except:\\n print(\\\"Error en la asignacion de staff\\\")\\n\\n \\n return render(request, self.template, self.context)\\n #return render(request, self.template, self.context)\",\n \"def save(self, *args, **kwargs):\\n create = self.id is None\\n if create:\\n self.username = u'lc_' + self.city.lower()\\n # Give the user a randomly generated password\\n password = generate_random_string()\\n self.password = password\\n # Compose the email to send to the LC\\n context = {\\n 'lc': self,\\n 'password': password,\\n }\\n subject = render_to_string(\\n 'registration/lc_register_email_subject.txt')\\n subject = subject.strip()\\n message = render_to_string(\\n 'registration/lc_register_email_text.txt',\\n context)\\n send_mail(subject,\\n message,\\n settings.DEFAULT_FROM_EMAIL,\\n [self.email])\\n # Delegate upward to the parent classes\\n super(LocalCommittee, self).save(*args, **kwargs)\",\n \"def contact():\\n if request.method == 'POST':\\n send_email()\\n return \\\"\\\"\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n # run email job\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\",\n \"def save_model(self, request, obj, form, change):\\n admin.ModelAdmin.save_model(self, request, obj, form, change)\\n\\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\\n post_title = obj.title\\n post_url = 'http://www.aptitudeworld.com.au/blog'\\n\\n # send an email if we are adding a new post\\n if not change:\\n # send an email to all users in the system to notify them that a new post is available\\n email_title = 'Aptitude World ANZ - A new post is available'\\n email_from = 'info@aptitudeworld.com.au'\\n\\n try:\\n # get all user emails in the database\\n users = apps.users.models.User.objects.all()\\n confirmation_email_txt_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\\n confirmation_email_html_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\\n for u in users:\\n email_to = u.email_address\\n full_name = u.first_name + \\\" \\\" + u.last_name\\n\\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\\n\\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\\n email_message.attach_alternative(email_html, 'text/html')\\n email_message.send()\\n except BadHeaderError:\\n pass\",\n \"def send_email():\\n send_mail(\\\"You've got some problem.\\\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\\n ['dimazarj2009@rambler.ru'], fail_silently=False,)\",\n \"def create_sent_email(self, *args, **kwargs):\\n receiver = kwargs['receiver']\\n sender = kwargs['sender']\\n user = kwargs['user']\\n body = kwargs['body']\\n subject = kwargs['subject']\\n tracker_id = kwargs['tracker_id']\\n if receiver and sender and subject and body:\\n sent_email = SentEmail()\\n if tracker_id:\\n sent_email.tracker_id = tracker_id\\n sent_email.read_status = 'Tracking'\\n else:\\n sent_email.read_status = 'Not Tracking'\\n sent_email.receiver = receiver\\n sent_email.subject = subject\\n sent_email.sender = sender\\n sent_email.status = 'sent'\\n sent_email.user = user\\n sent_email.body = body\\n sent_email.save()\\n return sent_email\\n else:\\n return False\",\n \"def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\\n\\n if is_new:\\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\\n\\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))\",\n \"def sendEmail(householdID):\\n contactID = mdb.getContact(householdID)\\n sqlq = \\\"\\\"\\\"\\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\\n FROM Contact\\n WHERE idContact = '{}';\\n \\\"\\\"\\\".format(contactID)\\n result = mdb.getSQL(sqlq)[0]\\n\\n # thisName = (\\\"%s %s\\\" % (result['Name'], result['Surname']))\\n thisName = (\\\"%s\\\" % (result['Name']))\\n thisAddress = (\\\"%s</br>%s</br>%s %s\\\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\\n thisAddress = thisAddress.replace(\\\"None </br>\\\", \\\"\\\")\\n thisEmail = (\\\"%s\\\" % (result['email']))\\n thisStatus = (\\\"%s\\\" % (result['status']))\\n thisAddress = thisAddress.replace(\\\"None</br>\\\", \\\"\\\")\\n participantCount = (\\\"%s\\\" % mdb.getParticipantCount(str(householdID)))\\n # prepare the custom email\\n\\n thisPath = os.path.dirname(os.path.abspath(__file__))\\n if (thisStatus == 'de'):\\n # DOESN'T happen yet - de is excluded from query for now\\n # emailPath = os.path.join(thisPath, \\\"emails/email_confirm_de.html\\\")\\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\\n else:\\n emailPath = os.path.join(thisPath, \\\"emails/email_automated_date.html\\\")\\n\\n templateFile = open(emailPath, \\\"r\\\")\\n templateText = templateFile.read()\\n templateFile.close()\\n templateText = templateText.replace(\\\"[householdID]\\\", householdID)\\n templateText = templateText.replace(\\\"[contactID]\\\", contactID)\\n templateText = templateText.replace(\\\"[name]\\\", thisName)\\n templateText = templateText.replace(\\\"[address]\\\", thisAddress)\\n templateText = templateText.replace(\\\"[securityCode]\\\", mdb.getSecurityCode(householdID))\\n templateText = templateText.replace(\\\"[participantCount]\\\", participantCount)\\n\\n # Subject\\n subjectLine = templateText.splitlines()[0]\\n templateText = templateText[templateText.find('\\\\n') + 1:] # find line break and return all from there - i.e. remove first line\\n \\n # email file\\n emailFilePath = os.path.join(thisPath, \\\"tempEmail.htmail\\\")\\n\\n emailFile = open(emailFilePath, \\\"w+\\\")\\n emailFile.write(templateText)\\n emailFile.close()\\n\\n # call('mutt -e \\\"set content_type=text/html\\\" -s \\\"[TESTING]' + subjectLine + '\\\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\\n call('mutt -e \\\"set content_type=text/html\\\" -s \\\"' + subjectLine + '\\\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)\",\n \"def save(self,owner=None):\\n\\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\\n\\n t = Ticket(\\n title = self.cleaned_data['title'],\\n owner=owner,\\n submitter_email = self.cleaned_data['submitter_email'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n account = owner.account\\n )\\n\\n t.save()\\n\\n f = FollowUp(\\n ticket = t,\\n title = _('Ticket Opened Via Web'),\\n date = datetime.now(),\\n public = True,\\n comment = self.cleaned_data['body'],\\n account = owner.account\\n )\\n\\n f.save()\\n\\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n }\\n\\n messages_sent_to = []\\n\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self, user):\\n\\n q = self.cleaned_data['queue']\\n\\n t = Ticket( title = self.cleaned_data['title'],\\n submitter_email = self.cleaned_data['submitter_email'],\\n account = self.cleaned_data['account'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n owner = self.cleaned_data['owner']\\n )\\n\\n if HAS_TAG_SUPPORT:\\n t.tags = self.cleaned_data['tags']\\n\\n if self.cleaned_data['assigned_to']:\\n try:\\n t.assigned_to = self.cleaned_data['assigned_to']\\n except User.DoesNotExist:\\n t.assigned_to = None\\n t.save()\\n\\n f = FollowUp( ticket = t,\\n title = _('Ticket Opened'),\\n date = datetime.now(),\\n public = False,\\n comment = self.cleaned_data['body'],\\n systemuser = user.account,\\n )\\n if self.cleaned_data['assigned_to']:\\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\\n 'name': t.get_assigned_to\\n }\\n\\n f.save()\\n \\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n 'comment': f.comment,\\n }\\n \\n messages_sent_to = []\\n\\n if t.submitter_email:\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n #FIX USERSETTINGS\\n #=======================================================================\\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\\n # send_templated_mail(\\n # 'assigned_to',\\n # context,\\n # recipients=t.assigned_to.email,\\n # sender=q.from_address,\\n # fail_silently=True,\\n # files=files,\\n # )\\n # messages_sent_to.append(t.assigned_to.email)\\n #=======================================================================\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self):\\n email = self.validated_data['email']\\n name = self.validated_data['name']\\n password = self.validated_data['password']\\n\\n email_query = models.Email.objects.filter(address=email)\\n if email_query.exists():\\n email_instance = email_query.get()\\n\\n # If the email is already verified, we send a duplicate\\n # notification and exit.\\n if email_instance.is_verified:\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"is already verified.\\\",\\n email_instance,\\n )\\n email_instance.send_duplicate_notification()\\n\\n return\\n\\n # If the email is not verified, we send a new verification\\n # token to the address.\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"already exists. Sending a new verification token instead.\\\"\\n )\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\\n\\n return\\n\\n # The email doesn't exist, so we create a new user and email,\\n # then send a verification token to the email.\\n user = models.User.objects.create_user(name, password)\\n email_instance = models.Email.objects.create(address=email, user=user)\\n\\n # The user's primary email is their only email. This is the only\\n # time the primary email can be unverified.\\n user.primary_email = email_instance\\n user.save()\\n\\n logger.info(\\n \\\"Registered new user %r with email address %r\\\",\\n user,\\n email_instance,\\n )\\n\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\",\n \"def send_email(self, message):\\n pass\",\n \"def send_mail(self, subject):\\r\\n pass\",\n \"def post():\\n contactus_json = request.get_json()\\n\\n try:\\n dict_data = ContactUsSchema().load(contactus_json)\\n dict_data['description'] = escape(dict_data['description'])\\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\\n response, status = 'Received successfully', HTTPStatus.OK\\n except ValidationError as project_err:\\n response, status = {'systemErrors': project_err.messages}, \\\\\\n HTTPStatus.BAD_REQUEST\\n return response, status\",\n \"async def handle_email(self, email):\\n\\t\\tif 'logdir' in self.log_settings:\\n\\t\\t\\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\\n\\t\\t\\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\\n\\t\\t\\t\\tf.write(email.email.as_bytes())\\n\\n\\t\\tawait self.log('You got mail!')\",\n \"def set_email_sent(self, instance_id, creator, type):\\n table = self.type_map[type]\\n item = {\\n \\\"InstanceID\\\": instance_id,\\n \\\"Creator\\\": creator,\\n \\\"Scheduled For Deletion\\\": True,\\n \\\"EmailSent\\\": True\\n }\\n return table.put_item(item)\",\n \"def order_success(request, job_id, token):\\n if request.user.is_active:\\n job_obj = Job.objects.get(job_id= job_id)\\n if job_obj.employeer == request.user:\\n if verifiers.payment_token_is_valid(token, job_obj):\\n if job_obj.accepted == False:\\n job_obj.status = Status.objects.create(status_name=\\\"PN\\\", comment=\\\"Awaiting payment confirmation from Stripe, this might take a few minutes.\\\")\\n job_obj.accepted = True\\n job_obj.save()\\n \\n messages.add_message(request, messages.INFO, 'Your order has been created.')\\n return redirect('/') \\n else:\\n #someone tries to manually enter the URL with wrong payment_token\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n #the person making the request is not the same as the person who created the job\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n return redirect(\\\"/login\\\")\",\n \"def order_created(order_id):\\n order = Order.objects.get(id=order_id)\\n subject = f'Order nr. {order.id}'\\n message = f'Dear {order.first_name},\\\\n\\\\n' \\\\\\n f'You have successfully placed an order.' \\\\\\n f'Your order ID is {order.id}.'\\n mail_sent = send_mail(subject,\\n message,\\n 'info@tishman.com.ng',\\n [order.email])\\n return mail_sent\",\n \"def send_email_alert_slack(email_pk: str):\\n logging.info(\\\"Sending alert for Email<%s>\\\", email_pk)\\n email = Email.objects.get(pk=email_pk)\\n issue = email.issue\\n assert issue, f\\\"Email<{email_pk}> does not have an associated Issue\\\"\\n paralegal = issue.paralegal\\n alert_sent = False\\n case_email_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-email-list\\\", args=(str(issue.pk),)\\n )\\n case_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-detail-view\\\", args=(str(issue.pk),)\\n )\\n msg = (\\n \\\"*New Email Notification*\\\\n\\\"\\n f\\\"A new email has been received for case <{case_url}|{issue.fileref}> .\\\\n\\\"\\n f\\\"You can view this case's emails here: <{case_email_url}|here>.\\\\n\\\"\\n )\\n if paralegal and issue.is_open:\\n # Send alert directly to paralegal for open issues.\\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\\n logging.info(\\\"Looking up %s in Slack\\\", alert_email)\\n slack_user = get_slack_user_by_email(alert_email)\\n if slack_user:\\n logging.info(\\\"Notifying %s of Email<%s> via Slack\\\", alert_email, email_pk)\\n send_slack_direct_message(msg, slack_user[\\\"id\\\"])\\n alert_sent = True\\n\\n if not alert_sent:\\n # In all other cases send to alerts channel.\\n logging.info(\\n \\\"Could not find someone to DM in Slack, sending generic alert for Email<%s>\\\",\\n email_pk,\\n )\\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\\n\\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\\n logging.info(\\\"Alert sent sucessfully for Email<%s>\\\", email_pk)\",\n \"def send_sbemail(email_addy, city):\\n print(\\\"Single email send attempt: %s, %s\\\" % (email_addy, city))\\n given_pair, weather = subject_phrase_picker(city)\\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\\n print(\\\"Going to try a mailgun post to: %s with subject %s\\\" % (email_addy, given_pair['subject']))\\n response = requests.post(\\n MG_API_URL,\\n auth=(\\\"api\\\", MG_API_KEY),\\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\\n 'to': [email_addy],\\n 'subject': given_pair['subject'],\\n 'text': mailtext\\n })\\n print(\\\"status: %s | %s and response: %s\\\" % (response.status_code, response.reason, response.text))\\n print(\\\"Attempted send %s to: %s!\\\" % (mailtext, email_addy))\\n return \\\"Sent %s to: %s!\\\" % (mailtext, email_addy)\",\n \"def on_update(self):\\n\\t\\tfor email_account in frappe.get_all(\\\"Email Account\\\", filters={\\\"domain\\\": self.name}):\\n\\t\\t\\ttry:\\n\\t\\t\\t\\temail_account = frappe.get_doc(\\\"Email Account\\\", email_account.name)\\n\\t\\t\\t\\tfor attr in [\\\"email_server\\\", \\\"use_imap\\\", \\\"use_ssl\\\", \\\"use_tls\\\", \\\"attachment_limit\\\", \\\"smtp_server\\\", \\\"smtp_port\\\", \\\"use_ssl_for_outgoing\\\", \\\"append_emails_to_sent_folder\\\", \\\"incoming_port\\\"]:\\n\\t\\t\\t\\t\\temail_account.set(attr, self.get(attr, default=0))\\n\\t\\t\\t\\temail_account.save()\\n\\n\\t\\t\\texcept Exception as e:\\n\\t\\t\\t\\tfrappe.msgprint(_(\\\"Error has occurred in {0}\\\").format(email_account.name), raise_exception=e.__class__)\",\n \"def put(self, campaign_id):\\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\\n if campaign.candidates.count() > 0:\\n campaign.launch_task('generate_mailer_file')\\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\\n else:\\n api.abort(404, message='Campaign does not exist', success=False)\",\n \"def post_save_moderation(self, sender, comment, request, **kwargs):\\n model = comment.content_type.model_class()\\n if model not in self._registry:\\n return\\n self._registry[model].email(comment, comment.content_object, request)\",\n \"def delegate_last_day():\\n\\n regs = Registration.objects.all()\\n\\n template = 'notifications/last_day_mail.html'\\n\\n for reg in regs:\\n subject = 'SciPy.in 2011: Schedule and other details'\\n message = loader.render_to_string(\\n template, dictionary={'name': reg.registrant.username})\\n\\n reg.registrant.email_user(subject=subject, message=message,\\n from_email='madhusudancs@gmail.com')\",\n \"def send_email(email_dict, appointment_id):\\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)\",\n \"def post(self):\\n # se captura y se parsea a json el body del request recibido por el\\n # webhook\\n request_body = json.loads(self.request.body)\\n\\n for body in request_body:\\n \\\"\\\"\\\" Evaluar el tipo de evento ya que trae campos diferentes \\\"\\\"\\\"\\n logging.info(request_body)\\n\\n event = str(body['event'])\\n correo = str(body['email'])\\n numero_folio = str(body['numero_folio'])\\n tipo_dte = str(body['tipo_dte'])\\n\\n logging.info(event)\\n\\n if event and correo and numero_folio and tipo_dte:\\n\\n if event == 'processed':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.processed_event = event\\n email_model.processed_sg_event_id = body['sg_event_id']\\n email_model.processed_sg_message_id = body['sg_message_id']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.processed_event = event\\n e.processed_sg_event_id = body['sg_event_id']\\n e.processed_sg_message_id = body['sg_message_id']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'delivered':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.delivered_event = event\\n email_model.delivered_sg_event_id = body['sg_event_id']\\n email_model.delivered_sg_message_id = body['sg_message_id']\\n email_model.delivered_response = body['response']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.delivered_event = event\\n e.delivered_sg_event_id = body['sg_event_id']\\n e.delivered_sg_message_id = body['sg_message_id']\\n e.delivered_response = body['response']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'open':\\n model = EmailModel()\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n if email_model.opened_first_date == None:\\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_event = event\\n email_model.opened_ip = body['ip']\\n email_model.opened_user_agent = body['useragent']\\n email_model.opened_sg_event_id = body['sg_event_id']\\n email_model.opened_sg_message_id = body['sg_message_id']\\n model.email_add_count(email_model)\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n if e.opened_first_date == None:\\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_event = event\\n e.opened_ip = body['ip']\\n e.opened_user_agent = body['useragent']\\n e.opened_sg_event_id = body['sg_event_id']\\n e.opened_sg_message_id = body['sg_message_id']\\n e.email_add_count(e)\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'dropped':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.dropped_sg_event_id = body['sg_event_id']\\n email_model.dropped_sg_message_id = body['sg_message_id']\\n email_model.dropped_reason = body['reason']\\n email_model.dropped_event = event\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.dropped_sg_event_id = body['sg_event_id']\\n e.dropped_sg_message_id = body['sg_message_id']\\n e.dropped_reason = body['reason']\\n e.dropped_event = event\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'bounce':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.bounce_event = event\\n email_model.bounce_sg_event_id = body['sg_event_id']\\n email_model.bounce_sg_message_id = body['sg_message_id']\\n email_model.bounce_reason = body['reason']\\n email_model.bounce_status = body['status']\\n email_model.bounce_type = body['type']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.bounce_event = event\\n e.bounce_sg_event_id = body['sg_event_id']\\n e.bounce_sg_message_id = body['sg_message_id']\\n e.bounce_reason = str(body['reason']).decode(\\\"utf-8\\\")\\n e.bounce_status = body['status']\\n e.bounce_type = body['type']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'unsubscribe':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.unsubscribe_uid = body['uid']\\n email_model.unsubscribe_purchase = body['purchase']\\n email_model.unsubscribe_id = body['id']\\n email_model.unsubscribe_event = body['event']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.unsubscribe_uid = body['uid']\\n e.unsubscribe_purchase = body['purchase']\\n e.unsubscribe_id = body['id']\\n e.unsubscribe_event = body['event']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n else:\\n logging.info('body con campos vacios')\",\n \"def send_mail(self, address, title, message):\\n pass\",\n \"def mail_sent():\\n\\n url = settings.SITE_URL + '\\\\charts'\\n subject = 'Анализ запрошенного ресурса'\\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\\n mail_sent = send_mail(subject,\\n message,\\n 'admin@myshop.com',\\n ['user@mail.ru,'])\\n print(message)\\n return mail_sent\",\n \"def emailNote(self, authenticationToken, parameters):\\r\\n pass\",\n \"def send_new_email(user):\\n token = user.get_token()\\n message = Message(\\n 'Verify Your New Email',\\n sender='storcwebsite@gmail.com',\\n recipients=[user.temp_email])\\n message.body = f\\\"The email address associated with your Storc \\\" \\\\\\n f\\\"account has changed.\\\\n\\\\nTo verify your new email address, \\\" \\\\\\n f\\\"please click the link below:\\\\n\\\\n\\\" \\\\\\n f\\\"{url_for('users.new_email', token=token, _external=True)}\\\"\\n mail.send(message)\",\n \"def report_save(message, bot):\\n text = message.text\\n reports = parse_report(text)\\n author_id = message.from_user.id\\n author_username = message.from_user.username\\n to_save = []\\n for report in reports:\\n obj_to_save = {\\n 'author': author_id,\\n 'author_name': author_username,\\n 'text': report,\\n }\\n to_save.append(obj_to_save)\\n try:\\n request = requests.post(\\n url='http://127.0.0.1:8000/api/v1/report/',\\n headers={\\n 'Authorization': f'Bearer {jwt_token}',\\n },\\n json=to_save,\\n )\\n bot.send_message(\\n message.from_user.id,\\n request.status_code,\\n )\\n except requests.exceptions.ConnectionError:\\n bot.send_message(\\n message.from_user.id,\\n 'Ошибка соединения с сервером',\\n )\",\n \"def perform_create(self, serializer):\\n user = serializer.save()\\n signals.user_registered.send(\\n sender=self.__class__, user=user, request=self.request\\n )\\n\\n context = get_email_context(user)\\n to = [get_user_email(user)]\\n if djconf.SEND_ACTIVATION_EMAIL:\\n djconf.EMAIL.activation(self.request, context).send(to)\\n elif djconf.SEND_CONFIRMATION_EMAIL:\\n djconf.EMAIL.confirmation(self.request, context).send(to)\",\n \"def send(self):\\n ReferralActivity.objects.create(\\n actor=self.user,\\n verb=ReferralActivityVerb.CREATED,\\n referral=self,\\n )\\n # Confirm the referral has been sent to the requester by email\\n Mailer.send_referral_saved(self)\\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\\n # email notifications)\\n for unit in self.units.all():\\n contacts = unit.members.filter(\\n unitmembership__role=UnitMembershipRole.OWNER\\n )\\n for contact in contacts:\\n Mailer.send_referral_received(self, contact=contact, unit=unit)\",\n \"def execute_automatic_email(self, request, pk=None):\\n try:\\n retreat = Retreat.objects.get(pk=pk)\\n except Exception:\\n response_data = {\\n 'detail': \\\"Retreat not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n try:\\n email = AutomaticEmail.objects.get(\\n id=int(request.GET.get('email'))\\n )\\n except Exception:\\n response_data = {\\n 'detail': \\\"AutomaticEmail not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n # Notify a user for every reserved seat\\n emails = []\\n for reservation in retreat.reservations.filter(is_active=True):\\n if reservation.automatic_email_logs.filter(email=email):\\n pass\\n else:\\n send_automatic_email(reservation.user, retreat, email)\\n AutomaticEmailLog.objects.create(\\n reservation=reservation,\\n email=email\\n )\\n emails.append(reservation.user.email)\\n\\n response_data = {\\n 'stop': True,\\n 'emails': emails\\n }\\n return Response(response_data, status=status.HTTP_200_OK)\",\n \"def save(self):\\n owner = self.context[\\\"request\\\"].user\\n recipient = self._recipient_email_inst.user\\n\\n logger.info(\\n \\\"Transferring subscription %r from user %r to user %r\\\",\\n owner.know_me_subscription,\\n owner.pk,\\n recipient.pk,\\n )\\n\\n with transaction.atomic():\\n models.Subscription.objects.filter(user=recipient).delete()\\n models.Subscription.objects.filter(user=owner).update(\\n user=recipient\\n )\",\n \"def send_message(self):\\r\\n return \\\"success\\\"\",\n \"def alert_for_pending_mails_1(request):\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tThread(target=alert_for_pending_mails_1_worker).start()\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tresponse = {}\\n\\n\\tresponse[\\\"info_to_contact\\\"] = \\\"Ok\\\"\\n\\n\\treturn response\",\n \"def save(self, request, item, linked_item, linked_item_form):\\n\\t\\titem.description = linked_item_form.cleaned_data['subject']\\n\\t\\tlinked_item.text = linked_item_form.cleaned_data['text']\\n\\t\\tlinked_item.delivery_notes = linked_item_form.cleaned_data['delivery_notes']\\n\\t\\titem.priority = Priority.objects.get(id = linked_item_form.cleaned_data['priority'])\\n\\t\\t#\\n\\t\\t# Look for actions other than save\\n\\t\\t#\\n\\t\\tif request.POST.get('update', '') == 'Completed':\\n\\t\\t\\titem.fixed = True\\n\\t\\tif request.POST.get('update', '') == 'Failed':\\n\\t\\t\\titem.fixed = False\\n\\t\\t\\titem.validated = True\\n\\t\\tif request.POST.get('update', '') == 'Verified':\\n\\t\\t\\titem.fixed = True\\n\\t\\t\\titem.validated = True\\n\\n\\t\\titem.save()\\n\\t\\tlinked_item.save()\\n\\n\\t\\t#\\n\\t\\t# Check for comments and add\\n\\t\\t#\\n\\t\\tcomment_text = linked_item_form.cleaned_data['comments'].strip()\\n\\t\\tif comment_text != '':\\n\\t\\t\\tpost_comment(request, item, comment_text)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.67478585","0.64101136","0.6403527","0.61686337","0.61469436","0.61077154","0.61048436","0.60561985","0.6044016","0.60433817","0.5962845","0.59153545","0.5879812","0.5870707","0.5869888","0.58417296","0.5837118","0.5825478","0.58196616","0.57915473","0.5787865","0.57669544","0.574968","0.57293534","0.5725928","0.57227284","0.5714649","0.5712621","0.56969035","0.5689443","0.5681858","0.56777817","0.5677773","0.5672467","0.5663979","0.566298","0.5644856","0.5630909","0.56081456","0.56014544","0.5594918","0.5574689","0.55703187","0.5560041","0.5559436","0.55421066","0.55217606","0.55094653","0.55061436","0.55015546","0.5498323","0.5486167","0.54831564","0.5477915","0.5476625","0.54750633","0.54570115","0.5435575","0.5427781","0.5424537","0.5414667","0.5401194","0.5398877","0.53959084","0.53936905","0.53778183","0.5365919","0.53633624","0.5341283","0.5337631","0.532041","0.5315671","0.53116167","0.5303751","0.5300678","0.52967966","0.5288687","0.5277436","0.52724516","0.52620757","0.5260432","0.5252903","0.52519655","0.52476853","0.5245848","0.5243782","0.5242832","0.5242335","0.52420616","0.52394414","0.52360696","0.523167","0.5231379","0.52313536","0.52290714","0.5227354","0.5226612","0.52210313","0.521271","0.5211467"],"string":"[\n \"0.67478585\",\n \"0.64101136\",\n \"0.6403527\",\n \"0.61686337\",\n \"0.61469436\",\n \"0.61077154\",\n \"0.61048436\",\n \"0.60561985\",\n \"0.6044016\",\n \"0.60433817\",\n \"0.5962845\",\n \"0.59153545\",\n \"0.5879812\",\n \"0.5870707\",\n \"0.5869888\",\n \"0.58417296\",\n \"0.5837118\",\n \"0.5825478\",\n \"0.58196616\",\n \"0.57915473\",\n \"0.5787865\",\n \"0.57669544\",\n \"0.574968\",\n \"0.57293534\",\n \"0.5725928\",\n \"0.57227284\",\n \"0.5714649\",\n \"0.5712621\",\n \"0.56969035\",\n \"0.5689443\",\n \"0.5681858\",\n \"0.56777817\",\n \"0.5677773\",\n \"0.5672467\",\n \"0.5663979\",\n \"0.566298\",\n \"0.5644856\",\n \"0.5630909\",\n \"0.56081456\",\n \"0.56014544\",\n \"0.5594918\",\n \"0.5574689\",\n \"0.55703187\",\n \"0.5560041\",\n \"0.5559436\",\n \"0.55421066\",\n \"0.55217606\",\n \"0.55094653\",\n \"0.55061436\",\n \"0.55015546\",\n \"0.5498323\",\n \"0.5486167\",\n \"0.54831564\",\n \"0.5477915\",\n \"0.5476625\",\n \"0.54750633\",\n \"0.54570115\",\n \"0.5435575\",\n \"0.5427781\",\n \"0.5424537\",\n \"0.5414667\",\n \"0.5401194\",\n \"0.5398877\",\n \"0.53959084\",\n \"0.53936905\",\n \"0.53778183\",\n \"0.5365919\",\n \"0.53633624\",\n \"0.5341283\",\n \"0.5337631\",\n \"0.532041\",\n \"0.5315671\",\n \"0.53116167\",\n \"0.5303751\",\n \"0.5300678\",\n \"0.52967966\",\n \"0.5288687\",\n \"0.5277436\",\n \"0.52724516\",\n \"0.52620757\",\n \"0.5260432\",\n \"0.5252903\",\n \"0.52519655\",\n \"0.52476853\",\n \"0.5245848\",\n \"0.5243782\",\n \"0.5242832\",\n \"0.5242335\",\n \"0.52420616\",\n \"0.52394414\",\n \"0.52360696\",\n \"0.523167\",\n \"0.5231379\",\n \"0.52313536\",\n \"0.52290714\",\n \"0.5227354\",\n \"0.5226612\",\n \"0.52210313\",\n \"0.521271\",\n \"0.5211467\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.54872906"},"document_rank":{"kind":"string","value":"51"}}},{"rowIdx":90868,"cells":{"query":{"kind":"string","value":"sends an email when lead saved successfully"},"document":{"kind":"string","value":"def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())","def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has been submitted by %s.<br><br>\n\nGo here to view or edit the request: <br>\n<a href=\"%s\">%s</a>\n<br><br>\nSincerely,<br><br>\nThe Janelia Parking Permit Program\n \"\"\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\n subject = 'A new parking permit request has been entered'\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\n text_content = re.sub(r'<[^>]+>','',message)\n html_content = message\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\n\n logger.info(\"in send lead mail task\")\n return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)","def save(self, commit=False):\n mail_result = self.send_email()\n if mail_result:\n self.instance.is_admin_notified = True\n\n contact = super().save(commit=commit)\n\n return contact","def save(self, **kwargs):\n\t\tif (self.answer != \"\"):\n\t\t\tself.answered = True\n\t\t\tsuper().save(**kwargs)\n\t\t\tmail_subject = \"Your Question has been answered.\"\n\t\t\t# Do not break the following string or the email will get cut off\n\t\t\tmessage = f\"Hi {self.user.first_name},\\n\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\n\\n\"\n\n\t\t\temail = EmailMessage(\n\t\t\t\tmail_subject, message, to=[self.user.email]\n\t\t\t)\n\t\t\temail.send()\n\t\telse:\n\t\t\tsuper().save(**kwargs)","def save(self): # noqa\n self.bound_admin.message_success(self.bound_request, _('Done.'))","def action_my_payslip_sent(self):\n self.ensure_one()\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\n if template:\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\n self.flag = True","def test_email_sent_on_failure(self):\n self._authorize()\n data = {\n 'Subject_Number': '000-1111',\n 'Pin_Code': '1234',\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\n 'Mobile_Number': '2223334444',\n }\n patient = self.create_xml_patient(data)\n payload = self.create_xml_payload([patient])\n response = self._post(payload)\n self.assertEqual(response.status_code, 500)\n self.assertEqual(len(mail.outbox), 1)","def send_referral_saved(cls, referral, created_by):\n\n template_id = settings.SENDINBLUE[\"REFERRAL_SAVED_TEMPLATE_ID\"]\n\n data = {\n \"params\": {\"case_number\": referral.id},\n \"replyTo\": cls.reply_to,\n \"templateId\": template_id,\n \"to\": [{\"email\": created_by.email}],\n }\n\n cls.send(data)","def save_model(self, request, obj, form, change):\n if change and obj.status == 'V':\n if Organization.objects.get(pk=obj.pk).status == 'P':\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\n try:\n sender = Plugin_Contact.objects.all()[0].email\n except IndexError:\n sender = None\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\n send_mail(u'Validation de votre fiche sur la PASR',\n u'Bonjour,\\n\\nVotre fiche vient d\\'être validée sur la PASR.',\n sender, dests)\n super(OrganizationAdmin, self).save_model(request, obj, form, change)","def callback_success_message(request):\n msg = 'Sucessfully recorded form :)'\n logger.info(msg)\n messages.info(request._request, msg)","def save(self):\n super(Notification, self).save()\n # get user recipent\n us = self.notified_user\n # check that user has a valid email address\n if us.email.find('@') > 0 and us.email.find('.') > 0:\n # mandatory fields\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\n to = us.email\n from_email = settings.DEFAULT_FROM_EMAIL\n # get text version of the message\n text_content = self.get_email_content_from_type(\n self.notification_type\n )\n # FIXME: HTML version implementation pending\n html_content = self.get_email_content_from_type(\n self.notification_type\n )\n msg = EmailMultiAlternatives(\n subject, \n text_content,\n from_email,\n [to]\n )\n msg.attach_alternative(html_content, \"text/html\")\n msg.send()","def mail(request):\n email_admin.delay('testinggg')\n return JsonResponse({\"details\":\"working\"})","def on_post(self):\n return \"Ok, the stuff is being saved\"","def send_email(request):\n # send emails and return some manner of success response\n send(**request.params)\n return {'success': 'mail sent!'}","def unsuccessful(self, **post):\n cr, uid, context = request.cr, request.uid, request.context\n attchment_list = []\n mail_obj = request.env['mail.mail'].sudo()\n\n sale_order_id = request.session.get('sale_order_id')\n\n print(\"\\n\\n\\n\\n\\n=========sale order id===========\", sale_order_id)\n if sale_order_id:\n order = request.env['sale.order'].sudo().browse(sale_order_id)\n print(\"\\n\\n\\n\\n\\n=========sale order order===========\", order)\n else:\n return request.redirect('/enrolment_book')\n request.website.sale_reset()\n\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\n raise_if_not_found=False)\n if template_id:\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\n order.id)\n enroll_file_name = \"Pro-Forma \" + order.name\n if pdf_data_enroll:\n pdfvals = {'name': enroll_file_name,\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas': base64.b64encode(pdf_data_enroll[0]),\n 'datas_fname': enroll_file_name + '.pdf',\n 'res_model': 'sale.order',\n 'type': 'binary'}\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\n attchment_list.append(pdf_create)\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\n if agreement_id:\n attchment_list.append(agreement_id)\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\n if banking_detail_id:\n attchment_list.append(banking_detail_id)\n\n # mail_values = {\n # 'email_from': template_id.email_from,\n # 'reply_to': template_id.reply_to,\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # 'subject': \"Charterquest FreeQuote/Enrolment \" + order.name,\n # 'notification': True,\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\n # 'auto_delete': False,\n # }\n # msg_id = mail_obj.create(mail_values)\n # msg_id.send()\n\n template_id.sudo().write(\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\n template_id.sudo().with_context(\n # email_to=each_request.get('email'),\n # event_list=event_list,\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\n # prof_body=invoice_id.prof_body.name,\n ).send_mail(order.id, force_send=True)\n return request.render(\"cfo_snr_jnr.event_unsuccessful\", {'order': order})","def claim_email(request):\n email = request.POST.get('email', '')\n email_user = User.objects.filter(email=email)\n payload = {\n 'res': 'failed'\n }\n if email_user.exists() and \\\n not email_user[0].profile.send_mail:\n request.user.profile.add_email(email)\n payload['res'] = 'success'\n\n return payload","def test_admin_approval_complete_email(self):\n new_user = UserModel().objects.create_user(**self.user_info)\n profile = self.registration_profile.objects.create_profile(new_user)\n profile.send_admin_approve_complete_email(Site.objects.get_current())\n self.assertEqual(len(mail.outbox), 1)\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])","def add_manualpost_email(request, submission_id=None, access_token=None):\n\n if request.method == 'POST':\n try:\n button_text = request.POST.get('submit', '')\n if button_text == 'Cancel':\n return redirect(\"submit/manual_post.html\")\n \n form = SubmissionEmailForm(request.POST)\n if form.is_valid():\n submission_pk = form.cleaned_data['submission_pk']\n message = form.cleaned_data['message']\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n \n if form.cleaned_data['direction'] == 'incoming':\n msgtype = 'msgin'\n else:\n msgtype = 'msgout'\n \n submission, submission_email_event = (\n add_submission_email(request=request,\n remote_ip=request.META.get('REMOTE_ADDR', None),\n name = form.draft_name,\n rev=form.revision,\n submission_pk = submission_pk,\n message = message,\n by = request.user.person,\n msgtype = msgtype) )\n \n messages.success(request, 'Email added.')\n \n try:\n draft = Document.objects.get(name=submission.name)\n except Document.DoesNotExist:\n # Assume this is revision 00 - we'll do this later\n draft = None\n \n if (draft != None):\n e = AddedMessageEvent(type=\"added_message\", doc=draft)\n e.message = submission_email_event.submissionemailevent.message\n e.msgtype = submission_email_event.submissionemailevent.msgtype\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\n e.by = request.user.person\n e.desc = submission_email_event.desc\n e.time = submission_email_event.time\n e.save()\n \n return redirect(\"ietf.submit.views.manualpost\")\n except ValidationError as e:\n form = SubmissionEmailForm(request.POST)\n form._errors = {}\n form._errors[\"__all__\"] = form.error_class([\"There was a failure uploading your message. (%s)\" % e.message])\n else:\n initial = {\n }\n\n if (submission_id != None):\n submission = get_submission_or_404(submission_id, access_token)\n initial['name'] = \"{}-{}\".format(submission.name, submission.rev)\n initial['direction'] = 'incoming'\n initial['submission_pk'] = submission.pk\n else:\n initial['direction'] = 'incoming'\n \n form = SubmissionEmailForm(initial=initial)\n\n return render(request, 'submit/add_submit_email.html',dict(form=form))","def receive_email_view(request):\n save_inbound_email(request.POST, request.FILES)\n return HttpResponse(200)","def validation_email_sent(request):\n assert(settings.EMAIL_VALIDATION == True)\n logging.debug('')\n data = {\n 'email': request.user.email,\n 'change_email_url': reverse('user_changeemail'),\n 'action_type': 'validate'\n }\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))","def postprocess():\n if ERRORS:\n address = 'tamakoshihiroki@gmail.com'\n body = '\\n\\n'.join( ERRORS )\n msg = create_message( body, address )\n send_mail( msg, address )","def send_mail_when_failed(self, body):\r\n pass","def test_send_email_on_invite(self):\n\n league = self.create_league()\n\n season = self.create_season(league)\n team = self.create_team(season)\n\n player = self.create_player()\n\n send_user_email_on_join(player, team.id)\n\n self.assertEqual(len(mail.outbox), 1)\n\n # if testing manually:\n # import pathlib\n # pathlib.Path(\"test_email.html\").write_text(last_sent.body)","def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\n lead = Lead.objects.get(id=lead_id)\n created_by = lead.created_by\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\n for user in recipients:\n recipients_list = []\n user = User.objects.filter(id=user, is_active=True).first()\n if user:\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\n recipients_list.append(user.email)\n context = {}\n context[\"url\"] = protocol + '://' + domain + \\\n reverse('leads:view_lead', args=(lead.id,))\n context[\"user\"] = user\n context[\"lead\"] = lead\n context[\"created_by\"] = created_by\n context[\"source\"] = source\n subject = 'Assigned a lead for you. '\n html_content = render_to_string(\n 'assigned_to/leads_assigned.html', context=context)\n msg = EmailMessage(\n subject,\n html_content,\n to=recipients_list\n )\n msg.content_subtype = \"html\"\n msg.send()","def form_valid(self, form):\n self.object = form.save()\n self.send_verify_email()\n return super().form_valid(form)","def save_response(self):\r\n self.q(css='input.save-button').first.click()\r\n EmptyPromise(\r\n lambda: 'save' in self.alert_message.lower(),\r\n \"Status message saved\"\r\n ).fulfill()","def form_valid(self, form):\n\n ModelAccountVerification.objects.send_verification_email(form.user,\n verification_type=\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\n\n messages.info(self.request, \"Email sent successfully to the {0}\".format(form.user.email))\n\n return HttpResponseRedirect(self.success_url)","def execute(self):\n return LOGGER.info(f\"{datetime.datetime.now()} - Sending EMail to the configured email list\")","def test_post_external_attendance_successfull(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-03T19:00',\n 'attendee_email': 'student@external.mx',\n 'exit_datetime': '2021-08-03T20:00'}\n response = self.client.post(\n reverse('tracking:external-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def post(self):\n return send_email(request.args)","def save_model(self, request, obj, form, change):\n notify.send(\n request.user,\n recipient=obj.reported_by,\n verb='updated your report',\n action_object=obj,\n description=obj.notes,\n target=obj\n )\n obj.save()","def send_warning(self):\n\n # Check whether all the necessary parameters for SMS are present\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\n client = Client(self.account_sid, self.auth_token)\n\n try:\n sms = client.messages.create(\n body=\"\"\"Last will: It was at least 30 days since your last check in. \n This is a reminder to check in in the next 24 hours.\"\"\",\n from_=self.twilio_phone,\n to=self.your_phone)\n sms\n print(\"\\nSMS sent\")\n except Exception as e:\n print(f\"An error occurred while trying to send the SMS. Error: {e}\")\n\n else:\n print(\"\\nMissing SMS parameters. SMS not sent\")\n\n # Check whether all the necessary parameters for email are present\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\n message = f\"\"\"It has been at least 30 days since you last checked in. \nYou need to check in in the next 24 hours.\\n\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\n\nIn order to reset simply go to the working directory and run python3 last_will.py\"\"\"\n\n # send_email will return 0 if everything went ok, otherwise it will return an error message\n status = send_email(self.sender_name, self.your_email,\n self.email, self.email_pwd,\n subject='Last will: Reminder to check in', unencrypted_message=message)\n\n if status != 0:\n print(status)\n exit(1)\n else:\n print(\"Email sent\\n\")\n\n print(f\"You have until {self.deadline} to check in. \"\n f\"In order to do that simply go to the working directory and run ./last_will.sh\\n\")\n else:\n print(\"Missing email parameters. Email not sent.\\n\")\n exit(1)","def complete_allotment(request):\n reg_no = request.user.username\n user = get_object_or_404(User, username=reg_no)\n sec_email = user.get_profile().secondary_email\n options_chosen = get_chosen_options(user)\n context = {'username': reg_no, 'email': sec_email, \n 'options_chosen': options_chosen}\n ##Sending mail with allotment details \n admin = User.objects.get(pk=1) \n from_email = admin.email\n subject = \"JAM 2012 admissions\"\n content = \"The following options were chosen by you \\n \\n\"\n if options_chosen:\n counter = 1\n for option in options_chosen:\n content += \"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\n\" %(counter, option.opt_code, option.opt_name, option.opt_location)\n counter += 1 \n \n content += \"\\n \\n \\nPlease do not delete this email and keep it for reference purposes. \\n \\n \\n \\n Regards, \\n JAM Office, IIT Bombay\"\n send_mail(subject, content, from_email, [sec_email])\n admin_content = content\n admin_content +=\"\\n\\n\\n#%s:\" % (reg_no)\n counter = 1\n for option in options_chosen:\n admin_content += \"%s,%s:\" %(counter, option.opt_code) \n counter += 1\n admin_content +=\"#\"\n admin_content += time.ctime() \n mail_admins(subject, admin_content) \n return render(request, 'allotter/complete.html', context)","def send_reminder(self):\n pass","def create_response_and_send_mail(token, partner_id, email):\n \t #set url\n\t url = survey.public_url\n \t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\n\t\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\n\t\t template = self.pool.get('email.template').browse(cr,uid,template_id)\n\t\t body = template.body_html\n\t if token:\n \t url = url + '/' + token\n\n\t # post the message\n \t values = {\n \t'model': None,\n\t 'res_id': None,\n \t 'subject': template.subject,\n \t'body': body.replace(\"__URL__\", url),\n\t 'body_html': body.replace(\"__URL__\", url),\n \t 'parent_id': None,\n \t'partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\n\t 'email_from': 'info@sba.org' or None,\n \t 'email_to': partner.email,\n\t\t }\n\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\n \t mail_mail_obj.send(cr, uid, [mail_id], context=context)","def send_reminder(self):\n message_contents = \"This is a reminder that your event: \" + self.event_title + \" takes place on \" + self.event_date + \" in \" + self.event_location\n subject = \"Event Reminder\"\n attendees = self.gameplanuser_set.all()\n for attendee in attendees:\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\n remindermessage.save()","def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\n\n location = reverse(\"KursAnmeldung-verify_email\", kwargs={\"hash\":rnd_hash})\n verify_link = request.build_absolute_uri(location)\n\n # FIXME: convert to users local time.\n now = datetime.datetime.utcnow()\n\n email_context = {\n \"verify_link\": verify_link,\n \"db_entry\": db_entry,\n \"now\": now,\n }\n\n # Render the internal page\n emailtext = render_to_string(\"kurs_anmeldung/verify_mailtext.txt\", email_context)\n\n # Get the preferences from the database:\n raw_notify_list = preferences[\"notify\"]\n notify_list = raw_notify_list.splitlines()\n notify_list = [i.strip() for i in notify_list if i]\n\n email_kwargs = {\n \"from_email\": preferences[\"from_email\"],\n \"subject\": preferences[\"email_subject\"],\n \"body\": emailtext,\n \"to\": [db_entry.email],\n \"bcc\": notify_list,\n }\n\n if MAIL_DEBUG == True:\n msg = u\"MAIL_DEBUG is on: No Email was sended!\"\n request.page_msg(msg)\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n\n request.page_msg(\"django.core.mail.EmailMessage kwargs:\")\n request.page_msg(email_kwargs)\n\n request.page_msg(\"debug mail text:\")\n request.page_msg(mark_safe(\"<pre>%s</pre>\" % emailtext))\n return\n\n # We can't use django.core.mail.send_mail, because all members\n # of the recipient list will see the others in the 'To' field.\n # But we would like to notify the admins via 'Bcc' field.\n\n connection = SMTPConnection(fail_silently=False)\n email = EmailMessage(**email_kwargs)\n\n try:\n sended = email.send(fail_silently=False)\n except Exception, err:\n msg = \"Error sending mail: %s\" % err\n LogEntry.objects.log_action(app_label=\"kurs_anmeldung\", action=\"error\",\n message=msg\n )\n db_entry.log(request, msg)\n db_entry.mail_sended = False\n if settings.DEBUG or request.user.is_staff:\n db_entry.save()\n raise\n else:\n db_entry.mail_sended = sended\n db_entry.log(request, \"mail sended: %s\" % sended)","def confirm_email(self):\n self.active = True\n self.save()","def save(self, register=False):\n sha = sha1(self.email).hexdigest()\n infos = self.to_dict()\n infos[\"plan\"] = infos[\"plan\"][\"id\"] if infos[\"plan\"] else None\n\n if not redis.hmset(\"sl:account:{}\".format(sha), infos):\n raise SleekException(\"Could not save current user.\", 401)\n \n if register:\n try:\n send_email.delay(\n \"florent.esp@gmail.com\", \"Welcome to sleekstatus !\",\n \"Welcome message\", [self.email]\n )\n except:\n pass # Cannot send email","def task_send_reminder_email():\n send_reminder_email()\n logger.info(\"Sent reminder email\")","def account_activation_sent(request):\r\n\treturn render(request, 'account_activation_sent.html')","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n reminders.Patient.objects.filter(\n pk__in=[self.test_patient.pk, self.other_patient.pk]\n ).update(next_visit=appt_date)\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n self.startRouter()\n self.router.logger.setLevel(logging.DEBUG)\n\n # run email job\n from afrims.apps.reminders.app import daily_email_callback\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)\n self.stopRouter()","def save(self, *args, **kwargs):\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\n allUsers = Bruker.get_all_dict(Bruker)\n mottaker = allUsers[int(data['mottaker'])]\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\n melding.save()","def save_appointment_details(request, calendar_id):\n def schedule_mail(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n def schedule_sms(reminder_date, appointment):\n # Configure our scheduler for reminder\n try:\n trigger = DateTrigger(run_date=reminder_date)\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\n except Exception as exp:\n print(exp)\n \n start_time = request.GET['start_time'][:19]\n end_time = request.GET['end_time'][:19]\n \n start_time = datetime.strptime(start_time, \"%Y-%m-%dT%H:%M:%S\")\n end_time=datetime.strptime(end_time, \"%Y-%m-%dT%H:%M:%S\")\n \n calendar_obj = Calendar.objects.get(pk=calendar_id)\n # if this is a POST request we need to process the form data\n if request.method == 'POST':\n\n # create a form instance and populate it with data from the request:\n form = AppointmentForm(request.POST)\n\n # check whether it's valid and save it\n if form.is_valid():\n # Save appointment details\n \n mobilephone = form.data['mobilephone']\n email = form.data['email']\n first_name = form.data['first_name']\n last_name = form.data['last_name']\n notes = form.data['notes']\n\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\n \n appointment.calendar = calendar_obj\n appointment.save()\n\n try:\n send_appointment_mail(appointment) # send appointment details email\n except Exception as exp:\n print(exp)\n \n try:\n send_appointment_sms(appointment) # send appointment details sms\n except Exception as exp:\n print(exp)\n \n # Calculate reminder schedule dates\n reminder1 = start_time - timedelta(hours=2)\n reminder2 = start_time - timedelta(hours=24)\n reminder3 = start_time - timedelta(days=7)\n\n # Schedule mails\n schedule_mail(reminder1, appointment)\n schedule_mail(reminder2, appointment)\n schedule_mail(reminder3, appointment)\n \n # Schedule sms\n schedule_sms(reminder1, appointment)\n schedule_sms(reminder2, appointment)\n schedule_sms(reminder3, appointment)\n \n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\n \n # if a GET (or any other method) we'll create a blank form\n else:\n form = AppointmentForm()\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\n 'office_location': calendar_obj.office_location})","def post(\n self,\n email,\n company_name,\n location,\n job_profile,\n salary,\n username,\n password,\n security_question,\n security_answer,\n notes,\n date_applied,\n status,\n):","def post_save_order_receiver(sender, instance, created, *args, **kwargs):\n sender_email = instance.cargo.sender.email\n recepient_email = instance.cargo.recepient.email\n\n booking_agent = instance.cargo.booking_station.branch_agent.email\n\n if created:\n instance._set_order_price()\n instance._set_time_approximations()\n price = instance.price\n\n subject = \"Order Finalized and ready to go.\"\n message = f\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\"\n\n send_async_email(\n subject=subject,\n message=message,\n sender=booking_agent,\n recepients=[sender_email, recepient_email],\n )","def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\n\n #in_reply_to = form.cleaned_data['in_reply_to']\n # create Message\n parts = pyzmail.parse.get_mail_parts(message)\n body=''\n for part in parts:\n if part.is_body == 'text/plain' and part.disposition == None:\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\n body = body + payload + '\\n'\n\n msg = submit_message_from_message(message, body, by)\n\n if (submission_pk != None):\n # Must exist - we're adding a message to an existing submission\n submission = Submission.objects.get(pk=submission_pk)\n else:\n # Must not exist\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\n if submissions.count() > 0:\n raise ValidationError(\"Submission {} already exists\".format(name))\n \n # create Submission using the name\n try:\n submission = Submission.objects.create(\n state_id=\"waiting-for-draft\",\n remote_ip=remote_ip,\n name=name,\n rev=rev,\n title=name,\n note=\"\",\n submission_date=datetime.date.today(),\n replaces=\"\",\n )\n from ietf.submit.utils import create_submission_event, docevent_from_submission\n desc = \"Submission created for rev {} in response to email\".format(rev)\n create_submission_event(request, \n submission,\n desc)\n docevent_from_submission(request,\n submission,\n desc)\n except Exception as e:\n log(\"Exception: %s\\n\" % e)\n raise\n\n if msgtype == 'msgin':\n rs = \"Received\"\n else:\n rs = \"Sent\"\n\n desc = \"{} message - manual post - {}-{}\".format(rs, name, rev)\n submission_email_event = SubmissionEmailEvent.objects.create(\n desc = desc,\n submission = submission,\n msgtype = msgtype,\n by = by,\n message = msg)\n #in_reply_to = in_reply_to\n\n save_submission_email_attachments(submission_email_event, parts)\n return submission, submission_email_event","def handle_sent(self, instance):\n if not instance.created_by:\n return\n\n activity = Activity(\n actor=instance.created_by,\n verb=RestrictedMailSent,\n object=instance,\n time=instance.used,\n extra_context={},\n )\n self.manager.add_activity(\n activity, [instance.created_by.pk], [NotificationFeed]\n )\n\n # Send notification\n notification = RestrictedMailSentNotification(instance.created_by)\n notification.notify()","def handle_failure(self, instance, sender=None, reason=None):\n send_email.delay(\n to_email=sender,\n context={\"reason\": reason},\n subject=\"Kunne ikke sende ut begrenset epost\",\n plain_template=\"restricted/email/process_failure.txt\",\n html_template=\"restricted/email/process_failure.html\",\n )","def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):\n print(\"member email\",mem_mail)\n logger.info(\"in sending existing customer mail task\")\n return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website)","def handle(self, *args, **options):\n\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\n if candidate.contact_address and candidate.participating]\n\n\n print 'sending e-mails'\n conn = get_connection()\n for c in candidates_with_email:\n if c.should_send_reminder():\n\n print 'emailing', c\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\n c.last_reminder_sent = timezone.now()\n c.save()\n msg = make_email(c)\n conn.send_messages([msg])\n conn.close()","def send(self, request, id, tribe_slug):\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\n document = self.get_document(id, tribe_slug)\n \n form = self._set_user_email_address(request)\n email = self._get_user_email_address(request)\n if form or not email:\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\n \n # NB: Temporarily disabling actual email sending for development\n #email_document(document, to=[email], subject='Document: %s' % document.title)\n print \"Sending email to %s\" % email \n \n # Send a signal to let everyone know about this document interaction\n document_interaction.send(sender=self, document=document, mode=\"sent\", request=request, recipient=email)\n \n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))","def submit(self):\r\n try:\r\n response = self.send_request()\r\n if response.ok:\r\n self.submitted_at = datetime.now(pytz.UTC)\r\n self.status = \"submitted\"\r\n self.save()\r\n else:\r\n self.status = \"must_retry\"\r\n self.error_msg = response.text\r\n self.save()\r\n except Exception as error:\r\n log.exception(error)\r\n self.status = \"must_retry\"\r\n self.save()","def _send(self, email_message): \n\t\tif not email_message.recipients(): \n\t\t\treturn False \n\t\trecipients = map(self._sanitize, email_message.recipients()) \n\t\tMsg = o.CreateItem(0)\n \t\tMsg.To = recipients\n\t\tMsg.Subject = 'subject'\n\t\tMsg.Body = 'text'\n\t\tself.connection.SaveChanges(0)\n\n\t\tMsg.Send()\n\t\treturn True","def set_email():\n bid_fields = {'assertion':request.form['bid_assertion'],\n 'audience':settings.DOMAIN}\n headers = {'Content-type':'application/x-www-form-urlencoded'}\n h.disable_ssl_certificate_validation=True\n resp, content = h.request('https://browserid.org/verify',\n 'POST',\n body=urlencode(bid_fields),\n headers=headers)\n bid_data = json.loads(content)\n if bid_data['status'] == 'okay' and bid_data['email']:\n user = f.userFromBidOrNew(bid_data['email'])\n session['user_id'] = user.id\n session['user_email'] = user.email\n\n return redirect(url_for('main'))","def wasSent(self):\n\t\tself.sent = True\n\t\tself.save()","def action_send_email(self):\n self.ensure_one()\n if self.communication_channel != \"email\":\n return False # Maybe we should raise an error here.\n\n if self.state in (\"sent\", \"done\"):\n raise ValidationError(_(\"This communication is already sent.\"))\n\n lines_2be_processed = self.credit_control_line_ids.filtered(\n lambda line: line.state != \"sent\"\n )\n\n if not lines_2be_processed:\n raise ValidationError(_(\"There is no draft lines to send.\"))\n\n self = self.with_context(lang=self.partner_id.lang)\n\n partner = self.partner_id\n mail_template = self.policy_level_id.email_template_id\n # Send the email\n partner.with_context(credit_control_mail=True).message_post_with_template(\n template_id=mail_template.id,\n model=self._name,\n res_id=self.id,\n )\n # Set the state of the credit control lines to \"queued\"\n lines_2be_processed.write({\"state\": \"sent\"})\n self.state = \"sent\"","def test_activate_form(self, mock_sendmail):\r\n res = self.testapp.post('/api/v1/suspend',\r\n params={'email': u'testing@dummy.com'},\r\n status=200)\r\n\r\n success = json.loads(res.body)\r\n self.assertTrue(\r\n 'message' in success,\r\n \"Should be successful with admin email address: \" + str(res))\r\n self.assertTrue(mock_sendmail.called)","def test_post_internal_attendance_success(self):\n building_name = 'Building 1'\n room_name = 'Room 1'\n\n building = Building.objects.create(\n name=building_name\n )\n room = Room.objects.create(\n name=room_name,\n building=building\n )\n post_data = {'room': room.id,\n 'entry_datetime': '2021-08-04T10:00',\n 'attendee_email': 'student@anahuac.mx',\n 'exit_datetime': '2021-08-04T11:30'}\n response = self.client.post(\n reverse('tracking:internal-register'),\n post_data\n )\n\n self.assertEqual(response.status_code, HTTPStatus.OK)\n messages = list(response.context['messages'])\n success_message = str(messages[0])\n self.assertEqual(success_message, 'Registro realizado exitosamente')","def save(self, *args, **kwargs):\n self.expire_date = self.sent_date + timedelta(days=self.period)\n super(Referral, self).save(*args, **kwargs)","def post(self, request):\n\n try:\n eventoid = request.POST.get('id', '')\n correo = request.POST.get('correo', '')\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\n print(\"Exito en la asignación de staff\")\n except:\n print(\"Error en la asignacion de staff\")\n\n \n return render(request, self.template, self.context)\n #return render(request, self.template, self.context)","def save(self, *args, **kwargs):\n create = self.id is None\n if create:\n self.username = u'lc_' + self.city.lower()\n # Give the user a randomly generated password\n password = generate_random_string()\n self.password = password\n # Compose the email to send to the LC\n context = {\n 'lc': self,\n 'password': password,\n }\n subject = render_to_string(\n 'registration/lc_register_email_subject.txt')\n subject = subject.strip()\n message = render_to_string(\n 'registration/lc_register_email_text.txt',\n context)\n send_mail(subject,\n message,\n settings.DEFAULT_FROM_EMAIL,\n [self.email])\n # Delegate upward to the parent classes\n super(LocalCommittee, self).save(*args, **kwargs)","def contact():\n if request.method == 'POST':\n send_email()\n return \"\"","def test_sending_mail(self):\n\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\n\n # run email job\n daily_email_callback(self.router)\n\n self.assertEqual(len(mail.outbox), 1)\n message = mail.outbox[0]\n self.assertTrue(self.test_contact.email in message.to)","def save_model(self, request, obj, form, change):\n admin.ModelAdmin.save_model(self, request, obj, form, change)\n\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\n post_title = obj.title\n post_url = 'http://www.aptitudeworld.com.au/blog'\n\n # send an email if we are adding a new post\n if not change:\n # send an email to all users in the system to notify them that a new post is available\n email_title = 'Aptitude World ANZ - A new post is available'\n email_from = 'info@aptitudeworld.com.au'\n\n try:\n # get all user emails in the database\n users = apps.users.models.User.objects.all()\n confirmation_email_txt_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\n confirmation_email_html_template = open(\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\n for u in users:\n email_to = u.email_address\n full_name = u.first_name + \" \" + u.last_name\n\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\n\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\n email_message.attach_alternative(email_html, 'text/html')\n email_message.send()\n except BadHeaderError:\n pass","def send_email():\n send_mail(\"You've got some problem.\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\n ['dimazarj2009@rambler.ru'], fail_silently=False,)","def create_sent_email(self, *args, **kwargs):\n receiver = kwargs['receiver']\n sender = kwargs['sender']\n user = kwargs['user']\n body = kwargs['body']\n subject = kwargs['subject']\n tracker_id = kwargs['tracker_id']\n if receiver and sender and subject and body:\n sent_email = SentEmail()\n if tracker_id:\n sent_email.tracker_id = tracker_id\n sent_email.read_status = 'Tracking'\n else:\n sent_email.read_status = 'Not Tracking'\n sent_email.receiver = receiver\n sent_email.subject = subject\n sent_email.sender = sender\n sent_email.status = 'sent'\n sent_email.user = user\n sent_email.body = body\n sent_email.save()\n return sent_email\n else:\n return False","def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\n\n if is_new:\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\n\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))","def sendEmail(householdID):\n contactID = mdb.getContact(householdID)\n sqlq = \"\"\"\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\n FROM Contact\n WHERE idContact = '{}';\n \"\"\".format(contactID)\n result = mdb.getSQL(sqlq)[0]\n\n # thisName = (\"%s %s\" % (result['Name'], result['Surname']))\n thisName = (\"%s\" % (result['Name']))\n thisAddress = (\"%s</br>%s</br>%s %s\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\n thisAddress = thisAddress.replace(\"None </br>\", \"\")\n thisEmail = (\"%s\" % (result['email']))\n thisStatus = (\"%s\" % (result['status']))\n thisAddress = thisAddress.replace(\"None</br>\", \"\")\n participantCount = (\"%s\" % mdb.getParticipantCount(str(householdID)))\n # prepare the custom email\n\n thisPath = os.path.dirname(os.path.abspath(__file__))\n if (thisStatus == 'de'):\n # DOESN'T happen yet - de is excluded from query for now\n # emailPath = os.path.join(thisPath, \"emails/email_confirm_de.html\")\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\n else:\n emailPath = os.path.join(thisPath, \"emails/email_automated_date.html\")\n\n templateFile = open(emailPath, \"r\")\n templateText = templateFile.read()\n templateFile.close()\n templateText = templateText.replace(\"[householdID]\", householdID)\n templateText = templateText.replace(\"[contactID]\", contactID)\n templateText = templateText.replace(\"[name]\", thisName)\n templateText = templateText.replace(\"[address]\", thisAddress)\n templateText = templateText.replace(\"[securityCode]\", mdb.getSecurityCode(householdID))\n templateText = templateText.replace(\"[participantCount]\", participantCount)\n\n # Subject\n subjectLine = templateText.splitlines()[0]\n templateText = templateText[templateText.find('\\n') + 1:] # find line break and return all from there - i.e. remove first line\n \n # email file\n emailFilePath = os.path.join(thisPath, \"tempEmail.htmail\")\n\n emailFile = open(emailFilePath, \"w+\")\n emailFile.write(templateText)\n emailFile.close()\n\n # call('mutt -e \"set content_type=text/html\" -s \"[TESTING]' + subjectLine + '\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\n call('mutt -e \"set content_type=text/html\" -s \"' + subjectLine + '\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)","def save(self,owner=None):\n\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\n\n t = Ticket(\n title = self.cleaned_data['title'],\n owner=owner,\n submitter_email = self.cleaned_data['submitter_email'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n account = owner.account\n )\n\n t.save()\n\n f = FollowUp(\n ticket = t,\n title = _('Ticket Opened Via Web'),\n date = datetime.now(),\n public = True,\n comment = self.cleaned_data['body'],\n account = owner.account\n )\n\n f.save()\n\n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n }\n\n messages_sent_to = []\n\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self, user):\n\n q = self.cleaned_data['queue']\n\n t = Ticket( title = self.cleaned_data['title'],\n submitter_email = self.cleaned_data['submitter_email'],\n account = self.cleaned_data['account'],\n created = datetime.now(),\n status = Ticket.OPEN_STATUS,\n queue = q,\n description = self.cleaned_data['body'],\n priority = self.cleaned_data['priority'],\n owner = self.cleaned_data['owner']\n )\n\n if HAS_TAG_SUPPORT:\n t.tags = self.cleaned_data['tags']\n\n if self.cleaned_data['assigned_to']:\n try:\n t.assigned_to = self.cleaned_data['assigned_to']\n except User.DoesNotExist:\n t.assigned_to = None\n t.save()\n\n f = FollowUp( ticket = t,\n title = _('Ticket Opened'),\n date = datetime.now(),\n public = False,\n comment = self.cleaned_data['body'],\n systemuser = user.account,\n )\n if self.cleaned_data['assigned_to']:\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\n 'name': t.get_assigned_to\n }\n\n f.save()\n \n files = []\n if self.cleaned_data['attachment']:\n import mimetypes\n file = self.cleaned_data['attachment']\n filename = file.name.replace(' ', '_')\n a = Attachment(\n followup=f,\n filename=filename,\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\n size=file.size,\n )\n a.file.save(file.name, file, save=False)\n a.save()\n \n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\n # Only files smaller than 512kb (or as defined in \n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\n files.append(a.file.path)\n\n context = {\n 'ticket': t,\n 'queue': q,\n 'comment': f.comment,\n }\n \n messages_sent_to = []\n\n if t.submitter_email:\n send_templated_mail(\n 'newticket_owner',\n context,\n recipients=t.submitter_email,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(t.submitter_email)\n\n #FIX USERSETTINGS\n #=======================================================================\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\n # send_templated_mail(\n # 'assigned_to',\n # context,\n # recipients=t.assigned_to.email,\n # sender=q.from_address,\n # fail_silently=True,\n # files=files,\n # )\n # messages_sent_to.append(t.assigned_to.email)\n #=======================================================================\n\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.new_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n messages_sent_to.append(q.new_ticket_cc)\n\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\n send_templated_mail(\n 'newticket_cc',\n context,\n recipients=q.updated_ticket_cc,\n sender=q.from_address,\n fail_silently=True,\n files=files,\n )\n\n return t","def save(self):\n email = self.validated_data['email']\n name = self.validated_data['name']\n password = self.validated_data['password']\n\n email_query = models.Email.objects.filter(address=email)\n if email_query.exists():\n email_instance = email_query.get()\n\n # If the email is already verified, we send a duplicate\n # notification and exit.\n if email_instance.is_verified:\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"is already verified.\",\n email_instance,\n )\n email_instance.send_duplicate_notification()\n\n return\n\n # If the email is not verified, we send a new verification\n # token to the address.\n logger.info(\n \"Not registering a new user because the email address %r \"\n \"already exists. Sending a new verification token instead.\"\n )\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()\n\n return\n\n # The email doesn't exist, so we create a new user and email,\n # then send a verification token to the email.\n user = models.User.objects.create_user(name, password)\n email_instance = models.Email.objects.create(address=email, user=user)\n\n # The user's primary email is their only email. This is the only\n # time the primary email can be unverified.\n user.primary_email = email_instance\n user.save()\n\n logger.info(\n \"Registered new user %r with email address %r\",\n user,\n email_instance,\n )\n\n verification = models.EmailVerification.objects.create(\n email=email_instance,\n )\n verification.send_email()","def send_email(self, message):\n pass","def send_mail(self, subject):\r\n pass","def post():\n contactus_json = request.get_json()\n\n try:\n dict_data = ContactUsSchema().load(contactus_json)\n dict_data['description'] = escape(dict_data['description'])\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\n response, status = 'Received successfully', HTTPStatus.OK\n except ValidationError as project_err:\n response, status = {'systemErrors': project_err.messages}, \\\n HTTPStatus.BAD_REQUEST\n return response, status","async def handle_email(self, email):\n\t\tif 'logdir' in self.log_settings:\n\t\t\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\n\t\t\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\n\t\t\t\tf.write(email.email.as_bytes())\n\n\t\tawait self.log('You got mail!')","def set_email_sent(self, instance_id, creator, type):\n table = self.type_map[type]\n item = {\n \"InstanceID\": instance_id,\n \"Creator\": creator,\n \"Scheduled For Deletion\": True,\n \"EmailSent\": True\n }\n return table.put_item(item)","def order_success(request, job_id, token):\n if request.user.is_active:\n job_obj = Job.objects.get(job_id= job_id)\n if job_obj.employeer == request.user:\n if verifiers.payment_token_is_valid(token, job_obj):\n if job_obj.accepted == False:\n job_obj.status = Status.objects.create(status_name=\"PN\", comment=\"Awaiting payment confirmation from Stripe, this might take a few minutes.\")\n job_obj.accepted = True\n job_obj.save()\n \n messages.add_message(request, messages.INFO, 'Your order has been created.')\n return redirect('/') \n else:\n #someone tries to manually enter the URL with wrong payment_token\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n #the person making the request is not the same as the person who created the job\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \n return redirect(\"/\") \n else:\n return redirect(\"/login\")","def order_created(order_id):\n order = Order.objects.get(id=order_id)\n subject = f'Order nr. {order.id}'\n message = f'Dear {order.first_name},\\n\\n' \\\n f'You have successfully placed an order.' \\\n f'Your order ID is {order.id}.'\n mail_sent = send_mail(subject,\n message,\n 'info@tishman.com.ng',\n [order.email])\n return mail_sent","def send_email_alert_slack(email_pk: str):\n logging.info(\"Sending alert for Email<%s>\", email_pk)\n email = Email.objects.get(pk=email_pk)\n issue = email.issue\n assert issue, f\"Email<{email_pk}> does not have an associated Issue\"\n paralegal = issue.paralegal\n alert_sent = False\n case_email_url = settings.CLERK_BASE_URL + reverse(\n \"case-email-list\", args=(str(issue.pk),)\n )\n case_url = settings.CLERK_BASE_URL + reverse(\n \"case-detail-view\", args=(str(issue.pk),)\n )\n msg = (\n \"*New Email Notification*\\n\"\n f\"A new email has been received for case <{case_url}|{issue.fileref}> .\\n\"\n f\"You can view this case's emails here: <{case_email_url}|here>.\\n\"\n )\n if paralegal and issue.is_open:\n # Send alert directly to paralegal for open issues.\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\n logging.info(\"Looking up %s in Slack\", alert_email)\n slack_user = get_slack_user_by_email(alert_email)\n if slack_user:\n logging.info(\"Notifying %s of Email<%s> via Slack\", alert_email, email_pk)\n send_slack_direct_message(msg, slack_user[\"id\"])\n alert_sent = True\n\n if not alert_sent:\n # In all other cases send to alerts channel.\n logging.info(\n \"Could not find someone to DM in Slack, sending generic alert for Email<%s>\",\n email_pk,\n )\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\n\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\n logging.info(\"Alert sent sucessfully for Email<%s>\", email_pk)","def send_sbemail(email_addy, city):\n print(\"Single email send attempt: %s, %s\" % (email_addy, city))\n given_pair, weather = subject_phrase_picker(city)\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\n print(\"Going to try a mailgun post to: %s with subject %s\" % (email_addy, given_pair['subject']))\n response = requests.post(\n MG_API_URL,\n auth=(\"api\", MG_API_KEY),\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\n 'to': [email_addy],\n 'subject': given_pair['subject'],\n 'text': mailtext\n })\n print(\"status: %s | %s and response: %s\" % (response.status_code, response.reason, response.text))\n print(\"Attempted send %s to: %s!\" % (mailtext, email_addy))\n return \"Sent %s to: %s!\" % (mailtext, email_addy)","def on_update(self):\n\t\tfor email_account in frappe.get_all(\"Email Account\", filters={\"domain\": self.name}):\n\t\t\ttry:\n\t\t\t\temail_account = frappe.get_doc(\"Email Account\", email_account.name)\n\t\t\t\tfor attr in [\"email_server\", \"use_imap\", \"use_ssl\", \"use_tls\", \"attachment_limit\", \"smtp_server\", \"smtp_port\", \"use_ssl_for_outgoing\", \"append_emails_to_sent_folder\", \"incoming_port\"]:\n\t\t\t\t\temail_account.set(attr, self.get(attr, default=0))\n\t\t\t\temail_account.save()\n\n\t\t\texcept Exception as e:\n\t\t\t\tfrappe.msgprint(_(\"Error has occurred in {0}\").format(email_account.name), raise_exception=e.__class__)","def put(self, campaign_id):\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\n if campaign.candidates.count() > 0:\n campaign.launch_task('generate_mailer_file')\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\n else:\n api.abort(404, message='Campaign does not exist', success=False)","def post_save_moderation(self, sender, comment, request, **kwargs):\n model = comment.content_type.model_class()\n if model not in self._registry:\n return\n self._registry[model].email(comment, comment.content_object, request)","def send_email(email_dict, appointment_id):\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)","def delegate_last_day():\n\n regs = Registration.objects.all()\n\n template = 'notifications/last_day_mail.html'\n\n for reg in regs:\n subject = 'SciPy.in 2011: Schedule and other details'\n message = loader.render_to_string(\n template, dictionary={'name': reg.registrant.username})\n\n reg.registrant.email_user(subject=subject, message=message,\n from_email='madhusudancs@gmail.com')","def send_mail(self, address, title, message):\n pass","def post(self):\n # se captura y se parsea a json el body del request recibido por el\n # webhook\n request_body = json.loads(self.request.body)\n\n for body in request_body:\n \"\"\" Evaluar el tipo de evento ya que trae campos diferentes \"\"\"\n logging.info(request_body)\n\n event = str(body['event'])\n correo = str(body['email'])\n numero_folio = str(body['numero_folio'])\n tipo_dte = str(body['tipo_dte'])\n\n logging.info(event)\n\n if event and correo and numero_folio and tipo_dte:\n\n if event == 'processed':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.processed_event = event\n email_model.processed_sg_event_id = body['sg_event_id']\n email_model.processed_sg_message_id = body['sg_message_id']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.processed_event = event\n e.processed_sg_event_id = body['sg_event_id']\n e.processed_sg_message_id = body['sg_message_id']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'delivered':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.delivered_event = event\n email_model.delivered_sg_event_id = body['sg_event_id']\n email_model.delivered_sg_message_id = body['sg_message_id']\n email_model.delivered_response = body['response']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.delivered_event = event\n e.delivered_sg_event_id = body['sg_event_id']\n e.delivered_sg_message_id = body['sg_message_id']\n e.delivered_response = body['response']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'open':\n model = EmailModel()\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n if email_model.opened_first_date == None:\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.opened_event = event\n email_model.opened_ip = body['ip']\n email_model.opened_user_agent = body['useragent']\n email_model.opened_sg_event_id = body['sg_event_id']\n email_model.opened_sg_message_id = body['sg_message_id']\n model.email_add_count(email_model)\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n if e.opened_first_date == None:\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.opened_event = event\n e.opened_ip = body['ip']\n e.opened_user_agent = body['useragent']\n e.opened_sg_event_id = body['sg_event_id']\n e.opened_sg_message_id = body['sg_message_id']\n e.email_add_count(e)\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'dropped':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.smtp_id = body['smtp-id']\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.dropped_sg_event_id = body['sg_event_id']\n email_model.dropped_sg_message_id = body['sg_message_id']\n email_model.dropped_reason = body['reason']\n email_model.dropped_event = event\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.smtp_id = body['smtp-id']\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.dropped_sg_event_id = body['sg_event_id']\n e.dropped_sg_message_id = body['sg_message_id']\n e.dropped_reason = body['reason']\n e.dropped_event = event\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'bounce':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.bounce_event = event\n email_model.bounce_sg_event_id = body['sg_event_id']\n email_model.bounce_sg_message_id = body['sg_message_id']\n email_model.bounce_reason = body['reason']\n email_model.bounce_status = body['status']\n email_model.bounce_type = body['type']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.bounce_event = event\n e.bounce_sg_event_id = body['sg_event_id']\n e.bounce_sg_message_id = body['sg_message_id']\n e.bounce_reason = str(body['reason']).decode(\"utf-8\")\n e.bounce_status = body['status']\n e.bounce_type = body['type']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n\n elif event == 'unsubscribe':\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\n if not email_model == None:\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n email_model.unsubscribe_uid = body['uid']\n email_model.unsubscribe_purchase = body['purchase']\n email_model.unsubscribe_id = body['id']\n email_model.unsubscribe_event = body['event']\n email_model.correo = str(body['email'])\n email_model.numero_folio = str(body['numero_folio'])\n email_model.tipo_dte = str(body['tipo_dte'])\n email_model.put()\n else:\n e = EmailModel()\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\n e.unsubscribe_uid = body['uid']\n e.unsubscribe_purchase = body['purchase']\n e.unsubscribe_id = body['id']\n e.unsubscribe_event = body['event']\n e.correo = str(body['email'])\n e.numero_folio = str(body['numero_folio'])\n e.tipo_dte = str(body['tipo_dte'])\n e.put()\n else:\n logging.info('body con campos vacios')","def mail_sent():\n\n url = settings.SITE_URL + '\\charts'\n subject = 'Анализ запрошенного ресурса'\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\n mail_sent = send_mail(subject,\n message,\n 'admin@myshop.com',\n ['user@mail.ru,'])\n print(message)\n return mail_sent","def emailNote(self, authenticationToken, parameters):\r\n pass","def send_new_email(user):\n token = user.get_token()\n message = Message(\n 'Verify Your New Email',\n sender='storcwebsite@gmail.com',\n recipients=[user.temp_email])\n message.body = f\"The email address associated with your Storc \" \\\n f\"account has changed.\\n\\nTo verify your new email address, \" \\\n f\"please click the link below:\\n\\n\" \\\n f\"{url_for('users.new_email', token=token, _external=True)}\"\n mail.send(message)","def perform_create(self, serializer):\n user = serializer.save()\n signals.user_registered.send(\n sender=self.__class__, user=user, request=self.request\n )\n\n context = get_email_context(user)\n to = [get_user_email(user)]\n if djconf.SEND_ACTIVATION_EMAIL:\n djconf.EMAIL.activation(self.request, context).send(to)\n elif djconf.SEND_CONFIRMATION_EMAIL:\n djconf.EMAIL.confirmation(self.request, context).send(to)","def report_save(message, bot):\n text = message.text\n reports = parse_report(text)\n author_id = message.from_user.id\n author_username = message.from_user.username\n to_save = []\n for report in reports:\n obj_to_save = {\n 'author': author_id,\n 'author_name': author_username,\n 'text': report,\n }\n to_save.append(obj_to_save)\n try:\n request = requests.post(\n url='http://127.0.0.1:8000/api/v1/report/',\n headers={\n 'Authorization': f'Bearer {jwt_token}',\n },\n json=to_save,\n )\n bot.send_message(\n message.from_user.id,\n request.status_code,\n )\n except requests.exceptions.ConnectionError:\n bot.send_message(\n message.from_user.id,\n 'Ошибка соединения с сервером',\n )","def send(self):\n ReferralActivity.objects.create(\n actor=self.user,\n verb=ReferralActivityVerb.CREATED,\n referral=self,\n )\n # Confirm the referral has been sent to the requester by email\n Mailer.send_referral_saved(self)\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\n # email notifications)\n for unit in self.units.all():\n contacts = unit.members.filter(\n unitmembership__role=UnitMembershipRole.OWNER\n )\n for contact in contacts:\n Mailer.send_referral_received(self, contact=contact, unit=unit)","def execute_automatic_email(self, request, pk=None):\n try:\n retreat = Retreat.objects.get(pk=pk)\n except Exception:\n response_data = {\n 'detail': \"Retreat not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n try:\n email = AutomaticEmail.objects.get(\n id=int(request.GET.get('email'))\n )\n except Exception:\n response_data = {\n 'detail': \"AutomaticEmail not found\"\n }\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\n\n # Notify a user for every reserved seat\n emails = []\n for reservation in retreat.reservations.filter(is_active=True):\n if reservation.automatic_email_logs.filter(email=email):\n pass\n else:\n send_automatic_email(reservation.user, retreat, email)\n AutomaticEmailLog.objects.create(\n reservation=reservation,\n email=email\n )\n emails.append(reservation.user.email)\n\n response_data = {\n 'stop': True,\n 'emails': emails\n }\n return Response(response_data, status=status.HTTP_200_OK)","def save(self):\n owner = self.context[\"request\"].user\n recipient = self._recipient_email_inst.user\n\n logger.info(\n \"Transferring subscription %r from user %r to user %r\",\n owner.know_me_subscription,\n owner.pk,\n recipient.pk,\n )\n\n with transaction.atomic():\n models.Subscription.objects.filter(user=recipient).delete()\n models.Subscription.objects.filter(user=owner).update(\n user=recipient\n )","def send_message(self):\r\n return \"success\"","def alert_for_pending_mails_1(request):\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tThread(target=alert_for_pending_mails_1_worker).start()\n\n\tprint(\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\")\n\n\tresponse = {}\n\n\tresponse[\"info_to_contact\"] = \"Ok\"\n\n\treturn response","def save(self, domain_override=None,\n subject_template_name='registration/password_reset_subject.txt',\n email_template_name='registration/password_reset_email.html',\n use_https=False, token_generator=default_token_generator,\n from_email=None, request=None,\n html_email_template_name=None):\n email = self.cleaned_data[\"email\"]\n User = get_user_model()\n active_users = User.objects.filter(email__iexact=email, is_active=True)\n for user in active_users:\n subject = _('Flisol - Restore your password')\n # send_email(\n # subject,\n # [user.email],\n # email_template_name,\n # {\n # 'email': user.email,\n # 'uid': urlsafe_base64_encode(force_bytes(user.pk)),\n # 'user': user,\n # 'token': token_generator.make_token(user),\n # 'protocol': settings.PROTOCOL,\n # },\n # )"],"string":"[\n \"def save(self, fail_silently=False):\\n send_mail(fail_silently=fail_silently, **self.get_message_dict())\",\n \"def send_created_email(self):\\n if settings.NOTIFY_NEW_REG:\\n to = settings.NOTIFY_NEW_REG\\n message = \\\"\\\"\\\"\\\\\\nGreetings,<br><br>\\n\\nA new vehicle registration has been submitted by %s.<br><br>\\n\\nGo here to view or edit the request: <br>\\n<a href=\\\"%s\\\">%s</a>\\n<br><br>\\nSincerely,<br><br>\\nThe Janelia Parking Permit Program\\n \\\"\\\"\\\" % (self.user_display_name(), self.get_edit_url(True), self.get_edit_url(True))\\n subject = 'A new parking permit request has been entered'\\n from_email = 'parkingpermit-donotreply@janelia.hhmi.org'\\n text_content = re.sub(r'<[^>]+>','',message)\\n html_content = message\\n msg = EmailMultiAlternatives(subject, text_content, from_email, to)\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):\\n\\n logger.info(\\\"in send lead mail task\\\")\\n return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website)\",\n \"def save(self, commit=False):\\n mail_result = self.send_email()\\n if mail_result:\\n self.instance.is_admin_notified = True\\n\\n contact = super().save(commit=commit)\\n\\n return contact\",\n \"def save(self, **kwargs):\\n\\t\\tif (self.answer != \\\"\\\"):\\n\\t\\t\\tself.answered = True\\n\\t\\t\\tsuper().save(**kwargs)\\n\\t\\t\\tmail_subject = \\\"Your Question has been answered.\\\"\\n\\t\\t\\t# Do not break the following string or the email will get cut off\\n\\t\\t\\tmessage = f\\\"Hi {self.user.first_name},\\\\n\\\\n You're receiving this email because you asked a question about {self.hall.name}. The HonestHalls team has now answered your question! Revisit the {self.hall.name} hall page to see their response.\\\\n\\\\n\\\"\\n\\n\\t\\t\\temail = EmailMessage(\\n\\t\\t\\t\\tmail_subject, message, to=[self.user.email]\\n\\t\\t\\t)\\n\\t\\t\\temail.send()\\n\\t\\telse:\\n\\t\\t\\tsuper().save(**kwargs)\",\n \"def save(self): # noqa\\n self.bound_admin.message_success(self.bound_request, _('Done.'))\",\n \"def action_my_payslip_sent(self):\\n self.ensure_one()\\n template = self.env.ref('payroll_email.email_template_for_my_payroll')\\n if template:\\n self.env['mail.template'].browse(template.id).send_mail(self.id,force_send=True)\\n self.flag = True\",\n \"def test_email_sent_on_failure(self):\\n self._authorize()\\n data = {\\n 'Subject_Number': '000-1111',\\n 'Pin_Code': '1234',\\n 'Date_Enrolled': datetime.datetime.now().strftime('%b %d %Y '),\\n 'Mobile_Number': '2223334444',\\n }\\n patient = self.create_xml_patient(data)\\n payload = self.create_xml_payload([patient])\\n response = self._post(payload)\\n self.assertEqual(response.status_code, 500)\\n self.assertEqual(len(mail.outbox), 1)\",\n \"def send_referral_saved(cls, referral, created_by):\\n\\n template_id = settings.SENDINBLUE[\\\"REFERRAL_SAVED_TEMPLATE_ID\\\"]\\n\\n data = {\\n \\\"params\\\": {\\\"case_number\\\": referral.id},\\n \\\"replyTo\\\": cls.reply_to,\\n \\\"templateId\\\": template_id,\\n \\\"to\\\": [{\\\"email\\\": created_by.email}],\\n }\\n\\n cls.send(data)\",\n \"def save_model(self, request, obj, form, change):\\n if change and obj.status == 'V':\\n if Organization.objects.get(pk=obj.pk).status == 'P':\\n from ionyweb.plugin_app.plugin_contact.models import Plugin_Contact\\n try:\\n sender = Plugin_Contact.objects.all()[0].email\\n except IndexError:\\n sender = None\\n dests = Person.objects.filter(engagements__org_admin=True, engagements__organization=obj).values_list('email', flat=True)\\n send_mail(u'Validation de votre fiche sur la PASR',\\n u'Bonjour,\\\\n\\\\nVotre fiche vient d\\\\'être validée sur la PASR.',\\n sender, dests)\\n super(OrganizationAdmin, self).save_model(request, obj, form, change)\",\n \"def callback_success_message(request):\\n msg = 'Sucessfully recorded form :)'\\n logger.info(msg)\\n messages.info(request._request, msg)\",\n \"def save(self):\\n super(Notification, self).save()\\n # get user recipent\\n us = self.notified_user\\n # check that user has a valid email address\\n if us.email.find('@') > 0 and us.email.find('.') > 0:\\n # mandatory fields\\n subject = strings.EMAIL_NOTIFICATION_SUBJECT\\n to = us.email\\n from_email = settings.DEFAULT_FROM_EMAIL\\n # get text version of the message\\n text_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n # FIXME: HTML version implementation pending\\n html_content = self.get_email_content_from_type(\\n self.notification_type\\n )\\n msg = EmailMultiAlternatives(\\n subject, \\n text_content,\\n from_email,\\n [to]\\n )\\n msg.attach_alternative(html_content, \\\"text/html\\\")\\n msg.send()\",\n \"def mail(request):\\n email_admin.delay('testinggg')\\n return JsonResponse({\\\"details\\\":\\\"working\\\"})\",\n \"def on_post(self):\\n return \\\"Ok, the stuff is being saved\\\"\",\n \"def send_email(request):\\n # send emails and return some manner of success response\\n send(**request.params)\\n return {'success': 'mail sent!'}\",\n \"def unsuccessful(self, **post):\\n cr, uid, context = request.cr, request.uid, request.context\\n attchment_list = []\\n mail_obj = request.env['mail.mail'].sudo()\\n\\n sale_order_id = request.session.get('sale_order_id')\\n\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order id===========\\\", sale_order_id)\\n if sale_order_id:\\n order = request.env['sale.order'].sudo().browse(sale_order_id)\\n print(\\\"\\\\n\\\\n\\\\n\\\\n\\\\n=========sale order order===========\\\", order)\\n else:\\n return request.redirect('/enrolment_book')\\n request.website.sale_reset()\\n\\n template_id = request.env.ref('cfo_snr_jnr.unsuccessful_sponsored_regist_enrol_email_template',\\n raise_if_not_found=False)\\n if template_id:\\n pdf_data_enroll = request.env.ref('event_price_kt.report_sale_enrollment').sudo().render_qweb_pdf(\\n order.id)\\n enroll_file_name = \\\"Pro-Forma \\\" + order.name\\n if pdf_data_enroll:\\n pdfvals = {'name': enroll_file_name,\\n 'db_datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas': base64.b64encode(pdf_data_enroll[0]),\\n 'datas_fname': enroll_file_name + '.pdf',\\n 'res_model': 'sale.order',\\n 'type': 'binary'}\\n pdf_create = request.env['ir.attachment'].sudo().create(pdfvals)\\n attchment_list.append(pdf_create)\\n agreement_id = request.env.ref('cfo_snr_jnr.term_and_condition_pdf_enrolment')\\n if agreement_id:\\n attchment_list.append(agreement_id)\\n banking_detail_id = request.env.ref('cfo_snr_jnr.banking_data_pdf')\\n if banking_detail_id:\\n attchment_list.append(banking_detail_id)\\n\\n # mail_values = {\\n # 'email_from': template_id.email_from,\\n # 'reply_to': template_id.reply_to,\\n # 'email_to': order.partner_id.email if order.partner_id.email else '',\\n # 'email_cc': 'enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # 'subject': \\\"Charterquest FreeQuote/Enrolment \\\" + order.name,\\n # 'notification': True,\\n # 'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])],\\n # 'auto_delete': False,\\n # }\\n # msg_id = mail_obj.create(mail_values)\\n # msg_id.send()\\n\\n template_id.sudo().write(\\n {'attachment_ids': [(6, 0, [each_attachment.id for each_attachment in attchment_list])]})\\n template_id.sudo().with_context(\\n # email_to=each_request.get('email'),\\n # event_list=event_list,\\n email_cc='enquiries@charterquest.co.za,accounts@charterquest.co.za,cqops@charterquest.co.za',\\n # prof_body=invoice_id.prof_body.name,\\n ).send_mail(order.id, force_send=True)\\n return request.render(\\\"cfo_snr_jnr.event_unsuccessful\\\", {'order': order})\",\n \"def claim_email(request):\\n email = request.POST.get('email', '')\\n email_user = User.objects.filter(email=email)\\n payload = {\\n 'res': 'failed'\\n }\\n if email_user.exists() and \\\\\\n not email_user[0].profile.send_mail:\\n request.user.profile.add_email(email)\\n payload['res'] = 'success'\\n\\n return payload\",\n \"def test_admin_approval_complete_email(self):\\n new_user = UserModel().objects.create_user(**self.user_info)\\n profile = self.registration_profile.objects.create_profile(new_user)\\n profile.send_admin_approve_complete_email(Site.objects.get_current())\\n self.assertEqual(len(mail.outbox), 1)\\n self.assertEqual(mail.outbox[0].to, [self.user_info['email']])\",\n \"def add_manualpost_email(request, submission_id=None, access_token=None):\\n\\n if request.method == 'POST':\\n try:\\n button_text = request.POST.get('submit', '')\\n if button_text == 'Cancel':\\n return redirect(\\\"submit/manual_post.html\\\")\\n \\n form = SubmissionEmailForm(request.POST)\\n if form.is_valid():\\n submission_pk = form.cleaned_data['submission_pk']\\n message = form.cleaned_data['message']\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n \\n if form.cleaned_data['direction'] == 'incoming':\\n msgtype = 'msgin'\\n else:\\n msgtype = 'msgout'\\n \\n submission, submission_email_event = (\\n add_submission_email(request=request,\\n remote_ip=request.META.get('REMOTE_ADDR', None),\\n name = form.draft_name,\\n rev=form.revision,\\n submission_pk = submission_pk,\\n message = message,\\n by = request.user.person,\\n msgtype = msgtype) )\\n \\n messages.success(request, 'Email added.')\\n \\n try:\\n draft = Document.objects.get(name=submission.name)\\n except Document.DoesNotExist:\\n # Assume this is revision 00 - we'll do this later\\n draft = None\\n \\n if (draft != None):\\n e = AddedMessageEvent(type=\\\"added_message\\\", doc=draft)\\n e.message = submission_email_event.submissionemailevent.message\\n e.msgtype = submission_email_event.submissionemailevent.msgtype\\n e.in_reply_to = submission_email_event.submissionemailevent.in_reply_to\\n e.by = request.user.person\\n e.desc = submission_email_event.desc\\n e.time = submission_email_event.time\\n e.save()\\n \\n return redirect(\\\"ietf.submit.views.manualpost\\\")\\n except ValidationError as e:\\n form = SubmissionEmailForm(request.POST)\\n form._errors = {}\\n form._errors[\\\"__all__\\\"] = form.error_class([\\\"There was a failure uploading your message. (%s)\\\" % e.message])\\n else:\\n initial = {\\n }\\n\\n if (submission_id != None):\\n submission = get_submission_or_404(submission_id, access_token)\\n initial['name'] = \\\"{}-{}\\\".format(submission.name, submission.rev)\\n initial['direction'] = 'incoming'\\n initial['submission_pk'] = submission.pk\\n else:\\n initial['direction'] = 'incoming'\\n \\n form = SubmissionEmailForm(initial=initial)\\n\\n return render(request, 'submit/add_submit_email.html',dict(form=form))\",\n \"def receive_email_view(request):\\n save_inbound_email(request.POST, request.FILES)\\n return HttpResponse(200)\",\n \"def validation_email_sent(request):\\n assert(settings.EMAIL_VALIDATION == True)\\n logging.debug('')\\n data = {\\n 'email': request.user.email,\\n 'change_email_url': reverse('user_changeemail'),\\n 'action_type': 'validate'\\n }\\n return render_to_response('authenticator/changeemail.html', RequestContext(request, data))\",\n \"def postprocess():\\n if ERRORS:\\n address = 'tamakoshihiroki@gmail.com'\\n body = '\\\\n\\\\n'.join( ERRORS )\\n msg = create_message( body, address )\\n send_mail( msg, address )\",\n \"def send_mail_when_failed(self, body):\\r\\n pass\",\n \"def test_send_email_on_invite(self):\\n\\n league = self.create_league()\\n\\n season = self.create_season(league)\\n team = self.create_team(season)\\n\\n player = self.create_player()\\n\\n send_user_email_on_join(player, team.id)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n\\n # if testing manually:\\n # import pathlib\\n # pathlib.Path(\\\"test_email.html\\\").write_text(last_sent.body)\",\n \"def send_email_to_assigned_user(recipients, lead_id, domain='demo.django-crm.io', protocol='http', source=''):\\n lead = Lead.objects.get(id=lead_id)\\n created_by = lead.created_by\\n blocked_domains = BlockedDomain.objects.values_list('domain', flat=True)\\n blocked_emails = BlockedEmail.objects.values_list('email', flat=True)\\n for user in recipients:\\n recipients_list = []\\n user = User.objects.filter(id=user, is_active=True).first()\\n if user:\\n if (user.email not in blocked_emails) and (user.email.split('@')[-1] not in blocked_domains):\\n recipients_list.append(user.email)\\n context = {}\\n context[\\\"url\\\"] = protocol + '://' + domain + \\\\\\n reverse('leads:view_lead', args=(lead.id,))\\n context[\\\"user\\\"] = user\\n context[\\\"lead\\\"] = lead\\n context[\\\"created_by\\\"] = created_by\\n context[\\\"source\\\"] = source\\n subject = 'Assigned a lead for you. '\\n html_content = render_to_string(\\n 'assigned_to/leads_assigned.html', context=context)\\n msg = EmailMessage(\\n subject,\\n html_content,\\n to=recipients_list\\n )\\n msg.content_subtype = \\\"html\\\"\\n msg.send()\",\n \"def form_valid(self, form):\\n self.object = form.save()\\n self.send_verify_email()\\n return super().form_valid(form)\",\n \"def save_response(self):\\r\\n self.q(css='input.save-button').first.click()\\r\\n EmptyPromise(\\r\\n lambda: 'save' in self.alert_message.lower(),\\r\\n \\\"Status message saved\\\"\\r\\n ).fulfill()\",\n \"def form_valid(self, form):\\n\\n ModelAccountVerification.objects.send_verification_email(form.user,\\n verification_type=\\n ModelAccountVerification.VERIFICATION_TYPES[1][0])\\n\\n messages.info(self.request, \\\"Email sent successfully to the {0}\\\".format(form.user.email))\\n\\n return HttpResponseRedirect(self.success_url)\",\n \"def execute(self):\\n return LOGGER.info(f\\\"{datetime.datetime.now()} - Sending EMail to the configured email list\\\")\",\n \"def test_post_external_attendance_successfull(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-03T19:00',\\n 'attendee_email': 'student@external.mx',\\n 'exit_datetime': '2021-08-03T20:00'}\\n response = self.client.post(\\n reverse('tracking:external-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def post(self):\\n return send_email(request.args)\",\n \"def save_model(self, request, obj, form, change):\\n notify.send(\\n request.user,\\n recipient=obj.reported_by,\\n verb='updated your report',\\n action_object=obj,\\n description=obj.notes,\\n target=obj\\n )\\n obj.save()\",\n \"def send_warning(self):\\n\\n # Check whether all the necessary parameters for SMS are present\\n if self.your_phone != '' and self.twilio_phone != '' and self.account_sid != '' and self.auth_token != '':\\n client = Client(self.account_sid, self.auth_token)\\n\\n try:\\n sms = client.messages.create(\\n body=\\\"\\\"\\\"Last will: It was at least 30 days since your last check in. \\n This is a reminder to check in in the next 24 hours.\\\"\\\"\\\",\\n from_=self.twilio_phone,\\n to=self.your_phone)\\n sms\\n print(\\\"\\\\nSMS sent\\\")\\n except Exception as e:\\n print(f\\\"An error occurred while trying to send the SMS. Error: {e}\\\")\\n\\n else:\\n print(\\\"\\\\nMissing SMS parameters. SMS not sent\\\")\\n\\n # Check whether all the necessary parameters for email are present\\n if self.sender_name != '' and self.recipient_email != '' and self.email != '' and self.email_pwd != '':\\n message = f\\\"\\\"\\\"It has been at least 30 days since you last checked in. \\nYou need to check in in the next 24 hours.\\\\n\\nOtherwise at {self.deadline} the email with the important info will be sent to the designated recipient.\\\\n\\nIn order to reset simply go to the working directory and run python3 last_will.py\\\"\\\"\\\"\\n\\n # send_email will return 0 if everything went ok, otherwise it will return an error message\\n status = send_email(self.sender_name, self.your_email,\\n self.email, self.email_pwd,\\n subject='Last will: Reminder to check in', unencrypted_message=message)\\n\\n if status != 0:\\n print(status)\\n exit(1)\\n else:\\n print(\\\"Email sent\\\\n\\\")\\n\\n print(f\\\"You have until {self.deadline} to check in. \\\"\\n f\\\"In order to do that simply go to the working directory and run ./last_will.sh\\\\n\\\")\\n else:\\n print(\\\"Missing email parameters. Email not sent.\\\\n\\\")\\n exit(1)\",\n \"def complete_allotment(request):\\n reg_no = request.user.username\\n user = get_object_or_404(User, username=reg_no)\\n sec_email = user.get_profile().secondary_email\\n options_chosen = get_chosen_options(user)\\n context = {'username': reg_no, 'email': sec_email, \\n 'options_chosen': options_chosen}\\n ##Sending mail with allotment details \\n admin = User.objects.get(pk=1) \\n from_email = admin.email\\n subject = \\\"JAM 2012 admissions\\\"\\n content = \\\"The following options were chosen by you \\\\n \\\\n\\\"\\n if options_chosen:\\n counter = 1\\n for option in options_chosen:\\n content += \\\"Preference Number: %s, Option Code: %s, Option Name: %s, Location: %s \\\\n\\\" %(counter, option.opt_code, option.opt_name, option.opt_location)\\n counter += 1 \\n \\n content += \\\"\\\\n \\\\n \\\\nPlease do not delete this email and keep it for reference purposes. \\\\n \\\\n \\\\n \\\\n Regards, \\\\n JAM Office, IIT Bombay\\\"\\n send_mail(subject, content, from_email, [sec_email])\\n admin_content = content\\n admin_content +=\\\"\\\\n\\\\n\\\\n#%s:\\\" % (reg_no)\\n counter = 1\\n for option in options_chosen:\\n admin_content += \\\"%s,%s:\\\" %(counter, option.opt_code) \\n counter += 1\\n admin_content +=\\\"#\\\"\\n admin_content += time.ctime() \\n mail_admins(subject, admin_content) \\n return render(request, 'allotter/complete.html', context)\",\n \"def send_reminder(self):\\n pass\",\n \"def create_response_and_send_mail(token, partner_id, email):\\n \\t #set url\\n\\t url = survey.public_url\\n \\t # url = urlparse.urlparse(url).path[1:] # dirty hack to avoid incorrect urls\\n\\t\\t template_id = self.pool.get('email.template').search(cr,uid,[('model','=','survey.survey')])\\n\\t\\t template = self.pool.get('email.template').browse(cr,uid,template_id)\\n\\t\\t body = template.body_html\\n\\t if token:\\n \\t url = url + '/' + token\\n\\n\\t # post the message\\n \\t values = {\\n \\t'model': None,\\n\\t 'res_id': None,\\n \\t 'subject': template.subject,\\n \\t'body': body.replace(\\\"__URL__\\\", url),\\n\\t 'body_html': body.replace(\\\"__URL__\\\", url),\\n \\t 'parent_id': None,\\n \\t'partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'notified_partner_ids': partner_id and [(4, partner_id)] or None,\\n\\t 'email_from': 'info@sba.org' or None,\\n \\t 'email_to': partner.email,\\n\\t\\t }\\n\\t mail_id = mail_mail_obj.create(cr, uid, values, context=context)\\n \\t mail_mail_obj.send(cr, uid, [mail_id], context=context)\",\n \"def send_reminder(self):\\n message_contents = \\\"This is a reminder that your event: \\\" + self.event_title + \\\" takes place on \\\" + self.event_date + \\\" in \\\" + self.event_location\\n subject = \\\"Event Reminder\\\"\\n attendees = self.gameplanuser_set.all()\\n for attendee in attendees:\\n remindermessage = Message.objects.create(sender=self.event_manager, recipient=attendee, contents=message_contents)\\n remindermessage.save()\",\n \"def _send_verify_email(request, preferences, db_entry, rnd_hash, new_entry):\\n\\n location = reverse(\\\"KursAnmeldung-verify_email\\\", kwargs={\\\"hash\\\":rnd_hash})\\n verify_link = request.build_absolute_uri(location)\\n\\n # FIXME: convert to users local time.\\n now = datetime.datetime.utcnow()\\n\\n email_context = {\\n \\\"verify_link\\\": verify_link,\\n \\\"db_entry\\\": db_entry,\\n \\\"now\\\": now,\\n }\\n\\n # Render the internal page\\n emailtext = render_to_string(\\\"kurs_anmeldung/verify_mailtext.txt\\\", email_context)\\n\\n # Get the preferences from the database:\\n raw_notify_list = preferences[\\\"notify\\\"]\\n notify_list = raw_notify_list.splitlines()\\n notify_list = [i.strip() for i in notify_list if i]\\n\\n email_kwargs = {\\n \\\"from_email\\\": preferences[\\\"from_email\\\"],\\n \\\"subject\\\": preferences[\\\"email_subject\\\"],\\n \\\"body\\\": emailtext,\\n \\\"to\\\": [db_entry.email],\\n \\\"bcc\\\": notify_list,\\n }\\n\\n if MAIL_DEBUG == True:\\n msg = u\\\"MAIL_DEBUG is on: No Email was sended!\\\"\\n request.page_msg(msg)\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n\\n request.page_msg(\\\"django.core.mail.EmailMessage kwargs:\\\")\\n request.page_msg(email_kwargs)\\n\\n request.page_msg(\\\"debug mail text:\\\")\\n request.page_msg(mark_safe(\\\"<pre>%s</pre>\\\" % emailtext))\\n return\\n\\n # We can't use django.core.mail.send_mail, because all members\\n # of the recipient list will see the others in the 'To' field.\\n # But we would like to notify the admins via 'Bcc' field.\\n\\n connection = SMTPConnection(fail_silently=False)\\n email = EmailMessage(**email_kwargs)\\n\\n try:\\n sended = email.send(fail_silently=False)\\n except Exception, err:\\n msg = \\\"Error sending mail: %s\\\" % err\\n LogEntry.objects.log_action(app_label=\\\"kurs_anmeldung\\\", action=\\\"error\\\",\\n message=msg\\n )\\n db_entry.log(request, msg)\\n db_entry.mail_sended = False\\n if settings.DEBUG or request.user.is_staff:\\n db_entry.save()\\n raise\\n else:\\n db_entry.mail_sended = sended\\n db_entry.log(request, \\\"mail sended: %s\\\" % sended)\",\n \"def confirm_email(self):\\n self.active = True\\n self.save()\",\n \"def save(self, register=False):\\n sha = sha1(self.email).hexdigest()\\n infos = self.to_dict()\\n infos[\\\"plan\\\"] = infos[\\\"plan\\\"][\\\"id\\\"] if infos[\\\"plan\\\"] else None\\n\\n if not redis.hmset(\\\"sl:account:{}\\\".format(sha), infos):\\n raise SleekException(\\\"Could not save current user.\\\", 401)\\n \\n if register:\\n try:\\n send_email.delay(\\n \\\"florent.esp@gmail.com\\\", \\\"Welcome to sleekstatus !\\\",\\n \\\"Welcome message\\\", [self.email]\\n )\\n except:\\n pass # Cannot send email\",\n \"def task_send_reminder_email():\\n send_reminder_email()\\n logger.info(\\\"Sent reminder email\\\")\",\n \"def account_activation_sent(request):\\r\\n\\treturn render(request, 'account_activation_sent.html')\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n reminders.Patient.objects.filter(\\n pk__in=[self.test_patient.pk, self.other_patient.pk]\\n ).update(next_visit=appt_date)\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n self.startRouter()\\n self.router.logger.setLevel(logging.DEBUG)\\n\\n # run email job\\n from afrims.apps.reminders.app import daily_email_callback\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\\n self.stopRouter()\",\n \"def save(self, *args, **kwargs):\\n data = self.cleaned_data #Gets the data from the form, stores it as a dict\\n allUsers = Bruker.get_all_dict(Bruker)\\n mottaker = allUsers[int(data['mottaker'])]\\n melding = Messages(content=data['content'], author=self.getUser(), receiver=mottaker)\\n melding.save()\",\n \"def save_appointment_details(request, calendar_id):\\n def schedule_mail(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_mail, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n def schedule_sms(reminder_date, appointment):\\n # Configure our scheduler for reminder\\n try:\\n trigger = DateTrigger(run_date=reminder_date)\\n scheduler.add_job(send_appointment_sms, args=[appointment], trigger=trigger)\\n except Exception as exp:\\n print(exp)\\n \\n start_time = request.GET['start_time'][:19]\\n end_time = request.GET['end_time'][:19]\\n \\n start_time = datetime.strptime(start_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n end_time=datetime.strptime(end_time, \\\"%Y-%m-%dT%H:%M:%S\\\")\\n \\n calendar_obj = Calendar.objects.get(pk=calendar_id)\\n # if this is a POST request we need to process the form data\\n if request.method == 'POST':\\n\\n # create a form instance and populate it with data from the request:\\n form = AppointmentForm(request.POST)\\n\\n # check whether it's valid and save it\\n if form.is_valid():\\n # Save appointment details\\n \\n mobilephone = form.data['mobilephone']\\n email = form.data['email']\\n first_name = form.data['first_name']\\n last_name = form.data['last_name']\\n notes = form.data['notes']\\n\\n appointment = Appointment(start_time=start_time, end_time=end_time, first_name=first_name, \\n last_name=last_name, email=email, mobilephone=mobilephone, notes=notes)\\n \\n appointment.calendar = calendar_obj\\n appointment.save()\\n\\n try:\\n send_appointment_mail(appointment) # send appointment details email\\n except Exception as exp:\\n print(exp)\\n \\n try:\\n send_appointment_sms(appointment) # send appointment details sms\\n except Exception as exp:\\n print(exp)\\n \\n # Calculate reminder schedule dates\\n reminder1 = start_time - timedelta(hours=2)\\n reminder2 = start_time - timedelta(hours=24)\\n reminder3 = start_time - timedelta(days=7)\\n\\n # Schedule mails\\n schedule_mail(reminder1, appointment)\\n schedule_mail(reminder2, appointment)\\n schedule_mail(reminder3, appointment)\\n \\n # Schedule sms\\n schedule_sms(reminder1, appointment)\\n schedule_sms(reminder2, appointment)\\n schedule_sms(reminder3, appointment)\\n \\n return redirect(reverse('appointment:complete_appointment', args=[calendar_id]))\\n \\n # if a GET (or any other method) we'll create a blank form\\n else:\\n form = AppointmentForm()\\n return render(request, 'appointment_form.html', {'form': form, 'start_time': start_time, 'end_time': end_time,\\n 'office_location': calendar_obj.office_location})\",\n \"def post(\\n self,\\n email,\\n company_name,\\n location,\\n job_profile,\\n salary,\\n username,\\n password,\\n security_question,\\n security_answer,\\n notes,\\n date_applied,\\n status,\\n):\",\n \"def post_save_order_receiver(sender, instance, created, *args, **kwargs):\\n sender_email = instance.cargo.sender.email\\n recepient_email = instance.cargo.recepient.email\\n\\n booking_agent = instance.cargo.booking_station.branch_agent.email\\n\\n if created:\\n instance._set_order_price()\\n instance._set_time_approximations()\\n price = instance.price\\n\\n subject = \\\"Order Finalized and ready to go.\\\"\\n message = f\\\"Your cargo has been booked and is ready for delivery. You will be notified whenever the status changes. It is currently {instance.get_status_display().title()}. It cost a total of ${price:.3f}. Your booking agent is {booking_agent}\\\"\\n\\n send_async_email(\\n subject=subject,\\n message=message,\\n sender=booking_agent,\\n recepients=[sender_email, recepient_email],\\n )\",\n \"def add_submission_email(request, remote_ip, name, rev, submission_pk, message, by, msgtype):\\n\\n #in_reply_to = form.cleaned_data['in_reply_to']\\n # create Message\\n parts = pyzmail.parse.get_mail_parts(message)\\n body=''\\n for part in parts:\\n if part.is_body == 'text/plain' and part.disposition == None:\\n payload, used_charset = pyzmail.decode_text(part.get_payload(), part.charset, None)\\n body = body + payload + '\\\\n'\\n\\n msg = submit_message_from_message(message, body, by)\\n\\n if (submission_pk != None):\\n # Must exist - we're adding a message to an existing submission\\n submission = Submission.objects.get(pk=submission_pk)\\n else:\\n # Must not exist\\n submissions = Submission.objects.filter(name=name,rev=rev).exclude(state_id='cancel')\\n if submissions.count() > 0:\\n raise ValidationError(\\\"Submission {} already exists\\\".format(name))\\n \\n # create Submission using the name\\n try:\\n submission = Submission.objects.create(\\n state_id=\\\"waiting-for-draft\\\",\\n remote_ip=remote_ip,\\n name=name,\\n rev=rev,\\n title=name,\\n note=\\\"\\\",\\n submission_date=datetime.date.today(),\\n replaces=\\\"\\\",\\n )\\n from ietf.submit.utils import create_submission_event, docevent_from_submission\\n desc = \\\"Submission created for rev {} in response to email\\\".format(rev)\\n create_submission_event(request, \\n submission,\\n desc)\\n docevent_from_submission(request,\\n submission,\\n desc)\\n except Exception as e:\\n log(\\\"Exception: %s\\\\n\\\" % e)\\n raise\\n\\n if msgtype == 'msgin':\\n rs = \\\"Received\\\"\\n else:\\n rs = \\\"Sent\\\"\\n\\n desc = \\\"{} message - manual post - {}-{}\\\".format(rs, name, rev)\\n submission_email_event = SubmissionEmailEvent.objects.create(\\n desc = desc,\\n submission = submission,\\n msgtype = msgtype,\\n by = by,\\n message = msg)\\n #in_reply_to = in_reply_to\\n\\n save_submission_email_attachments(submission_email_event, parts)\\n return submission, submission_email_event\",\n \"def handle_sent(self, instance):\\n if not instance.created_by:\\n return\\n\\n activity = Activity(\\n actor=instance.created_by,\\n verb=RestrictedMailSent,\\n object=instance,\\n time=instance.used,\\n extra_context={},\\n )\\n self.manager.add_activity(\\n activity, [instance.created_by.pk], [NotificationFeed]\\n )\\n\\n # Send notification\\n notification = RestrictedMailSentNotification(instance.created_by)\\n notification.notify()\",\n \"def handle_failure(self, instance, sender=None, reason=None):\\n send_email.delay(\\n to_email=sender,\\n context={\\\"reason\\\": reason},\\n subject=\\\"Kunne ikke sende ut begrenset epost\\\",\\n plain_template=\\\"restricted/email/process_failure.txt\\\",\\n html_template=\\\"restricted/email/process_failure.html\\\",\\n )\",\n \"def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):\\n print(\\\"member email\\\",mem_mail)\\n logger.info(\\\"in sending existing customer mail task\\\")\\n return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website)\",\n \"def handle(self, *args, **options):\\n\\n candidates_with_email = [candidate for candidate in Candidate.objects.all()\\n if candidate.contact_address and candidate.participating]\\n\\n\\n print 'sending e-mails'\\n conn = get_connection()\\n for c in candidates_with_email:\\n if c.should_send_reminder():\\n\\n print 'emailing', c\\n # store timestamp for reminder email so that they don't get another one for <REMINDER_TIME_PERIOD> days\\n c.last_reminder_sent = timezone.now()\\n c.save()\\n msg = make_email(c)\\n conn.send_messages([msg])\\n conn.close()\",\n \"def send(self, request, id, tribe_slug):\\n tribe = get_object_or_404(Tribe, slug=tribe_slug)\\n document = self.get_document(id, tribe_slug)\\n \\n form = self._set_user_email_address(request)\\n email = self._get_user_email_address(request)\\n if form or not email:\\n return render_to_response('django_dms/send.html', locals(), context_instance=RequestContext(request))\\n \\n # NB: Temporarily disabling actual email sending for development\\n #email_document(document, to=[email], subject='Document: %s' % document.title)\\n print \\\"Sending email to %s\\\" % email \\n \\n # Send a signal to let everyone know about this document interaction\\n document_interaction.send(sender=self, document=document, mode=\\\"sent\\\", request=request, recipient=email)\\n \\n return HttpResponseRedirect(reverse('%s_document_list' % self.name, args=(tribe_slug, ) ))\",\n \"def submit(self):\\r\\n try:\\r\\n response = self.send_request()\\r\\n if response.ok:\\r\\n self.submitted_at = datetime.now(pytz.UTC)\\r\\n self.status = \\\"submitted\\\"\\r\\n self.save()\\r\\n else:\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.error_msg = response.text\\r\\n self.save()\\r\\n except Exception as error:\\r\\n log.exception(error)\\r\\n self.status = \\\"must_retry\\\"\\r\\n self.save()\",\n \"def _send(self, email_message): \\n\\t\\tif not email_message.recipients(): \\n\\t\\t\\treturn False \\n\\t\\trecipients = map(self._sanitize, email_message.recipients()) \\n\\t\\tMsg = o.CreateItem(0)\\n \\t\\tMsg.To = recipients\\n\\t\\tMsg.Subject = 'subject'\\n\\t\\tMsg.Body = 'text'\\n\\t\\tself.connection.SaveChanges(0)\\n\\n\\t\\tMsg.Send()\\n\\t\\treturn True\",\n \"def set_email():\\n bid_fields = {'assertion':request.form['bid_assertion'],\\n 'audience':settings.DOMAIN}\\n headers = {'Content-type':'application/x-www-form-urlencoded'}\\n h.disable_ssl_certificate_validation=True\\n resp, content = h.request('https://browserid.org/verify',\\n 'POST',\\n body=urlencode(bid_fields),\\n headers=headers)\\n bid_data = json.loads(content)\\n if bid_data['status'] == 'okay' and bid_data['email']:\\n user = f.userFromBidOrNew(bid_data['email'])\\n session['user_id'] = user.id\\n session['user_email'] = user.email\\n\\n return redirect(url_for('main'))\",\n \"def wasSent(self):\\n\\t\\tself.sent = True\\n\\t\\tself.save()\",\n \"def action_send_email(self):\\n self.ensure_one()\\n if self.communication_channel != \\\"email\\\":\\n return False # Maybe we should raise an error here.\\n\\n if self.state in (\\\"sent\\\", \\\"done\\\"):\\n raise ValidationError(_(\\\"This communication is already sent.\\\"))\\n\\n lines_2be_processed = self.credit_control_line_ids.filtered(\\n lambda line: line.state != \\\"sent\\\"\\n )\\n\\n if not lines_2be_processed:\\n raise ValidationError(_(\\\"There is no draft lines to send.\\\"))\\n\\n self = self.with_context(lang=self.partner_id.lang)\\n\\n partner = self.partner_id\\n mail_template = self.policy_level_id.email_template_id\\n # Send the email\\n partner.with_context(credit_control_mail=True).message_post_with_template(\\n template_id=mail_template.id,\\n model=self._name,\\n res_id=self.id,\\n )\\n # Set the state of the credit control lines to \\\"queued\\\"\\n lines_2be_processed.write({\\\"state\\\": \\\"sent\\\"})\\n self.state = \\\"sent\\\"\",\n \"def test_activate_form(self, mock_sendmail):\\r\\n res = self.testapp.post('/api/v1/suspend',\\r\\n params={'email': u'testing@dummy.com'},\\r\\n status=200)\\r\\n\\r\\n success = json.loads(res.body)\\r\\n self.assertTrue(\\r\\n 'message' in success,\\r\\n \\\"Should be successful with admin email address: \\\" + str(res))\\r\\n self.assertTrue(mock_sendmail.called)\",\n \"def test_post_internal_attendance_success(self):\\n building_name = 'Building 1'\\n room_name = 'Room 1'\\n\\n building = Building.objects.create(\\n name=building_name\\n )\\n room = Room.objects.create(\\n name=room_name,\\n building=building\\n )\\n post_data = {'room': room.id,\\n 'entry_datetime': '2021-08-04T10:00',\\n 'attendee_email': 'student@anahuac.mx',\\n 'exit_datetime': '2021-08-04T11:30'}\\n response = self.client.post(\\n reverse('tracking:internal-register'),\\n post_data\\n )\\n\\n self.assertEqual(response.status_code, HTTPStatus.OK)\\n messages = list(response.context['messages'])\\n success_message = str(messages[0])\\n self.assertEqual(success_message, 'Registro realizado exitosamente')\",\n \"def save(self, *args, **kwargs):\\n self.expire_date = self.sent_date + timedelta(days=self.period)\\n super(Referral, self).save(*args, **kwargs)\",\n \"def post(self, request):\\n\\n try:\\n eventoid = request.POST.get('id', '')\\n correo = request.POST.get('correo', '')\\n AsigStaff.objects.create(id_Evento = eventoid, email_staff = correo)\\n print(\\\"Exito en la asignación de staff\\\")\\n except:\\n print(\\\"Error en la asignacion de staff\\\")\\n\\n \\n return render(request, self.template, self.context)\\n #return render(request, self.template, self.context)\",\n \"def save(self, *args, **kwargs):\\n create = self.id is None\\n if create:\\n self.username = u'lc_' + self.city.lower()\\n # Give the user a randomly generated password\\n password = generate_random_string()\\n self.password = password\\n # Compose the email to send to the LC\\n context = {\\n 'lc': self,\\n 'password': password,\\n }\\n subject = render_to_string(\\n 'registration/lc_register_email_subject.txt')\\n subject = subject.strip()\\n message = render_to_string(\\n 'registration/lc_register_email_text.txt',\\n context)\\n send_mail(subject,\\n message,\\n settings.DEFAULT_FROM_EMAIL,\\n [self.email])\\n # Delegate upward to the parent classes\\n super(LocalCommittee, self).save(*args, **kwargs)\",\n \"def contact():\\n if request.method == 'POST':\\n send_email()\\n return \\\"\\\"\",\n \"def test_sending_mail(self):\\n\\n appt_date = datetime.date.today() + datetime.timedelta(days=7) # Default for email\\n confirmed = self.create_confirmed_notification(self.test_patient, appt_date)\\n\\n # run email job\\n daily_email_callback(self.router)\\n\\n self.assertEqual(len(mail.outbox), 1)\\n message = mail.outbox[0]\\n self.assertTrue(self.test_contact.email in message.to)\",\n \"def save_model(self, request, obj, form, change):\\n admin.ModelAdmin.save_model(self, request, obj, form, change)\\n\\n # pydevd.settrace('localhost', port=8001, stdoutToServer=True, stderrToServer=True)\\n post_title = obj.title\\n post_url = 'http://www.aptitudeworld.com.au/blog'\\n\\n # send an email if we are adding a new post\\n if not change:\\n # send an email to all users in the system to notify them that a new post is available\\n email_title = 'Aptitude World ANZ - A new post is available'\\n email_from = 'info@aptitudeworld.com.au'\\n\\n try:\\n # get all user emails in the database\\n users = apps.users.models.User.objects.all()\\n confirmation_email_txt_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.txt', 'r')\\n confirmation_email_html_template = open(\\n os.getcwd() + '/apps/landing/templates/landing/email/new_post_email.html', 'r')\\n for u in users:\\n email_to = u.email_address\\n full_name = u.first_name + \\\" \\\" + u.last_name\\n\\n email_txt = confirmation_email_txt_template.read() % (full_name, post_title, post_url)\\n email_html = confirmation_email_html_template.read() % (full_name, post_title, post_url)\\n\\n email_message = EmailMultiAlternatives(email_title, email_txt, email_from, [email_to])\\n email_message.attach_alternative(email_html, 'text/html')\\n email_message.send()\\n except BadHeaderError:\\n pass\",\n \"def send_email():\\n send_mail(\\\"You've got some problem.\\\", 'REPAIR IT', 'dimazarj2009@rambler.ru',\\n ['dimazarj2009@rambler.ru'], fail_silently=False,)\",\n \"def create_sent_email(self, *args, **kwargs):\\n receiver = kwargs['receiver']\\n sender = kwargs['sender']\\n user = kwargs['user']\\n body = kwargs['body']\\n subject = kwargs['subject']\\n tracker_id = kwargs['tracker_id']\\n if receiver and sender and subject and body:\\n sent_email = SentEmail()\\n if tracker_id:\\n sent_email.tracker_id = tracker_id\\n sent_email.read_status = 'Tracking'\\n else:\\n sent_email.read_status = 'Not Tracking'\\n sent_email.receiver = receiver\\n sent_email.subject = subject\\n sent_email.sender = sender\\n sent_email.status = 'sent'\\n sent_email.user = user\\n sent_email.body = body\\n sent_email.save()\\n return sent_email\\n else:\\n return False\",\n \"def send_welcome_mail(backend, details, response, user, is_new=False, *args, **kwargs):\\n\\n if is_new:\\n context = Context({'user': user, 'ga_campaign_params' : 'utm_source=unishared&utm_content=v1&utm_medium=e-mail&utm_campaign=welcome_mail'})\\n\\n email_task.apply_async([u'Welcome on UniShared!', context, 'welcome_mail', [user.email]], eta= datetime.utcnow() + timedelta(hours=1))\",\n \"def sendEmail(householdID):\\n contactID = mdb.getContact(householdID)\\n sqlq = \\\"\\\"\\\"\\n SELECT Name, Surname, Address1, Address2, Town, Postcode, email, status\\n FROM Contact\\n WHERE idContact = '{}';\\n \\\"\\\"\\\".format(contactID)\\n result = mdb.getSQL(sqlq)[0]\\n\\n # thisName = (\\\"%s %s\\\" % (result['Name'], result['Surname']))\\n thisName = (\\\"%s\\\" % (result['Name']))\\n thisAddress = (\\\"%s</br>%s</br>%s %s\\\" % (result['Address1'], result['Address2'], result['Town'], result['Postcode']))\\n thisAddress = thisAddress.replace(\\\"None </br>\\\", \\\"\\\")\\n thisEmail = (\\\"%s\\\" % (result['email']))\\n thisStatus = (\\\"%s\\\" % (result['status']))\\n thisAddress = thisAddress.replace(\\\"None</br>\\\", \\\"\\\")\\n participantCount = (\\\"%s\\\" % mdb.getParticipantCount(str(householdID)))\\n # prepare the custom email\\n\\n thisPath = os.path.dirname(os.path.abspath(__file__))\\n if (thisStatus == 'de'):\\n # DOESN'T happen yet - de is excluded from query for now\\n # emailPath = os.path.join(thisPath, \\\"emails/email_confirm_de.html\\\")\\n locale.setlocale(locale.LC_ALL, 'de_DE.utf8')\\n else:\\n emailPath = os.path.join(thisPath, \\\"emails/email_automated_date.html\\\")\\n\\n templateFile = open(emailPath, \\\"r\\\")\\n templateText = templateFile.read()\\n templateFile.close()\\n templateText = templateText.replace(\\\"[householdID]\\\", householdID)\\n templateText = templateText.replace(\\\"[contactID]\\\", contactID)\\n templateText = templateText.replace(\\\"[name]\\\", thisName)\\n templateText = templateText.replace(\\\"[address]\\\", thisAddress)\\n templateText = templateText.replace(\\\"[securityCode]\\\", mdb.getSecurityCode(householdID))\\n templateText = templateText.replace(\\\"[participantCount]\\\", participantCount)\\n\\n # Subject\\n subjectLine = templateText.splitlines()[0]\\n templateText = templateText[templateText.find('\\\\n') + 1:] # find line break and return all from there - i.e. remove first line\\n \\n # email file\\n emailFilePath = os.path.join(thisPath, \\\"tempEmail.htmail\\\")\\n\\n emailFile = open(emailFilePath, \\\"w+\\\")\\n emailFile.write(templateText)\\n emailFile.close()\\n\\n # call('mutt -e \\\"set content_type=text/html\\\" -s \\\"[TESTING]' + subjectLine + '\\\" philipp.grunewald@ouce.ox.ac.uk < ' + emailFilePath, shell=True)\\n call('mutt -e \\\"set content_type=text/html\\\" -s \\\"' + subjectLine + '\\\" ' + thisEmail + ' -b meter@energy.ox.ac.uk < ' + emailFilePath, shell=True)\",\n \"def save(self,owner=None):\\n\\n q = Queue.objects.get(id=int(self.cleaned_data['queue']))\\n\\n t = Ticket(\\n title = self.cleaned_data['title'],\\n owner=owner,\\n submitter_email = self.cleaned_data['submitter_email'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n account = owner.account\\n )\\n\\n t.save()\\n\\n f = FollowUp(\\n ticket = t,\\n title = _('Ticket Opened Via Web'),\\n date = datetime.now(),\\n public = True,\\n comment = self.cleaned_data['body'],\\n account = owner.account\\n )\\n\\n f.save()\\n\\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n }\\n\\n messages_sent_to = []\\n\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self, user):\\n\\n q = self.cleaned_data['queue']\\n\\n t = Ticket( title = self.cleaned_data['title'],\\n submitter_email = self.cleaned_data['submitter_email'],\\n account = self.cleaned_data['account'],\\n created = datetime.now(),\\n status = Ticket.OPEN_STATUS,\\n queue = q,\\n description = self.cleaned_data['body'],\\n priority = self.cleaned_data['priority'],\\n owner = self.cleaned_data['owner']\\n )\\n\\n if HAS_TAG_SUPPORT:\\n t.tags = self.cleaned_data['tags']\\n\\n if self.cleaned_data['assigned_to']:\\n try:\\n t.assigned_to = self.cleaned_data['assigned_to']\\n except User.DoesNotExist:\\n t.assigned_to = None\\n t.save()\\n\\n f = FollowUp( ticket = t,\\n title = _('Ticket Opened'),\\n date = datetime.now(),\\n public = False,\\n comment = self.cleaned_data['body'],\\n systemuser = user.account,\\n )\\n if self.cleaned_data['assigned_to']:\\n f.title = _('Ticket Opened & Assigned to %(name)s') % {\\n 'name': t.get_assigned_to\\n }\\n\\n f.save()\\n \\n files = []\\n if self.cleaned_data['attachment']:\\n import mimetypes\\n file = self.cleaned_data['attachment']\\n filename = file.name.replace(' ', '_')\\n a = Attachment(\\n followup=f,\\n filename=filename,\\n mime_type=mimetypes.guess_type(filename)[0] or 'application/octet-stream',\\n size=file.size,\\n )\\n a.file.save(file.name, file, save=False)\\n a.save()\\n \\n if file.size < getattr(settings, 'MAX_EMAIL_ATTACHMENT_SIZE', 512000):\\n # Only files smaller than 512kb (or as defined in \\n # settings.MAX_EMAIL_ATTACHMENT_SIZE) are sent via email.\\n files.append(a.file.path)\\n\\n context = {\\n 'ticket': t,\\n 'queue': q,\\n 'comment': f.comment,\\n }\\n \\n messages_sent_to = []\\n\\n if t.submitter_email:\\n send_templated_mail(\\n 'newticket_owner',\\n context,\\n recipients=t.submitter_email,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(t.submitter_email)\\n\\n #FIX USERSETTINGS\\n #=======================================================================\\n # if t.assigned_to and t.assigned_to != user and getattr(t.assigned_to.usersettings.settings, 'email_on_ticket_assign', False) and t.assigned_to.email and t.assigned_to.email not in messages_sent_to:\\n # send_templated_mail(\\n # 'assigned_to',\\n # context,\\n # recipients=t.assigned_to.email,\\n # sender=q.from_address,\\n # fail_silently=True,\\n # files=files,\\n # )\\n # messages_sent_to.append(t.assigned_to.email)\\n #=======================================================================\\n\\n if q.new_ticket_cc and q.new_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.new_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n messages_sent_to.append(q.new_ticket_cc)\\n\\n if q.updated_ticket_cc and q.updated_ticket_cc != q.new_ticket_cc and q.updated_ticket_cc not in messages_sent_to:\\n send_templated_mail(\\n 'newticket_cc',\\n context,\\n recipients=q.updated_ticket_cc,\\n sender=q.from_address,\\n fail_silently=True,\\n files=files,\\n )\\n\\n return t\",\n \"def save(self):\\n email = self.validated_data['email']\\n name = self.validated_data['name']\\n password = self.validated_data['password']\\n\\n email_query = models.Email.objects.filter(address=email)\\n if email_query.exists():\\n email_instance = email_query.get()\\n\\n # If the email is already verified, we send a duplicate\\n # notification and exit.\\n if email_instance.is_verified:\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"is already verified.\\\",\\n email_instance,\\n )\\n email_instance.send_duplicate_notification()\\n\\n return\\n\\n # If the email is not verified, we send a new verification\\n # token to the address.\\n logger.info(\\n \\\"Not registering a new user because the email address %r \\\"\\n \\\"already exists. Sending a new verification token instead.\\\"\\n )\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\\n\\n return\\n\\n # The email doesn't exist, so we create a new user and email,\\n # then send a verification token to the email.\\n user = models.User.objects.create_user(name, password)\\n email_instance = models.Email.objects.create(address=email, user=user)\\n\\n # The user's primary email is their only email. This is the only\\n # time the primary email can be unverified.\\n user.primary_email = email_instance\\n user.save()\\n\\n logger.info(\\n \\\"Registered new user %r with email address %r\\\",\\n user,\\n email_instance,\\n )\\n\\n verification = models.EmailVerification.objects.create(\\n email=email_instance,\\n )\\n verification.send_email()\",\n \"def send_email(self, message):\\n pass\",\n \"def send_mail(self, subject):\\r\\n pass\",\n \"def post():\\n contactus_json = request.get_json()\\n\\n try:\\n dict_data = ContactUsSchema().load(contactus_json)\\n dict_data['description'] = escape(dict_data['description'])\\n EmailService.save_and_send(EmailType.CONTACT_US, dict_data)\\n response, status = 'Received successfully', HTTPStatus.OK\\n except ValidationError as project_err:\\n response, status = {'systemErrors': project_err.messages}, \\\\\\n HTTPStatus.BAD_REQUEST\\n return response, status\",\n \"async def handle_email(self, email):\\n\\t\\tif 'logdir' in self.log_settings:\\n\\t\\t\\tfilename = 'email_%s_%s.eml' % (datetime.datetime.utcnow().isoformat(), str(uuid.uuid4()))\\n\\t\\t\\twith open(str(Path(self.log_settings['logdir'], 'emails', filename).resolve()), 'wb') as f:\\n\\t\\t\\t\\tf.write(email.email.as_bytes())\\n\\n\\t\\tawait self.log('You got mail!')\",\n \"def set_email_sent(self, instance_id, creator, type):\\n table = self.type_map[type]\\n item = {\\n \\\"InstanceID\\\": instance_id,\\n \\\"Creator\\\": creator,\\n \\\"Scheduled For Deletion\\\": True,\\n \\\"EmailSent\\\": True\\n }\\n return table.put_item(item)\",\n \"def order_success(request, job_id, token):\\n if request.user.is_active:\\n job_obj = Job.objects.get(job_id= job_id)\\n if job_obj.employeer == request.user:\\n if verifiers.payment_token_is_valid(token, job_obj):\\n if job_obj.accepted == False:\\n job_obj.status = Status.objects.create(status_name=\\\"PN\\\", comment=\\\"Awaiting payment confirmation from Stripe, this might take a few minutes.\\\")\\n job_obj.accepted = True\\n job_obj.save()\\n \\n messages.add_message(request, messages.INFO, 'Your order has been created.')\\n return redirect('/') \\n else:\\n #someone tries to manually enter the URL with wrong payment_token\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n #the person making the request is not the same as the person who created the job\\n messages.add_message(request, messages.INFO, 'Something went wrong, try again later.') \\n return redirect(\\\"/\\\") \\n else:\\n return redirect(\\\"/login\\\")\",\n \"def order_created(order_id):\\n order = Order.objects.get(id=order_id)\\n subject = f'Order nr. {order.id}'\\n message = f'Dear {order.first_name},\\\\n\\\\n' \\\\\\n f'You have successfully placed an order.' \\\\\\n f'Your order ID is {order.id}.'\\n mail_sent = send_mail(subject,\\n message,\\n 'info@tishman.com.ng',\\n [order.email])\\n return mail_sent\",\n \"def send_email_alert_slack(email_pk: str):\\n logging.info(\\\"Sending alert for Email<%s>\\\", email_pk)\\n email = Email.objects.get(pk=email_pk)\\n issue = email.issue\\n assert issue, f\\\"Email<{email_pk}> does not have an associated Issue\\\"\\n paralegal = issue.paralegal\\n alert_sent = False\\n case_email_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-email-list\\\", args=(str(issue.pk),)\\n )\\n case_url = settings.CLERK_BASE_URL + reverse(\\n \\\"case-detail-view\\\", args=(str(issue.pk),)\\n )\\n msg = (\\n \\\"*New Email Notification*\\\\n\\\"\\n f\\\"A new email has been received for case <{case_url}|{issue.fileref}> .\\\\n\\\"\\n f\\\"You can view this case's emails here: <{case_email_url}|here>.\\\\n\\\"\\n )\\n if paralegal and issue.is_open:\\n # Send alert directly to paralegal for open issues.\\n alert_email = settings.SLACK_EMAIL_ALERT_OVERRIDE or paralegal.email\\n logging.info(\\\"Looking up %s in Slack\\\", alert_email)\\n slack_user = get_slack_user_by_email(alert_email)\\n if slack_user:\\n logging.info(\\\"Notifying %s of Email<%s> via Slack\\\", alert_email, email_pk)\\n send_slack_direct_message(msg, slack_user[\\\"id\\\"])\\n alert_sent = True\\n\\n if not alert_sent:\\n # In all other cases send to alerts channel.\\n logging.info(\\n \\\"Could not find someone to DM in Slack, sending generic alert for Email<%s>\\\",\\n email_pk,\\n )\\n send_slack_message(settings.SLACK_MESSAGE.CLIENT_INTAKE, msg)\\n\\n Email.objects.filter(pk=email_pk).update(is_alert_sent=True)\\n logging.info(\\\"Alert sent sucessfully for Email<%s>\\\", email_pk)\",\n \"def send_sbemail(email_addy, city):\\n print(\\\"Single email send attempt: %s, %s\\\" % (email_addy, city))\\n given_pair, weather = subject_phrase_picker(city)\\n mailtext = '%s %s in %s!' % ( given_pair['phrasing'], weather, city)\\n print(\\\"Going to try a mailgun post to: %s with subject %s\\\" % (email_addy, given_pair['subject']))\\n response = requests.post(\\n MG_API_URL,\\n auth=(\\\"api\\\", MG_API_KEY),\\n data={'from': 'Nick Cage <nickcage@%s>' % (MG_DOMAIN),\\n 'to': [email_addy],\\n 'subject': given_pair['subject'],\\n 'text': mailtext\\n })\\n print(\\\"status: %s | %s and response: %s\\\" % (response.status_code, response.reason, response.text))\\n print(\\\"Attempted send %s to: %s!\\\" % (mailtext, email_addy))\\n return \\\"Sent %s to: %s!\\\" % (mailtext, email_addy)\",\n \"def on_update(self):\\n\\t\\tfor email_account in frappe.get_all(\\\"Email Account\\\", filters={\\\"domain\\\": self.name}):\\n\\t\\t\\ttry:\\n\\t\\t\\t\\temail_account = frappe.get_doc(\\\"Email Account\\\", email_account.name)\\n\\t\\t\\t\\tfor attr in [\\\"email_server\\\", \\\"use_imap\\\", \\\"use_ssl\\\", \\\"use_tls\\\", \\\"attachment_limit\\\", \\\"smtp_server\\\", \\\"smtp_port\\\", \\\"use_ssl_for_outgoing\\\", \\\"append_emails_to_sent_folder\\\", \\\"incoming_port\\\"]:\\n\\t\\t\\t\\t\\temail_account.set(attr, self.get(attr, default=0))\\n\\t\\t\\t\\temail_account.save()\\n\\n\\t\\t\\texcept Exception as e:\\n\\t\\t\\t\\tfrappe.msgprint(_(\\\"Error has occurred in {0}\\\").format(email_account.name), raise_exception=e.__class__)\",\n \"def put(self, campaign_id):\\n campaign = Campaign.query.filter_by(public_id=campaign_id).first()\\n if campaign.candidates.count() > 0:\\n campaign.launch_task('generate_mailer_file')\\n return {'success': True, 'message': 'Initiated generate mailer file'}, 200\\n else:\\n api.abort(404, message='Campaign does not exist', success=False)\",\n \"def post_save_moderation(self, sender, comment, request, **kwargs):\\n model = comment.content_type.model_class()\\n if model not in self._registry:\\n return\\n self._registry[model].email(comment, comment.content_object, request)\",\n \"def send_email(email_dict, appointment_id):\\n event_identifier = g_cal.send_invite_through_gcal(email_dict)\\n models.Appointments.objects.filter(id=appointment_id).update(event_identifier=event_identifier)\",\n \"def delegate_last_day():\\n\\n regs = Registration.objects.all()\\n\\n template = 'notifications/last_day_mail.html'\\n\\n for reg in regs:\\n subject = 'SciPy.in 2011: Schedule and other details'\\n message = loader.render_to_string(\\n template, dictionary={'name': reg.registrant.username})\\n\\n reg.registrant.email_user(subject=subject, message=message,\\n from_email='madhusudancs@gmail.com')\",\n \"def send_mail(self, address, title, message):\\n pass\",\n \"def post(self):\\n # se captura y se parsea a json el body del request recibido por el\\n # webhook\\n request_body = json.loads(self.request.body)\\n\\n for body in request_body:\\n \\\"\\\"\\\" Evaluar el tipo de evento ya que trae campos diferentes \\\"\\\"\\\"\\n logging.info(request_body)\\n\\n event = str(body['event'])\\n correo = str(body['email'])\\n numero_folio = str(body['numero_folio'])\\n tipo_dte = str(body['tipo_dte'])\\n\\n logging.info(event)\\n\\n if event and correo and numero_folio and tipo_dte:\\n\\n if event == 'processed':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.processed_event = event\\n email_model.processed_sg_event_id = body['sg_event_id']\\n email_model.processed_sg_message_id = body['sg_message_id']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.processed_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.processed_event = event\\n e.processed_sg_event_id = body['sg_event_id']\\n e.processed_sg_message_id = body['sg_message_id']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'delivered':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.delivered_event = event\\n email_model.delivered_sg_event_id = body['sg_event_id']\\n email_model.delivered_sg_message_id = body['sg_message_id']\\n email_model.delivered_response = body['response']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.delivered_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.delivered_event = event\\n e.delivered_sg_event_id = body['sg_event_id']\\n e.delivered_sg_message_id = body['sg_message_id']\\n e.delivered_response = body['response']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'open':\\n model = EmailModel()\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n if email_model.opened_first_date == None:\\n email_model.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.opened_event = event\\n email_model.opened_ip = body['ip']\\n email_model.opened_user_agent = body['useragent']\\n email_model.opened_sg_event_id = body['sg_event_id']\\n email_model.opened_sg_message_id = body['sg_message_id']\\n model.email_add_count(email_model)\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n if e.opened_first_date == None:\\n e.opened_first_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_last_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.opened_event = event\\n e.opened_ip = body['ip']\\n e.opened_user_agent = body['useragent']\\n e.opened_sg_event_id = body['sg_event_id']\\n e.opened_sg_message_id = body['sg_message_id']\\n e.email_add_count(e)\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'dropped':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.smtp_id = body['smtp-id']\\n email_model.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.dropped_sg_event_id = body['sg_event_id']\\n email_model.dropped_sg_message_id = body['sg_message_id']\\n email_model.dropped_reason = body['reason']\\n email_model.dropped_event = event\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.smtp_id = body['smtp-id']\\n e.dropped_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.dropped_sg_event_id = body['sg_event_id']\\n e.dropped_sg_message_id = body['sg_message_id']\\n e.dropped_reason = body['reason']\\n e.dropped_event = event\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'bounce':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.bounce_event = event\\n email_model.bounce_sg_event_id = body['sg_event_id']\\n email_model.bounce_sg_message_id = body['sg_message_id']\\n email_model.bounce_reason = body['reason']\\n email_model.bounce_status = body['status']\\n email_model.bounce_type = body['type']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.bounce_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.bounce_event = event\\n e.bounce_sg_event_id = body['sg_event_id']\\n e.bounce_sg_message_id = body['sg_message_id']\\n e.bounce_reason = str(body['reason']).decode(\\\"utf-8\\\")\\n e.bounce_status = body['status']\\n e.bounce_type = body['type']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n\\n elif event == 'unsubscribe':\\n email_model = EmailModel.search_email(correo, numero_folio, tipo_dte)\\n if not email_model == None:\\n email_model.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n email_model.unsubscribe_uid = body['uid']\\n email_model.unsubscribe_purchase = body['purchase']\\n email_model.unsubscribe_id = body['id']\\n email_model.unsubscribe_event = body['event']\\n email_model.correo = str(body['email'])\\n email_model.numero_folio = str(body['numero_folio'])\\n email_model.tipo_dte = str(body['tipo_dte'])\\n email_model.put()\\n else:\\n e = EmailModel()\\n e.unsubscribe_date = datetime.datetime.fromtimestamp(body['timestamp'])\\n e.unsubscribe_uid = body['uid']\\n e.unsubscribe_purchase = body['purchase']\\n e.unsubscribe_id = body['id']\\n e.unsubscribe_event = body['event']\\n e.correo = str(body['email'])\\n e.numero_folio = str(body['numero_folio'])\\n e.tipo_dte = str(body['tipo_dte'])\\n e.put()\\n else:\\n logging.info('body con campos vacios')\",\n \"def mail_sent():\\n\\n url = settings.SITE_URL + '\\\\charts'\\n subject = 'Анализ запрошенного ресурса'\\n message = 'Графики популярного часа дня и дня недели {}'.format(url)\\n mail_sent = send_mail(subject,\\n message,\\n 'admin@myshop.com',\\n ['user@mail.ru,'])\\n print(message)\\n return mail_sent\",\n \"def emailNote(self, authenticationToken, parameters):\\r\\n pass\",\n \"def send_new_email(user):\\n token = user.get_token()\\n message = Message(\\n 'Verify Your New Email',\\n sender='storcwebsite@gmail.com',\\n recipients=[user.temp_email])\\n message.body = f\\\"The email address associated with your Storc \\\" \\\\\\n f\\\"account has changed.\\\\n\\\\nTo verify your new email address, \\\" \\\\\\n f\\\"please click the link below:\\\\n\\\\n\\\" \\\\\\n f\\\"{url_for('users.new_email', token=token, _external=True)}\\\"\\n mail.send(message)\",\n \"def perform_create(self, serializer):\\n user = serializer.save()\\n signals.user_registered.send(\\n sender=self.__class__, user=user, request=self.request\\n )\\n\\n context = get_email_context(user)\\n to = [get_user_email(user)]\\n if djconf.SEND_ACTIVATION_EMAIL:\\n djconf.EMAIL.activation(self.request, context).send(to)\\n elif djconf.SEND_CONFIRMATION_EMAIL:\\n djconf.EMAIL.confirmation(self.request, context).send(to)\",\n \"def report_save(message, bot):\\n text = message.text\\n reports = parse_report(text)\\n author_id = message.from_user.id\\n author_username = message.from_user.username\\n to_save = []\\n for report in reports:\\n obj_to_save = {\\n 'author': author_id,\\n 'author_name': author_username,\\n 'text': report,\\n }\\n to_save.append(obj_to_save)\\n try:\\n request = requests.post(\\n url='http://127.0.0.1:8000/api/v1/report/',\\n headers={\\n 'Authorization': f'Bearer {jwt_token}',\\n },\\n json=to_save,\\n )\\n bot.send_message(\\n message.from_user.id,\\n request.status_code,\\n )\\n except requests.exceptions.ConnectionError:\\n bot.send_message(\\n message.from_user.id,\\n 'Ошибка соединения с сервером',\\n )\",\n \"def send(self):\\n ReferralActivity.objects.create(\\n actor=self.user,\\n verb=ReferralActivityVerb.CREATED,\\n referral=self,\\n )\\n # Confirm the referral has been sent to the requester by email\\n Mailer.send_referral_saved(self)\\n # Send this email to all owners of the unit(s) (admins are not supposed to receive\\n # email notifications)\\n for unit in self.units.all():\\n contacts = unit.members.filter(\\n unitmembership__role=UnitMembershipRole.OWNER\\n )\\n for contact in contacts:\\n Mailer.send_referral_received(self, contact=contact, unit=unit)\",\n \"def execute_automatic_email(self, request, pk=None):\\n try:\\n retreat = Retreat.objects.get(pk=pk)\\n except Exception:\\n response_data = {\\n 'detail': \\\"Retreat not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n try:\\n email = AutomaticEmail.objects.get(\\n id=int(request.GET.get('email'))\\n )\\n except Exception:\\n response_data = {\\n 'detail': \\\"AutomaticEmail not found\\\"\\n }\\n return Response(response_data, status=status.HTTP_400_BAD_REQUEST)\\n\\n # Notify a user for every reserved seat\\n emails = []\\n for reservation in retreat.reservations.filter(is_active=True):\\n if reservation.automatic_email_logs.filter(email=email):\\n pass\\n else:\\n send_automatic_email(reservation.user, retreat, email)\\n AutomaticEmailLog.objects.create(\\n reservation=reservation,\\n email=email\\n )\\n emails.append(reservation.user.email)\\n\\n response_data = {\\n 'stop': True,\\n 'emails': emails\\n }\\n return Response(response_data, status=status.HTTP_200_OK)\",\n \"def save(self):\\n owner = self.context[\\\"request\\\"].user\\n recipient = self._recipient_email_inst.user\\n\\n logger.info(\\n \\\"Transferring subscription %r from user %r to user %r\\\",\\n owner.know_me_subscription,\\n owner.pk,\\n recipient.pk,\\n )\\n\\n with transaction.atomic():\\n models.Subscription.objects.filter(user=recipient).delete()\\n models.Subscription.objects.filter(user=owner).update(\\n user=recipient\\n )\",\n \"def send_message(self):\\r\\n return \\\"success\\\"\",\n \"def alert_for_pending_mails_1(request):\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>Beginning of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tThread(target=alert_for_pending_mails_1_worker).start()\\n\\n\\tprint(\\\">>>>>>>>>>>>>>>>>>>>>End of alert_for_pending_mails_1<<<<<<<<<<<<<<<<<<<<\\\")\\n\\n\\tresponse = {}\\n\\n\\tresponse[\\\"info_to_contact\\\"] = \\\"Ok\\\"\\n\\n\\treturn response\",\n \"def save(self, domain_override=None,\\n subject_template_name='registration/password_reset_subject.txt',\\n email_template_name='registration/password_reset_email.html',\\n use_https=False, token_generator=default_token_generator,\\n from_email=None, request=None,\\n html_email_template_name=None):\\n email = self.cleaned_data[\\\"email\\\"]\\n User = get_user_model()\\n active_users = User.objects.filter(email__iexact=email, is_active=True)\\n for user in active_users:\\n subject = _('Flisol - Restore your password')\\n # send_email(\\n # subject,\\n # [user.email],\\n # email_template_name,\\n # {\\n # 'email': user.email,\\n # 'uid': urlsafe_base64_encode(force_bytes(user.pk)),\\n # 'user': user,\\n # 'token': token_generator.make_token(user),\\n # 'protocol': settings.PROTOCOL,\\n # },\\n # )\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.67476535","0.6409794","0.64040166","0.6167846","0.61461926","0.61059725","0.61046255","0.605578","0.6043292","0.6042979","0.5960896","0.59149146","0.5868797","0.5867584","0.58403814","0.58367157","0.58244336","0.5819981","0.57925284","0.57854337","0.57668483","0.574957","0.5728815","0.57259405","0.5724137","0.57138467","0.5710851","0.56969476","0.5688985","0.5680687","0.5676845","0.5676211","0.5671401","0.56634235","0.56632507","0.5644984","0.5631534","0.5607752","0.56002873","0.55931866","0.5574163","0.5570142","0.5558944","0.5558735","0.5541035","0.5520252","0.55084115","0.5507227","0.5501676","0.54985225","0.54873294","0.5485911","0.54814804","0.54767025","0.54762274","0.5475122","0.5455393","0.5435693","0.54271823","0.5424069","0.54148173","0.54000676","0.5397839","0.5394657","0.53932303","0.537716","0.53655","0.53633374","0.5340855","0.53378004","0.53204244","0.53155553","0.5310433","0.530372","0.52996755","0.529565","0.528736","0.5277796","0.52709603","0.52611226","0.52590024","0.52533823","0.52522105","0.5247004","0.5245502","0.5243052","0.5242859","0.5242779","0.5240156","0.5237603","0.52355456","0.52316964","0.5230732","0.5230335","0.52298045","0.52260375","0.5225771","0.5218891","0.5211341","0.52110124"],"string":"[\n \"0.67476535\",\n \"0.6409794\",\n \"0.64040166\",\n \"0.6167846\",\n \"0.61461926\",\n \"0.61059725\",\n \"0.61046255\",\n \"0.605578\",\n \"0.6043292\",\n \"0.6042979\",\n \"0.5960896\",\n \"0.59149146\",\n \"0.5868797\",\n \"0.5867584\",\n \"0.58403814\",\n \"0.58367157\",\n \"0.58244336\",\n \"0.5819981\",\n \"0.57925284\",\n \"0.57854337\",\n \"0.57668483\",\n \"0.574957\",\n \"0.5728815\",\n \"0.57259405\",\n \"0.5724137\",\n \"0.57138467\",\n \"0.5710851\",\n \"0.56969476\",\n \"0.5688985\",\n \"0.5680687\",\n \"0.5676845\",\n \"0.5676211\",\n \"0.5671401\",\n \"0.56634235\",\n \"0.56632507\",\n \"0.5644984\",\n \"0.5631534\",\n \"0.5607752\",\n \"0.56002873\",\n \"0.55931866\",\n \"0.5574163\",\n \"0.5570142\",\n \"0.5558944\",\n \"0.5558735\",\n \"0.5541035\",\n \"0.5520252\",\n \"0.55084115\",\n \"0.5507227\",\n \"0.5501676\",\n \"0.54985225\",\n \"0.54873294\",\n \"0.5485911\",\n \"0.54814804\",\n \"0.54767025\",\n \"0.54762274\",\n \"0.5475122\",\n \"0.5455393\",\n \"0.5435693\",\n \"0.54271823\",\n \"0.5424069\",\n \"0.54148173\",\n \"0.54000676\",\n \"0.5397839\",\n \"0.5394657\",\n \"0.53932303\",\n \"0.537716\",\n \"0.53655\",\n \"0.53633374\",\n \"0.5340855\",\n \"0.53378004\",\n \"0.53204244\",\n \"0.53155553\",\n \"0.5310433\",\n \"0.530372\",\n \"0.52996755\",\n \"0.529565\",\n \"0.528736\",\n \"0.5277796\",\n \"0.52709603\",\n \"0.52611226\",\n \"0.52590024\",\n \"0.52533823\",\n \"0.52522105\",\n \"0.5247004\",\n \"0.5245502\",\n \"0.5243052\",\n \"0.5242859\",\n \"0.5242779\",\n \"0.5240156\",\n \"0.5237603\",\n \"0.52355456\",\n \"0.52316964\",\n \"0.5230732\",\n \"0.5230335\",\n \"0.52298045\",\n \"0.52260375\",\n \"0.5225771\",\n \"0.5218891\",\n \"0.5211341\",\n \"0.52110124\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.5880153"},"document_rank":{"kind":"string","value":"12"}}},{"rowIdx":90869,"cells":{"query":{"kind":"string","value":"Render each drawing by converting the drawing to image ink and then applying broaden & blur filters"},"document":{"kind":"string","value":"def forward(self, drawings, blur_sigma=None, epsilon=None):\n # draw the strokes (this part on cpu)\n if not isinstance(drawings[0], list):\n single = True\n drawings = [drawings]\n else:\n single = False\n\n pimgs = self.painter(drawings)\n pimgs = self.broaden_and_blur(pimgs, blur_sigma, epsilon) # (n,H,W)\n\n if single:\n pimgs = pimgs[0]\n\n return pimgs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def draw(self):\n self.bufferX = (self.appWidth/2) - self.viewX\n self.bufferY = (self.appHeight/2) - self.viewY\n anwp.sl.engine.clear()\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\n self.drawWarpLines()\n \n # render engine\n anwp.sl.engine.render()\n self.drawSystemInfo()\n self.drawWarpGateInfo()\n self.drawWarpTradeInfo()","def draw(self):\n\n self.updateLazyImageLoading()\n\n image(self.baseMap, 0, 0)\n\n for layer in self.layers:\n layer.draw()\n\n for marker in self.markers:\n marker.draw()","def _pre_draw_bge(self):\r\n self._pre_draw_common()\r\n # draw rays\r\n self._drawRays()","def draw(self):\n self.write_image()\n self.update()","def forward_partial(self, drawing):\n assert not self.is_cuda\n drawing = drawings_to_cpu(drawing)\n ns = len(drawing)\n pimgs = torch.zeros(ns+1, *self.imsize)\n canvas = torch.zeros(*self.imsize)\n for i, stk in enumerate(drawing):\n canvas, _ = self.add_stroke(canvas, stk)\n pimgs[i+1] = canvas\n\n return pimgs","def draw(self,image):\n\t\tfor square in self.squares:\n\t\t\tsquare.draw(image, (0,0,255))\n\t\t\tsquare.classify(image)","def draw(self):\n\n for item in self.vis:\n item.undraw()\n self.render()\n for item in self.vis:\n item.draw(self.win)\n self.drawn = True","def render(self):\n self.axial.Render()\n self.coronal.Render()\n self.sagittal.Render()\n #self.isosurface.Render()\n #self.rwi_pcp.Render()","def draw(self):\r\n\r\n\r\n\t\tself.predict()\r\n\t\t#print np.shape(self.gray)\r\n\t\t#cv2.rectangle(self.gray, (self.bb[0], self.bb[1]), (self.bb[0] + self.bb[2], self.bb[1] + self.bb[3]))\r\n\r\n\t\t# draw points as green circles\r\n\t\tfor point in self.features:\r\n\t\t\tcv2.circle(self.gray,(int(point[0][0]),int(point[0][1])),3,(255),-1)\r\n\t\t\t\r\n\t\tcv2.imshow('image',self.gray)\r\n\t\tcv2.waitKey(1)","def draw():","def draw(self):\n for obj in self.objects:\n obj.draw()","def draw(self,surface):\n surface.blit(self.image, self.rect)\n for moving in self.shots.values():\n moving.draw()","def pipeline(image):\n # undistort image\n undistorted_image = undistort_image(image)\n superimposed_image = find_lanes(undistorted_image)\n labels = find_vehicles(undistorted_image)\n\n draw_img = draw_labeled_bboxes(superimposed_image, labels)\n\n \n return draw_img","def render_map(self):\n # first we create a blank image, on which we will draw the base map\n width = self.image_size[0]\n height = self.image_size[1]\n # ex: size of the image 1080 height, 1920 width, 3 channels of colour\n base_map = np.zeros((height, width, 3), np.uint8)\n base_map[:, :] = self.background_color\n\n # we draw each shape of the dictionary on the blank image\n for shape_id in self.shape_dict_filt:\n shape = self.shape_dict_filt[shape_id]\n points = shape.points\n pts = np.array(points, np.int32)\n cv2.polylines(base_map, [pts], True, shape.color_line,\n shape.line_thick, cv2.LINE_AA)\n\n self.map_file = base_map","def render(self, ctx):\n image = np.zeros((self.height, self.width, 3), np.uint8)\n for shape in ctx.shapes:\n if isinstance(shape, context.Rectangle):\n x = int(shape.width / 2)\n y = int(shape.height / 2)\n rad = np.radians(shape.rotation)\n rotation = np.array([[np.cos(rad), -np.sin(rad)],\n [np.sin(rad), np.cos(rad)]])\n translation = np.array([[shape.center.x], [shape.center.y]])\n corners = np.array([[-x, x, x, -x], [y, y, -y, -y]])\n transformed_corners = rotation.dot(corners) + translation\n transformed_corners = transformed_corners.T.astype(int)\n cv2.fillPoly(image, pts=[transformed_corners],\n color=shape.color)\n elif isinstance(shape, context.Circle):\n center = (int(shape.center.x), int(shape.center.y))\n image = cv2.circle(image, center, int(shape.radius),\n color=shape.color, thickness=shape.thickness)\n elif isinstance(shape, context.Text):\n center = (int(shape.center.x), int(shape.center.y))\n image = cv2.putText(image, shape.content, center,\n cv2.FONT_HERSHEY_SIMPLEX, shape.size,\n shape.color, 3, cv2.LINE_AA)\n elif isinstance(shape, context.Image):\n file_image = cv2.imread(shape.filepath, cv2.IMREAD_UNCHANGED)\n file_image = cv2.resize(file_image, (shape.width, shape.height))\n\n y1 = int(shape.center.y - shape.height / 2)\n y2 = int(y1 + file_image.shape[0])\n x1 = int(shape.center.x - shape.width / 2)\n x2 = int(x1 + file_image.shape[1])\n\n rgba = cv2.cvtColor(file_image, cv2.COLOR_RGB2RGBA)\n alpha_s = rgba[:, :, 3] / 255.0\n alpha_l = 1.0 - alpha_s\n\n image_save = image.copy()\n for c in range(0, 3):\n try:\n image[y1:y2, x1:x2, c] = (\n alpha_s * file_image[:, :, c] +\n alpha_l * image[y1:y2, x1:x2, c])\n except ValueError:\n image = image_save\n\n self._display_frame(image)","def render(self, window):\n body = pygame.image.load(IMAGE_SNAKE).convert_alpha() # loading image\n for block in self.body:\n window.blit(body, (block[0]*SPRITE_SIZE, block[1]*SPRITE_SIZE)) # painting a beautiful snek\n if self.neural_net: # calls for neural net rendering\n self.neural_net.render(window, self.vision)","def render_and_save():\n\n rendering_config = configuration.get_config()\n rendering_config = ml_collections.FrozenConfigDict(rendering_config)\n aspect_ratio = rendering_config.aspect_ratio\n height = rendering_config.height\n width = int(aspect_ratio * height)\n\n scene_camera = build_camera(rendering_config, aspect_ratio)\n world = build_world(rendering_config)\n\n # Render.\n logging.info(\"Tracing rays...\")\n render_image_fn = jax.jit(\n render.generate_image,\n static_argnames=[\"height\", \"width\", \"config\"])\n image = render_image_fn(height, width, scene_camera, world, rendering_config)\n\n image = render.correct_gamma(image, gamma=rendering_config.gamma_correction)\n\n logging.info(\"Saving to file...\")\n output.export_as_ppm(image, rendering_config.output_file)\n\n return image","def draw(self):\n self.screen.fill((0,51,102))\n # get the new drawables\n self.drawables = (self.game_model.get_background_drawables()\n + self.game_model.get_plane_drawables()\n + self.game_model.get_bullet_drawables()\n + self.game_model.get_enemy_drawables())\n for d in self.drawables:\n rect = d.get_rect()\n surf = d.get_surface()\n surf.set_colorkey((255,255,255))\n self.screen.blit(surf, rect)","def draw(self, surface):\n temp = pygame.Surface(self.renderer.pixel_size)\n self.renderer.render_map(temp)\n pygame.transform.smoothscale(temp, surface.get_size(), surface)","def draw(self):","def draw(self, surface):\n surface.blit(self.image, (0,0))\n for widget in self.widgets:\n widget.draw(surface)","def main():\n parser = argparse.ArgumentParser(description=\"Process a wavefront object file\")\n parser.add_argument('--width', help=\"Width of output image\", dest='width',\n type=int, default=800)\n parser.add_argument('--height', help=\"Height of output image\", dest='height',\n type=int, default=600)\n parser.add_argument('--out', help=\"Name of output image file\", dest='output', type=str,\n default='output.png')\n parser.add_argument('filename', help=\"Alias Wavefront file to read as input\", type=str)\n args = parser.parse_args()\n\n back_buffer = Buffer(args.width, args.height)\n screen = Screen(back_buffer)\n\n obj = Wavefront(args.filename)\n print(\"Processing {}\".format(args.filename))\n\n max_extent = max(obj.v_extent[0], obj.v_extent[1])\n scale = min(args.width / max_extent, args.height / max_extent)\n translate = ((args.width / 2) + obj.v_min[0] + obj.v_max[0],\n (args.height / 2) + obj.v_min[1] + obj.v_max[1])\n print(\"Using scale: {}\".format(scale))\n\n def conv_x(x):\n \"\"\" Converts x coordinate.\"\"\"\n return int(x * scale + translate[0])\n def conv_y(y):\n \"\"\" Converts y coordinate.\"\"\"\n return int(y * scale + translate[1])\n\n for f in obj.f:\n v1 = obj.v[f[0]]\n v2 = obj.v[f[1]]\n v3 = obj.v[f[2]]\n screen.draw_line((conv_x(v1[0]), conv_y(v1[1])),\n (conv_x(v2[0]), conv_y(v2[1])),\n (255, 255, 255))\n screen.draw_line((conv_x(v2[0]), conv_y(v2[1])),\n (conv_x(v3[0]), conv_y(v3[1])),\n (255, 255, 255))\n screen.draw_line((conv_x(v3[0]), conv_y(v3[1])),\n (conv_x(v1[0]), conv_y(v1[1])),\n (255, 255, 255))\n\n print(\"Minimum vector: {}\".format(obj.v_min))\n print(\"Maximum vector: {}\".format(obj.v_max))\n print(\"Calculated extent: {}\".format(obj.v_extent))\n\n back_buffer.write_to_png(args.output)\n print(\"Written output to {}\".format(args.output))","def draw(self):\n return ImageDraw.Draw(self.buffer)","def renderCap( self, scale = 400.0):\n vertices = [\n vertex.Vertex( (x/scale,y/scale,0) ) for (x,y) in outline[::2]\n ]\n vertices.reverse()\n for type, vertices in self.tess.tessellate( vertices, forceTriangles=0 ):\n glNormal( 0,0,-1 )\n glColor3f( 1,0,0 )\n glBegin( type )\n for v in vertices:\n glVertex2dv( v.point[:2] )\n glEnd()","def draw(self):\n self.batch.draw()","def render(pictures):\n # Display the background\n display.fill(BLACK)\n\n # Draw each picture\n i = 0 # Number of pictures drawn\n for picture in pictures:\n offsets = get_offsets(i)\n\n for polygon in picture.polygons:\n # Adjust vertices so pictures aren't overlapping\n adjusted_vertices = []\n\n for v in polygon.vertices:\n adjusted_vertices.append((v[0] + offsets[0], v[1] + offsets[1]))\n pygame.draw.polygon(display, polygon.color, adjusted_vertices)\n\n i += 1\n\n pygame.display.update()","def draw_layers(self):\n\t\tfor z in xrange(0,16):\n\t\t\t#create surface for this layer\n\t\t\tsrf = pygame.Surface((16,128))\n\t\t\tfor x in xrange(0,16):\n\t\t\t\tfor y in xrange(0,128):\n\t\t\t\t\tv = self.data[ self.xyz_to_offset( x,y,z) ]\n\t\t\t\t\tif v != 0:\n\t\t\t\t\t\tsrf.fill( BLOCKS.get(v, [0,0])[1], \t(x, 127 -y, 1, 1 ))\n\t\t\t#save layer to dict for this chunk\n\t\t\tself.layers[z] = srf","def drawImage(self):\n \n self.retrieveThemeConfig()\n \n self.initImg()\n self.colsDone = 1\n \n queue = mp.Queue()\n procs = []\n for i in range(0,self.strandNum):\n procs.append(mp.Process(target=self.strandWorker,args=(queue,i,\n self.branches)))\n procs[i].start()\n\n self.allCords = []\n for i in range(0,self.strandNum):\n self.allCords.append(queue.get())\n \n logging.info(\"Strand workers finished\")\n\n self.curCol = 0\n for k in range(0,len(self.allCords)):\n for l in range(0,len(self.allCords[k])):\n for m in range(0,len(self.allCords[k][l]['cords'])):\n self.draw.line(self.allCords[k][l]['cords'][m],\n self.allCords[k][l]['col'],self.thickness)\n del self.draw\n self.exportImg()","def draw(self, frame):\n for prediction in self.predictions:\n prediction.draw(frame)","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n\t\tpass","def on_draw_over_image(self):","def extract_intrinsic_images_from_lighting_passes(self, outdir, vis=False):\n logger_name = thisfile + '->extract_intrinsic_images_from_lighting_passes()'\n xm.general.makedirs(outdir)\n data = self.data\n\n def collapse_passes(components):\n ch_arrays = []\n for ch in ['R', 'G', 'B']:\n comp_arrs = []\n for comp in components:\n comp_arrs.append(data[comp + '.' + ch])\n ch_array = np.sum(comp_arrs, axis=0) # sum components\n ch_arrays.append(ch_array)\n # Handle alpha channel\n first_alpha = data[components[0] + '.A']\n for ci in range(1, len(components)):\n assert (first_alpha == data[components[ci] + '.A']).all(), \\\n \"Alpha channels of all passes must be the same\"\n ch_arrays.append(first_alpha)\n return np.dstack(ch_arrays)\n\n # Albedo\n albedo = collapse_passes(['diffuse_color', 'glossy_color'])\n np.save(join(outdir, 'albedo.npy'), albedo)\n if vis:\n xm.vis.matrix_as_image(albedo, outpath=join(outdir, 'albedo.png'))\n # Shading\n shading = collapse_passes(['diffuse_indirect', 'diffuse_direct'])\n np.save(join(outdir, 'shading.npy'), shading)\n if vis:\n xm.vis.matrix_as_image(shading, join(outdir, 'shading.png'))\n # Specularity\n specularity = collapse_passes(['glossy_indirect', 'glossy_direct'])\n np.save(join(outdir, 'specularity.npy'), specularity)\n if vis:\n xm.vis.matrix_as_image(specularity, join(outdir, 'specularity.png'))\n # Reconstruction vs. ...\n recon = np.multiply(albedo, shading) + specularity\n recon[:, :, 3] = albedo[:, :, 3] # can't add up alpha channels\n np.save(join(outdir, 'recon.npy'), recon)\n if vis:\n xm.vis.matrix_as_image(recon, join(outdir, 'recon.png'))\n # ... composite from Blender, just for sanity check\n composite = collapse_passes(['composite'])\n np.save(join(outdir, 'composite.npy'), composite)\n if vis:\n xm.vis.matrix_as_image(composite, join(outdir, 'composite.png'))\n logger.name = logger_name\n logger.info(\"Intrinsic images extracted to %s\", outdir)","def on_draw( self ):\r\n self.clear()\r\n self.setup_3D()\r\n print \"DEBUG:\" , \"There are\" , len( self.renderlist ) , \"items in 'self.renderlist'\"\r\n for obj in self.renderlist:\r\n obj.draw()","def draw(font_string,font_size,lang,alphabets,outdir=\".\"): # language, font file name, font full path, font size, characters\n \n \n image_dir=lang+\".\"+\"images\"\n if(os.path.exists(image_dir)):\n pass\n else:\n os.mkdir(image_dir)\n \n #Using a font\n #font= ImageFont.truetype(font,fsz)\n boxfile=image_dir+\"/\"+\"bigimage.box\"\n f=open(boxfile,\"w\")\n wt = 4000\n ht = 4000 #modified later using a separate script\n\t\n bigimage=Image.new(\"L\",(wt,ht),255)\t#change here for inverting\n bigdraw=ImageDraw.Draw(bigimage)\n x=y=10\n count=0\n for akshar in alphabets:\n akshar.strip() #remove nasty characters\n \n #I shall now create an image with black bgc and white font color. One\n #getbbox() determines the bounding box values I shall invert the image.\n #This has to be done since getbbox() only finds bounding box values for\n #non-zero pixels (read as white), but tesseract-ocr runs on the exact\n #opposite bgc fgc combination. Contact debayanin@gmail.com.\n \n \n #The lines below are pango/cairo code \n surface = cairo.ImageSurface(cairo.FORMAT_A8, font_size*4, font_size*3)\n context = cairo.Context(surface)\n\n pc = pangocairo.CairoContext(context)\n\n layout = pc.create_layout()\n layout.set_font_description(pango.FontDescription(font_string))\n layout.set_text(akshar)\n print akshar\n\n # lines take care of centering the text.\n width, height = surface.get_width(), surface.get_height()\n w, h = layout.get_pixel_size()\n position = (10,10) #most likely this part messes up when you try to change the size within this script. It is suggested to use the separate script.\n context.move_to(*position)\n pc.show_layout(layout)\n surface.write_to_png(\"pango.png\")\n\t\n #Here we open the generated image using PIL functions\n temp_image=Image.open(\"pango.png\") #black background, white text\n draw = ImageDraw.Draw(temp_image)\n bbox = temp_image.getbbox()\n deltax=bbox[2]-bbox[0]\n deltay=bbox[3]-bbox[1]\n\n \n print bbox\n new_image=temp_image.crop(bbox)\n temp_image=temp_image.load()\n inverted_image = ImageChops.invert(new_image) #White background, black text\n\t\n\tinverted_image.save(image_dir+\"/\"+str(count)+\".png\")\n\tbigimage.paste(inverted_image,(x,y))\n\tos.unlink(image_dir+\"/\"+str(count)+\".png\")\n\tcount = count+1\n\t#bigimage.load()\n bigbox=(x,y,x+deltax,y+deltay)\n print bigbox\n draw=ImageDraw.Draw(bigimage)\n\t#draw.rectangle(bigbox,None,100)\n x=bigbox[2]+5\n if x>(wt-10):\n x=10; y=y+40\n\n os.unlink(\"pango.png\") #delete the pango generated png\n\n line=akshar+\" \"+str(bigbox[0]-1)+\" \"+str(ht-(bigbox[1]+deltay)-1)+\" \"+str(bigbox[2]+1)+\" \"+str(ht-(bigbox[3]-deltay)+1) # this is the line to be added to the box file\n\tf.write(line+'\\n')\n\n\t#degrade code starts\n\tstrip=[deltax*.2,deltax*.4,deltax*.7]\n\tfor values in range(0,2):\n\t\tdistort2=inverted_image\n\t\tfor wai in range(0,deltay):\n\t\t\tfor ex in range(strip[values],strip[values]+1):\n\t\t\t\tdistort2.putpixel((ex,wai),255)\n\t\tbigbox=(x,y,x+deltax,y+deltay)\n\t\t#draw.rectangle(bigbox,None,10)\n\t\tline=akshar+\" \"+str(bigbox[0]-1)+\" \"+str(ht-(bigbox[1]+deltay)-1)+\" \"+str(bigbox[2]+1)+\" \"+str(ht-(bigbox[3]-deltay)+1) # this is the line to be added to the box file\n \tf.write(line+'\\n')\n\t\tbigimage.paste(distort2,(x,y))\n\t\tx=bigbox[2]+5\n \tif x>(wt-10):\n \t\tx=10; y=y+40\n\t\t\n\t\t\t\n\t#degrade code ends\n \n #distort.distort(filename2,bbox,fsz,akshar)\n \n \n \n #bigimage.save(image_dir+\"/\"+\"bigimage.tif\",\"TIFF\") #useful to generate merged file for all images when using default sizes.\n f.close()\n train.train(lang,outdir)","def _drawOnCanvas(self):\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\n for key in self.elements:\n graphElement=self.elements[key]\n graphElement.draw(self.canvas)\n self.sync=True","def draw(self, view):\n for r in self._aliens:\n for alien in r:\n if alien != None:\n alien.draw(view)\n if self._ship != None:\n self._ship.draw(view)\n self._dline.draw(view)\n for bolt in self._bolts:\n bolt.draw(view)","def draw(self):\r\n scalex,scaley = self.getScale()\r\n try:\r\n self.clear()\r\n # Draw Graph Background\r\n self.drawLayout()\r\n if self.app.data == None:# If no data, break\r\n return\r\n # How much each pixel represents\r\n if scalex[1]-scalex[0] == 0:\r\n return\r\n step = (scalex[1]-scalex[0])/self.w# Draw lines at pixel level resolution\r\n self.fitYScale()\r\n sens_index = [0]# If one sensor displayed in this data player\r\n if len(self.sensor_ids) == 2:# If two sensors displayed in this data player\r\n sens_index = [1,0]# Draw order blue then red to make blue line on top\r\n for s in sens_index:\r\n i = scalex[0]\r\n x = 0\r\n trackcol = self.app.getSensorCol(self.sensors[self.sensor_ids[s]])\r\n while i < scalex[1]:\r\n i += step# i Is data\r\n x += 1# x is iteration/pixel-coordinate\r\n if i<0:# Skip data for t<0\r\n continue\r\n try:\r\n # Data retrieved from xml\r\n y = float(self.app.data[int(i)][self.sensor_ids[s]].text)\r\n y2 = float(self.app.data[int(i+step)][self.sensor_ids[s]].text)\r\n # Normalize into range 0 to 1 and multiply by height\r\n y = ((y-scaley[0])/(scaley[1]-scaley[0])) * self.h\r\n y2 = ((y2-scaley[0])/(scaley[1]-scaley[0])) * self.h\r\n except IndexError:# Missing data is skipped\r\n continue\r\n self.c.create_line(x,-y+self.h,x+1,-y2+self.h,fill=trackcol,width=1)\r\n self.drawScrubber()\r\n self.drawPeekScrubber()\r\n self.c.update()\r\n except tk.TclError:# If canvas destroyed, cancel draw operation\r\n return","def draw(self, surface):\n surface.blit(self.image, self.rect)","def draw(self, surface):\n surface.blit(self.image, self.rect)","def draw(self, surface):\r\n surface.blit(self.image, self.rect)","def draw(iiter):\n from matplotlib import pyplot as plt\n fig = plt.gcf()\n fig.canvas.draw()","def draw(self):\n\t\tif self.image != None:\n\t\t\treturn\n\n\t\tself.image = Image.new(\"RGBA\", \n\t\t\t(self.__class__._width, self.__class__._height), (0,0,0,0))\n\t\tdr = aggdraw.Draw(self.image)\n\n\t\tzIndexList = [] # list of z-indexes\n\t\tgIndexList = [] # list of gene indexes\n\n\t\tgeneIdx = 0\n\t\tfor gene in self.chromosome: \n\t\t\tzIndexList.append(gene[0])\n\t\t\tgIndexList.append(geneIdx)\n\t\t\tgeneIdx+=1\n\n\t\twhile len(zIndexList) > 0:\n\t\t\tzIndexMin = zIndexList.index(min(zIndexList))\n\t\t\tgeneIdx = gIndexList[zIndexMin]\n\t\t\tz = self.chromosome[geneIdx,0]\n\t\t\tcolor = self.chromosome[geneIdx,1]\n\t\t\tpoly = self.chromosome[geneIdx,2]\n\n\t\t\tcol = color.getColor()\n\t\t\tcords = poly.getCords()\n\t\t\tbrush = aggdraw.Brush(col)\n\n\t\t\t#dr.ellipse((x, y, x+rad, y+rad), None, brush)\n\t\t\tdr.polygon(cords, None, brush)\n\t\t\tdr.flush()\n\n\t\t\tif 0: # Debug - draw dot in center of polygon\n\t\t\t\tcenter = (poly.xAnchor, poly.yAnchor, poly.xAnchor+5, poly.yAnchor+5)\n\t\t\t\tdr.ellipse(center, None, aggdraw.Brush((255,0,0,255)))\n\t\t\t\tdr.flush()\n\n\t\t\tzIndexList.pop(zIndexMin)\n\t\t\tgIndexList.pop(zIndexMin)\n\n\t\tself.image = self.image.convert(\"RGB\")","def draw(self, surface):\n for box in self.checkboxes:\n box.draw(surface)","def generate_images_pred(self, inputs, outputs):\n for scale in self.scales:\n disp = outputs[(\"disp\", scale)]\n disp = F.interpolate(\n disp, [self.height, self.width], mode=\"bilinear\", align_corners=False)\n source_scale = 0\n\n _, depth = disp_to_depth(disp, self.min_depth, self.max_depth)\n\n outputs[(\"depth\", 0, scale)] = depth\n\n for i, frame_id in enumerate(self.frame_ids[1:]):\n\n T = outputs[(\"cam_T_cam\", 0, frame_id)]\n\n # from the authors of https://arxiv.org/abs/1712.00175\n # mean-normalized inverse depth from [62] to discourage shrinking of the estimated depth\n\n axisangle = outputs[(\"axisangle\", 0, frame_id)]\n translation = outputs[(\"translation\", 0, frame_id)]\n\n inv_depth = 1 / depth\n mean_inv_depth = inv_depth.mean(3, True).mean(2, True)\n\n T = transformation_from_parameters(\n axisangle[:, 0], translation[:, 0] * mean_inv_depth[:, 0], frame_id < 0)\n\n cam_points = self.backproject_depth[source_scale](\n depth, inputs[(\"inv_K\", source_scale)])\n pix_coords = self.project_3d[source_scale](\n cam_points, inputs[(\"K\", source_scale)], T)\n\n outputs[(\"sample\", frame_id, scale)] = pix_coords\n\n outputs[(\"color\", frame_id, scale)] = F.grid_sample(\n inputs[(\"color\", frame_id, source_scale)],\n outputs[(\"sample\", frame_id, scale)],\n padding_mode=\"border\")\n\n outputs[(\"color_identity\", frame_id, scale)] = \\\n inputs[(\"color\", frame_id, source_scale)]","def on_render(self):\n\n # set text font for in visualization\n pygame.font.init()\n font = pygame.font.SysFont('Arial', 24)\n\n # continue running algorithm until done\n if self.algorithm.isDone is False:\n self.algorithm.execute()\n\n # draw background\n for i in range(0, 3):\n for j in range(0, 3):\n self.screen.blit(self.img_grass, (i * 270, j * 270))\n\n # get water instances to draw and draw them\n allWatersList = []\n allWatersList.extend(self.area.allWatersList)\n\n for water in allWatersList:\n pygame.draw.rect(\n self.screen, (0, 0, 128),\n (water.x * 2,\n water.y * 2,\n water.width * 2,\n water.height * 2)\n )\n pygame.draw.rect(\n self.screen,\n (0, 0, 0),\n (0, self.height-50, self.width, 50)\n )\n\n # get house instances to draw and draw them\n housesToPlace = []\n housesToPlace.extend(self.area.mansionList)\n housesToPlace.extend(self.area.familyHomeList)\n housesToPlace.extend(self.area.bungalowList)\n\n for house in housesToPlace:\n # draw free space\n space = pygame.Surface((house.space * 4 + house.width * 2,\n house.space * 4 + house.height * 2))\n space.set_alpha(64)\n space.fill((180, 0, 0))\n self.screen.blit(space,\n (house.x * 2 - house.space * 2,\n house.y * 2 - house.space * 2))\n\n # draw minimum free space\n space = pygame.Surface((house.minimumSpace * 4 + house.width * 2,\n house.minimumSpace * 4 + house.height * 2))\n space.set_alpha(110)\n space.fill((100, 0, 0))\n self.screen.blit(space,\n (house.x * 2 - house.minimumSpace * 2,\n house.y * 2 - house.minimumSpace * 2))\n\n for house in housesToPlace:\n # draw house, colored based on type\n kind = type(house).__name__\n if kind == \"Mansion\":\n pygame.draw.rect(self.screen,\n (200, 255, 40),\n (house.x * 2,\n house.y * 2,\n house.width * 2,\n house.height * 2))\n elif kind == \"Bungalow\":\n pygame.draw.rect(self.screen,\n (255, 40, 200),\n (house.x * 2,\n house.y * 2,\n house.width * 2,\n house.height * 2))\n elif kind == \"FamilyHome\":\n pygame.draw.rect(self.screen,\n (0, 255, 0),\n (house.x * 2,\n house.y * 2,\n house.width * 2,\n house.height * 2))\n\n # Draw black bar at bottom of screen for extra info\n pygame.draw.rect(self.screen,\n (0, 0, 0),\n (0, self.height - 50, self.width, 50))\n\n # Draw area value and last value increase in infobox\n textSurface = font.render('Score: '\n + str(self.area.price),\n True, (255, 255, 255))\n self.screen.blit(textSurface, (10, self.height-35))\n\n # create distinct color for value decreases\n increaseColor = (255, 255, 255)\n if (self.area.price - self.lastPrice < 1):\n increaseColor = (80, 80, 80)\n textSurface = font.render('Increase: ' +\n str(self.area.price - self.lastPrice),\n True, increaseColor)\n self.screen.blit(textSurface, (330, self.height-35))\n pygame.draw.rect(self.screen,\n (0, 0, 0),\n (640, 0, 400, self.height))\n\n # save area values to draw graph\n if (self.area.price >= self.lastPrice and\n self.algorithm.isDone is False and\n self.showDecrease is False):\n\n self.scores.append(self.area.price)\n self.lastPrice = self.area.price\n\n if self.showDecrease is True and self.algorithm.isDone is False:\n self.scores.append(self.area.price)\n self.lastPrice = self.area.price\n\n # draw graph with area values\n fig = pylab.figure(figsize=[4, 4], # Inches\n dpi=100) # 100 dots per inch\n ax = fig.gca()\n ax.plot(self.scores)\n\n canvas = agg.FigureCanvasAgg(fig)\n canvas.draw()\n renderer = canvas.get_renderer()\n raw_data = renderer.tostring_rgb()\n\n surf = pygame.image.fromstring(raw_data, (400, 400), \"RGB\")\n self.screen.blit(surf, (640, 0))\n matplotlib.pyplot.close(fig)\n\n # Draw all time highest score if that's set\n if self.allTimeHigh is not 0:\n textSurface = font.render('Highest score: ' +\n str(self.allTimeHigh),\n True, (255, 255, 255))\n self.screen.blit(textSurface, (650, 410))\n\n pygame.display.flip()\n pass","def draw(self, **kwargs):\n for o in sorted(self._drawables, key=default_itemgetter(\"z\", default=0)):\n o.draw(**kwargs)","def _render(self):\n self.dirty = False\n self.image = self.font.render(self._text, self.aa, self.color_fg)\n self.rect = self.image.get_rect()","def blit(self):\n # self.ax1.draw_artist(self.lines[:2])\n # self.ax2.draw_artist(self.lines[2:4])\n self.ax1.autoscale()\n self.ax2.autoscale()\n self.ax1.redraw_in_frame()\n self.ax2.redraw_in_frame()\n self.fig.canvas.blit(self.fig.bbox)","def rasterize(self):\n\n for primitive in self._scene:\n bbox = primitive[\"bounding_box\"]\n # Loop through all pixels\n # You MUST use bounding boxes in order to speed up this loop\n for w in range(bbox[0][0], bbox[1][0]):\n x = w + 0.5\n for h in range(bbox[0][1], bbox[1][1]):\n y = h + 0.5\n # First, we check if the pixel center is inside the primitive\n im_x, im_y = w, self._height - (h + 1)\n if inside(x, y, primitive):\n # apply affine xfrom if needed\n if \"xform\" in primitive.keys():\n result = np.matmul(primitive[\"xform\"],\n [[im_x], [im_y], [1]])\n im_x, im_y = int(result[0][0]), int(result[1][0])\n\n self._image[im_y, im_x] = primitive[\"color\"]\n # break\n # break\n # break","def draw(self, background=COLOUR_BLACK, show=False, save=False,\n generation=None):\n size = self.img_size\n img = Image.new('RGB', size, background)\n draw = Image.new('RGBA', size)\n pdraw = ImageDraw.Draw(draw)\n for polygon in self.polygons:\n colour = polygon.colour\n points = polygon.points\n pdraw.polygon(points, fill=colour, outline=colour)\n img.paste(draw, mask=draw)\n\n if show:\n img.show()\n\n if save:\n # TODO use self.generation\n temp_dir = tempfile.gettempdir()\n temp_name = u\"0000000000{}\".format(generation)[-10:]\n out_path = u\"{}/{}.png\".format(temp_dir, temp_name)\n img = img.filter(ImageFilter.GaussianBlur(radius=3))\n img.save(out_path)\n print u\"saving image to {}\".format(out_path)\n\n return img","def wrapped_render(self, render_modes, front_and_back=False):\n\n # Render raw images\n render_color = False\n for mode in render_modes:\n if mode != RenderMode.DEPTH and mode != RenderMode.SCALED_DEPTH:\n render_color = True\n break\n\n color_im, depth_im = None, None\n if render_color:\n color_im, depth_im = self.render(render_color, front_and_back=front_and_back)\n else:\n depth_im = self.render(render_color)\n\n # For each specified render mode, add an Image object of the appropriate type\n images = []\n for render_mode in render_modes:\n # Then, convert them to an image wrapper class\n if render_mode == RenderMode.SEGMASK:\n images.append(BinaryImage((depth_im > 0.0).astype(np.uint8), frame=self.camera.intrinsics.frame, threshold=0))\n\n elif render_mode == RenderMode.COLOR:\n images.append(ColorImage(color_im, frame=self.camera.intrinsics.frame))\n\n elif render_mode == RenderMode.GRAYSCALE:\n images.append(ColorImage(color_im, frame=self.camera.intrinsics.frame).to_grayscale())\n\n elif render_mode == RenderMode.DEPTH:\n images.append(DepthImage(depth_im, frame=self.camera.intrinsics.frame))\n\n elif render_mode == RenderMode.SCALED_DEPTH:\n images.append(DepthImage(depth_im, frame=self.camera.intrinsics.frame).to_color())\n\n elif render_mode == RenderMode.RGBD:\n c = ColorImage(color_im, frame=self.camera.intrinsics.frame)\n d = DepthImage(depth_im, frame=self.camera.intrinsics.frame)\n images.append(RgbdImage.from_color_and_depth(c, d))\n\n elif render_mode == RenderMode.GD:\n g = ColorImage(color_im, frame=self.camera.intrinsics.frame).to_grayscale()\n d = DepthImage(depth_im, frame=self.camera.intrinsics.frame)\n images.append(GdImage.from_grayscale_and_depth(g, d))\n else:\n raise ValueError('Render mode {} not supported'.format(render_mode))\n\n return images","def on_draw(self, widget, cr):\n #print \"starting to draw\"\n if self.double_buffer is not None:\n self.draw_tiles()\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\n cr.paint()\n else:\n print('Invalid double buffer')\n #print \"done drawing\"\n return False","def draw(self):\n self.figure.canvas.draw_idle()","def draw_kame(self):\r\n #pygame.draw.rect(self.screen, self.color, self.rect, self.image)\r\n self.screen.blit(self.image, self.rect)","def on_draw(self):\n # draw everything","def example_BSR():\n pts = [(1,1),(2,2),(3,3)]\n lines = [ [ (1,1), (1,2), (2,1)], [ (6,1), (1,6), (5,-1)] ]\n\n bloody_simple_2drender('2d_render.png', pts=pts, vecs=pts, lines=lines )","def visualize(self):\n colors = {'outline': (220, 220, 220),\n 'inlier': (0, 255, 0),\n 'outlier': (0, 0, 255),\n 'lines': (128, 220, 128)}\n # Create output image for visualization\n gap = 5\n h1, w1 = self.target.image.shape[:2]\n h2, w2 = self.image.shape[:2]\n vis = np.zeros((max(h1, h2), w1 + w2 + gap, 3), np.uint8)\n vis[:h1, :w1, :] = self.target.image\n w1 += gap\n vis[:h2, w1:w1+w2, :] = self.image\n \n # Draw the located object \n quad = np.float32(self.quad) + np.float32([w1, 0])\n self.draw(vis, colors['outline'], 2, quad)\n \n # draw point details\n inliers = [(x0, y0, x1 + w1, y1) for (x0, y0), (x1, y1) in self.inliers]\n outliers = [(x0, y0, x1 + w1, y1) for (x0, y0), (x1, y1) in self.outliers]\n if colors['outlier'] is not None: # draw x on each point\n r = 2 # radius\n thickness = 2\n for x0, y0, x1, y1 in outliers:\n cv2.line(vis, (x0 - r, y0 - r), (x0 + r, y0 + r), colors['outlier'], thickness)\n cv2.line(vis, (x0 + r, y0 - r), (x0 - r, y0 + r), colors['outlier'], thickness)\n cv2.line(vis, (x1 - r, y1 - r), (x1 + r, y1 + r), colors['outlier'], thickness)\n cv2.line(vis, (x1 + r, y1 - r), (x1 - r, y1 + r), colors['outlier'], thickness)\n if colors['lines'] is not None:\n for x0, y0, x1, y1 in inliers:\n cv2.line(vis, (x0, y0), (x1, y1), colors['lines'], 1)\n if colors['inlier'] is not None:\n for x0, y0, x1, y1 in inliers:\n cv2.circle(vis, (x0, y0), 2, colors['inlier'], -1)\n cv2.circle(vis, (x1, y1), 2, colors['inlier'], -1)\n return vis","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def on_draw(self):\n\n # Clear the screen and start drawing\n arcade.start_render()\n\n # Draw the rectangles\n for shape in self.shapes:\n shape.draw()","def blending_example1():\n pic_desert = read_image(relpath(\"./externals/pic_desert.jpg\"), 2)\n pic_pool = read_image(relpath(\"./externals/pic_swim.jpg\"), 2)\n mask = read_image(relpath(\"./externals/mask_desert.jpg\"), 1)\n # making the mask binary (normalizing 2 original values)\n mask = strech_helper(mask).astype(np.bool)\n print(pic_desert.shape[2])\n [R1, G1, B1] = np.dsplit(pic_desert, pic_desert.shape[2])\n [R2, G2, B2] = np.dsplit(pic_pool, pic_pool.shape[2])\n R1 = np.reshape(R1, (512,1024))\n R2 = np.reshape(R2, (512,1024))\n G1 = np.reshape(G1, (512,1024))\n G2 = np.reshape(G2, (512,1024))\n B1 = np.reshape(B1, (512,1024))\n B2 = np.reshape(B2, (512,1024))\n\n blend1 = pyramid_blending(R2, R1, mask, 3, 3, 3)\n blend2 = pyramid_blending(G2, G1, mask, 3, 3, 3)\n blend3 = pyramid_blending(B2, B1, mask, 3, 3, 3)\n\n blend1 = np.reshape(blend1, (blend1.shape[0], blend1.shape[1], 1))\n blend2 = np.reshape(blend2, (blend2.shape[0], blend3.shape[1], 1))\n blend3 = np.reshape(blend3, (blend3.shape[0], blend3.shape[1], 1))\n\n new_pic = np.concatenate((blend1, blend2, blend3), axis=2)\n # plotting the images\n fig = plt.figure()\n ax1 = fig.add_subplot(221)\n ax2 = fig.add_subplot(222)\n ax3 = fig.add_subplot(223)\n ax4 = fig.add_subplot(224)\n ax1.imshow(pic_desert)\n ax2.imshow(pic_pool)\n ax3.imshow(mask, cmap='gray')\n ax4.imshow(new_pic)\n plt.show()\n\n return pic_desert, pic_pool, mask, new_pic","def draw(self):\n for section in self.sections:\n canvas_reset(self.canvas)\n section.draw(self.canvas)","def draw(self):\r\n self.screen.blit(self.image, self.image.get_rect())","def render_foreground(img: Image) -> Image:\n canvas = Image.new(\"RGBA\", img.size)\n shadow = img.filter(ImageFilter.GaussianBlur(4))\n shadow = invert_with_alpha(shadow, 120)\n\n canvas.paste(shadow, (4, 4))\n canvas = Image.alpha_composite(canvas, img)\n return canvas","def test_draw(self):\n image_name = filename(sys._getframe().f_code.co_name)\n result_file, reference_file = get_path(image_name)\n\n ''' This function is to create an empty image with a specific dimension\n with white background, and black/white colored '''\n\n image, canvas = get_image('L', (15,90),'white')\n\n for i in range(len(draw_points) - 1):\n draw(canvas, (draw_points[i + 0], draw_points[i + 1]), 'A')\n\n \"\"\" saving the file and closing it \"\"\"\n\n image.save(result_file)\n image.close()\n\n \"\"\" validate the resultant file against the reference images\"\"\"\n\n validate(reference_file, result_file)","def filter(self, drawable):\n pass","def draw(self, grille):\n grille.clear_highlight()\n for x in range(8):\n for y in range(8):\n self.draw_c(x, y, grille)","def render(self, rstate):\n if rstate.picking:\n for textures, group in self.texGroupItems:\n group.pickingDraw(rstate)\n else:\n for textures, group in self.texGroupItems:\n if not rstate.picking:\n try:\n self.bindTextures(textures, rstate)\n except Texture.DynamicTextureException:\n # The textures this drawable requires aren't ready yet\n return\n group.draw(rstate)","def draw (self):\n screen = self.screen\n dirty = False\n for z, displays in self.layers.iteritems():\n for display in displays:\n drew = display.draw(screen)\n # if made changes to the surface\n if drew:\n # set any displays that overlap this one dirty\n for d in display.overlapped:\n d.dirty = True\n dirty |= drew\n return dirty","def draw_img(self, i, j, k):\n if k < len(self.images):\n img = self.images[k]\n r = self.get_rect(i, j)\n self.screen.blit(img, r)","def draw(self):\n self._view.clear()\n for h in range(len(self._bricks)):\n self._bricks[h].draw(self._view)\n \n self._paddle.draw(self._view)\n \n for w in self._FP_list:\n w.draw(self._view)\n \n # draw ball if not None\n if not self._ball is None:\n self._ball.draw(self._view)","def draw(self, screen):","def draw(img, p):\n\t\t\n\t########### First of all we gotta define which point connects together, there are points for the head(5), the shoulders, elbows, hands, top of thighs, knees, feet\n\t########### top of thighs, knees and feet. We also gonna add 2 extra points for the neck and the pelvis, to make everythonh look better.\n\t########### We also provide the color of each line here\n\tconnexions = [\t\n\t\t\t\t\t(5, 7, 'navy'),\t\t# shoulder => elbow\n\t\t\t\t\t(7, 9, 'navy'),\t\t# elbow => hand\n\t\t\t\t\t(6, 8, 'navy'),\t\t# same on the other side\n\t\t\t\t\t(8, 10, 'navy'),\n\t\t\t\t\t(11, 13, 'lime'),\t# thigh => knee\n\t\t\t\t\t(13, 15, 'lime'),\t# knee => foot\n\t\t\t\t\t(12, 14, 'lime'),\t# same on the other side\n\t\t\t\t\t(14, 16, 'lime'),\n\n\t\t\t\t\t###### With The Extra points :\n\n\t\t\t\t\t(0, 17, 'aqua'),\t# head => neck\n\t\t\t\t\t(17, 5, 'aqua'),\t# neck => shoulders\n\t\t\t\t\t(17, 6, 'aqua'),\n\t\t\t\t\t(17, 18, 'teal'),\t# neck => pelvis\n\t\t\t\t\t(18, 11, 'teal'),\t# pelvis => thighs\n\t\t\t\t\t(18, 12, 'teal')\n\t\t\t\t\t]\n\n\t###### now let's find out how many objects were detected \n\t\n\tl = len(p[0][\"scores\"])\n\n\t##### time to draw now, we'll only select objects with a score over .9\n\n\td = idw.Draw(img)\n\n\tfor k in range(l):\n\n\t\tif p[0][\"scores\"][k] > 0.98:\n\n\t\t\t##### Let's add the neck and pelvis:\n\t\t\tneck = (p[0][\"keypoints\"][k][5] + p[0][\"keypoints\"][k][6])/2\n\t\t\tpelv = (p[0][\"keypoints\"][k][11] + p[0][\"keypoints\"][k][12])/2\n\n\t\t\t#### it's getting tricky here\n\n\t\t\tnepe = t.zeros((2, 3))\n\t\t\tnepe[0] = neck ; nepe[1] = pelv \n\n\t\t\t### Now let's put everything into a single tensor\n\t\t\tbody = t.cat((p[0][\"keypoints\"][k], nepe))\n\n\t\t\t#### We can start drawing now, for real\n\n\t\t\tfor tp in connexions:\n\n\t\t\t\tp0 = (int(body[tp[0], 0]), int(body[tp[0], 1]))\n\t\t\t\tp1 = (int(body[tp[1], 0]), int(body[tp[1], 1]))\n\t\t\t\td.line([p0, p1], fill=tp[2], width=2)\n\n\t\t\t#### Now the points\n\n\t\t\tfor ts in t.cat((body[0:1], body[5:])):\n\t\t\t\td.ellipse((int(ts[0]-2), int(ts[1]-2), int(ts[0]+2), int(ts[1]+2)), 'fuchsia')\n\n\t### and finally\n\t#plt.imshow(np.asarray(img)) Not Like That\n\timg.show()","def draw_image(self):\n self.PDF.saveState()\n self.PDF.scale(1, -1)\n # self.PDF.drawImage(\n # LOGO, 490, -78, width=80, preserveAspectRatio=True, mask=\"auto\"\n # )\n self.PDF.restoreState()","def pygame_render(objects_to_draw: DrawingObjects, surface):\n\n for arc in objects_to_draw.arcs:\n pygame.draw.arc(surface, arc.color, arc.enclosing_rect, arc.angle_begin, arc.angle_end, arc.lines_thickness)\n for rect in objects_to_draw.rects:\n pygame.draw.rect(surface, rect.color, rect.shape, rect.lines_thickness)\n for circle in objects_to_draw.circles:\n pygame.draw.circle(surface, circle.color, circle.center, circle.radius, circle.line_thickness)\n for a_line in objects_to_draw.lines:\n pygame.draw.line(surface, a_line.color, a_line.begin, a_line.end, a_line.line_thickness)","def __smoothen_color(self, outer, inner):\n outer_curve = zip(outer[0], outer[1])\n inner_curve = zip(inner[0], inner[1])\n x_points = []\n y_points = []\n for point in outer_curve:\n x_points.append(point[0])\n y_points.append(point[1])\n for point in inner_curve:\n x_points.append(point[0])\n y_points.append(point[1])\n img_base = np.zeros((self.height, self.width))\n cv2.fillConvexPoly(img_base, np.array(np.c_[x_points, y_points], dtype='int32'), 1)\n img_mask = cv2.GaussianBlur(img_base, (81, 81), 0) #51,51\n img_blur_3d = np.ndarray([self.height, self.width, 3], dtype='float')\n img_blur_3d[:, :, 0] = img_mask\n img_blur_3d[:, :, 1] = img_mask\n img_blur_3d[:, :, 2] = img_mask\n self.im_copy = (img_blur_3d * self.image * 0.7 + (1 - img_blur_3d * 0.7) * self.im_copy).astype('uint8')","def draw(self, surface):\n\n\t\tsurface.blit(self.image, self.rect.topleft)","def run(self, src, dest):\n self.logger.debug('Start the blurring. src=\"%s\", dest=\"%s\"', src, dest)\n create_frames(src, self.work['frames'])\n data = self.analyze()\n\n for frame_no, frame in enumerate(self.work.files('frames')):\n basename = os.path.basename(frame)\n areas = []\n for values in data:\n for sector in values['sectors']:\n if frame_no in range(sector[0]-self.offset, sector[1]+self.offset):\n areas.append(values['area'])\n break\n if areas:\n self.blur.blur_image(frame, areas, os.path.join(self.work['cleaned'], basename))\n else:\n copyfile(frame, os.path.join(self.work['cleaned'], basename))\n save_frames(self.work['cleaned'], dest)","def visualize_output(\n self,\n img: np.ndarray,\n output_data: List[SegmObject]) -> np.ndarray:\n draw_layer_tag = f'draw_layer_{self.id}_{self.draw_layer_index^1}'\n mix_factor = .3\n\n self.layer = dpg.add_draw_layer(\n parent=f'main_window_{self.id}',\n tag=draw_layer_tag,\n show=False\n )\n\n for out in output_data:\n if out.score < self.score_threshold:\n continue\n\n mask = cv2.resize(\n out.mask,\n (self.width, self.height),\n cv2.INTER_NEAREST\n ).astype(np.uint8)\n color_mask = np.tile(mask[..., None]/255., [1, 1, 4])\n color_image = np.tile(\n self.class_colors[out.clsname],\n [self.height, self.width, 1]\n )\n img += color_image * color_mask * mix_factor\n contour, _ = cv2.findContours(\n mask.astype(np.uint8),\n cv2.RETR_CCOMP,\n cv2.CHAIN_APPROX_NONE\n )\n color = np.array(self.class_colors[out.clsname])\n color = tuple((255*color).astype(np.uint8))\n for c in contour:\n dpg.draw_polyline(\n parent=draw_layer_tag,\n points=list(c[:, 0, :]),\n color=color,\n thickness=2\n )\n description = f'{out.clsname} [{out.score * 100:.2f}%]'\n dpg.draw_text(\n parent=draw_layer_tag,\n text=description,\n pos=(c[:, 0, 0].min() + _PADDING,\n c[:, 0, 1].min() + _PADDING),\n color=color,\n size=_FONT_SIZE\n )\n\n return img","def end_cast(self):\r\n #draw the actual map\r\n self.emap.draw(shader=self.mshader, camera=self.camera)\r\n super(ShadowCaster, self)._end()\r\n # set third texture to this ShadowCaster texture\r\n texs = self.emap.buf[0].textures\r\n if len(texs) == 2:\r\n texs.append(self)\r\n else:\r\n texs[2] = self\r\n # change background back to blue\r\n opengles.glClearColor(ctypes.c_float(0.4), ctypes.c_float(0.8), \r\n ctypes.c_float(0.8), ctypes.c_float(1.0))\r\n # work out left, top, right, bottom for shader\r\n self.emap.unif[48] = 0.5 * (1.0 + self.scaleu) # left [16][0]\r\n self.emap.unif[49] = 0.5 * (1.0 + self.scalev) # top [16][1]\r\n self.emap.unif[51] = 1.0 - self.emap.unif[48] # right [17][0]\r\n self.emap.unif[52] = 1.0 - self.emap.unif[49] # bottom [17][1]\r\n \r\n du = float(self.location[0] / self.emap.width)\r\n dv = float(self.location[2] / self.emap.depth)\r\n self.emap.unif[48] -= self.scaleu * (du if self.emap.unif[50] == 1.0 else dv)\r\n self.emap.unif[49] += self.scalev * (dv if self.emap.unif[50] == 1.0 else du)\r\n self.emap.unif[51] -= self.scaleu * (du if self.emap.unif[50] == 1.0 else dv)\r\n self.emap.unif[52] += self.scalev * (dv if self.emap.unif[50] == 1.0 else du)","def draw(self, view):\n for i in self.get_aliens():\n for n in i:\n if not n is None:\n n.draw(view)\n self.get_ship().draw(view)\n self.get_ship().get_linevalue().draw(view)\n for i in self.get_PU():\n i.draw(view)\n for i in self.get_bolts():\n i.draw(view)","def add_blurring(self):\n self._render_passes.add_blur_pass()","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self) -> Any:","def __init__(self, mapfile, xpos, zpos, emap, width=10.0, depth=10.0, height=10.0, name=\"building\", draw_details=None, yoff=0.0, scheme=None):\r\n self.xpos = xpos\r\n self.zpos = zpos\r\n self.width = width\r\n self.depth = depth\r\n self.height = height\r\n self.name = name\r\n self.ceilingthickness = 1.0\r\n self.walls = []\r\n\r\n if scheme == None:\r\n self.scheme = Building.baseScheme\r\n else:\r\n self.scheme = scheme\r\n\r\n # We don't have to be rigorous here, this should only be a draw_details or an iterable of draw_details.\r\n if hasattr(draw_details, \"__getitem__\") or hasattr(draw_details, \"__iter__\"):\r\n assert (len(draw_details) == self.scheme[\"#models\"])\r\n self.details = draw_details\r\n else:\r\n self.details = [draw_details for x in range(self.scheme[\"#models\"])]\r\n # having a method like this allows draw details to be set later\r\n\r\n self.yoff = yoff\r\n\r\n self.model = [MergeShape(name=name+\".\"+str(x)) for x in range(self.scheme[\"#models\"])]\r\n\r\n if mapfile[0] != '/':\r\n mapfile = sys.path[0] + '/' + mapfile\r\n print(\"Loading building map ...\", mapfile)\r\n\r\n im = Image.open(mapfile)\r\n im = ImageOps.invert(im)\r\n ix,iy = im.size\r\n\r\n print(\"image size\", ix, \",\", iy)\r\n\r\n startx = xpos - ix / 2 * width\r\n starty = zpos - ix / 2 * depth\r\n\r\n yoff += emap.calcHeight(-xpos,-zpos)\r\n\r\n if not im.mode == \"P\":\r\n im = im.convert('P', palette=Image.ADAPTIVE)\r\n im = im.transpose(Image.FLIP_TOP_BOTTOM)\r\n im = im.transpose(Image.FLIP_LEFT_RIGHT)\r\n pixels = im.load()\r\n\r\n for y in range(1,iy-1):\r\n print(\".\", end='')\r\n for x in range(1,ix-1):\r\n colour = pixels[x,y]\r\n\r\n if x == 1:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x-1,y], \"edge\"), wallfunc=self.west_wall, ceilingedgefunc=self.west_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n else:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x-1,y]), wallfunc=self.west_wall, ceilingedgefunc=self.west_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n\r\n if x == ix-2:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x+1,y], \"edge\"), wallfunc=self.east_wall, ceilingedgefunc=self.east_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n else:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x+1,y]), wallfunc=self.east_wall, ceilingedgefunc=self.east_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n\r\n if y == 1:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y-1], \"edge\"), wallfunc=self.south_wall, ceilingedgefunc=self.south_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n else:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y-1]), wallfunc=self.south_wall, ceilingedgefunc=self.south_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n\r\n if y == iy-2:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x, y+1], \"edge\"), wallfunc=self.north_wall, ceilingedgefunc=self.north_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n else:\r\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y+1]), wallfunc=self.north_wall, ceilingedgefunc=self.north_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n\r\n self._executeScheme(x, y, startx, starty, (colour, None), wallfunc=None, ceilingedgefunc=None, ceilingfunc=self.ceiling, rooffunc=self.roof)\r\n\r\n self.set_draw_details(self.details) # after models created otherwise\r\n # details lost by merging\r","def pixelize(\n layer: vp.LineCollection,\n pen_width: float | None,\n mode: str,\n keep_lines: bool,\n overdraw: float,\n):\n\n if pen_width is None:\n pen_width = layer.property(vp.METADATA_FIELD_PEN_WIDTH) or vp.convert_length(\"0.6mm\")\n color = layer.property(vp.METADATA_FIELD_COLOR) or vp.Color(\"black\")\n color = color.red, color.green, color.blue\n\n pixel_pitch = pen_width\n if mode == \"big\":\n pixel_pitch *= 5\n\n bounds = layer.bounds()\n if bounds is None:\n logging.warning(\"!!! pixelize: layer is empty\")\n return layer\n x_min, y_min, x_max, y_max = bounds\n\n img_size = (\n math.ceil((x_max - x_min) / pixel_pitch),\n math.ceil((y_max - y_min) / pixel_pitch),\n )\n img = Image.new(\"RGBA\", img_size)\n draw = ImageDraw.Draw(img)\n for line in layer:\n draw.line(\n [(pt.real, pt.imag) for pt in (line - x_min - 1j * y_min) / pixel_pitch],\n fill=color,\n )\n\n doc = vp.Document()\n if mode == \"big\":\n # noinspection PyTypeChecker\n big_mode(doc, np.array(img), np.array([color]), pen_width)\n elif mode == \"line\":\n # noinspection PyTypeChecker\n line_mode(doc, np.array(img), np.array([color]), pen_width, overdraw)\n elif mode == \"snake\":\n # noinspection PyTypeChecker\n snake_mode(doc, np.array(img), np.array([color]), pen_width)\n\n doc.translate(x_min, y_min)\n\n if keep_lines:\n layer.extend(doc.layers[1])\n layer.set_property(vp.METADATA_FIELD_PEN_WIDTH, pen_width)\n return layer\n else:\n result = layer.clone(doc.layers[1])\n result.set_property(vp.METADATA_FIELD_PEN_WIDTH, pen_width)\n return result","def render(self, output, image_size,\n face_colors=(\"#477984\", \"#EEAA4D\", \"#74C3F2\")):\n pass","def blending_example2():\n pic_earth = read_image(relpath(\"./externals/pic_earth.jpg\"), 2)\n pic_asteroid = read_image(relpath(\"./externals/pic_asteroid.jpg\"), 2)\n mask = read_image(relpath(\"./externals/mask_asteroid.jpg\"), 1)\n # making the mask binary (normalizing 2 original values)\n mask = strech_helper(mask).astype(np.bool)\n [R1, G1, B1] = np.dsplit(pic_earth, pic_earth.shape[2])\n [R2, G2, B2] = np.dsplit(pic_asteroid, pic_asteroid.shape[2])\n R1 = np.reshape(R1, (1024,1024))\n R2 = np.reshape(R2, (1024,1024))\n G1 = np.reshape(G1, (1024,1024))\n G2 = np.reshape(G2, (1024,1024))\n B1 = np.reshape(B1, (1024,1024))\n B2 = np.reshape(B2, (1024,1024))\n\n blend1 = pyramid_blending(R2, R1, mask, 3, 3, 3)\n blend2 = pyramid_blending(G2, G1, mask, 3, 3, 3)\n blend3 = pyramid_blending(B2, B1, mask, 3, 3, 3)\n\n blend1 = np.reshape(blend1, (blend1.shape[0], blend1.shape[1], 1))\n blend2 = np.reshape(blend2, (blend2.shape[0], blend3.shape[1], 1))\n blend3 = np.reshape(blend3, (blend3.shape[0], blend3.shape[1], 1))\n\n new_pic = np.concatenate((blend1, blend2, blend3), axis=2)\n # plotting the images\n fig = plt.figure()\n ax1 = fig.add_subplot(221)\n ax2 = fig.add_subplot(222)\n ax3 = fig.add_subplot(223)\n ax4 = fig.add_subplot(224)\n ax1.imshow(pic_earth)\n ax2.imshow(pic_asteroid)\n ax3.imshow(mask, cmap='gray')\n ax4.imshow(new_pic)\n plt.show()\n\n return pic_earth, pic_asteroid, mask, new_pic"],"string":"[\n \"def draw(self):\\n self.bufferX = (self.appWidth/2) - self.viewX\\n self.bufferY = (self.appHeight/2) - self.viewY\\n anwp.sl.engine.clear()\\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\\n self.drawWarpLines()\\n \\n # render engine\\n anwp.sl.engine.render()\\n self.drawSystemInfo()\\n self.drawWarpGateInfo()\\n self.drawWarpTradeInfo()\",\n \"def draw(self):\\n\\n self.updateLazyImageLoading()\\n\\n image(self.baseMap, 0, 0)\\n\\n for layer in self.layers:\\n layer.draw()\\n\\n for marker in self.markers:\\n marker.draw()\",\n \"def _pre_draw_bge(self):\\r\\n self._pre_draw_common()\\r\\n # draw rays\\r\\n self._drawRays()\",\n \"def draw(self):\\n self.write_image()\\n self.update()\",\n \"def forward_partial(self, drawing):\\n assert not self.is_cuda\\n drawing = drawings_to_cpu(drawing)\\n ns = len(drawing)\\n pimgs = torch.zeros(ns+1, *self.imsize)\\n canvas = torch.zeros(*self.imsize)\\n for i, stk in enumerate(drawing):\\n canvas, _ = self.add_stroke(canvas, stk)\\n pimgs[i+1] = canvas\\n\\n return pimgs\",\n \"def draw(self,image):\\n\\t\\tfor square in self.squares:\\n\\t\\t\\tsquare.draw(image, (0,0,255))\\n\\t\\t\\tsquare.classify(image)\",\n \"def draw(self):\\n\\n for item in self.vis:\\n item.undraw()\\n self.render()\\n for item in self.vis:\\n item.draw(self.win)\\n self.drawn = True\",\n \"def render(self):\\n self.axial.Render()\\n self.coronal.Render()\\n self.sagittal.Render()\\n #self.isosurface.Render()\\n #self.rwi_pcp.Render()\",\n \"def draw(self):\\r\\n\\r\\n\\r\\n\\t\\tself.predict()\\r\\n\\t\\t#print np.shape(self.gray)\\r\\n\\t\\t#cv2.rectangle(self.gray, (self.bb[0], self.bb[1]), (self.bb[0] + self.bb[2], self.bb[1] + self.bb[3]))\\r\\n\\r\\n\\t\\t# draw points as green circles\\r\\n\\t\\tfor point in self.features:\\r\\n\\t\\t\\tcv2.circle(self.gray,(int(point[0][0]),int(point[0][1])),3,(255),-1)\\r\\n\\t\\t\\t\\r\\n\\t\\tcv2.imshow('image',self.gray)\\r\\n\\t\\tcv2.waitKey(1)\",\n \"def draw():\",\n \"def draw(self):\\n for obj in self.objects:\\n obj.draw()\",\n \"def draw(self,surface):\\n surface.blit(self.image, self.rect)\\n for moving in self.shots.values():\\n moving.draw()\",\n \"def pipeline(image):\\n # undistort image\\n undistorted_image = undistort_image(image)\\n superimposed_image = find_lanes(undistorted_image)\\n labels = find_vehicles(undistorted_image)\\n\\n draw_img = draw_labeled_bboxes(superimposed_image, labels)\\n\\n \\n return draw_img\",\n \"def render_map(self):\\n # first we create a blank image, on which we will draw the base map\\n width = self.image_size[0]\\n height = self.image_size[1]\\n # ex: size of the image 1080 height, 1920 width, 3 channels of colour\\n base_map = np.zeros((height, width, 3), np.uint8)\\n base_map[:, :] = self.background_color\\n\\n # we draw each shape of the dictionary on the blank image\\n for shape_id in self.shape_dict_filt:\\n shape = self.shape_dict_filt[shape_id]\\n points = shape.points\\n pts = np.array(points, np.int32)\\n cv2.polylines(base_map, [pts], True, shape.color_line,\\n shape.line_thick, cv2.LINE_AA)\\n\\n self.map_file = base_map\",\n \"def render(self, ctx):\\n image = np.zeros((self.height, self.width, 3), np.uint8)\\n for shape in ctx.shapes:\\n if isinstance(shape, context.Rectangle):\\n x = int(shape.width / 2)\\n y = int(shape.height / 2)\\n rad = np.radians(shape.rotation)\\n rotation = np.array([[np.cos(rad), -np.sin(rad)],\\n [np.sin(rad), np.cos(rad)]])\\n translation = np.array([[shape.center.x], [shape.center.y]])\\n corners = np.array([[-x, x, x, -x], [y, y, -y, -y]])\\n transformed_corners = rotation.dot(corners) + translation\\n transformed_corners = transformed_corners.T.astype(int)\\n cv2.fillPoly(image, pts=[transformed_corners],\\n color=shape.color)\\n elif isinstance(shape, context.Circle):\\n center = (int(shape.center.x), int(shape.center.y))\\n image = cv2.circle(image, center, int(shape.radius),\\n color=shape.color, thickness=shape.thickness)\\n elif isinstance(shape, context.Text):\\n center = (int(shape.center.x), int(shape.center.y))\\n image = cv2.putText(image, shape.content, center,\\n cv2.FONT_HERSHEY_SIMPLEX, shape.size,\\n shape.color, 3, cv2.LINE_AA)\\n elif isinstance(shape, context.Image):\\n file_image = cv2.imread(shape.filepath, cv2.IMREAD_UNCHANGED)\\n file_image = cv2.resize(file_image, (shape.width, shape.height))\\n\\n y1 = int(shape.center.y - shape.height / 2)\\n y2 = int(y1 + file_image.shape[0])\\n x1 = int(shape.center.x - shape.width / 2)\\n x2 = int(x1 + file_image.shape[1])\\n\\n rgba = cv2.cvtColor(file_image, cv2.COLOR_RGB2RGBA)\\n alpha_s = rgba[:, :, 3] / 255.0\\n alpha_l = 1.0 - alpha_s\\n\\n image_save = image.copy()\\n for c in range(0, 3):\\n try:\\n image[y1:y2, x1:x2, c] = (\\n alpha_s * file_image[:, :, c] +\\n alpha_l * image[y1:y2, x1:x2, c])\\n except ValueError:\\n image = image_save\\n\\n self._display_frame(image)\",\n \"def render(self, window):\\n body = pygame.image.load(IMAGE_SNAKE).convert_alpha() # loading image\\n for block in self.body:\\n window.blit(body, (block[0]*SPRITE_SIZE, block[1]*SPRITE_SIZE)) # painting a beautiful snek\\n if self.neural_net: # calls for neural net rendering\\n self.neural_net.render(window, self.vision)\",\n \"def render_and_save():\\n\\n rendering_config = configuration.get_config()\\n rendering_config = ml_collections.FrozenConfigDict(rendering_config)\\n aspect_ratio = rendering_config.aspect_ratio\\n height = rendering_config.height\\n width = int(aspect_ratio * height)\\n\\n scene_camera = build_camera(rendering_config, aspect_ratio)\\n world = build_world(rendering_config)\\n\\n # Render.\\n logging.info(\\\"Tracing rays...\\\")\\n render_image_fn = jax.jit(\\n render.generate_image,\\n static_argnames=[\\\"height\\\", \\\"width\\\", \\\"config\\\"])\\n image = render_image_fn(height, width, scene_camera, world, rendering_config)\\n\\n image = render.correct_gamma(image, gamma=rendering_config.gamma_correction)\\n\\n logging.info(\\\"Saving to file...\\\")\\n output.export_as_ppm(image, rendering_config.output_file)\\n\\n return image\",\n \"def draw(self):\\n self.screen.fill((0,51,102))\\n # get the new drawables\\n self.drawables = (self.game_model.get_background_drawables()\\n + self.game_model.get_plane_drawables()\\n + self.game_model.get_bullet_drawables()\\n + self.game_model.get_enemy_drawables())\\n for d in self.drawables:\\n rect = d.get_rect()\\n surf = d.get_surface()\\n surf.set_colorkey((255,255,255))\\n self.screen.blit(surf, rect)\",\n \"def draw(self, surface):\\n temp = pygame.Surface(self.renderer.pixel_size)\\n self.renderer.render_map(temp)\\n pygame.transform.smoothscale(temp, surface.get_size(), surface)\",\n \"def draw(self):\",\n \"def draw(self, surface):\\n surface.blit(self.image, (0,0))\\n for widget in self.widgets:\\n widget.draw(surface)\",\n \"def main():\\n parser = argparse.ArgumentParser(description=\\\"Process a wavefront object file\\\")\\n parser.add_argument('--width', help=\\\"Width of output image\\\", dest='width',\\n type=int, default=800)\\n parser.add_argument('--height', help=\\\"Height of output image\\\", dest='height',\\n type=int, default=600)\\n parser.add_argument('--out', help=\\\"Name of output image file\\\", dest='output', type=str,\\n default='output.png')\\n parser.add_argument('filename', help=\\\"Alias Wavefront file to read as input\\\", type=str)\\n args = parser.parse_args()\\n\\n back_buffer = Buffer(args.width, args.height)\\n screen = Screen(back_buffer)\\n\\n obj = Wavefront(args.filename)\\n print(\\\"Processing {}\\\".format(args.filename))\\n\\n max_extent = max(obj.v_extent[0], obj.v_extent[1])\\n scale = min(args.width / max_extent, args.height / max_extent)\\n translate = ((args.width / 2) + obj.v_min[0] + obj.v_max[0],\\n (args.height / 2) + obj.v_min[1] + obj.v_max[1])\\n print(\\\"Using scale: {}\\\".format(scale))\\n\\n def conv_x(x):\\n \\\"\\\"\\\" Converts x coordinate.\\\"\\\"\\\"\\n return int(x * scale + translate[0])\\n def conv_y(y):\\n \\\"\\\"\\\" Converts y coordinate.\\\"\\\"\\\"\\n return int(y * scale + translate[1])\\n\\n for f in obj.f:\\n v1 = obj.v[f[0]]\\n v2 = obj.v[f[1]]\\n v3 = obj.v[f[2]]\\n screen.draw_line((conv_x(v1[0]), conv_y(v1[1])),\\n (conv_x(v2[0]), conv_y(v2[1])),\\n (255, 255, 255))\\n screen.draw_line((conv_x(v2[0]), conv_y(v2[1])),\\n (conv_x(v3[0]), conv_y(v3[1])),\\n (255, 255, 255))\\n screen.draw_line((conv_x(v3[0]), conv_y(v3[1])),\\n (conv_x(v1[0]), conv_y(v1[1])),\\n (255, 255, 255))\\n\\n print(\\\"Minimum vector: {}\\\".format(obj.v_min))\\n print(\\\"Maximum vector: {}\\\".format(obj.v_max))\\n print(\\\"Calculated extent: {}\\\".format(obj.v_extent))\\n\\n back_buffer.write_to_png(args.output)\\n print(\\\"Written output to {}\\\".format(args.output))\",\n \"def draw(self):\\n return ImageDraw.Draw(self.buffer)\",\n \"def renderCap( self, scale = 400.0):\\n vertices = [\\n vertex.Vertex( (x/scale,y/scale,0) ) for (x,y) in outline[::2]\\n ]\\n vertices.reverse()\\n for type, vertices in self.tess.tessellate( vertices, forceTriangles=0 ):\\n glNormal( 0,0,-1 )\\n glColor3f( 1,0,0 )\\n glBegin( type )\\n for v in vertices:\\n glVertex2dv( v.point[:2] )\\n glEnd()\",\n \"def draw(self):\\n self.batch.draw()\",\n \"def render(pictures):\\n # Display the background\\n display.fill(BLACK)\\n\\n # Draw each picture\\n i = 0 # Number of pictures drawn\\n for picture in pictures:\\n offsets = get_offsets(i)\\n\\n for polygon in picture.polygons:\\n # Adjust vertices so pictures aren't overlapping\\n adjusted_vertices = []\\n\\n for v in polygon.vertices:\\n adjusted_vertices.append((v[0] + offsets[0], v[1] + offsets[1]))\\n pygame.draw.polygon(display, polygon.color, adjusted_vertices)\\n\\n i += 1\\n\\n pygame.display.update()\",\n \"def draw_layers(self):\\n\\t\\tfor z in xrange(0,16):\\n\\t\\t\\t#create surface for this layer\\n\\t\\t\\tsrf = pygame.Surface((16,128))\\n\\t\\t\\tfor x in xrange(0,16):\\n\\t\\t\\t\\tfor y in xrange(0,128):\\n\\t\\t\\t\\t\\tv = self.data[ self.xyz_to_offset( x,y,z) ]\\n\\t\\t\\t\\t\\tif v != 0:\\n\\t\\t\\t\\t\\t\\tsrf.fill( BLOCKS.get(v, [0,0])[1], \\t(x, 127 -y, 1, 1 ))\\n\\t\\t\\t#save layer to dict for this chunk\\n\\t\\t\\tself.layers[z] = srf\",\n \"def drawImage(self):\\n \\n self.retrieveThemeConfig()\\n \\n self.initImg()\\n self.colsDone = 1\\n \\n queue = mp.Queue()\\n procs = []\\n for i in range(0,self.strandNum):\\n procs.append(mp.Process(target=self.strandWorker,args=(queue,i,\\n self.branches)))\\n procs[i].start()\\n\\n self.allCords = []\\n for i in range(0,self.strandNum):\\n self.allCords.append(queue.get())\\n \\n logging.info(\\\"Strand workers finished\\\")\\n\\n self.curCol = 0\\n for k in range(0,len(self.allCords)):\\n for l in range(0,len(self.allCords[k])):\\n for m in range(0,len(self.allCords[k][l]['cords'])):\\n self.draw.line(self.allCords[k][l]['cords'][m],\\n self.allCords[k][l]['col'],self.thickness)\\n del self.draw\\n self.exportImg()\",\n \"def draw(self, frame):\\n for prediction in self.predictions:\\n prediction.draw(frame)\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n\\t\\tpass\",\n \"def on_draw_over_image(self):\",\n \"def extract_intrinsic_images_from_lighting_passes(self, outdir, vis=False):\\n logger_name = thisfile + '->extract_intrinsic_images_from_lighting_passes()'\\n xm.general.makedirs(outdir)\\n data = self.data\\n\\n def collapse_passes(components):\\n ch_arrays = []\\n for ch in ['R', 'G', 'B']:\\n comp_arrs = []\\n for comp in components:\\n comp_arrs.append(data[comp + '.' + ch])\\n ch_array = np.sum(comp_arrs, axis=0) # sum components\\n ch_arrays.append(ch_array)\\n # Handle alpha channel\\n first_alpha = data[components[0] + '.A']\\n for ci in range(1, len(components)):\\n assert (first_alpha == data[components[ci] + '.A']).all(), \\\\\\n \\\"Alpha channels of all passes must be the same\\\"\\n ch_arrays.append(first_alpha)\\n return np.dstack(ch_arrays)\\n\\n # Albedo\\n albedo = collapse_passes(['diffuse_color', 'glossy_color'])\\n np.save(join(outdir, 'albedo.npy'), albedo)\\n if vis:\\n xm.vis.matrix_as_image(albedo, outpath=join(outdir, 'albedo.png'))\\n # Shading\\n shading = collapse_passes(['diffuse_indirect', 'diffuse_direct'])\\n np.save(join(outdir, 'shading.npy'), shading)\\n if vis:\\n xm.vis.matrix_as_image(shading, join(outdir, 'shading.png'))\\n # Specularity\\n specularity = collapse_passes(['glossy_indirect', 'glossy_direct'])\\n np.save(join(outdir, 'specularity.npy'), specularity)\\n if vis:\\n xm.vis.matrix_as_image(specularity, join(outdir, 'specularity.png'))\\n # Reconstruction vs. ...\\n recon = np.multiply(albedo, shading) + specularity\\n recon[:, :, 3] = albedo[:, :, 3] # can't add up alpha channels\\n np.save(join(outdir, 'recon.npy'), recon)\\n if vis:\\n xm.vis.matrix_as_image(recon, join(outdir, 'recon.png'))\\n # ... composite from Blender, just for sanity check\\n composite = collapse_passes(['composite'])\\n np.save(join(outdir, 'composite.npy'), composite)\\n if vis:\\n xm.vis.matrix_as_image(composite, join(outdir, 'composite.png'))\\n logger.name = logger_name\\n logger.info(\\\"Intrinsic images extracted to %s\\\", outdir)\",\n \"def on_draw( self ):\\r\\n self.clear()\\r\\n self.setup_3D()\\r\\n print \\\"DEBUG:\\\" , \\\"There are\\\" , len( self.renderlist ) , \\\"items in 'self.renderlist'\\\"\\r\\n for obj in self.renderlist:\\r\\n obj.draw()\",\n \"def draw(font_string,font_size,lang,alphabets,outdir=\\\".\\\"): # language, font file name, font full path, font size, characters\\n \\n \\n image_dir=lang+\\\".\\\"+\\\"images\\\"\\n if(os.path.exists(image_dir)):\\n pass\\n else:\\n os.mkdir(image_dir)\\n \\n #Using a font\\n #font= ImageFont.truetype(font,fsz)\\n boxfile=image_dir+\\\"/\\\"+\\\"bigimage.box\\\"\\n f=open(boxfile,\\\"w\\\")\\n wt = 4000\\n ht = 4000 #modified later using a separate script\\n\\t\\n bigimage=Image.new(\\\"L\\\",(wt,ht),255)\\t#change here for inverting\\n bigdraw=ImageDraw.Draw(bigimage)\\n x=y=10\\n count=0\\n for akshar in alphabets:\\n akshar.strip() #remove nasty characters\\n \\n #I shall now create an image with black bgc and white font color. One\\n #getbbox() determines the bounding box values I shall invert the image.\\n #This has to be done since getbbox() only finds bounding box values for\\n #non-zero pixels (read as white), but tesseract-ocr runs on the exact\\n #opposite bgc fgc combination. Contact debayanin@gmail.com.\\n \\n \\n #The lines below are pango/cairo code \\n surface = cairo.ImageSurface(cairo.FORMAT_A8, font_size*4, font_size*3)\\n context = cairo.Context(surface)\\n\\n pc = pangocairo.CairoContext(context)\\n\\n layout = pc.create_layout()\\n layout.set_font_description(pango.FontDescription(font_string))\\n layout.set_text(akshar)\\n print akshar\\n\\n # lines take care of centering the text.\\n width, height = surface.get_width(), surface.get_height()\\n w, h = layout.get_pixel_size()\\n position = (10,10) #most likely this part messes up when you try to change the size within this script. It is suggested to use the separate script.\\n context.move_to(*position)\\n pc.show_layout(layout)\\n surface.write_to_png(\\\"pango.png\\\")\\n\\t\\n #Here we open the generated image using PIL functions\\n temp_image=Image.open(\\\"pango.png\\\") #black background, white text\\n draw = ImageDraw.Draw(temp_image)\\n bbox = temp_image.getbbox()\\n deltax=bbox[2]-bbox[0]\\n deltay=bbox[3]-bbox[1]\\n\\n \\n print bbox\\n new_image=temp_image.crop(bbox)\\n temp_image=temp_image.load()\\n inverted_image = ImageChops.invert(new_image) #White background, black text\\n\\t\\n\\tinverted_image.save(image_dir+\\\"/\\\"+str(count)+\\\".png\\\")\\n\\tbigimage.paste(inverted_image,(x,y))\\n\\tos.unlink(image_dir+\\\"/\\\"+str(count)+\\\".png\\\")\\n\\tcount = count+1\\n\\t#bigimage.load()\\n bigbox=(x,y,x+deltax,y+deltay)\\n print bigbox\\n draw=ImageDraw.Draw(bigimage)\\n\\t#draw.rectangle(bigbox,None,100)\\n x=bigbox[2]+5\\n if x>(wt-10):\\n x=10; y=y+40\\n\\n os.unlink(\\\"pango.png\\\") #delete the pango generated png\\n\\n line=akshar+\\\" \\\"+str(bigbox[0]-1)+\\\" \\\"+str(ht-(bigbox[1]+deltay)-1)+\\\" \\\"+str(bigbox[2]+1)+\\\" \\\"+str(ht-(bigbox[3]-deltay)+1) # this is the line to be added to the box file\\n\\tf.write(line+'\\\\n')\\n\\n\\t#degrade code starts\\n\\tstrip=[deltax*.2,deltax*.4,deltax*.7]\\n\\tfor values in range(0,2):\\n\\t\\tdistort2=inverted_image\\n\\t\\tfor wai in range(0,deltay):\\n\\t\\t\\tfor ex in range(strip[values],strip[values]+1):\\n\\t\\t\\t\\tdistort2.putpixel((ex,wai),255)\\n\\t\\tbigbox=(x,y,x+deltax,y+deltay)\\n\\t\\t#draw.rectangle(bigbox,None,10)\\n\\t\\tline=akshar+\\\" \\\"+str(bigbox[0]-1)+\\\" \\\"+str(ht-(bigbox[1]+deltay)-1)+\\\" \\\"+str(bigbox[2]+1)+\\\" \\\"+str(ht-(bigbox[3]-deltay)+1) # this is the line to be added to the box file\\n \\tf.write(line+'\\\\n')\\n\\t\\tbigimage.paste(distort2,(x,y))\\n\\t\\tx=bigbox[2]+5\\n \\tif x>(wt-10):\\n \\t\\tx=10; y=y+40\\n\\t\\t\\n\\t\\t\\t\\n\\t#degrade code ends\\n \\n #distort.distort(filename2,bbox,fsz,akshar)\\n \\n \\n \\n #bigimage.save(image_dir+\\\"/\\\"+\\\"bigimage.tif\\\",\\\"TIFF\\\") #useful to generate merged file for all images when using default sizes.\\n f.close()\\n train.train(lang,outdir)\",\n \"def _drawOnCanvas(self):\\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\\n for key in self.elements:\\n graphElement=self.elements[key]\\n graphElement.draw(self.canvas)\\n self.sync=True\",\n \"def draw(self, view):\\n for r in self._aliens:\\n for alien in r:\\n if alien != None:\\n alien.draw(view)\\n if self._ship != None:\\n self._ship.draw(view)\\n self._dline.draw(view)\\n for bolt in self._bolts:\\n bolt.draw(view)\",\n \"def draw(self):\\r\\n scalex,scaley = self.getScale()\\r\\n try:\\r\\n self.clear()\\r\\n # Draw Graph Background\\r\\n self.drawLayout()\\r\\n if self.app.data == None:# If no data, break\\r\\n return\\r\\n # How much each pixel represents\\r\\n if scalex[1]-scalex[0] == 0:\\r\\n return\\r\\n step = (scalex[1]-scalex[0])/self.w# Draw lines at pixel level resolution\\r\\n self.fitYScale()\\r\\n sens_index = [0]# If one sensor displayed in this data player\\r\\n if len(self.sensor_ids) == 2:# If two sensors displayed in this data player\\r\\n sens_index = [1,0]# Draw order blue then red to make blue line on top\\r\\n for s in sens_index:\\r\\n i = scalex[0]\\r\\n x = 0\\r\\n trackcol = self.app.getSensorCol(self.sensors[self.sensor_ids[s]])\\r\\n while i < scalex[1]:\\r\\n i += step# i Is data\\r\\n x += 1# x is iteration/pixel-coordinate\\r\\n if i<0:# Skip data for t<0\\r\\n continue\\r\\n try:\\r\\n # Data retrieved from xml\\r\\n y = float(self.app.data[int(i)][self.sensor_ids[s]].text)\\r\\n y2 = float(self.app.data[int(i+step)][self.sensor_ids[s]].text)\\r\\n # Normalize into range 0 to 1 and multiply by height\\r\\n y = ((y-scaley[0])/(scaley[1]-scaley[0])) * self.h\\r\\n y2 = ((y2-scaley[0])/(scaley[1]-scaley[0])) * self.h\\r\\n except IndexError:# Missing data is skipped\\r\\n continue\\r\\n self.c.create_line(x,-y+self.h,x+1,-y2+self.h,fill=trackcol,width=1)\\r\\n self.drawScrubber()\\r\\n self.drawPeekScrubber()\\r\\n self.c.update()\\r\\n except tk.TclError:# If canvas destroyed, cancel draw operation\\r\\n return\",\n \"def draw(self, surface):\\n surface.blit(self.image, self.rect)\",\n \"def draw(self, surface):\\n surface.blit(self.image, self.rect)\",\n \"def draw(self, surface):\\r\\n surface.blit(self.image, self.rect)\",\n \"def draw(iiter):\\n from matplotlib import pyplot as plt\\n fig = plt.gcf()\\n fig.canvas.draw()\",\n \"def draw(self):\\n\\t\\tif self.image != None:\\n\\t\\t\\treturn\\n\\n\\t\\tself.image = Image.new(\\\"RGBA\\\", \\n\\t\\t\\t(self.__class__._width, self.__class__._height), (0,0,0,0))\\n\\t\\tdr = aggdraw.Draw(self.image)\\n\\n\\t\\tzIndexList = [] # list of z-indexes\\n\\t\\tgIndexList = [] # list of gene indexes\\n\\n\\t\\tgeneIdx = 0\\n\\t\\tfor gene in self.chromosome: \\n\\t\\t\\tzIndexList.append(gene[0])\\n\\t\\t\\tgIndexList.append(geneIdx)\\n\\t\\t\\tgeneIdx+=1\\n\\n\\t\\twhile len(zIndexList) > 0:\\n\\t\\t\\tzIndexMin = zIndexList.index(min(zIndexList))\\n\\t\\t\\tgeneIdx = gIndexList[zIndexMin]\\n\\t\\t\\tz = self.chromosome[geneIdx,0]\\n\\t\\t\\tcolor = self.chromosome[geneIdx,1]\\n\\t\\t\\tpoly = self.chromosome[geneIdx,2]\\n\\n\\t\\t\\tcol = color.getColor()\\n\\t\\t\\tcords = poly.getCords()\\n\\t\\t\\tbrush = aggdraw.Brush(col)\\n\\n\\t\\t\\t#dr.ellipse((x, y, x+rad, y+rad), None, brush)\\n\\t\\t\\tdr.polygon(cords, None, brush)\\n\\t\\t\\tdr.flush()\\n\\n\\t\\t\\tif 0: # Debug - draw dot in center of polygon\\n\\t\\t\\t\\tcenter = (poly.xAnchor, poly.yAnchor, poly.xAnchor+5, poly.yAnchor+5)\\n\\t\\t\\t\\tdr.ellipse(center, None, aggdraw.Brush((255,0,0,255)))\\n\\t\\t\\t\\tdr.flush()\\n\\n\\t\\t\\tzIndexList.pop(zIndexMin)\\n\\t\\t\\tgIndexList.pop(zIndexMin)\\n\\n\\t\\tself.image = self.image.convert(\\\"RGB\\\")\",\n \"def draw(self, surface):\\n for box in self.checkboxes:\\n box.draw(surface)\",\n \"def generate_images_pred(self, inputs, outputs):\\n for scale in self.scales:\\n disp = outputs[(\\\"disp\\\", scale)]\\n disp = F.interpolate(\\n disp, [self.height, self.width], mode=\\\"bilinear\\\", align_corners=False)\\n source_scale = 0\\n\\n _, depth = disp_to_depth(disp, self.min_depth, self.max_depth)\\n\\n outputs[(\\\"depth\\\", 0, scale)] = depth\\n\\n for i, frame_id in enumerate(self.frame_ids[1:]):\\n\\n T = outputs[(\\\"cam_T_cam\\\", 0, frame_id)]\\n\\n # from the authors of https://arxiv.org/abs/1712.00175\\n # mean-normalized inverse depth from [62] to discourage shrinking of the estimated depth\\n\\n axisangle = outputs[(\\\"axisangle\\\", 0, frame_id)]\\n translation = outputs[(\\\"translation\\\", 0, frame_id)]\\n\\n inv_depth = 1 / depth\\n mean_inv_depth = inv_depth.mean(3, True).mean(2, True)\\n\\n T = transformation_from_parameters(\\n axisangle[:, 0], translation[:, 0] * mean_inv_depth[:, 0], frame_id < 0)\\n\\n cam_points = self.backproject_depth[source_scale](\\n depth, inputs[(\\\"inv_K\\\", source_scale)])\\n pix_coords = self.project_3d[source_scale](\\n cam_points, inputs[(\\\"K\\\", source_scale)], T)\\n\\n outputs[(\\\"sample\\\", frame_id, scale)] = pix_coords\\n\\n outputs[(\\\"color\\\", frame_id, scale)] = F.grid_sample(\\n inputs[(\\\"color\\\", frame_id, source_scale)],\\n outputs[(\\\"sample\\\", frame_id, scale)],\\n padding_mode=\\\"border\\\")\\n\\n outputs[(\\\"color_identity\\\", frame_id, scale)] = \\\\\\n inputs[(\\\"color\\\", frame_id, source_scale)]\",\n \"def on_render(self):\\n\\n # set text font for in visualization\\n pygame.font.init()\\n font = pygame.font.SysFont('Arial', 24)\\n\\n # continue running algorithm until done\\n if self.algorithm.isDone is False:\\n self.algorithm.execute()\\n\\n # draw background\\n for i in range(0, 3):\\n for j in range(0, 3):\\n self.screen.blit(self.img_grass, (i * 270, j * 270))\\n\\n # get water instances to draw and draw them\\n allWatersList = []\\n allWatersList.extend(self.area.allWatersList)\\n\\n for water in allWatersList:\\n pygame.draw.rect(\\n self.screen, (0, 0, 128),\\n (water.x * 2,\\n water.y * 2,\\n water.width * 2,\\n water.height * 2)\\n )\\n pygame.draw.rect(\\n self.screen,\\n (0, 0, 0),\\n (0, self.height-50, self.width, 50)\\n )\\n\\n # get house instances to draw and draw them\\n housesToPlace = []\\n housesToPlace.extend(self.area.mansionList)\\n housesToPlace.extend(self.area.familyHomeList)\\n housesToPlace.extend(self.area.bungalowList)\\n\\n for house in housesToPlace:\\n # draw free space\\n space = pygame.Surface((house.space * 4 + house.width * 2,\\n house.space * 4 + house.height * 2))\\n space.set_alpha(64)\\n space.fill((180, 0, 0))\\n self.screen.blit(space,\\n (house.x * 2 - house.space * 2,\\n house.y * 2 - house.space * 2))\\n\\n # draw minimum free space\\n space = pygame.Surface((house.minimumSpace * 4 + house.width * 2,\\n house.minimumSpace * 4 + house.height * 2))\\n space.set_alpha(110)\\n space.fill((100, 0, 0))\\n self.screen.blit(space,\\n (house.x * 2 - house.minimumSpace * 2,\\n house.y * 2 - house.minimumSpace * 2))\\n\\n for house in housesToPlace:\\n # draw house, colored based on type\\n kind = type(house).__name__\\n if kind == \\\"Mansion\\\":\\n pygame.draw.rect(self.screen,\\n (200, 255, 40),\\n (house.x * 2,\\n house.y * 2,\\n house.width * 2,\\n house.height * 2))\\n elif kind == \\\"Bungalow\\\":\\n pygame.draw.rect(self.screen,\\n (255, 40, 200),\\n (house.x * 2,\\n house.y * 2,\\n house.width * 2,\\n house.height * 2))\\n elif kind == \\\"FamilyHome\\\":\\n pygame.draw.rect(self.screen,\\n (0, 255, 0),\\n (house.x * 2,\\n house.y * 2,\\n house.width * 2,\\n house.height * 2))\\n\\n # Draw black bar at bottom of screen for extra info\\n pygame.draw.rect(self.screen,\\n (0, 0, 0),\\n (0, self.height - 50, self.width, 50))\\n\\n # Draw area value and last value increase in infobox\\n textSurface = font.render('Score: '\\n + str(self.area.price),\\n True, (255, 255, 255))\\n self.screen.blit(textSurface, (10, self.height-35))\\n\\n # create distinct color for value decreases\\n increaseColor = (255, 255, 255)\\n if (self.area.price - self.lastPrice < 1):\\n increaseColor = (80, 80, 80)\\n textSurface = font.render('Increase: ' +\\n str(self.area.price - self.lastPrice),\\n True, increaseColor)\\n self.screen.blit(textSurface, (330, self.height-35))\\n pygame.draw.rect(self.screen,\\n (0, 0, 0),\\n (640, 0, 400, self.height))\\n\\n # save area values to draw graph\\n if (self.area.price >= self.lastPrice and\\n self.algorithm.isDone is False and\\n self.showDecrease is False):\\n\\n self.scores.append(self.area.price)\\n self.lastPrice = self.area.price\\n\\n if self.showDecrease is True and self.algorithm.isDone is False:\\n self.scores.append(self.area.price)\\n self.lastPrice = self.area.price\\n\\n # draw graph with area values\\n fig = pylab.figure(figsize=[4, 4], # Inches\\n dpi=100) # 100 dots per inch\\n ax = fig.gca()\\n ax.plot(self.scores)\\n\\n canvas = agg.FigureCanvasAgg(fig)\\n canvas.draw()\\n renderer = canvas.get_renderer()\\n raw_data = renderer.tostring_rgb()\\n\\n surf = pygame.image.fromstring(raw_data, (400, 400), \\\"RGB\\\")\\n self.screen.blit(surf, (640, 0))\\n matplotlib.pyplot.close(fig)\\n\\n # Draw all time highest score if that's set\\n if self.allTimeHigh is not 0:\\n textSurface = font.render('Highest score: ' +\\n str(self.allTimeHigh),\\n True, (255, 255, 255))\\n self.screen.blit(textSurface, (650, 410))\\n\\n pygame.display.flip()\\n pass\",\n \"def draw(self, **kwargs):\\n for o in sorted(self._drawables, key=default_itemgetter(\\\"z\\\", default=0)):\\n o.draw(**kwargs)\",\n \"def _render(self):\\n self.dirty = False\\n self.image = self.font.render(self._text, self.aa, self.color_fg)\\n self.rect = self.image.get_rect()\",\n \"def blit(self):\\n # self.ax1.draw_artist(self.lines[:2])\\n # self.ax2.draw_artist(self.lines[2:4])\\n self.ax1.autoscale()\\n self.ax2.autoscale()\\n self.ax1.redraw_in_frame()\\n self.ax2.redraw_in_frame()\\n self.fig.canvas.blit(self.fig.bbox)\",\n \"def rasterize(self):\\n\\n for primitive in self._scene:\\n bbox = primitive[\\\"bounding_box\\\"]\\n # Loop through all pixels\\n # You MUST use bounding boxes in order to speed up this loop\\n for w in range(bbox[0][0], bbox[1][0]):\\n x = w + 0.5\\n for h in range(bbox[0][1], bbox[1][1]):\\n y = h + 0.5\\n # First, we check if the pixel center is inside the primitive\\n im_x, im_y = w, self._height - (h + 1)\\n if inside(x, y, primitive):\\n # apply affine xfrom if needed\\n if \\\"xform\\\" in primitive.keys():\\n result = np.matmul(primitive[\\\"xform\\\"],\\n [[im_x], [im_y], [1]])\\n im_x, im_y = int(result[0][0]), int(result[1][0])\\n\\n self._image[im_y, im_x] = primitive[\\\"color\\\"]\\n # break\\n # break\\n # break\",\n \"def draw(self, background=COLOUR_BLACK, show=False, save=False,\\n generation=None):\\n size = self.img_size\\n img = Image.new('RGB', size, background)\\n draw = Image.new('RGBA', size)\\n pdraw = ImageDraw.Draw(draw)\\n for polygon in self.polygons:\\n colour = polygon.colour\\n points = polygon.points\\n pdraw.polygon(points, fill=colour, outline=colour)\\n img.paste(draw, mask=draw)\\n\\n if show:\\n img.show()\\n\\n if save:\\n # TODO use self.generation\\n temp_dir = tempfile.gettempdir()\\n temp_name = u\\\"0000000000{}\\\".format(generation)[-10:]\\n out_path = u\\\"{}/{}.png\\\".format(temp_dir, temp_name)\\n img = img.filter(ImageFilter.GaussianBlur(radius=3))\\n img.save(out_path)\\n print u\\\"saving image to {}\\\".format(out_path)\\n\\n return img\",\n \"def wrapped_render(self, render_modes, front_and_back=False):\\n\\n # Render raw images\\n render_color = False\\n for mode in render_modes:\\n if mode != RenderMode.DEPTH and mode != RenderMode.SCALED_DEPTH:\\n render_color = True\\n break\\n\\n color_im, depth_im = None, None\\n if render_color:\\n color_im, depth_im = self.render(render_color, front_and_back=front_and_back)\\n else:\\n depth_im = self.render(render_color)\\n\\n # For each specified render mode, add an Image object of the appropriate type\\n images = []\\n for render_mode in render_modes:\\n # Then, convert them to an image wrapper class\\n if render_mode == RenderMode.SEGMASK:\\n images.append(BinaryImage((depth_im > 0.0).astype(np.uint8), frame=self.camera.intrinsics.frame, threshold=0))\\n\\n elif render_mode == RenderMode.COLOR:\\n images.append(ColorImage(color_im, frame=self.camera.intrinsics.frame))\\n\\n elif render_mode == RenderMode.GRAYSCALE:\\n images.append(ColorImage(color_im, frame=self.camera.intrinsics.frame).to_grayscale())\\n\\n elif render_mode == RenderMode.DEPTH:\\n images.append(DepthImage(depth_im, frame=self.camera.intrinsics.frame))\\n\\n elif render_mode == RenderMode.SCALED_DEPTH:\\n images.append(DepthImage(depth_im, frame=self.camera.intrinsics.frame).to_color())\\n\\n elif render_mode == RenderMode.RGBD:\\n c = ColorImage(color_im, frame=self.camera.intrinsics.frame)\\n d = DepthImage(depth_im, frame=self.camera.intrinsics.frame)\\n images.append(RgbdImage.from_color_and_depth(c, d))\\n\\n elif render_mode == RenderMode.GD:\\n g = ColorImage(color_im, frame=self.camera.intrinsics.frame).to_grayscale()\\n d = DepthImage(depth_im, frame=self.camera.intrinsics.frame)\\n images.append(GdImage.from_grayscale_and_depth(g, d))\\n else:\\n raise ValueError('Render mode {} not supported'.format(render_mode))\\n\\n return images\",\n \"def on_draw(self, widget, cr):\\n #print \\\"starting to draw\\\"\\n if self.double_buffer is not None:\\n self.draw_tiles()\\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\\n cr.paint()\\n else:\\n print('Invalid double buffer')\\n #print \\\"done drawing\\\"\\n return False\",\n \"def draw(self):\\n self.figure.canvas.draw_idle()\",\n \"def draw_kame(self):\\r\\n #pygame.draw.rect(self.screen, self.color, self.rect, self.image)\\r\\n self.screen.blit(self.image, self.rect)\",\n \"def on_draw(self):\\n # draw everything\",\n \"def example_BSR():\\n pts = [(1,1),(2,2),(3,3)]\\n lines = [ [ (1,1), (1,2), (2,1)], [ (6,1), (1,6), (5,-1)] ]\\n\\n bloody_simple_2drender('2d_render.png', pts=pts, vecs=pts, lines=lines )\",\n \"def visualize(self):\\n colors = {'outline': (220, 220, 220),\\n 'inlier': (0, 255, 0),\\n 'outlier': (0, 0, 255),\\n 'lines': (128, 220, 128)}\\n # Create output image for visualization\\n gap = 5\\n h1, w1 = self.target.image.shape[:2]\\n h2, w2 = self.image.shape[:2]\\n vis = np.zeros((max(h1, h2), w1 + w2 + gap, 3), np.uint8)\\n vis[:h1, :w1, :] = self.target.image\\n w1 += gap\\n vis[:h2, w1:w1+w2, :] = self.image\\n \\n # Draw the located object \\n quad = np.float32(self.quad) + np.float32([w1, 0])\\n self.draw(vis, colors['outline'], 2, quad)\\n \\n # draw point details\\n inliers = [(x0, y0, x1 + w1, y1) for (x0, y0), (x1, y1) in self.inliers]\\n outliers = [(x0, y0, x1 + w1, y1) for (x0, y0), (x1, y1) in self.outliers]\\n if colors['outlier'] is not None: # draw x on each point\\n r = 2 # radius\\n thickness = 2\\n for x0, y0, x1, y1 in outliers:\\n cv2.line(vis, (x0 - r, y0 - r), (x0 + r, y0 + r), colors['outlier'], thickness)\\n cv2.line(vis, (x0 + r, y0 - r), (x0 - r, y0 + r), colors['outlier'], thickness)\\n cv2.line(vis, (x1 - r, y1 - r), (x1 + r, y1 + r), colors['outlier'], thickness)\\n cv2.line(vis, (x1 + r, y1 - r), (x1 - r, y1 + r), colors['outlier'], thickness)\\n if colors['lines'] is not None:\\n for x0, y0, x1, y1 in inliers:\\n cv2.line(vis, (x0, y0), (x1, y1), colors['lines'], 1)\\n if colors['inlier'] is not None:\\n for x0, y0, x1, y1 in inliers:\\n cv2.circle(vis, (x0, y0), 2, colors['inlier'], -1)\\n cv2.circle(vis, (x1, y1), 2, colors['inlier'], -1)\\n return vis\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def on_draw(self):\\n\\n # Clear the screen and start drawing\\n arcade.start_render()\\n\\n # Draw the rectangles\\n for shape in self.shapes:\\n shape.draw()\",\n \"def blending_example1():\\n pic_desert = read_image(relpath(\\\"./externals/pic_desert.jpg\\\"), 2)\\n pic_pool = read_image(relpath(\\\"./externals/pic_swim.jpg\\\"), 2)\\n mask = read_image(relpath(\\\"./externals/mask_desert.jpg\\\"), 1)\\n # making the mask binary (normalizing 2 original values)\\n mask = strech_helper(mask).astype(np.bool)\\n print(pic_desert.shape[2])\\n [R1, G1, B1] = np.dsplit(pic_desert, pic_desert.shape[2])\\n [R2, G2, B2] = np.dsplit(pic_pool, pic_pool.shape[2])\\n R1 = np.reshape(R1, (512,1024))\\n R2 = np.reshape(R2, (512,1024))\\n G1 = np.reshape(G1, (512,1024))\\n G2 = np.reshape(G2, (512,1024))\\n B1 = np.reshape(B1, (512,1024))\\n B2 = np.reshape(B2, (512,1024))\\n\\n blend1 = pyramid_blending(R2, R1, mask, 3, 3, 3)\\n blend2 = pyramid_blending(G2, G1, mask, 3, 3, 3)\\n blend3 = pyramid_blending(B2, B1, mask, 3, 3, 3)\\n\\n blend1 = np.reshape(blend1, (blend1.shape[0], blend1.shape[1], 1))\\n blend2 = np.reshape(blend2, (blend2.shape[0], blend3.shape[1], 1))\\n blend3 = np.reshape(blend3, (blend3.shape[0], blend3.shape[1], 1))\\n\\n new_pic = np.concatenate((blend1, blend2, blend3), axis=2)\\n # plotting the images\\n fig = plt.figure()\\n ax1 = fig.add_subplot(221)\\n ax2 = fig.add_subplot(222)\\n ax3 = fig.add_subplot(223)\\n ax4 = fig.add_subplot(224)\\n ax1.imshow(pic_desert)\\n ax2.imshow(pic_pool)\\n ax3.imshow(mask, cmap='gray')\\n ax4.imshow(new_pic)\\n plt.show()\\n\\n return pic_desert, pic_pool, mask, new_pic\",\n \"def draw(self):\\n for section in self.sections:\\n canvas_reset(self.canvas)\\n section.draw(self.canvas)\",\n \"def draw(self):\\r\\n self.screen.blit(self.image, self.image.get_rect())\",\n \"def render_foreground(img: Image) -> Image:\\n canvas = Image.new(\\\"RGBA\\\", img.size)\\n shadow = img.filter(ImageFilter.GaussianBlur(4))\\n shadow = invert_with_alpha(shadow, 120)\\n\\n canvas.paste(shadow, (4, 4))\\n canvas = Image.alpha_composite(canvas, img)\\n return canvas\",\n \"def test_draw(self):\\n image_name = filename(sys._getframe().f_code.co_name)\\n result_file, reference_file = get_path(image_name)\\n\\n ''' This function is to create an empty image with a specific dimension\\n with white background, and black/white colored '''\\n\\n image, canvas = get_image('L', (15,90),'white')\\n\\n for i in range(len(draw_points) - 1):\\n draw(canvas, (draw_points[i + 0], draw_points[i + 1]), 'A')\\n\\n \\\"\\\"\\\" saving the file and closing it \\\"\\\"\\\"\\n\\n image.save(result_file)\\n image.close()\\n\\n \\\"\\\"\\\" validate the resultant file against the reference images\\\"\\\"\\\"\\n\\n validate(reference_file, result_file)\",\n \"def filter(self, drawable):\\n pass\",\n \"def draw(self, grille):\\n grille.clear_highlight()\\n for x in range(8):\\n for y in range(8):\\n self.draw_c(x, y, grille)\",\n \"def render(self, rstate):\\n if rstate.picking:\\n for textures, group in self.texGroupItems:\\n group.pickingDraw(rstate)\\n else:\\n for textures, group in self.texGroupItems:\\n if not rstate.picking:\\n try:\\n self.bindTextures(textures, rstate)\\n except Texture.DynamicTextureException:\\n # The textures this drawable requires aren't ready yet\\n return\\n group.draw(rstate)\",\n \"def draw (self):\\n screen = self.screen\\n dirty = False\\n for z, displays in self.layers.iteritems():\\n for display in displays:\\n drew = display.draw(screen)\\n # if made changes to the surface\\n if drew:\\n # set any displays that overlap this one dirty\\n for d in display.overlapped:\\n d.dirty = True\\n dirty |= drew\\n return dirty\",\n \"def draw_img(self, i, j, k):\\n if k < len(self.images):\\n img = self.images[k]\\n r = self.get_rect(i, j)\\n self.screen.blit(img, r)\",\n \"def draw(self):\\n self._view.clear()\\n for h in range(len(self._bricks)):\\n self._bricks[h].draw(self._view)\\n \\n self._paddle.draw(self._view)\\n \\n for w in self._FP_list:\\n w.draw(self._view)\\n \\n # draw ball if not None\\n if not self._ball is None:\\n self._ball.draw(self._view)\",\n \"def draw(self, screen):\",\n \"def draw(img, p):\\n\\t\\t\\n\\t########### First of all we gotta define which point connects together, there are points for the head(5), the shoulders, elbows, hands, top of thighs, knees, feet\\n\\t########### top of thighs, knees and feet. We also gonna add 2 extra points for the neck and the pelvis, to make everythonh look better.\\n\\t########### We also provide the color of each line here\\n\\tconnexions = [\\t\\n\\t\\t\\t\\t\\t(5, 7, 'navy'),\\t\\t# shoulder => elbow\\n\\t\\t\\t\\t\\t(7, 9, 'navy'),\\t\\t# elbow => hand\\n\\t\\t\\t\\t\\t(6, 8, 'navy'),\\t\\t# same on the other side\\n\\t\\t\\t\\t\\t(8, 10, 'navy'),\\n\\t\\t\\t\\t\\t(11, 13, 'lime'),\\t# thigh => knee\\n\\t\\t\\t\\t\\t(13, 15, 'lime'),\\t# knee => foot\\n\\t\\t\\t\\t\\t(12, 14, 'lime'),\\t# same on the other side\\n\\t\\t\\t\\t\\t(14, 16, 'lime'),\\n\\n\\t\\t\\t\\t\\t###### With The Extra points :\\n\\n\\t\\t\\t\\t\\t(0, 17, 'aqua'),\\t# head => neck\\n\\t\\t\\t\\t\\t(17, 5, 'aqua'),\\t# neck => shoulders\\n\\t\\t\\t\\t\\t(17, 6, 'aqua'),\\n\\t\\t\\t\\t\\t(17, 18, 'teal'),\\t# neck => pelvis\\n\\t\\t\\t\\t\\t(18, 11, 'teal'),\\t# pelvis => thighs\\n\\t\\t\\t\\t\\t(18, 12, 'teal')\\n\\t\\t\\t\\t\\t]\\n\\n\\t###### now let's find out how many objects were detected \\n\\t\\n\\tl = len(p[0][\\\"scores\\\"])\\n\\n\\t##### time to draw now, we'll only select objects with a score over .9\\n\\n\\td = idw.Draw(img)\\n\\n\\tfor k in range(l):\\n\\n\\t\\tif p[0][\\\"scores\\\"][k] > 0.98:\\n\\n\\t\\t\\t##### Let's add the neck and pelvis:\\n\\t\\t\\tneck = (p[0][\\\"keypoints\\\"][k][5] + p[0][\\\"keypoints\\\"][k][6])/2\\n\\t\\t\\tpelv = (p[0][\\\"keypoints\\\"][k][11] + p[0][\\\"keypoints\\\"][k][12])/2\\n\\n\\t\\t\\t#### it's getting tricky here\\n\\n\\t\\t\\tnepe = t.zeros((2, 3))\\n\\t\\t\\tnepe[0] = neck ; nepe[1] = pelv \\n\\n\\t\\t\\t### Now let's put everything into a single tensor\\n\\t\\t\\tbody = t.cat((p[0][\\\"keypoints\\\"][k], nepe))\\n\\n\\t\\t\\t#### We can start drawing now, for real\\n\\n\\t\\t\\tfor tp in connexions:\\n\\n\\t\\t\\t\\tp0 = (int(body[tp[0], 0]), int(body[tp[0], 1]))\\n\\t\\t\\t\\tp1 = (int(body[tp[1], 0]), int(body[tp[1], 1]))\\n\\t\\t\\t\\td.line([p0, p1], fill=tp[2], width=2)\\n\\n\\t\\t\\t#### Now the points\\n\\n\\t\\t\\tfor ts in t.cat((body[0:1], body[5:])):\\n\\t\\t\\t\\td.ellipse((int(ts[0]-2), int(ts[1]-2), int(ts[0]+2), int(ts[1]+2)), 'fuchsia')\\n\\n\\t### and finally\\n\\t#plt.imshow(np.asarray(img)) Not Like That\\n\\timg.show()\",\n \"def draw_image(self):\\n self.PDF.saveState()\\n self.PDF.scale(1, -1)\\n # self.PDF.drawImage(\\n # LOGO, 490, -78, width=80, preserveAspectRatio=True, mask=\\\"auto\\\"\\n # )\\n self.PDF.restoreState()\",\n \"def pygame_render(objects_to_draw: DrawingObjects, surface):\\n\\n for arc in objects_to_draw.arcs:\\n pygame.draw.arc(surface, arc.color, arc.enclosing_rect, arc.angle_begin, arc.angle_end, arc.lines_thickness)\\n for rect in objects_to_draw.rects:\\n pygame.draw.rect(surface, rect.color, rect.shape, rect.lines_thickness)\\n for circle in objects_to_draw.circles:\\n pygame.draw.circle(surface, circle.color, circle.center, circle.radius, circle.line_thickness)\\n for a_line in objects_to_draw.lines:\\n pygame.draw.line(surface, a_line.color, a_line.begin, a_line.end, a_line.line_thickness)\",\n \"def __smoothen_color(self, outer, inner):\\n outer_curve = zip(outer[0], outer[1])\\n inner_curve = zip(inner[0], inner[1])\\n x_points = []\\n y_points = []\\n for point in outer_curve:\\n x_points.append(point[0])\\n y_points.append(point[1])\\n for point in inner_curve:\\n x_points.append(point[0])\\n y_points.append(point[1])\\n img_base = np.zeros((self.height, self.width))\\n cv2.fillConvexPoly(img_base, np.array(np.c_[x_points, y_points], dtype='int32'), 1)\\n img_mask = cv2.GaussianBlur(img_base, (81, 81), 0) #51,51\\n img_blur_3d = np.ndarray([self.height, self.width, 3], dtype='float')\\n img_blur_3d[:, :, 0] = img_mask\\n img_blur_3d[:, :, 1] = img_mask\\n img_blur_3d[:, :, 2] = img_mask\\n self.im_copy = (img_blur_3d * self.image * 0.7 + (1 - img_blur_3d * 0.7) * self.im_copy).astype('uint8')\",\n \"def draw(self, surface):\\n\\n\\t\\tsurface.blit(self.image, self.rect.topleft)\",\n \"def run(self, src, dest):\\n self.logger.debug('Start the blurring. src=\\\"%s\\\", dest=\\\"%s\\\"', src, dest)\\n create_frames(src, self.work['frames'])\\n data = self.analyze()\\n\\n for frame_no, frame in enumerate(self.work.files('frames')):\\n basename = os.path.basename(frame)\\n areas = []\\n for values in data:\\n for sector in values['sectors']:\\n if frame_no in range(sector[0]-self.offset, sector[1]+self.offset):\\n areas.append(values['area'])\\n break\\n if areas:\\n self.blur.blur_image(frame, areas, os.path.join(self.work['cleaned'], basename))\\n else:\\n copyfile(frame, os.path.join(self.work['cleaned'], basename))\\n save_frames(self.work['cleaned'], dest)\",\n \"def visualize_output(\\n self,\\n img: np.ndarray,\\n output_data: List[SegmObject]) -> np.ndarray:\\n draw_layer_tag = f'draw_layer_{self.id}_{self.draw_layer_index^1}'\\n mix_factor = .3\\n\\n self.layer = dpg.add_draw_layer(\\n parent=f'main_window_{self.id}',\\n tag=draw_layer_tag,\\n show=False\\n )\\n\\n for out in output_data:\\n if out.score < self.score_threshold:\\n continue\\n\\n mask = cv2.resize(\\n out.mask,\\n (self.width, self.height),\\n cv2.INTER_NEAREST\\n ).astype(np.uint8)\\n color_mask = np.tile(mask[..., None]/255., [1, 1, 4])\\n color_image = np.tile(\\n self.class_colors[out.clsname],\\n [self.height, self.width, 1]\\n )\\n img += color_image * color_mask * mix_factor\\n contour, _ = cv2.findContours(\\n mask.astype(np.uint8),\\n cv2.RETR_CCOMP,\\n cv2.CHAIN_APPROX_NONE\\n )\\n color = np.array(self.class_colors[out.clsname])\\n color = tuple((255*color).astype(np.uint8))\\n for c in contour:\\n dpg.draw_polyline(\\n parent=draw_layer_tag,\\n points=list(c[:, 0, :]),\\n color=color,\\n thickness=2\\n )\\n description = f'{out.clsname} [{out.score * 100:.2f}%]'\\n dpg.draw_text(\\n parent=draw_layer_tag,\\n text=description,\\n pos=(c[:, 0, 0].min() + _PADDING,\\n c[:, 0, 1].min() + _PADDING),\\n color=color,\\n size=_FONT_SIZE\\n )\\n\\n return img\",\n \"def end_cast(self):\\r\\n #draw the actual map\\r\\n self.emap.draw(shader=self.mshader, camera=self.camera)\\r\\n super(ShadowCaster, self)._end()\\r\\n # set third texture to this ShadowCaster texture\\r\\n texs = self.emap.buf[0].textures\\r\\n if len(texs) == 2:\\r\\n texs.append(self)\\r\\n else:\\r\\n texs[2] = self\\r\\n # change background back to blue\\r\\n opengles.glClearColor(ctypes.c_float(0.4), ctypes.c_float(0.8), \\r\\n ctypes.c_float(0.8), ctypes.c_float(1.0))\\r\\n # work out left, top, right, bottom for shader\\r\\n self.emap.unif[48] = 0.5 * (1.0 + self.scaleu) # left [16][0]\\r\\n self.emap.unif[49] = 0.5 * (1.0 + self.scalev) # top [16][1]\\r\\n self.emap.unif[51] = 1.0 - self.emap.unif[48] # right [17][0]\\r\\n self.emap.unif[52] = 1.0 - self.emap.unif[49] # bottom [17][1]\\r\\n \\r\\n du = float(self.location[0] / self.emap.width)\\r\\n dv = float(self.location[2] / self.emap.depth)\\r\\n self.emap.unif[48] -= self.scaleu * (du if self.emap.unif[50] == 1.0 else dv)\\r\\n self.emap.unif[49] += self.scalev * (dv if self.emap.unif[50] == 1.0 else du)\\r\\n self.emap.unif[51] -= self.scaleu * (du if self.emap.unif[50] == 1.0 else dv)\\r\\n self.emap.unif[52] += self.scalev * (dv if self.emap.unif[50] == 1.0 else du)\",\n \"def draw(self, view):\\n for i in self.get_aliens():\\n for n in i:\\n if not n is None:\\n n.draw(view)\\n self.get_ship().draw(view)\\n self.get_ship().get_linevalue().draw(view)\\n for i in self.get_PU():\\n i.draw(view)\\n for i in self.get_bolts():\\n i.draw(view)\",\n \"def add_blurring(self):\\n self._render_passes.add_blur_pass()\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self) -> Any:\",\n \"def __init__(self, mapfile, xpos, zpos, emap, width=10.0, depth=10.0, height=10.0, name=\\\"building\\\", draw_details=None, yoff=0.0, scheme=None):\\r\\n self.xpos = xpos\\r\\n self.zpos = zpos\\r\\n self.width = width\\r\\n self.depth = depth\\r\\n self.height = height\\r\\n self.name = name\\r\\n self.ceilingthickness = 1.0\\r\\n self.walls = []\\r\\n\\r\\n if scheme == None:\\r\\n self.scheme = Building.baseScheme\\r\\n else:\\r\\n self.scheme = scheme\\r\\n\\r\\n # We don't have to be rigorous here, this should only be a draw_details or an iterable of draw_details.\\r\\n if hasattr(draw_details, \\\"__getitem__\\\") or hasattr(draw_details, \\\"__iter__\\\"):\\r\\n assert (len(draw_details) == self.scheme[\\\"#models\\\"])\\r\\n self.details = draw_details\\r\\n else:\\r\\n self.details = [draw_details for x in range(self.scheme[\\\"#models\\\"])]\\r\\n # having a method like this allows draw details to be set later\\r\\n\\r\\n self.yoff = yoff\\r\\n\\r\\n self.model = [MergeShape(name=name+\\\".\\\"+str(x)) for x in range(self.scheme[\\\"#models\\\"])]\\r\\n\\r\\n if mapfile[0] != '/':\\r\\n mapfile = sys.path[0] + '/' + mapfile\\r\\n print(\\\"Loading building map ...\\\", mapfile)\\r\\n\\r\\n im = Image.open(mapfile)\\r\\n im = ImageOps.invert(im)\\r\\n ix,iy = im.size\\r\\n\\r\\n print(\\\"image size\\\", ix, \\\",\\\", iy)\\r\\n\\r\\n startx = xpos - ix / 2 * width\\r\\n starty = zpos - ix / 2 * depth\\r\\n\\r\\n yoff += emap.calcHeight(-xpos,-zpos)\\r\\n\\r\\n if not im.mode == \\\"P\\\":\\r\\n im = im.convert('P', palette=Image.ADAPTIVE)\\r\\n im = im.transpose(Image.FLIP_TOP_BOTTOM)\\r\\n im = im.transpose(Image.FLIP_LEFT_RIGHT)\\r\\n pixels = im.load()\\r\\n\\r\\n for y in range(1,iy-1):\\r\\n print(\\\".\\\", end='')\\r\\n for x in range(1,ix-1):\\r\\n colour = pixels[x,y]\\r\\n\\r\\n if x == 1:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x-1,y], \\\"edge\\\"), wallfunc=self.west_wall, ceilingedgefunc=self.west_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n else:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x-1,y]), wallfunc=self.west_wall, ceilingedgefunc=self.west_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n\\r\\n if x == ix-2:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x+1,y], \\\"edge\\\"), wallfunc=self.east_wall, ceilingedgefunc=self.east_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n else:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x+1,y]), wallfunc=self.east_wall, ceilingedgefunc=self.east_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n\\r\\n if y == 1:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y-1], \\\"edge\\\"), wallfunc=self.south_wall, ceilingedgefunc=self.south_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n else:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y-1]), wallfunc=self.south_wall, ceilingedgefunc=self.south_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n\\r\\n if y == iy-2:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x, y+1], \\\"edge\\\"), wallfunc=self.north_wall, ceilingedgefunc=self.north_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n else:\\r\\n self._executeScheme(x, y, startx, starty, (colour, pixels[x,y+1]), wallfunc=self.north_wall, ceilingedgefunc=self.north_edge, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n\\r\\n self._executeScheme(x, y, startx, starty, (colour, None), wallfunc=None, ceilingedgefunc=None, ceilingfunc=self.ceiling, rooffunc=self.roof)\\r\\n\\r\\n self.set_draw_details(self.details) # after models created otherwise\\r\\n # details lost by merging\\r\",\n \"def pixelize(\\n layer: vp.LineCollection,\\n pen_width: float | None,\\n mode: str,\\n keep_lines: bool,\\n overdraw: float,\\n):\\n\\n if pen_width is None:\\n pen_width = layer.property(vp.METADATA_FIELD_PEN_WIDTH) or vp.convert_length(\\\"0.6mm\\\")\\n color = layer.property(vp.METADATA_FIELD_COLOR) or vp.Color(\\\"black\\\")\\n color = color.red, color.green, color.blue\\n\\n pixel_pitch = pen_width\\n if mode == \\\"big\\\":\\n pixel_pitch *= 5\\n\\n bounds = layer.bounds()\\n if bounds is None:\\n logging.warning(\\\"!!! pixelize: layer is empty\\\")\\n return layer\\n x_min, y_min, x_max, y_max = bounds\\n\\n img_size = (\\n math.ceil((x_max - x_min) / pixel_pitch),\\n math.ceil((y_max - y_min) / pixel_pitch),\\n )\\n img = Image.new(\\\"RGBA\\\", img_size)\\n draw = ImageDraw.Draw(img)\\n for line in layer:\\n draw.line(\\n [(pt.real, pt.imag) for pt in (line - x_min - 1j * y_min) / pixel_pitch],\\n fill=color,\\n )\\n\\n doc = vp.Document()\\n if mode == \\\"big\\\":\\n # noinspection PyTypeChecker\\n big_mode(doc, np.array(img), np.array([color]), pen_width)\\n elif mode == \\\"line\\\":\\n # noinspection PyTypeChecker\\n line_mode(doc, np.array(img), np.array([color]), pen_width, overdraw)\\n elif mode == \\\"snake\\\":\\n # noinspection PyTypeChecker\\n snake_mode(doc, np.array(img), np.array([color]), pen_width)\\n\\n doc.translate(x_min, y_min)\\n\\n if keep_lines:\\n layer.extend(doc.layers[1])\\n layer.set_property(vp.METADATA_FIELD_PEN_WIDTH, pen_width)\\n return layer\\n else:\\n result = layer.clone(doc.layers[1])\\n result.set_property(vp.METADATA_FIELD_PEN_WIDTH, pen_width)\\n return result\",\n \"def render(self, output, image_size,\\n face_colors=(\\\"#477984\\\", \\\"#EEAA4D\\\", \\\"#74C3F2\\\")):\\n pass\",\n \"def blending_example2():\\n pic_earth = read_image(relpath(\\\"./externals/pic_earth.jpg\\\"), 2)\\n pic_asteroid = read_image(relpath(\\\"./externals/pic_asteroid.jpg\\\"), 2)\\n mask = read_image(relpath(\\\"./externals/mask_asteroid.jpg\\\"), 1)\\n # making the mask binary (normalizing 2 original values)\\n mask = strech_helper(mask).astype(np.bool)\\n [R1, G1, B1] = np.dsplit(pic_earth, pic_earth.shape[2])\\n [R2, G2, B2] = np.dsplit(pic_asteroid, pic_asteroid.shape[2])\\n R1 = np.reshape(R1, (1024,1024))\\n R2 = np.reshape(R2, (1024,1024))\\n G1 = np.reshape(G1, (1024,1024))\\n G2 = np.reshape(G2, (1024,1024))\\n B1 = np.reshape(B1, (1024,1024))\\n B2 = np.reshape(B2, (1024,1024))\\n\\n blend1 = pyramid_blending(R2, R1, mask, 3, 3, 3)\\n blend2 = pyramid_blending(G2, G1, mask, 3, 3, 3)\\n blend3 = pyramid_blending(B2, B1, mask, 3, 3, 3)\\n\\n blend1 = np.reshape(blend1, (blend1.shape[0], blend1.shape[1], 1))\\n blend2 = np.reshape(blend2, (blend2.shape[0], blend3.shape[1], 1))\\n blend3 = np.reshape(blend3, (blend3.shape[0], blend3.shape[1], 1))\\n\\n new_pic = np.concatenate((blend1, blend2, blend3), axis=2)\\n # plotting the images\\n fig = plt.figure()\\n ax1 = fig.add_subplot(221)\\n ax2 = fig.add_subplot(222)\\n ax3 = fig.add_subplot(223)\\n ax4 = fig.add_subplot(224)\\n ax1.imshow(pic_earth)\\n ax2.imshow(pic_asteroid)\\n ax3.imshow(mask, cmap='gray')\\n ax4.imshow(new_pic)\\n plt.show()\\n\\n return pic_earth, pic_asteroid, mask, new_pic\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.62051916","0.608807","0.60164183","0.5994042","0.5974199","0.5956731","0.59398204","0.5916923","0.5856512","0.582398","0.5776995","0.57500255","0.57297295","0.57178104","0.570983","0.5693524","0.5684857","0.5681157","0.5680937","0.5669405","0.5637934","0.562625","0.5624656","0.56058574","0.5598345","0.55960697","0.55954355","0.55895686","0.55881375","0.556966","0.556966","0.556966","0.556966","0.55671185","0.5562524","0.5560224","0.5558146","0.55548227","0.55528337","0.5539285","0.5538307","0.5532236","0.5532236","0.55309254","0.5524841","0.5498432","0.54965967","0.5487996","0.54877627","0.54518455","0.54467946","0.54463756","0.5444206","0.543837","0.54378825","0.5432342","0.5429866","0.5429728","0.5427452","0.54207045","0.54147476","0.54083234","0.54083234","0.54083234","0.5402452","0.5399269","0.5399034","0.53888863","0.5388239","0.5378729","0.5376633","0.53755856","0.5366437","0.53655607","0.5359801","0.5353391","0.53513175","0.53478384","0.5342143","0.5338164","0.5335753","0.5316574","0.53160495","0.5315662","0.53133667","0.53112406","0.5297302","0.52955914","0.52955914","0.52955914","0.52955914","0.52955914","0.52955914","0.52955914","0.52955914","0.52878064","0.52842885","0.5273809","0.5263196","0.52589136"],"string":"[\n \"0.62051916\",\n \"0.608807\",\n \"0.60164183\",\n \"0.5994042\",\n \"0.5974199\",\n \"0.5956731\",\n \"0.59398204\",\n \"0.5916923\",\n \"0.5856512\",\n \"0.582398\",\n \"0.5776995\",\n \"0.57500255\",\n \"0.57297295\",\n \"0.57178104\",\n \"0.570983\",\n \"0.5693524\",\n \"0.5684857\",\n \"0.5681157\",\n \"0.5680937\",\n \"0.5669405\",\n \"0.5637934\",\n \"0.562625\",\n \"0.5624656\",\n \"0.56058574\",\n \"0.5598345\",\n \"0.55960697\",\n \"0.55954355\",\n \"0.55895686\",\n \"0.55881375\",\n \"0.556966\",\n \"0.556966\",\n \"0.556966\",\n \"0.556966\",\n \"0.55671185\",\n \"0.5562524\",\n \"0.5560224\",\n \"0.5558146\",\n \"0.55548227\",\n \"0.55528337\",\n \"0.5539285\",\n \"0.5538307\",\n \"0.5532236\",\n \"0.5532236\",\n \"0.55309254\",\n \"0.5524841\",\n \"0.5498432\",\n \"0.54965967\",\n \"0.5487996\",\n \"0.54877627\",\n \"0.54518455\",\n \"0.54467946\",\n \"0.54463756\",\n \"0.5444206\",\n \"0.543837\",\n \"0.54378825\",\n \"0.5432342\",\n \"0.5429866\",\n \"0.5429728\",\n \"0.5427452\",\n \"0.54207045\",\n \"0.54147476\",\n \"0.54083234\",\n \"0.54083234\",\n \"0.54083234\",\n \"0.5402452\",\n \"0.5399269\",\n \"0.5399034\",\n \"0.53888863\",\n \"0.5388239\",\n \"0.5378729\",\n \"0.5376633\",\n \"0.53755856\",\n \"0.5366437\",\n \"0.53655607\",\n \"0.5359801\",\n \"0.5353391\",\n \"0.53513175\",\n \"0.53478384\",\n \"0.5342143\",\n \"0.5338164\",\n \"0.5335753\",\n \"0.5316574\",\n \"0.53160495\",\n \"0.5315662\",\n \"0.53133667\",\n \"0.53112406\",\n \"0.5297302\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52955914\",\n \"0.52878064\",\n \"0.52842885\",\n \"0.5273809\",\n \"0.5263196\",\n \"0.52589136\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.69175744"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90870,"cells":{"query":{"kind":"string","value":"In this version, we include all partial canvas renders in addition to the final renders"},"document":{"kind":"string","value":"def forward_partial(self, drawing, blur_sigma=None, epsilon=None, concat=False):\n if isinstance(drawing[0], list):\n pimgs = [self.painter.forward_partial(d) for d in drawing]\n lengths = [len(p) for p in pimgs]\n pimgs = torch.cat(pimgs)\n pimgs = self.broaden_and_blur(pimgs, blur_sigma, epsilon)\n if concat:\n return pimgs\n pimgs = torch.split(pimgs, lengths, 0)\n return list(pimgs)\n else:\n pimgs = self.painter.forward_partial(drawing)\n pimgs = self.broaden_and_blur(pimgs, blur_sigma, epsilon)\n return pimgs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def on_draw(self):\n\t\tself.render()","def on_draw(self):\n\t\tself.render()","def dspyRender(self):\n pass","def on_draw(self):\n # draw everything","def render(self):\n self.axial.Render()\n self.coronal.Render()\n self.sagittal.Render()\n #self.isosurface.Render()\n #self.rwi_pcp.Render()","def pre_draw(self):","def _render_callback(self, _sim, _viewer):\n pass","def on_draw( self ):\r\n self.clear()\r\n self.setup_3D()\r\n print \"DEBUG:\" , \"There are\" , len( self.renderlist ) , \"items in 'self.renderlist'\"\r\n for obj in self.renderlist:\r\n obj.draw()","def _automatic_rendering(self):\n if self.viewer is not None and self.enable_auto_render:\n self.should_update_rendering = True\n\n if self.automatic_rendering_callback:\n self.automatic_rendering_callback()\n else:\n self.render(self.rendering_mode)","def _automatic_rendering(self) -> None:\n if self.viewer is not None and self.enable_auto_render:\n\n if self._record_video_wrapper and self._record_video_wrapper.video_recorder:\n self._record_video_wrapper.video_recorder.capture_frame()\n else:\n self.render()","def _prepare_draw(self, view=None):\n pass","def add_canvas():\n return render_template('courses/add_canvas.html')","def draw(self):\n\t\tpass","def draw(self):\n self.bufferX = (self.appWidth/2) - self.viewX\n self.bufferY = (self.appHeight/2) - self.viewY\n anwp.sl.engine.clear()\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\n self.drawWarpLines()\n \n # render engine\n anwp.sl.engine.render()\n self.drawSystemInfo()\n self.drawWarpGateInfo()\n self.drawWarpTradeInfo()","def render(self, visual_elements, canvas):\n self.coordinate_system = visual_elements.coordinate_system\n self.canvas = canvas\n for ve in visual_elements.elements: self.add(ve)\n return","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def _render(self) -> None:\n pass","def on_draw_overlay(self):","def _drawOnCanvas(self):\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\n for key in self.elements:\n graphElement=self.elements[key]\n graphElement.draw(self.canvas)\n self.sync=True","def redraw(self, **kwargs):\n #src_dict = self.data_sources\n #self.remove_sources(src_dict.keys())\n self.renderers = {}\n #self.renderers = {}\n self.figure = self.draw_figure(**kwargs)\n #self.add_sources(src_dict)\n # todo does the old figure linger on?\n self.render_sources(self.data_sources)\n self.bk_pane.object = self.figure","def render(self):\n canvas_id = 'zdog_{}'.format(self.CANVAS_INDEX)\n illo_id = 'illo_{}'.format(self.CANVAS_INDEX)\n Scene.CANVAS_INDEX += 1\n\n html_lines = []\n\n js_lines = []\n\n euler = -rowan.to_euler(\n self.rotation, convention='xyz', axis_type='intrinsic')\n translation = self.translation*(1, -1, 1)\n\n pan_cfg = self.get_feature_config('pan')\n pan = pan_cfg.get('value', True) if pan_cfg is not None else False\n\n js_lines.append(\"\"\"\n let {illo_id} = new Zdog.Illustration({{\n element: '#{canvas_id}',\n zoom: {zoom},\n dragRotate: {rotation_enabled},\n rotate: {{x: {angle[0]}, y: {angle[1]}, z: {angle[2]}}},\n translate: {{x: {pos[0]}, y: {pos[1]}, z: {pos[2]}}},\n }});\n \"\"\".format(\n illo_id=illo_id, canvas_id=canvas_id, zoom=self.zoom*self.pixel_scale,\n angle=euler, pos=translation,\n rotation_enabled=('false' if pan else 'true')))\n\n config = self.get_feature_config('ambient_light')\n ambient_light = 0 if config is None else config.get('value', .4)\n\n config = self.get_feature_config('directional_light')\n directional_light = ([(0, 0, 0)] if config is None else\n config.get('value', [(0, 0, 0)]))\n directional_light = np.atleast_2d(directional_light)\n\n shapeIndex = 0\n for i, prim in enumerate(self._primitives):\n js_lines.extend(prim.render(\n rotation=self.rotation, illo_id=illo_id,\n name_suffix=i, ambient_light=ambient_light,\n directional_light=directional_light))\n\n (width, height) = map(int, self.size_pixels)\n html_lines.append(\"\"\"\n <canvas id=\"{canvas_id}\" width=\"{width}\" height=\"{height}\"></canvas>\n \"\"\".format(canvas_id=canvas_id, width=width, height=height))\n\n html_lines.append(\"\"\"<script>\n var fill_{canvas_id} = function() {{\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append(LOCAL_HELPER_SCRIPT)\n html_lines.extend(js_lines)\n\n pan_snippet = \"\"\"\n new Zdog.Dragger({{\n startElement: {illo_id}.element,\n onDragStart: function( pointer, moveX, moveY) {{\n this.lastX = 0;\n this.lastY = 0;\n }},\n onDragMove: function( pointer, moveX, moveY ) {{\n let deltax = moveX - this.lastX;\n let deltay = moveY - this.lastY;\n let scale = 1.0/{illo_id}.zoom;\n {illo_id}.translate.x += deltax*scale;\n {illo_id}.translate.y += deltay*scale;\n this.lastX = moveX;\n this.lastY = moveY;\n }}\n }});\"\"\".format(illo_id=illo_id)\n if pan:\n html_lines.append(pan_snippet)\n\n html_lines.append(\"\"\"\n let this_canvas = document.querySelector(\"#{canvas_id}\");\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append(\"\"\"\n let animate_{canvas_id} = function() {{\n if(is_in_view(this_canvas))\n {{\n {illo_id}.updateRenderGraph();\n }}\n if(document.contains(this_canvas))\n {{\n requestAnimationFrame(animate_{canvas_id});\n }}\n }};\n animate_{canvas_id}();\"\"\".format(canvas_id=canvas_id, illo_id=illo_id))\n # remove the global reference to this function after using it\n html_lines.append('fill_{canvas_id} = null;'.format(canvas_id=canvas_id))\n html_lines.append('};') # end of fill_{canvas_id}\n # now call fill_{canvas_id}, possibly after loading zdog\n html_lines.append(\"\"\"\n if (typeof Zdog == 'undefined')\n {{\n var script = document.createElement('script');\n script.addEventListener('load', fill_{canvas_id}, false);\n script.src = 'https://unpkg.com/zdog@1/dist/zdog.dist.min.js';\n document.getElementsByTagName('head')[0].appendChild(script);\n }}\n else\n fill_{canvas_id}();\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append('</script>')\n\n return '\\n'.join(html_lines)","def update(self, events: list[event.Event]) -> None:\n if self.buffer.get_width() != self.render_width or self.buffer.get_height() != self.render_height:\n self.pre_render()","def render(self):\n\n self.desert_image.render()\n self.cannon_image.render()\n self.play_button.render()\n self.escape_label.render()","def on_draw(self, widget, cr):\n #print \"starting to draw\"\n if self.double_buffer is not None:\n self.draw_tiles()\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\n cr.paint()\n else:\n print('Invalid double buffer')\n #print \"done drawing\"\n return False","def render(self):\n self.rendering = True\n self.env.render()","def _prepare_draw(self, view=None):\n return True","def pre_render(self) -> None:\n self.buffer = Surface((self.render_width, self.render_height), SRCALPHA)\n self.buffer.fill(list(self.halo_texture.surfaces.values())[0].get_at((0, 0)))\n\n self.buffer.fill((0, 0, 0, 0), Rect(\n (self.render_width - self.halo_texture.get_width()) // 2,\n (self.render_height - self.halo_texture.get_height()) // 2,\n self.halo_texture.get_width(),\n self.halo_texture.get_height()\n ))","def sgnRenderAllFrames(self):\n\n self.uiSaveSelectedRenders()\n\n self.core.render(self.lstSelectedRenders, bEverything = True)","def refresh_canvas(self):\n self.canvas.delete('all')\n self.draw_handler(self)\n self.canvas.after(CANVAS[\"REFRESH_TIME\"], self.refresh_canvas)","def refresh_view(self):\n if self._step_number % 2 == 0:\n self._view.draw_enemies(self._game.enemies)\n self._view.draw_towers(self._game.towers)\n self._view.draw_obstacles(self._game.obstacles)","def setup_draw(self):\n pass","def paint(self):\r\n pass","def paintGL(self):\n self._sceneviewer.renderScene()\n # paintGL end","def draw(self):\n for section in self.sections:\n canvas_reset(self.canvas)\n section.draw(self.canvas)","def draw(self):","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):","def _render(self, gc, points):\n with gc:\n gc.set_antialias(True)\n self._draw_default_axes(gc)\n self._draw_default_grid(gc)\n if len(points)>0:\n gc.clip_to_rect(self.x, self.y, self.width, self.height)\n gc.set_stroke_color(self.color_)\n gc.set_line_width(self.line_width)\n gc.set_line_dash(self.line_style_)\n\n gc.begin_path()\n gc.lines(points)\n gc.stroke_path()\n\n return","def prepareTestCanvas(self):\r\n loadLayers()\r\n setCanvasCrs(4326, True)\r\n CANVAS.resize(QtCore.QSize(400, 400))\r\n CANVAS.zoomToFullExtent()","def _draw(self, canvas, options):\n pass # must override in subclass","def blit(self):\n # self.ax1.draw_artist(self.lines[:2])\n # self.ax2.draw_artist(self.lines[2:4])\n self.ax1.autoscale()\n self.ax2.autoscale()\n self.ax1.redraw_in_frame()\n self.ax2.redraw_in_frame()\n self.fig.canvas.blit(self.fig.bbox)","def getImmediateRendering():\n\n\treturn False","def on_paint(self, event):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the viewport, modelview and projection matrices\n self.init_gl()\n self.init = True\n\n size = self.GetClientSize()\n text = \"\".join([_(u\"Canvas redrawn on paint event, size is \"),\n str(size.width), \", \", str(size.height)])\n self.render(text)","def render(self, screen):\n pass","def render(self, screen):\n pass","def prepareForDraw(self, obj_path, camera_path, frame_context, old_data):\n pass","def _refresh_render(self):\n current_frame = self.frame\n self.frame = int(1E6)\n self.frame = current_frame","def draw(self, *args, **kwargs):\n self.window.clear()\n self.batch.draw()","def test_visualisations_perform_visualisation_render_data_refresh_synchronously(self):\n pass","def renderall(self):\n\n if not self.isinitialized:\n return\n # clear display\n self.screen.fill(BGCOLOR)\n # draw the board\n self.drawBoard()\n # flip the display to show whatever we drew\n pygame.display.flip()","def pysweep_before_finish_init(self):\n self.displaycanvas = DisplayCanvas(self.pysweep.master, self.boardsize, self.lcounterlength, self.rcounterlength, self.images)\n self.displaycanvas.pack()\n\n self.pysweep.master.update_idletasks()\n self.displaycanvas.update_idletasks()\n # enode = self.arbitrary()\n # print('DisplayCanvas:', enode)","def render(self):\r\n super().render()","def render(self) -> None:\n if self.native_rendering:\n self._render()\n else:\n self.renderer.render_image(self.get_rendered_image())","def draw(self):\n\n self.updateLazyImageLoading()\n\n image(self.baseMap, 0, 0)\n\n for layer in self.layers:\n layer.draw()\n\n for marker in self.markers:\n marker.draw()","def sgnRenderCurrentFrame(self):\n\n self.uiSaveSelectedRenders()\n\n self.core.render(self.lstSelectedRenders, bCurrentFrame = True)","def draw(self):\n self.batch.draw()","def on_draw_event(self, widget, ctx):\n # the _need_redraw flag doesnt work. it sometimes prevents\n # the rendering and leaving the canvas blank\n #if self._need_redraw:\n self._renderer.set_context(ctx)\n allocation = self.get_allocation()\n x, y, w, h = allocation.x, allocation.y, allocation.width, allocation.height\n self._render_figure(w, h)\n #self._need_redraw = False\n\n return False # finish event propagation?","def _drawFrame(self):\n\n self._clearScreen()\n \n for object in Object.Objects:\n self._drawObject(object)\n\n for entity in Entity.Entities:\n self._drawObject(entity)\n\n self._drawObject(Game.Player)","def make(self):\n\t\tif RENDER_VIEWS > 1:\n\t\t\tself._make()","def draw(self):\n self.write_image()\n self.update()","def StartDraw(self):\r\n self.zoom = self.test.viewZoom\r\n self.center = self.test.viewCenter\r\n self.offset = self.test.viewOffset\r\n self.screenSize = self.test.screenSize","def canvas_api():\n pass","def draw(self) -> Any:","def renderCanvas(self, plot, painter, canvasRect, maps):\r\n canvas = plot.canvas()\r\n r = canvasRect.adjusted(0., 0., -1., 1.)\r\n if self.__data.layoutFlags & self.FrameWithScales:\r\n painter.save()\r\n r.adjust(-1., -1., 1., 1.)\r\n painter.setPen(QPen(Qt.black))\r\n if not (self.__data.discardFlags & self.DiscardCanvasBackground):\r\n bgBrush = canvas.palette().brush(plot.backgroundRole())\r\n painter.setBrush(bgBrush)\r\n painter.drawRect(r)\r\n painter.restore()\r\n painter.save()\r\n painter.setClipRect(canvasRect)\r\n plot.drawItems(painter, canvasRect, maps)\r\n painter.restore()\r\n elif canvas.testAttribute(Qt.WA_StyledBackground):\r\n clipPath = QPainterPath()\r\n painter.save()\r\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\r\n QwtPainter.drawBackground(painter, r, canvas)\r\n clipPath = qwtCanvasClip(canvas, canvasRect)\r\n painter.restore()\r\n painter.save()\r\n if clipPath.isEmpty():\r\n painter.setClipRect(canvasRect)\r\n else:\r\n painter.setClipPath(clipPath)\r\n plot.drawItems(painter, canvasRect, maps)\r\n painter.restore()\r\n else:\r\n clipPath = QPainterPath()\r\n frameWidth = 0\r\n if not self.__data.discardFlags & self.DiscardCanvasFrame:\r\n frameWidth = canvas.frameWidth()\r\n clipPath = qwtCanvasClip(canvas, canvasRect)\r\n innerRect = canvasRect.adjusted(frameWidth, frameWidth,\r\n -frameWidth, -frameWidth)\r\n painter.save()\r\n if clipPath.isEmpty():\r\n painter.setClipRect(innerRect)\r\n else:\r\n painter.setClipPath(clipPath)\r\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\r\n QwtPainter.drawBackground(painter, innerRect, canvas)\r\n plot.drawItems(painter, innerRect, maps)\r\n painter.restore()\r\n if frameWidth > 0:\r\n painter.save()\r\n frameStyle = canvas.frameShadow() | canvas.frameShape()\r\n frameWidth = canvas.frameWidth()\r\n borderRadius = canvas.borderRadius()\r\n if borderRadius > 0.:\r\n QwtPainter.drawRoundedFrame(painter, canvasRect, r, r,\r\n canvas.palette(), frameWidth,\r\n frameStyle)\r\n else:\r\n midLineWidth = canvas.midLineWidth()\r\n QwtPainter.drawFrame(painter, canvasRect, canvas.palette(),\r\n canvas.foregroundRole(), frameWidth,\r\n midLineWidth, frameStyle)\r\n painter.restore()","def reindex_graphics(self):\n for obj in self.context.static_objects:\n self.canvas.children.remove(obj.widget.canvas)\n # fill _objects_z_index\n _objects_z_index = {}\n for obj in self.context.static_objects:\n y = obj.widget.pos[1]\n if not y in _objects_z_index:\n _objects_z_index[y] = []\n _objects_z_index[y].append(obj)\n _keys = _objects_z_index.keys()\n _keys.sort()\n _keys.reverse()\n for k in _keys:\n objs = _objects_z_index[k]\n for obj in objs:\n self.canvas.add(obj.widget.canvas)","def draw(self, canvas):\n super().draw(canvas, self.__path)","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def on_paint(self, event):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the OpenGL rendering context\n self.init_gl()\n self.init = True\n\n size = self.GetClientSize()\n text = \"\".join([_(u\"Canvas redrawn on paint event, size is \"),\n str(size.width), \", \", str(size.height)])\n self.render()","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def render(self):\n\n # self.screen.blit(self.background, (0, 0))\n\n \"\"\"work with ever game state\"\"\"\n if self.state == FIRST_ENTER:\n \"\"\"welcome screen))\"\"\"\n text = self.font.render(\"PETERSBURG VAMPIRES\", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 3))\n text = self.font.render(\"press ENTER to start\", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_WIDTH / 4, SCREEN_HEIGHT / 2))\n elif self.state == GAME:\n \"\"\"rendering enemies and player\"\"\"\n old_time = time.time()\n self.levels[self.curient_level].render(self.screen)\n # self.levels[self.curient_level].render(self.screen)\n for obj in self.levels[self.curient_level].objects:\n \"\"\"debug mode\"\"\"\n # text = self.debug_font.render(\"koor:\" + str(obj.position_np), True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]))\n\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]))\n #\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+obj.image_size[Y]))\n #\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]+obj.image_size[Y]))\n #\n # x, y = get_grid_xy(obj.position_np, ZOMBIE_SIZE)\n # text = self.debug_font.render(\"x_grid: \" + str(x) + \",y_grid: \" + str(y), True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+20))\n\n obj.render(self.screen)\n \"\"\"rendering HP status\"\"\"\n text = self.font.render(\"HP: \" + str(self.player.HP), True, (255, 0, 0))\n self.screen.blit(text, (0, 10))\n new_time = time.time()\n self.render_time.append(new_time - old_time)\n elif self.state == END:\n \"\"\"The end screen\"\"\"\n self.screen.blit(pygame.image.load(END_BUTTON_PATH), (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 8))\n elif self.state == WIN:\n text = self.font.render(\"YOU WIN, REPEAT? \", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_HEIGHT / 2, SCREEN_WIDTH / 2))\n\n pygame.display.flip()","def forward_partial(self, drawing):\n assert not self.is_cuda\n drawing = drawings_to_cpu(drawing)\n ns = len(drawing)\n pimgs = torch.zeros(ns+1, *self.imsize)\n canvas = torch.zeros(*self.imsize)\n for i, stk in enumerate(drawing):\n canvas, _ = self.add_stroke(canvas, stk)\n pimgs[i+1] = canvas\n\n return pimgs","def render(self):\n gl.glEnableClientState(gl.GL_VERTEX_ARRAY);\n gl.glEnableClientState(gl.GL_NORMAL_ARRAY);\n gl.glVertexPointer(3, gl.GL_DOUBLE, 0, self._vertices);\n gl.glNormalPointer(gl.GL_DOUBLE, 0, self._normals);\n for patch in self._patches:\n patch.render()\n gl.glDisableClientState(gl.GL_VERTEX_ARRAY);\n gl.glDisableClientState(gl.GL_NORMAL_ARRAY);","def _before_render_event(self, *args, **kwargs):\n for chart in self._charts:\n chart._render_event(*args, **kwargs)","def draw(self):\n for obj in self.objects:\n obj.draw()","def render(self):\n self.screen.fill(prepare.BACKGROUND_COLOR)\n self.health_bar()\n # self.enemy_health()\n self.energy_bar()\n self.level.draw(self.screen)\n pg.display.update()","def get(self):\n self.render_front()","def render(self):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the OpenGL rendering context\n self.init_gl()\n self.init = True\n\n # Clear everything\n GL.glClear(GL.GL_COLOR_BUFFER_BIT | GL.GL_DEPTH_BUFFER_BIT)\n\n for device_id, output_id in self.monitors.monitors_dictionary:\n i = list(\n self.monitors.monitors_dictionary.keys()).index(\n (device_id, output_id))\n GL.glColor3f(self.colours[i%8][0], self.colours[i%8][1], self.colours[i%8][2])\n x = (i - len(self.monitors.monitors_dictionary)/2)*20\n signal_list = self.monitors.monitors_dictionary[(\n device_id, output_id)]\n\n signal_name = self.devices.get_signal_name(device_id, output_id)\n\n self.render_text(signal_name, x, 0, 10*len(signal_list)+10)\n\n length = len(signal_list)\n for j in range(length):\n signal = signal_list[j]\n z = (j-length/2)*20\n if signal == self.devices.HIGH:\n y = 11\n if signal == self.devices.LOW:\n y = 1\n if signal != self.devices.BLANK:\n self.draw_cuboid(x, z, 5, 10, y)\n\n # We have been drawing to the back buffer, flush the graphics pipeline\n # and swap the back buffer to the front\n GL.glFlush()\n self.SwapBuffers()","def display(self):\n\t\tfor c in self.canvas.values():\n\t\t\tc.update()\n\t\t\tself.superficie.blit(c.superficie, c.origen)\n\t\tpygame.display.flip()","def render(self, mode='human'):\n if self.microgridRender == None : self.microgridRender = MicrogridRender(self.microgrid, self.clock);\t\n self.microgridRender.render(mode);","def finish_render():\n get_window().static_display = True\n get_window().flip_count = 0\n get_window().flip()","def _pre_draw_bge(self):\r\n self._pre_draw_common()\r\n # draw rays\r\n self._drawRays()","def on_draw(self):\n self.clear()\n self.manager.draw()","def render(self, chart):\n chart.create_visualization_files(self.__outputpath)","def draw(self, screen):"],"string":"[\n \"def on_draw(self):\\n\\t\\tself.render()\",\n \"def on_draw(self):\\n\\t\\tself.render()\",\n \"def dspyRender(self):\\n pass\",\n \"def on_draw(self):\\n # draw everything\",\n \"def render(self):\\n self.axial.Render()\\n self.coronal.Render()\\n self.sagittal.Render()\\n #self.isosurface.Render()\\n #self.rwi_pcp.Render()\",\n \"def pre_draw(self):\",\n \"def _render_callback(self, _sim, _viewer):\\n pass\",\n \"def on_draw( self ):\\r\\n self.clear()\\r\\n self.setup_3D()\\r\\n print \\\"DEBUG:\\\" , \\\"There are\\\" , len( self.renderlist ) , \\\"items in 'self.renderlist'\\\"\\r\\n for obj in self.renderlist:\\r\\n obj.draw()\",\n \"def _automatic_rendering(self):\\n if self.viewer is not None and self.enable_auto_render:\\n self.should_update_rendering = True\\n\\n if self.automatic_rendering_callback:\\n self.automatic_rendering_callback()\\n else:\\n self.render(self.rendering_mode)\",\n \"def _automatic_rendering(self) -> None:\\n if self.viewer is not None and self.enable_auto_render:\\n\\n if self._record_video_wrapper and self._record_video_wrapper.video_recorder:\\n self._record_video_wrapper.video_recorder.capture_frame()\\n else:\\n self.render()\",\n \"def _prepare_draw(self, view=None):\\n pass\",\n \"def add_canvas():\\n return render_template('courses/add_canvas.html')\",\n \"def draw(self):\\n\\t\\tpass\",\n \"def draw(self):\\n self.bufferX = (self.appWidth/2) - self.viewX\\n self.bufferY = (self.appHeight/2) - self.viewY\\n anwp.sl.engine.clear()\\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\\n self.drawWarpLines()\\n \\n # render engine\\n anwp.sl.engine.render()\\n self.drawSystemInfo()\\n self.drawWarpGateInfo()\\n self.drawWarpTradeInfo()\",\n \"def render(self, visual_elements, canvas):\\n self.coordinate_system = visual_elements.coordinate_system\\n self.canvas = canvas\\n for ve in visual_elements.elements: self.add(ve)\\n return\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def _render(self) -> None:\\n pass\",\n \"def on_draw_overlay(self):\",\n \"def _drawOnCanvas(self):\\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\\n for key in self.elements:\\n graphElement=self.elements[key]\\n graphElement.draw(self.canvas)\\n self.sync=True\",\n \"def redraw(self, **kwargs):\\n #src_dict = self.data_sources\\n #self.remove_sources(src_dict.keys())\\n self.renderers = {}\\n #self.renderers = {}\\n self.figure = self.draw_figure(**kwargs)\\n #self.add_sources(src_dict)\\n # todo does the old figure linger on?\\n self.render_sources(self.data_sources)\\n self.bk_pane.object = self.figure\",\n \"def render(self):\\n canvas_id = 'zdog_{}'.format(self.CANVAS_INDEX)\\n illo_id = 'illo_{}'.format(self.CANVAS_INDEX)\\n Scene.CANVAS_INDEX += 1\\n\\n html_lines = []\\n\\n js_lines = []\\n\\n euler = -rowan.to_euler(\\n self.rotation, convention='xyz', axis_type='intrinsic')\\n translation = self.translation*(1, -1, 1)\\n\\n pan_cfg = self.get_feature_config('pan')\\n pan = pan_cfg.get('value', True) if pan_cfg is not None else False\\n\\n js_lines.append(\\\"\\\"\\\"\\n let {illo_id} = new Zdog.Illustration({{\\n element: '#{canvas_id}',\\n zoom: {zoom},\\n dragRotate: {rotation_enabled},\\n rotate: {{x: {angle[0]}, y: {angle[1]}, z: {angle[2]}}},\\n translate: {{x: {pos[0]}, y: {pos[1]}, z: {pos[2]}}},\\n }});\\n \\\"\\\"\\\".format(\\n illo_id=illo_id, canvas_id=canvas_id, zoom=self.zoom*self.pixel_scale,\\n angle=euler, pos=translation,\\n rotation_enabled=('false' if pan else 'true')))\\n\\n config = self.get_feature_config('ambient_light')\\n ambient_light = 0 if config is None else config.get('value', .4)\\n\\n config = self.get_feature_config('directional_light')\\n directional_light = ([(0, 0, 0)] if config is None else\\n config.get('value', [(0, 0, 0)]))\\n directional_light = np.atleast_2d(directional_light)\\n\\n shapeIndex = 0\\n for i, prim in enumerate(self._primitives):\\n js_lines.extend(prim.render(\\n rotation=self.rotation, illo_id=illo_id,\\n name_suffix=i, ambient_light=ambient_light,\\n directional_light=directional_light))\\n\\n (width, height) = map(int, self.size_pixels)\\n html_lines.append(\\\"\\\"\\\"\\n <canvas id=\\\"{canvas_id}\\\" width=\\\"{width}\\\" height=\\\"{height}\\\"></canvas>\\n \\\"\\\"\\\".format(canvas_id=canvas_id, width=width, height=height))\\n\\n html_lines.append(\\\"\\\"\\\"<script>\\n var fill_{canvas_id} = function() {{\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append(LOCAL_HELPER_SCRIPT)\\n html_lines.extend(js_lines)\\n\\n pan_snippet = \\\"\\\"\\\"\\n new Zdog.Dragger({{\\n startElement: {illo_id}.element,\\n onDragStart: function( pointer, moveX, moveY) {{\\n this.lastX = 0;\\n this.lastY = 0;\\n }},\\n onDragMove: function( pointer, moveX, moveY ) {{\\n let deltax = moveX - this.lastX;\\n let deltay = moveY - this.lastY;\\n let scale = 1.0/{illo_id}.zoom;\\n {illo_id}.translate.x += deltax*scale;\\n {illo_id}.translate.y += deltay*scale;\\n this.lastX = moveX;\\n this.lastY = moveY;\\n }}\\n }});\\\"\\\"\\\".format(illo_id=illo_id)\\n if pan:\\n html_lines.append(pan_snippet)\\n\\n html_lines.append(\\\"\\\"\\\"\\n let this_canvas = document.querySelector(\\\"#{canvas_id}\\\");\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append(\\\"\\\"\\\"\\n let animate_{canvas_id} = function() {{\\n if(is_in_view(this_canvas))\\n {{\\n {illo_id}.updateRenderGraph();\\n }}\\n if(document.contains(this_canvas))\\n {{\\n requestAnimationFrame(animate_{canvas_id});\\n }}\\n }};\\n animate_{canvas_id}();\\\"\\\"\\\".format(canvas_id=canvas_id, illo_id=illo_id))\\n # remove the global reference to this function after using it\\n html_lines.append('fill_{canvas_id} = null;'.format(canvas_id=canvas_id))\\n html_lines.append('};') # end of fill_{canvas_id}\\n # now call fill_{canvas_id}, possibly after loading zdog\\n html_lines.append(\\\"\\\"\\\"\\n if (typeof Zdog == 'undefined')\\n {{\\n var script = document.createElement('script');\\n script.addEventListener('load', fill_{canvas_id}, false);\\n script.src = 'https://unpkg.com/zdog@1/dist/zdog.dist.min.js';\\n document.getElementsByTagName('head')[0].appendChild(script);\\n }}\\n else\\n fill_{canvas_id}();\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append('</script>')\\n\\n return '\\\\n'.join(html_lines)\",\n \"def update(self, events: list[event.Event]) -> None:\\n if self.buffer.get_width() != self.render_width or self.buffer.get_height() != self.render_height:\\n self.pre_render()\",\n \"def render(self):\\n\\n self.desert_image.render()\\n self.cannon_image.render()\\n self.play_button.render()\\n self.escape_label.render()\",\n \"def on_draw(self, widget, cr):\\n #print \\\"starting to draw\\\"\\n if self.double_buffer is not None:\\n self.draw_tiles()\\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\\n cr.paint()\\n else:\\n print('Invalid double buffer')\\n #print \\\"done drawing\\\"\\n return False\",\n \"def render(self):\\n self.rendering = True\\n self.env.render()\",\n \"def _prepare_draw(self, view=None):\\n return True\",\n \"def pre_render(self) -> None:\\n self.buffer = Surface((self.render_width, self.render_height), SRCALPHA)\\n self.buffer.fill(list(self.halo_texture.surfaces.values())[0].get_at((0, 0)))\\n\\n self.buffer.fill((0, 0, 0, 0), Rect(\\n (self.render_width - self.halo_texture.get_width()) // 2,\\n (self.render_height - self.halo_texture.get_height()) // 2,\\n self.halo_texture.get_width(),\\n self.halo_texture.get_height()\\n ))\",\n \"def sgnRenderAllFrames(self):\\n\\n self.uiSaveSelectedRenders()\\n\\n self.core.render(self.lstSelectedRenders, bEverything = True)\",\n \"def refresh_canvas(self):\\n self.canvas.delete('all')\\n self.draw_handler(self)\\n self.canvas.after(CANVAS[\\\"REFRESH_TIME\\\"], self.refresh_canvas)\",\n \"def refresh_view(self):\\n if self._step_number % 2 == 0:\\n self._view.draw_enemies(self._game.enemies)\\n self._view.draw_towers(self._game.towers)\\n self._view.draw_obstacles(self._game.obstacles)\",\n \"def setup_draw(self):\\n pass\",\n \"def paint(self):\\r\\n pass\",\n \"def paintGL(self):\\n self._sceneviewer.renderScene()\\n # paintGL end\",\n \"def draw(self):\\n for section in self.sections:\\n canvas_reset(self.canvas)\\n section.draw(self.canvas)\",\n \"def draw(self):\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\",\n \"def _render(self, gc, points):\\n with gc:\\n gc.set_antialias(True)\\n self._draw_default_axes(gc)\\n self._draw_default_grid(gc)\\n if len(points)>0:\\n gc.clip_to_rect(self.x, self.y, self.width, self.height)\\n gc.set_stroke_color(self.color_)\\n gc.set_line_width(self.line_width)\\n gc.set_line_dash(self.line_style_)\\n\\n gc.begin_path()\\n gc.lines(points)\\n gc.stroke_path()\\n\\n return\",\n \"def prepareTestCanvas(self):\\r\\n loadLayers()\\r\\n setCanvasCrs(4326, True)\\r\\n CANVAS.resize(QtCore.QSize(400, 400))\\r\\n CANVAS.zoomToFullExtent()\",\n \"def _draw(self, canvas, options):\\n pass # must override in subclass\",\n \"def blit(self):\\n # self.ax1.draw_artist(self.lines[:2])\\n # self.ax2.draw_artist(self.lines[2:4])\\n self.ax1.autoscale()\\n self.ax2.autoscale()\\n self.ax1.redraw_in_frame()\\n self.ax2.redraw_in_frame()\\n self.fig.canvas.blit(self.fig.bbox)\",\n \"def getImmediateRendering():\\n\\n\\treturn False\",\n \"def on_paint(self, event):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the viewport, modelview and projection matrices\\n self.init_gl()\\n self.init = True\\n\\n size = self.GetClientSize()\\n text = \\\"\\\".join([_(u\\\"Canvas redrawn on paint event, size is \\\"),\\n str(size.width), \\\", \\\", str(size.height)])\\n self.render(text)\",\n \"def render(self, screen):\\n pass\",\n \"def render(self, screen):\\n pass\",\n \"def prepareForDraw(self, obj_path, camera_path, frame_context, old_data):\\n pass\",\n \"def _refresh_render(self):\\n current_frame = self.frame\\n self.frame = int(1E6)\\n self.frame = current_frame\",\n \"def draw(self, *args, **kwargs):\\n self.window.clear()\\n self.batch.draw()\",\n \"def test_visualisations_perform_visualisation_render_data_refresh_synchronously(self):\\n pass\",\n \"def renderall(self):\\n\\n if not self.isinitialized:\\n return\\n # clear display\\n self.screen.fill(BGCOLOR)\\n # draw the board\\n self.drawBoard()\\n # flip the display to show whatever we drew\\n pygame.display.flip()\",\n \"def pysweep_before_finish_init(self):\\n self.displaycanvas = DisplayCanvas(self.pysweep.master, self.boardsize, self.lcounterlength, self.rcounterlength, self.images)\\n self.displaycanvas.pack()\\n\\n self.pysweep.master.update_idletasks()\\n self.displaycanvas.update_idletasks()\\n # enode = self.arbitrary()\\n # print('DisplayCanvas:', enode)\",\n \"def render(self):\\r\\n super().render()\",\n \"def render(self) -> None:\\n if self.native_rendering:\\n self._render()\\n else:\\n self.renderer.render_image(self.get_rendered_image())\",\n \"def draw(self):\\n\\n self.updateLazyImageLoading()\\n\\n image(self.baseMap, 0, 0)\\n\\n for layer in self.layers:\\n layer.draw()\\n\\n for marker in self.markers:\\n marker.draw()\",\n \"def sgnRenderCurrentFrame(self):\\n\\n self.uiSaveSelectedRenders()\\n\\n self.core.render(self.lstSelectedRenders, bCurrentFrame = True)\",\n \"def draw(self):\\n self.batch.draw()\",\n \"def on_draw_event(self, widget, ctx):\\n # the _need_redraw flag doesnt work. it sometimes prevents\\n # the rendering and leaving the canvas blank\\n #if self._need_redraw:\\n self._renderer.set_context(ctx)\\n allocation = self.get_allocation()\\n x, y, w, h = allocation.x, allocation.y, allocation.width, allocation.height\\n self._render_figure(w, h)\\n #self._need_redraw = False\\n\\n return False # finish event propagation?\",\n \"def _drawFrame(self):\\n\\n self._clearScreen()\\n \\n for object in Object.Objects:\\n self._drawObject(object)\\n\\n for entity in Entity.Entities:\\n self._drawObject(entity)\\n\\n self._drawObject(Game.Player)\",\n \"def make(self):\\n\\t\\tif RENDER_VIEWS > 1:\\n\\t\\t\\tself._make()\",\n \"def draw(self):\\n self.write_image()\\n self.update()\",\n \"def StartDraw(self):\\r\\n self.zoom = self.test.viewZoom\\r\\n self.center = self.test.viewCenter\\r\\n self.offset = self.test.viewOffset\\r\\n self.screenSize = self.test.screenSize\",\n \"def canvas_api():\\n pass\",\n \"def draw(self) -> Any:\",\n \"def renderCanvas(self, plot, painter, canvasRect, maps):\\r\\n canvas = plot.canvas()\\r\\n r = canvasRect.adjusted(0., 0., -1., 1.)\\r\\n if self.__data.layoutFlags & self.FrameWithScales:\\r\\n painter.save()\\r\\n r.adjust(-1., -1., 1., 1.)\\r\\n painter.setPen(QPen(Qt.black))\\r\\n if not (self.__data.discardFlags & self.DiscardCanvasBackground):\\r\\n bgBrush = canvas.palette().brush(plot.backgroundRole())\\r\\n painter.setBrush(bgBrush)\\r\\n painter.drawRect(r)\\r\\n painter.restore()\\r\\n painter.save()\\r\\n painter.setClipRect(canvasRect)\\r\\n plot.drawItems(painter, canvasRect, maps)\\r\\n painter.restore()\\r\\n elif canvas.testAttribute(Qt.WA_StyledBackground):\\r\\n clipPath = QPainterPath()\\r\\n painter.save()\\r\\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\\r\\n QwtPainter.drawBackground(painter, r, canvas)\\r\\n clipPath = qwtCanvasClip(canvas, canvasRect)\\r\\n painter.restore()\\r\\n painter.save()\\r\\n if clipPath.isEmpty():\\r\\n painter.setClipRect(canvasRect)\\r\\n else:\\r\\n painter.setClipPath(clipPath)\\r\\n plot.drawItems(painter, canvasRect, maps)\\r\\n painter.restore()\\r\\n else:\\r\\n clipPath = QPainterPath()\\r\\n frameWidth = 0\\r\\n if not self.__data.discardFlags & self.DiscardCanvasFrame:\\r\\n frameWidth = canvas.frameWidth()\\r\\n clipPath = qwtCanvasClip(canvas, canvasRect)\\r\\n innerRect = canvasRect.adjusted(frameWidth, frameWidth,\\r\\n -frameWidth, -frameWidth)\\r\\n painter.save()\\r\\n if clipPath.isEmpty():\\r\\n painter.setClipRect(innerRect)\\r\\n else:\\r\\n painter.setClipPath(clipPath)\\r\\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\\r\\n QwtPainter.drawBackground(painter, innerRect, canvas)\\r\\n plot.drawItems(painter, innerRect, maps)\\r\\n painter.restore()\\r\\n if frameWidth > 0:\\r\\n painter.save()\\r\\n frameStyle = canvas.frameShadow() | canvas.frameShape()\\r\\n frameWidth = canvas.frameWidth()\\r\\n borderRadius = canvas.borderRadius()\\r\\n if borderRadius > 0.:\\r\\n QwtPainter.drawRoundedFrame(painter, canvasRect, r, r,\\r\\n canvas.palette(), frameWidth,\\r\\n frameStyle)\\r\\n else:\\r\\n midLineWidth = canvas.midLineWidth()\\r\\n QwtPainter.drawFrame(painter, canvasRect, canvas.palette(),\\r\\n canvas.foregroundRole(), frameWidth,\\r\\n midLineWidth, frameStyle)\\r\\n painter.restore()\",\n \"def reindex_graphics(self):\\n for obj in self.context.static_objects:\\n self.canvas.children.remove(obj.widget.canvas)\\n # fill _objects_z_index\\n _objects_z_index = {}\\n for obj in self.context.static_objects:\\n y = obj.widget.pos[1]\\n if not y in _objects_z_index:\\n _objects_z_index[y] = []\\n _objects_z_index[y].append(obj)\\n _keys = _objects_z_index.keys()\\n _keys.sort()\\n _keys.reverse()\\n for k in _keys:\\n objs = _objects_z_index[k]\\n for obj in objs:\\n self.canvas.add(obj.widget.canvas)\",\n \"def draw(self, canvas):\\n super().draw(canvas, self.__path)\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def on_paint(self, event):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the OpenGL rendering context\\n self.init_gl()\\n self.init = True\\n\\n size = self.GetClientSize()\\n text = \\\"\\\".join([_(u\\\"Canvas redrawn on paint event, size is \\\"),\\n str(size.width), \\\", \\\", str(size.height)])\\n self.render()\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def render(self):\\n\\n # self.screen.blit(self.background, (0, 0))\\n\\n \\\"\\\"\\\"work with ever game state\\\"\\\"\\\"\\n if self.state == FIRST_ENTER:\\n \\\"\\\"\\\"welcome screen))\\\"\\\"\\\"\\n text = self.font.render(\\\"PETERSBURG VAMPIRES\\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 3))\\n text = self.font.render(\\\"press ENTER to start\\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_WIDTH / 4, SCREEN_HEIGHT / 2))\\n elif self.state == GAME:\\n \\\"\\\"\\\"rendering enemies and player\\\"\\\"\\\"\\n old_time = time.time()\\n self.levels[self.curient_level].render(self.screen)\\n # self.levels[self.curient_level].render(self.screen)\\n for obj in self.levels[self.curient_level].objects:\\n \\\"\\\"\\\"debug mode\\\"\\\"\\\"\\n # text = self.debug_font.render(\\\"koor:\\\" + str(obj.position_np), True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]))\\n\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]))\\n #\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+obj.image_size[Y]))\\n #\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]+obj.image_size[Y]))\\n #\\n # x, y = get_grid_xy(obj.position_np, ZOMBIE_SIZE)\\n # text = self.debug_font.render(\\\"x_grid: \\\" + str(x) + \\\",y_grid: \\\" + str(y), True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+20))\\n\\n obj.render(self.screen)\\n \\\"\\\"\\\"rendering HP status\\\"\\\"\\\"\\n text = self.font.render(\\\"HP: \\\" + str(self.player.HP), True, (255, 0, 0))\\n self.screen.blit(text, (0, 10))\\n new_time = time.time()\\n self.render_time.append(new_time - old_time)\\n elif self.state == END:\\n \\\"\\\"\\\"The end screen\\\"\\\"\\\"\\n self.screen.blit(pygame.image.load(END_BUTTON_PATH), (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 8))\\n elif self.state == WIN:\\n text = self.font.render(\\\"YOU WIN, REPEAT? \\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_HEIGHT / 2, SCREEN_WIDTH / 2))\\n\\n pygame.display.flip()\",\n \"def forward_partial(self, drawing):\\n assert not self.is_cuda\\n drawing = drawings_to_cpu(drawing)\\n ns = len(drawing)\\n pimgs = torch.zeros(ns+1, *self.imsize)\\n canvas = torch.zeros(*self.imsize)\\n for i, stk in enumerate(drawing):\\n canvas, _ = self.add_stroke(canvas, stk)\\n pimgs[i+1] = canvas\\n\\n return pimgs\",\n \"def render(self):\\n gl.glEnableClientState(gl.GL_VERTEX_ARRAY);\\n gl.glEnableClientState(gl.GL_NORMAL_ARRAY);\\n gl.glVertexPointer(3, gl.GL_DOUBLE, 0, self._vertices);\\n gl.glNormalPointer(gl.GL_DOUBLE, 0, self._normals);\\n for patch in self._patches:\\n patch.render()\\n gl.glDisableClientState(gl.GL_VERTEX_ARRAY);\\n gl.glDisableClientState(gl.GL_NORMAL_ARRAY);\",\n \"def _before_render_event(self, *args, **kwargs):\\n for chart in self._charts:\\n chart._render_event(*args, **kwargs)\",\n \"def draw(self):\\n for obj in self.objects:\\n obj.draw()\",\n \"def render(self):\\n self.screen.fill(prepare.BACKGROUND_COLOR)\\n self.health_bar()\\n # self.enemy_health()\\n self.energy_bar()\\n self.level.draw(self.screen)\\n pg.display.update()\",\n \"def get(self):\\n self.render_front()\",\n \"def render(self):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the OpenGL rendering context\\n self.init_gl()\\n self.init = True\\n\\n # Clear everything\\n GL.glClear(GL.GL_COLOR_BUFFER_BIT | GL.GL_DEPTH_BUFFER_BIT)\\n\\n for device_id, output_id in self.monitors.monitors_dictionary:\\n i = list(\\n self.monitors.monitors_dictionary.keys()).index(\\n (device_id, output_id))\\n GL.glColor3f(self.colours[i%8][0], self.colours[i%8][1], self.colours[i%8][2])\\n x = (i - len(self.monitors.monitors_dictionary)/2)*20\\n signal_list = self.monitors.monitors_dictionary[(\\n device_id, output_id)]\\n\\n signal_name = self.devices.get_signal_name(device_id, output_id)\\n\\n self.render_text(signal_name, x, 0, 10*len(signal_list)+10)\\n\\n length = len(signal_list)\\n for j in range(length):\\n signal = signal_list[j]\\n z = (j-length/2)*20\\n if signal == self.devices.HIGH:\\n y = 11\\n if signal == self.devices.LOW:\\n y = 1\\n if signal != self.devices.BLANK:\\n self.draw_cuboid(x, z, 5, 10, y)\\n\\n # We have been drawing to the back buffer, flush the graphics pipeline\\n # and swap the back buffer to the front\\n GL.glFlush()\\n self.SwapBuffers()\",\n \"def display(self):\\n\\t\\tfor c in self.canvas.values():\\n\\t\\t\\tc.update()\\n\\t\\t\\tself.superficie.blit(c.superficie, c.origen)\\n\\t\\tpygame.display.flip()\",\n \"def render(self, mode='human'):\\n if self.microgridRender == None : self.microgridRender = MicrogridRender(self.microgrid, self.clock);\\t\\n self.microgridRender.render(mode);\",\n \"def finish_render():\\n get_window().static_display = True\\n get_window().flip_count = 0\\n get_window().flip()\",\n \"def _pre_draw_bge(self):\\r\\n self._pre_draw_common()\\r\\n # draw rays\\r\\n self._drawRays()\",\n \"def on_draw(self):\\n self.clear()\\n self.manager.draw()\",\n \"def render(self, chart):\\n chart.create_visualization_files(self.__outputpath)\",\n \"def draw(self, screen):\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.62798715","0.62798715","0.6229959","0.62175214","0.62031156","0.6023987","0.5880108","0.58327454","0.58257514","0.5803622","0.5789395","0.57692194","0.5759491","0.57456064","0.57041544","0.5677518","0.5677518","0.5677518","0.5677518","0.56731313","0.5672394","0.5657822","0.565429","0.562781","0.5625311","0.56140006","0.56121695","0.55965394","0.55932885","0.55922323","0.5589038","0.55629504","0.55365866","0.5531392","0.5521091","0.5502241","0.54994035","0.54946756","0.54896766","0.54896766","0.54896766","0.54896766","0.54896766","0.54896766","0.54680127","0.54665613","0.54640025","0.5456918","0.5451559","0.54473037","0.542622","0.5422615","0.5422615","0.5410161","0.5407036","0.54055995","0.54025245","0.53979194","0.53894186","0.53868896","0.5386106","0.53830636","0.5346759","0.53420955","0.5334694","0.5328274","0.5325468","0.5322357","0.5320627","0.53074044","0.53065264","0.5305843","0.5303125","0.5301566","0.52959347","0.52959347","0.52959347","0.52919894","0.52879196","0.52879196","0.52879196","0.52879196","0.52879196","0.52879196","0.52879196","0.52879196","0.52852386","0.52825654","0.5281756","0.528146","0.5281257","0.5277416","0.5271197","0.52675444","0.5259245","0.5258945","0.52587706","0.5256811","0.5253516","0.5253171","0.52386785"],"string":"[\n \"0.62798715\",\n \"0.62798715\",\n \"0.6229959\",\n \"0.62175214\",\n \"0.62031156\",\n \"0.6023987\",\n \"0.5880108\",\n \"0.58327454\",\n \"0.58257514\",\n \"0.5803622\",\n \"0.5789395\",\n \"0.57692194\",\n \"0.5759491\",\n \"0.57456064\",\n \"0.57041544\",\n \"0.5677518\",\n \"0.5677518\",\n \"0.5677518\",\n \"0.5677518\",\n \"0.56731313\",\n \"0.5672394\",\n \"0.5657822\",\n \"0.565429\",\n \"0.562781\",\n \"0.5625311\",\n \"0.56140006\",\n \"0.56121695\",\n \"0.55965394\",\n \"0.55932885\",\n \"0.55922323\",\n \"0.5589038\",\n \"0.55629504\",\n \"0.55365866\",\n \"0.5531392\",\n \"0.5521091\",\n \"0.5502241\",\n \"0.54994035\",\n \"0.54946756\",\n \"0.54896766\",\n \"0.54896766\",\n \"0.54896766\",\n \"0.54896766\",\n \"0.54896766\",\n \"0.54896766\",\n \"0.54680127\",\n \"0.54665613\",\n \"0.54640025\",\n \"0.5456918\",\n \"0.5451559\",\n \"0.54473037\",\n \"0.542622\",\n \"0.5422615\",\n \"0.5422615\",\n \"0.5410161\",\n \"0.5407036\",\n \"0.54055995\",\n \"0.54025245\",\n \"0.53979194\",\n \"0.53894186\",\n \"0.53868896\",\n \"0.5386106\",\n \"0.53830636\",\n \"0.5346759\",\n \"0.53420955\",\n \"0.5334694\",\n \"0.5328274\",\n \"0.5325468\",\n \"0.5322357\",\n \"0.5320627\",\n \"0.53074044\",\n \"0.53065264\",\n \"0.5305843\",\n \"0.5303125\",\n \"0.5301566\",\n \"0.52959347\",\n \"0.52959347\",\n \"0.52959347\",\n \"0.52919894\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52879196\",\n \"0.52852386\",\n \"0.52825654\",\n \"0.5281756\",\n \"0.528146\",\n \"0.5281257\",\n \"0.5277416\",\n \"0.5271197\",\n \"0.52675444\",\n \"0.5259245\",\n \"0.5258945\",\n \"0.52587706\",\n \"0.5256811\",\n \"0.5253516\",\n \"0.5253171\",\n \"0.52386785\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90871,"cells":{"query":{"kind":"string","value":"In this version, we include all partial canvas renders in addition to the final renders"},"document":{"kind":"string","value":"def forward_partial(self, drawing):\n assert not self.is_cuda\n drawing = drawings_to_cpu(drawing)\n ns = len(drawing)\n pimgs = torch.zeros(ns+1, *self.imsize)\n canvas = torch.zeros(*self.imsize)\n for i, stk in enumerate(drawing):\n canvas, _ = self.add_stroke(canvas, stk)\n pimgs[i+1] = canvas\n\n return pimgs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def on_draw(self):\n\t\tself.render()","def on_draw(self):\n\t\tself.render()","def dspyRender(self):\n pass","def on_draw(self):\n # draw everything","def render(self):\n self.axial.Render()\n self.coronal.Render()\n self.sagittal.Render()\n #self.isosurface.Render()\n #self.rwi_pcp.Render()","def pre_draw(self):","def _render_callback(self, _sim, _viewer):\n pass","def on_draw( self ):\r\n self.clear()\r\n self.setup_3D()\r\n print \"DEBUG:\" , \"There are\" , len( self.renderlist ) , \"items in 'self.renderlist'\"\r\n for obj in self.renderlist:\r\n obj.draw()","def _automatic_rendering(self):\n if self.viewer is not None and self.enable_auto_render:\n self.should_update_rendering = True\n\n if self.automatic_rendering_callback:\n self.automatic_rendering_callback()\n else:\n self.render(self.rendering_mode)","def _automatic_rendering(self) -> None:\n if self.viewer is not None and self.enable_auto_render:\n\n if self._record_video_wrapper and self._record_video_wrapper.video_recorder:\n self._record_video_wrapper.video_recorder.capture_frame()\n else:\n self.render()","def _prepare_draw(self, view=None):\n pass","def add_canvas():\n return render_template('courses/add_canvas.html')","def draw(self):\n\t\tpass","def draw(self):\n self.bufferX = (self.appWidth/2) - self.viewX\n self.bufferY = (self.appHeight/2) - self.viewY\n anwp.sl.engine.clear()\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\n self.drawWarpLines()\n \n # render engine\n anwp.sl.engine.render()\n self.drawSystemInfo()\n self.drawWarpGateInfo()\n self.drawWarpTradeInfo()","def render(self, visual_elements, canvas):\n self.coordinate_system = visual_elements.coordinate_system\n self.canvas = canvas\n for ve in visual_elements.elements: self.add(ve)\n return","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def draw(self):\n pass","def on_draw_overlay(self):","def _render(self) -> None:\n pass","def _drawOnCanvas(self):\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\n for key in self.elements:\n graphElement=self.elements[key]\n graphElement.draw(self.canvas)\n self.sync=True","def redraw(self, **kwargs):\n #src_dict = self.data_sources\n #self.remove_sources(src_dict.keys())\n self.renderers = {}\n #self.renderers = {}\n self.figure = self.draw_figure(**kwargs)\n #self.add_sources(src_dict)\n # todo does the old figure linger on?\n self.render_sources(self.data_sources)\n self.bk_pane.object = self.figure","def render(self):\n canvas_id = 'zdog_{}'.format(self.CANVAS_INDEX)\n illo_id = 'illo_{}'.format(self.CANVAS_INDEX)\n Scene.CANVAS_INDEX += 1\n\n html_lines = []\n\n js_lines = []\n\n euler = -rowan.to_euler(\n self.rotation, convention='xyz', axis_type='intrinsic')\n translation = self.translation*(1, -1, 1)\n\n pan_cfg = self.get_feature_config('pan')\n pan = pan_cfg.get('value', True) if pan_cfg is not None else False\n\n js_lines.append(\"\"\"\n let {illo_id} = new Zdog.Illustration({{\n element: '#{canvas_id}',\n zoom: {zoom},\n dragRotate: {rotation_enabled},\n rotate: {{x: {angle[0]}, y: {angle[1]}, z: {angle[2]}}},\n translate: {{x: {pos[0]}, y: {pos[1]}, z: {pos[2]}}},\n }});\n \"\"\".format(\n illo_id=illo_id, canvas_id=canvas_id, zoom=self.zoom*self.pixel_scale,\n angle=euler, pos=translation,\n rotation_enabled=('false' if pan else 'true')))\n\n config = self.get_feature_config('ambient_light')\n ambient_light = 0 if config is None else config.get('value', .4)\n\n config = self.get_feature_config('directional_light')\n directional_light = ([(0, 0, 0)] if config is None else\n config.get('value', [(0, 0, 0)]))\n directional_light = np.atleast_2d(directional_light)\n\n shapeIndex = 0\n for i, prim in enumerate(self._primitives):\n js_lines.extend(prim.render(\n rotation=self.rotation, illo_id=illo_id,\n name_suffix=i, ambient_light=ambient_light,\n directional_light=directional_light))\n\n (width, height) = map(int, self.size_pixels)\n html_lines.append(\"\"\"\n <canvas id=\"{canvas_id}\" width=\"{width}\" height=\"{height}\"></canvas>\n \"\"\".format(canvas_id=canvas_id, width=width, height=height))\n\n html_lines.append(\"\"\"<script>\n var fill_{canvas_id} = function() {{\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append(LOCAL_HELPER_SCRIPT)\n html_lines.extend(js_lines)\n\n pan_snippet = \"\"\"\n new Zdog.Dragger({{\n startElement: {illo_id}.element,\n onDragStart: function( pointer, moveX, moveY) {{\n this.lastX = 0;\n this.lastY = 0;\n }},\n onDragMove: function( pointer, moveX, moveY ) {{\n let deltax = moveX - this.lastX;\n let deltay = moveY - this.lastY;\n let scale = 1.0/{illo_id}.zoom;\n {illo_id}.translate.x += deltax*scale;\n {illo_id}.translate.y += deltay*scale;\n this.lastX = moveX;\n this.lastY = moveY;\n }}\n }});\"\"\".format(illo_id=illo_id)\n if pan:\n html_lines.append(pan_snippet)\n\n html_lines.append(\"\"\"\n let this_canvas = document.querySelector(\"#{canvas_id}\");\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append(\"\"\"\n let animate_{canvas_id} = function() {{\n if(is_in_view(this_canvas))\n {{\n {illo_id}.updateRenderGraph();\n }}\n if(document.contains(this_canvas))\n {{\n requestAnimationFrame(animate_{canvas_id});\n }}\n }};\n animate_{canvas_id}();\"\"\".format(canvas_id=canvas_id, illo_id=illo_id))\n # remove the global reference to this function after using it\n html_lines.append('fill_{canvas_id} = null;'.format(canvas_id=canvas_id))\n html_lines.append('};') # end of fill_{canvas_id}\n # now call fill_{canvas_id}, possibly after loading zdog\n html_lines.append(\"\"\"\n if (typeof Zdog == 'undefined')\n {{\n var script = document.createElement('script');\n script.addEventListener('load', fill_{canvas_id}, false);\n script.src = 'https://unpkg.com/zdog@1/dist/zdog.dist.min.js';\n document.getElementsByTagName('head')[0].appendChild(script);\n }}\n else\n fill_{canvas_id}();\n \"\"\".format(canvas_id=canvas_id))\n html_lines.append('</script>')\n\n return '\\n'.join(html_lines)","def update(self, events: list[event.Event]) -> None:\n if self.buffer.get_width() != self.render_width or self.buffer.get_height() != self.render_height:\n self.pre_render()","def render(self):\n\n self.desert_image.render()\n self.cannon_image.render()\n self.play_button.render()\n self.escape_label.render()","def on_draw(self, widget, cr):\n #print \"starting to draw\"\n if self.double_buffer is not None:\n self.draw_tiles()\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\n cr.paint()\n else:\n print('Invalid double buffer')\n #print \"done drawing\"\n return False","def render(self):\n self.rendering = True\n self.env.render()","def _prepare_draw(self, view=None):\n return True","def pre_render(self) -> None:\n self.buffer = Surface((self.render_width, self.render_height), SRCALPHA)\n self.buffer.fill(list(self.halo_texture.surfaces.values())[0].get_at((0, 0)))\n\n self.buffer.fill((0, 0, 0, 0), Rect(\n (self.render_width - self.halo_texture.get_width()) // 2,\n (self.render_height - self.halo_texture.get_height()) // 2,\n self.halo_texture.get_width(),\n self.halo_texture.get_height()\n ))","def sgnRenderAllFrames(self):\n\n self.uiSaveSelectedRenders()\n\n self.core.render(self.lstSelectedRenders, bEverything = True)","def refresh_canvas(self):\n self.canvas.delete('all')\n self.draw_handler(self)\n self.canvas.after(CANVAS[\"REFRESH_TIME\"], self.refresh_canvas)","def refresh_view(self):\n if self._step_number % 2 == 0:\n self._view.draw_enemies(self._game.enemies)\n self._view.draw_towers(self._game.towers)\n self._view.draw_obstacles(self._game.obstacles)","def setup_draw(self):\n pass","def paint(self):\r\n pass","def paintGL(self):\n self._sceneviewer.renderScene()\n # paintGL end","def draw(self):\n for section in self.sections:\n canvas_reset(self.canvas)\n section.draw(self.canvas)","def draw(self):","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def render(self):\n pass","def _render(self, gc, points):\n with gc:\n gc.set_antialias(True)\n self._draw_default_axes(gc)\n self._draw_default_grid(gc)\n if len(points)>0:\n gc.clip_to_rect(self.x, self.y, self.width, self.height)\n gc.set_stroke_color(self.color_)\n gc.set_line_width(self.line_width)\n gc.set_line_dash(self.line_style_)\n\n gc.begin_path()\n gc.lines(points)\n gc.stroke_path()\n\n return","def render(self):","def prepareTestCanvas(self):\r\n loadLayers()\r\n setCanvasCrs(4326, True)\r\n CANVAS.resize(QtCore.QSize(400, 400))\r\n CANVAS.zoomToFullExtent()","def _draw(self, canvas, options):\n pass # must override in subclass","def blit(self):\n # self.ax1.draw_artist(self.lines[:2])\n # self.ax2.draw_artist(self.lines[2:4])\n self.ax1.autoscale()\n self.ax2.autoscale()\n self.ax1.redraw_in_frame()\n self.ax2.redraw_in_frame()\n self.fig.canvas.blit(self.fig.bbox)","def getImmediateRendering():\n\n\treturn False","def on_paint(self, event):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the viewport, modelview and projection matrices\n self.init_gl()\n self.init = True\n\n size = self.GetClientSize()\n text = \"\".join([_(u\"Canvas redrawn on paint event, size is \"),\n str(size.width), \", \", str(size.height)])\n self.render(text)","def render(self, screen):\n pass","def render(self, screen):\n pass","def prepareForDraw(self, obj_path, camera_path, frame_context, old_data):\n pass","def draw(self, *args, **kwargs):\n self.window.clear()\n self.batch.draw()","def _refresh_render(self):\n current_frame = self.frame\n self.frame = int(1E6)\n self.frame = current_frame","def test_visualisations_perform_visualisation_render_data_refresh_synchronously(self):\n pass","def renderall(self):\n\n if not self.isinitialized:\n return\n # clear display\n self.screen.fill(BGCOLOR)\n # draw the board\n self.drawBoard()\n # flip the display to show whatever we drew\n pygame.display.flip()","def pysweep_before_finish_init(self):\n self.displaycanvas = DisplayCanvas(self.pysweep.master, self.boardsize, self.lcounterlength, self.rcounterlength, self.images)\n self.displaycanvas.pack()\n\n self.pysweep.master.update_idletasks()\n self.displaycanvas.update_idletasks()\n # enode = self.arbitrary()\n # print('DisplayCanvas:', enode)","def render(self) -> None:\n if self.native_rendering:\n self._render()\n else:\n self.renderer.render_image(self.get_rendered_image())","def render(self):\r\n super().render()","def draw(self):\n\n self.updateLazyImageLoading()\n\n image(self.baseMap, 0, 0)\n\n for layer in self.layers:\n layer.draw()\n\n for marker in self.markers:\n marker.draw()","def sgnRenderCurrentFrame(self):\n\n self.uiSaveSelectedRenders()\n\n self.core.render(self.lstSelectedRenders, bCurrentFrame = True)","def draw(self):\n self.batch.draw()","def on_draw_event(self, widget, ctx):\n # the _need_redraw flag doesnt work. it sometimes prevents\n # the rendering and leaving the canvas blank\n #if self._need_redraw:\n self._renderer.set_context(ctx)\n allocation = self.get_allocation()\n x, y, w, h = allocation.x, allocation.y, allocation.width, allocation.height\n self._render_figure(w, h)\n #self._need_redraw = False\n\n return False # finish event propagation?","def _drawFrame(self):\n\n self._clearScreen()\n \n for object in Object.Objects:\n self._drawObject(object)\n\n for entity in Entity.Entities:\n self._drawObject(entity)\n\n self._drawObject(Game.Player)","def make(self):\n\t\tif RENDER_VIEWS > 1:\n\t\t\tself._make()","def draw(self):\n self.write_image()\n self.update()","def StartDraw(self):\r\n self.zoom = self.test.viewZoom\r\n self.center = self.test.viewCenter\r\n self.offset = self.test.viewOffset\r\n self.screenSize = self.test.screenSize","def canvas_api():\n pass","def draw(self) -> Any:","def renderCanvas(self, plot, painter, canvasRect, maps):\r\n canvas = plot.canvas()\r\n r = canvasRect.adjusted(0., 0., -1., 1.)\r\n if self.__data.layoutFlags & self.FrameWithScales:\r\n painter.save()\r\n r.adjust(-1., -1., 1., 1.)\r\n painter.setPen(QPen(Qt.black))\r\n if not (self.__data.discardFlags & self.DiscardCanvasBackground):\r\n bgBrush = canvas.palette().brush(plot.backgroundRole())\r\n painter.setBrush(bgBrush)\r\n painter.drawRect(r)\r\n painter.restore()\r\n painter.save()\r\n painter.setClipRect(canvasRect)\r\n plot.drawItems(painter, canvasRect, maps)\r\n painter.restore()\r\n elif canvas.testAttribute(Qt.WA_StyledBackground):\r\n clipPath = QPainterPath()\r\n painter.save()\r\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\r\n QwtPainter.drawBackground(painter, r, canvas)\r\n clipPath = qwtCanvasClip(canvas, canvasRect)\r\n painter.restore()\r\n painter.save()\r\n if clipPath.isEmpty():\r\n painter.setClipRect(canvasRect)\r\n else:\r\n painter.setClipPath(clipPath)\r\n plot.drawItems(painter, canvasRect, maps)\r\n painter.restore()\r\n else:\r\n clipPath = QPainterPath()\r\n frameWidth = 0\r\n if not self.__data.discardFlags & self.DiscardCanvasFrame:\r\n frameWidth = canvas.frameWidth()\r\n clipPath = qwtCanvasClip(canvas, canvasRect)\r\n innerRect = canvasRect.adjusted(frameWidth, frameWidth,\r\n -frameWidth, -frameWidth)\r\n painter.save()\r\n if clipPath.isEmpty():\r\n painter.setClipRect(innerRect)\r\n else:\r\n painter.setClipPath(clipPath)\r\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\r\n QwtPainter.drawBackground(painter, innerRect, canvas)\r\n plot.drawItems(painter, innerRect, maps)\r\n painter.restore()\r\n if frameWidth > 0:\r\n painter.save()\r\n frameStyle = canvas.frameShadow() | canvas.frameShape()\r\n frameWidth = canvas.frameWidth()\r\n borderRadius = canvas.borderRadius()\r\n if borderRadius > 0.:\r\n QwtPainter.drawRoundedFrame(painter, canvasRect, r, r,\r\n canvas.palette(), frameWidth,\r\n frameStyle)\r\n else:\r\n midLineWidth = canvas.midLineWidth()\r\n QwtPainter.drawFrame(painter, canvasRect, canvas.palette(),\r\n canvas.foregroundRole(), frameWidth,\r\n midLineWidth, frameStyle)\r\n painter.restore()","def reindex_graphics(self):\n for obj in self.context.static_objects:\n self.canvas.children.remove(obj.widget.canvas)\n # fill _objects_z_index\n _objects_z_index = {}\n for obj in self.context.static_objects:\n y = obj.widget.pos[1]\n if not y in _objects_z_index:\n _objects_z_index[y] = []\n _objects_z_index[y].append(obj)\n _keys = _objects_z_index.keys()\n _keys.sort()\n _keys.reverse()\n for k in _keys:\n objs = _objects_z_index[k]\n for obj in objs:\n self.canvas.add(obj.widget.canvas)","def draw(self, canvas):\n super().draw(canvas, self.__path)","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def draw(self):\n raise NotImplementedError","def on_paint(self, event):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the OpenGL rendering context\n self.init_gl()\n self.init = True\n\n size = self.GetClientSize()\n text = \"\".join([_(u\"Canvas redrawn on paint event, size is \"),\n str(size.width), \", \", str(size.height)])\n self.render()","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def draw(self): \n [component.draw() for component in self.components]","def render(self):\n\n # self.screen.blit(self.background, (0, 0))\n\n \"\"\"work with ever game state\"\"\"\n if self.state == FIRST_ENTER:\n \"\"\"welcome screen))\"\"\"\n text = self.font.render(\"PETERSBURG VAMPIRES\", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 3))\n text = self.font.render(\"press ENTER to start\", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_WIDTH / 4, SCREEN_HEIGHT / 2))\n elif self.state == GAME:\n \"\"\"rendering enemies and player\"\"\"\n old_time = time.time()\n self.levels[self.curient_level].render(self.screen)\n # self.levels[self.curient_level].render(self.screen)\n for obj in self.levels[self.curient_level].objects:\n \"\"\"debug mode\"\"\"\n # text = self.debug_font.render(\"koor:\" + str(obj.position_np), True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]))\n\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]))\n #\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+obj.image_size[Y]))\n #\n # text = self.debug_font.render(\"x\", True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]+obj.image_size[Y]))\n #\n # x, y = get_grid_xy(obj.position_np, ZOMBIE_SIZE)\n # text = self.debug_font.render(\"x_grid: \" + str(x) + \",y_grid: \" + str(y), True, (255, 0, 0))\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+20))\n\n obj.render(self.screen)\n \"\"\"rendering HP status\"\"\"\n text = self.font.render(\"HP: \" + str(self.player.HP), True, (255, 0, 0))\n self.screen.blit(text, (0, 10))\n new_time = time.time()\n self.render_time.append(new_time - old_time)\n elif self.state == END:\n \"\"\"The end screen\"\"\"\n self.screen.blit(pygame.image.load(END_BUTTON_PATH), (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 8))\n elif self.state == WIN:\n text = self.font.render(\"YOU WIN, REPEAT? \", True, (255, 0, 0))\n self.screen.blit(text, (SCREEN_HEIGHT / 2, SCREEN_WIDTH / 2))\n\n pygame.display.flip()","def render(self):\n gl.glEnableClientState(gl.GL_VERTEX_ARRAY);\n gl.glEnableClientState(gl.GL_NORMAL_ARRAY);\n gl.glVertexPointer(3, gl.GL_DOUBLE, 0, self._vertices);\n gl.glNormalPointer(gl.GL_DOUBLE, 0, self._normals);\n for patch in self._patches:\n patch.render()\n gl.glDisableClientState(gl.GL_VERTEX_ARRAY);\n gl.glDisableClientState(gl.GL_NORMAL_ARRAY);","def draw(self):\n for obj in self.objects:\n obj.draw()","def _before_render_event(self, *args, **kwargs):\n for chart in self._charts:\n chart._render_event(*args, **kwargs)","def render(self):\n self.screen.fill(prepare.BACKGROUND_COLOR)\n self.health_bar()\n # self.enemy_health()\n self.energy_bar()\n self.level.draw(self.screen)\n pg.display.update()","def get(self):\n self.render_front()","def render(self):\n self.SetCurrent(self.context)\n if not self.init:\n # Configure the OpenGL rendering context\n self.init_gl()\n self.init = True\n\n # Clear everything\n GL.glClear(GL.GL_COLOR_BUFFER_BIT | GL.GL_DEPTH_BUFFER_BIT)\n\n for device_id, output_id in self.monitors.monitors_dictionary:\n i = list(\n self.monitors.monitors_dictionary.keys()).index(\n (device_id, output_id))\n GL.glColor3f(self.colours[i%8][0], self.colours[i%8][1], self.colours[i%8][2])\n x = (i - len(self.monitors.monitors_dictionary)/2)*20\n signal_list = self.monitors.monitors_dictionary[(\n device_id, output_id)]\n\n signal_name = self.devices.get_signal_name(device_id, output_id)\n\n self.render_text(signal_name, x, 0, 10*len(signal_list)+10)\n\n length = len(signal_list)\n for j in range(length):\n signal = signal_list[j]\n z = (j-length/2)*20\n if signal == self.devices.HIGH:\n y = 11\n if signal == self.devices.LOW:\n y = 1\n if signal != self.devices.BLANK:\n self.draw_cuboid(x, z, 5, 10, y)\n\n # We have been drawing to the back buffer, flush the graphics pipeline\n # and swap the back buffer to the front\n GL.glFlush()\n self.SwapBuffers()","def display(self):\n\t\tfor c in self.canvas.values():\n\t\t\tc.update()\n\t\t\tself.superficie.blit(c.superficie, c.origen)\n\t\tpygame.display.flip()","def finish_render():\n get_window().static_display = True\n get_window().flip_count = 0\n get_window().flip()","def render(self, mode='human'):\n if self.microgridRender == None : self.microgridRender = MicrogridRender(self.microgrid, self.clock);\t\n self.microgridRender.render(mode);","def _pre_draw_bge(self):\r\n self._pre_draw_common()\r\n # draw rays\r\n self._drawRays()","def on_draw(self):\n self.clear()\n self.manager.draw()","def render(self, chart):\n chart.create_visualization_files(self.__outputpath)","def draw(self, screen):"],"string":"[\n \"def on_draw(self):\\n\\t\\tself.render()\",\n \"def on_draw(self):\\n\\t\\tself.render()\",\n \"def dspyRender(self):\\n pass\",\n \"def on_draw(self):\\n # draw everything\",\n \"def render(self):\\n self.axial.Render()\\n self.coronal.Render()\\n self.sagittal.Render()\\n #self.isosurface.Render()\\n #self.rwi_pcp.Render()\",\n \"def pre_draw(self):\",\n \"def _render_callback(self, _sim, _viewer):\\n pass\",\n \"def on_draw( self ):\\r\\n self.clear()\\r\\n self.setup_3D()\\r\\n print \\\"DEBUG:\\\" , \\\"There are\\\" , len( self.renderlist ) , \\\"items in 'self.renderlist'\\\"\\r\\n for obj in self.renderlist:\\r\\n obj.draw()\",\n \"def _automatic_rendering(self):\\n if self.viewer is not None and self.enable_auto_render:\\n self.should_update_rendering = True\\n\\n if self.automatic_rendering_callback:\\n self.automatic_rendering_callback()\\n else:\\n self.render(self.rendering_mode)\",\n \"def _automatic_rendering(self) -> None:\\n if self.viewer is not None and self.enable_auto_render:\\n\\n if self._record_video_wrapper and self._record_video_wrapper.video_recorder:\\n self._record_video_wrapper.video_recorder.capture_frame()\\n else:\\n self.render()\",\n \"def _prepare_draw(self, view=None):\\n pass\",\n \"def add_canvas():\\n return render_template('courses/add_canvas.html')\",\n \"def draw(self):\\n\\t\\tpass\",\n \"def draw(self):\\n self.bufferX = (self.appWidth/2) - self.viewX\\n self.bufferY = (self.appHeight/2) - self.viewY\\n anwp.sl.engine.clear()\\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\\n self.drawWarpLines()\\n \\n # render engine\\n anwp.sl.engine.render()\\n self.drawSystemInfo()\\n self.drawWarpGateInfo()\\n self.drawWarpTradeInfo()\",\n \"def render(self, visual_elements, canvas):\\n self.coordinate_system = visual_elements.coordinate_system\\n self.canvas = canvas\\n for ve in visual_elements.elements: self.add(ve)\\n return\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def draw(self):\\n pass\",\n \"def on_draw_overlay(self):\",\n \"def _render(self) -> None:\\n pass\",\n \"def _drawOnCanvas(self):\\n self.canvas=np.ones(self.canvas.shape,dtype=np.uint8)*255\\n for key in self.elements:\\n graphElement=self.elements[key]\\n graphElement.draw(self.canvas)\\n self.sync=True\",\n \"def redraw(self, **kwargs):\\n #src_dict = self.data_sources\\n #self.remove_sources(src_dict.keys())\\n self.renderers = {}\\n #self.renderers = {}\\n self.figure = self.draw_figure(**kwargs)\\n #self.add_sources(src_dict)\\n # todo does the old figure linger on?\\n self.render_sources(self.data_sources)\\n self.bk_pane.object = self.figure\",\n \"def render(self):\\n canvas_id = 'zdog_{}'.format(self.CANVAS_INDEX)\\n illo_id = 'illo_{}'.format(self.CANVAS_INDEX)\\n Scene.CANVAS_INDEX += 1\\n\\n html_lines = []\\n\\n js_lines = []\\n\\n euler = -rowan.to_euler(\\n self.rotation, convention='xyz', axis_type='intrinsic')\\n translation = self.translation*(1, -1, 1)\\n\\n pan_cfg = self.get_feature_config('pan')\\n pan = pan_cfg.get('value', True) if pan_cfg is not None else False\\n\\n js_lines.append(\\\"\\\"\\\"\\n let {illo_id} = new Zdog.Illustration({{\\n element: '#{canvas_id}',\\n zoom: {zoom},\\n dragRotate: {rotation_enabled},\\n rotate: {{x: {angle[0]}, y: {angle[1]}, z: {angle[2]}}},\\n translate: {{x: {pos[0]}, y: {pos[1]}, z: {pos[2]}}},\\n }});\\n \\\"\\\"\\\".format(\\n illo_id=illo_id, canvas_id=canvas_id, zoom=self.zoom*self.pixel_scale,\\n angle=euler, pos=translation,\\n rotation_enabled=('false' if pan else 'true')))\\n\\n config = self.get_feature_config('ambient_light')\\n ambient_light = 0 if config is None else config.get('value', .4)\\n\\n config = self.get_feature_config('directional_light')\\n directional_light = ([(0, 0, 0)] if config is None else\\n config.get('value', [(0, 0, 0)]))\\n directional_light = np.atleast_2d(directional_light)\\n\\n shapeIndex = 0\\n for i, prim in enumerate(self._primitives):\\n js_lines.extend(prim.render(\\n rotation=self.rotation, illo_id=illo_id,\\n name_suffix=i, ambient_light=ambient_light,\\n directional_light=directional_light))\\n\\n (width, height) = map(int, self.size_pixels)\\n html_lines.append(\\\"\\\"\\\"\\n <canvas id=\\\"{canvas_id}\\\" width=\\\"{width}\\\" height=\\\"{height}\\\"></canvas>\\n \\\"\\\"\\\".format(canvas_id=canvas_id, width=width, height=height))\\n\\n html_lines.append(\\\"\\\"\\\"<script>\\n var fill_{canvas_id} = function() {{\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append(LOCAL_HELPER_SCRIPT)\\n html_lines.extend(js_lines)\\n\\n pan_snippet = \\\"\\\"\\\"\\n new Zdog.Dragger({{\\n startElement: {illo_id}.element,\\n onDragStart: function( pointer, moveX, moveY) {{\\n this.lastX = 0;\\n this.lastY = 0;\\n }},\\n onDragMove: function( pointer, moveX, moveY ) {{\\n let deltax = moveX - this.lastX;\\n let deltay = moveY - this.lastY;\\n let scale = 1.0/{illo_id}.zoom;\\n {illo_id}.translate.x += deltax*scale;\\n {illo_id}.translate.y += deltay*scale;\\n this.lastX = moveX;\\n this.lastY = moveY;\\n }}\\n }});\\\"\\\"\\\".format(illo_id=illo_id)\\n if pan:\\n html_lines.append(pan_snippet)\\n\\n html_lines.append(\\\"\\\"\\\"\\n let this_canvas = document.querySelector(\\\"#{canvas_id}\\\");\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append(\\\"\\\"\\\"\\n let animate_{canvas_id} = function() {{\\n if(is_in_view(this_canvas))\\n {{\\n {illo_id}.updateRenderGraph();\\n }}\\n if(document.contains(this_canvas))\\n {{\\n requestAnimationFrame(animate_{canvas_id});\\n }}\\n }};\\n animate_{canvas_id}();\\\"\\\"\\\".format(canvas_id=canvas_id, illo_id=illo_id))\\n # remove the global reference to this function after using it\\n html_lines.append('fill_{canvas_id} = null;'.format(canvas_id=canvas_id))\\n html_lines.append('};') # end of fill_{canvas_id}\\n # now call fill_{canvas_id}, possibly after loading zdog\\n html_lines.append(\\\"\\\"\\\"\\n if (typeof Zdog == 'undefined')\\n {{\\n var script = document.createElement('script');\\n script.addEventListener('load', fill_{canvas_id}, false);\\n script.src = 'https://unpkg.com/zdog@1/dist/zdog.dist.min.js';\\n document.getElementsByTagName('head')[0].appendChild(script);\\n }}\\n else\\n fill_{canvas_id}();\\n \\\"\\\"\\\".format(canvas_id=canvas_id))\\n html_lines.append('</script>')\\n\\n return '\\\\n'.join(html_lines)\",\n \"def update(self, events: list[event.Event]) -> None:\\n if self.buffer.get_width() != self.render_width or self.buffer.get_height() != self.render_height:\\n self.pre_render()\",\n \"def render(self):\\n\\n self.desert_image.render()\\n self.cannon_image.render()\\n self.play_button.render()\\n self.escape_label.render()\",\n \"def on_draw(self, widget, cr):\\n #print \\\"starting to draw\\\"\\n if self.double_buffer is not None:\\n self.draw_tiles()\\n cr.set_source_surface(self.double_buffer, 0.0, 0.0)\\n cr.paint()\\n else:\\n print('Invalid double buffer')\\n #print \\\"done drawing\\\"\\n return False\",\n \"def render(self):\\n self.rendering = True\\n self.env.render()\",\n \"def _prepare_draw(self, view=None):\\n return True\",\n \"def pre_render(self) -> None:\\n self.buffer = Surface((self.render_width, self.render_height), SRCALPHA)\\n self.buffer.fill(list(self.halo_texture.surfaces.values())[0].get_at((0, 0)))\\n\\n self.buffer.fill((0, 0, 0, 0), Rect(\\n (self.render_width - self.halo_texture.get_width()) // 2,\\n (self.render_height - self.halo_texture.get_height()) // 2,\\n self.halo_texture.get_width(),\\n self.halo_texture.get_height()\\n ))\",\n \"def sgnRenderAllFrames(self):\\n\\n self.uiSaveSelectedRenders()\\n\\n self.core.render(self.lstSelectedRenders, bEverything = True)\",\n \"def refresh_canvas(self):\\n self.canvas.delete('all')\\n self.draw_handler(self)\\n self.canvas.after(CANVAS[\\\"REFRESH_TIME\\\"], self.refresh_canvas)\",\n \"def refresh_view(self):\\n if self._step_number % 2 == 0:\\n self._view.draw_enemies(self._game.enemies)\\n self._view.draw_towers(self._game.towers)\\n self._view.draw_obstacles(self._game.obstacles)\",\n \"def setup_draw(self):\\n pass\",\n \"def paint(self):\\r\\n pass\",\n \"def paintGL(self):\\n self._sceneviewer.renderScene()\\n # paintGL end\",\n \"def draw(self):\\n for section in self.sections:\\n canvas_reset(self.canvas)\\n section.draw(self.canvas)\",\n \"def draw(self):\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def render(self):\\n pass\",\n \"def _render(self, gc, points):\\n with gc:\\n gc.set_antialias(True)\\n self._draw_default_axes(gc)\\n self._draw_default_grid(gc)\\n if len(points)>0:\\n gc.clip_to_rect(self.x, self.y, self.width, self.height)\\n gc.set_stroke_color(self.color_)\\n gc.set_line_width(self.line_width)\\n gc.set_line_dash(self.line_style_)\\n\\n gc.begin_path()\\n gc.lines(points)\\n gc.stroke_path()\\n\\n return\",\n \"def render(self):\",\n \"def prepareTestCanvas(self):\\r\\n loadLayers()\\r\\n setCanvasCrs(4326, True)\\r\\n CANVAS.resize(QtCore.QSize(400, 400))\\r\\n CANVAS.zoomToFullExtent()\",\n \"def _draw(self, canvas, options):\\n pass # must override in subclass\",\n \"def blit(self):\\n # self.ax1.draw_artist(self.lines[:2])\\n # self.ax2.draw_artist(self.lines[2:4])\\n self.ax1.autoscale()\\n self.ax2.autoscale()\\n self.ax1.redraw_in_frame()\\n self.ax2.redraw_in_frame()\\n self.fig.canvas.blit(self.fig.bbox)\",\n \"def getImmediateRendering():\\n\\n\\treturn False\",\n \"def on_paint(self, event):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the viewport, modelview and projection matrices\\n self.init_gl()\\n self.init = True\\n\\n size = self.GetClientSize()\\n text = \\\"\\\".join([_(u\\\"Canvas redrawn on paint event, size is \\\"),\\n str(size.width), \\\", \\\", str(size.height)])\\n self.render(text)\",\n \"def render(self, screen):\\n pass\",\n \"def render(self, screen):\\n pass\",\n \"def prepareForDraw(self, obj_path, camera_path, frame_context, old_data):\\n pass\",\n \"def draw(self, *args, **kwargs):\\n self.window.clear()\\n self.batch.draw()\",\n \"def _refresh_render(self):\\n current_frame = self.frame\\n self.frame = int(1E6)\\n self.frame = current_frame\",\n \"def test_visualisations_perform_visualisation_render_data_refresh_synchronously(self):\\n pass\",\n \"def renderall(self):\\n\\n if not self.isinitialized:\\n return\\n # clear display\\n self.screen.fill(BGCOLOR)\\n # draw the board\\n self.drawBoard()\\n # flip the display to show whatever we drew\\n pygame.display.flip()\",\n \"def pysweep_before_finish_init(self):\\n self.displaycanvas = DisplayCanvas(self.pysweep.master, self.boardsize, self.lcounterlength, self.rcounterlength, self.images)\\n self.displaycanvas.pack()\\n\\n self.pysweep.master.update_idletasks()\\n self.displaycanvas.update_idletasks()\\n # enode = self.arbitrary()\\n # print('DisplayCanvas:', enode)\",\n \"def render(self) -> None:\\n if self.native_rendering:\\n self._render()\\n else:\\n self.renderer.render_image(self.get_rendered_image())\",\n \"def render(self):\\r\\n super().render()\",\n \"def draw(self):\\n\\n self.updateLazyImageLoading()\\n\\n image(self.baseMap, 0, 0)\\n\\n for layer in self.layers:\\n layer.draw()\\n\\n for marker in self.markers:\\n marker.draw()\",\n \"def sgnRenderCurrentFrame(self):\\n\\n self.uiSaveSelectedRenders()\\n\\n self.core.render(self.lstSelectedRenders, bCurrentFrame = True)\",\n \"def draw(self):\\n self.batch.draw()\",\n \"def on_draw_event(self, widget, ctx):\\n # the _need_redraw flag doesnt work. it sometimes prevents\\n # the rendering and leaving the canvas blank\\n #if self._need_redraw:\\n self._renderer.set_context(ctx)\\n allocation = self.get_allocation()\\n x, y, w, h = allocation.x, allocation.y, allocation.width, allocation.height\\n self._render_figure(w, h)\\n #self._need_redraw = False\\n\\n return False # finish event propagation?\",\n \"def _drawFrame(self):\\n\\n self._clearScreen()\\n \\n for object in Object.Objects:\\n self._drawObject(object)\\n\\n for entity in Entity.Entities:\\n self._drawObject(entity)\\n\\n self._drawObject(Game.Player)\",\n \"def make(self):\\n\\t\\tif RENDER_VIEWS > 1:\\n\\t\\t\\tself._make()\",\n \"def draw(self):\\n self.write_image()\\n self.update()\",\n \"def StartDraw(self):\\r\\n self.zoom = self.test.viewZoom\\r\\n self.center = self.test.viewCenter\\r\\n self.offset = self.test.viewOffset\\r\\n self.screenSize = self.test.screenSize\",\n \"def canvas_api():\\n pass\",\n \"def draw(self) -> Any:\",\n \"def renderCanvas(self, plot, painter, canvasRect, maps):\\r\\n canvas = plot.canvas()\\r\\n r = canvasRect.adjusted(0., 0., -1., 1.)\\r\\n if self.__data.layoutFlags & self.FrameWithScales:\\r\\n painter.save()\\r\\n r.adjust(-1., -1., 1., 1.)\\r\\n painter.setPen(QPen(Qt.black))\\r\\n if not (self.__data.discardFlags & self.DiscardCanvasBackground):\\r\\n bgBrush = canvas.palette().brush(plot.backgroundRole())\\r\\n painter.setBrush(bgBrush)\\r\\n painter.drawRect(r)\\r\\n painter.restore()\\r\\n painter.save()\\r\\n painter.setClipRect(canvasRect)\\r\\n plot.drawItems(painter, canvasRect, maps)\\r\\n painter.restore()\\r\\n elif canvas.testAttribute(Qt.WA_StyledBackground):\\r\\n clipPath = QPainterPath()\\r\\n painter.save()\\r\\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\\r\\n QwtPainter.drawBackground(painter, r, canvas)\\r\\n clipPath = qwtCanvasClip(canvas, canvasRect)\\r\\n painter.restore()\\r\\n painter.save()\\r\\n if clipPath.isEmpty():\\r\\n painter.setClipRect(canvasRect)\\r\\n else:\\r\\n painter.setClipPath(clipPath)\\r\\n plot.drawItems(painter, canvasRect, maps)\\r\\n painter.restore()\\r\\n else:\\r\\n clipPath = QPainterPath()\\r\\n frameWidth = 0\\r\\n if not self.__data.discardFlags & self.DiscardCanvasFrame:\\r\\n frameWidth = canvas.frameWidth()\\r\\n clipPath = qwtCanvasClip(canvas, canvasRect)\\r\\n innerRect = canvasRect.adjusted(frameWidth, frameWidth,\\r\\n -frameWidth, -frameWidth)\\r\\n painter.save()\\r\\n if clipPath.isEmpty():\\r\\n painter.setClipRect(innerRect)\\r\\n else:\\r\\n painter.setClipPath(clipPath)\\r\\n if not self.__data.discardFlags & self.DiscardCanvasBackground:\\r\\n QwtPainter.drawBackground(painter, innerRect, canvas)\\r\\n plot.drawItems(painter, innerRect, maps)\\r\\n painter.restore()\\r\\n if frameWidth > 0:\\r\\n painter.save()\\r\\n frameStyle = canvas.frameShadow() | canvas.frameShape()\\r\\n frameWidth = canvas.frameWidth()\\r\\n borderRadius = canvas.borderRadius()\\r\\n if borderRadius > 0.:\\r\\n QwtPainter.drawRoundedFrame(painter, canvasRect, r, r,\\r\\n canvas.palette(), frameWidth,\\r\\n frameStyle)\\r\\n else:\\r\\n midLineWidth = canvas.midLineWidth()\\r\\n QwtPainter.drawFrame(painter, canvasRect, canvas.palette(),\\r\\n canvas.foregroundRole(), frameWidth,\\r\\n midLineWidth, frameStyle)\\r\\n painter.restore()\",\n \"def reindex_graphics(self):\\n for obj in self.context.static_objects:\\n self.canvas.children.remove(obj.widget.canvas)\\n # fill _objects_z_index\\n _objects_z_index = {}\\n for obj in self.context.static_objects:\\n y = obj.widget.pos[1]\\n if not y in _objects_z_index:\\n _objects_z_index[y] = []\\n _objects_z_index[y].append(obj)\\n _keys = _objects_z_index.keys()\\n _keys.sort()\\n _keys.reverse()\\n for k in _keys:\\n objs = _objects_z_index[k]\\n for obj in objs:\\n self.canvas.add(obj.widget.canvas)\",\n \"def draw(self, canvas):\\n super().draw(canvas, self.__path)\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def draw(self):\\n raise NotImplementedError\",\n \"def on_paint(self, event):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the OpenGL rendering context\\n self.init_gl()\\n self.init = True\\n\\n size = self.GetClientSize()\\n text = \\\"\\\".join([_(u\\\"Canvas redrawn on paint event, size is \\\"),\\n str(size.width), \\\", \\\", str(size.height)])\\n self.render()\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def draw(self): \\n [component.draw() for component in self.components]\",\n \"def render(self):\\n\\n # self.screen.blit(self.background, (0, 0))\\n\\n \\\"\\\"\\\"work with ever game state\\\"\\\"\\\"\\n if self.state == FIRST_ENTER:\\n \\\"\\\"\\\"welcome screen))\\\"\\\"\\\"\\n text = self.font.render(\\\"PETERSBURG VAMPIRES\\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 3))\\n text = self.font.render(\\\"press ENTER to start\\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_WIDTH / 4, SCREEN_HEIGHT / 2))\\n elif self.state == GAME:\\n \\\"\\\"\\\"rendering enemies and player\\\"\\\"\\\"\\n old_time = time.time()\\n self.levels[self.curient_level].render(self.screen)\\n # self.levels[self.curient_level].render(self.screen)\\n for obj in self.levels[self.curient_level].objects:\\n \\\"\\\"\\\"debug mode\\\"\\\"\\\"\\n # text = self.debug_font.render(\\\"koor:\\\" + str(obj.position_np), True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]))\\n\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]))\\n #\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+obj.image_size[Y]))\\n #\\n # text = self.debug_font.render(\\\"x\\\", True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X]+obj.image_size[X], obj.position_np[Y]+obj.image_size[Y]))\\n #\\n # x, y = get_grid_xy(obj.position_np, ZOMBIE_SIZE)\\n # text = self.debug_font.render(\\\"x_grid: \\\" + str(x) + \\\",y_grid: \\\" + str(y), True, (255, 0, 0))\\n # self.screen.blit(text, (obj.position_np[X], obj.position_np[Y]+20))\\n\\n obj.render(self.screen)\\n \\\"\\\"\\\"rendering HP status\\\"\\\"\\\"\\n text = self.font.render(\\\"HP: \\\" + str(self.player.HP), True, (255, 0, 0))\\n self.screen.blit(text, (0, 10))\\n new_time = time.time()\\n self.render_time.append(new_time - old_time)\\n elif self.state == END:\\n \\\"\\\"\\\"The end screen\\\"\\\"\\\"\\n self.screen.blit(pygame.image.load(END_BUTTON_PATH), (SCREEN_WIDTH / 5, SCREEN_HEIGHT / 8))\\n elif self.state == WIN:\\n text = self.font.render(\\\"YOU WIN, REPEAT? \\\", True, (255, 0, 0))\\n self.screen.blit(text, (SCREEN_HEIGHT / 2, SCREEN_WIDTH / 2))\\n\\n pygame.display.flip()\",\n \"def render(self):\\n gl.glEnableClientState(gl.GL_VERTEX_ARRAY);\\n gl.glEnableClientState(gl.GL_NORMAL_ARRAY);\\n gl.glVertexPointer(3, gl.GL_DOUBLE, 0, self._vertices);\\n gl.glNormalPointer(gl.GL_DOUBLE, 0, self._normals);\\n for patch in self._patches:\\n patch.render()\\n gl.glDisableClientState(gl.GL_VERTEX_ARRAY);\\n gl.glDisableClientState(gl.GL_NORMAL_ARRAY);\",\n \"def draw(self):\\n for obj in self.objects:\\n obj.draw()\",\n \"def _before_render_event(self, *args, **kwargs):\\n for chart in self._charts:\\n chart._render_event(*args, **kwargs)\",\n \"def render(self):\\n self.screen.fill(prepare.BACKGROUND_COLOR)\\n self.health_bar()\\n # self.enemy_health()\\n self.energy_bar()\\n self.level.draw(self.screen)\\n pg.display.update()\",\n \"def get(self):\\n self.render_front()\",\n \"def render(self):\\n self.SetCurrent(self.context)\\n if not self.init:\\n # Configure the OpenGL rendering context\\n self.init_gl()\\n self.init = True\\n\\n # Clear everything\\n GL.glClear(GL.GL_COLOR_BUFFER_BIT | GL.GL_DEPTH_BUFFER_BIT)\\n\\n for device_id, output_id in self.monitors.monitors_dictionary:\\n i = list(\\n self.monitors.monitors_dictionary.keys()).index(\\n (device_id, output_id))\\n GL.glColor3f(self.colours[i%8][0], self.colours[i%8][1], self.colours[i%8][2])\\n x = (i - len(self.monitors.monitors_dictionary)/2)*20\\n signal_list = self.monitors.monitors_dictionary[(\\n device_id, output_id)]\\n\\n signal_name = self.devices.get_signal_name(device_id, output_id)\\n\\n self.render_text(signal_name, x, 0, 10*len(signal_list)+10)\\n\\n length = len(signal_list)\\n for j in range(length):\\n signal = signal_list[j]\\n z = (j-length/2)*20\\n if signal == self.devices.HIGH:\\n y = 11\\n if signal == self.devices.LOW:\\n y = 1\\n if signal != self.devices.BLANK:\\n self.draw_cuboid(x, z, 5, 10, y)\\n\\n # We have been drawing to the back buffer, flush the graphics pipeline\\n # and swap the back buffer to the front\\n GL.glFlush()\\n self.SwapBuffers()\",\n \"def display(self):\\n\\t\\tfor c in self.canvas.values():\\n\\t\\t\\tc.update()\\n\\t\\t\\tself.superficie.blit(c.superficie, c.origen)\\n\\t\\tpygame.display.flip()\",\n \"def finish_render():\\n get_window().static_display = True\\n get_window().flip_count = 0\\n get_window().flip()\",\n \"def render(self, mode='human'):\\n if self.microgridRender == None : self.microgridRender = MicrogridRender(self.microgrid, self.clock);\\t\\n self.microgridRender.render(mode);\",\n \"def _pre_draw_bge(self):\\r\\n self._pre_draw_common()\\r\\n # draw rays\\r\\n self._drawRays()\",\n \"def on_draw(self):\\n self.clear()\\n self.manager.draw()\",\n \"def render(self, chart):\\n chart.create_visualization_files(self.__outputpath)\",\n \"def draw(self, screen):\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.628021","0.628021","0.62287074","0.6217737","0.6203357","0.60232645","0.5878731","0.58328944","0.58255154","0.5802863","0.5789068","0.57696724","0.5759943","0.5747099","0.5704188","0.56780964","0.56780964","0.56780964","0.56780964","0.5672039","0.5671816","0.56596327","0.5655339","0.56275475","0.5626634","0.56148493","0.5612647","0.5596928","0.55931777","0.55920345","0.5588081","0.55644345","0.55371624","0.5531814","0.55212396","0.5502409","0.550047","0.54949445","0.54885125","0.54885125","0.54885125","0.54885125","0.54885125","0.54885125","0.5466787","0.5466383","0.5464405","0.5458059","0.5451571","0.54461306","0.5428607","0.5422306","0.5422306","0.5410208","0.54071915","0.54066515","0.54018235","0.53987134","0.53901416","0.53865117","0.5386171","0.5383457","0.53463894","0.5342882","0.53352463","0.532959","0.53244734","0.53242147","0.53210354","0.53078264","0.53065974","0.53062123","0.5302961","0.5302713","0.52969915","0.52969915","0.52969915","0.5294216","0.52879965","0.52879965","0.52879965","0.52879965","0.52879965","0.52879965","0.52879965","0.52879965","0.52855414","0.5282988","0.5281306","0.52798694","0.52775496","0.5269974","0.52686334","0.52604836","0.5259117","0.5258392","0.52558696","0.5253976","0.5253753","0.52396464"],"string":"[\n \"0.628021\",\n \"0.628021\",\n \"0.62287074\",\n \"0.6217737\",\n \"0.6203357\",\n \"0.60232645\",\n \"0.5878731\",\n \"0.58328944\",\n \"0.58255154\",\n \"0.5802863\",\n \"0.5789068\",\n \"0.57696724\",\n \"0.5759943\",\n \"0.5747099\",\n \"0.5704188\",\n \"0.56780964\",\n \"0.56780964\",\n \"0.56780964\",\n \"0.56780964\",\n \"0.5672039\",\n \"0.5671816\",\n \"0.56596327\",\n \"0.5655339\",\n \"0.56275475\",\n \"0.5626634\",\n \"0.56148493\",\n \"0.5612647\",\n \"0.5596928\",\n \"0.55931777\",\n \"0.55920345\",\n \"0.5588081\",\n \"0.55644345\",\n \"0.55371624\",\n \"0.5531814\",\n \"0.55212396\",\n \"0.5502409\",\n \"0.550047\",\n \"0.54949445\",\n \"0.54885125\",\n \"0.54885125\",\n \"0.54885125\",\n \"0.54885125\",\n \"0.54885125\",\n \"0.54885125\",\n \"0.5466787\",\n \"0.5466383\",\n \"0.5464405\",\n \"0.5458059\",\n \"0.5451571\",\n \"0.54461306\",\n \"0.5428607\",\n \"0.5422306\",\n \"0.5422306\",\n \"0.5410208\",\n \"0.54071915\",\n \"0.54066515\",\n \"0.54018235\",\n \"0.53987134\",\n \"0.53901416\",\n \"0.53865117\",\n \"0.5386171\",\n \"0.5383457\",\n \"0.53463894\",\n \"0.5342882\",\n \"0.53352463\",\n \"0.532959\",\n \"0.53244734\",\n \"0.53242147\",\n \"0.53210354\",\n \"0.53078264\",\n \"0.53065974\",\n \"0.53062123\",\n \"0.5302961\",\n \"0.5302713\",\n \"0.52969915\",\n \"0.52969915\",\n \"0.52969915\",\n \"0.5294216\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52879965\",\n \"0.52855414\",\n \"0.5282988\",\n \"0.5281306\",\n \"0.52798694\",\n \"0.52775496\",\n \"0.5269974\",\n \"0.52686334\",\n \"0.52604836\",\n \"0.5259117\",\n \"0.5258392\",\n \"0.52558696\",\n \"0.5253976\",\n \"0.5253753\",\n \"0.52396464\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.5282165"},"document_rank":{"kind":"string","value":"88"}}},{"rowIdx":90872,"cells":{"query":{"kind":"string","value":"Get property value from object."},"document":{"kind":"string","value":"def get(self, obj):\n raise NotImplementedError"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def get_value_from_object(obj, key):\n if is_dict(obj):\n return obj.get(key)\n return getattr(obj, key, None)","def get(self, prop):\r\n prop_parts = prop.split(\".\")\r\n val = None\r\n for part in prop_parts:\r\n if val is None:\r\n val = self.obj.get(part)\r\n else:\r\n val = val.get(part)\r\n return val","def get_value(obj, name):\n if isinstance(obj, dict):\n return obj.get(name)\n\n return getattr(obj, name, obj)","def __call__(self, arg):\n return self.get_property(arg)","def get_property_value(self, property, db):\n try:\n for p in self.properties:\n if p.idProperty == int(property):\n return p.get_value()\n except:\n return None","def do_get_property(self, spec):\n attribute = self.find_attribute(spec.name)\n if attribute is not None and isinstance(attribute, property):\n return attribute.fget(self)\n else:\n raise ValueError(\"No such property\", spec.name)","def get_property(self, property):\n return self.shell([\"getprop\", property])","def get_property(self,obj_property):\n return self.redis_server.hget(self.frbr_key,obj_property)","def get_value (property):\n assert is_iterable_typed(property, basestring) or isinstance(property, basestring)\n return replace_grist (property, '')","def get_property_value(self, property, db):\n if property == 'resultTime':\n return self.get_time()\n elif property == 'wallTime':\n return self.wallTime\n elif property == 'cost':\n return self.cost\n else:\n try:\n for pv in self.properties:\n if pv.idProperty == int(property):\n return pv.get_value()\n except:\n return None","def get_property(self, key):\n return self.properties.get(key)","def get_property(self, name):\n if (not name in self.properties):\n raise KeyError(\"Key '\" + name + \"' not found\")\n return self.properties[name]","def get_property(self,name):\n return self.dp.get_property(name)","def get_value(self):\n attr = getattr(self.obj, self.attr)\n try:\n args = self.args\n kwargs = self.kwargs\n return attr(*args, **kwargs)\n except AttributeError:\n return attr","def _get(self):\n var = self.variables.get(prop_name)\n if var is None:\n result = default() if callable(default) else default\n if objtype:\n self.variables[prop_name] = result\n return result\n if not isinstance(var, prop_type):\n raise ValueError(\"unexpected property type\")\n if objtype:\n return cast(T, var)\n return cast(T, var.value)","def getProperty(self, identifier):\n if identifier in self.properties.keys():\n return self.properties[identifier]\n else:\n return None","def get_raw_value ( self, object ):\n try:\n return xgetattr( self.get_object( object ), self.name )\n except:\n return None","def get_value(o, object_name, value_name, value_type):\n try:\n return o[value_name]\n except KeyError:\n l.warning(\"The field {} doesn't exists in {}: {}\".format(value_name, object_name, ujson.dumps(o)))\n return value_type()","def getProperty(propname):","def _get_value(o):\n return value(o, exception=False)","def GetProperty(self, propertyname, arg = None):\n if self.serviceimplementation == 'basic':\n # Conventionally properties starting with X (and only them) may return a UNO object\n calltype = self.vbGet + (self.flgUno if propertyname[0] == 'X' else 0)\n if arg is None:\n return self.EXEC(self.objectreference, calltype, propertyname)\n else: # There are a few cases (Calc ...) where GetProperty accepts an argument\n return self.EXEC(self.objectreference, calltype, propertyname, arg)\n return None","def _getPropertyValue(self, name, target = ''):\n value = self._getPropertyValue(name)\n return value.getData() if value else None","def getValue(self, name):\n\n return getattr(self, name)","def getp(self,p):\n return property_dict[p]","def getValue(self, *args):\n return _libsbml.ConversionProperties_getValue(self, *args)","def get_property_value(prop, paths):\n\n data = parse_config(paths)\n return data.get(prop)","def prop(self):\n return getattr(self, name)","def get_dynamic_property(vim, mobj, type, property_name):\n properties = get_dynamic_properties(vim, mobj, [property_name], type)\n property_value = None\n if property_name in properties:\n property_value = properties.get(property_name)\n return property_value","def get_value(self, obj, attr, accessor=None, default=missing_):\n accessor_func = accessor or utils.get_value\n check_key = attr if self.attribute is None else self.attribute\n return accessor_func(obj, check_key, default)","def GetProp(self, name):\n return self._props.get(name)","def _get_value(self):\n return self.__value","def __getattr__(self, name):\r\n\t\treturn self.properties[name]","def __getattr__(self, name):\r\n\t\treturn self.properties[name]","def get_property_value(self, reply):\n self.logger.debug(\"Attempting to parse %s's value\", reply)\n if isinstance(reply, GetPropertyReply):\n if 8 == reply.format:\n if hasattr(reply.value, 'buf'):\n value = str(reply.value.buf())\n else:\n value = ''\n for chunk in reply.value:\n value += chr(chunk)\n self.logger.silly(\"Parsed %s\", value)\n return value\n elif reply.format in (16, 32):\n value = list(\n unpack(\n 'I' * reply.value_len,\n reply.value.buf()\n )\n )\n self.logger.silly(\"Parsed %s\", value)\n return value\n self.logger.warning(\"The reply might not be valid: %s\", reply)\n return None","def get_property_value(self):\n if live_object_is_valid(self._parent):\n return int(getattr(self._parent, self._property_name) / self.marker_factor)\n else:\n return","def get_val(self, **kwargs):\n return self._value","def get_item(obj, key):\n val = None\n if obj and type(obj) == dict:\n val = obj.get(key)\n elif obj and hasattr(obj, key):\n val = getattr(obj, key)\n val = val or ''\n return val","def get_val(self):\n if self.val_obj is None:\n return None\n else:\n return self.val_obj.val","def get_value ( self, object ):\n return unicode( object[ self.index ] )","def get_value ( self, object ):\n try:\n value = getattr( object, self.name )\n try:\n return self.format % ( value, )\n except:\n return 'Format!'\n except:\n return 'Undefined!'","def __get__(self, obj, objtype=None):\n print('Hello from __get__')\n # print(self, obj, objtype)\n # print(type(self), type(obj), type(objtype))\n return self.val","def _get_value(self):\n \n return self._value","def getProperty(*args):","def getProperty(*args):","def getProperty(*args):","def getProperty(*args):","def getProperty(*args):","def do_get_property(self, pspec):\n\n if pspec.name == 'joined':\n return self._joined\n\n if self._get_properties_call is not None:\n print('%r: Blocking on GetProperties() because someone '\n 'wants property %s', self, pspec.name)\n self._get_properties_call.block()\n\n if pspec.name == 'id':\n return self._id\n elif pspec.name == 'name':\n return self._name\n elif pspec.name == 'color':\n return self._color\n elif pspec.name == 'type':\n return self._type\n elif pspec.name == 'tags':\n return self._tags\n elif pspec.name == 'private':\n return self._private","def getValuesFor(self, obj):\n val = valueInContext(self.value, obj)\n assert isinstance(val, dict)\n return val","def __get__(self, obj, cls):\n if obj is None:\n return self\n\n value = self.fget(obj)\n setattr(obj, self.fget.__name__, value)\n return value","def getProperty(self, propertyName: unicode) -> unicode:\n ...","def _GetPropertyValue(entity, property):\n if property in datastore_types._SPECIAL_PROPERTIES:\n assert property == datastore_types._KEY_SPECIAL_PROPERTY\n return entity.key()\n else:\n return entity[property]","def _get(obj, name):\n try:\n # try to get value using dict's __getitem__ descriptor first\n return dict.__getitem__(obj, name)\n except TypeError:\n # if it's a dict, then preserve the TypeError\n if isinstance(obj, dict):\n raise\n # otherwise try one last time, relying on __getitem__ if any\n return obj[name]","def get_dynamic_property(vim, mobj, type, property_name):\r\n obj_content = \\\r\n get_object_properties(vim, None, mobj, type, [property_name])\r\n property_value = None\r\n if obj_content:\r\n dynamic_property = obj_content[0].propSet\r\n if dynamic_property:\r\n property_value = dynamic_property[0].val\r\n return property_value","def getProperty(unique_name):","def get(self, item):\n try:\n data = copy.deepcopy(getattr(self, item))\n except AttributeError:\n raise M2Error('Error while trying to get non-existent property `%s`' % item, False)\n except SQLAlchemyError:\n self.s.rollback()\n raise\n return data","def value_of_css_property(self, name):\n return self.element.value_of_css_property(name)","def get_value(v):\n # try to get attribute from user\n obj = getattr(user, v, None)\n # if no, search for corresponding method in MailgunBatchMessage\n if not obj:\n obj = getattr(self, 'get_var_' + v, None)\n if obj:\n obj = functools.partial(obj, user)\n\n # is there is no function to get var value raise or return empty string\n else:\n if strict:\n raise ValueError('There is no way to get `{0}` from user.'\n 'user.{0} and MailgunBatchMessage.get_var_{0} tried'.format(v))\n return ''\n\n return obj() if callable(obj) else obj","def get(self):\n return self._value","def get_value(self):\n return self._value","def getValue(self, name: unicode) -> object:\n ...","def get_value(self, reference_path):\n\n if (reference_path.start_protocol is not None and\n reference_path.start_protocol != self.id):\n\n raise ValueError('The reference path does not target this protocol.')\n\n if reference_path.property_name is None or reference_path.property_name == '':\n raise ValueError('The reference path does specify a property to return.')\n\n return get_nested_attribute(self, reference_path.property_name)","def property(self) -> Optional[pulumi.Input[str]]:\n return pulumi.get(self, \"property\")","def get_property_value(self, name):\n # Return table field as property\n tableName, primKey = self.provider._split_path(self.path)\n if primKey is not None:\n ns, localName = util.split_namespace(name)\n if ns == (tableName + \":\"):\n conn = self.provider._init_connection()\n fieldlist = self.provider._get_field_list(conn, tableName)\n if localName in fieldlist:\n val = self.provider._get_field_by_primary_key(\n conn, tableName, primKey, localName\n )\n conn.close()\n return val\n conn.close()\n # else, let default implementation return supported live and dead properties\n return super().get_property_value(name)","def get(key, nodename=None):\n return _get_property(key, nodename, None)","def _value_for(self, object, name, value):\n try:\n return self.value_for(value)\n except TraitError:\n self.error(object, name, value)","def _value_for(self, object, name, value):\n try:\n return self.value_for(value)\n except TraitError:\n self.error(object, name, value)","def _val(obj):\n if isinstance(obj, ctypes._SimpleCData):\n return obj.value\n else:\n return obj","def callPropertyGet(self, name = \"__value\", index = None):\n\t\tif name == 'IID':\n\t\t\treturn CSLValue(typeid = \"string\", value = self.callerInfo.IID)\n\t\tEntry = self.vtbl['p_' + name + \"_get\"]\n\t\t#localTbl = { 'vars':{}, 'status':0, 'props':{}, 'alias':{}, 'persistent':{}, 'instance':{}}\n\t\tpropEntry = Entry.parent\n\t\tprmName = propEntry.data['prm']\n\t\tlocalTbl = self.CSLCreateLocalTbl({}, {}, {}, copy.copy(propEntry.data['persistent']), copy.copy(propEntry.data['instance']))\n\t\tif prmName != \"\":\n\t\t\tif index == None:\n\t\t\t\tdefault = propEntry.data['default']\n\t\t\t\tif default != \"\":\n\t\t\t\t\tdefault = self.CSLCheckValue(default, localTbl)\n\t\t\t\telse:\n\t\t\t\t\tdefault = CSLValue(typeid = \"NULL\", value = None)\n\t\t\telse:\n\t\t\t\tdefault = index\n\n\t\t\tlocalTbl['vars'][prmName] = default\n\n\t\tself.procStack.append('p_' + name + '_get')\n\t\tself.lastLTbl = self.CSLInterpreter(Entry.child, localTbl)\n\t\tself.procStack.pop()\n\t\tl = self.lastLTbl['vars']\n\n\t\tself.debug(DEBUG_CALL, \"\\n\\nGetProp result: (\", name, \")\", l, \"haskey:\", l.has_key(name))\n\n\t\tif l != None and l.has_key(name):\n\t\t\tself.debug(DEBUG_CALL, \"Get Property return:\", l[name])\n\t\t\treturn copy.deepcopy(l[name])\n\t\telse:\n\t\t\treturn CSLValue(typeid = \"NULL\", value = None)","def get_Value(self):\n obj_p=object.__getattribute__(self, 'StoredValue')\n if isinstance(obj_p, type(True)):\n return obj_p & self.statusOn \n else: \n if self.statusOn: \n return obj_p \n else: \n self._log.warning('You are attempting to access the job'+ \n 'property %s which has not been set', self._context_name) \n return None","def getattribute(objeto, name: str):\r\n # Get internal dict value matching name.\r\n value = objeto.__dict__.get(name)\r\n if not value:\r\n # Raise AttributeError if attribute value not found.\r\n return None\r\n # Return attribute value.\r\n return value","def propertyValueFromNodeGetter(instance):\n\n subspace = nodeGetter(instance)\n\n if SynchronousPropertyAccess.SynchronousPropertyAccess.getCurrent() is not None:\n if not subspace.loaded:\n subspace.keyspace.waitLoaded()\n else:\n subspace.keyspace.ensureSubscribed()\n\n val = subspace.value\n\n if val is None:\n return default()\n\n return val[0]","def get(self, key):\n\n return self.dict_props.get(key)","def getPropertie(self, propname):\n if propname in self._getPropName():\n return self.properties[propname]\n else:\n raise Exception(\"la propiedad no existe\")","def get_val_obj(self):\n return self.val_obj","def _column_value_getter(self, obj):\n if is_namedtuple(obj):\n # Get values using properties\n def get(obj, column):\n return getattr(obj, column, None)\n\n elif is_dict_like(obj):\n # Get values using dictionary keys\n def get(obj, column):\n return obj.get(column)\n\n elif is_list_like(obj):\n # Get values using list indexes\n def get(obj, column):\n col = self.column_location(column)\n try:\n return obj[col]\n except IndexError:\n return None\n\n else:\n # Fallback to single column\n def get(obj, _):\n return obj\n\n return get","def getProperty(self, propID, time, objectID=0):\n key = (propID, objectID, time)\n if propID not in self.properties.index:\n\n raise APIError.APIError('Unknown property ID')\n\n if time not in self.properties.index:\n prop = interP.interpolateProperty(self.properties,\n time=time,\n propertyID=propID,\n objectID=objectID,\n method='linear')\n else:\n prop = self.properties[key]\n\n # Check pyro registering if applicaple\n if hasattr(self, '_pyroDaemon') and not hasattr(prop, '_PyroURI'):\n uri = self._pyroDaemon.register(prop)\n prop._PyroURI = uri\n\n return(prop)","def getprop(self, prop_name):\n return self.shell(\"getprop %s\" % prop_name)","def get_variable(self, name):\n return self._properties[name]","def Get(self, interface_name, property_name):\n logger.debug(\n \"%r.Get(%r, %r) -> ...\",\n self, interface_name, property_name)\n try:\n props = self._dct_entry[interface_name]\n except KeyError:\n raise dbus.exceptions.DBusException(\n dbus.PROPERTIES_IFACE,\n \"No such interface {}\".format(interface_name))\n try:\n prop = props[property_name]\n except KeyError:\n raise dbus.exceptions.DBusException(\n dbus.PROPERTIES_IFACE,\n \"No such property {}:{}\".format(\n interface_name, property_name))\n try:\n value = prop.__get__(self, self.__class__)\n except dbus.exceptions.DBusException as exc:\n logger.error(\n \"%r.Get(%r, %r) -> (exception) %r\",\n self, interface_name, property_name, exc)\n raise\n except Exception as exc:\n logger.exception(\n \"runaway exception from Get(%r, %r)\",\n interface_name, property_name)\n raise dbus.exceptions.DBusException(\n dbus.PROPERTIES_IFACE,\n \"Unable to get property interface/property {}:{}: {!r}\".format(\n interface_name, property_name, exc))\n else:\n logger.debug(\n \"%r.Get(%r, %r) -> %r\",\n self, interface_name, property_name, value)\n return value","def get_value(key, obj, default=missing):\n if isinstance(key, int):\n return _get_value_for_key(key, obj, default)\n return _get_value_for_keys(key.split('.'), obj, default)","def get_property(self, client):\r\n client.getProperty()","def get_value(self, key):\n return self[key]","def get_device_property(self, client, prop):\r\n value = client.getDeviceProperty(prop)\r\n return value","def read_value(self):\n return self.load_attr(\"value\")","def __get__(self,obj,objtype):\n result = super(Number,self).__get__(obj,objtype)\n # CEBALERT: results in extra lookups (_value_is_dynamic() is\n # also looking up 'result' - should just pass it in). Note\n # that this method is called often.\n if self._value_is_dynamic(obj,objtype): self._check_value(result)\n return result","def get(self, obj):\n obj = self._to_obj_tuple(obj)\n rows = self.query(object=obj)\n if rows:\n return rows[0]","def GetValue(self):\n return self._value","def GetValue(self):\n return self._value","def _get(self, name):\n return object.__getattribute__(self, name)","def _get(self, name):\n return object.__getattribute__(self, name)","def getValue(self):\n \n return self._value","def get_value(self):\n return self.value","def get_value(self):\n return self.value","def get_value(self):\n return self.value","def get_value(self):\n return self._value","def get_value(self):\n return self._value","def get_value(self):\n return self._value","def __getattr__ (self, name):\n\t\ttry:\n\t\t\treturn self.__dict__[name]\n\t\texcept KeyError:\n\t\t\treturn self.__dict__[\"value\"][name]","def GetWirelessProperty(self, networkid, prop):\n try:\n value = self.LastScan[networkid].get(prop)\n except IndexError:\n if self.debug_mode:\n print \"GetWirelessProperty: Index error occured trying to \" + \\\n \"retrieve property %s\" % prop\n value = \"\"\n try:\n value = misc.to_unicode(value)\n except:\n pass\n return value"],"string":"[\n \"def get_value_from_object(obj, key):\\n if is_dict(obj):\\n return obj.get(key)\\n return getattr(obj, key, None)\",\n \"def get(self, prop):\\r\\n prop_parts = prop.split(\\\".\\\")\\r\\n val = None\\r\\n for part in prop_parts:\\r\\n if val is None:\\r\\n val = self.obj.get(part)\\r\\n else:\\r\\n val = val.get(part)\\r\\n return val\",\n \"def get_value(obj, name):\\n if isinstance(obj, dict):\\n return obj.get(name)\\n\\n return getattr(obj, name, obj)\",\n \"def __call__(self, arg):\\n return self.get_property(arg)\",\n \"def get_property_value(self, property, db):\\n try:\\n for p in self.properties:\\n if p.idProperty == int(property):\\n return p.get_value()\\n except:\\n return None\",\n \"def do_get_property(self, spec):\\n attribute = self.find_attribute(spec.name)\\n if attribute is not None and isinstance(attribute, property):\\n return attribute.fget(self)\\n else:\\n raise ValueError(\\\"No such property\\\", spec.name)\",\n \"def get_property(self, property):\\n return self.shell([\\\"getprop\\\", property])\",\n \"def get_property(self,obj_property):\\n return self.redis_server.hget(self.frbr_key,obj_property)\",\n \"def get_value (property):\\n assert is_iterable_typed(property, basestring) or isinstance(property, basestring)\\n return replace_grist (property, '')\",\n \"def get_property_value(self, property, db):\\n if property == 'resultTime':\\n return self.get_time()\\n elif property == 'wallTime':\\n return self.wallTime\\n elif property == 'cost':\\n return self.cost\\n else:\\n try:\\n for pv in self.properties:\\n if pv.idProperty == int(property):\\n return pv.get_value()\\n except:\\n return None\",\n \"def get_property(self, key):\\n return self.properties.get(key)\",\n \"def get_property(self, name):\\n if (not name in self.properties):\\n raise KeyError(\\\"Key '\\\" + name + \\\"' not found\\\")\\n return self.properties[name]\",\n \"def get_property(self,name):\\n return self.dp.get_property(name)\",\n \"def get_value(self):\\n attr = getattr(self.obj, self.attr)\\n try:\\n args = self.args\\n kwargs = self.kwargs\\n return attr(*args, **kwargs)\\n except AttributeError:\\n return attr\",\n \"def _get(self):\\n var = self.variables.get(prop_name)\\n if var is None:\\n result = default() if callable(default) else default\\n if objtype:\\n self.variables[prop_name] = result\\n return result\\n if not isinstance(var, prop_type):\\n raise ValueError(\\\"unexpected property type\\\")\\n if objtype:\\n return cast(T, var)\\n return cast(T, var.value)\",\n \"def getProperty(self, identifier):\\n if identifier in self.properties.keys():\\n return self.properties[identifier]\\n else:\\n return None\",\n \"def get_raw_value ( self, object ):\\n try:\\n return xgetattr( self.get_object( object ), self.name )\\n except:\\n return None\",\n \"def get_value(o, object_name, value_name, value_type):\\n try:\\n return o[value_name]\\n except KeyError:\\n l.warning(\\\"The field {} doesn't exists in {}: {}\\\".format(value_name, object_name, ujson.dumps(o)))\\n return value_type()\",\n \"def getProperty(propname):\",\n \"def _get_value(o):\\n return value(o, exception=False)\",\n \"def GetProperty(self, propertyname, arg = None):\\n if self.serviceimplementation == 'basic':\\n # Conventionally properties starting with X (and only them) may return a UNO object\\n calltype = self.vbGet + (self.flgUno if propertyname[0] == 'X' else 0)\\n if arg is None:\\n return self.EXEC(self.objectreference, calltype, propertyname)\\n else: # There are a few cases (Calc ...) where GetProperty accepts an argument\\n return self.EXEC(self.objectreference, calltype, propertyname, arg)\\n return None\",\n \"def _getPropertyValue(self, name, target = ''):\\n value = self._getPropertyValue(name)\\n return value.getData() if value else None\",\n \"def getValue(self, name):\\n\\n return getattr(self, name)\",\n \"def getp(self,p):\\n return property_dict[p]\",\n \"def getValue(self, *args):\\n return _libsbml.ConversionProperties_getValue(self, *args)\",\n \"def get_property_value(prop, paths):\\n\\n data = parse_config(paths)\\n return data.get(prop)\",\n \"def prop(self):\\n return getattr(self, name)\",\n \"def get_dynamic_property(vim, mobj, type, property_name):\\n properties = get_dynamic_properties(vim, mobj, [property_name], type)\\n property_value = None\\n if property_name in properties:\\n property_value = properties.get(property_name)\\n return property_value\",\n \"def get_value(self, obj, attr, accessor=None, default=missing_):\\n accessor_func = accessor or utils.get_value\\n check_key = attr if self.attribute is None else self.attribute\\n return accessor_func(obj, check_key, default)\",\n \"def GetProp(self, name):\\n return self._props.get(name)\",\n \"def _get_value(self):\\n return self.__value\",\n \"def __getattr__(self, name):\\r\\n\\t\\treturn self.properties[name]\",\n \"def __getattr__(self, name):\\r\\n\\t\\treturn self.properties[name]\",\n \"def get_property_value(self, reply):\\n self.logger.debug(\\\"Attempting to parse %s's value\\\", reply)\\n if isinstance(reply, GetPropertyReply):\\n if 8 == reply.format:\\n if hasattr(reply.value, 'buf'):\\n value = str(reply.value.buf())\\n else:\\n value = ''\\n for chunk in reply.value:\\n value += chr(chunk)\\n self.logger.silly(\\\"Parsed %s\\\", value)\\n return value\\n elif reply.format in (16, 32):\\n value = list(\\n unpack(\\n 'I' * reply.value_len,\\n reply.value.buf()\\n )\\n )\\n self.logger.silly(\\\"Parsed %s\\\", value)\\n return value\\n self.logger.warning(\\\"The reply might not be valid: %s\\\", reply)\\n return None\",\n \"def get_property_value(self):\\n if live_object_is_valid(self._parent):\\n return int(getattr(self._parent, self._property_name) / self.marker_factor)\\n else:\\n return\",\n \"def get_val(self, **kwargs):\\n return self._value\",\n \"def get_item(obj, key):\\n val = None\\n if obj and type(obj) == dict:\\n val = obj.get(key)\\n elif obj and hasattr(obj, key):\\n val = getattr(obj, key)\\n val = val or ''\\n return val\",\n \"def get_val(self):\\n if self.val_obj is None:\\n return None\\n else:\\n return self.val_obj.val\",\n \"def get_value ( self, object ):\\n return unicode( object[ self.index ] )\",\n \"def get_value ( self, object ):\\n try:\\n value = getattr( object, self.name )\\n try:\\n return self.format % ( value, )\\n except:\\n return 'Format!'\\n except:\\n return 'Undefined!'\",\n \"def __get__(self, obj, objtype=None):\\n print('Hello from __get__')\\n # print(self, obj, objtype)\\n # print(type(self), type(obj), type(objtype))\\n return self.val\",\n \"def _get_value(self):\\n \\n return self._value\",\n \"def getProperty(*args):\",\n \"def getProperty(*args):\",\n \"def getProperty(*args):\",\n \"def getProperty(*args):\",\n \"def getProperty(*args):\",\n \"def do_get_property(self, pspec):\\n\\n if pspec.name == 'joined':\\n return self._joined\\n\\n if self._get_properties_call is not None:\\n print('%r: Blocking on GetProperties() because someone '\\n 'wants property %s', self, pspec.name)\\n self._get_properties_call.block()\\n\\n if pspec.name == 'id':\\n return self._id\\n elif pspec.name == 'name':\\n return self._name\\n elif pspec.name == 'color':\\n return self._color\\n elif pspec.name == 'type':\\n return self._type\\n elif pspec.name == 'tags':\\n return self._tags\\n elif pspec.name == 'private':\\n return self._private\",\n \"def getValuesFor(self, obj):\\n val = valueInContext(self.value, obj)\\n assert isinstance(val, dict)\\n return val\",\n \"def __get__(self, obj, cls):\\n if obj is None:\\n return self\\n\\n value = self.fget(obj)\\n setattr(obj, self.fget.__name__, value)\\n return value\",\n \"def getProperty(self, propertyName: unicode) -> unicode:\\n ...\",\n \"def _GetPropertyValue(entity, property):\\n if property in datastore_types._SPECIAL_PROPERTIES:\\n assert property == datastore_types._KEY_SPECIAL_PROPERTY\\n return entity.key()\\n else:\\n return entity[property]\",\n \"def _get(obj, name):\\n try:\\n # try to get value using dict's __getitem__ descriptor first\\n return dict.__getitem__(obj, name)\\n except TypeError:\\n # if it's a dict, then preserve the TypeError\\n if isinstance(obj, dict):\\n raise\\n # otherwise try one last time, relying on __getitem__ if any\\n return obj[name]\",\n \"def get_dynamic_property(vim, mobj, type, property_name):\\r\\n obj_content = \\\\\\r\\n get_object_properties(vim, None, mobj, type, [property_name])\\r\\n property_value = None\\r\\n if obj_content:\\r\\n dynamic_property = obj_content[0].propSet\\r\\n if dynamic_property:\\r\\n property_value = dynamic_property[0].val\\r\\n return property_value\",\n \"def getProperty(unique_name):\",\n \"def get(self, item):\\n try:\\n data = copy.deepcopy(getattr(self, item))\\n except AttributeError:\\n raise M2Error('Error while trying to get non-existent property `%s`' % item, False)\\n except SQLAlchemyError:\\n self.s.rollback()\\n raise\\n return data\",\n \"def value_of_css_property(self, name):\\n return self.element.value_of_css_property(name)\",\n \"def get_value(v):\\n # try to get attribute from user\\n obj = getattr(user, v, None)\\n # if no, search for corresponding method in MailgunBatchMessage\\n if not obj:\\n obj = getattr(self, 'get_var_' + v, None)\\n if obj:\\n obj = functools.partial(obj, user)\\n\\n # is there is no function to get var value raise or return empty string\\n else:\\n if strict:\\n raise ValueError('There is no way to get `{0}` from user.'\\n 'user.{0} and MailgunBatchMessage.get_var_{0} tried'.format(v))\\n return ''\\n\\n return obj() if callable(obj) else obj\",\n \"def get(self):\\n return self._value\",\n \"def get_value(self):\\n return self._value\",\n \"def getValue(self, name: unicode) -> object:\\n ...\",\n \"def get_value(self, reference_path):\\n\\n if (reference_path.start_protocol is not None and\\n reference_path.start_protocol != self.id):\\n\\n raise ValueError('The reference path does not target this protocol.')\\n\\n if reference_path.property_name is None or reference_path.property_name == '':\\n raise ValueError('The reference path does specify a property to return.')\\n\\n return get_nested_attribute(self, reference_path.property_name)\",\n \"def property(self) -> Optional[pulumi.Input[str]]:\\n return pulumi.get(self, \\\"property\\\")\",\n \"def get_property_value(self, name):\\n # Return table field as property\\n tableName, primKey = self.provider._split_path(self.path)\\n if primKey is not None:\\n ns, localName = util.split_namespace(name)\\n if ns == (tableName + \\\":\\\"):\\n conn = self.provider._init_connection()\\n fieldlist = self.provider._get_field_list(conn, tableName)\\n if localName in fieldlist:\\n val = self.provider._get_field_by_primary_key(\\n conn, tableName, primKey, localName\\n )\\n conn.close()\\n return val\\n conn.close()\\n # else, let default implementation return supported live and dead properties\\n return super().get_property_value(name)\",\n \"def get(key, nodename=None):\\n return _get_property(key, nodename, None)\",\n \"def _value_for(self, object, name, value):\\n try:\\n return self.value_for(value)\\n except TraitError:\\n self.error(object, name, value)\",\n \"def _value_for(self, object, name, value):\\n try:\\n return self.value_for(value)\\n except TraitError:\\n self.error(object, name, value)\",\n \"def _val(obj):\\n if isinstance(obj, ctypes._SimpleCData):\\n return obj.value\\n else:\\n return obj\",\n \"def callPropertyGet(self, name = \\\"__value\\\", index = None):\\n\\t\\tif name == 'IID':\\n\\t\\t\\treturn CSLValue(typeid = \\\"string\\\", value = self.callerInfo.IID)\\n\\t\\tEntry = self.vtbl['p_' + name + \\\"_get\\\"]\\n\\t\\t#localTbl = { 'vars':{}, 'status':0, 'props':{}, 'alias':{}, 'persistent':{}, 'instance':{}}\\n\\t\\tpropEntry = Entry.parent\\n\\t\\tprmName = propEntry.data['prm']\\n\\t\\tlocalTbl = self.CSLCreateLocalTbl({}, {}, {}, copy.copy(propEntry.data['persistent']), copy.copy(propEntry.data['instance']))\\n\\t\\tif prmName != \\\"\\\":\\n\\t\\t\\tif index == None:\\n\\t\\t\\t\\tdefault = propEntry.data['default']\\n\\t\\t\\t\\tif default != \\\"\\\":\\n\\t\\t\\t\\t\\tdefault = self.CSLCheckValue(default, localTbl)\\n\\t\\t\\t\\telse:\\n\\t\\t\\t\\t\\tdefault = CSLValue(typeid = \\\"NULL\\\", value = None)\\n\\t\\t\\telse:\\n\\t\\t\\t\\tdefault = index\\n\\n\\t\\t\\tlocalTbl['vars'][prmName] = default\\n\\n\\t\\tself.procStack.append('p_' + name + '_get')\\n\\t\\tself.lastLTbl = self.CSLInterpreter(Entry.child, localTbl)\\n\\t\\tself.procStack.pop()\\n\\t\\tl = self.lastLTbl['vars']\\n\\n\\t\\tself.debug(DEBUG_CALL, \\\"\\\\n\\\\nGetProp result: (\\\", name, \\\")\\\", l, \\\"haskey:\\\", l.has_key(name))\\n\\n\\t\\tif l != None and l.has_key(name):\\n\\t\\t\\tself.debug(DEBUG_CALL, \\\"Get Property return:\\\", l[name])\\n\\t\\t\\treturn copy.deepcopy(l[name])\\n\\t\\telse:\\n\\t\\t\\treturn CSLValue(typeid = \\\"NULL\\\", value = None)\",\n \"def get_Value(self):\\n obj_p=object.__getattribute__(self, 'StoredValue')\\n if isinstance(obj_p, type(True)):\\n return obj_p & self.statusOn \\n else: \\n if self.statusOn: \\n return obj_p \\n else: \\n self._log.warning('You are attempting to access the job'+ \\n 'property %s which has not been set', self._context_name) \\n return None\",\n \"def getattribute(objeto, name: str):\\r\\n # Get internal dict value matching name.\\r\\n value = objeto.__dict__.get(name)\\r\\n if not value:\\r\\n # Raise AttributeError if attribute value not found.\\r\\n return None\\r\\n # Return attribute value.\\r\\n return value\",\n \"def propertyValueFromNodeGetter(instance):\\n\\n subspace = nodeGetter(instance)\\n\\n if SynchronousPropertyAccess.SynchronousPropertyAccess.getCurrent() is not None:\\n if not subspace.loaded:\\n subspace.keyspace.waitLoaded()\\n else:\\n subspace.keyspace.ensureSubscribed()\\n\\n val = subspace.value\\n\\n if val is None:\\n return default()\\n\\n return val[0]\",\n \"def get(self, key):\\n\\n return self.dict_props.get(key)\",\n \"def getPropertie(self, propname):\\n if propname in self._getPropName():\\n return self.properties[propname]\\n else:\\n raise Exception(\\\"la propiedad no existe\\\")\",\n \"def get_val_obj(self):\\n return self.val_obj\",\n \"def _column_value_getter(self, obj):\\n if is_namedtuple(obj):\\n # Get values using properties\\n def get(obj, column):\\n return getattr(obj, column, None)\\n\\n elif is_dict_like(obj):\\n # Get values using dictionary keys\\n def get(obj, column):\\n return obj.get(column)\\n\\n elif is_list_like(obj):\\n # Get values using list indexes\\n def get(obj, column):\\n col = self.column_location(column)\\n try:\\n return obj[col]\\n except IndexError:\\n return None\\n\\n else:\\n # Fallback to single column\\n def get(obj, _):\\n return obj\\n\\n return get\",\n \"def getProperty(self, propID, time, objectID=0):\\n key = (propID, objectID, time)\\n if propID not in self.properties.index:\\n\\n raise APIError.APIError('Unknown property ID')\\n\\n if time not in self.properties.index:\\n prop = interP.interpolateProperty(self.properties,\\n time=time,\\n propertyID=propID,\\n objectID=objectID,\\n method='linear')\\n else:\\n prop = self.properties[key]\\n\\n # Check pyro registering if applicaple\\n if hasattr(self, '_pyroDaemon') and not hasattr(prop, '_PyroURI'):\\n uri = self._pyroDaemon.register(prop)\\n prop._PyroURI = uri\\n\\n return(prop)\",\n \"def getprop(self, prop_name):\\n return self.shell(\\\"getprop %s\\\" % prop_name)\",\n \"def get_variable(self, name):\\n return self._properties[name]\",\n \"def Get(self, interface_name, property_name):\\n logger.debug(\\n \\\"%r.Get(%r, %r) -> ...\\\",\\n self, interface_name, property_name)\\n try:\\n props = self._dct_entry[interface_name]\\n except KeyError:\\n raise dbus.exceptions.DBusException(\\n dbus.PROPERTIES_IFACE,\\n \\\"No such interface {}\\\".format(interface_name))\\n try:\\n prop = props[property_name]\\n except KeyError:\\n raise dbus.exceptions.DBusException(\\n dbus.PROPERTIES_IFACE,\\n \\\"No such property {}:{}\\\".format(\\n interface_name, property_name))\\n try:\\n value = prop.__get__(self, self.__class__)\\n except dbus.exceptions.DBusException as exc:\\n logger.error(\\n \\\"%r.Get(%r, %r) -> (exception) %r\\\",\\n self, interface_name, property_name, exc)\\n raise\\n except Exception as exc:\\n logger.exception(\\n \\\"runaway exception from Get(%r, %r)\\\",\\n interface_name, property_name)\\n raise dbus.exceptions.DBusException(\\n dbus.PROPERTIES_IFACE,\\n \\\"Unable to get property interface/property {}:{}: {!r}\\\".format(\\n interface_name, property_name, exc))\\n else:\\n logger.debug(\\n \\\"%r.Get(%r, %r) -> %r\\\",\\n self, interface_name, property_name, value)\\n return value\",\n \"def get_value(key, obj, default=missing):\\n if isinstance(key, int):\\n return _get_value_for_key(key, obj, default)\\n return _get_value_for_keys(key.split('.'), obj, default)\",\n \"def get_property(self, client):\\r\\n client.getProperty()\",\n \"def get_value(self, key):\\n return self[key]\",\n \"def get_device_property(self, client, prop):\\r\\n value = client.getDeviceProperty(prop)\\r\\n return value\",\n \"def read_value(self):\\n return self.load_attr(\\\"value\\\")\",\n \"def __get__(self,obj,objtype):\\n result = super(Number,self).__get__(obj,objtype)\\n # CEBALERT: results in extra lookups (_value_is_dynamic() is\\n # also looking up 'result' - should just pass it in). Note\\n # that this method is called often.\\n if self._value_is_dynamic(obj,objtype): self._check_value(result)\\n return result\",\n \"def get(self, obj):\\n obj = self._to_obj_tuple(obj)\\n rows = self.query(object=obj)\\n if rows:\\n return rows[0]\",\n \"def GetValue(self):\\n return self._value\",\n \"def GetValue(self):\\n return self._value\",\n \"def _get(self, name):\\n return object.__getattribute__(self, name)\",\n \"def _get(self, name):\\n return object.__getattribute__(self, name)\",\n \"def getValue(self):\\n \\n return self._value\",\n \"def get_value(self):\\n return self.value\",\n \"def get_value(self):\\n return self.value\",\n \"def get_value(self):\\n return self.value\",\n \"def get_value(self):\\n return self._value\",\n \"def get_value(self):\\n return self._value\",\n \"def get_value(self):\\n return self._value\",\n \"def __getattr__ (self, name):\\n\\t\\ttry:\\n\\t\\t\\treturn self.__dict__[name]\\n\\t\\texcept KeyError:\\n\\t\\t\\treturn self.__dict__[\\\"value\\\"][name]\",\n \"def GetWirelessProperty(self, networkid, prop):\\n try:\\n value = self.LastScan[networkid].get(prop)\\n except IndexError:\\n if self.debug_mode:\\n print \\\"GetWirelessProperty: Index error occured trying to \\\" + \\\\\\n \\\"retrieve property %s\\\" % prop\\n value = \\\"\\\"\\n try:\\n value = misc.to_unicode(value)\\n except:\\n pass\\n return value\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.7294242","0.7260294","0.72458637","0.7177455","0.71202344","0.6993076","0.69180053","0.6906648","0.6891664","0.68564713","0.68375313","0.6823808","0.6799708","0.6789528","0.67803353","0.67259824","0.67239094","0.6705578","0.6705456","0.66965705","0.6668577","0.66025084","0.6584929","0.6561236","0.65505","0.6515804","0.6498194","0.6462627","0.644257","0.64325255","0.6410394","0.6408052","0.6408052","0.63819456","0.63810456","0.6380277","0.6375951","0.6371185","0.6361211","0.6343597","0.63352215","0.6321889","0.6306985","0.6306985","0.6306985","0.6306985","0.6306985","0.6296232","0.6280495","0.62680507","0.6242114","0.6235011","0.6226821","0.6220504","0.6206488","0.6203724","0.6188937","0.61866325","0.6181314","0.6178142","0.6153527","0.61501807","0.61469465","0.61411756","0.6133938","0.6115981","0.6115981","0.61074245","0.61015165","0.60992455","0.60905075","0.60886526","0.6084259","0.60827565","0.6082152","0.6062388","0.6062295","0.6043301","0.6043141","0.60292405","0.6015712","0.59755373","0.5970953","0.59686345","0.5966111","0.5955677","0.59491926","0.59456074","0.59456074","0.594509","0.594509","0.5926986","0.592019","0.592019","0.592019","0.5917917","0.5917917","0.5917917","0.5916588","0.59141695"],"string":"[\n \"0.7294242\",\n \"0.7260294\",\n \"0.72458637\",\n \"0.7177455\",\n \"0.71202344\",\n \"0.6993076\",\n \"0.69180053\",\n \"0.6906648\",\n \"0.6891664\",\n \"0.68564713\",\n \"0.68375313\",\n \"0.6823808\",\n \"0.6799708\",\n \"0.6789528\",\n \"0.67803353\",\n \"0.67259824\",\n \"0.67239094\",\n \"0.6705578\",\n \"0.6705456\",\n \"0.66965705\",\n \"0.6668577\",\n \"0.66025084\",\n \"0.6584929\",\n \"0.6561236\",\n \"0.65505\",\n \"0.6515804\",\n \"0.6498194\",\n \"0.6462627\",\n \"0.644257\",\n \"0.64325255\",\n \"0.6410394\",\n \"0.6408052\",\n \"0.6408052\",\n \"0.63819456\",\n \"0.63810456\",\n \"0.6380277\",\n \"0.6375951\",\n \"0.6371185\",\n \"0.6361211\",\n \"0.6343597\",\n \"0.63352215\",\n \"0.6321889\",\n \"0.6306985\",\n \"0.6306985\",\n \"0.6306985\",\n \"0.6306985\",\n \"0.6306985\",\n \"0.6296232\",\n \"0.6280495\",\n \"0.62680507\",\n \"0.6242114\",\n \"0.6235011\",\n \"0.6226821\",\n \"0.6220504\",\n \"0.6206488\",\n \"0.6203724\",\n \"0.6188937\",\n \"0.61866325\",\n \"0.6181314\",\n \"0.6178142\",\n \"0.6153527\",\n \"0.61501807\",\n \"0.61469465\",\n \"0.61411756\",\n \"0.6133938\",\n \"0.6115981\",\n \"0.6115981\",\n \"0.61074245\",\n \"0.61015165\",\n \"0.60992455\",\n \"0.60905075\",\n \"0.60886526\",\n \"0.6084259\",\n \"0.60827565\",\n \"0.6082152\",\n \"0.6062388\",\n \"0.6062295\",\n \"0.6043301\",\n \"0.6043141\",\n \"0.60292405\",\n \"0.6015712\",\n \"0.59755373\",\n \"0.5970953\",\n \"0.59686345\",\n \"0.5966111\",\n \"0.5955677\",\n \"0.59491926\",\n \"0.59456074\",\n \"0.59456074\",\n \"0.594509\",\n \"0.594509\",\n \"0.5926986\",\n \"0.592019\",\n \"0.592019\",\n \"0.592019\",\n \"0.5917917\",\n \"0.5917917\",\n \"0.5917917\",\n \"0.5916588\",\n \"0.59141695\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6502106"},"document_rank":{"kind":"string","value":"26"}}},{"rowIdx":90873,"cells":{"query":{"kind":"string","value":"Set property value on object."},"document":{"kind":"string","value":"def set(self, obj, value):\n raise NotImplementedError"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def set(self, obj, value):\n pass","def set_value ( self, object, value ):\n target, name = self.target_name( object )\n setattr( target, name, value )","def __set__(self, obj, value):\r\n pass","def set_property(self, name, value):\n self.properties[name] = value","def __set__(self, obj, value):\n\n return setattr(obj, '_' + self.name, value)","def set(self,obj,value):\r\n\t\tvalue = self.parse(value)\r\n\t\tsetattr(obj,self.name,value)","def do_set_property(self, spec, value):\n attribute = self.find_attribute(spec.name)\n if attribute is not None and isinstance(attribute, property):\n attribute.fset(self, value)\n else:\n raise ValueError(\"No such property\", spec.name)","def set_value ( self, object, value ):\n object[ self.index ] = value","def do_set_property(self, pspec, val):\n # FIXME: need an asynchronous API to set these properties,\n # particularly 'private'\n\n if pspec.name == 'name':\n self._name = val\n elif pspec.name == 'color':\n self._color = val\n elif pspec.name == 'tags':\n self._tags = val\n elif pspec.name == 'private':\n self._private = val\n else:\n raise ValueError('Unknown property %r', pspec.name)\n\n self._publish_properties()","def set_property(key, value):\n return impl.set_property(**locals())","def set_user_property(self, property, value):\n self._user_properties.update({property: value})","def SetProperty(self, propertyname, value):\n if self.serviceimplementation == 'basic':\n flag = self.vbLet\n if isinstance(value, datetime.datetime):\n value = SFScriptForge.SF_Basic.CDateToUnoDateTime(value)\n flag += self.flgDateArg\n if repr(type(value)) == \"<class 'pyuno'>\":\n flag += self.flgUno\n return self.EXEC(self.objectreference, flag, propertyname, value)","def value(self, val: Any) -> None:\n self.component.set_property_value(self, val)\n self._value = val","def prop_set(prop, value, extra_args=None, cibfile=None):\n return item_create(\n item=\"property\",\n item_id=\"{}={}\".format(prop, value),\n item_type=None,\n create=\"set\",\n extra_args=extra_args,\n cibfile=cibfile,\n )","def set_property(self,obj_property,value):\n existing_properties = self.get_property(obj_property)\n if existing_properties is None:\n if type(value) == list:\n if len(value) == 1:\n self.redis_server.hset(self.frbr_key,\n obj_property,\n value[0])\n else:\n new_redis_key = \"%s:%s\" % (self.frbr_key,\n obj_property)\n for row in value:\n self.redis_server.sadd(new_redis_key,\n row)\n self.redis_server.hset(self.frbr_key,\n obj_property,\n new_redis_key)\n else:\n self.redis_server.hset(self.frbr_key,\n obj_property,\n value)","def setprop(name, value):\n _slp.setprop(name, value)","def set_value(self, object, name, value):\n cname = TraitsCache + name\n old = object.__dict__.get(cname, Undefined)\n if value != old:\n object.__dict__[cname] = value\n object.trait_property_changed(name, old, value)","def set_value(self, object, name, value):\n cname = TraitsCache + name\n old = object.__dict__.get(cname, Undefined)\n if value != old:\n object.__dict__[cname] = value\n object.trait_property_changed(name, old, value)","def set(self, obj, value):\n assert self.setter is not None, \"Setter accessor is not specified.\"\n if callable(self.setter):\n return self.setter(obj, value)\n\n assert isinstance(self.setter, string_types), \"Accessor must be a function or a dot-separated string.\"\n\n def _set(obj, attr, value):\n if isinstance(obj, dict):\n obj[attr] = value\n else:\n setattr(obj, attr, value)\n return value\n\n path = self.setter.split(\".\")\n for attr in path[:-1]:\n obj = _set(obj, attr, {})\n\n _set(obj, path[-1], value)","def set_role_property(self, obj, property_name, property_value): # pylint: disable=no-self-use\n if isinstance(obj, dict):\n if 'properties' in obj:\n obj = obj['properties']\n obj[property_name] = property_value\n else:\n if hasattr(obj, 'properties'):\n obj = obj.properties\n obj.property_name = property_value","async def ufp_set(self, obj: T, value: Any) -> None:\n _LOGGER.debug(\"Setting %s to %s for %s\", self.name, value, obj.display_name)\n if self.ufp_set_method is not None:\n await getattr(obj, self.ufp_set_method)(value)\n elif self.ufp_set_method_fn is not None:\n await self.ufp_set_method_fn(obj, value)","def _update_property(self, field, value):\n if value is not None:\n self.properties[field] = value\n elif field in self.properties:\n del self.properties[field]","def __set__(self, instance, value):\n instance._values[self.name] = self.process(value)","def __setattr__(self, name, value):\n if self.serviceimplementation == 'basic':\n if name in ('serviceproperties', 'localProperties', 'internal_attributes', 'propertysynonyms',\n 'forceGetProperty'):\n pass\n elif name[0:2] == '__' or name in self.internal_attributes or name in self.localProperties:\n pass\n elif name in self.serviceproperties or name in self.propertysynonyms:\n if name in self.propertysynonyms: # Reset real name if argument provided in lower or camel case\n name = self.propertysynonyms[name]\n if self.internal: # internal = True forces property local setting even if property is read-only\n pass\n elif self.serviceproperties[name] is True: # True == Editable\n self.SetProperty(name, value)\n return\n else:\n raise AttributeError(\n \"type object '\" + self.objecttype + \"' has no editable property '\" + name + \"'\")\n else:\n raise AttributeError(\"type object '\" + self.objecttype + \"' has no property '\" + name + \"'\")\n object.__setattr__(self, name, value)\n return","def set_value(obj, key, value):\r\n try:\r\n obj[key] = value\r\n except TypeError:\r\n setattr(obj, key, value)","def SetProp(self, name, value):\n self._props[name] = value\n self._changed(name, value)","def __set__(self, instance, value):\n if self.mode == READONLY:\n raise AttributeError(\"This is a READONLY attribute\")\n value = self.validate(value)\n if self.name is None : self.name = Property.search(instance, None,self)\n if self.name is not None:\n instance.__dict__[self.name] = value\n if hasattr(instance, \"__store__\"):\n instance.__store__[self.name] = value\n self.deleted = False\n else:\n raise AttributeError(\"Cannot find %s in %s \" % (self,instance))","def assign(self, value):\n self.value = value","def __setattr__(self, name, value):\n self.set(**{name: value})","def __set__(self, instance, value):\n if value is None and not self.allow_none:\n raise TypeError('None is not a valid value for this property.')\n\n if value is not None and not isinstance(value, self.valid_types):\n raise TypeError('%s (%r) is not a valid type for this property.'\n % (type(value).__name__, value))\n\n attr_name = self.get_attr_name(instance)\n instance.__dict__['_%s_typed' % attr_name] = value","def set_property(self, client, key, value):\r\n client.setProperty(key, value)\r\n return True","def set_property(self, key, value):\n self.properties[key] = value\n return self.properties[key]","def __set__(self, model_instance, value):\r\n raise ValueError, 'Virtual property is read-only'","def _setPropertyValue(self, name, value, typeString = ''):\n method = getattr(self.__class__, \"_setPropertyValue\" + getTypeString(value))\n return method(self, name, value, typeString)","def set_value (self):\n raise NotImplementedError","def __set__(self, instance, val):\n raise AttributeError(\"Can't set attribute\")","def _setitem_impl(self, index, obj, value):\n obj.set_value(value)\n return obj","def __setattr__(self, key, value):\n return setattr(self.__dict__['_obj'], key, value)","def __setattr__ (self, attr, value):\n self.set_value (attr, value)","def XPSetWidgetProperty(inWidget, inProperty, inValue):\n pass","def put_prop(self, obj_type, obj_id, prop_name, value):\n ierr = exolib.py_expp(self.exoid, obj_type, obj_id, prop_name, value)\n if ierr:\n raise ExodusIIWriterError(\"Error putting prop value\")","def setValue(self, name: unicode, value: object) -> None:\n ...","def set_property(self, key, value):\n if not key:\n raise Exception('A key should be provided.')\n if not value:\n raise Exception('A value should be provided.')\n self._properties[key] = value","def setProperty(*args):","def setProperty(*args):","def setProperty(*args):","def setProperty(*args):","def setProperty(*args):","def set(self, attr, val):\r\n self.__dict__[attr] = val","def set(self, prop, value):\r\n\r\n prop_parts = prop.split(\".\")\r\n if self.copy_dict:\r\n new_dict = copy.deepcopy(self.obj)\r\n else:\r\n new_dict = self.obj\r\n pointer = None\r\n parts_length = len(prop_parts) - 1\r\n for i, part in enumerate(prop_parts):\r\n if pointer is None and i == parts_length:\r\n new_dict[part] = value\r\n elif pointer is None:\r\n pointer = new_dict.get(part)\r\n elif i == parts_length:\r\n pointer[part] = value\r\n else:\r\n pointer = pointer.get(part)\r\n return new_dict","def setProperty(self, path, key, value):\n \n try:\n self._client.propset(key, value, self._workingCopyPath + path)\n self.checkin(path)\n except ClientError, error:\n raise SubversionError(error)","def run_setprop(self, expanded, unexpanded) :\n\t\toptions, args = self.getopt([\"name=\", \"value=\"], expanded)\n\t\tif (options is None) and (args is None) :\n\t\t\treturn -1\t# message was already displayed in self.getopt()\n\t\tif not args :\n\t\t\treturn self.errormessage(\"Needs at least one object to change this property\")\n\t\tif not (options.has_key(\"name\") and options.has_key(\"value\")) :\n\t\t\treturn self.errormessage(\"You must supply a property name and value\")\n\t\tproperty = options[\"name\"]\n\t\tif options.has_key(\"value\") :\n\t\t\tpropvalue = options[\"value\"]\n\t\t\ttry :\n\t\t\t\t# maybe it's a list in a string, e.g. \"['e', 'f']\"\n\t\t\t\t# or something like that\n\t\t\t\tnewvalue = eval(propvalue)\n\t\t\t\tif (type(newvalue) != type(0)) and (type(newvalue) != type(0.0)) :\n\t\t\t\t\t# we mustn't convert numeric to string\n\t\t\t\t\tpropvalue = newvalue\n\t\t\texcept NameError :\n\t\t\t\tpass\t# normal string\n\t\telse :\n\t\t\tpropvalue = \"\"\n\n\t\tstatus = 0\n\t\tfor arg in args :\n\t\t\tobject = self.toObject(self.__context, arg)\n\t\t\tif object is not None :\n\t\t\t\tif not self.HasPerms(object, 'Manage properties') :\n\t\t\t\t\tstatus = status - 1\n\t\t\t\telif hasattr(object, 'hasProperty') :\n\t\t\t\t\tif not object.hasProperty(property) :\n\t\t\t\t\t\tstatus = status + self.errormessage(\"Object %s has no property %s\" % (self.ObjectPath(object), property))\n\t\t\t\t\telse :\n\t\t\t\t\t\t# in the following lines the absence of a _updateProperty\n\t\t\t\t\t\t# attribute indicates an object without properties (e.g. a method)\n\t\t\t\t\t\t# which indicates an object for which setting properties is a nonsense\n\t\t\t\t\t\tif hasattr(object, \"_updateProperty\") :\n\t\t\t\t\t\t\tobject._updateProperty(property, propvalue)\n\t\t\t\t\t\t\tself.htmlmessage(\"Object %s property %s was modified to %s\" % (self.ObjectPath(object), property, str(propvalue)))\n\t\treturn status","def set_prop(prop, value, config_type=\"\", config_path=\"\"):\n\n path = get_config_path(config_type, config_path)\n set_property_value(prop, value, path)","def set_value(self, _, prop_value_pair):\n (prop, val) = prop_value_pair\n\n _, _, obj = self.popup_data\n if not isinstance(obj, value_model.Value):\n raise TypeError(\"Expected %s\" % type(value_model.Value))\n\n if not prop == obj.parent:\n raise ValueError(\"Property '%s' is not the parent of '%s'\" % (prop, obj))\n\n # To enable undo redo for this we need a bit of trickery\n new_prop = prop.clone(keep_id=True)\n create_pseudo_values([new_prop])\n if new_prop.pseudo_values[obj.index].pseudo_values != obj.pseudo_values:\n raise ValueError(\"Cannot find replacement value\")\n\n # Update the value in the new property\n new_prop.pseudo_values[obj.index].pseudo_values = val\n\n # Lets replace the old property with the new and updated one\n cmd = commands.ReplaceObject(obj=prop, repl=new_prop)\n self.execute(cmd)\n\n # Reset the view to make sure the changes are properly displayed.\n self.select_object(new_prop)\n self.reset_value_view(None)","def __setattr__(self, attr, value):\n field = getattr(self, attr)\n if isinstance(field, BaseType):\n # Check the value type\n check = field.accept_value(value)\n \n old_value = getattr(self, attr)\n object.__setattr__(self, attr, value)\n if isinstance(old_value, BaseType):\n # Not set yet\n old_value = None\n \n if Model.data_connector and Model.data_connector.running:\n with Model.data_connector.u_lock:\n Model.data_connector.update_object(self, attr, old_value)","def value(self, value):\n\n\t\tself.__value = value","def changeProperty(self, node, name, propertyName, value, setIfNotExist=False):","def __set__(self, obj, value):\r\n if value != None and (obj.id == value or (hasattr(value, \"id\") and obj.id == value.id)):\r\n raise ValueError, \"Can not associate an object with itself!\"\r\n return super(ReferenceProperty, self).__set__(obj,value)","def set(self, database_name, property_, value):\n return isempty(self._send_command(database_name, \"%s=%s\" % (property_, value)))","def set_value(self,x):\n self._value = x","def set_value(self,x):\n self._value = x","def proxy_set(self, value):\n setter = getattr(self, self.proxy_setter)\n return setter(value)","def setitem(obj, attr, value):\n from functools import partial\n _setattr = partial(setattr, obj)\n setter = getattr(obj, '__setitem__', None) or _setattr\n setter(attr, value)","def set(self, properties):\n raise NotImplementedError","def setValue(self, value):\n self._value = value","def set(self, obj, success=True):\n self.value = obj\n self.success = success\n if self.callback and success:\n self.callback(obj)\n if self.errback and not success:\n self.errback(obj)\n with self.condition:\n self.ready = True\n self.condition.notify()","def setter(self, func):\n self.fset = func\n self.set_setattr_mode(SetAttr.CallObject_ObjectValue, func)\n return self","def set_value(self, value):\n self.value = value","def set_value(self, value):\n self.value = value","def set_value(self, value):\n self.value = value","def _platformix_set(self, context, fake_reply, prop, value):\r\n if hasattr(self.host, prop):\r\n if not callable(getattr(self.host, prop)):\r\n try:\r\n setattr(self.host, prop, value)\r\n except Exception as e:\r\n eprint(\"Platformix protocol: failed to set attribute {} of {} to value {} \"\r\n \"due to exception {}\".format(prop, self.host.name, value, e))\r\n exprint()\r\n self._reply(context, proto_failure(\r\n \"Failed to set attribute {} of {} to value {} \"\r\n \"due to exception {}\".format(prop, self.host.name, value, e)), fake_reply)\r\n return\r\n self._reply(context, proto_success(getattr(self.host, prop), prop), fake_reply)\r\n else:\r\n self._reply(context, proto_failure(\"Attribute {} of {} is a method\".format(\r\n prop, self.host.name)), fake_reply)\r\n else:\r\n self._reply(context, proto_failure(\"Property {} not found on {}\".format(prop, self.host.name)), fake_reply)","def set(self, name, value):\n pass","def set_property(wellorcontainer, property_name, value):\n wells = convert_to_wellgroup(wellorcontainer)\n \n if not isinstance(value, str):\n value = str(value)\n \n for well in wells:\n assert isinstance(well, Well)\n well.properties[property_name] = value","def setDynamicProperty(self, widget, property_name, property_value):\n\n\t\twidget.setProperty(property_name, property_value)\n\t\twidget.style().unpolish(widget)\n\t\twidget.style().polish(widget)","def set_value(self, val):\n self.value = val","def setProperty(self, newProp, objectID=0):\n\n # Set the new property to container\n key = (newProp.getPropertyID(),\n newProp.getObjectID(),\n newProp.getTime())\n self.properties.set_value(key, newProp)","def setProperty(self, child, key, value):\n\n # First get the child's dictionary\n childDict = self.getInfoDict(child)\n if childDict:\n childDict[key] = value","def put_prop(self, obj_type, obj_id, prop_name, value):\n ierr = exolib.py_expp(self.exoid, obj_type, obj_id, prop_name, value)\n if ierr:\n raise ExodusIIWriterError(\"Error putting prop\")","def set_property_value(self, current_value, new_value):\n if self._enabled:\n current_value *= self.marker_factor\n new_value *= self.marker_factor\n if live_object_is_valid(self._parent) and current_value != new_value:\n new_param_value = max(self._absolute_range[0] * self.marker_factor, min(self._absolute_range[1] * self.marker_factor, new_value))\n setattr(self._parent, self._property_name, new_param_value)\n if self._should_zoom:\n self._parent.view.show_loop()\n self._parent.view.show_loop()","def set_value(self, val):\n self._value = val","def set_value(self, val):\n self._value = val","def _set_value(o, d):\n if isinstance(o, Param) and not o._mutable:\n return # ignore requests to set immutable params\n else:\n try:\n o.value = d\n except AttributeError:\n o = d # this would be an indexed parameter","def set_property(self, vtk_property):\n self.vtk_act.SetProperty(vtk_property)","def set(self, attr_name, value):\n\n get_logger().debug('Going to set XML attribute property: %s', attr_name)\n\n if self.textproperty is not None:\n # TODO: potentially programming error\n raise MarshallingError()\n\n try:\n # pylint: disable=unnecessary-dunder-call\n self.attributes[attr_name].__set__(self.obj, value)\n get_logger().debug('Set XML attribute property: %s', attr_name)\n except AttributeError as ex:\n get_logger().error('XML property %s: %s', attr_name, str(ex))\n except KeyError:\n get_logger().debug('Not an XML attribute property: %s', attr_name)","def SetPropObject(self, obj):\n\t\tif self.propObj is not None:\n\t\t\t# deregister notification\n\t\t\tif hasattr(self.propObj, \"DropPropsNotify\") :\n\t\t\t\tself.propObj.DropPropsNotify()\n\t\tif hasattr(obj, \"GetPropList\") :\n\t\t\tself.propObj = obj\n\t\t\tself.UpdateGrid()\n\t\telse :\n\t\t\twx.MessageBox(_(\"Can't display properties for %s, because it hasn't GetPropList interface\") % obj.__class__.__name__,\n\t\t\t\t\t\t\t_(\"Sampo Framework\"), wx.OK | wx.CENTER | wx.ICON_ERROR)","def set(self, attribute, value):\n self.__setattr__(attribute, value)","def autoprops_generated_setter(self, **kwargs):\n setattr(self, private_property_name, kwargs[property_name])","def __set__(self, instance, value):\n if self.deprecated:\n cls_name = type(instance).__name__\n warnings.warn('%s.%s is deprecated. Please set %s.%s instead.'\n % (cls_name, self.get_attr_name(instance),\n cls_name, self.prop_name),\n self.deprecation_warning,\n stacklevel=2)\n\n if value is not None and self._convert_to_func is not None:\n value = self._convert_to_func(value)\n\n setattr(instance, self.prop_name, value)","def set_property(self, name, value, persist_changes=True):\n self._properties_metadata[name] = {'readonly': not persist_changes}\n self._properties[name] = value","def setValue(self, *args):\n return _libsbml.ConversionProperties_setValue(self, *args)","def setter(self, value):\n if field.readonly:\n raise fields.ValidationError(f\"'{name}' is a read only attribute\")\n\n # accept if this is a raw value too\n value = field.from_raw(value)\n\n # validate\n field.validate(value)\n\n # set current instance as parent for embedded objects/instances\n if isinstance(field, fields.Object):\n value.parent = self\n\n # se attribute\n setattr(self, inner_name, value)\n self._attr_updated(name, value)","def __setattr__(self, name, value):\n if not hasattr(self, name):\n raise AttributeError(\"'%s' object has no attribute '%s'\" % (type(self).__name__, name))\n object.__setattr__(self, name, value)","def setProperty(self,prop=None):\n objects = self.check()\n if objects:\n if prop is None:\n res = askItems([['property',0]],\n caption = 'Set Property Number for Selection (negative value to remove)')\n if res:\n prop = int(res['property'])\n if prop < 0:\n prop = None\n for o in objects:\n if hasattr(o,'setProp'):\n o.setProp(prop)\n self.draw()","def set_property_value(self, name, value, dry_run=False):\n raise DAVError(\n HTTP_FORBIDDEN, err_condition=PRECONDITION_CODE_ProtectedProperty\n )","def __set__( self, client, value ):\n\t\tif value is None:\n\t\t\ttry:\n\t\t\t\treturn self.__delete__( client )\n\t\t\texcept AttributeError, err:\n\t\t\t\t# wasn't defined anyway\n\t\t\t\treturn None\n\t\telse:\n\t\t\treturn super( DBProperty, self ).__set__( client, value )","def set_value(self, field, value):\n field = self.find_first(field)\n if field is not None:\n field.value = value","def __setattr__(self, attr, obj):\n\n # Underscored-prefixed/volatile/stateful slots can be set\n # without affecting other such slots.\n if (\n attr not in self.__volatile_slots__\n # Are we trying to set a custom property?\n and attr in getattr(self, \"__all_props__\", ())\n # Is it a custom property that's already been set?\n and hasattr(self, attr)\n # Are we setting it to a new value?\n and getattr(self, attr) is not obj\n ):\n self.reset()\n\n object.__setattr__(self, attr, obj)","def __setattr__(self, attr, value):\n super().__setattr__(attr, value)","def __set__(self, instance, value):\n raise AttributeError(\"A Default Property is Read only\")","def setval(self, val):\r\n self.value = val"],"string":"[\n \"def set(self, obj, value):\\n pass\",\n \"def set_value ( self, object, value ):\\n target, name = self.target_name( object )\\n setattr( target, name, value )\",\n \"def __set__(self, obj, value):\\r\\n pass\",\n \"def set_property(self, name, value):\\n self.properties[name] = value\",\n \"def __set__(self, obj, value):\\n\\n return setattr(obj, '_' + self.name, value)\",\n \"def set(self,obj,value):\\r\\n\\t\\tvalue = self.parse(value)\\r\\n\\t\\tsetattr(obj,self.name,value)\",\n \"def do_set_property(self, spec, value):\\n attribute = self.find_attribute(spec.name)\\n if attribute is not None and isinstance(attribute, property):\\n attribute.fset(self, value)\\n else:\\n raise ValueError(\\\"No such property\\\", spec.name)\",\n \"def set_value ( self, object, value ):\\n object[ self.index ] = value\",\n \"def do_set_property(self, pspec, val):\\n # FIXME: need an asynchronous API to set these properties,\\n # particularly 'private'\\n\\n if pspec.name == 'name':\\n self._name = val\\n elif pspec.name == 'color':\\n self._color = val\\n elif pspec.name == 'tags':\\n self._tags = val\\n elif pspec.name == 'private':\\n self._private = val\\n else:\\n raise ValueError('Unknown property %r', pspec.name)\\n\\n self._publish_properties()\",\n \"def set_property(key, value):\\n return impl.set_property(**locals())\",\n \"def set_user_property(self, property, value):\\n self._user_properties.update({property: value})\",\n \"def SetProperty(self, propertyname, value):\\n if self.serviceimplementation == 'basic':\\n flag = self.vbLet\\n if isinstance(value, datetime.datetime):\\n value = SFScriptForge.SF_Basic.CDateToUnoDateTime(value)\\n flag += self.flgDateArg\\n if repr(type(value)) == \\\"<class 'pyuno'>\\\":\\n flag += self.flgUno\\n return self.EXEC(self.objectreference, flag, propertyname, value)\",\n \"def value(self, val: Any) -> None:\\n self.component.set_property_value(self, val)\\n self._value = val\",\n \"def prop_set(prop, value, extra_args=None, cibfile=None):\\n return item_create(\\n item=\\\"property\\\",\\n item_id=\\\"{}={}\\\".format(prop, value),\\n item_type=None,\\n create=\\\"set\\\",\\n extra_args=extra_args,\\n cibfile=cibfile,\\n )\",\n \"def set_property(self,obj_property,value):\\n existing_properties = self.get_property(obj_property)\\n if existing_properties is None:\\n if type(value) == list:\\n if len(value) == 1:\\n self.redis_server.hset(self.frbr_key,\\n obj_property,\\n value[0])\\n else:\\n new_redis_key = \\\"%s:%s\\\" % (self.frbr_key,\\n obj_property)\\n for row in value:\\n self.redis_server.sadd(new_redis_key,\\n row)\\n self.redis_server.hset(self.frbr_key,\\n obj_property,\\n new_redis_key)\\n else:\\n self.redis_server.hset(self.frbr_key,\\n obj_property,\\n value)\",\n \"def setprop(name, value):\\n _slp.setprop(name, value)\",\n \"def set_value(self, object, name, value):\\n cname = TraitsCache + name\\n old = object.__dict__.get(cname, Undefined)\\n if value != old:\\n object.__dict__[cname] = value\\n object.trait_property_changed(name, old, value)\",\n \"def set_value(self, object, name, value):\\n cname = TraitsCache + name\\n old = object.__dict__.get(cname, Undefined)\\n if value != old:\\n object.__dict__[cname] = value\\n object.trait_property_changed(name, old, value)\",\n \"def set(self, obj, value):\\n assert self.setter is not None, \\\"Setter accessor is not specified.\\\"\\n if callable(self.setter):\\n return self.setter(obj, value)\\n\\n assert isinstance(self.setter, string_types), \\\"Accessor must be a function or a dot-separated string.\\\"\\n\\n def _set(obj, attr, value):\\n if isinstance(obj, dict):\\n obj[attr] = value\\n else:\\n setattr(obj, attr, value)\\n return value\\n\\n path = self.setter.split(\\\".\\\")\\n for attr in path[:-1]:\\n obj = _set(obj, attr, {})\\n\\n _set(obj, path[-1], value)\",\n \"def set_role_property(self, obj, property_name, property_value): # pylint: disable=no-self-use\\n if isinstance(obj, dict):\\n if 'properties' in obj:\\n obj = obj['properties']\\n obj[property_name] = property_value\\n else:\\n if hasattr(obj, 'properties'):\\n obj = obj.properties\\n obj.property_name = property_value\",\n \"async def ufp_set(self, obj: T, value: Any) -> None:\\n _LOGGER.debug(\\\"Setting %s to %s for %s\\\", self.name, value, obj.display_name)\\n if self.ufp_set_method is not None:\\n await getattr(obj, self.ufp_set_method)(value)\\n elif self.ufp_set_method_fn is not None:\\n await self.ufp_set_method_fn(obj, value)\",\n \"def _update_property(self, field, value):\\n if value is not None:\\n self.properties[field] = value\\n elif field in self.properties:\\n del self.properties[field]\",\n \"def __set__(self, instance, value):\\n instance._values[self.name] = self.process(value)\",\n \"def __setattr__(self, name, value):\\n if self.serviceimplementation == 'basic':\\n if name in ('serviceproperties', 'localProperties', 'internal_attributes', 'propertysynonyms',\\n 'forceGetProperty'):\\n pass\\n elif name[0:2] == '__' or name in self.internal_attributes or name in self.localProperties:\\n pass\\n elif name in self.serviceproperties or name in self.propertysynonyms:\\n if name in self.propertysynonyms: # Reset real name if argument provided in lower or camel case\\n name = self.propertysynonyms[name]\\n if self.internal: # internal = True forces property local setting even if property is read-only\\n pass\\n elif self.serviceproperties[name] is True: # True == Editable\\n self.SetProperty(name, value)\\n return\\n else:\\n raise AttributeError(\\n \\\"type object '\\\" + self.objecttype + \\\"' has no editable property '\\\" + name + \\\"'\\\")\\n else:\\n raise AttributeError(\\\"type object '\\\" + self.objecttype + \\\"' has no property '\\\" + name + \\\"'\\\")\\n object.__setattr__(self, name, value)\\n return\",\n \"def set_value(obj, key, value):\\r\\n try:\\r\\n obj[key] = value\\r\\n except TypeError:\\r\\n setattr(obj, key, value)\",\n \"def SetProp(self, name, value):\\n self._props[name] = value\\n self._changed(name, value)\",\n \"def __set__(self, instance, value):\\n if self.mode == READONLY:\\n raise AttributeError(\\\"This is a READONLY attribute\\\")\\n value = self.validate(value)\\n if self.name is None : self.name = Property.search(instance, None,self)\\n if self.name is not None:\\n instance.__dict__[self.name] = value\\n if hasattr(instance, \\\"__store__\\\"):\\n instance.__store__[self.name] = value\\n self.deleted = False\\n else:\\n raise AttributeError(\\\"Cannot find %s in %s \\\" % (self,instance))\",\n \"def assign(self, value):\\n self.value = value\",\n \"def __setattr__(self, name, value):\\n self.set(**{name: value})\",\n \"def __set__(self, instance, value):\\n if value is None and not self.allow_none:\\n raise TypeError('None is not a valid value for this property.')\\n\\n if value is not None and not isinstance(value, self.valid_types):\\n raise TypeError('%s (%r) is not a valid type for this property.'\\n % (type(value).__name__, value))\\n\\n attr_name = self.get_attr_name(instance)\\n instance.__dict__['_%s_typed' % attr_name] = value\",\n \"def set_property(self, client, key, value):\\r\\n client.setProperty(key, value)\\r\\n return True\",\n \"def set_property(self, key, value):\\n self.properties[key] = value\\n return self.properties[key]\",\n \"def __set__(self, model_instance, value):\\r\\n raise ValueError, 'Virtual property is read-only'\",\n \"def _setPropertyValue(self, name, value, typeString = ''):\\n method = getattr(self.__class__, \\\"_setPropertyValue\\\" + getTypeString(value))\\n return method(self, name, value, typeString)\",\n \"def set_value (self):\\n raise NotImplementedError\",\n \"def __set__(self, instance, val):\\n raise AttributeError(\\\"Can't set attribute\\\")\",\n \"def _setitem_impl(self, index, obj, value):\\n obj.set_value(value)\\n return obj\",\n \"def __setattr__(self, key, value):\\n return setattr(self.__dict__['_obj'], key, value)\",\n \"def __setattr__ (self, attr, value):\\n self.set_value (attr, value)\",\n \"def XPSetWidgetProperty(inWidget, inProperty, inValue):\\n pass\",\n \"def put_prop(self, obj_type, obj_id, prop_name, value):\\n ierr = exolib.py_expp(self.exoid, obj_type, obj_id, prop_name, value)\\n if ierr:\\n raise ExodusIIWriterError(\\\"Error putting prop value\\\")\",\n \"def setValue(self, name: unicode, value: object) -> None:\\n ...\",\n \"def set_property(self, key, value):\\n if not key:\\n raise Exception('A key should be provided.')\\n if not value:\\n raise Exception('A value should be provided.')\\n self._properties[key] = value\",\n \"def setProperty(*args):\",\n \"def setProperty(*args):\",\n \"def setProperty(*args):\",\n \"def setProperty(*args):\",\n \"def setProperty(*args):\",\n \"def set(self, attr, val):\\r\\n self.__dict__[attr] = val\",\n \"def set(self, prop, value):\\r\\n\\r\\n prop_parts = prop.split(\\\".\\\")\\r\\n if self.copy_dict:\\r\\n new_dict = copy.deepcopy(self.obj)\\r\\n else:\\r\\n new_dict = self.obj\\r\\n pointer = None\\r\\n parts_length = len(prop_parts) - 1\\r\\n for i, part in enumerate(prop_parts):\\r\\n if pointer is None and i == parts_length:\\r\\n new_dict[part] = value\\r\\n elif pointer is None:\\r\\n pointer = new_dict.get(part)\\r\\n elif i == parts_length:\\r\\n pointer[part] = value\\r\\n else:\\r\\n pointer = pointer.get(part)\\r\\n return new_dict\",\n \"def setProperty(self, path, key, value):\\n \\n try:\\n self._client.propset(key, value, self._workingCopyPath + path)\\n self.checkin(path)\\n except ClientError, error:\\n raise SubversionError(error)\",\n \"def run_setprop(self, expanded, unexpanded) :\\n\\t\\toptions, args = self.getopt([\\\"name=\\\", \\\"value=\\\"], expanded)\\n\\t\\tif (options is None) and (args is None) :\\n\\t\\t\\treturn -1\\t# message was already displayed in self.getopt()\\n\\t\\tif not args :\\n\\t\\t\\treturn self.errormessage(\\\"Needs at least one object to change this property\\\")\\n\\t\\tif not (options.has_key(\\\"name\\\") and options.has_key(\\\"value\\\")) :\\n\\t\\t\\treturn self.errormessage(\\\"You must supply a property name and value\\\")\\n\\t\\tproperty = options[\\\"name\\\"]\\n\\t\\tif options.has_key(\\\"value\\\") :\\n\\t\\t\\tpropvalue = options[\\\"value\\\"]\\n\\t\\t\\ttry :\\n\\t\\t\\t\\t# maybe it's a list in a string, e.g. \\\"['e', 'f']\\\"\\n\\t\\t\\t\\t# or something like that\\n\\t\\t\\t\\tnewvalue = eval(propvalue)\\n\\t\\t\\t\\tif (type(newvalue) != type(0)) and (type(newvalue) != type(0.0)) :\\n\\t\\t\\t\\t\\t# we mustn't convert numeric to string\\n\\t\\t\\t\\t\\tpropvalue = newvalue\\n\\t\\t\\texcept NameError :\\n\\t\\t\\t\\tpass\\t# normal string\\n\\t\\telse :\\n\\t\\t\\tpropvalue = \\\"\\\"\\n\\n\\t\\tstatus = 0\\n\\t\\tfor arg in args :\\n\\t\\t\\tobject = self.toObject(self.__context, arg)\\n\\t\\t\\tif object is not None :\\n\\t\\t\\t\\tif not self.HasPerms(object, 'Manage properties') :\\n\\t\\t\\t\\t\\tstatus = status - 1\\n\\t\\t\\t\\telif hasattr(object, 'hasProperty') :\\n\\t\\t\\t\\t\\tif not object.hasProperty(property) :\\n\\t\\t\\t\\t\\t\\tstatus = status + self.errormessage(\\\"Object %s has no property %s\\\" % (self.ObjectPath(object), property))\\n\\t\\t\\t\\t\\telse :\\n\\t\\t\\t\\t\\t\\t# in the following lines the absence of a _updateProperty\\n\\t\\t\\t\\t\\t\\t# attribute indicates an object without properties (e.g. a method)\\n\\t\\t\\t\\t\\t\\t# which indicates an object for which setting properties is a nonsense\\n\\t\\t\\t\\t\\t\\tif hasattr(object, \\\"_updateProperty\\\") :\\n\\t\\t\\t\\t\\t\\t\\tobject._updateProperty(property, propvalue)\\n\\t\\t\\t\\t\\t\\t\\tself.htmlmessage(\\\"Object %s property %s was modified to %s\\\" % (self.ObjectPath(object), property, str(propvalue)))\\n\\t\\treturn status\",\n \"def set_prop(prop, value, config_type=\\\"\\\", config_path=\\\"\\\"):\\n\\n path = get_config_path(config_type, config_path)\\n set_property_value(prop, value, path)\",\n \"def set_value(self, _, prop_value_pair):\\n (prop, val) = prop_value_pair\\n\\n _, _, obj = self.popup_data\\n if not isinstance(obj, value_model.Value):\\n raise TypeError(\\\"Expected %s\\\" % type(value_model.Value))\\n\\n if not prop == obj.parent:\\n raise ValueError(\\\"Property '%s' is not the parent of '%s'\\\" % (prop, obj))\\n\\n # To enable undo redo for this we need a bit of trickery\\n new_prop = prop.clone(keep_id=True)\\n create_pseudo_values([new_prop])\\n if new_prop.pseudo_values[obj.index].pseudo_values != obj.pseudo_values:\\n raise ValueError(\\\"Cannot find replacement value\\\")\\n\\n # Update the value in the new property\\n new_prop.pseudo_values[obj.index].pseudo_values = val\\n\\n # Lets replace the old property with the new and updated one\\n cmd = commands.ReplaceObject(obj=prop, repl=new_prop)\\n self.execute(cmd)\\n\\n # Reset the view to make sure the changes are properly displayed.\\n self.select_object(new_prop)\\n self.reset_value_view(None)\",\n \"def __setattr__(self, attr, value):\\n field = getattr(self, attr)\\n if isinstance(field, BaseType):\\n # Check the value type\\n check = field.accept_value(value)\\n \\n old_value = getattr(self, attr)\\n object.__setattr__(self, attr, value)\\n if isinstance(old_value, BaseType):\\n # Not set yet\\n old_value = None\\n \\n if Model.data_connector and Model.data_connector.running:\\n with Model.data_connector.u_lock:\\n Model.data_connector.update_object(self, attr, old_value)\",\n \"def value(self, value):\\n\\n\\t\\tself.__value = value\",\n \"def changeProperty(self, node, name, propertyName, value, setIfNotExist=False):\",\n \"def __set__(self, obj, value):\\r\\n if value != None and (obj.id == value or (hasattr(value, \\\"id\\\") and obj.id == value.id)):\\r\\n raise ValueError, \\\"Can not associate an object with itself!\\\"\\r\\n return super(ReferenceProperty, self).__set__(obj,value)\",\n \"def set(self, database_name, property_, value):\\n return isempty(self._send_command(database_name, \\\"%s=%s\\\" % (property_, value)))\",\n \"def set_value(self,x):\\n self._value = x\",\n \"def set_value(self,x):\\n self._value = x\",\n \"def proxy_set(self, value):\\n setter = getattr(self, self.proxy_setter)\\n return setter(value)\",\n \"def setitem(obj, attr, value):\\n from functools import partial\\n _setattr = partial(setattr, obj)\\n setter = getattr(obj, '__setitem__', None) or _setattr\\n setter(attr, value)\",\n \"def set(self, properties):\\n raise NotImplementedError\",\n \"def setValue(self, value):\\n self._value = value\",\n \"def set(self, obj, success=True):\\n self.value = obj\\n self.success = success\\n if self.callback and success:\\n self.callback(obj)\\n if self.errback and not success:\\n self.errback(obj)\\n with self.condition:\\n self.ready = True\\n self.condition.notify()\",\n \"def setter(self, func):\\n self.fset = func\\n self.set_setattr_mode(SetAttr.CallObject_ObjectValue, func)\\n return self\",\n \"def set_value(self, value):\\n self.value = value\",\n \"def set_value(self, value):\\n self.value = value\",\n \"def set_value(self, value):\\n self.value = value\",\n \"def _platformix_set(self, context, fake_reply, prop, value):\\r\\n if hasattr(self.host, prop):\\r\\n if not callable(getattr(self.host, prop)):\\r\\n try:\\r\\n setattr(self.host, prop, value)\\r\\n except Exception as e:\\r\\n eprint(\\\"Platformix protocol: failed to set attribute {} of {} to value {} \\\"\\r\\n \\\"due to exception {}\\\".format(prop, self.host.name, value, e))\\r\\n exprint()\\r\\n self._reply(context, proto_failure(\\r\\n \\\"Failed to set attribute {} of {} to value {} \\\"\\r\\n \\\"due to exception {}\\\".format(prop, self.host.name, value, e)), fake_reply)\\r\\n return\\r\\n self._reply(context, proto_success(getattr(self.host, prop), prop), fake_reply)\\r\\n else:\\r\\n self._reply(context, proto_failure(\\\"Attribute {} of {} is a method\\\".format(\\r\\n prop, self.host.name)), fake_reply)\\r\\n else:\\r\\n self._reply(context, proto_failure(\\\"Property {} not found on {}\\\".format(prop, self.host.name)), fake_reply)\",\n \"def set(self, name, value):\\n pass\",\n \"def set_property(wellorcontainer, property_name, value):\\n wells = convert_to_wellgroup(wellorcontainer)\\n \\n if not isinstance(value, str):\\n value = str(value)\\n \\n for well in wells:\\n assert isinstance(well, Well)\\n well.properties[property_name] = value\",\n \"def setDynamicProperty(self, widget, property_name, property_value):\\n\\n\\t\\twidget.setProperty(property_name, property_value)\\n\\t\\twidget.style().unpolish(widget)\\n\\t\\twidget.style().polish(widget)\",\n \"def set_value(self, val):\\n self.value = val\",\n \"def setProperty(self, newProp, objectID=0):\\n\\n # Set the new property to container\\n key = (newProp.getPropertyID(),\\n newProp.getObjectID(),\\n newProp.getTime())\\n self.properties.set_value(key, newProp)\",\n \"def setProperty(self, child, key, value):\\n\\n # First get the child's dictionary\\n childDict = self.getInfoDict(child)\\n if childDict:\\n childDict[key] = value\",\n \"def put_prop(self, obj_type, obj_id, prop_name, value):\\n ierr = exolib.py_expp(self.exoid, obj_type, obj_id, prop_name, value)\\n if ierr:\\n raise ExodusIIWriterError(\\\"Error putting prop\\\")\",\n \"def set_property_value(self, current_value, new_value):\\n if self._enabled:\\n current_value *= self.marker_factor\\n new_value *= self.marker_factor\\n if live_object_is_valid(self._parent) and current_value != new_value:\\n new_param_value = max(self._absolute_range[0] * self.marker_factor, min(self._absolute_range[1] * self.marker_factor, new_value))\\n setattr(self._parent, self._property_name, new_param_value)\\n if self._should_zoom:\\n self._parent.view.show_loop()\\n self._parent.view.show_loop()\",\n \"def set_value(self, val):\\n self._value = val\",\n \"def set_value(self, val):\\n self._value = val\",\n \"def _set_value(o, d):\\n if isinstance(o, Param) and not o._mutable:\\n return # ignore requests to set immutable params\\n else:\\n try:\\n o.value = d\\n except AttributeError:\\n o = d # this would be an indexed parameter\",\n \"def set_property(self, vtk_property):\\n self.vtk_act.SetProperty(vtk_property)\",\n \"def set(self, attr_name, value):\\n\\n get_logger().debug('Going to set XML attribute property: %s', attr_name)\\n\\n if self.textproperty is not None:\\n # TODO: potentially programming error\\n raise MarshallingError()\\n\\n try:\\n # pylint: disable=unnecessary-dunder-call\\n self.attributes[attr_name].__set__(self.obj, value)\\n get_logger().debug('Set XML attribute property: %s', attr_name)\\n except AttributeError as ex:\\n get_logger().error('XML property %s: %s', attr_name, str(ex))\\n except KeyError:\\n get_logger().debug('Not an XML attribute property: %s', attr_name)\",\n \"def SetPropObject(self, obj):\\n\\t\\tif self.propObj is not None:\\n\\t\\t\\t# deregister notification\\n\\t\\t\\tif hasattr(self.propObj, \\\"DropPropsNotify\\\") :\\n\\t\\t\\t\\tself.propObj.DropPropsNotify()\\n\\t\\tif hasattr(obj, \\\"GetPropList\\\") :\\n\\t\\t\\tself.propObj = obj\\n\\t\\t\\tself.UpdateGrid()\\n\\t\\telse :\\n\\t\\t\\twx.MessageBox(_(\\\"Can't display properties for %s, because it hasn't GetPropList interface\\\") % obj.__class__.__name__,\\n\\t\\t\\t\\t\\t\\t\\t_(\\\"Sampo Framework\\\"), wx.OK | wx.CENTER | wx.ICON_ERROR)\",\n \"def set(self, attribute, value):\\n self.__setattr__(attribute, value)\",\n \"def autoprops_generated_setter(self, **kwargs):\\n setattr(self, private_property_name, kwargs[property_name])\",\n \"def __set__(self, instance, value):\\n if self.deprecated:\\n cls_name = type(instance).__name__\\n warnings.warn('%s.%s is deprecated. Please set %s.%s instead.'\\n % (cls_name, self.get_attr_name(instance),\\n cls_name, self.prop_name),\\n self.deprecation_warning,\\n stacklevel=2)\\n\\n if value is not None and self._convert_to_func is not None:\\n value = self._convert_to_func(value)\\n\\n setattr(instance, self.prop_name, value)\",\n \"def set_property(self, name, value, persist_changes=True):\\n self._properties_metadata[name] = {'readonly': not persist_changes}\\n self._properties[name] = value\",\n \"def setValue(self, *args):\\n return _libsbml.ConversionProperties_setValue(self, *args)\",\n \"def setter(self, value):\\n if field.readonly:\\n raise fields.ValidationError(f\\\"'{name}' is a read only attribute\\\")\\n\\n # accept if this is a raw value too\\n value = field.from_raw(value)\\n\\n # validate\\n field.validate(value)\\n\\n # set current instance as parent for embedded objects/instances\\n if isinstance(field, fields.Object):\\n value.parent = self\\n\\n # se attribute\\n setattr(self, inner_name, value)\\n self._attr_updated(name, value)\",\n \"def __setattr__(self, name, value):\\n if not hasattr(self, name):\\n raise AttributeError(\\\"'%s' object has no attribute '%s'\\\" % (type(self).__name__, name))\\n object.__setattr__(self, name, value)\",\n \"def setProperty(self,prop=None):\\n objects = self.check()\\n if objects:\\n if prop is None:\\n res = askItems([['property',0]],\\n caption = 'Set Property Number for Selection (negative value to remove)')\\n if res:\\n prop = int(res['property'])\\n if prop < 0:\\n prop = None\\n for o in objects:\\n if hasattr(o,'setProp'):\\n o.setProp(prop)\\n self.draw()\",\n \"def set_property_value(self, name, value, dry_run=False):\\n raise DAVError(\\n HTTP_FORBIDDEN, err_condition=PRECONDITION_CODE_ProtectedProperty\\n )\",\n \"def __set__( self, client, value ):\\n\\t\\tif value is None:\\n\\t\\t\\ttry:\\n\\t\\t\\t\\treturn self.__delete__( client )\\n\\t\\t\\texcept AttributeError, err:\\n\\t\\t\\t\\t# wasn't defined anyway\\n\\t\\t\\t\\treturn None\\n\\t\\telse:\\n\\t\\t\\treturn super( DBProperty, self ).__set__( client, value )\",\n \"def set_value(self, field, value):\\n field = self.find_first(field)\\n if field is not None:\\n field.value = value\",\n \"def __setattr__(self, attr, obj):\\n\\n # Underscored-prefixed/volatile/stateful slots can be set\\n # without affecting other such slots.\\n if (\\n attr not in self.__volatile_slots__\\n # Are we trying to set a custom property?\\n and attr in getattr(self, \\\"__all_props__\\\", ())\\n # Is it a custom property that's already been set?\\n and hasattr(self, attr)\\n # Are we setting it to a new value?\\n and getattr(self, attr) is not obj\\n ):\\n self.reset()\\n\\n object.__setattr__(self, attr, obj)\",\n \"def __setattr__(self, attr, value):\\n super().__setattr__(attr, value)\",\n \"def __set__(self, instance, value):\\n raise AttributeError(\\\"A Default Property is Read only\\\")\",\n \"def setval(self, val):\\r\\n self.value = val\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.80547476","0.77362704","0.76287323","0.76190275","0.75847745","0.7426509","0.73941714","0.72769517","0.726904","0.72673047","0.7146816","0.7140176","0.7101052","0.7087965","0.7087463","0.7041664","0.69987345","0.69987345","0.6997965","0.69783264","0.6958644","0.69510746","0.68757504","0.6826675","0.6824359","0.68217134","0.6810024","0.6785712","0.67760295","0.6765161","0.6757913","0.67286384","0.6720421","0.66904914","0.6688953","0.6673924","0.6664124","0.66613615","0.6634168","0.6631756","0.66218543","0.66186684","0.6591585","0.65896904","0.65896904","0.65896904","0.65896904","0.65896904","0.65649176","0.6561349","0.6557601","0.65527314","0.6517892","0.65148294","0.6514115","0.65105635","0.6509635","0.65052897","0.65040034","0.64915806","0.64915806","0.64890295","0.6473472","0.6458243","0.6454401","0.64459723","0.6445705","0.643957","0.643957","0.643957","0.6422582","0.64028317","0.63945955","0.6390556","0.6372176","0.637174","0.6368701","0.6362085","0.63572925","0.63499093","0.63499093","0.63458914","0.6344987","0.6339965","0.6334685","0.6330212","0.6328096","0.6315735","0.63153124","0.6314864","0.63137037","0.63119817","0.6306431","0.6306395","0.6303508","0.6298098","0.6295941","0.62907434","0.6289114","0.6285625"],"string":"[\n \"0.80547476\",\n \"0.77362704\",\n \"0.76287323\",\n \"0.76190275\",\n \"0.75847745\",\n \"0.7426509\",\n \"0.73941714\",\n \"0.72769517\",\n \"0.726904\",\n \"0.72673047\",\n \"0.7146816\",\n \"0.7140176\",\n \"0.7101052\",\n \"0.7087965\",\n \"0.7087463\",\n \"0.7041664\",\n \"0.69987345\",\n \"0.69987345\",\n \"0.6997965\",\n \"0.69783264\",\n \"0.6958644\",\n \"0.69510746\",\n \"0.68757504\",\n \"0.6826675\",\n \"0.6824359\",\n \"0.68217134\",\n \"0.6810024\",\n \"0.6785712\",\n \"0.67760295\",\n \"0.6765161\",\n \"0.6757913\",\n \"0.67286384\",\n \"0.6720421\",\n \"0.66904914\",\n \"0.6688953\",\n \"0.6673924\",\n \"0.6664124\",\n \"0.66613615\",\n \"0.6634168\",\n \"0.6631756\",\n \"0.66218543\",\n \"0.66186684\",\n \"0.6591585\",\n \"0.65896904\",\n \"0.65896904\",\n \"0.65896904\",\n \"0.65896904\",\n \"0.65896904\",\n \"0.65649176\",\n \"0.6561349\",\n \"0.6557601\",\n \"0.65527314\",\n \"0.6517892\",\n \"0.65148294\",\n \"0.6514115\",\n \"0.65105635\",\n \"0.6509635\",\n \"0.65052897\",\n \"0.65040034\",\n \"0.64915806\",\n \"0.64915806\",\n \"0.64890295\",\n \"0.6473472\",\n \"0.6458243\",\n \"0.6454401\",\n \"0.64459723\",\n \"0.6445705\",\n \"0.643957\",\n \"0.643957\",\n \"0.643957\",\n \"0.6422582\",\n \"0.64028317\",\n \"0.63945955\",\n \"0.6390556\",\n \"0.6372176\",\n \"0.637174\",\n \"0.6368701\",\n \"0.6362085\",\n \"0.63572925\",\n \"0.63499093\",\n \"0.63499093\",\n \"0.63458914\",\n \"0.6344987\",\n \"0.6339965\",\n \"0.6334685\",\n \"0.6330212\",\n \"0.6328096\",\n \"0.6315735\",\n \"0.63153124\",\n \"0.6314864\",\n \"0.63137037\",\n \"0.63119817\",\n \"0.6306431\",\n \"0.6306395\",\n \"0.6303508\",\n \"0.6298098\",\n \"0.6295941\",\n \"0.62907434\",\n \"0.6289114\",\n \"0.6285625\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.79016274"},"document_rank":{"kind":"string","value":"1"}}},{"rowIdx":90874,"cells":{"query":{"kind":"string","value":"Set `password_hash` only when `password` is not set."},"document":{"kind":"string","value":"def _create(cls, model_class, *args, **kwargs):\n assert 'password' in kwargs\n assert 'passwd_hash' in kwargs\n if kwargs.get('password'):\n kwargs.pop('passwd_hash')\n return super()._create(model_class, *args, **kwargs)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def set_password_hash(self, password):\n salt = bcrypt.gensalt()\n self.password_hash = bcrypt.hashpw(password.encode(), salt)","def hash_password(self, password):\n self.password_hash = generate_password_hash(password)","def set_password(self, password):\n self.password = self.hash_password(password)","def setPassword(self, password, hashed=False):\n if hashed or self.hashed:\n self.hashed = True\n self.password = utils.saltHash(password)\n else:\n self.password = password","def password(self, password):\n self.password_hash = generate_password_hash(password)\n self.password_set = True","def setPassword(self, unhashPass):\n\t\tself.passHash = generate_password_hash(unhashPass)","def set_password(self, password):\n self.password_hash = generate_password_hash(password)","def set_password(self, password):\n self.password_hash = generate_password_hash(password)","def password(self, password):\n if password is None:\n self._password = None\n else:\n self._password = generate_password_hash(password)","def set_password(self, password):\n self.password_hash = generate_password_hash(str(password))","def set_password(self, password):\n self.password_hash = generate_password_hash(f\"{password}{self.user_salt}\")","def hash_password(self, password):\n self.password = pwd_context.encrypt(password)","def password(self, password):\n self.password_hash = generate_password_hash(password)","def password(self, password):\n self.password_hash = generate_password_hash(password)","def password(self, password):\n self.password_hash = generate_password_hash(password)","def password(self, password):\n self.password_hash = generate_password_hash(password)","def set_password(self, password: Optional[str]) -> None:\n if password is None:\n self.set_unusable_password()\n return\n\n if not check_password_strength(password):\n raise PasswordTooWeakError\n\n super().set_password(password)","def password(self, password):\n self.password_hash = generate_password_hash(password)","def hash_password(self):\n self.__password = self.str_to_hash(self.__password)","def _set_password(self, password):\n self._password = generate_password_hash(password)","def password(self, password):\n\n self.password_hash = generate_password_hash(password)","def set_password(self, password):\n self.password = generate_password_hash(password)","def set_password(self, password):\n self.password = generate_password_hash(password, method='pbkdf2:sha256')","def check_password(self, password):\n if not self.password or not password:\n return False\n ok, new_hash = pwd_context.verify_and_update(password, self.password)\n if not ok:\n return False\n if new_hash:\n # hashed password was using a deprecated scheme, update it\n log.info(\"Updating user's password.\")\n self.password = new_hash\n self.save()\n return True","def set_password(self, password):\n\n self.password = bcrypt.generate_password_hash(password)","def hash_password(self, original_password):\n self.password = generate_password_hash(original_password)","def check_hash(self, password):\n return bcrypt.check_password_hash(self.password, password)","def _set_password(self, password):\r\n hashed_password = password\r\n\r\n if isinstance(password, unicode):\r\n password_8bit = password.encode('UTF-8')\r\n else:\r\n password_8bit = password\r\n\r\n # Hash a password for the first time, with a randomly-generated salt\r\n salt = bcrypt.gensalt(10)\r\n hashed_password = bcrypt.hashpw(password_8bit, salt)\r\n\r\n # Make sure the hased password is an UTF-8 object at the end of the\r\n # process because SQLAlchemy _wants_ a unicode object for Unicode\r\n # fields\r\n if not isinstance(hashed_password, unicode):\r\n hashed_password = hashed_password.decode('UTF-8')\r\n\r\n self._password = hashed_password","def verify_password(self, hash, password):\r\n try:\r\n PasswordHasher().verify(hash, password)\r\n return True\r\n except:\r\n return False","def set_password(self, user, password):\n hashed_password = self.hash_password(password)\n server_name = self.get_server_name()\n hookenv.log(\"Storing hash: {}\".format(hashed_password), hookenv.DEBUG)\n result = self.pgsql_query(\n \"UPDATE users SET password_hash = '{}' WHERE name = '@{}:{}';\".format(\n hashed_password, user, server_name\n )\n )\n return result","def check_password(self, password):\n\n if self.password is None:\n return False\n return check_password_hash(self.password, password)","def set_pass(self, pw):\n\t\tself.passhash = generate_password_hash(pw)","def test_password_setter(self):\n self.user.password = '123456'\n self.assertIsNotNone(self.user.password_hash)","def check_password(self, raw_password):\n\t\tdef setter(raw_password):\n\t\t\tself.set_password(raw_password)\n\t\t\t# Password hash upgrades shouldn't be considered password changes.\n\t\t\tself._password = None\n\t\t\tself.save(update_fields=[\"password\"])\n\t\treturn check_password(raw_password, self.password, setter)","def check_password(self, password):\n check = False\n if self.password is not None:\n logger.critical('Old style password exists.')\n if check_password_hash(self.password, password):\n self.passwd = password\n self.password = None\n db.session.add(self)\n db.session.commit()\n logger.critical('Old style password replaced.')\n else:\n return check\n try:\n check = bcrypt.check_password_hash(self._password.encode('utf8'), password.encode('utf8'))\n except:\n logger.critical('Error in password check.')\n finally:\n return check","def check_password(self, password=None):\n return check_password_hash(self.password, password)","def hash_password(self, password):\n new_password = ['*' for i in password if password is not None]\n return ''.join(new_password)","def clean(self):\n if self.email is None or self.password is None:\n raise ValidationError\n self.password = generate_password_hash(self.password)","def password_hash_iterations(self, password_hash_iterations):\n\n self._password_hash_iterations = password_hash_iterations","def update_password(self, user, password):\n user.password = hashers.make_password(password)","def set_password(self, password):\n self.password = password","def test_set_password_without_save(self):\n user = UserModel()\n user.password = \"is_god\"\n assert isinstance(user.password, HashedPassword), \"password should be hashed\"\n assert user.password == \"is_god\"","def set_password(self, password):\n self.authentication.password = password","def checkPassword(self, unhashPass):\n\t\treturn check_password_hash(self.passHash, unhashPass)","def password_hash_algorithm(self, password_hash_algorithm):\n\n self._password_hash_algorithm = password_hash_algorithm","def check_password(self, password):\n return check_password_hash(self.password, password)","def check_password(self, password):\n return check_password_hash(self.password_hash, f\"{password}{self.user_salt}\")","def check_password(self, password):\n return check_password_hash(self.password_hash, password)","def check_password(self, password):\n return check_password_hash(self.password_hash, password)","def check_password(self, password):\n return check_password_hash(self.password_hash, password)","def check_password(self, password):\n return check_password_hash(self.password_hash, password)","async def hash_password(self, password: str | None) -> str | None:\n if password is not None:\n fut = self._hash_password(password)\n return await self.cpu_subsystem.execute(fut)\n return None","def setpassword(self, pwd):\n pass","def set_password(self, raw_password):\n self.password = security.generate_password_hash(raw_password, length=12)","def set_password(self, raw_password):\n self.password = security.generate_password_hash(raw_password, length=12)","def _validate_password(self, user, password):\n valid, new_hash = self.pwd_context.verify_and_update(str(password), user['users_auth'])\n if not valid:\n logger.error('Invalid password for %(users_name)s', user)\n raise YKAuthError(\"Invalid password for %(users_name)s\" % user)\n if new_hash:\n # TODO: update user's hash with new_hash\n logger.warning(\"User %(users_name)s's hash needs update\", user)\n return True","def verify_password(self, password):\n if (self.password_set and\n check_password_hash(self.password_hash, password)):\n self.failed_logins = 0\n return True\n else:\n self.failed_logins += 1\n if self.failed_logins > 2:\n self.blocked = True\n db.session.commit()\n return False","def check_password(self, password):\n if not self._password or not password:\n return False\n return check_password_hash(self._password, password)","def get_password_hash(password):\n if not Utility.check_empty_string(password):\n return Utility.pwd_context.hash(password)","def verify_password(self, password):\n self.password_hash = generate_password_hash(password)\n\n return check_password_hash(self.password_hash, password)","def _hash_password(self, password):\n passwordhash = bcrypt.hashpw(password.encode('utf8'), bcrypt.gensalt())\n return passwordhash","def set_password(self, password):\n self.password = md5crypt(password, gen_salt())","def password(self, password) :\n\t\ttry :\n\t\t\tself._password = password\n\t\texcept Exception as e:\n\t\t\traise e","def valid_login_password(username, password, pw_hash):\n salt = pw_hash.split(',')[1]\n return pw_hash == make_pw_hash(username, password, salt)","def verify_password(self, password):\n return check_password_hash(self.password_hash, password)","def check_password(hashed_password, password):\n return check_password_hash(hashed_password, password)","def check_password(self, password):\n\n return check_password_hash(self.password_hash, password)","def validate_login(password_hash, password):\n print(password_hash, password)\n return check_password_hash(password_hash, password)","def validate_password(self, password):\n return Bcrypt().check_password_hash(self.password, password)","def set_password(self, password):\n self.PASSWORD = password","def has_password(self, password):\n return self.password_hash == util.password_hash(password)","def password(self, password):\n if password is not None and len(password) > 30:\n raise ValueError(\"Invalid value for `password`, length must be less than or equal to `30`\")\n\n self._password = password","def _check_password(self, body):\n if not self.config.security_initialize:\n raise RuntimeError(\"First set a password\")\n\n password = hash_password(body[ATTR_PASSWORD])\n if password != self.config.security_password:\n raise RuntimeError(\"Wrong password\")","def check_password(self, password):\n\n return bcrypt.check_password_hash(self.password, password)","def verify_password(self, password, password_hash):\n\n # Print deprecation warning if called with (password, user) instead of (password, user.password)\n if isinstance(password_hash, self.user_manager.db_manager.UserClass):\n print(\n 'Deprecation warning: verify_password(password, user) has been changed'\\\n ' to: verify_password(password, password_hash). The user param will be deprecated.'\\\n ' Please change your call with verify_password(password, user) into'\\\n ' a call with verify_password(password, user.password)'\n ' as soon as possible.')\n password_hash = password_hash.password # effectively user.password\n\n # Use passlib's CryptContext to verify a password\n return self.password_crypt_context.verify(password, password_hash)","def update_password(self, username, password): #WORKS\n password_hash = generate_password_hash(password)\n try:\n self.cur.execute(\"UPDATE users SET password = \\\"{}\\\" WHERE username = \\\"{}\\\"\".format(password_hash, username))\n self.db.commit()\n except:\n self.db.rollback()","def passsword(self, password):\n self.passwor_harsh = generate_password_hash(password)","def check_password(self, raw_password):\n def setter(raw_password):\n self.set_password(raw_password)\n self.save(update_fields=[\"password\"])\n return check_password(raw_password, self.password, setter)","def set_password(self, raw_password):\n if raw_password is None:\n self.set_unusable_password()\n else:\n import random\n algo = PASSWORD_ALGO\n salt = get_hexdigest(algo, str(random.random()), str(random.random()))[:5]\n hsh = get_hexdigest(algo, salt, raw_password)\n self.password = '%s$%s$%s' % (algo, salt, hsh)","def update_password(self, pwd):\n self.password = bcrypt.generate_password_hash(pwd).decode('utf8')","def test_that_an_incorrect_password_returns_false(new_user):\n user, user_data = new_user\n password = 'x{}x'.format(user_data.get('password'))\n\n assert_that(user.verify_password(password)).is_false()","def test_no_forced_password_change(self):\r\n\r\n email, password = self._setup_user()\r\n self._login(email, password)\r\n\r\n email, password = self._setup_user(is_staff=True)\r\n self._login(email, password)","def checkPassword(self, password):\n if password is None:\n return False\n if self.hashed:\n (salt, _) = self.password.split('|')\n return (self.password == utils.saltHash(password, salt=salt))\n else:\n return (self.password == password)","def check_hashed_password(self, input_password):\n return check_password_hash(self.password, input_password)","def verify_password(password, password_hash):\n\n password_entered = hash_password(password)\n return secrets.compare_digest(password_entered, password_hash)","def valid_password(self, password):\n return check_hash(password, self.password)","def update_password(self, new_password=None):\n\n self.password = generate_password_hash(new_password)\n\n if self.save(verbose=False):\n self.logger.warn('Updated password! %s' % self)\n else:\n raise AttributeError('Password update failed!')","def __generate_hash(password):\n if password is None:\n return None\n return bcrypt.generate_password_hash(password, rounds=10).decode(\"utf8\")","def verify_password(password, hash):\n\n method = get_hash_algorithm(flask.current_app.config['HASH_ALGORITHM'])\n\n return method.verify(password, hash)","def check_password(self, raw_password):\n def setter(raw_password):\n self.set_password(raw_password)\n self.save(update_fields=[\"password\"])\n return check_password(raw_password, self.password, setter)","def test_set_password(matrix, mock_check_output, mock_psycopg2):\n matrix.save_pgsql_conf(db)\n matrix.set_password(\"testuser\", \"testpassword\")\n assert mock_check_output.called_with(\n [\n \"snap\",\n \"run\",\n \"{}.hash_password\".format(matrix.synapse_snap),\n \"-c\",\n matrix.synapse_config,\n \"-p\",\n \"testpassword\",\n ]\n )\n assert mock_check_output.call_count == 1\n assert mock_psycopg2.Cursor.execute.called_with(\n \"UPDATE users SET password_hash='testhash' WHERE name='testuser';\"\n )\n assert mock_psycopg2.Cursor.execute.call_count == 1","def set_password(self, password):\n self.__init__(password=password)","def set_password(self, password):\n self.PASS = password","def set_password(self, raw_password: str):\n self.new_password = raw_password","def update_password(self, password):\n self.password = scryptsalsa208sha256_str(password.encode('utf-8')).decode('utf-8')\n return True","def hash_passwd(password, hash_method=\"sha256\"):\n\n return generate_password_hash(password, hash_method)","def _authenticate_password(user_input: str, hash_: str) -> bool:\n return pbkdf2_sha256.verify(user_input, hash_)","def has_password(self, has_password):\n if has_password is None:\n raise ValueError(\"Invalid value for `has_password`, must not be `None`\") # noqa: E501\n\n self._has_password = has_password","def verify_password(self, password):\n return check_password_hash(self.password_hash, password)","def verify_password(self, password):\n return check_password_hash(self.password_hash, password)","def verify_password(self, password):\n return check_password_hash(self.password_hash, password)"],"string":"[\n \"def set_password_hash(self, password):\\n salt = bcrypt.gensalt()\\n self.password_hash = bcrypt.hashpw(password.encode(), salt)\",\n \"def hash_password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def set_password(self, password):\\n self.password = self.hash_password(password)\",\n \"def setPassword(self, password, hashed=False):\\n if hashed or self.hashed:\\n self.hashed = True\\n self.password = utils.saltHash(password)\\n else:\\n self.password = password\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\\n self.password_set = True\",\n \"def setPassword(self, unhashPass):\\n\\t\\tself.passHash = generate_password_hash(unhashPass)\",\n \"def set_password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def set_password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def password(self, password):\\n if password is None:\\n self._password = None\\n else:\\n self._password = generate_password_hash(password)\",\n \"def set_password(self, password):\\n self.password_hash = generate_password_hash(str(password))\",\n \"def set_password(self, password):\\n self.password_hash = generate_password_hash(f\\\"{password}{self.user_salt}\\\")\",\n \"def hash_password(self, password):\\n self.password = pwd_context.encrypt(password)\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def set_password(self, password: Optional[str]) -> None:\\n if password is None:\\n self.set_unusable_password()\\n return\\n\\n if not check_password_strength(password):\\n raise PasswordTooWeakError\\n\\n super().set_password(password)\",\n \"def password(self, password):\\n self.password_hash = generate_password_hash(password)\",\n \"def hash_password(self):\\n self.__password = self.str_to_hash(self.__password)\",\n \"def _set_password(self, password):\\n self._password = generate_password_hash(password)\",\n \"def password(self, password):\\n\\n self.password_hash = generate_password_hash(password)\",\n \"def set_password(self, password):\\n self.password = generate_password_hash(password)\",\n \"def set_password(self, password):\\n self.password = generate_password_hash(password, method='pbkdf2:sha256')\",\n \"def check_password(self, password):\\n if not self.password or not password:\\n return False\\n ok, new_hash = pwd_context.verify_and_update(password, self.password)\\n if not ok:\\n return False\\n if new_hash:\\n # hashed password was using a deprecated scheme, update it\\n log.info(\\\"Updating user's password.\\\")\\n self.password = new_hash\\n self.save()\\n return True\",\n \"def set_password(self, password):\\n\\n self.password = bcrypt.generate_password_hash(password)\",\n \"def hash_password(self, original_password):\\n self.password = generate_password_hash(original_password)\",\n \"def check_hash(self, password):\\n return bcrypt.check_password_hash(self.password, password)\",\n \"def _set_password(self, password):\\r\\n hashed_password = password\\r\\n\\r\\n if isinstance(password, unicode):\\r\\n password_8bit = password.encode('UTF-8')\\r\\n else:\\r\\n password_8bit = password\\r\\n\\r\\n # Hash a password for the first time, with a randomly-generated salt\\r\\n salt = bcrypt.gensalt(10)\\r\\n hashed_password = bcrypt.hashpw(password_8bit, salt)\\r\\n\\r\\n # Make sure the hased password is an UTF-8 object at the end of the\\r\\n # process because SQLAlchemy _wants_ a unicode object for Unicode\\r\\n # fields\\r\\n if not isinstance(hashed_password, unicode):\\r\\n hashed_password = hashed_password.decode('UTF-8')\\r\\n\\r\\n self._password = hashed_password\",\n \"def verify_password(self, hash, password):\\r\\n try:\\r\\n PasswordHasher().verify(hash, password)\\r\\n return True\\r\\n except:\\r\\n return False\",\n \"def set_password(self, user, password):\\n hashed_password = self.hash_password(password)\\n server_name = self.get_server_name()\\n hookenv.log(\\\"Storing hash: {}\\\".format(hashed_password), hookenv.DEBUG)\\n result = self.pgsql_query(\\n \\\"UPDATE users SET password_hash = '{}' WHERE name = '@{}:{}';\\\".format(\\n hashed_password, user, server_name\\n )\\n )\\n return result\",\n \"def check_password(self, password):\\n\\n if self.password is None:\\n return False\\n return check_password_hash(self.password, password)\",\n \"def set_pass(self, pw):\\n\\t\\tself.passhash = generate_password_hash(pw)\",\n \"def test_password_setter(self):\\n self.user.password = '123456'\\n self.assertIsNotNone(self.user.password_hash)\",\n \"def check_password(self, raw_password):\\n\\t\\tdef setter(raw_password):\\n\\t\\t\\tself.set_password(raw_password)\\n\\t\\t\\t# Password hash upgrades shouldn't be considered password changes.\\n\\t\\t\\tself._password = None\\n\\t\\t\\tself.save(update_fields=[\\\"password\\\"])\\n\\t\\treturn check_password(raw_password, self.password, setter)\",\n \"def check_password(self, password):\\n check = False\\n if self.password is not None:\\n logger.critical('Old style password exists.')\\n if check_password_hash(self.password, password):\\n self.passwd = password\\n self.password = None\\n db.session.add(self)\\n db.session.commit()\\n logger.critical('Old style password replaced.')\\n else:\\n return check\\n try:\\n check = bcrypt.check_password_hash(self._password.encode('utf8'), password.encode('utf8'))\\n except:\\n logger.critical('Error in password check.')\\n finally:\\n return check\",\n \"def check_password(self, password=None):\\n return check_password_hash(self.password, password)\",\n \"def hash_password(self, password):\\n new_password = ['*' for i in password if password is not None]\\n return ''.join(new_password)\",\n \"def clean(self):\\n if self.email is None or self.password is None:\\n raise ValidationError\\n self.password = generate_password_hash(self.password)\",\n \"def password_hash_iterations(self, password_hash_iterations):\\n\\n self._password_hash_iterations = password_hash_iterations\",\n \"def update_password(self, user, password):\\n user.password = hashers.make_password(password)\",\n \"def set_password(self, password):\\n self.password = password\",\n \"def test_set_password_without_save(self):\\n user = UserModel()\\n user.password = \\\"is_god\\\"\\n assert isinstance(user.password, HashedPassword), \\\"password should be hashed\\\"\\n assert user.password == \\\"is_god\\\"\",\n \"def set_password(self, password):\\n self.authentication.password = password\",\n \"def checkPassword(self, unhashPass):\\n\\t\\treturn check_password_hash(self.passHash, unhashPass)\",\n \"def password_hash_algorithm(self, password_hash_algorithm):\\n\\n self._password_hash_algorithm = password_hash_algorithm\",\n \"def check_password(self, password):\\n return check_password_hash(self.password, password)\",\n \"def check_password(self, password):\\n return check_password_hash(self.password_hash, f\\\"{password}{self.user_salt}\\\")\",\n \"def check_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def check_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def check_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def check_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"async def hash_password(self, password: str | None) -> str | None:\\n if password is not None:\\n fut = self._hash_password(password)\\n return await self.cpu_subsystem.execute(fut)\\n return None\",\n \"def setpassword(self, pwd):\\n pass\",\n \"def set_password(self, raw_password):\\n self.password = security.generate_password_hash(raw_password, length=12)\",\n \"def set_password(self, raw_password):\\n self.password = security.generate_password_hash(raw_password, length=12)\",\n \"def _validate_password(self, user, password):\\n valid, new_hash = self.pwd_context.verify_and_update(str(password), user['users_auth'])\\n if not valid:\\n logger.error('Invalid password for %(users_name)s', user)\\n raise YKAuthError(\\\"Invalid password for %(users_name)s\\\" % user)\\n if new_hash:\\n # TODO: update user's hash with new_hash\\n logger.warning(\\\"User %(users_name)s's hash needs update\\\", user)\\n return True\",\n \"def verify_password(self, password):\\n if (self.password_set and\\n check_password_hash(self.password_hash, password)):\\n self.failed_logins = 0\\n return True\\n else:\\n self.failed_logins += 1\\n if self.failed_logins > 2:\\n self.blocked = True\\n db.session.commit()\\n return False\",\n \"def check_password(self, password):\\n if not self._password or not password:\\n return False\\n return check_password_hash(self._password, password)\",\n \"def get_password_hash(password):\\n if not Utility.check_empty_string(password):\\n return Utility.pwd_context.hash(password)\",\n \"def verify_password(self, password):\\n self.password_hash = generate_password_hash(password)\\n\\n return check_password_hash(self.password_hash, password)\",\n \"def _hash_password(self, password):\\n passwordhash = bcrypt.hashpw(password.encode('utf8'), bcrypt.gensalt())\\n return passwordhash\",\n \"def set_password(self, password):\\n self.password = md5crypt(password, gen_salt())\",\n \"def password(self, password) :\\n\\t\\ttry :\\n\\t\\t\\tself._password = password\\n\\t\\texcept Exception as e:\\n\\t\\t\\traise e\",\n \"def valid_login_password(username, password, pw_hash):\\n salt = pw_hash.split(',')[1]\\n return pw_hash == make_pw_hash(username, password, salt)\",\n \"def verify_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def check_password(hashed_password, password):\\n return check_password_hash(hashed_password, password)\",\n \"def check_password(self, password):\\n\\n return check_password_hash(self.password_hash, password)\",\n \"def validate_login(password_hash, password):\\n print(password_hash, password)\\n return check_password_hash(password_hash, password)\",\n \"def validate_password(self, password):\\n return Bcrypt().check_password_hash(self.password, password)\",\n \"def set_password(self, password):\\n self.PASSWORD = password\",\n \"def has_password(self, password):\\n return self.password_hash == util.password_hash(password)\",\n \"def password(self, password):\\n if password is not None and len(password) > 30:\\n raise ValueError(\\\"Invalid value for `password`, length must be less than or equal to `30`\\\")\\n\\n self._password = password\",\n \"def _check_password(self, body):\\n if not self.config.security_initialize:\\n raise RuntimeError(\\\"First set a password\\\")\\n\\n password = hash_password(body[ATTR_PASSWORD])\\n if password != self.config.security_password:\\n raise RuntimeError(\\\"Wrong password\\\")\",\n \"def check_password(self, password):\\n\\n return bcrypt.check_password_hash(self.password, password)\",\n \"def verify_password(self, password, password_hash):\\n\\n # Print deprecation warning if called with (password, user) instead of (password, user.password)\\n if isinstance(password_hash, self.user_manager.db_manager.UserClass):\\n print(\\n 'Deprecation warning: verify_password(password, user) has been changed'\\\\\\n ' to: verify_password(password, password_hash). The user param will be deprecated.'\\\\\\n ' Please change your call with verify_password(password, user) into'\\\\\\n ' a call with verify_password(password, user.password)'\\n ' as soon as possible.')\\n password_hash = password_hash.password # effectively user.password\\n\\n # Use passlib's CryptContext to verify a password\\n return self.password_crypt_context.verify(password, password_hash)\",\n \"def update_password(self, username, password): #WORKS\\n password_hash = generate_password_hash(password)\\n try:\\n self.cur.execute(\\\"UPDATE users SET password = \\\\\\\"{}\\\\\\\" WHERE username = \\\\\\\"{}\\\\\\\"\\\".format(password_hash, username))\\n self.db.commit()\\n except:\\n self.db.rollback()\",\n \"def passsword(self, password):\\n self.passwor_harsh = generate_password_hash(password)\",\n \"def check_password(self, raw_password):\\n def setter(raw_password):\\n self.set_password(raw_password)\\n self.save(update_fields=[\\\"password\\\"])\\n return check_password(raw_password, self.password, setter)\",\n \"def set_password(self, raw_password):\\n if raw_password is None:\\n self.set_unusable_password()\\n else:\\n import random\\n algo = PASSWORD_ALGO\\n salt = get_hexdigest(algo, str(random.random()), str(random.random()))[:5]\\n hsh = get_hexdigest(algo, salt, raw_password)\\n self.password = '%s$%s$%s' % (algo, salt, hsh)\",\n \"def update_password(self, pwd):\\n self.password = bcrypt.generate_password_hash(pwd).decode('utf8')\",\n \"def test_that_an_incorrect_password_returns_false(new_user):\\n user, user_data = new_user\\n password = 'x{}x'.format(user_data.get('password'))\\n\\n assert_that(user.verify_password(password)).is_false()\",\n \"def test_no_forced_password_change(self):\\r\\n\\r\\n email, password = self._setup_user()\\r\\n self._login(email, password)\\r\\n\\r\\n email, password = self._setup_user(is_staff=True)\\r\\n self._login(email, password)\",\n \"def checkPassword(self, password):\\n if password is None:\\n return False\\n if self.hashed:\\n (salt, _) = self.password.split('|')\\n return (self.password == utils.saltHash(password, salt=salt))\\n else:\\n return (self.password == password)\",\n \"def check_hashed_password(self, input_password):\\n return check_password_hash(self.password, input_password)\",\n \"def verify_password(password, password_hash):\\n\\n password_entered = hash_password(password)\\n return secrets.compare_digest(password_entered, password_hash)\",\n \"def valid_password(self, password):\\n return check_hash(password, self.password)\",\n \"def update_password(self, new_password=None):\\n\\n self.password = generate_password_hash(new_password)\\n\\n if self.save(verbose=False):\\n self.logger.warn('Updated password! %s' % self)\\n else:\\n raise AttributeError('Password update failed!')\",\n \"def __generate_hash(password):\\n if password is None:\\n return None\\n return bcrypt.generate_password_hash(password, rounds=10).decode(\\\"utf8\\\")\",\n \"def verify_password(password, hash):\\n\\n method = get_hash_algorithm(flask.current_app.config['HASH_ALGORITHM'])\\n\\n return method.verify(password, hash)\",\n \"def check_password(self, raw_password):\\n def setter(raw_password):\\n self.set_password(raw_password)\\n self.save(update_fields=[\\\"password\\\"])\\n return check_password(raw_password, self.password, setter)\",\n \"def test_set_password(matrix, mock_check_output, mock_psycopg2):\\n matrix.save_pgsql_conf(db)\\n matrix.set_password(\\\"testuser\\\", \\\"testpassword\\\")\\n assert mock_check_output.called_with(\\n [\\n \\\"snap\\\",\\n \\\"run\\\",\\n \\\"{}.hash_password\\\".format(matrix.synapse_snap),\\n \\\"-c\\\",\\n matrix.synapse_config,\\n \\\"-p\\\",\\n \\\"testpassword\\\",\\n ]\\n )\\n assert mock_check_output.call_count == 1\\n assert mock_psycopg2.Cursor.execute.called_with(\\n \\\"UPDATE users SET password_hash='testhash' WHERE name='testuser';\\\"\\n )\\n assert mock_psycopg2.Cursor.execute.call_count == 1\",\n \"def set_password(self, password):\\n self.__init__(password=password)\",\n \"def set_password(self, password):\\n self.PASS = password\",\n \"def set_password(self, raw_password: str):\\n self.new_password = raw_password\",\n \"def update_password(self, password):\\n self.password = scryptsalsa208sha256_str(password.encode('utf-8')).decode('utf-8')\\n return True\",\n \"def hash_passwd(password, hash_method=\\\"sha256\\\"):\\n\\n return generate_password_hash(password, hash_method)\",\n \"def _authenticate_password(user_input: str, hash_: str) -> bool:\\n return pbkdf2_sha256.verify(user_input, hash_)\",\n \"def has_password(self, has_password):\\n if has_password is None:\\n raise ValueError(\\\"Invalid value for `has_password`, must not be `None`\\\") # noqa: E501\\n\\n self._has_password = has_password\",\n \"def verify_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def verify_password(self, password):\\n return check_password_hash(self.password_hash, password)\",\n \"def verify_password(self, password):\\n return check_password_hash(self.password_hash, password)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.7201298","0.7091728","0.70867187","0.6961345","0.69605434","0.69136447","0.6894837","0.6894837","0.68855447","0.6867788","0.68643236","0.6862712","0.68401206","0.68401206","0.68401206","0.68401206","0.67581147","0.6730123","0.6710391","0.6706543","0.6686309","0.66579616","0.6592467","0.65007746","0.64854056","0.648194","0.6362343","0.63571465","0.6336486","0.6293629","0.6273566","0.62694126","0.6268601","0.62603796","0.62519634","0.62309736","0.6172302","0.6155126","0.6154504","0.612065","0.60920024","0.60842884","0.60738456","0.60655934","0.6059644","0.6058584","0.6054388","0.6052507","0.6052507","0.6052507","0.6052507","0.6045436","0.6034001","0.6031313","0.6031313","0.60294145","0.6027881","0.60270596","0.60245925","0.6012174","0.60095686","0.6004385","0.6000069","0.59991777","0.59977543","0.5997209","0.5995166","0.5961534","0.596107","0.5956953","0.5943412","0.5932634","0.5919886","0.59000707","0.5888688","0.58865815","0.5854362","0.5854286","0.5853513","0.58299977","0.58277375","0.5826349","0.5826251","0.5821161","0.582002","0.5819606","0.5817649","0.58164245","0.58131385","0.5789562","0.578262","0.578247","0.5781","0.57788116","0.57704884","0.57636833","0.57610905","0.57467747","0.573398","0.573398","0.573398"],"string":"[\n \"0.7201298\",\n \"0.7091728\",\n \"0.70867187\",\n \"0.6961345\",\n \"0.69605434\",\n \"0.69136447\",\n \"0.6894837\",\n \"0.6894837\",\n \"0.68855447\",\n \"0.6867788\",\n \"0.68643236\",\n \"0.6862712\",\n \"0.68401206\",\n \"0.68401206\",\n \"0.68401206\",\n \"0.68401206\",\n \"0.67581147\",\n \"0.6730123\",\n \"0.6710391\",\n \"0.6706543\",\n \"0.6686309\",\n \"0.66579616\",\n \"0.6592467\",\n \"0.65007746\",\n \"0.64854056\",\n \"0.648194\",\n \"0.6362343\",\n \"0.63571465\",\n \"0.6336486\",\n \"0.6293629\",\n \"0.6273566\",\n \"0.62694126\",\n \"0.6268601\",\n \"0.62603796\",\n \"0.62519634\",\n \"0.62309736\",\n \"0.6172302\",\n \"0.6155126\",\n \"0.6154504\",\n \"0.612065\",\n \"0.60920024\",\n \"0.60842884\",\n \"0.60738456\",\n \"0.60655934\",\n \"0.6059644\",\n \"0.6058584\",\n \"0.6054388\",\n \"0.6052507\",\n \"0.6052507\",\n \"0.6052507\",\n \"0.6052507\",\n \"0.6045436\",\n \"0.6034001\",\n \"0.6031313\",\n \"0.6031313\",\n \"0.60294145\",\n \"0.6027881\",\n \"0.60270596\",\n \"0.60245925\",\n \"0.6012174\",\n \"0.60095686\",\n \"0.6004385\",\n \"0.6000069\",\n \"0.59991777\",\n \"0.59977543\",\n \"0.5997209\",\n \"0.5995166\",\n \"0.5961534\",\n \"0.596107\",\n \"0.5956953\",\n \"0.5943412\",\n \"0.5932634\",\n \"0.5919886\",\n \"0.59000707\",\n \"0.5888688\",\n \"0.58865815\",\n \"0.5854362\",\n \"0.5854286\",\n \"0.5853513\",\n \"0.58299977\",\n \"0.58277375\",\n \"0.5826349\",\n \"0.5826251\",\n \"0.5821161\",\n \"0.582002\",\n \"0.5819606\",\n \"0.5817649\",\n \"0.58164245\",\n \"0.58131385\",\n \"0.5789562\",\n \"0.578262\",\n \"0.578247\",\n \"0.5781\",\n \"0.57788116\",\n \"0.57704884\",\n \"0.57636833\",\n \"0.57610905\",\n \"0.57467747\",\n \"0.573398\",\n \"0.573398\",\n \"0.573398\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90875,"cells":{"query":{"kind":"string","value":"Create a room history entry, which is required for consistency reasons."},"document":{"kind":"string","value":"def room_history_entries(self, create, extracted, **kwargs):\n if create and self.room is not None:\n # Set room history entry begin to registration date\n rhe = RoomHistoryEntry.q.filter_by(user=self, room=self.room).one()\n rhe.active_during = starting_from(self.registered_at)\n\n for key, value in kwargs.items():\n setattr(rhe, key, value)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def get_room_history(self, room):\n pass","def create_historical_record(self, instance, history_type):\n history_date = getattr(instance, '_history_date', now())\n history_changeset = self.get_history_changeset(instance)\n manager = getattr(instance, self.manager_name)\n attrs = {}\n for field in instance._meta.fields:\n attrs[field.attname] = getattr(instance, field.attname)\n\n for field_name in self.additional_fields:\n loader = getattr(self, 'get_%s_value' % field_name)\n value = loader(instance, type)\n attrs[field_name] = value\n\n manager.create(\n history_date=history_date, history_type=history_type,\n history_changeset=history_changeset, **attrs)","async def _build_room_entry(self, room_id: str) -> JsonDict:\n stats = await self._store.get_room_with_stats(room_id)\n\n # currently this should be impossible because we call\n # check_user_in_room_or_world_readable on the room before we get here, so\n # there should always be an entry\n assert stats is not None, \"unable to retrieve stats for %s\" % (room_id,)\n\n current_state_ids = await self._store.get_current_state_ids(room_id)\n create_event = await self._store.get_event(\n current_state_ids[(EventTypes.Create, \"\")]\n )\n\n # TODO: update once MSC1772 lands\n room_type = create_event.content.get(EventContentFields.ROOM_TYPE)\n if not room_type:\n room_type = create_event.content.get(EventContentFields.MSC1772_ROOM_TYPE)\n\n room_version = await self._store.get_room_version(room_id)\n allowed_spaces = None\n if await self._event_auth_handler.has_restricted_join_rules(\n current_state_ids, room_version\n ):\n allowed_spaces = await self._event_auth_handler.get_spaces_that_allow_join(\n current_state_ids\n )\n\n entry = {\n \"room_id\": stats[\"room_id\"],\n \"name\": stats[\"name\"],\n \"topic\": stats[\"topic\"],\n \"canonical_alias\": stats[\"canonical_alias\"],\n \"num_joined_members\": stats[\"joined_members\"],\n \"avatar_url\": stats[\"avatar\"],\n \"world_readable\": (\n stats[\"history_visibility\"] == HistoryVisibility.WORLD_READABLE\n ),\n \"guest_can_join\": stats[\"guest_access\"] == \"can_join\",\n \"creation_ts\": create_event.origin_server_ts,\n \"room_type\": room_type,\n \"allowed_spaces\": allowed_spaces,\n }\n\n # Filter out Nones – rather omit the field altogether\n room_entry = {k: v for k, v in entry.items() if v is not None}\n\n return room_entry","def add_history(self):\n # add separator, if there already are history entries\n if self.parentApp.History != '':\n self.parentApp.History += (\n '\\n\\n--- --- --- --- --- --- --- --- --- --- --- ---\\n\\n'\n )\n\n # add the transaction to it\n self.parentApp.History += self.parentApp.tmpTransC.to_str()","def __init__(self, *args):\n this = _ida_hexrays.new_history_item_t(*args)\n try: self.this.append(this)\n except: self.this = this","def create_test_history(**kw):\n history = get_test_history(**kw)\n # Let DB generate ID if it isn't specified explicitly\n if 'id' not in kw:\n del history['id']\n dbapi = db_api.get_instance()\n return dbapi.create_node_history(history)","def addHistory(self):\r\n \r\n data = self.get()\r\n \r\n if data == '':\r\n return\r\n elif len(self.history) != 0 and self.history[0] == data:\r\n return\r\n \r\n if len(self.history) == self.historySize:\r\n self.history.pop()\r\n \r\n self.history.insert(0, data)","def create_entry(entry):\n Entry.create(**entry)\n return entry","def save(self, *args, **kwargs):\n changesDict = self.diff\n\n # Let's construct the history record entry and add it to the dict.\n # Each record is further going to be dictionary stored in the DB as a string.\n\n # We need to keep track of which user updated the data. Check if the user has\n # assigned a callable to fetch it. Else we will initialize it to be empty.\n who = None\n if self.history__get_user_hook is not None:\n if callable(self.history__get_user_hook):\n who = self.history__get_user_hook()\n else:\n logger.error(\"history_get_user initialized with a non-callable\")\n\n # only add the 'who' field if we have a string value.\n if who is not None and isinstance(who,str):\n entry = {\n \"who\" : str(who),\n \"when\" : timezone.now().isoformat(),\n \"what\" : changesDict,\n }\n else:\n entry = {\n \"when\" : timezone.now().isoformat(),\n \"what\" : changesDict,\n }\n\n # prepend this entry into the list\n if isinstance(self.history, list):\n self.history.insert(0,entry)\n if isinstance(self.history__max_entry_count, int):\n if self.history__max_entry_count >= 0:\n # Value has been specified. Trim the list. Value of 0 deletes the whole list.\n del self.history[self.history__max_entry_count: ]\n logger.debug('Added the following record to history:\\n %s'%(str(entry),))\n else:\n logger.error(\"History field not saved initially as list!!\")\n\n # Let's not hold up the save. Save and reset the initial state.\n super(ChangeHistoryMixin, self).save(*args, **kwargs)\n self.__initial = self._dict","def _push_history(self):\n self._history.append(self._state)","def new_room(request):\n print(\"new room\")\n new_room = None\n while not new_room:\n with transaction.atomic():\n label = haikunator.haikunate()\n if Room.objects.filter(label=label).exists():\n continue\n new_room = Room.objects.create(label=label)\n return redirect('chat:chat_room', label=label)","def append_history_record(self, history_record):\n self.history.append(history_record)\n self._increment_history_pointer()","def history():","def create_and_add_entry(self, **attrs):\n return self.add_entry(self.create_entry(**attrs))","def add_history(self, author, string, data_var=None):\n\n attr = 'history'\n self.add_string(attr, author, string, data_var=data_var)","def __build_history(self, obj: Object) -> dict:\n previous_history = obj.history\n return {**previous_history, self.__get_timestamp(): {'type_id': obj.type_id, 'redshift': obj.redshift}}","def add_entry(self, *args, **kwargs):\n entry = Entry(*args, **kwargs) # NOTE: not sure this is good\n self._entries[entry.uuid] = entry\n return entry","def creating_entry(self):\n response = \"\"\n today = str(date.today())\n curent_time = str(datetime.time(datetime.now()))\n entry = Diary(self.entry_id, self.title, self.body)\n lst = {}\n lst[\"entry_id\"] = entry.entry_id\n lst[\"title\"] = entry.title\n lst[\"date\"] = today\n lst[\"time\"] = curent_time\n lst[\"body\"] = entry.body\n lst[\"updated\"] = entry.updated\n if Validate.validate_entry(Diary.entries, entry):\n response = jsonify({\"message\": \"Duplicate data,Try again\"})\n response.status_code = 409\n else:\n Diary.entries.append(lst)\n response = jsonify({\"message\": \"Entry saved\", \"data\": lst})\n response.status_code = 201\n return response","def __build_history(self, obj: Object) -> dict:\n previous_history = dict(obj.history)\n return {**previous_history, self.__get_timestamp(): {'type_id': obj.type_id, 'redshift': obj.redshift}}","def add_history(self,date,what,note):\r\n note = '.'.join(note.split(','))\r\n self.history.append([date,what,note])","def new_history_entry(self, answer):\r\n answer = OpenEndedChild.sanitize_html(answer)\r\n self.child_history.append({'answer': answer})\r\n self.stored_answer = None","def new_room(request):\n new_room = None\n while not new_room:\n with transaction.atomic():\n label = id_generator(10)\n if Room.objects.filter(label=label).exists():\n continue\n new_room = Room.objects.create(label=label)\n return redirect(chat_room, label=label)","def add_room(self, data):\n room_id = data['room_id']\n x, y = literal_eval(data['coordinates'])\n room_data = {'id': data['room_id'],\n 'title': data['title'],\n 'description' : data['description'],\n 'coordinates': literal_eval(data['coordinates']),\n 'elevation': data['elevation'],\n 'terrain': data['terrain'],\n 'exits' : {direction: '?' for direction in data['exits']}\n }\n self.rooms.setdefault(room_id, room_data)","def create_base_entry(vin=\"INVALID\", time_unix=None):\n\t\treturn LogEntry(vin=vin, app_id=\"INVALID\", time_unix=time_unix)","def onRegisterHistory(self):\n pass","def new_entry():\n clear_screen()\n entry = {}\n entry['id'] = get_next_id()\n entry['name'] = input_name()\n print(\"How many minutes did you spend on {}?\".format(entry['name']))\n print(\"Or you may specify a format after the time, seperated by a comma\")\n entry['time_spent'] = input_time_spent()\n add_notes = input(\"Add notes? Y/n \").lower()\n if add_notes != 'n':\n entry['notes'] = input_notes()\n entry['date'] = datetime.now().strftime(FMT_MONTH_DAY_YEAR)\n with open(WORK_LOG_FILENAME, 'a', newline='') as work_log:\n work_log_writer = csv.DictWriter(work_log, fieldnames=FIELDNAMES)\n work_log_writer.writerow(entry)","def room_add():\n if not json_param_exists('number'):\n abort(400, 'invalid room number')\n\n if not isinstance(request.json['number'], str):\n abort(400, 'room number must be string')\n\n num = request.json['number']\n room = Room(number=num)\n\n try:\n get_db().add(room)\n get_db().commit()\n except IntegrityError:\n abort(409, 'room number is already in use')\n return json.dumps(room.as_dict(include_features=False)), 201","def add_to_game_history(self, tile, flag=False):\n move = {\n 'tile': tile,\n 'flag':flag,\n 'coordinate': self.stack[tile]['coordinate'],\n 'value': self.stack[tile]['value']}\n self.history.append(move)","def save_record_state_to_history(result=None, **kwargs):\n\n # No need to store history array within the history table data\n data = copy.copy(result)\n if 'history' in data:\n del data['history']\n\n if 'ratings' in data:\n del data['ratings']\n\n history = RecordHistory(record_id=result['id'], user_id=result['user_id'],\n data=json.dumps(data))\n\n db.session.add(history)\n db.session.commit()\n\n if 'history' not in result:\n result['history'] = []\n\n result['history'].insert(0, to_dict(history))","def history():\n return apology(\"TODO\")","def history():\n return apology(\"TODO\")","def history():\n return apology(\"TODO\")","def test_tracker_addHistory():\n\n trackers, cap = init_tracker()\n tr = trackers[0]\n tr.addHistory([1, 1, 1, 1])\n\n assert len(tr.history) >= 1","def add_history(self, text, author='', date=None):\n date = date or datetime.date.today()\n version = self.header['SRC_V']\n ttext = '[%s] %s (%s %s)' % (version, text, author, str(date))\n if len(ttext) > 68:\n n = len(text) - len(ttext) + 68\n ttext = '[%s] %s (%s %s)' % (version, text[:n], author, str(date))\n self.header['HISTORY'] = ttext","def instantiate_history(self):\n serialized_history = self.cache.get('history')\n history = ast.literal_eval(serialized_history.decode('utf-8'))\n return history","def add_new_entry(self):\n clear_screen()\n new_entry = Entry.create()\n if new_entry is None:\n print(\"Add new entry cancelled. Returning to main menu...\")\n time.sleep(1)\n return None\n self.entries.append(new_entry)\n with open(self.file_name, \"a\") as file:\n writer = csv.writer(file)\n writer.writerow([new_entry.date, new_entry.name, new_entry.minutes, new_entry.note])","def _append_history(ds):\n try:\n history = ds.attrs['history']\n except KeyError:\n history = \"\"\n now = datetime.datetime.now()\n prog = __file__ # os.path.basename(__file__)\n history = (now.strftime(\"%a %b %d %H:%M:%S %Y\") +\n \": {} {}\\n\".format(prog, \" \".join(sys.argv[1:])) +\n history)\n ds.attrs['history'] = history","def log_create(sender, instance, created, **kwargs):\n if created:\n changes = model_instance_diff(None, instance)\n\n log_entry = LogEntry.objects.log_create(\n instance,\n action=LogEntry.Action.CREATE,\n changes=json.dumps(changes),\n )\n log_created.send(\n sender=LogEntry,\n old_instance=None,\n new_instance=instance,\n log_instance=log_entry,\n )","def create(self):\n logging.warning(\n \"IRC back-end does not support explicit creation, joining room \"\n \"instead to ensure it exists.\"\n )\n self.join()","def test_hist_append():\n FNAME = 'xonsh-SESSIONID.json'\n FNAME += '.append'\n hist = History(filename=FNAME, here='yup', **HIST_TEST_KWARGS)\n with mock_xonsh_env({'HISTCONTROL': set()}):\n hf = hist.append({'joco': 'still alive'})\n yield assert_is_none, hf\n yield assert_equal, 'still alive', hist.buffer[0]['joco']\n os.remove(FNAME)","def _grab_history(self):\n self.data['history_lines'] = []\n self.data['history_file'] = None\n self.data['history_encoding'] = None\n self.data['headings'] = []\n self.data['history_last_release'] = ''\n self.data['history_insert_line_here'] = 0\n default_location = None\n config = self.setup_cfg.config\n if config and config.has_option('zest.releaser', 'history_file'):\n default_location = config.get('zest.releaser', 'history_file')\n history_file = self.vcs.history_file(location=default_location)\n self.data['history_file'] = history_file\n if not history_file:\n logger.warn(\"No history file found\")\n return\n logger.debug(\"Checking %s\", history_file)\n history_lines, history_encoding = read_text_file(history_file)\n history_lines = history_lines.split('\\n')\n headings = utils.extract_headings_from_history(history_lines)\n if not headings:\n logger.warn(\"No detectable version heading in the history \"\n \"file %s\", history_file)\n return\n self.data['history_lines'] = history_lines\n self.data['history_encoding'] = history_encoding\n self.data['headings'] = headings\n\n # Grab last header.\n start = headings[0]['line']\n if len(headings) > 1:\n # Include the next header plus underline, as this is nice\n # to show in the history_last_release.\n end = headings[1]['line'] + 2\n else:\n end = len(history_lines)\n history_last_release = '\\n'.join(history_lines[start:end])\n self.data['history_last_release'] = history_last_release\n\n # Add line number where an extra changelog entry can be inserted. Can\n # be useful for entry points. 'start' is the header, +1 is the\n # underline, +2 is probably an empty line, so then we should take +3.\n # Or rather: the first non-empty line.\n insert = start + 2\n while insert < end:\n if history_lines[insert].strip():\n break\n insert += 1\n self.data['history_insert_line_here'] = insert","def add_entry(self, scenario_info):\n scenario_id, status = scenario_info[\"id\"], \"created\"\n sql = self.insert()\n self.cur.execute(\n sql,\n (\n scenario_id,\n status,\n ),\n )","def add_entry(self, scenario_info):\n scenario_id, status = scenario_info[\"id\"], \"created\"\n sql = self.insert()\n self.cur.execute(\n sql,\n (\n scenario_id,\n status,\n ),\n )","def add_vertex(self, room):\r\n if room['room_id'] not in self.rooms:\r\n self.rooms[room['room_id']] = room\r\n # self.rooms[room['room_id']]['exits'] = {\r\n # d: '?' for d in room['exits']}\r","def history():\n db.execute(\"CREATE TABLE IF NOT EXISTS hist(user_id INTEGER NOT NULL, typ TEXT NOT NULL, symbol TEXT NOT NULL, name TEXT NOT NULL, price NUMERIC NOT NULL, nos INTEGER NOT NULL, cost NUMERIC NOT NULL, time DATETIME NOT NULL, FOREIGN KEY(user_id) REFERENCES users(id))\")\n stocks = db.execute(\"SELECT * FROM hist WHERE user_id = ?\", session[\"user_id\"])\n return render_template(\"history.html\", stocks = stocks)","def _write_history(self):\n if self.data['history_file'] is None:\n return\n contents = '\\n'.join(self.data['history_lines'])\n history = self.data['history_file']\n write_text_file(\n history, contents, encoding=self.data['history_encoding'])\n logger.info(\"History file %s updated.\", history)","def test_add_entry_creates_db_entry_with_correct_details(self):\n example_inputs = [\n 'Example Employee',\n '2018-05-01',\n 'Example Task',\n 100,\n 'Example Note'\n ]\n with patch('builtins.input', side_effect=example_inputs):\n self.menu.add_entry()\n\n query = (db_manager\n .LogEntry\n .select()\n .join(db_manager.Employee)\n .where(\n db_manager.Employee.name == example_inputs[0],\n db_manager.LogEntry.date == example_inputs[1],\n db_manager.LogEntry.task_name == example_inputs[2],\n db_manager.LogEntry.duration == example_inputs[3],\n db_manager.LogEntry.notes == example_inputs[4]\n ))\n record = query[0]\n record_data = [\n record.employee.name,\n record.date.strftime(\"%Y-%m-%d\"),\n record.task_name,\n record.duration,\n record.notes\n ]\n self.assertEqual(example_inputs, record_data)","def save_history(sender, instance, **kwargs): # pylint: disable=no-self-argument, unused-argument\r\n if instance.module_type in StudentModuleHistory.HISTORY_SAVING_TYPES:\r\n history_entry = StudentModuleHistory(student_module=instance,\r\n version=None,\r\n created=instance.modified,\r\n state=instance.state,\r\n grade=instance.grade,\r\n max_grade=instance.max_grade)\r\n history_entry.save()","def update_history(self, move):\r\n player_number = self.player_numbers[self.current_player]\r\n heaps = tuple(self.heaps)\r\n self.history.append([player_number, heaps, move])","def add_entry(name, title, duration, notes):\n clear()\n print('Entry added to work log!')\n return Entry.create(\n employee_name=name,\n task_title=title,\n time_spent=duration,\n task_notes=notes\n )","def func(self):\n\n caller = self.caller\n\n if not self.lhs:\n string = \"Usage: @dig[/teleport] roomname[;alias;alias...][:parent] [= exit_there\"\n string += \"[;alias;alias..][:parent]] \"\n string += \"[, exit_back_here[;alias;alias..][:parent]]\"\n caller.msg(string)\n return\n\n room = self.lhs_objs[0]\n\n if not room[\"name\"]:\n caller.msg(\"You must supply a new room name.\")\n return\n location = caller.location\n\n # Create the new room\n typeclass = room[\"option\"]\n if not typeclass:\n typeclass = settings.BASE_ROOM_TYPECLASS\n\n # create room\n lockstring = \"control:id(%s) or perm(Immortals); delete:id(%s) or perm(Wizards); edit:id(%s) or perm(Wizards)\"\n lockstring = lockstring % (caller.dbref, caller.dbref, caller.dbref)\n\n new_room = create.create_object(\n typeclass, room[\"name\"], aliases=room[\"aliases\"], report_to=caller\n )\n new_room.locks.add(lockstring)\n alias_string = \"\"\n if new_room.aliases.all():\n alias_string = \" (%s)\" % \", \".join(new_room.aliases.all())\n room_string = \"Created room %s(%s)%s of type %s.\" % (\n new_room,\n new_room.dbref,\n alias_string,\n typeclass,\n )\n\n # create exit to room\n\n exit_to_string = \"\"\n exit_back_string = \"\"\n\n if self.rhs_objs:\n to_exit = self.rhs_objs[0]\n if not to_exit[\"name\"]:\n exit_to_string = \"\\nNo exit created to new room.\"\n elif not location:\n exit_to_string = \"\\nYou cannot create an exit from a None-location.\"\n else:\n # Build the exit to the new room from the current one\n typeclass = to_exit[\"option\"]\n if not typeclass:\n typeclass = settings.BASE_EXIT_TYPECLASS\n\n new_to_exit = create.create_object(\n typeclass,\n to_exit[\"name\"],\n location,\n aliases=to_exit[\"aliases\"],\n locks=lockstring,\n destination=new_room,\n report_to=caller,\n )\n alias_string = \"\"\n if new_to_exit.aliases.all():\n alias_string = \" (%s)\" % \", \".join(new_to_exit.aliases.all())\n exit_to_string = \"\\nCreated Exit from %s to %s: %s(%s)%s.\"\n exit_to_string = exit_to_string % (\n location.name,\n new_room.name,\n new_to_exit,\n new_to_exit.dbref,\n alias_string,\n )\n\n # Create exit back from new room\n\n if len(self.rhs_objs) > 1:\n # Building the exit back to the current room\n back_exit = self.rhs_objs[1]\n if not back_exit[\"name\"]:\n exit_back_string = \"\\nNo back exit created.\"\n elif not location:\n exit_back_string = (\n \"\\nYou cannot create an exit back to a None-location.\"\n )\n else:\n typeclass = back_exit[\"option\"]\n if not typeclass:\n typeclass = settings.BASE_EXIT_TYPECLASS\n new_back_exit = create.create_object(\n typeclass,\n back_exit[\"name\"],\n new_room,\n aliases=back_exit[\"aliases\"],\n locks=lockstring,\n destination=location,\n report_to=caller,\n )\n alias_string = \"\"\n if new_back_exit.aliases.all():\n alias_string = \" (%s)\" % \", \".join(new_back_exit.aliases.all())\n exit_back_string = \"\\nCreated Exit back from %s to %s: %s(%s)%s.\"\n exit_back_string = exit_back_string % (\n new_room.name,\n location.name,\n new_back_exit,\n new_back_exit.dbref,\n alias_string,\n )\n caller.msg(\"%s%s%s\" % (room_string, exit_to_string, exit_back_string))\n if new_room and (\"teleport\" in self.switches or \"tel\" in self.switches):\n caller.move_to(new_room)\n # use property to set colored name attribute and strip ansi markup from key\n new_room.name = new_room.name\n return new_room","def save_history(cube, field, filename): \n\n history.append(cube.attributes['history'])","def __init__(self, *args):\n this = _libsbml.new_ModelHistory(*args)\n try: self.this.append(this)\n except: self.this = this","def history(self, history):\n self._history = history","def save_history():\n\n mid = get_mid()\n back_file = contact_name + \"_\" + today\n\n if not os.path.isdir(back_path):\n print('WARNING: o {} directory found, creating.').format(back_path)\n os.mkdir(back_path)\n else:\n print(\"OK: {} found.\".format(back_path))\n\n os.chdir(back_path)\n with open(back_file, 'w') as bf:\n for mes in get_todays_history(mid):\n data = \"{}\\n\".format(mes)\n bf.write(data)","async def history(message, client, extra_args):\n\n if not extra_args or not (user_id := utils.from_mention(extra_args[0])):\n user_id = message.author.id\n\n @database.query\n def get_transactions(conn):\n cursor = conn.cursor()\n cursor.execute(\n \"SELECT * FROM funnypts WHERE awarder = ? OR awardee = ? ORDER BY date DESC\", (user_id, user_id))\n transactions = cursor.fetchall()\n cursor.close()\n conn.close()\n return transactions\n\n if not (transactions := get_transactions()):\n await message.channel.send(\"THIS USER HAS NO HISTORY, THEY SHOULD THOUGH\")\n return\n\n @utils.paginated_embeds\n def populate(embed, entry, entry_number):\n awarder = client.get_user(entry[0]).name.split(\"#\", 1)[0]\n awardee = client.get_user(entry[1]).name.split(\"#\", 1)[0]\n transaction = \"GIVEN TO\" if entry[3] > 0 else \"TAKEN FROM\"\n date = entry[4].split(\" \", 1)[0]\n reason = \"\\\"{0}\\\"\".format(entry[2])\n\n embed.add_field(\n name=\"{0} — {2} — {1} • {3}\".format(awarder, awardee, transaction, date), value=reason, inline=False)\n\n title = f\"{client.get_user(user_id).name}'s FUNNYPOINT HISTORY\"\n embeds = populate(title, transactions, page_length=5)\n await utils.sauce_pages(embeds, message, client)","def from_history_event(expected_remote, h):\n event = eptStaleEvent()\n event.expected_remote = expected_remote\n event.ts = h.ts\n event.remote = h.remote\n event.pctag = h.pctag\n event.encap = h.encap\n event.flags = h.flags\n event.intf_id = h.intf_id\n event.intf_name = h.intf_name\n event.epg_name = h.epg_name\n event.vnid_name = h.vnid_name\n return event","def create_entry(cls, title, date, timeSpent, learned, resources):\n try:\n with DATABASE.transaction():\n cls.create(\n title=title,\n date=date,\n timeSpent=timeSpent,\n learned=learned,\n resources=resources\n )\n except IntegrityError:\n raise ValueError(\"Entry already exists\")","def history():\n rows = db.execute(\"SELECT * FROM histories WHERE id=:id\", id=session[\"user_id\"])\n\n return render_template(\"history.html\", rows=rows)","def get_history():\n return response_texts_to_entries(make_post_request(HISTORY_API, data={\"k\": config[\"api_key\"]}))","def update_history(self, guess='', result=''):\n item = json.dumps({'guess': guess, 'result': result})\n self.history.append(item)","def insertHistoryRecord(self, record, date=None, note=\" \"):\n if date == None:\n date = self.today()\n \n if self.historyTableName == None:\n return []\n\n else:\n \n fieldValues = list(record) + [date, note]\n fieldString = \"(\" + \",\".join(self.getColumnHeads(self.historyTableName)) + \")\"\n questionMarks = \"(\" + \",\".join([\"?\"] * len(fieldValues)) + \")\"\n\n self._db._c.execute(\"INSERT INTO \" + self.historyTableName +\n fieldString + \" VALUES \" + questionMarks,\n fieldValues)\n self._db._conn.commit()\n\n return fieldValues[0]","def record_history_on_commit(self,\n clocked: 'Clocked',\n changes: dict,\n session: orm.Session,\n timestamp: dt.datetime):\n new_tick = self._get_new_tick(clocked)\n\n new_clock = self.make_clock(timestamp, new_tick)\n attr = {'entity': clocked}\n\n for prop, cls in self.history_models.items():\n if prop in changes:\n value = changes[prop]\n\n self._cap_previous_history_row(clocked, new_clock, cls)\n\n # Add new history row\n hist = attr.copy()\n hist[prop.key] = value\n session.add(\n cls(\n vclock=new_clock.vclock,\n effective=new_clock.effective,\n **hist,\n ),\n )","def handle_create_room(self, lobby_command, client_socket):\n msg = \"Handling room creation of {0}\".format(lobby_command)\n print(msg)\n user = self.clients[client_socket]['data'].decode('utf-8')\n roomname = lobby_command.split()[1]\n\n if roomname == \"mine\" or roomname == \"all\":\n msg = f'Client {user} error! reserved word that cannot be a room name.'\n self.log_and_send(client_socket, msg)\n return\n\n for room in self.rooms:\n if room.name == roomname:\n msg = f\"Invalid request from client: {roomname} already exists!\"\n self.log_and_send(client_socket, msg)\n return\n\n self.rooms.append(Room(name=roomname, creator=user))\n msg = f\"Room {roomname} created for client {user} (creator/Admin).\"\n self.log_and_send(client_socket, msg)\n return","def add_room(self, room):\n self.rooms[room.name] = room","def history(self, history):\n\n self._history = history","def insertRoom(cursor, room: Room) -> None:\n\n # TODO: Calculate the values. room object may not contain the creator\n # or maybe remove that feature\n cursor.execute(\n \"INSERT INTO rooms VALUES (?, ?, ?)\",\n (roomID, roomName, creator)\n )","def add_activity(self, record):\n # this brings shame to me but meh, faster to hardcode (see model_activity)\n # id = str(record.value(\"id\"))\n id = record.value(\"id\")\n room = record.value(\"room\")\n if room == 0:\n room = \"Nenhuma\"\n else:\n room = str(room)\n weekday = constants.days_of_the_week[record.value(\"weekday\")]\n weektime = record.value(\"weektime\").toString(\"HH:mm\")\n entry = (id, record.value(\"description\"), room,\n weekday, weektime)\n\n if entry in self._activity_list:\n return False\n else:\n if self.is_in_del_queue(id):\n # no real need to add as it was queued to be removed\n self._removed_activities.remove(id)\n else:\n self._added_activities.append(id)\n # print self._added_activities\n\n self._activity_list.append(entry)\n # sorts by day/time\n self._activity_list.sort(key=operator.itemgetter(3,4))\n # self._activity_list = sorted(self._activity_list, key=lambda dia_hora: (dia_hora[3], dia_hora[2]))\n return True","def add(\n description: str = typer.Argument(\n ...,\n help=\"Description of the log entry\"\n ),\n date: datetime = typer.Option(\n datetime.now().strftime(\"%Y-%m-%d\"), '--date', '-d',\n help=\"Date of the log entry\"\n ),\n time: datetime = typer.Option(\n datetime.now().strftime(\"%I:%M %p\"), '--time', '-t',\n formats=[\"%H:%M:%S\", \"%I:%M %p\"],\n help=\"Time of the log entry\"\n )\n):\n log_entry_time = time.time()\n log_datetime = datetime.combine(date, log_entry_time)\n\n manager = LogBookManager()\n created, message = manager.create(description, log_datetime)\n\n if created:\n typer.echo(\n typer.style(message, fg=typer.colors.GREEN, bold=True)\n )\n else:\n typer.echo(\n typer.style(message, fg=typer.colors.RED, bold=True)\n )","def create(data):\n db = core.connect()\n theTime = utils.utctime()\n data[\"created\"] = theTime\n data[\"modified\"] = theTime\n data[\"domain\"] = utils.domain(data[\"href\"])\n newShift = schema.shift()\n newShift.update(data)\n newShift[\"type\"] = \"shift\"\n id = db.create(newShift)\n newShift = db[id]\n return joinData(newShift, newShift[\"createdBy\"])","def log_create(action, *args, **kw):\n from olympia.activity.models import ActivityLog\n\n return ActivityLog.create(action, *args, **kw)","def create_history_form(self):\n history_items = [json.loads(item) for item in self.history]\n\n history_form_items = []\n for item in history_items:\n history_form_items.append(GameHistoryForm(guess=item.get('guess'),\n result=item.get('result')))\n\n return GameHistoryForms(history=history_form_items)","def new_room(self):\r\n return Room()","def new_archive_record(self, event):\n dbmanager = self.engine.db_binder.get_manager(self.data_binding)\n dbmanager.addRecord(event.record)","def add_entry(self, account):\n def txn():\n entry = self.entries.filter('account =', account).get()\n if not entry:\n entry = Entry(account=account, parent=self)\n entry.put()\n created = True\n else:\n created = False\n return entry, created\n return db.run_in_transaction(txn)","def history():\n history = db.execute(\"SELECT * from history WHERE id=:id\", id=session[\"user_id\"])\n\n return render_template(\"history.html\", history = history)","def __insert_history_module(self, form_data):\n\n sql_fields = []\n sql_fields.extend([\"`User_ID`=%(user_id)s\",\n \"`Moderated`=%(moderated)s\",\n \"`Moderator_ID`=%(moderator_id)s\",\n \"`Type_ID`=%(type_id)s\",\n \"`Country_ID`=%(country_id)s\",\n \"`State_ID`=%(state_id)s\",\n \"`Accepted`=%(accepted)s\"])\n\n sql_values = {}\n\n parent_plant_id = form_data.get(\"Parent_Plant_ID\", 0)\n\n if int(parent_plant_id) > 0:\n sql_fields.append(\"`Parent_Plant_ID`=%(parent_plant_id)s\")\n sql_values['parent_plant_id'] = parent_plant_id\n\n sql_statement = \"INSERT INTO History SET \" + \",\".join(sql_fields)\n\n sql_values['user_id'] = str(form_data.get('User_ID', 0))\n sql_values['moderated'] = form_data.get('Moderated', 0)\n sql_values['moderator_id'] = form_data.get('Moderator_ID', None)\n sql_values['type_id'] = form_data.get('Type_ID')\n sql_values['country_id'] = form_data.get('Country_ID')\n sql_values['state_id'] = form_data.get('State_ID')\n sql_values['accepted'] = form_data.get('Accepted')\n\n session = self.db_conn.cursor()\n try:\n session.execute(sql_statement, sql_values)\n self.db_conn.commit()\n except Exception:\n #session.rollback()\n raise\n\n description_id = session.lastrowid\n\n if int(parent_plant_id) == 0:\n parent_update_stmt = \"UPDATE HISTORY SET \\\n Parent_Plant_ID=%(parent_plant_id)s \\\n WHERE Description_ID=%(description_id)s\"\n try:\n session.execute(parent_update_stmt,\n {'parent_plant_id': description_id,\n 'description_id': description_id})\n self.db_conn.commit()\n except Exception:\n #session.rollback()\n raise\n\n session.close()\n return description_id","def add(self, record):\n self._hist_records[record.uid] = record","def save_history(self, request, scenario, user):\n if isinstance(scenario, Scenario):\n s = scenario.to_dict()\n else:\n s = scenario\n h = History(request=request, scenario=s, user=user)\n h.save()\n return h","def history():\n history = db.execute(\"SELECT * from history WHERE id=:id\", id=session[\"user_id\"])\n \n return render_template(\"history.html\", history = history)","def __init__(self, history=None):\n\n self.__history = history if history else []","def create_history_application(python_input, original_document):\n history_mapping = HistoryMapping(python_input.history, original_document)\n\n def default_buffer_pos_changed():\n \"\"\" When the cursor changes in the default buffer. Synchronize with\n history buffer. \"\"\"\n # Only when this buffer has the focus.\n if buffer_mapping.focus_stack[-1] == DEFAULT_BUFFER:\n try:\n line_no = default_buffer.document.cursor_position_row - \\\n history_mapping.result_line_offset\n\n if line_no < 0: # When the cursor is above the inserted region.\n raise IndexError\n\n history_lineno = sorted(history_mapping.selected_lines)[line_no]\n except IndexError:\n pass\n else:\n history_buffer.cursor_position = \\\n history_buffer.document.translate_row_col_to_index(history_lineno, 0)\n\n def history_buffer_pos_changed():\n \"\"\" When the cursor changes in the history buffer. Synchronize. \"\"\"\n # Only when this buffer has the focus.\n if buffer_mapping.focus_stack[-1] == HISTORY_BUFFER:\n line_no = history_buffer.document.cursor_position_row\n\n if line_no in history_mapping.selected_lines:\n default_lineno = sorted(history_mapping.selected_lines).index(line_no) + \\\n history_mapping.result_line_offset\n\n default_buffer.cursor_position = \\\n default_buffer.document.translate_row_col_to_index(default_lineno, 0)\n\n history_buffer = Buffer(\n initial_document=Document(history_mapping.concatenated_history),\n on_cursor_position_changed=Callback(history_buffer_pos_changed),\n accept_action=AcceptAction(\n lambda cli, buffer: cli.set_return_value(default_buffer.document)),\n read_only=True)\n\n default_buffer = Buffer(\n initial_document=history_mapping.get_new_document(),\n on_cursor_position_changed=Callback(default_buffer_pos_changed),\n read_only=True)\n\n help_buffer = Buffer(\n initial_document=Document(HELP_TEXT, 0),\n accept_action=AcceptAction.IGNORE,\n read_only=True\n )\n\n buffer_mapping = BufferMapping({\n HISTORY_BUFFER: history_buffer,\n DEFAULT_BUFFER: default_buffer,\n HELP_BUFFER: help_buffer,\n }, initial=HISTORY_BUFFER)\n\n application = Application(\n layout=create_layout(python_input, history_mapping),\n use_alternate_screen=True,\n buffers=buffer_mapping,\n style=python_input._current_style,\n mouse_support=Condition(lambda cli: python_input.enable_mouse_support),\n key_bindings_registry=create_key_bindings(python_input, history_mapping)\n )\n return application","def create_new_hist(gi, galaxyemail, galaxypass, server, workflowid, files, new_hist):\n if workflowid != \"0\":\n if len(filter(None, files)) > 0:\n workflow = gi.workflows.show_workflow(workflowid)\n if new_hist is None or new_hist == \"\":\n new_hist_name = strftime(workflow['name'] + \"_%d_%b_%Y_%H:%M:%S\", gmtime())\n else:\n new_hist_name = new_hist\n gi.histories.create_history(name=new_hist_name)\n history_id = get_history_id(galaxyemail, galaxypass, server)\n else:\n pass\n else:\n if len(filter(None, files)) > 0:\n if new_hist is None or new_hist == \"\":\n new_hist_name = strftime(\"Use_Galaxy_%d_%b_%Y_%H:%M:%S\", gmtime())\n else:\n new_hist_name = new_hist\n gi.histories.create_history(name=new_hist_name)\n history_id = get_history_id(galaxyemail, galaxypass, server)\n else:\n pass\n return history_id","def logbook_create(name, lb_id=None):\n return IMPL.logbook_create(name, lb_id)","def createRecord(self):\n self.dto.getRecord().append(self.controller.createNewObj())\n print(\"Record added.\")","def generate_history(self):\n self.reporter.generate()","def create_history_line(field_name: str, old_value: str, new_value: str) -> str:\n return 'Updated field \"' + field_name + '\" from \"' + old_value + '\" to \"' + new_value + '\". '","def append_to_history(self, message, plugin):\n datefmt = '%Y-%m-%dT%H:%M:%S.%fZ%z' # 2018-03-27T13:29:12.623222Z\n now = datetime.datetime.utcnow()\n log_item = {\n 'timestamp': now.strftime(datefmt),\n 'plugin': plugin,\n 'message': message\n }\n self.history_log.append(log_item)","def add_room(self, room):\n self.rooms.append(room)","def RecordHistory( self ):\n if not self.restoringHistory:\n record = self.activated_node\n if self.historyIndex < -1:\n try:\n del self.history[self.historyIndex+1:]\n except AttributeError, err:\n pass\n if (not self.history) or record != self.history[-1]:\n self.history.append( record )\n del self.history[:-200]\n self.historyIndex = -1","def push_history(self, submission, action, program_compiled):\n\n history_data = {\n 'user': self.user.get().name,\n 'score': self.score,\n 'action': action,\n 'submission': submission,\n 'program_compiled': program_compiled,\n 'level': self.current_level\n }\n taskqueue.add(url='/tasks/push_game_history', params=history_data)","def test_create_log(self):\n message = \"Message is {0}\".format(random.random())\n resp = gracedb.writeLog(eventId, message)\n self.assertEqual(resp.status, 201)\n new_log_uri = resp.getheader('Location')\n new_log = resp.json()\n self.assertEqual(new_log_uri, new_log['self'])\n check_new_log = gracedb.get(new_log_uri).json()\n self.assertEqual(check_new_log['comment'], message)","def create_new_event(self):\n pass","def _add_change_log(self):\n wiz = self.machine_email_id\n change_log = self.env['machine.instance.change_log'].create({\n 'name': wiz.sub_subject,\n 'date': wiz.date,\n 'duration': wiz.duration,\n 'user_id': wiz.user_id.id,\n 'priority': wiz.priority,\n 'machine_instance_id': self.machine_instance_id.id,\n\n })\n self.change_log_id = change_log.id","def Modifier_History(self):\n\t\tpass","def push(self, value):\n self.history.append(value)","def create_sync_entry(ts, coll, idx):\n sync_log = connection.ElasticLogs()\n sync_log.ts = ts\n sync_log.coll = unicode(coll)\n sync_log.idx = unicode(idx)\n sync_log.save()\n return True","def slot_history_changed(self, _sender, _data):\r\n last_candle = self.history.last_candle()\r\n if last_candle:\r\n self.client.history_last_candle = last_candle.tim","def rec_history(self, epoch, iter, mode, batch_size, lr, loss_val, metrics_val):\n with closing(sqlite3.connect(self.dbfile, detect_types=self.time_setting)) as con:\n cursor = con.cursor()\n sql = 'insert into history(time, exid, epoch, iter, mode, batch_size, lr, loss, metrics) values(?, ?, ?, ?, ?, ?, ?, ?, ?)'\n now = datetime.datetime.now()\n data = (now, self.ex_id, epoch, iter, mode,\n batch_size, lr, loss_val, metrics_val)\n cursor.execute(sql, data)\n con.commit()","def history(command):\n namespace = app.main(command)\n assert namespace.command == 'h' or namespace.command == \"history\""],"string":"[\n \"def get_room_history(self, room):\\n pass\",\n \"def create_historical_record(self, instance, history_type):\\n history_date = getattr(instance, '_history_date', now())\\n history_changeset = self.get_history_changeset(instance)\\n manager = getattr(instance, self.manager_name)\\n attrs = {}\\n for field in instance._meta.fields:\\n attrs[field.attname] = getattr(instance, field.attname)\\n\\n for field_name in self.additional_fields:\\n loader = getattr(self, 'get_%s_value' % field_name)\\n value = loader(instance, type)\\n attrs[field_name] = value\\n\\n manager.create(\\n history_date=history_date, history_type=history_type,\\n history_changeset=history_changeset, **attrs)\",\n \"async def _build_room_entry(self, room_id: str) -> JsonDict:\\n stats = await self._store.get_room_with_stats(room_id)\\n\\n # currently this should be impossible because we call\\n # check_user_in_room_or_world_readable on the room before we get here, so\\n # there should always be an entry\\n assert stats is not None, \\\"unable to retrieve stats for %s\\\" % (room_id,)\\n\\n current_state_ids = await self._store.get_current_state_ids(room_id)\\n create_event = await self._store.get_event(\\n current_state_ids[(EventTypes.Create, \\\"\\\")]\\n )\\n\\n # TODO: update once MSC1772 lands\\n room_type = create_event.content.get(EventContentFields.ROOM_TYPE)\\n if not room_type:\\n room_type = create_event.content.get(EventContentFields.MSC1772_ROOM_TYPE)\\n\\n room_version = await self._store.get_room_version(room_id)\\n allowed_spaces = None\\n if await self._event_auth_handler.has_restricted_join_rules(\\n current_state_ids, room_version\\n ):\\n allowed_spaces = await self._event_auth_handler.get_spaces_that_allow_join(\\n current_state_ids\\n )\\n\\n entry = {\\n \\\"room_id\\\": stats[\\\"room_id\\\"],\\n \\\"name\\\": stats[\\\"name\\\"],\\n \\\"topic\\\": stats[\\\"topic\\\"],\\n \\\"canonical_alias\\\": stats[\\\"canonical_alias\\\"],\\n \\\"num_joined_members\\\": stats[\\\"joined_members\\\"],\\n \\\"avatar_url\\\": stats[\\\"avatar\\\"],\\n \\\"world_readable\\\": (\\n stats[\\\"history_visibility\\\"] == HistoryVisibility.WORLD_READABLE\\n ),\\n \\\"guest_can_join\\\": stats[\\\"guest_access\\\"] == \\\"can_join\\\",\\n \\\"creation_ts\\\": create_event.origin_server_ts,\\n \\\"room_type\\\": room_type,\\n \\\"allowed_spaces\\\": allowed_spaces,\\n }\\n\\n # Filter out Nones – rather omit the field altogether\\n room_entry = {k: v for k, v in entry.items() if v is not None}\\n\\n return room_entry\",\n \"def add_history(self):\\n # add separator, if there already are history entries\\n if self.parentApp.History != '':\\n self.parentApp.History += (\\n '\\\\n\\\\n--- --- --- --- --- --- --- --- --- --- --- ---\\\\n\\\\n'\\n )\\n\\n # add the transaction to it\\n self.parentApp.History += self.parentApp.tmpTransC.to_str()\",\n \"def __init__(self, *args):\\n this = _ida_hexrays.new_history_item_t(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def create_test_history(**kw):\\n history = get_test_history(**kw)\\n # Let DB generate ID if it isn't specified explicitly\\n if 'id' not in kw:\\n del history['id']\\n dbapi = db_api.get_instance()\\n return dbapi.create_node_history(history)\",\n \"def addHistory(self):\\r\\n \\r\\n data = self.get()\\r\\n \\r\\n if data == '':\\r\\n return\\r\\n elif len(self.history) != 0 and self.history[0] == data:\\r\\n return\\r\\n \\r\\n if len(self.history) == self.historySize:\\r\\n self.history.pop()\\r\\n \\r\\n self.history.insert(0, data)\",\n \"def create_entry(entry):\\n Entry.create(**entry)\\n return entry\",\n \"def save(self, *args, **kwargs):\\n changesDict = self.diff\\n\\n # Let's construct the history record entry and add it to the dict.\\n # Each record is further going to be dictionary stored in the DB as a string.\\n\\n # We need to keep track of which user updated the data. Check if the user has\\n # assigned a callable to fetch it. Else we will initialize it to be empty.\\n who = None\\n if self.history__get_user_hook is not None:\\n if callable(self.history__get_user_hook):\\n who = self.history__get_user_hook()\\n else:\\n logger.error(\\\"history_get_user initialized with a non-callable\\\")\\n\\n # only add the 'who' field if we have a string value.\\n if who is not None and isinstance(who,str):\\n entry = {\\n \\\"who\\\" : str(who),\\n \\\"when\\\" : timezone.now().isoformat(),\\n \\\"what\\\" : changesDict,\\n }\\n else:\\n entry = {\\n \\\"when\\\" : timezone.now().isoformat(),\\n \\\"what\\\" : changesDict,\\n }\\n\\n # prepend this entry into the list\\n if isinstance(self.history, list):\\n self.history.insert(0,entry)\\n if isinstance(self.history__max_entry_count, int):\\n if self.history__max_entry_count >= 0:\\n # Value has been specified. Trim the list. Value of 0 deletes the whole list.\\n del self.history[self.history__max_entry_count: ]\\n logger.debug('Added the following record to history:\\\\n %s'%(str(entry),))\\n else:\\n logger.error(\\\"History field not saved initially as list!!\\\")\\n\\n # Let's not hold up the save. Save and reset the initial state.\\n super(ChangeHistoryMixin, self).save(*args, **kwargs)\\n self.__initial = self._dict\",\n \"def _push_history(self):\\n self._history.append(self._state)\",\n \"def new_room(request):\\n print(\\\"new room\\\")\\n new_room = None\\n while not new_room:\\n with transaction.atomic():\\n label = haikunator.haikunate()\\n if Room.objects.filter(label=label).exists():\\n continue\\n new_room = Room.objects.create(label=label)\\n return redirect('chat:chat_room', label=label)\",\n \"def append_history_record(self, history_record):\\n self.history.append(history_record)\\n self._increment_history_pointer()\",\n \"def history():\",\n \"def create_and_add_entry(self, **attrs):\\n return self.add_entry(self.create_entry(**attrs))\",\n \"def add_history(self, author, string, data_var=None):\\n\\n attr = 'history'\\n self.add_string(attr, author, string, data_var=data_var)\",\n \"def __build_history(self, obj: Object) -> dict:\\n previous_history = obj.history\\n return {**previous_history, self.__get_timestamp(): {'type_id': obj.type_id, 'redshift': obj.redshift}}\",\n \"def add_entry(self, *args, **kwargs):\\n entry = Entry(*args, **kwargs) # NOTE: not sure this is good\\n self._entries[entry.uuid] = entry\\n return entry\",\n \"def creating_entry(self):\\n response = \\\"\\\"\\n today = str(date.today())\\n curent_time = str(datetime.time(datetime.now()))\\n entry = Diary(self.entry_id, self.title, self.body)\\n lst = {}\\n lst[\\\"entry_id\\\"] = entry.entry_id\\n lst[\\\"title\\\"] = entry.title\\n lst[\\\"date\\\"] = today\\n lst[\\\"time\\\"] = curent_time\\n lst[\\\"body\\\"] = entry.body\\n lst[\\\"updated\\\"] = entry.updated\\n if Validate.validate_entry(Diary.entries, entry):\\n response = jsonify({\\\"message\\\": \\\"Duplicate data,Try again\\\"})\\n response.status_code = 409\\n else:\\n Diary.entries.append(lst)\\n response = jsonify({\\\"message\\\": \\\"Entry saved\\\", \\\"data\\\": lst})\\n response.status_code = 201\\n return response\",\n \"def __build_history(self, obj: Object) -> dict:\\n previous_history = dict(obj.history)\\n return {**previous_history, self.__get_timestamp(): {'type_id': obj.type_id, 'redshift': obj.redshift}}\",\n \"def add_history(self,date,what,note):\\r\\n note = '.'.join(note.split(','))\\r\\n self.history.append([date,what,note])\",\n \"def new_history_entry(self, answer):\\r\\n answer = OpenEndedChild.sanitize_html(answer)\\r\\n self.child_history.append({'answer': answer})\\r\\n self.stored_answer = None\",\n \"def new_room(request):\\n new_room = None\\n while not new_room:\\n with transaction.atomic():\\n label = id_generator(10)\\n if Room.objects.filter(label=label).exists():\\n continue\\n new_room = Room.objects.create(label=label)\\n return redirect(chat_room, label=label)\",\n \"def add_room(self, data):\\n room_id = data['room_id']\\n x, y = literal_eval(data['coordinates'])\\n room_data = {'id': data['room_id'],\\n 'title': data['title'],\\n 'description' : data['description'],\\n 'coordinates': literal_eval(data['coordinates']),\\n 'elevation': data['elevation'],\\n 'terrain': data['terrain'],\\n 'exits' : {direction: '?' for direction in data['exits']}\\n }\\n self.rooms.setdefault(room_id, room_data)\",\n \"def create_base_entry(vin=\\\"INVALID\\\", time_unix=None):\\n\\t\\treturn LogEntry(vin=vin, app_id=\\\"INVALID\\\", time_unix=time_unix)\",\n \"def onRegisterHistory(self):\\n pass\",\n \"def new_entry():\\n clear_screen()\\n entry = {}\\n entry['id'] = get_next_id()\\n entry['name'] = input_name()\\n print(\\\"How many minutes did you spend on {}?\\\".format(entry['name']))\\n print(\\\"Or you may specify a format after the time, seperated by a comma\\\")\\n entry['time_spent'] = input_time_spent()\\n add_notes = input(\\\"Add notes? Y/n \\\").lower()\\n if add_notes != 'n':\\n entry['notes'] = input_notes()\\n entry['date'] = datetime.now().strftime(FMT_MONTH_DAY_YEAR)\\n with open(WORK_LOG_FILENAME, 'a', newline='') as work_log:\\n work_log_writer = csv.DictWriter(work_log, fieldnames=FIELDNAMES)\\n work_log_writer.writerow(entry)\",\n \"def room_add():\\n if not json_param_exists('number'):\\n abort(400, 'invalid room number')\\n\\n if not isinstance(request.json['number'], str):\\n abort(400, 'room number must be string')\\n\\n num = request.json['number']\\n room = Room(number=num)\\n\\n try:\\n get_db().add(room)\\n get_db().commit()\\n except IntegrityError:\\n abort(409, 'room number is already in use')\\n return json.dumps(room.as_dict(include_features=False)), 201\",\n \"def add_to_game_history(self, tile, flag=False):\\n move = {\\n 'tile': tile,\\n 'flag':flag,\\n 'coordinate': self.stack[tile]['coordinate'],\\n 'value': self.stack[tile]['value']}\\n self.history.append(move)\",\n \"def save_record_state_to_history(result=None, **kwargs):\\n\\n # No need to store history array within the history table data\\n data = copy.copy(result)\\n if 'history' in data:\\n del data['history']\\n\\n if 'ratings' in data:\\n del data['ratings']\\n\\n history = RecordHistory(record_id=result['id'], user_id=result['user_id'],\\n data=json.dumps(data))\\n\\n db.session.add(history)\\n db.session.commit()\\n\\n if 'history' not in result:\\n result['history'] = []\\n\\n result['history'].insert(0, to_dict(history))\",\n \"def history():\\n return apology(\\\"TODO\\\")\",\n \"def history():\\n return apology(\\\"TODO\\\")\",\n \"def history():\\n return apology(\\\"TODO\\\")\",\n \"def test_tracker_addHistory():\\n\\n trackers, cap = init_tracker()\\n tr = trackers[0]\\n tr.addHistory([1, 1, 1, 1])\\n\\n assert len(tr.history) >= 1\",\n \"def add_history(self, text, author='', date=None):\\n date = date or datetime.date.today()\\n version = self.header['SRC_V']\\n ttext = '[%s] %s (%s %s)' % (version, text, author, str(date))\\n if len(ttext) > 68:\\n n = len(text) - len(ttext) + 68\\n ttext = '[%s] %s (%s %s)' % (version, text[:n], author, str(date))\\n self.header['HISTORY'] = ttext\",\n \"def instantiate_history(self):\\n serialized_history = self.cache.get('history')\\n history = ast.literal_eval(serialized_history.decode('utf-8'))\\n return history\",\n \"def add_new_entry(self):\\n clear_screen()\\n new_entry = Entry.create()\\n if new_entry is None:\\n print(\\\"Add new entry cancelled. Returning to main menu...\\\")\\n time.sleep(1)\\n return None\\n self.entries.append(new_entry)\\n with open(self.file_name, \\\"a\\\") as file:\\n writer = csv.writer(file)\\n writer.writerow([new_entry.date, new_entry.name, new_entry.minutes, new_entry.note])\",\n \"def _append_history(ds):\\n try:\\n history = ds.attrs['history']\\n except KeyError:\\n history = \\\"\\\"\\n now = datetime.datetime.now()\\n prog = __file__ # os.path.basename(__file__)\\n history = (now.strftime(\\\"%a %b %d %H:%M:%S %Y\\\") +\\n \\\": {} {}\\\\n\\\".format(prog, \\\" \\\".join(sys.argv[1:])) +\\n history)\\n ds.attrs['history'] = history\",\n \"def log_create(sender, instance, created, **kwargs):\\n if created:\\n changes = model_instance_diff(None, instance)\\n\\n log_entry = LogEntry.objects.log_create(\\n instance,\\n action=LogEntry.Action.CREATE,\\n changes=json.dumps(changes),\\n )\\n log_created.send(\\n sender=LogEntry,\\n old_instance=None,\\n new_instance=instance,\\n log_instance=log_entry,\\n )\",\n \"def create(self):\\n logging.warning(\\n \\\"IRC back-end does not support explicit creation, joining room \\\"\\n \\\"instead to ensure it exists.\\\"\\n )\\n self.join()\",\n \"def test_hist_append():\\n FNAME = 'xonsh-SESSIONID.json'\\n FNAME += '.append'\\n hist = History(filename=FNAME, here='yup', **HIST_TEST_KWARGS)\\n with mock_xonsh_env({'HISTCONTROL': set()}):\\n hf = hist.append({'joco': 'still alive'})\\n yield assert_is_none, hf\\n yield assert_equal, 'still alive', hist.buffer[0]['joco']\\n os.remove(FNAME)\",\n \"def _grab_history(self):\\n self.data['history_lines'] = []\\n self.data['history_file'] = None\\n self.data['history_encoding'] = None\\n self.data['headings'] = []\\n self.data['history_last_release'] = ''\\n self.data['history_insert_line_here'] = 0\\n default_location = None\\n config = self.setup_cfg.config\\n if config and config.has_option('zest.releaser', 'history_file'):\\n default_location = config.get('zest.releaser', 'history_file')\\n history_file = self.vcs.history_file(location=default_location)\\n self.data['history_file'] = history_file\\n if not history_file:\\n logger.warn(\\\"No history file found\\\")\\n return\\n logger.debug(\\\"Checking %s\\\", history_file)\\n history_lines, history_encoding = read_text_file(history_file)\\n history_lines = history_lines.split('\\\\n')\\n headings = utils.extract_headings_from_history(history_lines)\\n if not headings:\\n logger.warn(\\\"No detectable version heading in the history \\\"\\n \\\"file %s\\\", history_file)\\n return\\n self.data['history_lines'] = history_lines\\n self.data['history_encoding'] = history_encoding\\n self.data['headings'] = headings\\n\\n # Grab last header.\\n start = headings[0]['line']\\n if len(headings) > 1:\\n # Include the next header plus underline, as this is nice\\n # to show in the history_last_release.\\n end = headings[1]['line'] + 2\\n else:\\n end = len(history_lines)\\n history_last_release = '\\\\n'.join(history_lines[start:end])\\n self.data['history_last_release'] = history_last_release\\n\\n # Add line number where an extra changelog entry can be inserted. Can\\n # be useful for entry points. 'start' is the header, +1 is the\\n # underline, +2 is probably an empty line, so then we should take +3.\\n # Or rather: the first non-empty line.\\n insert = start + 2\\n while insert < end:\\n if history_lines[insert].strip():\\n break\\n insert += 1\\n self.data['history_insert_line_here'] = insert\",\n \"def add_entry(self, scenario_info):\\n scenario_id, status = scenario_info[\\\"id\\\"], \\\"created\\\"\\n sql = self.insert()\\n self.cur.execute(\\n sql,\\n (\\n scenario_id,\\n status,\\n ),\\n )\",\n \"def add_entry(self, scenario_info):\\n scenario_id, status = scenario_info[\\\"id\\\"], \\\"created\\\"\\n sql = self.insert()\\n self.cur.execute(\\n sql,\\n (\\n scenario_id,\\n status,\\n ),\\n )\",\n \"def add_vertex(self, room):\\r\\n if room['room_id'] not in self.rooms:\\r\\n self.rooms[room['room_id']] = room\\r\\n # self.rooms[room['room_id']]['exits'] = {\\r\\n # d: '?' for d in room['exits']}\\r\",\n \"def history():\\n db.execute(\\\"CREATE TABLE IF NOT EXISTS hist(user_id INTEGER NOT NULL, typ TEXT NOT NULL, symbol TEXT NOT NULL, name TEXT NOT NULL, price NUMERIC NOT NULL, nos INTEGER NOT NULL, cost NUMERIC NOT NULL, time DATETIME NOT NULL, FOREIGN KEY(user_id) REFERENCES users(id))\\\")\\n stocks = db.execute(\\\"SELECT * FROM hist WHERE user_id = ?\\\", session[\\\"user_id\\\"])\\n return render_template(\\\"history.html\\\", stocks = stocks)\",\n \"def _write_history(self):\\n if self.data['history_file'] is None:\\n return\\n contents = '\\\\n'.join(self.data['history_lines'])\\n history = self.data['history_file']\\n write_text_file(\\n history, contents, encoding=self.data['history_encoding'])\\n logger.info(\\\"History file %s updated.\\\", history)\",\n \"def test_add_entry_creates_db_entry_with_correct_details(self):\\n example_inputs = [\\n 'Example Employee',\\n '2018-05-01',\\n 'Example Task',\\n 100,\\n 'Example Note'\\n ]\\n with patch('builtins.input', side_effect=example_inputs):\\n self.menu.add_entry()\\n\\n query = (db_manager\\n .LogEntry\\n .select()\\n .join(db_manager.Employee)\\n .where(\\n db_manager.Employee.name == example_inputs[0],\\n db_manager.LogEntry.date == example_inputs[1],\\n db_manager.LogEntry.task_name == example_inputs[2],\\n db_manager.LogEntry.duration == example_inputs[3],\\n db_manager.LogEntry.notes == example_inputs[4]\\n ))\\n record = query[0]\\n record_data = [\\n record.employee.name,\\n record.date.strftime(\\\"%Y-%m-%d\\\"),\\n record.task_name,\\n record.duration,\\n record.notes\\n ]\\n self.assertEqual(example_inputs, record_data)\",\n \"def save_history(sender, instance, **kwargs): # pylint: disable=no-self-argument, unused-argument\\r\\n if instance.module_type in StudentModuleHistory.HISTORY_SAVING_TYPES:\\r\\n history_entry = StudentModuleHistory(student_module=instance,\\r\\n version=None,\\r\\n created=instance.modified,\\r\\n state=instance.state,\\r\\n grade=instance.grade,\\r\\n max_grade=instance.max_grade)\\r\\n history_entry.save()\",\n \"def update_history(self, move):\\r\\n player_number = self.player_numbers[self.current_player]\\r\\n heaps = tuple(self.heaps)\\r\\n self.history.append([player_number, heaps, move])\",\n \"def add_entry(name, title, duration, notes):\\n clear()\\n print('Entry added to work log!')\\n return Entry.create(\\n employee_name=name,\\n task_title=title,\\n time_spent=duration,\\n task_notes=notes\\n )\",\n \"def func(self):\\n\\n caller = self.caller\\n\\n if not self.lhs:\\n string = \\\"Usage: @dig[/teleport] roomname[;alias;alias...][:parent] [= exit_there\\\"\\n string += \\\"[;alias;alias..][:parent]] \\\"\\n string += \\\"[, exit_back_here[;alias;alias..][:parent]]\\\"\\n caller.msg(string)\\n return\\n\\n room = self.lhs_objs[0]\\n\\n if not room[\\\"name\\\"]:\\n caller.msg(\\\"You must supply a new room name.\\\")\\n return\\n location = caller.location\\n\\n # Create the new room\\n typeclass = room[\\\"option\\\"]\\n if not typeclass:\\n typeclass = settings.BASE_ROOM_TYPECLASS\\n\\n # create room\\n lockstring = \\\"control:id(%s) or perm(Immortals); delete:id(%s) or perm(Wizards); edit:id(%s) or perm(Wizards)\\\"\\n lockstring = lockstring % (caller.dbref, caller.dbref, caller.dbref)\\n\\n new_room = create.create_object(\\n typeclass, room[\\\"name\\\"], aliases=room[\\\"aliases\\\"], report_to=caller\\n )\\n new_room.locks.add(lockstring)\\n alias_string = \\\"\\\"\\n if new_room.aliases.all():\\n alias_string = \\\" (%s)\\\" % \\\", \\\".join(new_room.aliases.all())\\n room_string = \\\"Created room %s(%s)%s of type %s.\\\" % (\\n new_room,\\n new_room.dbref,\\n alias_string,\\n typeclass,\\n )\\n\\n # create exit to room\\n\\n exit_to_string = \\\"\\\"\\n exit_back_string = \\\"\\\"\\n\\n if self.rhs_objs:\\n to_exit = self.rhs_objs[0]\\n if not to_exit[\\\"name\\\"]:\\n exit_to_string = \\\"\\\\nNo exit created to new room.\\\"\\n elif not location:\\n exit_to_string = \\\"\\\\nYou cannot create an exit from a None-location.\\\"\\n else:\\n # Build the exit to the new room from the current one\\n typeclass = to_exit[\\\"option\\\"]\\n if not typeclass:\\n typeclass = settings.BASE_EXIT_TYPECLASS\\n\\n new_to_exit = create.create_object(\\n typeclass,\\n to_exit[\\\"name\\\"],\\n location,\\n aliases=to_exit[\\\"aliases\\\"],\\n locks=lockstring,\\n destination=new_room,\\n report_to=caller,\\n )\\n alias_string = \\\"\\\"\\n if new_to_exit.aliases.all():\\n alias_string = \\\" (%s)\\\" % \\\", \\\".join(new_to_exit.aliases.all())\\n exit_to_string = \\\"\\\\nCreated Exit from %s to %s: %s(%s)%s.\\\"\\n exit_to_string = exit_to_string % (\\n location.name,\\n new_room.name,\\n new_to_exit,\\n new_to_exit.dbref,\\n alias_string,\\n )\\n\\n # Create exit back from new room\\n\\n if len(self.rhs_objs) > 1:\\n # Building the exit back to the current room\\n back_exit = self.rhs_objs[1]\\n if not back_exit[\\\"name\\\"]:\\n exit_back_string = \\\"\\\\nNo back exit created.\\\"\\n elif not location:\\n exit_back_string = (\\n \\\"\\\\nYou cannot create an exit back to a None-location.\\\"\\n )\\n else:\\n typeclass = back_exit[\\\"option\\\"]\\n if not typeclass:\\n typeclass = settings.BASE_EXIT_TYPECLASS\\n new_back_exit = create.create_object(\\n typeclass,\\n back_exit[\\\"name\\\"],\\n new_room,\\n aliases=back_exit[\\\"aliases\\\"],\\n locks=lockstring,\\n destination=location,\\n report_to=caller,\\n )\\n alias_string = \\\"\\\"\\n if new_back_exit.aliases.all():\\n alias_string = \\\" (%s)\\\" % \\\", \\\".join(new_back_exit.aliases.all())\\n exit_back_string = \\\"\\\\nCreated Exit back from %s to %s: %s(%s)%s.\\\"\\n exit_back_string = exit_back_string % (\\n new_room.name,\\n location.name,\\n new_back_exit,\\n new_back_exit.dbref,\\n alias_string,\\n )\\n caller.msg(\\\"%s%s%s\\\" % (room_string, exit_to_string, exit_back_string))\\n if new_room and (\\\"teleport\\\" in self.switches or \\\"tel\\\" in self.switches):\\n caller.move_to(new_room)\\n # use property to set colored name attribute and strip ansi markup from key\\n new_room.name = new_room.name\\n return new_room\",\n \"def save_history(cube, field, filename): \\n\\n history.append(cube.attributes['history'])\",\n \"def __init__(self, *args):\\n this = _libsbml.new_ModelHistory(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def history(self, history):\\n self._history = history\",\n \"def save_history():\\n\\n mid = get_mid()\\n back_file = contact_name + \\\"_\\\" + today\\n\\n if not os.path.isdir(back_path):\\n print('WARNING: o {} directory found, creating.').format(back_path)\\n os.mkdir(back_path)\\n else:\\n print(\\\"OK: {} found.\\\".format(back_path))\\n\\n os.chdir(back_path)\\n with open(back_file, 'w') as bf:\\n for mes in get_todays_history(mid):\\n data = \\\"{}\\\\n\\\".format(mes)\\n bf.write(data)\",\n \"async def history(message, client, extra_args):\\n\\n if not extra_args or not (user_id := utils.from_mention(extra_args[0])):\\n user_id = message.author.id\\n\\n @database.query\\n def get_transactions(conn):\\n cursor = conn.cursor()\\n cursor.execute(\\n \\\"SELECT * FROM funnypts WHERE awarder = ? OR awardee = ? ORDER BY date DESC\\\", (user_id, user_id))\\n transactions = cursor.fetchall()\\n cursor.close()\\n conn.close()\\n return transactions\\n\\n if not (transactions := get_transactions()):\\n await message.channel.send(\\\"THIS USER HAS NO HISTORY, THEY SHOULD THOUGH\\\")\\n return\\n\\n @utils.paginated_embeds\\n def populate(embed, entry, entry_number):\\n awarder = client.get_user(entry[0]).name.split(\\\"#\\\", 1)[0]\\n awardee = client.get_user(entry[1]).name.split(\\\"#\\\", 1)[0]\\n transaction = \\\"GIVEN TO\\\" if entry[3] > 0 else \\\"TAKEN FROM\\\"\\n date = entry[4].split(\\\" \\\", 1)[0]\\n reason = \\\"\\\\\\\"{0}\\\\\\\"\\\".format(entry[2])\\n\\n embed.add_field(\\n name=\\\"{0} — {2} — {1} • {3}\\\".format(awarder, awardee, transaction, date), value=reason, inline=False)\\n\\n title = f\\\"{client.get_user(user_id).name}'s FUNNYPOINT HISTORY\\\"\\n embeds = populate(title, transactions, page_length=5)\\n await utils.sauce_pages(embeds, message, client)\",\n \"def from_history_event(expected_remote, h):\\n event = eptStaleEvent()\\n event.expected_remote = expected_remote\\n event.ts = h.ts\\n event.remote = h.remote\\n event.pctag = h.pctag\\n event.encap = h.encap\\n event.flags = h.flags\\n event.intf_id = h.intf_id\\n event.intf_name = h.intf_name\\n event.epg_name = h.epg_name\\n event.vnid_name = h.vnid_name\\n return event\",\n \"def create_entry(cls, title, date, timeSpent, learned, resources):\\n try:\\n with DATABASE.transaction():\\n cls.create(\\n title=title,\\n date=date,\\n timeSpent=timeSpent,\\n learned=learned,\\n resources=resources\\n )\\n except IntegrityError:\\n raise ValueError(\\\"Entry already exists\\\")\",\n \"def history():\\n rows = db.execute(\\\"SELECT * FROM histories WHERE id=:id\\\", id=session[\\\"user_id\\\"])\\n\\n return render_template(\\\"history.html\\\", rows=rows)\",\n \"def get_history():\\n return response_texts_to_entries(make_post_request(HISTORY_API, data={\\\"k\\\": config[\\\"api_key\\\"]}))\",\n \"def update_history(self, guess='', result=''):\\n item = json.dumps({'guess': guess, 'result': result})\\n self.history.append(item)\",\n \"def insertHistoryRecord(self, record, date=None, note=\\\" \\\"):\\n if date == None:\\n date = self.today()\\n \\n if self.historyTableName == None:\\n return []\\n\\n else:\\n \\n fieldValues = list(record) + [date, note]\\n fieldString = \\\"(\\\" + \\\",\\\".join(self.getColumnHeads(self.historyTableName)) + \\\")\\\"\\n questionMarks = \\\"(\\\" + \\\",\\\".join([\\\"?\\\"] * len(fieldValues)) + \\\")\\\"\\n\\n self._db._c.execute(\\\"INSERT INTO \\\" + self.historyTableName +\\n fieldString + \\\" VALUES \\\" + questionMarks,\\n fieldValues)\\n self._db._conn.commit()\\n\\n return fieldValues[0]\",\n \"def record_history_on_commit(self,\\n clocked: 'Clocked',\\n changes: dict,\\n session: orm.Session,\\n timestamp: dt.datetime):\\n new_tick = self._get_new_tick(clocked)\\n\\n new_clock = self.make_clock(timestamp, new_tick)\\n attr = {'entity': clocked}\\n\\n for prop, cls in self.history_models.items():\\n if prop in changes:\\n value = changes[prop]\\n\\n self._cap_previous_history_row(clocked, new_clock, cls)\\n\\n # Add new history row\\n hist = attr.copy()\\n hist[prop.key] = value\\n session.add(\\n cls(\\n vclock=new_clock.vclock,\\n effective=new_clock.effective,\\n **hist,\\n ),\\n )\",\n \"def handle_create_room(self, lobby_command, client_socket):\\n msg = \\\"Handling room creation of {0}\\\".format(lobby_command)\\n print(msg)\\n user = self.clients[client_socket]['data'].decode('utf-8')\\n roomname = lobby_command.split()[1]\\n\\n if roomname == \\\"mine\\\" or roomname == \\\"all\\\":\\n msg = f'Client {user} error! reserved word that cannot be a room name.'\\n self.log_and_send(client_socket, msg)\\n return\\n\\n for room in self.rooms:\\n if room.name == roomname:\\n msg = f\\\"Invalid request from client: {roomname} already exists!\\\"\\n self.log_and_send(client_socket, msg)\\n return\\n\\n self.rooms.append(Room(name=roomname, creator=user))\\n msg = f\\\"Room {roomname} created for client {user} (creator/Admin).\\\"\\n self.log_and_send(client_socket, msg)\\n return\",\n \"def add_room(self, room):\\n self.rooms[room.name] = room\",\n \"def history(self, history):\\n\\n self._history = history\",\n \"def insertRoom(cursor, room: Room) -> None:\\n\\n # TODO: Calculate the values. room object may not contain the creator\\n # or maybe remove that feature\\n cursor.execute(\\n \\\"INSERT INTO rooms VALUES (?, ?, ?)\\\",\\n (roomID, roomName, creator)\\n )\",\n \"def add_activity(self, record):\\n # this brings shame to me but meh, faster to hardcode (see model_activity)\\n # id = str(record.value(\\\"id\\\"))\\n id = record.value(\\\"id\\\")\\n room = record.value(\\\"room\\\")\\n if room == 0:\\n room = \\\"Nenhuma\\\"\\n else:\\n room = str(room)\\n weekday = constants.days_of_the_week[record.value(\\\"weekday\\\")]\\n weektime = record.value(\\\"weektime\\\").toString(\\\"HH:mm\\\")\\n entry = (id, record.value(\\\"description\\\"), room,\\n weekday, weektime)\\n\\n if entry in self._activity_list:\\n return False\\n else:\\n if self.is_in_del_queue(id):\\n # no real need to add as it was queued to be removed\\n self._removed_activities.remove(id)\\n else:\\n self._added_activities.append(id)\\n # print self._added_activities\\n\\n self._activity_list.append(entry)\\n # sorts by day/time\\n self._activity_list.sort(key=operator.itemgetter(3,4))\\n # self._activity_list = sorted(self._activity_list, key=lambda dia_hora: (dia_hora[3], dia_hora[2]))\\n return True\",\n \"def add(\\n description: str = typer.Argument(\\n ...,\\n help=\\\"Description of the log entry\\\"\\n ),\\n date: datetime = typer.Option(\\n datetime.now().strftime(\\\"%Y-%m-%d\\\"), '--date', '-d',\\n help=\\\"Date of the log entry\\\"\\n ),\\n time: datetime = typer.Option(\\n datetime.now().strftime(\\\"%I:%M %p\\\"), '--time', '-t',\\n formats=[\\\"%H:%M:%S\\\", \\\"%I:%M %p\\\"],\\n help=\\\"Time of the log entry\\\"\\n )\\n):\\n log_entry_time = time.time()\\n log_datetime = datetime.combine(date, log_entry_time)\\n\\n manager = LogBookManager()\\n created, message = manager.create(description, log_datetime)\\n\\n if created:\\n typer.echo(\\n typer.style(message, fg=typer.colors.GREEN, bold=True)\\n )\\n else:\\n typer.echo(\\n typer.style(message, fg=typer.colors.RED, bold=True)\\n )\",\n \"def create(data):\\n db = core.connect()\\n theTime = utils.utctime()\\n data[\\\"created\\\"] = theTime\\n data[\\\"modified\\\"] = theTime\\n data[\\\"domain\\\"] = utils.domain(data[\\\"href\\\"])\\n newShift = schema.shift()\\n newShift.update(data)\\n newShift[\\\"type\\\"] = \\\"shift\\\"\\n id = db.create(newShift)\\n newShift = db[id]\\n return joinData(newShift, newShift[\\\"createdBy\\\"])\",\n \"def log_create(action, *args, **kw):\\n from olympia.activity.models import ActivityLog\\n\\n return ActivityLog.create(action, *args, **kw)\",\n \"def create_history_form(self):\\n history_items = [json.loads(item) for item in self.history]\\n\\n history_form_items = []\\n for item in history_items:\\n history_form_items.append(GameHistoryForm(guess=item.get('guess'),\\n result=item.get('result')))\\n\\n return GameHistoryForms(history=history_form_items)\",\n \"def new_room(self):\\r\\n return Room()\",\n \"def new_archive_record(self, event):\\n dbmanager = self.engine.db_binder.get_manager(self.data_binding)\\n dbmanager.addRecord(event.record)\",\n \"def add_entry(self, account):\\n def txn():\\n entry = self.entries.filter('account =', account).get()\\n if not entry:\\n entry = Entry(account=account, parent=self)\\n entry.put()\\n created = True\\n else:\\n created = False\\n return entry, created\\n return db.run_in_transaction(txn)\",\n \"def history():\\n history = db.execute(\\\"SELECT * from history WHERE id=:id\\\", id=session[\\\"user_id\\\"])\\n\\n return render_template(\\\"history.html\\\", history = history)\",\n \"def __insert_history_module(self, form_data):\\n\\n sql_fields = []\\n sql_fields.extend([\\\"`User_ID`=%(user_id)s\\\",\\n \\\"`Moderated`=%(moderated)s\\\",\\n \\\"`Moderator_ID`=%(moderator_id)s\\\",\\n \\\"`Type_ID`=%(type_id)s\\\",\\n \\\"`Country_ID`=%(country_id)s\\\",\\n \\\"`State_ID`=%(state_id)s\\\",\\n \\\"`Accepted`=%(accepted)s\\\"])\\n\\n sql_values = {}\\n\\n parent_plant_id = form_data.get(\\\"Parent_Plant_ID\\\", 0)\\n\\n if int(parent_plant_id) > 0:\\n sql_fields.append(\\\"`Parent_Plant_ID`=%(parent_plant_id)s\\\")\\n sql_values['parent_plant_id'] = parent_plant_id\\n\\n sql_statement = \\\"INSERT INTO History SET \\\" + \\\",\\\".join(sql_fields)\\n\\n sql_values['user_id'] = str(form_data.get('User_ID', 0))\\n sql_values['moderated'] = form_data.get('Moderated', 0)\\n sql_values['moderator_id'] = form_data.get('Moderator_ID', None)\\n sql_values['type_id'] = form_data.get('Type_ID')\\n sql_values['country_id'] = form_data.get('Country_ID')\\n sql_values['state_id'] = form_data.get('State_ID')\\n sql_values['accepted'] = form_data.get('Accepted')\\n\\n session = self.db_conn.cursor()\\n try:\\n session.execute(sql_statement, sql_values)\\n self.db_conn.commit()\\n except Exception:\\n #session.rollback()\\n raise\\n\\n description_id = session.lastrowid\\n\\n if int(parent_plant_id) == 0:\\n parent_update_stmt = \\\"UPDATE HISTORY SET \\\\\\n Parent_Plant_ID=%(parent_plant_id)s \\\\\\n WHERE Description_ID=%(description_id)s\\\"\\n try:\\n session.execute(parent_update_stmt,\\n {'parent_plant_id': description_id,\\n 'description_id': description_id})\\n self.db_conn.commit()\\n except Exception:\\n #session.rollback()\\n raise\\n\\n session.close()\\n return description_id\",\n \"def add(self, record):\\n self._hist_records[record.uid] = record\",\n \"def save_history(self, request, scenario, user):\\n if isinstance(scenario, Scenario):\\n s = scenario.to_dict()\\n else:\\n s = scenario\\n h = History(request=request, scenario=s, user=user)\\n h.save()\\n return h\",\n \"def history():\\n history = db.execute(\\\"SELECT * from history WHERE id=:id\\\", id=session[\\\"user_id\\\"])\\n \\n return render_template(\\\"history.html\\\", history = history)\",\n \"def __init__(self, history=None):\\n\\n self.__history = history if history else []\",\n \"def create_history_application(python_input, original_document):\\n history_mapping = HistoryMapping(python_input.history, original_document)\\n\\n def default_buffer_pos_changed():\\n \\\"\\\"\\\" When the cursor changes in the default buffer. Synchronize with\\n history buffer. \\\"\\\"\\\"\\n # Only when this buffer has the focus.\\n if buffer_mapping.focus_stack[-1] == DEFAULT_BUFFER:\\n try:\\n line_no = default_buffer.document.cursor_position_row - \\\\\\n history_mapping.result_line_offset\\n\\n if line_no < 0: # When the cursor is above the inserted region.\\n raise IndexError\\n\\n history_lineno = sorted(history_mapping.selected_lines)[line_no]\\n except IndexError:\\n pass\\n else:\\n history_buffer.cursor_position = \\\\\\n history_buffer.document.translate_row_col_to_index(history_lineno, 0)\\n\\n def history_buffer_pos_changed():\\n \\\"\\\"\\\" When the cursor changes in the history buffer. Synchronize. \\\"\\\"\\\"\\n # Only when this buffer has the focus.\\n if buffer_mapping.focus_stack[-1] == HISTORY_BUFFER:\\n line_no = history_buffer.document.cursor_position_row\\n\\n if line_no in history_mapping.selected_lines:\\n default_lineno = sorted(history_mapping.selected_lines).index(line_no) + \\\\\\n history_mapping.result_line_offset\\n\\n default_buffer.cursor_position = \\\\\\n default_buffer.document.translate_row_col_to_index(default_lineno, 0)\\n\\n history_buffer = Buffer(\\n initial_document=Document(history_mapping.concatenated_history),\\n on_cursor_position_changed=Callback(history_buffer_pos_changed),\\n accept_action=AcceptAction(\\n lambda cli, buffer: cli.set_return_value(default_buffer.document)),\\n read_only=True)\\n\\n default_buffer = Buffer(\\n initial_document=history_mapping.get_new_document(),\\n on_cursor_position_changed=Callback(default_buffer_pos_changed),\\n read_only=True)\\n\\n help_buffer = Buffer(\\n initial_document=Document(HELP_TEXT, 0),\\n accept_action=AcceptAction.IGNORE,\\n read_only=True\\n )\\n\\n buffer_mapping = BufferMapping({\\n HISTORY_BUFFER: history_buffer,\\n DEFAULT_BUFFER: default_buffer,\\n HELP_BUFFER: help_buffer,\\n }, initial=HISTORY_BUFFER)\\n\\n application = Application(\\n layout=create_layout(python_input, history_mapping),\\n use_alternate_screen=True,\\n buffers=buffer_mapping,\\n style=python_input._current_style,\\n mouse_support=Condition(lambda cli: python_input.enable_mouse_support),\\n key_bindings_registry=create_key_bindings(python_input, history_mapping)\\n )\\n return application\",\n \"def create_new_hist(gi, galaxyemail, galaxypass, server, workflowid, files, new_hist):\\n if workflowid != \\\"0\\\":\\n if len(filter(None, files)) > 0:\\n workflow = gi.workflows.show_workflow(workflowid)\\n if new_hist is None or new_hist == \\\"\\\":\\n new_hist_name = strftime(workflow['name'] + \\\"_%d_%b_%Y_%H:%M:%S\\\", gmtime())\\n else:\\n new_hist_name = new_hist\\n gi.histories.create_history(name=new_hist_name)\\n history_id = get_history_id(galaxyemail, galaxypass, server)\\n else:\\n pass\\n else:\\n if len(filter(None, files)) > 0:\\n if new_hist is None or new_hist == \\\"\\\":\\n new_hist_name = strftime(\\\"Use_Galaxy_%d_%b_%Y_%H:%M:%S\\\", gmtime())\\n else:\\n new_hist_name = new_hist\\n gi.histories.create_history(name=new_hist_name)\\n history_id = get_history_id(galaxyemail, galaxypass, server)\\n else:\\n pass\\n return history_id\",\n \"def logbook_create(name, lb_id=None):\\n return IMPL.logbook_create(name, lb_id)\",\n \"def createRecord(self):\\n self.dto.getRecord().append(self.controller.createNewObj())\\n print(\\\"Record added.\\\")\",\n \"def generate_history(self):\\n self.reporter.generate()\",\n \"def create_history_line(field_name: str, old_value: str, new_value: str) -> str:\\n return 'Updated field \\\"' + field_name + '\\\" from \\\"' + old_value + '\\\" to \\\"' + new_value + '\\\". '\",\n \"def append_to_history(self, message, plugin):\\n datefmt = '%Y-%m-%dT%H:%M:%S.%fZ%z' # 2018-03-27T13:29:12.623222Z\\n now = datetime.datetime.utcnow()\\n log_item = {\\n 'timestamp': now.strftime(datefmt),\\n 'plugin': plugin,\\n 'message': message\\n }\\n self.history_log.append(log_item)\",\n \"def add_room(self, room):\\n self.rooms.append(room)\",\n \"def RecordHistory( self ):\\n if not self.restoringHistory:\\n record = self.activated_node\\n if self.historyIndex < -1:\\n try:\\n del self.history[self.historyIndex+1:]\\n except AttributeError, err:\\n pass\\n if (not self.history) or record != self.history[-1]:\\n self.history.append( record )\\n del self.history[:-200]\\n self.historyIndex = -1\",\n \"def push_history(self, submission, action, program_compiled):\\n\\n history_data = {\\n 'user': self.user.get().name,\\n 'score': self.score,\\n 'action': action,\\n 'submission': submission,\\n 'program_compiled': program_compiled,\\n 'level': self.current_level\\n }\\n taskqueue.add(url='/tasks/push_game_history', params=history_data)\",\n \"def test_create_log(self):\\n message = \\\"Message is {0}\\\".format(random.random())\\n resp = gracedb.writeLog(eventId, message)\\n self.assertEqual(resp.status, 201)\\n new_log_uri = resp.getheader('Location')\\n new_log = resp.json()\\n self.assertEqual(new_log_uri, new_log['self'])\\n check_new_log = gracedb.get(new_log_uri).json()\\n self.assertEqual(check_new_log['comment'], message)\",\n \"def create_new_event(self):\\n pass\",\n \"def _add_change_log(self):\\n wiz = self.machine_email_id\\n change_log = self.env['machine.instance.change_log'].create({\\n 'name': wiz.sub_subject,\\n 'date': wiz.date,\\n 'duration': wiz.duration,\\n 'user_id': wiz.user_id.id,\\n 'priority': wiz.priority,\\n 'machine_instance_id': self.machine_instance_id.id,\\n\\n })\\n self.change_log_id = change_log.id\",\n \"def Modifier_History(self):\\n\\t\\tpass\",\n \"def push(self, value):\\n self.history.append(value)\",\n \"def create_sync_entry(ts, coll, idx):\\n sync_log = connection.ElasticLogs()\\n sync_log.ts = ts\\n sync_log.coll = unicode(coll)\\n sync_log.idx = unicode(idx)\\n sync_log.save()\\n return True\",\n \"def slot_history_changed(self, _sender, _data):\\r\\n last_candle = self.history.last_candle()\\r\\n if last_candle:\\r\\n self.client.history_last_candle = last_candle.tim\",\n \"def rec_history(self, epoch, iter, mode, batch_size, lr, loss_val, metrics_val):\\n with closing(sqlite3.connect(self.dbfile, detect_types=self.time_setting)) as con:\\n cursor = con.cursor()\\n sql = 'insert into history(time, exid, epoch, iter, mode, batch_size, lr, loss, metrics) values(?, ?, ?, ?, ?, ?, ?, ?, ?)'\\n now = datetime.datetime.now()\\n data = (now, self.ex_id, epoch, iter, mode,\\n batch_size, lr, loss_val, metrics_val)\\n cursor.execute(sql, data)\\n con.commit()\",\n \"def history(command):\\n namespace = app.main(command)\\n assert namespace.command == 'h' or namespace.command == \\\"history\\\"\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.66256154","0.62829894","0.62376595","0.6034511","0.6028819","0.6023105","0.6000676","0.5989488","0.59509176","0.5908362","0.582105","0.5779776","0.5768183","0.5764454","0.5743833","0.57111114","0.57085574","0.56982267","0.56964606","0.5662375","0.56458884","0.56297135","0.56260294","0.5588034","0.55702764","0.55458367","0.5520868","0.5493872","0.54862016","0.54689825","0.54689825","0.54689825","0.54568696","0.54542387","0.5453748","0.5451503","0.54512864","0.54448587","0.5429529","0.54254013","0.53988993","0.5395591","0.5395591","0.5378117","0.53693926","0.5367795","0.5364187","0.53639555","0.53633016","0.5361774","0.53531706","0.53390396","0.5336222","0.53259546","0.5315659","0.5314026","0.53024745","0.5296045","0.52640855","0.5253614","0.5252504","0.5248911","0.5240126","0.52377355","0.52372086","0.5235558","0.5215482","0.5214961","0.52084327","0.52049494","0.51991516","0.51968753","0.51909894","0.51836765","0.5176978","0.5173988","0.51705295","0.5154755","0.51519763","0.51399535","0.51394105","0.51377267","0.5130946","0.51291096","0.5123175","0.51121473","0.51036406","0.5096592","0.5085603","0.5084619","0.5082582","0.50799644","0.50732046","0.5067754","0.5062005","0.50597465","0.50514925","0.5046194","0.5028631","0.5023263"],"string":"[\n \"0.66256154\",\n \"0.62829894\",\n \"0.62376595\",\n \"0.6034511\",\n \"0.6028819\",\n \"0.6023105\",\n \"0.6000676\",\n \"0.5989488\",\n \"0.59509176\",\n \"0.5908362\",\n \"0.582105\",\n \"0.5779776\",\n \"0.5768183\",\n \"0.5764454\",\n \"0.5743833\",\n \"0.57111114\",\n \"0.57085574\",\n \"0.56982267\",\n \"0.56964606\",\n \"0.5662375\",\n \"0.56458884\",\n \"0.56297135\",\n \"0.56260294\",\n \"0.5588034\",\n \"0.55702764\",\n \"0.55458367\",\n \"0.5520868\",\n \"0.5493872\",\n \"0.54862016\",\n \"0.54689825\",\n \"0.54689825\",\n \"0.54689825\",\n \"0.54568696\",\n \"0.54542387\",\n \"0.5453748\",\n \"0.5451503\",\n \"0.54512864\",\n \"0.54448587\",\n \"0.5429529\",\n \"0.54254013\",\n \"0.53988993\",\n \"0.5395591\",\n \"0.5395591\",\n \"0.5378117\",\n \"0.53693926\",\n \"0.5367795\",\n \"0.5364187\",\n \"0.53639555\",\n \"0.53633016\",\n \"0.5361774\",\n \"0.53531706\",\n \"0.53390396\",\n \"0.5336222\",\n \"0.53259546\",\n \"0.5315659\",\n \"0.5314026\",\n \"0.53024745\",\n \"0.5296045\",\n \"0.52640855\",\n \"0.5253614\",\n \"0.5252504\",\n \"0.5248911\",\n \"0.5240126\",\n \"0.52377355\",\n \"0.52372086\",\n \"0.5235558\",\n \"0.5215482\",\n \"0.5214961\",\n \"0.52084327\",\n \"0.52049494\",\n \"0.51991516\",\n \"0.51968753\",\n \"0.51909894\",\n \"0.51836765\",\n \"0.5176978\",\n \"0.5173988\",\n \"0.51705295\",\n \"0.5154755\",\n \"0.51519763\",\n \"0.51399535\",\n \"0.51394105\",\n \"0.51377267\",\n \"0.5130946\",\n \"0.51291096\",\n \"0.5123175\",\n \"0.51121473\",\n \"0.51036406\",\n \"0.5096592\",\n \"0.5085603\",\n \"0.5084619\",\n \"0.5082582\",\n \"0.50799644\",\n \"0.50732046\",\n \"0.5067754\",\n \"0.5062005\",\n \"0.50597465\",\n \"0.50514925\",\n \"0.5046194\",\n \"0.5028631\",\n \"0.5023263\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.7202129"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90876,"cells":{"query":{"kind":"string","value":"If the input URL contains a port number, strip it out"},"document":{"kind":"string","value":"def rewrite_url(line):\n url_list = line.split(' ')\n old_url = url_list[0]\n new_url = '\\n'\n # http://192.168.1.13/biedronka_logo.jpg 192.168.1.100/192.168.1.100 - GET myip=- myport=3128 \n if old_url.endswith('jpg') or old_url.endswith('.jpeg'):\n filename = str(uuid.uuid4())+\".jpg\"\n \n wget_cmd = \"/usr/bin/wget -q -O /var/www/html/img/\" +filename+\" \"+ old_url\n log(wget_cmd)\n os.system(wget_cmd)\n \n mogrify_cmd = \"/usr/bin/mogrify -flip /var/www/html/img/\" +filename\n log(mogrify_cmd)\n os.system(mogrify_cmd)\n\n new_url = \"http://127.0.0.1/img/\"+filename+new_url\n log('line: '+line+' | new_url: '+new_url)\n\n return new_url"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def test_remove_empty_port():\n assert (normalize_url(\"http://www.example.com:/\") ==\n \"http://www.example.com/\")","def test_remove_default_port():\n assert (normalize_url(\"http://www.example.com:80/bar.html\") ==\n \"http://www.example.com/bar.html\")\n assert (normalize_url(\"HTTPS://example.com:443/abc/\") ==\n \"https://example.com/abc\")","def strip(url):\r\n split = list(urlsplit(url))\r\n split[4]=''\r\n return urlunsplit(split)","def url_unsplit(parts):\n if parts[2] == default_ports.get(parts[0]):\n return \"%s://%s%s\" % (parts[0], parts[1], parts[3])\n return \"%s://%s:%d%s\" % parts","def extract_port(url):\n port = urlsplit(url).port\n if port is None:\n port = cookiejar.DEFAULT_HTTP_PORT\n return port","def addPortToURL(url, port):\n (scheme, netloc, url, query, fragment) = urllib2.urlparse.urlsplit(url)\n netloc = netloc.split(':')[0]\n netloc = \"%s:%s\" % (netloc, port)\n url = urllib2.urlparse.urlunsplit((scheme, netloc, url, query, fragment))\n return url","def _parseurl(url):\n tracker1=url\n port=int(re.findall(\"[0-9]+\",tracker1)[0])\n host=re.findall(\"[^0-9]+\",tracker1)[0]\n host=host[:-1]\n host=host[6:]\n return host,port","def test_url_port(self):\n assert ct.url_port(\"https://www.bad-actor.services:5000/\") == \"5000\"\n assert ct.url_port(\"https://www.bad-actor.services:502/\") == \"502\"\n assert ct.url_port(\"https://www.bad-actor.services/\") == \"443\"\n assert ct.url_port(\"http://www.bad-actor.services/\") == \"80\"\n assert ct.url_port(\"http://www.bad-actor.services:50200\") == \"50200\"","def api_host_port(hostport):\n if hostport is None:\n return (None, None)\n formatted_host = __host_per_rfc_2732(hostport)\n # The \"bogus\" is to make it look like a real, parseable URL.\n parsed = urlparse(\"bogus://%s\" % (formatted_host)) \n return (None if parsed.hostname == \"none\" else parsed.hostname,\n parsed.port)","def pruneURL(url):\n match = URL_GROUPER.match(url)\n if match is None:\n return url\n else:\n url_parts = match.groupdict()\n protocol = url_parts['protocol']\n if protocol is None:\n protocol = ''\n tail = url_parts['tail']\n if tail is None:\n tail = ''\n return \"%s://%s\" % (protocol, tail)","def _clean_url(self, url):\n return \"\".join(url.split(\"?\")[:1])","def __find_port(self, url):\n match = self.__REGEX_PORT.search(url)\n if match:\n port_num = match.group(0).split(':')[1]\n return port_num\n return None","def clean_url(url):\n o = urlsplit(url)\n return \"{scheme}://{netloc}{path}\".format(\n scheme=o[0], netloc=o[1], path=o[2],\n )","def clean_url(url):\n return url[:url.find('?')]","def clean_url(url):\n return re.sub(';jsessionid=.+\\?', '?', url).strip()","def make_clean_url(url):\n return urlparse.urldefrag(url)[0]","def scrub_url(self, url):\n return self.__url_scrubber(url)","def cut_url(url):\n if len(url) > 50:\n return f\"...{url[-45:]}\"\n return url","def remove_url(tweet):\n return re.sub(r\"http\\S+\", \"URL\", tweet)","def url_truncate(url):\n \n url_tuple = urlparse.urlparse(url)\n return url_tuple[0] + '://' + url_tuple[1]","def _fix_url(url):\n\n if not url.startswith('http'):\n url = 'http://' + url\n\n return url","def url_split(url):\n scheme, netloc = urllib.splittype(url)\n host, document = urllib.splithost(netloc)\n port = default_ports.get(scheme, 0)\n if host:\n host = host.lower()\n host, port = splitport(host, port=port)\n return scheme, host, port, document","def process_url(url: str) -> str:\n split_url = urlsplit(url.strip())\n if split_url.scheme == 'amqp+ssl':\n split_url = split_url._replace(scheme='amqps')\n\n if ((not split_url.username or not split_url.password) and\n 'username' in config and 'password' in config):\n user_pass = f\"{config['username']}:{config['password']}@\"\n new_netloc = user_pass + split_url.netloc\n split_url = split_url._replace(netloc=new_netloc)\n\n return urlunsplit(split_url)","def _strip_version(endpoint):\n if endpoint.endswith('/'):\n endpoint = endpoint[:-1]\n url_bits = endpoint.split('/')\n if re.match(r'v\\d+\\.?\\d*', url_bits[-1]):\n endpoint = '/'.join(url_bits[:-1])\n return endpoint","def _update_url_scheme(self, url):\n if self.base_scheme and not url.startswith(\"%s://\" % self.base_scheme):\n # url_split = urlparse.urlsplit(url)\n url_split = urlsplit(url)\n # url = urlparse.urlunsplit(\n url = urlunsplit(\n [\n self.base_scheme,\n url_split.netloc,\n url_split.path,\n url_split.query,\n url_split.fragment\n ]\n )\n return url","def sanitize_url(url, require_scheme = False):\r\n if not url or ' ' in url:\r\n return\r\n\r\n url = url.strip()\r\n if url.lower() == 'self':\r\n return url\r\n\r\n u = urlparse(url)\r\n # first pass: make sure a scheme has been specified\r\n if not require_scheme and not u.scheme:\r\n url = 'http://' + url\r\n u = urlparse(url)\r\n\r\n if (u.scheme and u.scheme in valid_schemes\r\n and u.hostname and len(u.hostname) < 255\r\n and '%' not in u.netloc):\r\n return url","def remove_URL(sample):\n return re.sub(r\"http\\S+\", \"\", sample)","def normalize_url(node):\n if not node:\n node = DEFAULT_NODE\n elif '://' not in node:\n node = '//{}'.format(node)\n parts = urlparse(node, scheme='http', allow_fragments=False)\n port = parts.port if parts.port else _get_default_port(parts.scheme)\n netloc = '{}:{}'.format(parts.hostname, port)\n return urlunparse((parts.scheme, netloc, parts.path, '', '', ''))","def clean_url(url):\n scheme, netloc, path, query, fragment = url_parse.urlsplit(url)\n path = url_parse.quote(path)\n url = url_parse.urlunsplit((scheme, netloc, path, query, fragment))\n return url","def clean_url(url: str) -> str:\n r = urlparse(url)\n parts = list(r)\n # Add a / to the end of the path if it isn't there\n if not parts[2].endswith(\"/\"):\n parts[2] += \"/\"\n return urlunparse(parts)","def url_fix_common_typos(url):\n if url.startswith(\"http//\"):\n url = \"http://\" + url[6:]\n elif url.startswith(\"https//\"):\n url = \"https://\" + url[7:]\n return url","def process_url(url):\n # only get url path, remove host,params.\n url = urlparse(url).path\n # url = list(url)\n # for i in range(len(url)):\n # if _is_punctuation(url[i]):\n # url[i] = \" \"\n # url = ''.join(url)\n # url = ' '.join(url.split())\n return url","def host_cleanup(host):\n if not host.startswith('https://'):\n host = 'https://' + host # Add schema\n host = strip_end(host, '/')\n host = strip_end(host, '/api/v1')\n host = strip_end(host, '/')\n return host","def deprotocolise(url):\n return PROTORE.sub('', url)","def sanitize_url(url: str, protocol: str = 'https://') -> str:\n sanitized = url[0:-1] if url[-1] == '/' else url\n with_protocol = sanitized if sanitized.startswith('http') else f'{protocol}{sanitized}'\n return with_protocol","def normalize_url(url):\n # print(url)\n if not url.startswith('http://') and not url.startswith('https://'):\n return 'https://{}/{}'.format(zone_name, url.replace('//', '/'))\n return url","def normalize_url(self, url):\n pass","def asinGeturl(url):\n asin = url.split('/')\n for i in asin:\n asinNum = i.strip()\n if len(asinNum) != 10:\n continue\n else:\n asinN = asinNum\n\n return asinN","def sanitizeUrl(url):\n return url.split('?')[0]","def normalize_url(url):\n if not url.startswith((\"git+\", \"hg+\")):\n return url\n return url[4:]","def clean_port(self):\n port = self.cleaned_data['port']\n if not port:\n port = 0\n return port","def format_url(url):\n if not (url.startswith(\"//\") or url.startswith(\"http\")):\n url = \"http://\" + url\n return url","def _current_ip_port(is_secure: bool, host: str, url: str) -> str:\n\n protocol = 'https://' if is_secure else 'http://'\n web_url = protocol + host\n return web_url + url","def splitport(host, port=0):\n if \":\" in host:\n shost, sport = host.split(\":\", 1)\n iport = is_numeric_port(sport)\n if iport:\n host, port = shost, iport\n elif not sport:\n # empty port, ie. the host was \"hostname:\"\n host = shost\n else:\n # For an invalid non-empty port leave the host name as is\n pass\n return host, port","def hostify_url(url):\n\tif url[0] == '/':\n\t\treturn HOST + url\n\telse:\n\t\treturn url","def clean_url(url):\n\n if url is None:\n return None\n\n if '??' in url:\n url = url.split('??')[0]\n\n if url.endswith('?'):\n url = url[:-1]\n\n if '`' in url:\n url = url.replace('`', '')\n\n return url","def extract_scheme(url):\n return urlsplit(url, \"http\").scheme","def test_normalize_xmlrpc_address_missing_port(self):\r\n input_val = 'http://10.20.30.40'\r\n expected_val = 'http://10.20.30.40:1471'\r\n actual_val = normalize_xmlrpc_address(input_val, 1471)\r\n self.assertEqual(expected_val, actual_val)","def _get_event_port(port: Optional[int], event_url: Optional[str]) -> Optional[int]:\n if port is not None:\n return port\n if event_url is not None:\n u = urlparse(event_url)\n return u.port\n return None","def cleanUri(uri):\n if not uri.startswith(\"/\") and not uri.startswith('http'):\n uri = \"/\" + uri\n\n if 'http://' in uri or 'https://' in uri:\n uri = uri.split('://')[0] + '://' + \\\n uri.split('://')[1].replace(\"//\", \"/\")\n else:\n uri = uri.replace(\"//\", \"/\")\n\n if uri.endswith(\"/\"):\n uri = uri[:-1]\n\n return uri","def url_validator_callback(url: str) -> str:\n if url is None:\n return url\n\n url = url.strip()\n try:\n result = urlparse(url)\n if result.scheme and result.netloc:\n return url\n except:\n pass\n raise typer.BadParameter(\"Please supply a valid url\")","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def preprocess_url(self, request_url: str):\n if re.fullmatch('.*/+', request_url) is None:\n return request_url\n else:\n return re.compile(request_url.rstrip('/') + '/+')","def getFilteredUrl(self, url):\n url = url.split('#')[0]\n if url.startswith('/wiki'):\n return ('https://en.wikipedia.org' + url)\n if 'en.wikipedia.org/wiki/' not in url:\n return ('https://en.wikipedia.org/wiki' + url)\n return url","def reflect(host, port):\n global URL\n URL = URL.replace('<ip>', host).replace('<port>', str(port))","def url_prepare(url):\n if 'http://' in url or 'https://' in url:\n return url\n try:\n if requests.get('https://' + url):\n return 'https://' + url\n except Exception as ex:\n pprint(ex)\n return 'http://' + url","def GetServerFromUrl(url):\n return urlunparse((GetSchemeFromUrl(url), GetNetLocFromUrl(url), '', '', '',\n ''))","def format_url(url):\n no_scheme = url.split('://', 1)[-1]\n return '[{0}]({1})'.format(no_scheme, url)","def processUrl(url):\n domain = 'http://www.gsmarena.com/'\n if domain not in url:\n url = urllib.parse.urljoin(domain, url)\n return url","def clean_url(self, url):\n if self.is_blacklisted(url):\n return None\n\n # If the URL was untinyfied we need to start over.\n extracted = self.untiny.extract(url)\n if extracted != url:\n return self.clean_url(extracted)\n\n redirects_to = self.follow_redirects(extracted)\n if not redirects_to:\n return None\n\n # If the URL redirects somewhere else we need to start over.\n if redirects_to != url:\n return self.clean_url(redirects_to)\n\n return self.clean_params(redirects_to)","def parse_url(url, port = 80):\n scheme = url[0:url.find(\"://\")]\n if scheme not in (\\\n 'file', 'ftp', 'gopher', 'hd1', 'http', 'https', \\\n 'imap', 'mailto', 'mms', \\\n 'news', 'nntp', 'prospero', 'rsync', 'rtsp', 'rtspu', \\\n 'sftp', 'shttp', \\\n 'sip', 'sips', 'snews', 'svn', 'svn+ssh', \\\n 'telnet', 'wais'):\n no_scheme = True\n url = url.replace(scheme, 'http', 1)\n else:\n no_scheme = False\n u = urlparse.urlparse(url)\n hasuser = u.netloc.find('@')\n d = {\n 'scheme' : (scheme if no_scheme else u.scheme),\n 'path' : u.path,\n 'query' : u.query,\n 'fragment' : u.fragment,\n 'user' : (u.username if u.username != None else ''),\n 'pass' : (u.password if u.password != None else ''),\n 'port' : (u.port if u.port != None else port),\n 'host' : u.netloc[((hasuser + 1) if (hasuser >= 0) else 0):]\n }\n return d","def clean_link(self, url: str) -> str:\n return self.CLEAN_REGEX.sub(lambda match: f\"%{ord(match.group(0)):02x}\", url)","def relURL(self, url):\n scheme = urlparse(url)[0]\n if self.restricted and scheme and scheme not in self.url_schemes:\n return '#'\n return url","def _decompose_host_port(hp):\n split_re = re.compile(r\"^([^[:][^:]*|\\[[^\\]]+\\])(:([0-9]+))?$\")\n match = split_re.match(hp)\n if match is None:\n return (\"\", None)\n return (match.group(1), match.group(3))","def sanitize_url(self, url):\r\n if not self.markdown.safeMode:\r\n # Return immediately bipassing parsing.\r\n return url\r\n \r\n try:\r\n scheme, netloc, path, params, query, fragment = url = urlparse(url)\r\n except ValueError:\r\n # Bad url - so bad it couldn't be parsed.\r\n return ''\r\n \r\n locless_schemes = ['', 'mailto', 'news']\r\n if netloc == '' and scheme not in locless_schemes:\r\n # This fails regardless of anything else. \r\n # Return immediately to save additional proccessing\r\n return ''\r\n\r\n for part in url[2:]:\r\n if \":\" in part:\r\n # Not a safe url\r\n return ''\r\n\r\n # Url passes all tests. Return url as-is.\r\n return urlunparse(url)","def remove_url(txt):\n\n return \" \".join(re.sub(\"([^0-9A-Za-z \\t])|(\\w+:\\/\\/\\S+)\", \"\", txt).split())","def parse_url(url):\n newurl = urlparse(url)\n return \"{0}://{1}\".format(newurl.scheme, newurl.netloc)","def protocolise(url):\n parsed = urlparse.urlparse(url)\n if not parsed.scheme and not PROTORE.search(url):\n url = 'http://{0}'.format(url)\n return url","def get_host(url):\n parts = url.split('/')\n if url.startswith('http'):\n return parts[2]\n else:\n return parts[0]","def extractHostname(self, url):\n return url.split(\"http://\")[1].split('/')[0]","def remove_url(text):\n return re.sub(r'http\\S+', ' ', text)","def remove_url(txt):\n print(txt['fields']['tweet'])\n return \" \".join(re.sub(\"([^0-9A-Za-z \\t])|(\\w+:\\/\\/\\S+)\", \"\", txt['fields']['tweet']).split())","def normalize_url(url):\n parse = urlparse(url)\n\n # netloc should be lowercase\n netloc = parse.netloc.lower()\n if parse.scheme == \"http\":\n if netloc.endswith(\":80\"):\n netloc = netloc[:-3]\n\n elif parse.scheme == \"https\" and netloc.endswith(\":443\"):\n netloc = netloc[:-4]\n\n # add a '/' at the end of the netloc if there in no path\n if not parse.path:\n netloc = netloc + \"/\"\n\n return \"{}://{}{}\".format(parse.scheme, netloc, parse.path)","def reformat_weburl(s):\n s = (s or '').strip()\n\n if s and '.' in s and 'notavailable' not in s:\n match = re.match(r'^http(s)?://', s)\n if not match:\n s = 'http://' + s\n\n return s","def modLocationPort(self, location):\r\n components = urlparse.urlparse(location)\r\n reverse_proxy_port = self.father.getHost().port\r\n reverse_proxy_host = self.father.getHost().host\r\n # returns an ordered dict of urlparse.ParseResult components\r\n _components = components._asdict()\r\n _components['netloc'] = '%s:%d' % (\r\n reverse_proxy_host, reverse_proxy_port\r\n )\r\n return urlparse.urlunparse(_components.values())","def get_normalized_url(url):\r\n scheme, netloc, path, params, query, fragment = urlparse(url)\r\n\r\n # Exclude default port numbers.\r\n if scheme == 'http' and netloc[-3:] == ':80':\r\n netloc = netloc[:-3]\r\n elif scheme == 'https' and netloc[-4:] == ':443':\r\n netloc = netloc[:-4]\r\n if scheme not in ('http', 'https'):\r\n raise ValueError(\"Unsupported URL %s (%s).\" % (url, scheme))\r\n\r\n # Normalized URL excludes params, query, and fragment.\r\n return urlunparse((scheme, netloc, path, None, None, None))","def remove_url(text):\r\n url = re.sub('https?://[A-Za-z0-9./]+', '', text)\r\n return url","def get_host_name(url):\n return urlparse.urlparse(url)[1]","def parse_port():\n port = 8484\n try:\n port = int(sys.argv[1])\n except Exception as e:\n print(\"CLI argument for port could not be parsed: \" + str(e))\n print(\"Fall back on default port: \" + str(port))\n pass\n return '{}'.format(port)","def test_having_port(self):\n url = 'https://user:12345@domain.com:8080/project/objects.inv'\n expected = 'https://domain.com:8080/project/objects.inv'\n actual = _strip_basic_auth(url)\n assert expected == actual","def _parse_host(host: str) -> str:\n urlparse_host = urlsplit(host).hostname\n if urlparse_host:\n # In this case, host = https://xx.cloud.databricks.com\n return urlparse_host\n else:\n # In this case, host = xx.cloud.databricks.com\n return host","def extract_api_name(url):\n host = RE_HOST.sub('\\\\1', url)\n return host","def _check_ip_port_split(self):\n if self._type == \"A\":\n formatted_value = self._value.split(':')\n self._ip = formatted_value[0]\n self._port = int(formatted_value[1])","def format_url(url: str) -> str:\n return urljoin(url.replace('https://app', 'https://api'), '')","def remove_url(txt):\n\n return \" \".join(re.sub(\"([^0-9A-Za-z \\t])|(\\w+:\\/\\/\\S+)\", \"\", txt).split())","def unparse(self):\r\n # only parse the query params if there is an update dict\r\n q = self.query\r\n if self._url_updates or self._query_dict is not None:\r\n q = self._query_dict or self.query_dict\r\n q.update(self._url_updates)\r\n q = query_string(q).lstrip('?')\r\n\r\n # make sure the port is not doubly specified \r\n if self.port and \":\" in self.hostname:\r\n self.hostname = self.hostname.split(':')[0]\r\n\r\n # if there is a netloc, there had better be a scheme\r\n if self.netloc and not self.scheme:\r\n self.scheme = \"http\"\r\n \r\n return urlunparse((self.scheme, self.netloc,\r\n self.path.replace('//', '/'),\r\n self.params, q, self.fragment))","def make_safe_url(self, url):\n\n # Split the URL into scheme, netloc, path, query and fragment\n parts = list(urlsplit(url))\n\n # Clear scheme and netloc and rebuild URL\n parts[0] = '' # Empty scheme\n parts[1] = '' # Empty netloc (hostname:port)\n safe_url = urlunsplit(parts)\n return safe_url","def item_url(url):\n return all(map(lambda x: str.isdigit(x), str(url.strip('/').split('/')[-1])))","def _urlparse_splitscheme(url):\r\n # The scheme is valid only if it contains these characters.\r\n scheme_chars = \\\r\n \"abcdefghijklmnopqrstuvwxyz0123456789+-.\"\r\n\r\n scheme = \"\"\r\n rest = url\r\n\r\n spart = url.split(\":\", 1)\r\n if len(spart) == 2:\r\n\r\n # Normalize the scheme.\r\n spart[0] = spart[0].lower()\r\n\r\n # A scheme is valid only if it starts with an alpha character.\r\n if spart[0] and spart[0][0].isalpha():\r\n for char in spart[0]:\r\n if char not in scheme_chars:\r\n break\r\n (scheme, rest) = spart\r\n\r\n return scheme, rest","def GetSchemeHostPort(environ):\n url = \"{0}://\".format(environ[\"wsgi.url_scheme\"])\n\n if environ.get(\"HTTP_HOST\"):\n url += environ[\"HTTP_HOST\"]\n else:\n url += environ[\"SERVER_NAME\"]\n\n if environ[\"wsgi.url_scheme\"] == \"https\":\n if environ[\"SERVER_PORT\"] != \"443\":\n url += \":{0}\".format(environ[\"SERVER_PORT\"])\n else:\n if environ[\"SERVER_PORT\"] != \"80\":\n url += \":{0}\".format(environ[\"SERVER_PORT\"])\n\n return url","def truncate_url(url):\n url = parse.unquote(url)\n if len(url) <= 60 :\n return url\n url = url[:-1] if url.endswith(\"/\") else url\n url = url.split(\"//\",1)[1].split(\"/\")\n url = \"%s/.../%s\"%(url[0],url[-1])\n return url[:60]+\"...\" if len(url) > 60 else url","def strip_src_url(self, url: str) -> str:\n strip_url = re.search(self.__URL_BASE_PATTERN, url)\n\n if not strip_url:\n raise ValueError(\"url `%s` does not match handler's base pattern\" % url)\n\n # noinspection PyUnresolvedReferences\n return strip_url[0]","def cleanUrl(url):\n\turl_clean = url.replace(' ','%20')\n\t\"\"\" add /index.html where necessary \"\"\"\n\tif (url[-1:]=='/'):\n\t\turl_clean += 'index.html'\n\telif (url[-5:].find('.') == -1):\n\t\t url_clean += '/index.html'\n\treturn url_clean","def _sanitize_url_prefix(url_prefix: Optional[str]) -> str:\n if not url_prefix:\n return ''\n\n while url_prefix.startswith('//'):\n url_prefix = url_prefix[1:]\n while url_prefix.endswith('/'):\n url_prefix = url_prefix[:-1]\n\n if url_prefix == '':\n return ''\n\n if url_prefix.startswith('/') \\\n or url_prefix.startswith('http://') \\\n or url_prefix.startswith('https://'):\n return url_prefix\n\n return '/' + url_prefix","def process_url(url):\n parsed = urlparse(url)\n if parsed.scheme:\n return parsed.netloc, parsed.path\n else:\n host_part = parsed.path\n hostname = host_part.partition(\"/\")[0]\n path = \"/\" + host_part.partition(\"/\")[2]\n return hostname, path","def prepare_url_to_request(url):\n return url\n # get server url from database\n cursor = Database.cursor()\n cursor.execute(\"SELECT url, id FROM servers ORDER BY used_at ASC LIMIT 1\")\n row = cursor.fetchone()\n server_url = row[0]\n identity = row[1]\n # update server usage time in database\n cursor.execute(\n \"UPDATE servers SET used_at='\" +\n datetime.datetime.today().strftime('%Y-%m-%d %H:%M:%S') +\n \"' where id = '\" + str(identity) + \"'\"\n )\n Database.get_connection().commit()\n return re.sub('^.*?\\.com', server_url, url)"],"string":"[\n \"def test_remove_empty_port():\\n assert (normalize_url(\\\"http://www.example.com:/\\\") ==\\n \\\"http://www.example.com/\\\")\",\n \"def test_remove_default_port():\\n assert (normalize_url(\\\"http://www.example.com:80/bar.html\\\") ==\\n \\\"http://www.example.com/bar.html\\\")\\n assert (normalize_url(\\\"HTTPS://example.com:443/abc/\\\") ==\\n \\\"https://example.com/abc\\\")\",\n \"def strip(url):\\r\\n split = list(urlsplit(url))\\r\\n split[4]=''\\r\\n return urlunsplit(split)\",\n \"def url_unsplit(parts):\\n if parts[2] == default_ports.get(parts[0]):\\n return \\\"%s://%s%s\\\" % (parts[0], parts[1], parts[3])\\n return \\\"%s://%s:%d%s\\\" % parts\",\n \"def extract_port(url):\\n port = urlsplit(url).port\\n if port is None:\\n port = cookiejar.DEFAULT_HTTP_PORT\\n return port\",\n \"def addPortToURL(url, port):\\n (scheme, netloc, url, query, fragment) = urllib2.urlparse.urlsplit(url)\\n netloc = netloc.split(':')[0]\\n netloc = \\\"%s:%s\\\" % (netloc, port)\\n url = urllib2.urlparse.urlunsplit((scheme, netloc, url, query, fragment))\\n return url\",\n \"def _parseurl(url):\\n tracker1=url\\n port=int(re.findall(\\\"[0-9]+\\\",tracker1)[0])\\n host=re.findall(\\\"[^0-9]+\\\",tracker1)[0]\\n host=host[:-1]\\n host=host[6:]\\n return host,port\",\n \"def test_url_port(self):\\n assert ct.url_port(\\\"https://www.bad-actor.services:5000/\\\") == \\\"5000\\\"\\n assert ct.url_port(\\\"https://www.bad-actor.services:502/\\\") == \\\"502\\\"\\n assert ct.url_port(\\\"https://www.bad-actor.services/\\\") == \\\"443\\\"\\n assert ct.url_port(\\\"http://www.bad-actor.services/\\\") == \\\"80\\\"\\n assert ct.url_port(\\\"http://www.bad-actor.services:50200\\\") == \\\"50200\\\"\",\n \"def api_host_port(hostport):\\n if hostport is None:\\n return (None, None)\\n formatted_host = __host_per_rfc_2732(hostport)\\n # The \\\"bogus\\\" is to make it look like a real, parseable URL.\\n parsed = urlparse(\\\"bogus://%s\\\" % (formatted_host)) \\n return (None if parsed.hostname == \\\"none\\\" else parsed.hostname,\\n parsed.port)\",\n \"def pruneURL(url):\\n match = URL_GROUPER.match(url)\\n if match is None:\\n return url\\n else:\\n url_parts = match.groupdict()\\n protocol = url_parts['protocol']\\n if protocol is None:\\n protocol = ''\\n tail = url_parts['tail']\\n if tail is None:\\n tail = ''\\n return \\\"%s://%s\\\" % (protocol, tail)\",\n \"def _clean_url(self, url):\\n return \\\"\\\".join(url.split(\\\"?\\\")[:1])\",\n \"def __find_port(self, url):\\n match = self.__REGEX_PORT.search(url)\\n if match:\\n port_num = match.group(0).split(':')[1]\\n return port_num\\n return None\",\n \"def clean_url(url):\\n o = urlsplit(url)\\n return \\\"{scheme}://{netloc}{path}\\\".format(\\n scheme=o[0], netloc=o[1], path=o[2],\\n )\",\n \"def clean_url(url):\\n return url[:url.find('?')]\",\n \"def clean_url(url):\\n return re.sub(';jsessionid=.+\\\\?', '?', url).strip()\",\n \"def make_clean_url(url):\\n return urlparse.urldefrag(url)[0]\",\n \"def scrub_url(self, url):\\n return self.__url_scrubber(url)\",\n \"def cut_url(url):\\n if len(url) > 50:\\n return f\\\"...{url[-45:]}\\\"\\n return url\",\n \"def remove_url(tweet):\\n return re.sub(r\\\"http\\\\S+\\\", \\\"URL\\\", tweet)\",\n \"def url_truncate(url):\\n \\n url_tuple = urlparse.urlparse(url)\\n return url_tuple[0] + '://' + url_tuple[1]\",\n \"def _fix_url(url):\\n\\n if not url.startswith('http'):\\n url = 'http://' + url\\n\\n return url\",\n \"def url_split(url):\\n scheme, netloc = urllib.splittype(url)\\n host, document = urllib.splithost(netloc)\\n port = default_ports.get(scheme, 0)\\n if host:\\n host = host.lower()\\n host, port = splitport(host, port=port)\\n return scheme, host, port, document\",\n \"def process_url(url: str) -> str:\\n split_url = urlsplit(url.strip())\\n if split_url.scheme == 'amqp+ssl':\\n split_url = split_url._replace(scheme='amqps')\\n\\n if ((not split_url.username or not split_url.password) and\\n 'username' in config and 'password' in config):\\n user_pass = f\\\"{config['username']}:{config['password']}@\\\"\\n new_netloc = user_pass + split_url.netloc\\n split_url = split_url._replace(netloc=new_netloc)\\n\\n return urlunsplit(split_url)\",\n \"def _strip_version(endpoint):\\n if endpoint.endswith('/'):\\n endpoint = endpoint[:-1]\\n url_bits = endpoint.split('/')\\n if re.match(r'v\\\\d+\\\\.?\\\\d*', url_bits[-1]):\\n endpoint = '/'.join(url_bits[:-1])\\n return endpoint\",\n \"def _update_url_scheme(self, url):\\n if self.base_scheme and not url.startswith(\\\"%s://\\\" % self.base_scheme):\\n # url_split = urlparse.urlsplit(url)\\n url_split = urlsplit(url)\\n # url = urlparse.urlunsplit(\\n url = urlunsplit(\\n [\\n self.base_scheme,\\n url_split.netloc,\\n url_split.path,\\n url_split.query,\\n url_split.fragment\\n ]\\n )\\n return url\",\n \"def sanitize_url(url, require_scheme = False):\\r\\n if not url or ' ' in url:\\r\\n return\\r\\n\\r\\n url = url.strip()\\r\\n if url.lower() == 'self':\\r\\n return url\\r\\n\\r\\n u = urlparse(url)\\r\\n # first pass: make sure a scheme has been specified\\r\\n if not require_scheme and not u.scheme:\\r\\n url = 'http://' + url\\r\\n u = urlparse(url)\\r\\n\\r\\n if (u.scheme and u.scheme in valid_schemes\\r\\n and u.hostname and len(u.hostname) < 255\\r\\n and '%' not in u.netloc):\\r\\n return url\",\n \"def remove_URL(sample):\\n return re.sub(r\\\"http\\\\S+\\\", \\\"\\\", sample)\",\n \"def normalize_url(node):\\n if not node:\\n node = DEFAULT_NODE\\n elif '://' not in node:\\n node = '//{}'.format(node)\\n parts = urlparse(node, scheme='http', allow_fragments=False)\\n port = parts.port if parts.port else _get_default_port(parts.scheme)\\n netloc = '{}:{}'.format(parts.hostname, port)\\n return urlunparse((parts.scheme, netloc, parts.path, '', '', ''))\",\n \"def clean_url(url):\\n scheme, netloc, path, query, fragment = url_parse.urlsplit(url)\\n path = url_parse.quote(path)\\n url = url_parse.urlunsplit((scheme, netloc, path, query, fragment))\\n return url\",\n \"def clean_url(url: str) -> str:\\n r = urlparse(url)\\n parts = list(r)\\n # Add a / to the end of the path if it isn't there\\n if not parts[2].endswith(\\\"/\\\"):\\n parts[2] += \\\"/\\\"\\n return urlunparse(parts)\",\n \"def url_fix_common_typos(url):\\n if url.startswith(\\\"http//\\\"):\\n url = \\\"http://\\\" + url[6:]\\n elif url.startswith(\\\"https//\\\"):\\n url = \\\"https://\\\" + url[7:]\\n return url\",\n \"def process_url(url):\\n # only get url path, remove host,params.\\n url = urlparse(url).path\\n # url = list(url)\\n # for i in range(len(url)):\\n # if _is_punctuation(url[i]):\\n # url[i] = \\\" \\\"\\n # url = ''.join(url)\\n # url = ' '.join(url.split())\\n return url\",\n \"def host_cleanup(host):\\n if not host.startswith('https://'):\\n host = 'https://' + host # Add schema\\n host = strip_end(host, '/')\\n host = strip_end(host, '/api/v1')\\n host = strip_end(host, '/')\\n return host\",\n \"def deprotocolise(url):\\n return PROTORE.sub('', url)\",\n \"def sanitize_url(url: str, protocol: str = 'https://') -> str:\\n sanitized = url[0:-1] if url[-1] == '/' else url\\n with_protocol = sanitized if sanitized.startswith('http') else f'{protocol}{sanitized}'\\n return with_protocol\",\n \"def normalize_url(url):\\n # print(url)\\n if not url.startswith('http://') and not url.startswith('https://'):\\n return 'https://{}/{}'.format(zone_name, url.replace('//', '/'))\\n return url\",\n \"def normalize_url(self, url):\\n pass\",\n \"def asinGeturl(url):\\n asin = url.split('/')\\n for i in asin:\\n asinNum = i.strip()\\n if len(asinNum) != 10:\\n continue\\n else:\\n asinN = asinNum\\n\\n return asinN\",\n \"def sanitizeUrl(url):\\n return url.split('?')[0]\",\n \"def normalize_url(url):\\n if not url.startswith((\\\"git+\\\", \\\"hg+\\\")):\\n return url\\n return url[4:]\",\n \"def clean_port(self):\\n port = self.cleaned_data['port']\\n if not port:\\n port = 0\\n return port\",\n \"def format_url(url):\\n if not (url.startswith(\\\"//\\\") or url.startswith(\\\"http\\\")):\\n url = \\\"http://\\\" + url\\n return url\",\n \"def _current_ip_port(is_secure: bool, host: str, url: str) -> str:\\n\\n protocol = 'https://' if is_secure else 'http://'\\n web_url = protocol + host\\n return web_url + url\",\n \"def splitport(host, port=0):\\n if \\\":\\\" in host:\\n shost, sport = host.split(\\\":\\\", 1)\\n iport = is_numeric_port(sport)\\n if iport:\\n host, port = shost, iport\\n elif not sport:\\n # empty port, ie. the host was \\\"hostname:\\\"\\n host = shost\\n else:\\n # For an invalid non-empty port leave the host name as is\\n pass\\n return host, port\",\n \"def hostify_url(url):\\n\\tif url[0] == '/':\\n\\t\\treturn HOST + url\\n\\telse:\\n\\t\\treturn url\",\n \"def clean_url(url):\\n\\n if url is None:\\n return None\\n\\n if '??' in url:\\n url = url.split('??')[0]\\n\\n if url.endswith('?'):\\n url = url[:-1]\\n\\n if '`' in url:\\n url = url.replace('`', '')\\n\\n return url\",\n \"def extract_scheme(url):\\n return urlsplit(url, \\\"http\\\").scheme\",\n \"def test_normalize_xmlrpc_address_missing_port(self):\\r\\n input_val = 'http://10.20.30.40'\\r\\n expected_val = 'http://10.20.30.40:1471'\\r\\n actual_val = normalize_xmlrpc_address(input_val, 1471)\\r\\n self.assertEqual(expected_val, actual_val)\",\n \"def _get_event_port(port: Optional[int], event_url: Optional[str]) -> Optional[int]:\\n if port is not None:\\n return port\\n if event_url is not None:\\n u = urlparse(event_url)\\n return u.port\\n return None\",\n \"def cleanUri(uri):\\n if not uri.startswith(\\\"/\\\") and not uri.startswith('http'):\\n uri = \\\"/\\\" + uri\\n\\n if 'http://' in uri or 'https://' in uri:\\n uri = uri.split('://')[0] + '://' + \\\\\\n uri.split('://')[1].replace(\\\"//\\\", \\\"/\\\")\\n else:\\n uri = uri.replace(\\\"//\\\", \\\"/\\\")\\n\\n if uri.endswith(\\\"/\\\"):\\n uri = uri[:-1]\\n\\n return uri\",\n \"def url_validator_callback(url: str) -> str:\\n if url is None:\\n return url\\n\\n url = url.strip()\\n try:\\n result = urlparse(url)\\n if result.scheme and result.netloc:\\n return url\\n except:\\n pass\\n raise typer.BadParameter(\\\"Please supply a valid url\\\")\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def preprocess_url(self, request_url: str):\\n if re.fullmatch('.*/+', request_url) is None:\\n return request_url\\n else:\\n return re.compile(request_url.rstrip('/') + '/+')\",\n \"def getFilteredUrl(self, url):\\n url = url.split('#')[0]\\n if url.startswith('/wiki'):\\n return ('https://en.wikipedia.org' + url)\\n if 'en.wikipedia.org/wiki/' not in url:\\n return ('https://en.wikipedia.org/wiki' + url)\\n return url\",\n \"def reflect(host, port):\\n global URL\\n URL = URL.replace('<ip>', host).replace('<port>', str(port))\",\n \"def url_prepare(url):\\n if 'http://' in url or 'https://' in url:\\n return url\\n try:\\n if requests.get('https://' + url):\\n return 'https://' + url\\n except Exception as ex:\\n pprint(ex)\\n return 'http://' + url\",\n \"def GetServerFromUrl(url):\\n return urlunparse((GetSchemeFromUrl(url), GetNetLocFromUrl(url), '', '', '',\\n ''))\",\n \"def format_url(url):\\n no_scheme = url.split('://', 1)[-1]\\n return '[{0}]({1})'.format(no_scheme, url)\",\n \"def processUrl(url):\\n domain = 'http://www.gsmarena.com/'\\n if domain not in url:\\n url = urllib.parse.urljoin(domain, url)\\n return url\",\n \"def clean_url(self, url):\\n if self.is_blacklisted(url):\\n return None\\n\\n # If the URL was untinyfied we need to start over.\\n extracted = self.untiny.extract(url)\\n if extracted != url:\\n return self.clean_url(extracted)\\n\\n redirects_to = self.follow_redirects(extracted)\\n if not redirects_to:\\n return None\\n\\n # If the URL redirects somewhere else we need to start over.\\n if redirects_to != url:\\n return self.clean_url(redirects_to)\\n\\n return self.clean_params(redirects_to)\",\n \"def parse_url(url, port = 80):\\n scheme = url[0:url.find(\\\"://\\\")]\\n if scheme not in (\\\\\\n 'file', 'ftp', 'gopher', 'hd1', 'http', 'https', \\\\\\n 'imap', 'mailto', 'mms', \\\\\\n 'news', 'nntp', 'prospero', 'rsync', 'rtsp', 'rtspu', \\\\\\n 'sftp', 'shttp', \\\\\\n 'sip', 'sips', 'snews', 'svn', 'svn+ssh', \\\\\\n 'telnet', 'wais'):\\n no_scheme = True\\n url = url.replace(scheme, 'http', 1)\\n else:\\n no_scheme = False\\n u = urlparse.urlparse(url)\\n hasuser = u.netloc.find('@')\\n d = {\\n 'scheme' : (scheme if no_scheme else u.scheme),\\n 'path' : u.path,\\n 'query' : u.query,\\n 'fragment' : u.fragment,\\n 'user' : (u.username if u.username != None else ''),\\n 'pass' : (u.password if u.password != None else ''),\\n 'port' : (u.port if u.port != None else port),\\n 'host' : u.netloc[((hasuser + 1) if (hasuser >= 0) else 0):]\\n }\\n return d\",\n \"def clean_link(self, url: str) -> str:\\n return self.CLEAN_REGEX.sub(lambda match: f\\\"%{ord(match.group(0)):02x}\\\", url)\",\n \"def relURL(self, url):\\n scheme = urlparse(url)[0]\\n if self.restricted and scheme and scheme not in self.url_schemes:\\n return '#'\\n return url\",\n \"def _decompose_host_port(hp):\\n split_re = re.compile(r\\\"^([^[:][^:]*|\\\\[[^\\\\]]+\\\\])(:([0-9]+))?$\\\")\\n match = split_re.match(hp)\\n if match is None:\\n return (\\\"\\\", None)\\n return (match.group(1), match.group(3))\",\n \"def sanitize_url(self, url):\\r\\n if not self.markdown.safeMode:\\r\\n # Return immediately bipassing parsing.\\r\\n return url\\r\\n \\r\\n try:\\r\\n scheme, netloc, path, params, query, fragment = url = urlparse(url)\\r\\n except ValueError:\\r\\n # Bad url - so bad it couldn't be parsed.\\r\\n return ''\\r\\n \\r\\n locless_schemes = ['', 'mailto', 'news']\\r\\n if netloc == '' and scheme not in locless_schemes:\\r\\n # This fails regardless of anything else. \\r\\n # Return immediately to save additional proccessing\\r\\n return ''\\r\\n\\r\\n for part in url[2:]:\\r\\n if \\\":\\\" in part:\\r\\n # Not a safe url\\r\\n return ''\\r\\n\\r\\n # Url passes all tests. Return url as-is.\\r\\n return urlunparse(url)\",\n \"def remove_url(txt):\\n\\n return \\\" \\\".join(re.sub(\\\"([^0-9A-Za-z \\\\t])|(\\\\w+:\\\\/\\\\/\\\\S+)\\\", \\\"\\\", txt).split())\",\n \"def parse_url(url):\\n newurl = urlparse(url)\\n return \\\"{0}://{1}\\\".format(newurl.scheme, newurl.netloc)\",\n \"def protocolise(url):\\n parsed = urlparse.urlparse(url)\\n if not parsed.scheme and not PROTORE.search(url):\\n url = 'http://{0}'.format(url)\\n return url\",\n \"def get_host(url):\\n parts = url.split('/')\\n if url.startswith('http'):\\n return parts[2]\\n else:\\n return parts[0]\",\n \"def extractHostname(self, url):\\n return url.split(\\\"http://\\\")[1].split('/')[0]\",\n \"def remove_url(text):\\n return re.sub(r'http\\\\S+', ' ', text)\",\n \"def remove_url(txt):\\n print(txt['fields']['tweet'])\\n return \\\" \\\".join(re.sub(\\\"([^0-9A-Za-z \\\\t])|(\\\\w+:\\\\/\\\\/\\\\S+)\\\", \\\"\\\", txt['fields']['tweet']).split())\",\n \"def normalize_url(url):\\n parse = urlparse(url)\\n\\n # netloc should be lowercase\\n netloc = parse.netloc.lower()\\n if parse.scheme == \\\"http\\\":\\n if netloc.endswith(\\\":80\\\"):\\n netloc = netloc[:-3]\\n\\n elif parse.scheme == \\\"https\\\" and netloc.endswith(\\\":443\\\"):\\n netloc = netloc[:-4]\\n\\n # add a '/' at the end of the netloc if there in no path\\n if not parse.path:\\n netloc = netloc + \\\"/\\\"\\n\\n return \\\"{}://{}{}\\\".format(parse.scheme, netloc, parse.path)\",\n \"def reformat_weburl(s):\\n s = (s or '').strip()\\n\\n if s and '.' in s and 'notavailable' not in s:\\n match = re.match(r'^http(s)?://', s)\\n if not match:\\n s = 'http://' + s\\n\\n return s\",\n \"def modLocationPort(self, location):\\r\\n components = urlparse.urlparse(location)\\r\\n reverse_proxy_port = self.father.getHost().port\\r\\n reverse_proxy_host = self.father.getHost().host\\r\\n # returns an ordered dict of urlparse.ParseResult components\\r\\n _components = components._asdict()\\r\\n _components['netloc'] = '%s:%d' % (\\r\\n reverse_proxy_host, reverse_proxy_port\\r\\n )\\r\\n return urlparse.urlunparse(_components.values())\",\n \"def get_normalized_url(url):\\r\\n scheme, netloc, path, params, query, fragment = urlparse(url)\\r\\n\\r\\n # Exclude default port numbers.\\r\\n if scheme == 'http' and netloc[-3:] == ':80':\\r\\n netloc = netloc[:-3]\\r\\n elif scheme == 'https' and netloc[-4:] == ':443':\\r\\n netloc = netloc[:-4]\\r\\n if scheme not in ('http', 'https'):\\r\\n raise ValueError(\\\"Unsupported URL %s (%s).\\\" % (url, scheme))\\r\\n\\r\\n # Normalized URL excludes params, query, and fragment.\\r\\n return urlunparse((scheme, netloc, path, None, None, None))\",\n \"def remove_url(text):\\r\\n url = re.sub('https?://[A-Za-z0-9./]+', '', text)\\r\\n return url\",\n \"def get_host_name(url):\\n return urlparse.urlparse(url)[1]\",\n \"def parse_port():\\n port = 8484\\n try:\\n port = int(sys.argv[1])\\n except Exception as e:\\n print(\\\"CLI argument for port could not be parsed: \\\" + str(e))\\n print(\\\"Fall back on default port: \\\" + str(port))\\n pass\\n return '{}'.format(port)\",\n \"def test_having_port(self):\\n url = 'https://user:12345@domain.com:8080/project/objects.inv'\\n expected = 'https://domain.com:8080/project/objects.inv'\\n actual = _strip_basic_auth(url)\\n assert expected == actual\",\n \"def _parse_host(host: str) -> str:\\n urlparse_host = urlsplit(host).hostname\\n if urlparse_host:\\n # In this case, host = https://xx.cloud.databricks.com\\n return urlparse_host\\n else:\\n # In this case, host = xx.cloud.databricks.com\\n return host\",\n \"def extract_api_name(url):\\n host = RE_HOST.sub('\\\\\\\\1', url)\\n return host\",\n \"def _check_ip_port_split(self):\\n if self._type == \\\"A\\\":\\n formatted_value = self._value.split(':')\\n self._ip = formatted_value[0]\\n self._port = int(formatted_value[1])\",\n \"def format_url(url: str) -> str:\\n return urljoin(url.replace('https://app', 'https://api'), '')\",\n \"def remove_url(txt):\\n\\n return \\\" \\\".join(re.sub(\\\"([^0-9A-Za-z \\\\t])|(\\\\w+:\\\\/\\\\/\\\\S+)\\\", \\\"\\\", txt).split())\",\n \"def unparse(self):\\r\\n # only parse the query params if there is an update dict\\r\\n q = self.query\\r\\n if self._url_updates or self._query_dict is not None:\\r\\n q = self._query_dict or self.query_dict\\r\\n q.update(self._url_updates)\\r\\n q = query_string(q).lstrip('?')\\r\\n\\r\\n # make sure the port is not doubly specified \\r\\n if self.port and \\\":\\\" in self.hostname:\\r\\n self.hostname = self.hostname.split(':')[0]\\r\\n\\r\\n # if there is a netloc, there had better be a scheme\\r\\n if self.netloc and not self.scheme:\\r\\n self.scheme = \\\"http\\\"\\r\\n \\r\\n return urlunparse((self.scheme, self.netloc,\\r\\n self.path.replace('//', '/'),\\r\\n self.params, q, self.fragment))\",\n \"def make_safe_url(self, url):\\n\\n # Split the URL into scheme, netloc, path, query and fragment\\n parts = list(urlsplit(url))\\n\\n # Clear scheme and netloc and rebuild URL\\n parts[0] = '' # Empty scheme\\n parts[1] = '' # Empty netloc (hostname:port)\\n safe_url = urlunsplit(parts)\\n return safe_url\",\n \"def item_url(url):\\n return all(map(lambda x: str.isdigit(x), str(url.strip('/').split('/')[-1])))\",\n \"def _urlparse_splitscheme(url):\\r\\n # The scheme is valid only if it contains these characters.\\r\\n scheme_chars = \\\\\\r\\n \\\"abcdefghijklmnopqrstuvwxyz0123456789+-.\\\"\\r\\n\\r\\n scheme = \\\"\\\"\\r\\n rest = url\\r\\n\\r\\n spart = url.split(\\\":\\\", 1)\\r\\n if len(spart) == 2:\\r\\n\\r\\n # Normalize the scheme.\\r\\n spart[0] = spart[0].lower()\\r\\n\\r\\n # A scheme is valid only if it starts with an alpha character.\\r\\n if spart[0] and spart[0][0].isalpha():\\r\\n for char in spart[0]:\\r\\n if char not in scheme_chars:\\r\\n break\\r\\n (scheme, rest) = spart\\r\\n\\r\\n return scheme, rest\",\n \"def GetSchemeHostPort(environ):\\n url = \\\"{0}://\\\".format(environ[\\\"wsgi.url_scheme\\\"])\\n\\n if environ.get(\\\"HTTP_HOST\\\"):\\n url += environ[\\\"HTTP_HOST\\\"]\\n else:\\n url += environ[\\\"SERVER_NAME\\\"]\\n\\n if environ[\\\"wsgi.url_scheme\\\"] == \\\"https\\\":\\n if environ[\\\"SERVER_PORT\\\"] != \\\"443\\\":\\n url += \\\":{0}\\\".format(environ[\\\"SERVER_PORT\\\"])\\n else:\\n if environ[\\\"SERVER_PORT\\\"] != \\\"80\\\":\\n url += \\\":{0}\\\".format(environ[\\\"SERVER_PORT\\\"])\\n\\n return url\",\n \"def truncate_url(url):\\n url = parse.unquote(url)\\n if len(url) <= 60 :\\n return url\\n url = url[:-1] if url.endswith(\\\"/\\\") else url\\n url = url.split(\\\"//\\\",1)[1].split(\\\"/\\\")\\n url = \\\"%s/.../%s\\\"%(url[0],url[-1])\\n return url[:60]+\\\"...\\\" if len(url) > 60 else url\",\n \"def strip_src_url(self, url: str) -> str:\\n strip_url = re.search(self.__URL_BASE_PATTERN, url)\\n\\n if not strip_url:\\n raise ValueError(\\\"url `%s` does not match handler's base pattern\\\" % url)\\n\\n # noinspection PyUnresolvedReferences\\n return strip_url[0]\",\n \"def cleanUrl(url):\\n\\turl_clean = url.replace(' ','%20')\\n\\t\\\"\\\"\\\" add /index.html where necessary \\\"\\\"\\\"\\n\\tif (url[-1:]=='/'):\\n\\t\\turl_clean += 'index.html'\\n\\telif (url[-5:].find('.') == -1):\\n\\t\\t url_clean += '/index.html'\\n\\treturn url_clean\",\n \"def _sanitize_url_prefix(url_prefix: Optional[str]) -> str:\\n if not url_prefix:\\n return ''\\n\\n while url_prefix.startswith('//'):\\n url_prefix = url_prefix[1:]\\n while url_prefix.endswith('/'):\\n url_prefix = url_prefix[:-1]\\n\\n if url_prefix == '':\\n return ''\\n\\n if url_prefix.startswith('/') \\\\\\n or url_prefix.startswith('http://') \\\\\\n or url_prefix.startswith('https://'):\\n return url_prefix\\n\\n return '/' + url_prefix\",\n \"def process_url(url):\\n parsed = urlparse(url)\\n if parsed.scheme:\\n return parsed.netloc, parsed.path\\n else:\\n host_part = parsed.path\\n hostname = host_part.partition(\\\"/\\\")[0]\\n path = \\\"/\\\" + host_part.partition(\\\"/\\\")[2]\\n return hostname, path\",\n \"def prepare_url_to_request(url):\\n return url\\n # get server url from database\\n cursor = Database.cursor()\\n cursor.execute(\\\"SELECT url, id FROM servers ORDER BY used_at ASC LIMIT 1\\\")\\n row = cursor.fetchone()\\n server_url = row[0]\\n identity = row[1]\\n # update server usage time in database\\n cursor.execute(\\n \\\"UPDATE servers SET used_at='\\\" +\\n datetime.datetime.today().strftime('%Y-%m-%d %H:%M:%S') +\\n \\\"' where id = '\\\" + str(identity) + \\\"'\\\"\\n )\\n Database.get_connection().commit()\\n return re.sub('^.*?\\\\.com', server_url, url)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.7776715","0.7397051","0.6689824","0.6589313","0.650951","0.650275","0.6461937","0.6428155","0.64266","0.6361706","0.6282323","0.62341475","0.61968756","0.6188919","0.6155203","0.6127684","0.61132604","0.60064256","0.59795815","0.5977992","0.5977227","0.5968989","0.5967388","0.5922773","0.59153837","0.59056395","0.5902812","0.5899662","0.5897696","0.5895561","0.58948076","0.58918256","0.588476","0.5881303","0.58743846","0.58669955","0.5844319","0.58424795","0.5834531","0.5817526","0.5770801","0.57540065","0.5746494","0.57054776","0.5693405","0.56881726","0.56841534","0.5681658","0.56748194","0.56720793","0.56575865","0.5633133","0.5633133","0.5633133","0.5633133","0.5633133","0.5633133","0.5633133","0.56283706","0.5623719","0.56173414","0.56055784","0.56041354","0.56035024","0.5588007","0.5585122","0.5563702","0.5559533","0.5548687","0.5533714","0.5525886","0.55220413","0.5521912","0.55078715","0.5502418","0.5488507","0.5485783","0.5480068","0.54633296","0.54504395","0.54490066","0.5448046","0.5446375","0.5436536","0.5429912","0.54224503","0.54173243","0.54086995","0.54085535","0.54083127","0.5405399","0.5403679","0.5394769","0.5388446","0.5382798","0.5380738","0.53791606","0.53679734","0.5366008","0.53614205","0.53598046"],"string":"[\n \"0.7776715\",\n \"0.7397051\",\n \"0.6689824\",\n \"0.6589313\",\n \"0.650951\",\n \"0.650275\",\n \"0.6461937\",\n \"0.6428155\",\n \"0.64266\",\n \"0.6361706\",\n \"0.6282323\",\n \"0.62341475\",\n \"0.61968756\",\n \"0.6188919\",\n \"0.6155203\",\n \"0.6127684\",\n \"0.61132604\",\n \"0.60064256\",\n \"0.59795815\",\n \"0.5977992\",\n \"0.5977227\",\n \"0.5968989\",\n \"0.5967388\",\n \"0.5922773\",\n \"0.59153837\",\n \"0.59056395\",\n \"0.5902812\",\n \"0.5899662\",\n \"0.5897696\",\n \"0.5895561\",\n \"0.58948076\",\n \"0.58918256\",\n \"0.588476\",\n \"0.5881303\",\n \"0.58743846\",\n \"0.58669955\",\n \"0.5844319\",\n \"0.58424795\",\n \"0.5834531\",\n \"0.5817526\",\n \"0.5770801\",\n \"0.57540065\",\n \"0.5746494\",\n \"0.57054776\",\n \"0.5693405\",\n \"0.56881726\",\n \"0.56841534\",\n \"0.5681658\",\n \"0.56748194\",\n \"0.56720793\",\n \"0.56575865\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.5633133\",\n \"0.56283706\",\n \"0.5623719\",\n \"0.56173414\",\n \"0.56055784\",\n \"0.56041354\",\n \"0.56035024\",\n \"0.5588007\",\n \"0.5585122\",\n \"0.5563702\",\n \"0.5559533\",\n \"0.5548687\",\n \"0.5533714\",\n \"0.5525886\",\n \"0.55220413\",\n \"0.5521912\",\n \"0.55078715\",\n \"0.5502418\",\n \"0.5488507\",\n \"0.5485783\",\n \"0.5480068\",\n \"0.54633296\",\n \"0.54504395\",\n \"0.54490066\",\n \"0.5448046\",\n \"0.5446375\",\n \"0.5436536\",\n \"0.5429912\",\n \"0.54224503\",\n \"0.54173243\",\n \"0.54086995\",\n \"0.54085535\",\n \"0.54083127\",\n \"0.5405399\",\n \"0.5403679\",\n \"0.5394769\",\n \"0.5388446\",\n \"0.5382798\",\n \"0.5380738\",\n \"0.53791606\",\n \"0.53679734\",\n \"0.5366008\",\n \"0.53614205\",\n \"0.53598046\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90877,"cells":{"query":{"kind":"string","value":"MLP Block in Transformer Encoder"},"document":{"kind":"string","value":"def __init__(self, hidden_layers, dropout=0.1, activation='gelu'):\n super(MLPBlock, self).__init__()\n\n layers = []\n for num_units in hidden_layers:\n layers.extend([\n Dense(num_units, activation=activation),\n Dropout(dropout)\n ])\n\n self.mlp = Sequential(layers)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\n super(TransformerEncoder, self).__init__()\n\n # Create num_layers of TransformerBlock\n self.encoder = Sequential(\n [\n TransformerBlock(num_heads=num_heads,\n D=D,\n hidden_layers=[mlp_dim, D],\n dropout=dropout,\n norm_eps=norm_eps)\n for _ in range(num_layers)\n ]\n )","def forward(self, x):\n # Encoder1 --block1\n encode_block1 = self.conv_encoder1(x)\n if self.residus[0] == 1:\n encode_block1 += self.residual_shortcut1(x)\n encode_pool1 = self.max_pool_encoder1(encode_block1)\n\n # Encoder2 --block2\n encode_block2 = self.conv_encoder2(encode_pool1)\n if self.residus[1] == 1:\n encode_block2 += self.residual_shortcut2(encode_pool1)\n encode_pool2 = self.max_pool_encoder2(encode_block2)\n \n # Encoder3 --block3\n encode_block3 = self.conv_encoder3(encode_pool2)\n if self.residus[2] == 1:\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\n encode_pool3 = self.max_pool_encoder3(encode_block3)\n \n # Encoder4 --block4\n encode_block4 = self.conv_encoder4(encode_pool3) \n if self.residus[3] == 1:\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\n encode_pool4 = self.max_pool_encoder4(encode_block4)\n\n # Transitional block \n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\n \n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\n \n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\n\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\n \n convTranspose_transitional = self.convTranspose_transitional(middle_block) \n # Decoder4 --block5\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \n if self.residus[4] == 1:\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\n\n #--block6\n cat_layer3 = self.conv_decoder4(decode_block4) \n if self.residus[5] == 1:\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\n \n \n # Decoder3 --block7\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\n if self.residus[6] == 1:\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\n \n #--block8\n cat_layer2 = self.conv_decoder3(decode_block3) \n if self.residus[7] == 1:\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\n \n # Decoder2 --block9\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \n if self.residus[8] == 1:\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\n \n #--block10\n cat_layer1 = self.conv_decoder2(decode_block2)\n if self.residus[9] == 1:\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\n \n # Decoder1 --block11\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \n if self.residus[10] == 1:\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\n \n #--block12\n final_layer = self.final_layer(decode_block1)\n if self.residus[11] == 1:\n final_layer += self.residual_shortcut_final_layer(decode_block1)\n \n \n return final_layer","def __init__(self, unified_encoder, mlp):\r\n\r\n super().__init__()\r\n\r\n self.unified_encoder = unified_encoder\r\n self.mlp = mlp","def TransformerEncoderLayer(self, name, is_causal=True):\n # Hack to be compatible with ckpt generated by self._rep\n return self._Seq(\n name,\n self._Seq('block', self._Attention('self_atten', is_causal=is_causal),\n self.Feedforward('ff')))","def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):\n super(TransformerBlock, self).__init__()\n # Attention\n self.attention = MultiHeadAttention(\n num_heads=num_heads, key_dim=D, dropout=dropout\n )\n self.norm_attention = LayerNormalization(epsilon=norm_eps)\n\n # MLP\n self.mlp = MLPBlock(hidden_layers, dropout)\n self.norm_mlp = LayerNormalization(epsilon=norm_eps)","def __init__(self, enc, mlp):\r\n\r\n super().__init__()\r\n\r\n self.enc = enc\r\n self.mlp = mlp","def transform(self):","def __init__(self, dm, h, hidden, drop_rate=0.1):\r\n super(EncoderBlock, self).__init__()\r\n self.mha = MultiHeadAttention(dm, h)\r\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\r\n activation='relu')\r\n self.dense_output = tf.keras.layers.Dense(units=dm)\r\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\r\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)","def forward(self, x):\n # Encode\n encode_block1 = self.conv_encoder1(x)\n encode_pool1 = self.max_pool_encoder1(encode_block1)\n encode_block2 = self.conv_encoder2(encode_pool1)\n encode_pool2 = self.max_pool_encoder2(encode_block2)\n encode_block3 = self.conv_encoder3(encode_pool2)\n encode_pool3 = self.max_pool_encoder3(encode_block3)\n encode_block4 = self.conv_encoder4(encode_pool3)\n encode_pool4 = self.max_pool_encoder4(encode_block4)\n\n # Transitional block\n middle_block = self.transitional_block(encode_pool4)\n\n # Decode\n decode_block4 = torch.cat((middle_block, encode_block4), 1)\n cat_layer3 = self.conv_decoder4(decode_block4)\n decode_block3 = torch.cat((cat_layer3, encode_block3), 1)\n cat_layer2 = self.conv_decoder3(decode_block3)\n decode_block2 = torch.cat((cat_layer2, encode_block2), 1)\n cat_layer1 = self.conv_decoder2(decode_block2)\n decode_block1 = torch.cat((cat_layer1, encode_block1), 1)\n final_layer = self.final_layer(decode_block1)\n return final_layer","def mtf_bitransformer_base():\n hparams = mtf_transformer2_base()\n hparams.max_length = 256\n hparams.shared_embedding = True\n # HYPERPARAMETERS FOR THE LAYER STACKS\n hparams.add_hparam(\"encoder_layers\", [\"self_att\", \"drd\"] * 6)\n hparams.add_hparam(\"decoder_layers\", [\"self_att\", \"enc_att\", \"drd\"] * 6)\n hparams.add_hparam(\"encoder_num_layers\", 6)\n hparams.add_hparam(\"decoder_num_layers\", 6)\n # number of heads in multihead attention\n hparams.add_hparam(\"encoder_num_heads\", 8)\n hparams.add_hparam(\"decoder_num_heads\", 8)\n hparams.add_hparam(\"local_attention_radius\", 128)\n\n # default of 0 for standard transformer behavior\n # 1 means a single set of keys and values that are read by all query heads\n hparams.add_hparam(\"encoder_num_memory_heads\", 0)\n hparams.add_hparam(\"decoder_num_memory_heads\", 0)\n # share attention keys and values\n hparams.add_hparam(\"encoder_shared_kv\", False)\n hparams.add_hparam(\"decoder_shared_kv\", False)\n\n # Parameters for computing the maximum decode length in beam search.\n # Maximum decode length is:\n # min(max_length,\n # decode_length_multiplier * input_length + decode_length_constant)\n hparams.add_hparam(\"decode_length_multiplier\", 1.5)\n hparams.add_hparam(\"decode_length_constant\", 10.0)\n # used during decoding\n hparams.add_hparam(\"alpha\", 0.6)\n hparams.sampling_temp = 0.0\n return hparams","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n CustomTransformerBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n backward_compatible=self.backward_compatible,\n float_type=self.float_type,\n name=(\"layer_%d\" % i)))\n super(CustomTransformer, self).build(unused_input_shapes)","def __init__(self, momentum: float = .5):\n super(VanillaEncoder, self).__init__()\n self.conv1 = PointNetConv2Layer(64, momentum)\n self.conv2 = PointNetConv2Layer(64, momentum)\n self.conv3 = PointNetConv2Layer(64, momentum)\n self.conv4 = PointNetConv2Layer(128, momentum)\n self.conv5 = PointNetConv2Layer(1024, momentum)","def __init__(self, ):\n super().__init__()\n channels = 3\n\n # Initial convolution block\n out_features = 64\n # encoder\n self.input = nn.Sequential(\n nn.ReflectionPad2d(channels),\n nn.Conv2d(3, out_features, (7, 7)),\n nn.InstanceNorm2d(out_features),\n nn.MaxPool2d(2),\n nn.LeakyReLU(0.2, inplace=True),\n nn.Conv2d(out_features, out_features * 2, (3, 3), stride=(1, 1), padding=(1, 1)),\n nn.LeakyReLU(0.2, inplace=True),\n nn.MaxPool2d(2),\n nn.Conv2d(out_features * 2, out_features * 4, (3, 3), stride=(1, 1), padding=(1, 1)),\n nn.LeakyReLU(0.2, inplace=True),\n )\n\n self.in0 = nn.InstanceNorm2d(256)\n self.block0 = blocks()\n self.block1 = blocks()\n self.block2 = blocks()\n self.block3 = blocks()\n self.block4 = blocks()\n self.block5 = blocks()\n self.block6 = blocks()\n self.block7 = blocks()\n\n self.out = nn.Sequential(\n\n nn.Upsample(scale_factor=2),\n nn.Conv2d(out_features * 4, out_features * 2, 3, stride=1, padding=1),\n nn.InstanceNorm2d(out_features * 2),\n nn.LeakyReLU(0.2, inplace=True),\n\n nn.Upsample(scale_factor=2),\n nn.Conv2d(out_features * 2, out_features, 3, stride=1, padding=1),\n nn.InstanceNorm2d(out_features),\n nn.LeakyReLU(0.2, inplace=True),\n\n nn.ReflectionPad2d(channels),\n nn.Conv2d(out_features, channels, 7),\n nn.Tanh(),\n )","def mt_step(self, lang1, lang2, lambda_coeff):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n lang1_id = params.lang2id[lang1]\n lang2_id = params.lang2id[lang2]\n\n # generate batch\n if lang1 == lang2:\n (x1, len1) = self.get_batch('ae', lang1)\n (x2, len2) = (x1, len1)\n (x1, len1) = self.add_noise(x1, len1)\n else:\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\n langs1 = x1.clone().fill_(lang1_id)\n langs2 = x2.clone().fill_(lang2_id)\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n # key point generate label\n\n # cuda\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\n\n # encode source sentence\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\n enc1 = enc1.transpose(0, 1)\n\n # decode target sentence\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\n src_len=len1)\n\n # loss\n # the last word not apply prediction logic\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def bart_mlm_step(self, lang1, lang2, lambda_coeff):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n lang1_id = params.lang2id[lang1]\n lang2_id = params.lang2id[lang1]\n\n # generate batch / select words to predict\n x, lengths, positions, langs, _ = self.generate_batch(lang1, lang2, 'pred')\n x, lengths, positions, langs, _ = self.round_batch(x, lengths, positions, langs)\n\n (x1, len1, x2, len2, y, pred_mask, positions) = self.bart_token_mask_sent(x, lengths)\n\n if params.use_noise:\n x1,len1 = self.add_noise(x1,len1)\n langs1 = x1.clone().fill_(lang1_id)\n langs2 = x2.clone().fill_(lang2_id)\n\n # cuda\n x1,x2,len1,len2, y, pred_mask, positions,langs1,langs2 = to_cuda(x1,x2,len1,len2, y, pred_mask, positions,langs1,langs2)\n\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\n enc1 = enc1.transpose(0, 1)\n\n # enc_mask = x1.ne(params.mask_index)\n # enc_mask = enc_mask.transpose(0, 1)\n\n dec2 = model('crossfwd', stream_='text',\n x=x2, lengths=len2, langs=langs2, causal=True,\n src_enc=enc1, src_len=len1)\n\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n\n self.stats[('M-BART-%s' % lang1) if lang2 is None else ('MLM-%s-%s' % (lang1, lang2))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += lengths.size(0)\n self.stats['processed_w'] += pred_mask.sum().item()","def mt_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n # _lang1 = _lang2 = 'en'\n # lang2_id = params.lang2id[_lang2]\n\n # generate batch\n\n (x_src,len_src),(x2, len2),(x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\n # convert fp16\n #assign lang ids\n\n\n lang0_id = params.lang2id[params.ft_lgs[0]]\n lang_src = x_src.clone().fill_(lang0_id)\n lang1_id = params.lang2id[params.ft_lgs[1]]\n langs = x2.clone().fill_(lang1_id)\n\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n #\n len1 = x1_mask.sum(dim=1)\n x1 = x1.transpose(0, 1)\n img_loc = img_loc.transpose(0, 1)\n\n # cuda\n\n x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src)\n\n if params.mt_only_text:\n encoder_outputs = model('crossfwd', stream_='text', x=x_src, lengths=len_src, langs=lang_src, causal=False,refine_image=params.refine_image)\n len_all = len_src\n else:\n encoder_outputs = model('jointfwd', x=x_src, lengths=len_src, x_img=x1, lengths_img=len1, causal=False,\n langs=None,\n image_loc=img_loc, refine_image=params.refine_image)\n len_all = len_src + len1\n\n # encode source sentence\n # enc1 = model('jo', stream_='img', x=x1, lengths=len1, langs=None, causal=False, cross_modal=True,\n # image_loc=img_loc, refine_image=params.refine_image, refine_encoder=params.refine_encoder,\n # image_dist=None)\n enc1 = encoder_outputs.transpose(0, 1)\n\n\n # decode target sentence\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs, causal=True, src_enc=enc1,\n src_len=len_all)\n\n # loss\n # the last word not apply prediction logic\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n self.stats[('IC-%s-%s' % (dataset, input_stream))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def transform():","def __init__(self):\n super(Encoder2, self).__init__()\n self.lblocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n ]\n )\n\n self.blocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n ]\n )","def transform():\n pass","def forward(self, x):\n skips = {}\n skips[\"input\"] = x\n\n # Encoder\n skips[\"pool1_out\"] = self.enc_block1(x)\n skips[\"pool2_out\"] = self.enc_block2(skips[\"pool1_out\"])\n skips[\"pool3_out\"] = self.enc_block3(skips[\"pool2_out\"])\n skips[\"pool4_out\"] = self.enc_block4(skips[\"pool3_out\"])\n output = self.enc_block5(skips[\"pool4_out\"])\n output = self.enc_block6(output)\n\n # Decoder\n output = self.upsample5(output)\n output = self.dec_block1(torch.cat([skips[\"pool4_out\"], output], dim=1))\n output = self.dec_block2(torch.cat([skips[\"pool3_out\"], output], dim=1))\n output = self.dec_block3(torch.cat([skips[\"pool2_out\"], output], dim=1))\n output = self.dec_block4(torch.cat([skips[\"pool1_out\"], output], dim=1))\n output = self.dec_block5(torch.cat([skips[\"input\"], output], dim=1))\n\n return output","def preprocess(self):","def __init__(self):\n super(Encoder3, self).__init__()\n self.lblocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n ]\n )\n\n self.blocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n ]\n )","def to_basic_block(self) -> \"gr::basic_block_sptr\":\n return _beamforming_swig.beamformer_sptr_to_basic_block(self)","def mobilenetv2 (inputs, k, alpha = 1.0, train_bn = False):\n\n x = conv_block(inputs, 32, alpha, (3, 3), strides=(2, 2), block_id=0, train_bn=train_bn) # Input Res: 1\n\n x = inverted_residual_block(x, 16, (3, 3), t=1, strides=1, n=1, alpha=1.0, block_id=1, train_bn=train_bn)\t# Input Res: 1/2\n x = inverted_residual_block(x, 24, (3, 3), t=6, strides=2, n=2, alpha=1.0, block_id=2, train_bn=train_bn)\t# Input Res: 1/2\n x = inverted_residual_block(x, 32, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=4, train_bn=train_bn)\t# Input Res: 1/4\n x = inverted_residual_block(x, 64, (3, 3), t=6, strides=2, n=4, alpha=1.0, block_id=7, train_bn=train_bn)\t# Input Res: 1/8\n x = inverted_residual_block(x, 96, (3, 3), t=6, strides=1, n=3, alpha=1.0, block_id=11, train_bn=train_bn)\t# Input Res: 1/8\n x = inverted_residual_block(x, 160, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=14, train_bn=train_bn)\t# Input Res: 1/16\n x = inverted_residual_block(x, 320, (3, 3), t=6, strides=1, n=1, alpha=1.0, block_id=17, train_bn=train_bn)\t# Input Res: 1/32\n\n x = conv_block(x, 1280, alpha, (1, 1), strides=(1, 1), block_id=18, train_bn=train_bn) # Input Res: 1/32\n\n x = KL.GlobalAveragePooling2D()(x)\n x = KL.Reshape((1, 1, 1280))(x)\n x = KL.Dropout(0.3, name='Dropout')(x)\n x = KL.Conv2D(k, (1, 1), padding='same')(x)\n\n x = KL.Activation('softmax', name='softmax')(x)\n output = KL.Reshape((k,))(x)\n\n model = KM.Model(inputs, output)\n plot_model(model, to_file='MobileNetv2.png', show_shapes=True)\n\n return model","def transform(self, cfg_pipeline):\n return","def __init__(self, **kwargs):\n FeatureDefinition.__init__(self)\n\n nbTypes = self._getTypeNumber(kwargs)\n \n print(\"BETTER FEATURES\")\n \n \n block_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\n (\"xywh\", Pipeline([\n ('selector', NodeTransformerXYWH_v2()),\n #v1 ('xywh', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('xywh', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n , (\"neighbors\", Pipeline([\n ('selector', NodeTransformerNeighbors()),\n #v1 ('neighbors', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('neighbors', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n , (\"1hot\", Pipeline([\n ('1hot', Node1HotFeatures()) #does the 1-hot encoding directly\n ])\n )\n ])\n grid_line_transformer = GridLine_NodeTransformer_v2()\n \n self._node_transformer = TransformerListByType([block_transformer, grid_line_transformer]) \n \n edge_BB_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\n (\"1hot\", Pipeline([\n ('1hot', Edge1HotFeatures(PageNumberSimpleSequenciality()))\n ])\n )\n , (\"boolean\", Pipeline([\n ('boolean', EdgeBooleanFeatures_v2())\n ])\n )\n , (\"numerical\", Pipeline([\n ('selector', EdgeNumericalSelector()),\n #v1 ('numerical', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('numerical', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n ] )\n edge_BL_transformer = DU_ABPTableRG.Block2GridLine_EdgeTransformer()\n edge_LL_transformer = DU_ABPTableRG.GridLine2GridLine_EdgeTransformer()\n self._edge_transformer = TransformerListByType([edge_BB_transformer,\n edge_BL_transformer,\n edge_BL_transformer, # useless but required\n edge_LL_transformer \n ])\n \n self.tfidfNodeTextVectorizer = None #tdifNodeTextVectorizer","def convert_bmm(g, op, block):\n\n x = g.get_node(op.input(\"X\")[0])\n y = g.get_node(op.input(\"Y\")[0])\n y = _op.transpose(y, [0, 2, 1])\n out = _op.nn.batch_matmul(x, y)\n g.add_node(op.output(\"Out\")[0], out)","def build_label_transform():\n\n return NALabelEncoder()","def __init__(self, in_channel_list, out_channel_list, n_hr_blocks, n_blocks, block, bn_momentum=0.01):\n super(Stage, self).__init__()\n pre_branches = len(in_channel_list)\n cur_branches = len(out_channel_list)\n if cur_branches < pre_branches:\n raise ValueError('Current number of branch is smaller than previous number of branch')\n\n self.hr_blocks = nn.ModuleList()\n if in_channel_list != out_channel_list[:pre_branches]:\n use_transform = True\n else:\n use_transform = False\n for i in range(n_hr_blocks):\n self.hr_blocks.append(HighResolutionBlock(in_channel_list, out_channel_list[:pre_branches], block, n_blocks[0], use_transform))\n in_channel_list = [chs * block.expansion for chs in out_channel_list[:pre_branches]]\n self.downsample_layer = DownSampleLayer(in_channel_list, out_channel_list, bn_momentum)","def transform(self, data):","def __init__(self, args):\n super(Transformer, self).__init__()\n\n self.name = 'Transformer'\n if len(args.max_relative_pos) != args.nlayers:\n assert len(args.max_relative_pos) == 1\n args.max_relative_pos = args.max_relative_pos * args.nlayers\n\n self.embedder = Embedder(args)\n self.encoder = Encoder(args, self.embedder.enc_input_size)\n self.decoder_loc = nn.Linear(self.embedder.enc_input_size, 1)\n self.decoder_fix = nn.Linear(self.embedder.enc_input_size, 1)\n self.nobug_embedding = nn.Parameter(torch.zeros(1))\n self.criterion = nn.CrossEntropyLoss(reduction='none')","def __init__(self,\n input_dim,\n dec_seq_len,\n out_seq_len,\n d_model=512,\n nhead=8,\n num_encoder_layers=6,\n num_decoder_layers=6,\n dim_feedforward=2048,\n dropout=0.1,\n activation='relu',\n custom_encoder=None,\n custom_decoder=None):\n super(TransformerTS, self).__init__()\n self.transform = nn.Transformer(\n d_model=d_model,\n nhead=nhead,\n num_encoder_layers=num_encoder_layers,\n num_decoder_layers=num_decoder_layers,\n dim_feedforward=dim_feedforward,\n dropout=dropout,\n activation=activation,\n custom_encoder=custom_encoder,\n custom_decoder=custom_decoder\n )\n self.pos = PositionalEncoding(d_model)\n self.enc_input_fc = nn.Linear(input_dim, d_model)\n self.dec_input_fc = nn.Linear(input_dim, d_model)\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\n self.dec_seq_len = dec_seq_len","def create_pipeline(self, train: LAMLDataset) -> LAMLTransformer:\n raise NotImplementedError","def TextEncBlock(L,d,scope=\"TextEncBlock\"):\n e = L.shape.as_list()[2]\n\n with tf.variable_scope(scope):\n conv_params = {\"filters\":2*d,\"kernel_size\":1,\"padding\":\"same\"} \n with tf.variable_scope(\"C_block1\"):\n L1 = tf.nn.relu(conv1d(inputs=L,**conv_params)) # relu(conv)\n L2 = conv1d(inputs=L1,**conv_params) # conv\n # hc_dilation_blocks \n L3_HC1_1 = hc_dilation_block(L2,local_scope=\"HC_dilation_block1\")\n L3_HC1_2 = hc_dilation_block(L3_HC1_1,local_scope=\"HC_dilation_block2\")\n # highway_conv_block2\n with tf.variable_scope(\"HC_block2\"):\n L4_HC2_1 = highway_activation_conv(L3_HC1_2,kernel_size=3,scope=\"HC1\")\n L4_HC2_2 = highway_activation_conv(L4_HC2_1,kernel_size=3,scope=\"HC2\")\n # highway_conv_block3\n with tf.variable_scope(\"HC_block3\"):\n L5_HC3_1 = highway_activation_conv(L4_HC2_2,kernel_size=1,scope=\"HC1\")\n KV = highway_activation_conv(L5_HC3_1,kernel_size=1,scope=\"HC2\")\n\n return KV[:,:,:d], KV[:,:,d:] # key, value","def __init__(self):\n Transform.__init__(self)","def forward(self, x):\n # encode\n encode_block1 = self.conv_encode1(x)\n encode_pool1 = self.conv_maxpool1(encode_block1)\n encode_block2 = self.conv_encode2(encode_pool1)\n encode_pool2 = self.conv_maxpool2(encode_block2)\n encode_block3 = self.conv_encode3(encode_pool2)\n encode_pool3 = self.conv_maxpool3(encode_block3)\n # Bottleneck\n bottleneck1 = self.bottleneck(encode_pool3)\n # Decode\n decode_block3 = crop_and_concat(\n bottleneck1, encode_block3, crop=True)\n cat_layer2 = self.conv_decode3(decode_block3)\n decode_block2 = crop_and_concat(\n cat_layer2, encode_block2, crop=True)\n cat_layer1 = self.conv_decode2(decode_block2)\n decode_block1 = crop_and_concat(\n cat_layer1, encode_block1, crop=True)\n final_layer = self.final_layer(decode_block1)\n return final_layer","def _encode(self):\n with tf.variable_scope('encoding'):\n self.sep_p_encodes, _ = bilstm_layer(self.p_emb, self.p_length, self.hidden_size)\n tf.get_variable_scope().reuse_variables()\n self.sep_q_encodes, _ = bilstm_layer(self.q_emb, self.q_length, self.hidden_size)\n if self.use_dropout:\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, 1-self.dropout)\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, 1-self.dropout)","def __init__(self, mbart_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\n self.linear1_weight = mbart_layer.fc1.weight\n self.linear1_bias = mbart_layer.fc1.bias\n self.linear2_weight = mbart_layer.fc2.weight\n self.linear2_bias = mbart_layer.fc2.bias\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\n self.norm2_eps = mbart_layer.final_layer_norm.eps\n self.norm2_weight = mbart_layer.final_layer_norm.weight\n self.norm2_bias = mbart_layer.final_layer_norm.bias\n self.num_heads = mbart_layer.self_attn.num_heads\n self.embed_dim = mbart_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def convert_mish(g, op, block):\n\n x = g.get_node(op.input(\"X\")[0])\n dtype = infer_type(x).checked_type.dtype\n exp = _op.exp(x)\n add = _op.add(exp, _expr.const(1.0, dtype))\n log = _op.log(add)\n tanh = _op.tanh(log)\n out = _op.multiply(x, tanh)\n g.add_node(op.output(\"Out\")[0], out)","def lnlike(self, block):\n ##### Extract from datablock\n self.cosmology = {'Omega_m': block.get_double('cosmological_parameters', 'Omega_m'),\n 'Omega_l': block.get_double('cosmological_parameters', 'Omega_lambda'),\n 'Omega_b': block.get_double('cosmological_parameters', 'Omega_b'),\n 'h': block.get_double('cosmological_parameters', 'hubble')/100,\n 'ns': block.get_double('cosmological_parameters', 'n_s'),\n 'w0': block.get_double('cosmological_parameters', 'w'),\n 'wa': block.get_double('cosmological_parameters', 'wa'),\n 'sigma8': block.get_double('cosmological_parameters', 'sigma_8')}\n self.scaling = {\n # SZ\n 'Asz': block.get_double('mor_parameters', 'Asz'),\n 'Bsz': block.get_double('mor_parameters', 'Bsz'),\n 'Csz': block.get_double('mor_parameters', 'Csz'),\n 'Dsz': block.get_double('mor_parameters', 'Dsz'),\n # X-ray\n 'Ax': block.get_double('mor_parameters', 'Ax'),\n 'Bx': block.get_double('mor_parameters', 'Bx'),\n 'Cx': block.get_double('mor_parameters', 'Cx'),\n 'Dx': block.get_double('mor_parameters', 'Dx'),\n 'dlnMg_dlnr': block.get_double('mor_parameters', 'dlnMg_dlnr'),\n # WL\n 'WLbias': block.get_double('mor_parameters', 'WLbias'),\n 'WLscatter': block.get_double('mor_parameters', 'WLscatter'),\n 'HSTbias': block.get_double('mor_parameters', 'HSTbias'),\n 'HSTscatterLSS': block.get_double('mor_parameters', 'HSTscatterLSS'),\n 'MegacamBias': block.get_double('mor_parameters', 'MegacamBias'),\n 'MegacamScatterLSS': block.get_double('mor_parameters', 'MegacamScatterLSS'),\n # Correlation coefficients\n 'rhoSZWL': block.get_double('mor_parameters', 'rhoSZWL'),\n 'rhoWLX': block.get_double('mor_parameters', 'rhoWLX'),\n 'rhoSZX': block.get_double('mor_parameters', 'rhoSZX'),\n }\n # Halo mass function\n self.HMF = {'M_arr': block.get_double_array_1d('HMF', 'M_arr'),\n 'z_arr': block.get_double_array_1d('HMF', 'z_arr'),\n 'dNdlnM': block.get_double_array_nd('HMF', 'dNdlnM')}\n self.HMF['len_z'] = len(self.HMF['z_arr'])\n\n ##### Setup stuff for WL\n if self.todo['WL']:\n # Set bias and scatter for Megacam from sim calibration\n # and nuissance parameters\n self.WL.set_scaling(self.scaling)\n # Precompute array of angular diameter distances\n self.WL.get_dAs(self.cosmology)\n\n ##### Populate and check observable covariance matrices\n self.covmat = {'invertible': True}\n if not observablecovmat.set_covmats(self.todo, self.scaling, self.covmat):\n self.covmat['invertible'] = False\n return -np.inf\n\n ##### Set up interpolation for HMF\n HMF_in = self.HMF['dNdlnM'][1:,:]\n if np.any(HMF_in==0):\n HMF_in[np.where(HMF_in==0)] = np.nextafter(0, 1)\n self.HMF_interp = RectBivariateSpline(np.log(self.HMF['z_arr'][1:]), np.log(self.HMF['M_arr']), np.log(HMF_in))\n\n ##### Initialize mass-concentration relation class (for WL)\n if self.todo['WL']:\n self.MCrel = Mconversion_concentration.ConcentrationConversion(self.mcType, self.cosmology)\n\n ##### Compute interpolation table for M500-M200\n if self.todo['WL']:\n z_arr = np.linspace(.1, 2, 20)\n M500 = np.logspace(np.log10(self.HMF['M_arr'][0]), np.log10(self.HMF['M_arr'][-1]), 20)\n M200 = np.array([np.array([self.MCrel.MDelta_to_M200(m, 500., z) for m in M500]) for z in z_arr])\n self.lnM500_to_lnM200 = RectBivariateSpline(z_arr, np.log(M500), np.log(M200))\n\n ##### Evaluate the individual likelihoods\n len_data = len(self.catalog['SPT_ID'])\n\n if self.NPROC==0:\n # Iterate through cluster list\n likelihoods = np.array([self.clusterlike(i) for i in range(len_data)])\n else:\n # Launch a multiprocessing pool and get the likelihoods\n pool = Pool(processes=self.NPROC)\n argin = zip([self]*len_data, range(len_data))\n likelihoods = pool.map(unwrap_self_f, argin)\n pool.close()\n\n # If likelihood computation failed it returned 0\n if np.count_nonzero(likelihoods)<len_data:\n return -np.inf\n\n lnlike = np.sum(np.log(likelihoods))\n\n return lnlike","def forward(self, *args):\r\n enc_src, _, _ = self.unified_encoder(*args)\r\n enc_src = enc_src.view(enc_src.shape[0], -1)\r\n y_pred = self.mlp(enc_src)\r\n return y_pred","def make_mlp_model():\n return snt.Sequential([\n snt.nets.MLP([LATENT_SIZE] * NUM_LAYERS, activate_final=True),\n snt.LayerNorm()\n ])","def fit_transform(self):\n if self.enc_types == \"label\":\n return self._label_encoding()\n elif self.enc_types == \"ohe\":\n return self._one_hot_encoder()\n elif self.enc_types == \"binary\":\n return self._binarization()\n else:\n raise Exception(\"Encoding type not understood\")","def __transition_block(self, x, stage, nb_filter, compression=1.0, dropout_rate=None, weight_decay=1E-4):\n\n\t\teps = 1.1e-5\n\t\tconv_name_base = \"conv\" + str(stage) + \"_blk\"\n\t\trelu_name_base = \"relu\" + str(stage) + \"_blk\"\n\t\tpool_name_base = \"pool\" + str(stage) \n\n\t\tx = BatchNormalization(epsilon=eps, axis=self.concat_axis, name=conv_name_base+\"_bn\")(x)\n\t\tx = Scale(axis=self.concat_axis, name=conv_name_base+\"_scale\")(x)\n\t\tx = Activation(\"relu\", name=relu_name_base)(x)\n\t\tx = Conv2D(int(nb_filter * compression), (1, 1), name=conv_name_base, use_bias=False)(x)\n\n\t\tif dropout_rate:\n\t\t\tx = Dropout(dropout_rate)(x)\n\n\t\tx = AveragePooling2D((2, 2), strides=(2, 2), name=pool_name_base)(x)\n\n\t\treturn x","def __init__(self, in_channel_list, out_channel_list, block, n_blocks,\n use_transform=False):\n super(HighResolutionBlock, self).__init__()\n self.out_chs_list = out_channel_list\n self.branches = nn.ModuleList()\n\n for in_chs, out_chs in zip(in_channel_list, out_channel_list):\n branch = Branch(in_chs, out_chs, block, n_blocks)\n self.branches.append(branch)\n if len(out_channel_list) > 1:\n self.mul_resolution_layer = MultiResolutionLayer(out_channel_list)\n else:\n self.mul_resolution_layer = None\n pass","def forward(self, input):\n x = self.emb(input)\n output = self.attention(x)\n # print(\"Idggfghrf \", output.size())\n return output","def beam_init(self, encoder_outputs, K):\n encoder_outputs = TransformerEncoderOutput(\n repeat_batch(encoder_outputs.output, K),\n repeat_batch(encoder_outputs.src_mask, K)\n )\n return encoder_outputs","def __init__(\n self,\n num_branches,\n blocks,\n num_blocks,\n num_inchannels,\n num_channels,\n num_heads,\n num_window_sizes,\n num_mlp_ratios,\n multi_scale_output=True,\n drop_path=0.0,\n ):\n super(HighResolutionTransformerModule, self).__init__()\n self._check_branches(\n num_branches, blocks, num_blocks, num_inchannels, num_channels\n )\n\n self.num_inchannels = num_inchannels\n self.num_branches = num_branches\n\n self.multi_scale_output = multi_scale_output\n self.branches = self._make_branches(\n num_branches,\n blocks,\n num_blocks,\n num_channels,\n num_heads,\n num_window_sizes,\n num_mlp_ratios,\n drop_path,\n )\n self.fuse_layers = self._make_fuse_layers()\n self.relu = nn.ReLU(inplace=True)\n\n self.num_heads = num_heads\n self.num_window_sizes = num_window_sizes\n self.num_mlp_ratios = num_mlp_ratios","def __call__(self, **kwargs):\n segname = 'block_{}_expand_relu'\n blocks = [13, 6, 3, 1]\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\n backbone_out = self._backbone.get_layer('block_16_project')\n\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\n p4 = self._fpn_block(p5, skips[1].output)\n p3 = self._fpn_block(p4, skips[2].output)\n p2 = self._fpn_block(p3, skips[3].output)\n\n s5 = self._conv_block(p5, 128)\n s4 = self._conv_block(p4, 128)\n s3 = self._conv_block(p3, 128)\n s2 = self._conv_block(p2, 128)\n\n s5 = tf.keras.layers.UpSampling2D(\n size=(8, 8),\n interpolation='nearest'\n )(s5)\n\n s4 = tf.keras.layers.UpSampling2D(\n size=(4, 4),\n interpolation='nearest'\n )(s4)\n\n s3 = tf.keras.layers.UpSampling2D(\n size=(2, 2),\n interpolation='nearest'\n )(s3)\n\n concat = [s5, s4, s3, s2]\n x = tf.keras.layers.Concatenate()(concat)\n x = tf.keras.layers.Conv2D(\n 64,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n x = tf.keras.layers.BatchNormalization()(x)\n x = tf.keras.layers.Activation('relu')(x)\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\n\n x = tf.keras.layers.Conv2D(\n 1,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n out = tf.keras.layers.Activation('sigmoid')(x)\n model = tf.keras.models.Model(\n inputs=self._backbone.input,\n outputs=out\n )\n\n return model","def __init__(self, distance_function=euclidean_distance, depth=64, pool_last_two=True):\n super().__init__()\n self.depth = depth\n self.encoder = nn.Sequential(\n conv_block(1, depth, pool_size = 2),\n conv_block(depth, depth, pool_size = 2),\n conv_block(depth, depth, pool_size = 2),\n conv_block(depth, depth, pool_size = 2)\n )\n self.distance = distance_function","def _define_encoder(self):\n self.encoder = nn.Sequential(View((-1, 64 * 64 * 3)),\n nn.Linear(64 * 64 * 3, 5120, bias=False), nn.SELU(),\n nn.BatchNorm1d(5120),\n nn.Linear(5120, 2560, bias=False), nn.SELU(),\n nn.BatchNorm1d(2560),\n nn.Linear(2560, 512, bias=False), nn.SELU(),\n nn.BatchNorm1d(512),\n nn.Linear(512, 128, bias=False), nn.SELU(),\n nn.BatchNorm1d(128),\n nn.Linear(128, self.encoding_shape, bias=False), nn.SELU(),\n )","def free_lb_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n\n (x2, len2), (x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\n\n if len(params.ft_lgs) > 0:\n lang1_id = params.lang2id[params.ft_lgs[0]]\n langs = x2.clone().fill_(lang1_id)\n else:\n lang1_id = params.lang2id['en']\n langs = x2.clone().fill_(lang1_id)\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n #\n len1 = x1_mask.sum(dim=1)\n x1 = x1.transpose(0, 1)\n img_loc = img_loc.transpose(0, 1)\n if len(params.ft_lgs) > 0:\n lang1_id = params.lang2id[params.ft_lgs[0]]\n langs_img = x1_mask.transpose(0, 1).clone().fill_(lang1_id)\n else:\n lang1_id = params.lang2id['en']\n langs_img = x1_mask.transpose(0, 1).clone().fill_(lang1_id)\n\n # cuda\n\n x1, len1, img_loc, x2, len2, y, x1_mask, langs, langs_img = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,\n langs, langs_img)\n\n # convert fp16\n #assign lang ids\n\n #free lb\n if params.free_text:\n embeds_init, delta = self.deal_freelb_delta(model,x2.transpose(0,1),len2)\n if params.free_img:\n image_delta = self.deal_image_freelb_delta(x1)\n\n # the main loop\n dp_masks = None\n adv_steps=3\n tr_loss, tb_loss = 0.0, 0.0\n for astep in range(adv_steps):\n # (0) forward\n text_imb = None\n if params.free_text:\n delta.requires_grad_()\n text_imb = delta + embeds_init\n img_imb = x1\n if params.free_img:\n image_delta.requires_grad_()\n img_imb = x1 + image_delta\n\n output = self.get_ic_output(params,model,x2,len2,img_imb,len1,img_loc,langs,langs_img,pred_mask,y,text_imb)\n loss, dec_out = output[0], output[1]\n\n\n # (1) backward\n\n # if params.n_gpu_per_node > 1:\n # loss = loss.mean() # mean() to average on multi-gpu parallel training\n # if params.accumulate_gradients > 1:\n # loss = loss / params.accumulate_gradients\n\n loss = loss /(1.0*adv_steps)\n\n self.free_optimize(loss)\n\n tb_loss+=loss.item()\n #\n # self.free_optimize(loss,adv_steps)\n\n if astep == adv_steps - 1:\n # further updates on delta\n break\n\n # update delta\n if params.free_text:\n embeds_init, delta = self.update_freelb_delta(model, delta, embeds_init, x2.transpose(0,1))\n if params.free_img:\n image_delta = self.update_image_freelb_delta(x1, image_delta)\n\n # loss = tr_loss+tb_loss\n\n self.stats[('FRLB-IC-%s-%s' % (dataset, input_stream))].append(tb_loss)\n loss = lambda_coeff * loss\n\n # optimize\n # self.free_optimize(None,True)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def init_encoder(self):\n\n vgg = models.vgg16(pretrained=True)\n\n blocks = [self.layer_1,\n self.layer_2,\n self.layer_3,\n self.layer_4,\n self.layer_5]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data\n\n blocks = [self.layer_11,\n self.layer_12,\n self.layer_13,\n self.layer_14,\n self.layer_15]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data","def __init__(self, idim, odim, args, ignore_id=-1):\n torch.nn.Module.__init__(self)\n if args.transformer_attn_dropout_rate is None:\n args.transformer_attn_dropout_rate = args.dropout_rate\n\n self.mtlalpha = args.mtlalpha\n if not hasattr(args, \"beta\"):\n self.beta = 0.0\n else:\n self.beta = args.beta\n self.num_steps = args.num_steps\n self.tau = args.error_temperature\n self.idx_blank = 0\n self.masktoken = odim - 1\n self.sos = odim - 1\n self.eos = odim - 1\n self.odim = odim\n self.ignore_id = ignore_id\n self.lsm_weight = args.lsm_weight\n self.reporter = None\n\n self.subsample = [1]\n self.encoder = Encoder(\n idim=idim,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.eunits,\n num_blocks=args.elayers,\n input_layer=args.transformer_input_layer,\n dropout_rate=args.dropout_rate,\n death_rate=args.edeath_rate,\n positional_dropout_rate=args.dropout_rate,\n attention_dropout_rate=args.transformer_attn_dropout_rate\n )\n self.enc_odim = args.adim\n\n # Define alignment embedding module.\n self.embed = torch.nn.Embedding(odim, self.enc_odim)\n self.decoder = Decoder(\n odim=odim,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n dropout_rate=args.dropout_rate,\n death_rate=args.ddeath_rate,\n positional_dropout_rate=args.dropout_rate,\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\n input_layer=\"embed_linear\",\n pos_enc_class=ScaledPositionalEncoding\n )\n\n # self.verbose = args.verbose\n self.reset_parameters(args)\n self.adim = args.adim\n\n # CTC directly works on top of encoder outputs.\n self.ctc = CTC(odim, self.enc_odim, args.dropout_rate, ctc_type=args.ctc_type, reduce=True)\n self.rnnlm = None\n\n # Decoder masking params\n self.decoder_num_mask = args.decoder_num_mask\n self.decoder_mask_width = args.decoder_mask_width","def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\n latent_dimension):\n super(VAEEncoder, self).__init__()\n self.latent_dimension = latent_dimension\n\n # Reduce dimension up to second last layer of Encoder\n self.encode_nn = nn.Sequential(\n nn.Linear(in_dimension, layer_1d),\n nn.ReLU(),\n nn.Linear(layer_1d, layer_2d),\n nn.ReLU(),\n nn.Linear(layer_2d, layer_3d),\n nn.ReLU()\n )\n\n # Latent space mean\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\n\n # Latent space variance\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)","def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\n latent_dimension):\n super(VAEEncoder, self).__init__()\n self.latent_dimension = latent_dimension\n\n # Reduce dimension up to second last layer of Encoder\n self.encode_nn = nn.Sequential(\n nn.Linear(in_dimension, layer_1d),\n nn.ReLU(),\n nn.Linear(layer_1d, layer_2d),\n nn.ReLU(),\n nn.Linear(layer_2d, layer_3d),\n nn.ReLU()\n )\n\n # Latent space mean\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\n\n # Latent space variance\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)","def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\n ptm=\"bert\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\n super(PTMClassifier_PNN, self).__init__()\n \n if old_cols is None:\n old_cols = []\n \n self.old_cols = []\n \n self.output_size = output_size\n self.hidden_size = hidden_size # default\n self.feature_size = feature_size\n \n self.fc1 = nn.Linear(self.hidden_size, 100)\n self.classifier = nn.Linear(100, self.output_size)\n \n if len(old_cols) > 0:\n self.old_fc1s = ListModule()\n self.base_1 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n self.base_2 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n \n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_1)\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_2)\n \n for old_col in old_cols:\n self.old_fc1s.append(\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\n \n # ptm\n self.ptm = ptm.lower()\n assert self.ptm in supported_ptm\n ptm_ = import_from(\"transformers\", supported_ptm[self.ptm][0] + \"Model\")\n \n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\n if tokenizer is not None:\n self.encoder.resize_token_embeddings(len(tokenizer))\n self.prob_l = prob_l\n \n self.require_proto = require_proto\n \n # todo : modify net into one network\n if self.require_proto:\n # prototype-based classification\n self.net = nn.Sequential(\n self.fc1\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.fc1,\n self.encoder_adaptor,\n nn.ReLU(),\n )\n else:\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\n self.net = nn.Sequential(\n self.classifier\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.classifier\n )\n \n self.reset_parameters()","def __init__(\n self,\n encoder,\n num_classes=20,\n shape_size=16,\n mask_conv=True,\n min_confidence=0.2,\n min_remain=0.5,\n mask_threshold=0.4,\n ):\n super(PaPs, self).__init__()\n self.encoder = encoder\n self.shape_size = shape_size\n self.num_classes = num_classes\n self.min_scale = 1 / shape_size\n self.register_buffer(\"min_confidence\", torch.tensor([min_confidence]))\n self.min_remain = min_remain\n self.mask_threshold = mask_threshold\n self.center_extractor = CenterExtractor()\n\n enc_dim = encoder.enc_dim\n stack_dim = encoder.stack_dim\n self.heatmap_conv = nn.Sequential(\n ConvLayer(nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\n padding_mode=\"reflect\"),\n nn.Sigmoid(),\n )\n\n self.saliency_conv = ConvLayer(\n nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\n padding_mode=\"reflect\"\n )\n\n self.shape_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, shape_size ** 2),\n )\n\n self.size_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, stack_dim // 4),\n nn.BatchNorm1d(stack_dim // 4),\n nn.ReLU(),\n nn.Linear(stack_dim // 4, 2),\n nn.Softplus(),\n )\n\n self.class_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, stack_dim // 4),\n nn.Linear(stack_dim // 4, num_classes),\n )\n\n if mask_conv:\n self.mask_cnn = nn.Sequential(\n nn.Conv2d(in_channels=1, out_channels=16, kernel_size=3, padding=1),\n nn.GroupNorm(num_channels=16, num_groups=1),\n nn.ReLU(),\n nn.Conv2d(in_channels=16, out_channels=16, kernel_size=3, padding=1),\n nn.ReLU(),\n nn.Conv2d(in_channels=16, out_channels=1, kernel_size=3, padding=1),\n )\n else:\n self.mask_cnn = None","def transform(self, X):\n ...","def transform(self, X):\n ...","def transform(self, X):\n ...","def transform(self, X):\n ...","def transform(self, X):\n ...","def transform(self, X):\n ...","def transform(self, X):\n ...","def _define_encoder(self):\n raise NotImplementedError","def create_pipelines_lingspam():\n stop = ('stop', StopWordRemovalTransformer())\n lemma = ('lemma', LemmatizeTransformer())\n binz = ('binarizer', CountVectorizer())\n we = ('document embedding', DocEmbeddingVectorizer())\n sel = ('fsel', SelectKBest(score_func=mutual_info_classif, k=100))\n clf = ('cls', BernoulliNB()) # Binary features in the original paper. \n return Pipeline([binz, sel, clf]), \\\n Pipeline([stop, binz, sel, clf]), \\\n Pipeline([lemma, binz, sel, clf]), \\\n Pipeline([stop, lemma, binz, sel, clf]), \\\n Pipeline([stop, lemma, we, sel, clf])","def __init__(self, transformer):\n self.transformer = transformer","def __init__(\n self,\n in_channels=128,\n aux_channels=80,\n channels=64,\n out_channels=1,\n kernel_size=9,\n dilation=2,\n bias=True,\n noise_upsample_scales=[11, 2, 2, 2],\n noise_upsample_activation=\"LeakyReLU\",\n noise_upsample_activation_params={\"negative_slope\": 0.2},\n upsample_scales=[2, 2, 2, 2, 2, 2, 2, 2, 1],\n upsample_mode=\"nearest\",\n gated_function=\"softmax\",\n use_weight_norm=True,\n ):\n super().__init__()\n\n self.in_channels = in_channels\n\n noise_upsample = []\n in_chs = in_channels\n for noise_upsample_scale in noise_upsample_scales:\n # NOTE(kan-bayashi): How should we design noise upsampling part?\n noise_upsample += [\n torch.nn.ConvTranspose1d(\n in_chs,\n channels,\n noise_upsample_scale * 2,\n stride=noise_upsample_scale,\n padding=noise_upsample_scale // 2 + noise_upsample_scale % 2,\n output_padding=noise_upsample_scale % 2,\n bias=bias,\n )\n ]\n noise_upsample += [\n getattr(torch.nn, noise_upsample_activation)(\n **noise_upsample_activation_params\n )\n ]\n in_chs = channels\n self.noise_upsample = torch.nn.Sequential(*noise_upsample)\n self.noise_upsample_factor = np.prod(noise_upsample_scales)\n\n self.blocks = torch.nn.ModuleList()\n aux_chs = aux_channels\n for upsample_scale in upsample_scales:\n self.blocks += [\n TADEResBlock(\n in_channels=channels,\n aux_channels=aux_chs,\n kernel_size=kernel_size,\n dilation=dilation,\n bias=bias,\n upsample_factor=upsample_scale,\n upsample_mode=upsample_mode,\n gated_function=gated_function,\n ),\n ]\n aux_chs = channels\n self.upsample_factor = np.prod(upsample_scales)\n\n self.output_conv = torch.nn.Sequential(\n torch.nn.Conv1d(\n channels,\n out_channels,\n kernel_size,\n 1,\n bias=bias,\n padding=(kernel_size - 1) // 2,\n ),\n torch.nn.Tanh(),\n )\n\n # apply weight norm\n if use_weight_norm:\n self.apply_weight_norm()\n\n # reset parameters\n self.reset_parameters()","def __init__(self, c):\n super(DeepConv, self).__init__(self)\n\n self.block_1 = BaseBlock(c, 16)\n self.block_2 = BaseBlock(16, 32)\n self.block_3 = BaseBlock(32, 32)\n\n self._body = nn.Sequential(self.block_1, self.block_2, self.block_3)","def preprocess(item):\n item = feature_engineering(item)\n item = encode_features(item)\n return item","def mtf_transformer_lm_moe():\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\n hparams.decoder_layers = [\"att\", \"moe\"] * 4\n moe.set_default_moe_hparams(hparams)\n hparams.mesh_shape = \"all:8\"\n hparams.layout = \"batch:all;experts:all\"\n return hparams","def forward(self, x):\r\n x = self.conv1(x)\r\n x = self.conv1_BN(x)\r\n x = F.relu(x)\r\n x = self.conv1_dp(x)\r\n x = self.Block2_1(x)\r\n x = self.Block2_2(x)\r\n x = self.Block3_1(x)\r\n x = self.Block3_2(x)\r\n x = self.Block3_3(x)\r\n x = self.Block3_4(x)\r\n x = self.Block4_1(x)\r\n x = self.Block4_2(x)\r\n x = self.Block4_3(x)\r\n x = self.Block4_4(x)\r\n x = self.Block5_1(x)\r\n x = self.Block5_2(x)\r\n x = self.MP(x)\r\n x = x.view(x.size(0),-1)\r\n x = self.fc(x)\r\n \r\n return x","def __init__(self):\n super(Generator_Model, self).__init__()\n\n # audio encoder\n self.audio_encoder = nn.Sequential(\n conv_block(1, 32, kernel_size=1, stride=1, padding=1),\n conv_block(32, 32, kernel_size=1, stride=1, padding=1, residual=True),\n \n conv_block(32, 64, kernel_size=3, stride=(3, 1), padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(64, 128, kernel_size=3, stride=3, padding=1),\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(128, 256, kernel_size=3, stride=(3, 2), padding=1),\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(256, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 512, kernel_size=1, stride=1, padding=0)\n )\n\n # face encoder\n self.face_encoder = nn.ModuleList([\n nn.Sequential(conv_block(6, 16, kernel_size=7, stride=1, padding=3)),\n\n nn.Sequential(conv_block(16, 32, kernel_size=3, stride=2, padding=1),\n conv_block(32, 32, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(32, 64, kernel_size=3, stride=2, padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(64, 128, kernel_size=3, stride=2, padding=1),\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(128, 256, kernel_size=3, stride=2, padding=1),\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(256, 512, kernel_size=3, stride=2, padding=1),\n conv_block(512, 512, kernel_size=3, stride=1, padding=1, residual=True)),\n \n nn.Sequential(conv_block(512, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 512, kernel_size=1, stride=1, padding=0))\n ])\n\n # face decoder\n self.face_decoder = nn.ModuleList([\n nn.Sequential(conv_block(512, 512, kernel_size=1, stride=1, padding=0),),\n\n nn.Sequential(conv_t_block(512, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 1024, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(1024, 512, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(512, 768, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(768, 384, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(384, 512, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(512, 256, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(256, 320, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(320, 128, kernel_size=3, stride=2, padding=1, output_padding=1), \n conv_block(128, 160, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(160, 64, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True))\n ])\n\n # output blocks for face prediction\n self.pred_output = nn.Sequential(conv_block(64, 32, kernel_size=3, stride=1, padding=1),\n conv_block(32, 3, kernel_size=1, stride=1, padding=0),\n nn.Sigmoid())\n\n # optimizer\n self.optimizer = torch.optim.Adam(self.parameters(), lr=0.001, betas=(0.5, 0.999))","def forward(self, x):\n if self.training:\n x = self.input_pert(x)\n x = self.encoder(x)\n x = self.decoder(x)\n return x","def transform(self, x):","def prepare_emb(self):\n with tf.variable_scope(\"PrepEmb\", reuse=tf.AUTO_REUSE):\n self.src_ten = tf.cast(tf.convert_to_tensor(self.src_ten), tf.float32)\n self.tgt_ten = tf.cast(tf.convert_to_tensor(self.tgt_ten), tf.float32)\n # Mapping\n self.src_ten = tf.matmul(self.src_ten, self.W)\n # Normalization\n self.src_ten = tf.nn.l2_normalize(self.src_ten, axis=1)\n self.tgt_ten = tf.nn.l2_normalize(self.tgt_ten, axis=1)","def create_bem_flow(name='bem', out_format='stl'):\n \"\"\"\n Initialize the workflow\n \"\"\"\n\n bemflow = pe.Workflow(name=name)\n \"\"\"\n Define the inputs to the workflow.\n \"\"\"\n\n inputnode = pe.Node(\n niu.IdentityInterface(fields=['subject_id', 'subjects_dir']),\n name='inputspec')\n \"\"\"\n Define all the nodes of the workflow:\n\n fssource: used to retrieve aseg.mgz\n mri_convert : converts aseg.mgz to aseg.nii\n tessellate : tessellates regions in aseg.mgz\n surfconvert : converts regions to stereolithographic (.stl) format\n\n \"\"\"\n\n watershed_bem = pe.Node(interface=mne.WatershedBEM(), name='WatershedBEM')\n\n surfconvert = pe.MapNode(\n fs.MRIsConvert(out_datatype=out_format),\n iterfield=['in_file'],\n name='surfconvert')\n \"\"\"\n Connect the nodes\n \"\"\"\n\n bemflow.connect([\n (inputnode, watershed_bem, [('subject_id', 'subject_id'),\n ('subjects_dir', 'subjects_dir')]),\n (watershed_bem, surfconvert, [('mesh_files', 'in_file')]),\n ])\n \"\"\"\n Setup an outputnode that defines relevant inputs of the workflow.\n \"\"\"\n\n outputnode = pe.Node(\n niu.IdentityInterface(fields=[\"meshes\"]), name=\"outputspec\")\n bemflow.connect([\n (surfconvert, outputnode, [(\"converted\", \"meshes\")]),\n ])\n return bemflow","def __init__(self, inputs, outputs):\n super(ELM, self).__init__(inputs, outputs)","def _encode(self):\n with tf.variable_scope('passage_encoding'):\n self.sep_p_encodes, _ = rnn('bi-lstm', self.p_emb, self.p_length, self.hidden_size)\n with tf.variable_scope('question_encoding'):\n self.sep_q_encodes, _ = rnn('bi-lstm', self.q_emb, self.q_length, self.hidden_size)\n if self.use_dropout:\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, self.dropout_keep_prob)\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, self.dropout_keep_prob)","def mtf_transformer2_base():\n hparams = common_hparams.basic_params1()\n\n hparams.add_hparam(\"d_model\", 1024)\n hparams.batch_size = 4\n hparams.max_length = 1024\n hparams.label_smoothing = 0.0\n # a small positive value - this seems important for stability when training\n # with bfloat16 activations.\n hparams.add_hparam(\"z_loss\", 1e-4)\n\n # hparams applying to both encoder and decoder layer stacks.\n hparams.add_hparam(\"d_ff\", 2048)\n hparams.add_hparam(\"d_kv\", 128)\n hparams.add_hparam(\"attention_dropout\", 0.0)\n hparams.add_hparam(\"relu_dropout\", 0.0)\n hparams.del_hparam(\"num_heads\")\n hparams.del_hparam(\"num_hidden_layers\")\n hparams.layer_prepostprocess_dropout = 0.0\n hparams.add_hparam(\"extra_logit\", False)\n # number of experts for moe_1d\n hparams.moe_num_experts = 32\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\n hparams.add_hparam(\"moe_expert_x\", 8)\n hparams.add_hparam(\"moe_expert_y\", 4)\n hparams.add_hparam(\"moe_hidden_size\", 32768)\n\n # round up vocab sizes to be a multiple of this value\n hparams.vocab_divisor = 128\n\n hparams.optimizer = \"Adafactor\"\n hparams.learning_rate_schedule = \"rsqrt_decay*linear_decay\"\n hparams.learning_rate_warmup_steps = 10000\n hparams.add_hparam(\"master_dtype\", \"bfloat16\")\n hparams.add_hparam(\"slice_dtype\", \"float32\")\n hparams.activation_dtype = \"bfloat16\"\n\n # 8-way model-parallelism\n hparams.add_hparam(\"mesh_shape\", \"model:8\")\n hparams.add_hparam(\"layout\", \"batch:batch;vocab:model;d_ff:model;heads:model\")\n\n # If nonzero, we split the batch across two tensor-dimensions named\n # \"outer_batch\" and \"inner_batch\", allowing for splitting across two mesh\n # dimensions. This is necessary for hierarchical mixture of experts.\n # The two tensor dimensions have sizes hparams.outer_batch_size and\n # hparams.batch_size // hparams.outer_batch_size.\n hparams.add_hparam(\"outer_batch_size\", 0)\n\n hparams.shared_embedding_and_softmax_weights = False\n # length for training or decoding - defaults to max_length\n hparams.add_hparam(\"length\", 0)\n\n # These parameters make Transformer model compatible with mtf\n # Do not override these.\n hparams.no_data_parallelism = True\n hparams.use_fixed_batch_size = True\n hparams.add_hparam(\"mtf_mode\", True)\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\n hparams.bottom = {\n \"inputs\": modalities.identity_bottom,\n \"targets\": modalities.identity_bottom,\n }\n hparams.top = {\n \"targets\": modalities.identity_top,\n }\n hparams.add_hparam(\"beam_size\", 1)\n\n # If this is True, then in a packed dataset (where exaples are concatenated\n # to form longer examples) we use the global position (within the concatenated\n # sequence) to compute the positional embedding, instead of the position\n # within the individual sequence. This is counterintuitive, but for some\n # reason, it keeps the model from diverging.\n hparams.add_hparam(\"use_global_position_in_packed_sequence\", True)\n\n return hparams","def build(self, lang, linearInMLP=False):\n inputLayers, concLayers = [], []\n inputToken = Input((3 + configuration['embedding']['useB1'] + configuration['embedding']['useB-1'],))\n inputLayers.append(inputToken)\n tokenEmb = Embedding(len(self.vocabulary.tokenIndices), configuration['mlp']['tokenEmb'],\n trainable=configuration['mlp']['trainable'],\n weights=self.getWeightMatrix(self.vocabulary.tokenIndices, lang))(inputToken)\n tokenFlatten = Flatten()(tokenEmb)\n concLayers.append(tokenFlatten)\n inputPos = Input((3 + configuration['embedding']['useB1'] + configuration['embedding']['useB-1'],))\n inputLayers.append(inputPos)\n posEmb = Embedding(len(self.vocabulary.posIndices), configuration['mlp']['posEmb'],\n trainable=configuration['mlp']['trainable'])(inputPos)\n posFlatten = Flatten()(posEmb)\n concLayers.append(posFlatten)\n if linearInMLP:\n linearPredInput = Input(shape=(8,))\n inputLayers.append(linearPredInput)\n concLayers.append(linearPredInput)\n\n conc = keras.layers.concatenate(concLayers) if len(concLayers) > 1 else concLayers[0]\n dense1Layer = Dense(configuration['mlp']['dense1UnitNumber'],\n activation=configuration['nn']['dense1Activation'])(conc)\n lastLayer = Dropout(configuration['mlp']['dense1Dropout'])(dense1Layer)\n if configuration['mlp2']['dense2']:\n dense2Layer = Dense(configuration['mlp2']['dense2UnitNumber'],\n activation=configuration['mlp2']['dense2Activation'])(lastLayer)\n lastLayer = Dropout(configuration['mlp2']['dense2Dropout'])(dense2Layer)\n softmaxLayer = Dense(8 if enableCategorization else 4, activation='softmax')(lastLayer)\n return inputLayers, softmaxLayer","def init_encoder(self):\n\n vgg = models.vgg16(pretrained=True)\n\n blocks = [self.layer_1,\n self.layer_2,\n self.layer_3,\n self.layer_4,\n self.layer_5]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data","def mobilenetv2_block(x_in, ch_in, ch_out, expansion_factor, regularizer, strides=1):\n x = conv_2d_with_bn_relu(ch_out=expansion_factor*ch_in, kernel_size=1, regularizer=regularizer)(x_in)\n x = depthwise_conv_2d_with_bn_relu(strides, regularizer=regularizer)(x)\n x = BatchNormalization(beta_regularizer=regularizer, gamma_regularizer=regularizer)(\n Conv2D(ch_out, kernel_size=1, strides=1, padding=\"same\", use_bias=False, kernel_regularizer=regularizer)(x))\n return x if strides == 2 or ch_in != ch_out else Add()([x, x_in])","def forward(self, *args): # noqa: R0914\r\n encoder_out, (hn, cn) = self.unified_encoder(*args)\r\n device = hn.device\r\n non_sequential_cont_decoded = self.mlp_non_seq_cont(hn)\r\n non_sequential_cat_decoded = []\r\n for mlp_non_seq_cat in self.mlp_non_seq_cat_list:\r\n non_sequential_cat_decoded.append(mlp_non_seq_cat(hn))\r\n\r\n hn = torch.unsqueeze(hn, 0)\r\n cn = torch.unsqueeze(cn, 0)\r\n # decoded is the output prediction of timestep i-1 of the decoder\r\n decoded = torch.zeros(encoder_out.shape[0], int(\r\n self.unified_encoder.seq_cont_count + self.unified_encoder.no_of_embs_seq), device=device)\r\n seq_cont_decoded = torch.Tensor(device=device)\r\n seq_cat_decoded = []\r\n for _ in range(self.unified_encoder.seq_cat_count):\r\n seq_cat_decoded.append(torch.Tensor(device=device))\r\n\r\n for _ in range(encoder_out.shape[1]):\r\n decoded, (hn, cn), out_cont, out_cat = self.decoder(decoded, (hn, cn))\r\n # Predict all categorical columns\r\n out_cat_onehot = []\r\n if self.unified_encoder.seq_cat_count != 0:\r\n for idx, out in enumerate(out_cat):\r\n out_cat_onehot.append(torch.argmax(out, dim=1).unsqueeze(-1))\r\n seq_cat_decoded[idx] = torch.cat(\r\n [seq_cat_decoded[idx], out.view(out.shape[0], 1, -1)], dim=1)\r\n out_cat_onehot = torch.cat(out_cat_onehot, -1)\r\n out_cat_embedding = self.unified_encoder.seq_emb_layers(out_cat_onehot)\r\n decoded = torch.cat([out_cat_embedding, out_cont], dim=-1)\r\n else:\r\n decoded = out_cont\r\n seq_cont_decoded = torch.cat(\r\n [seq_cont_decoded, out_cont.view(out_cont.shape[0], 1, -1)], dim=1)\r\n\r\n return non_sequential_cont_decoded, non_sequential_cat_decoded, seq_cont_decoded, seq_cat_decoded","def slide_step(self, dataset='slide', input_stream='img', lambda_coeff=1):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n\n (x2, len2), (x1, x1_mask, img_loc, img_id), labels= self.get_batch('slide2img', dataset, input_stream)\n # convert fp16\n #assign lang ids\n # follow uvp\n\n img_len = x1_mask.sum(dim=1)\n x_img = x1.transpose(0, 1)\n img_loc = img_loc.transpose(0, 1)\n\n x2, len2, x_img, img_loc, img_len = to_cuda(x2, len2, x_img, img_loc, img_len)\n\n encoder_outputs = model('jointfwd', x=x2, lengths=len2, x_img=x_img, lengths_img=img_len, causal=False,\n langs=None,\n image_loc=img_loc, refine_image=params.refine_image)\n\n encoder_outputs = encoder_outputs.transpose(0, 1)\n\n relation_scores = model('predict', tensor=encoder_outputs, is_relation=True)\n\n # loss\n # the last word not apply prediction logic\n s2 = torch.from_numpy(np.array(labels, dtype='float32')).view(-1)\n loss = F.binary_cross_entropy_with_logits(relation_scores.view(-1).cpu(),\n s2)\n\n\n self.stats[('SLIDE-%s' % (input_stream))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def fit_transform(self, X, y, blk_start_loc=None):\n # normalize y\n self.scale = y.std()\n y = y / self.scale\n M, N = X.shape\n # automatically set block partition\n if blk_start_loc is None:\n blkLen = int(N / 16.0)\n blk_start_loc = np.arange(0, N, blkLen)\n self.blk_len_list, self.is_equal_block = block_parse(blk_start_loc, N)\n # init variables\n self.init(X, y, blk_start_loc)\n # bootstrap ADD one basis\n ml, A, theta = self.logobj_mapping()\n idx = ml.argmin(0)\n Sig, w, Xu = self.bootup(A, idx)\n # loops\n ML = np.zeros(self.max_iters)\n ML[0] = ml[idx]\n for count in range(1, self.max_iters):\n ml, A, theta = self.logobj_mapping()\n idx = ml.argmin(0)\n\n # check convergence now\n ML[count] = ml[idx]\n if ML[count] >= 0:\n break\n if count > 1:\n ml_ratio = np.abs(ML[count] - ML[count - 1]) / np.abs(ML[count] - ML[0])\n if ml_ratio < self.epsilon:\n break\n\n # operation on basis\n if self.index[idx] is True:\n if theta[idx] > self.prune_gamma:\n proc = self.estimate\n else:\n proc = self.delete\n else:\n proc = self.add\n # process Sig, w, Xu\n Sig, w, Xu = proc(Sig, w, Xu, A, idx)\n\n # exit\n self.count = count\n return self.w_format(w)","def __init__(self, params, model, name=\"ds2_encoder\", mode='train'):\n super(DeepSpeech2Encoder, self).__init__(params, model, name, mode)","def test_transform_prep_proba():\n args = get_layer('transform', 'manual', 'temporal', True, True, window=2, step_size=3)\n run_layer(*args)","def transform(self, inputs: list, stage: str) -> datapack.DataPack:","def __build_mol_to_latent_model(self):\n\n # Input tensor (MANDATORY)\n encoder_inputs = Input(shape=self.input_shape, name=\"Encoder_Inputs\")\n\n x = encoder_inputs\n\n # The two encoder layers, number of cells are halved as Bidirectional\n encoder = Bidirectional(\n LSTM(\n self.lstm_dim // 2,\n return_sequences=True,\n return_state=True, # Return the states at end of the batch\n name=\"Encoder_LSTM_1\",\n )\n )\n\n x, state_h, state_c, state_h_reverse, state_c_reverse = encoder(x)\n\n if self.bn:\n x = BatchNormalization(momentum=self.bn_momentum, name=\"BN_1\")(x)\n\n encoder2 = Bidirectional(\n LSTM(\n self.lstm_dim // 2,\n return_state=True, # Return the states at end of the batch\n name=\"Encoder_LSTM_2\",\n )\n )\n\n _, state_h2, state_c2, state_h2_reverse, state_c2_reverse = encoder2(x)\n\n # Concatenate all states of the forward and the backward LSTM layers\n states = Concatenate(axis=-1, name=\"Concatenate_1\")(\n [\n state_h,\n state_c,\n state_h2,\n state_c2,\n state_h_reverse,\n state_c_reverse,\n state_h2_reverse,\n state_c2_reverse,\n ]\n )\n\n if self.bn:\n states = BatchNormalization(momentum=self.bn_momentum, name=\"BN_2\")(states)\n\n # A non-linear recombination\n neck_relu = Dense(\n self.codelayer_dim, activation=self.h_activation, name=\"Codelayer_Relu\"\n )\n neck_outputs = neck_relu(states)\n\n if self.bn:\n neck_outputs = BatchNormalization(\n momentum=self.bn_momentum, name=\"BN_Codelayer\"\n )(neck_outputs)\n\n # Add Gaussian noise to \"spread\" the distribution of the latent variables during training\n neck_outputs = GaussianNoise(self.noise_std, name=\"Gaussian_Noise\")(\n neck_outputs\n )\n\n # Define the model\n self.__mol_to_latent_model = Model(encoder_inputs, neck_outputs)\n\n # Name it!\n self.mol_to_latent_model.name = \"mol_to_latent_model\"","def TransformerEncoderLayer(self, name, stride=1, first_n=None,\n ff_hidden_dim=None, num_heads=None,\n use_moe=False):\n p = self.params\n if ff_hidden_dim is None:\n ff_hidden_dim = p.ff_hidden_dim\n if num_heads is None:\n num_heads = p.num_heads\n if use_moe:\n ffw_p = self.MoE('moe', ff_hidden_dim=ff_hidden_dim)\n else:\n ffw_p = self.Feedforward('ff', ff_hidden_dim=ff_hidden_dim)\n return self._Seq(name, self._Seq(\n 'block',\n self._StridedAttention('self_atten', stride=stride,\n first_n=first_n, num_heads=num_heads),\n ffw_p))","def DontuseThis():\n BCM_outputs = ['phi','rho','theta',\n 'r_probabilityMaps','l_probabilityMaps',\n 'models']\n BCM_Models = pe.Node(interface=nio.DataGrabber(input_names=['structures'],\n outfields=BCM_outputs),\n name='10_BCM_Models')\n BCM_Models.inputs.base_directory = atlas_fname_wpath\n BCM_Models.inputs.template_args['phi'] = [['spatialImages','phi','nii.gz']]\n BCM_Models.inputs.template_args['rho'] = [['spatialImages','rho','nii.gz']]\n BCM_Models.inputs.template_args['theta'] = [['spatialImages','theta','nii.gz']]\n BCM_Models.inputs.template_args['r_probabilityMaps'] = [['structures']]\n BCM_Models.inputs.template_args['l_probabilityMaps'] = [['structures']]\n BCM_Models.inputs.template_args['models'] = [['structures']]\n\n BRAINSCut_structures = ['caudate','thalamus','putamen','hippocampus']\n #BRAINSCut_structures = ['caudate','thalamus']\n BCM_Models.iterables = ( 'structures', BRAINSCut_structures )\n BCM_Models.inputs.template = '%s/%s.%s'\n BCM_Models.inputs.field_template = dict(\n r_probabilityMaps='probabilityMaps/r_%s_ProbabilityMap.nii.gz',\n l_probabilityMaps='probabilityMaps/l_%s_ProbabilityMap.nii.gz',\n models='modelFiles/%sModel*',\n )\n\n \"\"\"\n The xml creation and BRAINSCut need to be their own mini-pipeline that gets\n executed once for each of the structures in BRAINSCut_structures. This can be\n accomplished with a map node and a new pipeline.\n \"\"\"\n \"\"\"\n Create xml file for BRAINSCut\n \"\"\"\n\n\n BFitAtlasToSubject = pe.Node(interface=BRAINSFit(),name=\"BFitAtlasToSubject\")\n BFitAtlasToSubject.inputs.costMetric=\"MMI\"\n BFitAtlasToSubject.inputs.maskProcessingMode=\"ROI\"\n BFitAtlasToSubject.inputs.numberOfSamples=100000\n BFitAtlasToSubject.inputs.numberOfIterations=[1500,1500]\n BFitAtlasToSubject.inputs.numberOfHistogramBins=50\n BFitAtlasToSubject.inputs.maximumStepLength=0.2\n BFitAtlasToSubject.inputs.minimumStepLength=[0.005,0.005]\n BFitAtlasToSubject.inputs.transformType= [\"Affine\",\"BSpline\"]\n BFitAtlasToSubject.inputs.maxBSplineDisplacement= 7\n BFitAtlasToSubject.inputs.maskInferiorCutOffFromCenter=65\n BFitAtlasToSubject.inputs.splineGridSize=[28,20,24]\n BFitAtlasToSubject.inputs.outputVolume=\"Trial_Initializer_Output.nii.gz\"\n BFitAtlasToSubject.inputs.outputTransform=\"Trial_Initializer_Output.mat\"\n cutWF.connect(SplitAvgBABC,'avgBABCT1',BFitAtlasToSubject,'fixedVolume')\n cutWF.connect(BABC,'outputLabels',BFitAtlasToSubject,'fixedBinaryVolume')\n cutWF.connect(BAtlas,'template_t1',BFitAtlasToSubject,'movingVolume')\n cutWF.connect(BAtlas,'template_brain',BFitAtlasToSubject,'movingBinaryVolume')\n cutWF.connect(BLI,'outputTransformFilename',BFitAtlasToSubject,'initialTransform')\n\n CreateBRAINSCutXML = pe.Node(Function(input_names=['rho','phi','theta',\n 'model',\n 'r_probabilityMap',\n 'l_probabilityMap',\n 'atlasT1','atlasBrain',\n 'subjT1','subjT2',\n 'subjT1GAD','subjT2GAD',\n 'subjSGGAD','subjBrain',\n 'atlasToSubj','output_dir'],\n output_names=['xml_filename','rl_structure_filename_list'],\n function = create_BRAINSCut_XML),\n overwrite = True,\n name=\"CreateBRAINSCutXML\")\n\n ## HACK Makde better directory\n CreateBRAINSCutXML.inputs.output_dir = \".\" #os.path.join(cutWF.base_dir, \"BRAINSCut_output\")\n cutWF.connect(BCM_Models,'models',CreateBRAINSCutXML,'model')\n cutWF.connect(BCM_Models,'rho',CreateBRAINSCutXML,'rho')\n cutWF.connect(BCM_Models,'phi',CreateBRAINSCutXML,'phi')\n cutWF.connect(BCM_Models,'theta',CreateBRAINSCutXML,'theta')\n cutWF.connect(BCM_Models,'r_probabilityMaps',CreateBRAINSCutXML,'r_probabilityMap')\n cutWF.connect(BCM_Models,'l_probabilityMaps',CreateBRAINSCutXML,'l_probabilityMap')\n cutWF.connect(BAtlas,'template_t1',CreateBRAINSCutXML,'atlasT1')\n cutWF.connect(BAtlas,'template_brain',CreateBRAINSCutXML,'atlasBrain')\n cutWF.connect(SplitAvgBABC,'avgBABCT1',CreateBRAINSCutXML,'subjT1')\n cutWF.connect(SplitAvgBABC,'avgBABCT2',CreateBRAINSCutXML,'subjT2')\n cutWF.connect(GADT1,'outputVolume',CreateBRAINSCutXML,'subjT1GAD')\n cutWF.connect(GADT2,'outputVolume',CreateBRAINSCutXML,'subjT2GAD')\n cutWF.connect(SGI,'outputFileName',CreateBRAINSCutXML,'subjSGGAD')\n cutWF.connect(BABC,'outputLabels',CreateBRAINSCutXML,'subjBrain')\n cutWF.connect(BFitAtlasToSubject,'outputTransform',CreateBRAINSCutXML,'atlasToSubj')\n #CreateBRAINSCutXML.inputs.atlasToSubj = \"INTERNAL_REGISTER.mat\"\n #cutWF.connect(BABC,'atlasToSubjectTransform',CreateBRAINSCutXML,'atlasToSubj')\n\n \"\"\"\n ResampleNACLabels\n \"\"\"\n ResampleAtlasNACLabels=pe.Node(interface=BRAINSResample(),name=\"ResampleAtlasNACLabels\")\n ResampleAtlasNACLabels.inputs.interpolationMode = \"NearestNeighbor\"\n ResampleAtlasNACLabels.inputs.outputVolume = \"atlasToSubjectNACLabels.nii.gz\"\n\n cutWF.connect(cutWF,'OutputSpec.atlasToSubjectTransform',ResampleAtlasNACLabels,'warpTransform')\n cutWF.connect(cutWF,'OutputSpec.t1_corrected',ResampleAtlasNACLabels,'referenceVolume')\n cutWF.connect(BAtlas,'template_nac_lables',ResampleAtlasNACLabels,'inputVolume')\n\n \"\"\"\n BRAINSMush\n \"\"\"\n BMUSH=pe.Node(interface=BRAINSMush(),name=\"BMUSH\")\n BMUSH.inputs.outputVolume = \"MushImage.nii.gz\"\n BMUSH.inputs.outputMask = \"MushMask.nii.gz\"\n BMUSH.inputs.lowerThresholdFactor = 1.2\n BMUSH.inputs.upperThresholdFactor = 0.55\n\n cutWF.connect(myLocalTCWF,'OutputSpec.t1_corrected',BMUSH,'inputFirstVolume')\n cutWF.connect(myLocalTCWF,'OutputSpec.t2_corrected',BMUSH,'inputSecondVolume')\n cutWF.connect(myLocalTCWF,'OutputSpec.outputLabels',BMUSH,'inputMaskVolume')\n\n \"\"\"\n BRAINSROIAuto\n \"\"\"\n BROI = pe.Node(interface=BRAINSROIAuto(), name=\"BRAINSROIAuto\")\n BROI.inputs.closingSize=12\n BROI.inputs.otsuPercentileThreshold=0.01\n BROI.inputs.thresholdCorrectionFactor=1.0\n BROI.inputs.outputROIMaskVolume = \"temproiAuto_t1_ACPC_corrected_BRAINSABC.nii.gz\"\n cutWF.connect(myLocalTCWF,'OutputSpec.t1_corrected',BROI,'inputVolume')\n\n \"\"\"\n Split the implicit outputs of BABCext\n \"\"\"\n SplitAvgBABC = pe.Node(Function(input_names=['in_files','T1_count'], output_names=['avgBABCT1','avgBABCT2'],\n function = get_first_T1_and_T2), run_without_submitting=True, name=\"99_SplitAvgBABC\")\n SplitAvgBABC.inputs.T1_count = 1 ## There is only 1 average T1 image.\n\n cutWF.connect(myLocalTCWF,'OutputSpec.outputAverageImages',SplitAvgBABC,'in_files')\n\n\n\n def printFullPath(outFileFullPath):\n print(\"=\"*80)\n print(\"=\"*80)\n print(\"=\"*80)\n print(\"=\"*80)\n print(\"{0}\".format(outFileFullPath))\n return outFileFullPath\n printOutImage = pe.Node( Function(function=printFullPath, input_names = ['outFileFullPath'], output_names = ['genoutFileFullPath']), run_without_submitting=True, name=\"99_printOutImage\")\n cutWF.connect( GADT2, 'outputVolume', printOutImage, 'outFileFullPath' )","def preprocessing_transform(self, x):\n if isinstance(self.column_text, int) and self.column_text not in x.columns:\n col = self.column_text\n else:\n col = list(x.columns).index(self.column_text)\n\n ct = x.shape[0]\n # INPUTS\n if self.method_embedding.lower() in ['roberta', \"camembert\", \"xlm-roberta\"]:\n ids = np.ones((ct, self.maxlen), dtype='int32')\n else:\n ids = np.zeros((ct, self.maxlen), dtype='int32')\n att = np.zeros((ct, self.maxlen), dtype='int32')\n tok = np.zeros((ct, self.maxlen), dtype='int32')\n\n for k in range(ct):\n text = \" \" + \" \".join(x.iloc[k, col].split())\n\n if self.method_embedding == 'RoBERTa':\n enc = self.tokenizer.encode(text)\n else:\n enc = self.tokenizer.encode(text, max_length=self.maxlen, truncation=True)\n\n # CREATE BERT INPUTS\n if self.method_embedding == 'RoBERTa':\n ids[k, :len(enc.ids)] = enc.ids[:self.maxlen]\n att[k, :len(enc.ids)] = 1\n else:\n ids[k, :len(enc)] = enc\n att[k, :len(enc)] = 1\n\n x_preprocessed = [ids, att, tok]\n if self.dimension_embedding == 'word_embedding':\n return x_preprocessed\n else:\n model_extractor = self.model_extract_document_embedding()\n document_embedding = model_extractor.predict(x_preprocessed)\n return document_embedding","def __init__(self, vocab_size, embedding_size, output_size, lat_dim):\n super(Encoder, self).__init__()\n\n self.vocab_size = vocab_size\n self.embedding = nn.Embedding(vocab_size, embedding_size)\n self.lstm = nn.LSTM(embedding_size, output_size)\n \n self.hidden2mean = nn.Linear(output_size, lat_dim)\n self.hidden2logv = nn.Linear(output_size, lat_dim)\n \n self.cell2mean = nn.Linear(output_size, lat_dim)\n self.cell2logv = nn.Linear(output_size, lat_dim)","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n TransformerDecoderBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n multi_channel_cross_attention=self.multi_channel_cross_attention,\n name=(\"layer_%d\" % i)))\n super(TransformerDecoder, self).build(unused_input_shapes)","def __build_ml_pipeline(self, clf: MultiOutputClassifier) -> Pipeline:\r\n pipeline = Pipeline([\r\n ('features', FeatureUnion(\r\n [('text', Pipeline(\r\n [('text_field_extractor', \r\n basic_utils.TextFieldExtractor('message')), \r\n #('vect', CountVectorizer(tokenizer=clean_text)),\r\n #('tfidf', TfidfTransformer())\r\n ('tfidf', TfidfVectorizer(tokenizer=basic_utils.tokenize, \r\n min_df=.0025, max_df=0.25, \r\n ngram_range=(1,2)))\r\n ])),\r\n \r\n ('numerics', FeatureUnion(\r\n [('text_len', \r\n Pipeline([('text_len_extractor', \r\n basic_utils.NumericFieldExtractor('text_len')), \r\n ('text_len_scaler', StandardScaler())\r\n ])),\r\n ('punt_perc', \r\n Pipeline([('punt_perc_extractor', \r\n basic_utils.NumericFieldExtractor('punt_perc')), \r\n ('punt_perc_scaler', StandardScaler())\r\n ]))\r\n ])),\r\n ('starting_verb', basic_utils.PosFieldExtractor('starting_verb_flag'))\r\n ])),\r\n \r\n ('clf', clf)\r\n ])\r\n \r\n return pipeline","def make_encoder(self, input_size: int, latent_size: int) -> nn.Module:\n pass","def _create_encoder(self):\n\n def _init_weights(layer):\n \"\"\"Initializes the weights of a layer based on type.\"\"\"\n if isinstance(layer, (nn.Conv2d, nn.Linear)):\n torch.nn.init.xavier_uniform_(layer.weight)\n try:\n # Some layers may not have biases, so catch the exception and pass.\n layer.bias.data.fill_(0.0)\n except AttributeError:\n pass\n\n kernel_size = 5\n pad = 2\n input_channels = 1\n first_conv_channels = 6\n second_conv_channels = 16\n max_pool_kernel = 2\n linear_size = 120\n n_pixels = 7\n\n encoder = nn.Sequential(\n nn.Conv2d(\n input_channels, first_conv_channels, kernel_size, padding=pad),\n nn.BatchNorm2d(first_conv_channels),\n nn.ReLU(),\n nn.MaxPool2d(max_pool_kernel),\n nn.Conv2d(\n first_conv_channels, second_conv_channels, kernel_size,\n padding=pad),\n nn.BatchNorm2d(second_conv_channels),\n nn.ReLU(),\n nn.MaxPool2d(max_pool_kernel),\n utils.Flatten(),\n nn.Linear(n_pixels * n_pixels * self.n_digits * second_conv_channels,\n linear_size),\n nn.BatchNorm1d(linear_size),\n nn.ReLU(),\n nn.Linear(linear_size, self.embedding_dim),\n nn.Linear(self.embedding_dim, self.n_classes, bias=False),\n )\n\n encoder.apply(_init_weights)\n\n # This is the empirical approximation for initialization the vMF\n # distributions for each class in the final layer.\n if self.use_vmf:\n utils.vmf_class_weight_init(encoder[-1].weight, self.kappa_confidence,\n self.embedding_dim)\n\n return encoder","def __init__(self,\n input_feature: int,\n output_feature: int = 19,\n n_permutation_list: int = 6,\n dropout: float = 0.2,\n UseBN: bool = True):\n super(LCNNBlock, self).__init__()\n\n self.conv_weights = nn.Linear(input_feature, output_feature)\n self.batch_norm = nn.BatchNorm1d(output_feature)\n self.UseBN = UseBN\n self.activation = Shifted_softplus()\n self.dropout = Custom_dropout(dropout, n_permutation_list)\n self.permutation = n_permutation_list"],"string":"[\n \"def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\\n super(TransformerEncoder, self).__init__()\\n\\n # Create num_layers of TransformerBlock\\n self.encoder = Sequential(\\n [\\n TransformerBlock(num_heads=num_heads,\\n D=D,\\n hidden_layers=[mlp_dim, D],\\n dropout=dropout,\\n norm_eps=norm_eps)\\n for _ in range(num_layers)\\n ]\\n )\",\n \"def forward(self, x):\\n # Encoder1 --block1\\n encode_block1 = self.conv_encoder1(x)\\n if self.residus[0] == 1:\\n encode_block1 += self.residual_shortcut1(x)\\n encode_pool1 = self.max_pool_encoder1(encode_block1)\\n\\n # Encoder2 --block2\\n encode_block2 = self.conv_encoder2(encode_pool1)\\n if self.residus[1] == 1:\\n encode_block2 += self.residual_shortcut2(encode_pool1)\\n encode_pool2 = self.max_pool_encoder2(encode_block2)\\n \\n # Encoder3 --block3\\n encode_block3 = self.conv_encoder3(encode_pool2)\\n if self.residus[2] == 1:\\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\\n encode_pool3 = self.max_pool_encoder3(encode_block3)\\n \\n # Encoder4 --block4\\n encode_block4 = self.conv_encoder4(encode_pool3) \\n if self.residus[3] == 1:\\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\\n encode_pool4 = self.max_pool_encoder4(encode_block4)\\n\\n # Transitional block \\n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\\n \\n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\\n \\n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\\n\\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\\n \\n convTranspose_transitional = self.convTranspose_transitional(middle_block) \\n # Decoder4 --block5\\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \\n if self.residus[4] == 1:\\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\\n\\n #--block6\\n cat_layer3 = self.conv_decoder4(decode_block4) \\n if self.residus[5] == 1:\\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\\n \\n \\n # Decoder3 --block7\\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\\n if self.residus[6] == 1:\\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\\n \\n #--block8\\n cat_layer2 = self.conv_decoder3(decode_block3) \\n if self.residus[7] == 1:\\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\\n \\n # Decoder2 --block9\\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \\n if self.residus[8] == 1:\\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\\n \\n #--block10\\n cat_layer1 = self.conv_decoder2(decode_block2)\\n if self.residus[9] == 1:\\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\\n \\n # Decoder1 --block11\\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \\n if self.residus[10] == 1:\\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\\n \\n #--block12\\n final_layer = self.final_layer(decode_block1)\\n if self.residus[11] == 1:\\n final_layer += self.residual_shortcut_final_layer(decode_block1)\\n \\n \\n return final_layer\",\n \"def __init__(self, unified_encoder, mlp):\\r\\n\\r\\n super().__init__()\\r\\n\\r\\n self.unified_encoder = unified_encoder\\r\\n self.mlp = mlp\",\n \"def TransformerEncoderLayer(self, name, is_causal=True):\\n # Hack to be compatible with ckpt generated by self._rep\\n return self._Seq(\\n name,\\n self._Seq('block', self._Attention('self_atten', is_causal=is_causal),\\n self.Feedforward('ff')))\",\n \"def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):\\n super(TransformerBlock, self).__init__()\\n # Attention\\n self.attention = MultiHeadAttention(\\n num_heads=num_heads, key_dim=D, dropout=dropout\\n )\\n self.norm_attention = LayerNormalization(epsilon=norm_eps)\\n\\n # MLP\\n self.mlp = MLPBlock(hidden_layers, dropout)\\n self.norm_mlp = LayerNormalization(epsilon=norm_eps)\",\n \"def __init__(self, enc, mlp):\\r\\n\\r\\n super().__init__()\\r\\n\\r\\n self.enc = enc\\r\\n self.mlp = mlp\",\n \"def transform(self):\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\r\\n super(EncoderBlock, self).__init__()\\r\\n self.mha = MultiHeadAttention(dm, h)\\r\\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\\r\\n activation='relu')\\r\\n self.dense_output = tf.keras.layers.Dense(units=dm)\\r\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\r\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\",\n \"def forward(self, x):\\n # Encode\\n encode_block1 = self.conv_encoder1(x)\\n encode_pool1 = self.max_pool_encoder1(encode_block1)\\n encode_block2 = self.conv_encoder2(encode_pool1)\\n encode_pool2 = self.max_pool_encoder2(encode_block2)\\n encode_block3 = self.conv_encoder3(encode_pool2)\\n encode_pool3 = self.max_pool_encoder3(encode_block3)\\n encode_block4 = self.conv_encoder4(encode_pool3)\\n encode_pool4 = self.max_pool_encoder4(encode_block4)\\n\\n # Transitional block\\n middle_block = self.transitional_block(encode_pool4)\\n\\n # Decode\\n decode_block4 = torch.cat((middle_block, encode_block4), 1)\\n cat_layer3 = self.conv_decoder4(decode_block4)\\n decode_block3 = torch.cat((cat_layer3, encode_block3), 1)\\n cat_layer2 = self.conv_decoder3(decode_block3)\\n decode_block2 = torch.cat((cat_layer2, encode_block2), 1)\\n cat_layer1 = self.conv_decoder2(decode_block2)\\n decode_block1 = torch.cat((cat_layer1, encode_block1), 1)\\n final_layer = self.final_layer(decode_block1)\\n return final_layer\",\n \"def mtf_bitransformer_base():\\n hparams = mtf_transformer2_base()\\n hparams.max_length = 256\\n hparams.shared_embedding = True\\n # HYPERPARAMETERS FOR THE LAYER STACKS\\n hparams.add_hparam(\\\"encoder_layers\\\", [\\\"self_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"decoder_layers\\\", [\\\"self_att\\\", \\\"enc_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"encoder_num_layers\\\", 6)\\n hparams.add_hparam(\\\"decoder_num_layers\\\", 6)\\n # number of heads in multihead attention\\n hparams.add_hparam(\\\"encoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"decoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"local_attention_radius\\\", 128)\\n\\n # default of 0 for standard transformer behavior\\n # 1 means a single set of keys and values that are read by all query heads\\n hparams.add_hparam(\\\"encoder_num_memory_heads\\\", 0)\\n hparams.add_hparam(\\\"decoder_num_memory_heads\\\", 0)\\n # share attention keys and values\\n hparams.add_hparam(\\\"encoder_shared_kv\\\", False)\\n hparams.add_hparam(\\\"decoder_shared_kv\\\", False)\\n\\n # Parameters for computing the maximum decode length in beam search.\\n # Maximum decode length is:\\n # min(max_length,\\n # decode_length_multiplier * input_length + decode_length_constant)\\n hparams.add_hparam(\\\"decode_length_multiplier\\\", 1.5)\\n hparams.add_hparam(\\\"decode_length_constant\\\", 10.0)\\n # used during decoding\\n hparams.add_hparam(\\\"alpha\\\", 0.6)\\n hparams.sampling_temp = 0.0\\n return hparams\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n CustomTransformerBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n backward_compatible=self.backward_compatible,\\n float_type=self.float_type,\\n name=(\\\"layer_%d\\\" % i)))\\n super(CustomTransformer, self).build(unused_input_shapes)\",\n \"def __init__(self, momentum: float = .5):\\n super(VanillaEncoder, self).__init__()\\n self.conv1 = PointNetConv2Layer(64, momentum)\\n self.conv2 = PointNetConv2Layer(64, momentum)\\n self.conv3 = PointNetConv2Layer(64, momentum)\\n self.conv4 = PointNetConv2Layer(128, momentum)\\n self.conv5 = PointNetConv2Layer(1024, momentum)\",\n \"def __init__(self, ):\\n super().__init__()\\n channels = 3\\n\\n # Initial convolution block\\n out_features = 64\\n # encoder\\n self.input = nn.Sequential(\\n nn.ReflectionPad2d(channels),\\n nn.Conv2d(3, out_features, (7, 7)),\\n nn.InstanceNorm2d(out_features),\\n nn.MaxPool2d(2),\\n nn.LeakyReLU(0.2, inplace=True),\\n nn.Conv2d(out_features, out_features * 2, (3, 3), stride=(1, 1), padding=(1, 1)),\\n nn.LeakyReLU(0.2, inplace=True),\\n nn.MaxPool2d(2),\\n nn.Conv2d(out_features * 2, out_features * 4, (3, 3), stride=(1, 1), padding=(1, 1)),\\n nn.LeakyReLU(0.2, inplace=True),\\n )\\n\\n self.in0 = nn.InstanceNorm2d(256)\\n self.block0 = blocks()\\n self.block1 = blocks()\\n self.block2 = blocks()\\n self.block3 = blocks()\\n self.block4 = blocks()\\n self.block5 = blocks()\\n self.block6 = blocks()\\n self.block7 = blocks()\\n\\n self.out = nn.Sequential(\\n\\n nn.Upsample(scale_factor=2),\\n nn.Conv2d(out_features * 4, out_features * 2, 3, stride=1, padding=1),\\n nn.InstanceNorm2d(out_features * 2),\\n nn.LeakyReLU(0.2, inplace=True),\\n\\n nn.Upsample(scale_factor=2),\\n nn.Conv2d(out_features * 2, out_features, 3, stride=1, padding=1),\\n nn.InstanceNorm2d(out_features),\\n nn.LeakyReLU(0.2, inplace=True),\\n\\n nn.ReflectionPad2d(channels),\\n nn.Conv2d(out_features, channels, 7),\\n nn.Tanh(),\\n )\",\n \"def mt_step(self, lang1, lang2, lambda_coeff):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n lang1_id = params.lang2id[lang1]\\n lang2_id = params.lang2id[lang2]\\n\\n # generate batch\\n if lang1 == lang2:\\n (x1, len1) = self.get_batch('ae', lang1)\\n (x2, len2) = (x1, len1)\\n (x1, len1) = self.add_noise(x1, len1)\\n else:\\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\\n langs1 = x1.clone().fill_(lang1_id)\\n langs2 = x2.clone().fill_(lang2_id)\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n # key point generate label\\n\\n # cuda\\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\\n\\n # encode source sentence\\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\\n enc1 = enc1.transpose(0, 1)\\n\\n # decode target sentence\\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\\n src_len=len1)\\n\\n # loss\\n # the last word not apply prediction logic\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def bart_mlm_step(self, lang1, lang2, lambda_coeff):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n lang1_id = params.lang2id[lang1]\\n lang2_id = params.lang2id[lang1]\\n\\n # generate batch / select words to predict\\n x, lengths, positions, langs, _ = self.generate_batch(lang1, lang2, 'pred')\\n x, lengths, positions, langs, _ = self.round_batch(x, lengths, positions, langs)\\n\\n (x1, len1, x2, len2, y, pred_mask, positions) = self.bart_token_mask_sent(x, lengths)\\n\\n if params.use_noise:\\n x1,len1 = self.add_noise(x1,len1)\\n langs1 = x1.clone().fill_(lang1_id)\\n langs2 = x2.clone().fill_(lang2_id)\\n\\n # cuda\\n x1,x2,len1,len2, y, pred_mask, positions,langs1,langs2 = to_cuda(x1,x2,len1,len2, y, pred_mask, positions,langs1,langs2)\\n\\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\\n enc1 = enc1.transpose(0, 1)\\n\\n # enc_mask = x1.ne(params.mask_index)\\n # enc_mask = enc_mask.transpose(0, 1)\\n\\n dec2 = model('crossfwd', stream_='text',\\n x=x2, lengths=len2, langs=langs2, causal=True,\\n src_enc=enc1, src_len=len1)\\n\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n\\n self.stats[('M-BART-%s' % lang1) if lang2 is None else ('MLM-%s-%s' % (lang1, lang2))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += lengths.size(0)\\n self.stats['processed_w'] += pred_mask.sum().item()\",\n \"def mt_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n # _lang1 = _lang2 = 'en'\\n # lang2_id = params.lang2id[_lang2]\\n\\n # generate batch\\n\\n (x_src,len_src),(x2, len2),(x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\\n # convert fp16\\n #assign lang ids\\n\\n\\n lang0_id = params.lang2id[params.ft_lgs[0]]\\n lang_src = x_src.clone().fill_(lang0_id)\\n lang1_id = params.lang2id[params.ft_lgs[1]]\\n langs = x2.clone().fill_(lang1_id)\\n\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n #\\n len1 = x1_mask.sum(dim=1)\\n x1 = x1.transpose(0, 1)\\n img_loc = img_loc.transpose(0, 1)\\n\\n # cuda\\n\\n x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src)\\n\\n if params.mt_only_text:\\n encoder_outputs = model('crossfwd', stream_='text', x=x_src, lengths=len_src, langs=lang_src, causal=False,refine_image=params.refine_image)\\n len_all = len_src\\n else:\\n encoder_outputs = model('jointfwd', x=x_src, lengths=len_src, x_img=x1, lengths_img=len1, causal=False,\\n langs=None,\\n image_loc=img_loc, refine_image=params.refine_image)\\n len_all = len_src + len1\\n\\n # encode source sentence\\n # enc1 = model('jo', stream_='img', x=x1, lengths=len1, langs=None, causal=False, cross_modal=True,\\n # image_loc=img_loc, refine_image=params.refine_image, refine_encoder=params.refine_encoder,\\n # image_dist=None)\\n enc1 = encoder_outputs.transpose(0, 1)\\n\\n\\n # decode target sentence\\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs, causal=True, src_enc=enc1,\\n src_len=len_all)\\n\\n # loss\\n # the last word not apply prediction logic\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n self.stats[('IC-%s-%s' % (dataset, input_stream))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def transform():\",\n \"def __init__(self):\\n super(Encoder2, self).__init__()\\n self.lblocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n ]\\n )\\n\\n self.blocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n ]\\n )\",\n \"def transform():\\n pass\",\n \"def forward(self, x):\\n skips = {}\\n skips[\\\"input\\\"] = x\\n\\n # Encoder\\n skips[\\\"pool1_out\\\"] = self.enc_block1(x)\\n skips[\\\"pool2_out\\\"] = self.enc_block2(skips[\\\"pool1_out\\\"])\\n skips[\\\"pool3_out\\\"] = self.enc_block3(skips[\\\"pool2_out\\\"])\\n skips[\\\"pool4_out\\\"] = self.enc_block4(skips[\\\"pool3_out\\\"])\\n output = self.enc_block5(skips[\\\"pool4_out\\\"])\\n output = self.enc_block6(output)\\n\\n # Decoder\\n output = self.upsample5(output)\\n output = self.dec_block1(torch.cat([skips[\\\"pool4_out\\\"], output], dim=1))\\n output = self.dec_block2(torch.cat([skips[\\\"pool3_out\\\"], output], dim=1))\\n output = self.dec_block3(torch.cat([skips[\\\"pool2_out\\\"], output], dim=1))\\n output = self.dec_block4(torch.cat([skips[\\\"pool1_out\\\"], output], dim=1))\\n output = self.dec_block5(torch.cat([skips[\\\"input\\\"], output], dim=1))\\n\\n return output\",\n \"def preprocess(self):\",\n \"def __init__(self):\\n super(Encoder3, self).__init__()\\n self.lblocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n ]\\n )\\n\\n self.blocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n ]\\n )\",\n \"def to_basic_block(self) -> \\\"gr::basic_block_sptr\\\":\\n return _beamforming_swig.beamformer_sptr_to_basic_block(self)\",\n \"def mobilenetv2 (inputs, k, alpha = 1.0, train_bn = False):\\n\\n x = conv_block(inputs, 32, alpha, (3, 3), strides=(2, 2), block_id=0, train_bn=train_bn) # Input Res: 1\\n\\n x = inverted_residual_block(x, 16, (3, 3), t=1, strides=1, n=1, alpha=1.0, block_id=1, train_bn=train_bn)\\t# Input Res: 1/2\\n x = inverted_residual_block(x, 24, (3, 3), t=6, strides=2, n=2, alpha=1.0, block_id=2, train_bn=train_bn)\\t# Input Res: 1/2\\n x = inverted_residual_block(x, 32, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=4, train_bn=train_bn)\\t# Input Res: 1/4\\n x = inverted_residual_block(x, 64, (3, 3), t=6, strides=2, n=4, alpha=1.0, block_id=7, train_bn=train_bn)\\t# Input Res: 1/8\\n x = inverted_residual_block(x, 96, (3, 3), t=6, strides=1, n=3, alpha=1.0, block_id=11, train_bn=train_bn)\\t# Input Res: 1/8\\n x = inverted_residual_block(x, 160, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=14, train_bn=train_bn)\\t# Input Res: 1/16\\n x = inverted_residual_block(x, 320, (3, 3), t=6, strides=1, n=1, alpha=1.0, block_id=17, train_bn=train_bn)\\t# Input Res: 1/32\\n\\n x = conv_block(x, 1280, alpha, (1, 1), strides=(1, 1), block_id=18, train_bn=train_bn) # Input Res: 1/32\\n\\n x = KL.GlobalAveragePooling2D()(x)\\n x = KL.Reshape((1, 1, 1280))(x)\\n x = KL.Dropout(0.3, name='Dropout')(x)\\n x = KL.Conv2D(k, (1, 1), padding='same')(x)\\n\\n x = KL.Activation('softmax', name='softmax')(x)\\n output = KL.Reshape((k,))(x)\\n\\n model = KM.Model(inputs, output)\\n plot_model(model, to_file='MobileNetv2.png', show_shapes=True)\\n\\n return model\",\n \"def transform(self, cfg_pipeline):\\n return\",\n \"def __init__(self, **kwargs):\\n FeatureDefinition.__init__(self)\\n\\n nbTypes = self._getTypeNumber(kwargs)\\n \\n print(\\\"BETTER FEATURES\\\")\\n \\n \\n block_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\\n (\\\"xywh\\\", Pipeline([\\n ('selector', NodeTransformerXYWH_v2()),\\n #v1 ('xywh', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('xywh', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n , (\\\"neighbors\\\", Pipeline([\\n ('selector', NodeTransformerNeighbors()),\\n #v1 ('neighbors', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('neighbors', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n , (\\\"1hot\\\", Pipeline([\\n ('1hot', Node1HotFeatures()) #does the 1-hot encoding directly\\n ])\\n )\\n ])\\n grid_line_transformer = GridLine_NodeTransformer_v2()\\n \\n self._node_transformer = TransformerListByType([block_transformer, grid_line_transformer]) \\n \\n edge_BB_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\\n (\\\"1hot\\\", Pipeline([\\n ('1hot', Edge1HotFeatures(PageNumberSimpleSequenciality()))\\n ])\\n )\\n , (\\\"boolean\\\", Pipeline([\\n ('boolean', EdgeBooleanFeatures_v2())\\n ])\\n )\\n , (\\\"numerical\\\", Pipeline([\\n ('selector', EdgeNumericalSelector()),\\n #v1 ('numerical', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('numerical', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n ] )\\n edge_BL_transformer = DU_ABPTableRG.Block2GridLine_EdgeTransformer()\\n edge_LL_transformer = DU_ABPTableRG.GridLine2GridLine_EdgeTransformer()\\n self._edge_transformer = TransformerListByType([edge_BB_transformer,\\n edge_BL_transformer,\\n edge_BL_transformer, # useless but required\\n edge_LL_transformer \\n ])\\n \\n self.tfidfNodeTextVectorizer = None #tdifNodeTextVectorizer\",\n \"def convert_bmm(g, op, block):\\n\\n x = g.get_node(op.input(\\\"X\\\")[0])\\n y = g.get_node(op.input(\\\"Y\\\")[0])\\n y = _op.transpose(y, [0, 2, 1])\\n out = _op.nn.batch_matmul(x, y)\\n g.add_node(op.output(\\\"Out\\\")[0], out)\",\n \"def build_label_transform():\\n\\n return NALabelEncoder()\",\n \"def __init__(self, in_channel_list, out_channel_list, n_hr_blocks, n_blocks, block, bn_momentum=0.01):\\n super(Stage, self).__init__()\\n pre_branches = len(in_channel_list)\\n cur_branches = len(out_channel_list)\\n if cur_branches < pre_branches:\\n raise ValueError('Current number of branch is smaller than previous number of branch')\\n\\n self.hr_blocks = nn.ModuleList()\\n if in_channel_list != out_channel_list[:pre_branches]:\\n use_transform = True\\n else:\\n use_transform = False\\n for i in range(n_hr_blocks):\\n self.hr_blocks.append(HighResolutionBlock(in_channel_list, out_channel_list[:pre_branches], block, n_blocks[0], use_transform))\\n in_channel_list = [chs * block.expansion for chs in out_channel_list[:pre_branches]]\\n self.downsample_layer = DownSampleLayer(in_channel_list, out_channel_list, bn_momentum)\",\n \"def transform(self, data):\",\n \"def __init__(self, args):\\n super(Transformer, self).__init__()\\n\\n self.name = 'Transformer'\\n if len(args.max_relative_pos) != args.nlayers:\\n assert len(args.max_relative_pos) == 1\\n args.max_relative_pos = args.max_relative_pos * args.nlayers\\n\\n self.embedder = Embedder(args)\\n self.encoder = Encoder(args, self.embedder.enc_input_size)\\n self.decoder_loc = nn.Linear(self.embedder.enc_input_size, 1)\\n self.decoder_fix = nn.Linear(self.embedder.enc_input_size, 1)\\n self.nobug_embedding = nn.Parameter(torch.zeros(1))\\n self.criterion = nn.CrossEntropyLoss(reduction='none')\",\n \"def __init__(self,\\n input_dim,\\n dec_seq_len,\\n out_seq_len,\\n d_model=512,\\n nhead=8,\\n num_encoder_layers=6,\\n num_decoder_layers=6,\\n dim_feedforward=2048,\\n dropout=0.1,\\n activation='relu',\\n custom_encoder=None,\\n custom_decoder=None):\\n super(TransformerTS, self).__init__()\\n self.transform = nn.Transformer(\\n d_model=d_model,\\n nhead=nhead,\\n num_encoder_layers=num_encoder_layers,\\n num_decoder_layers=num_decoder_layers,\\n dim_feedforward=dim_feedforward,\\n dropout=dropout,\\n activation=activation,\\n custom_encoder=custom_encoder,\\n custom_decoder=custom_decoder\\n )\\n self.pos = PositionalEncoding(d_model)\\n self.enc_input_fc = nn.Linear(input_dim, d_model)\\n self.dec_input_fc = nn.Linear(input_dim, d_model)\\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\\n self.dec_seq_len = dec_seq_len\",\n \"def create_pipeline(self, train: LAMLDataset) -> LAMLTransformer:\\n raise NotImplementedError\",\n \"def TextEncBlock(L,d,scope=\\\"TextEncBlock\\\"):\\n e = L.shape.as_list()[2]\\n\\n with tf.variable_scope(scope):\\n conv_params = {\\\"filters\\\":2*d,\\\"kernel_size\\\":1,\\\"padding\\\":\\\"same\\\"} \\n with tf.variable_scope(\\\"C_block1\\\"):\\n L1 = tf.nn.relu(conv1d(inputs=L,**conv_params)) # relu(conv)\\n L2 = conv1d(inputs=L1,**conv_params) # conv\\n # hc_dilation_blocks \\n L3_HC1_1 = hc_dilation_block(L2,local_scope=\\\"HC_dilation_block1\\\")\\n L3_HC1_2 = hc_dilation_block(L3_HC1_1,local_scope=\\\"HC_dilation_block2\\\")\\n # highway_conv_block2\\n with tf.variable_scope(\\\"HC_block2\\\"):\\n L4_HC2_1 = highway_activation_conv(L3_HC1_2,kernel_size=3,scope=\\\"HC1\\\")\\n L4_HC2_2 = highway_activation_conv(L4_HC2_1,kernel_size=3,scope=\\\"HC2\\\")\\n # highway_conv_block3\\n with tf.variable_scope(\\\"HC_block3\\\"):\\n L5_HC3_1 = highway_activation_conv(L4_HC2_2,kernel_size=1,scope=\\\"HC1\\\")\\n KV = highway_activation_conv(L5_HC3_1,kernel_size=1,scope=\\\"HC2\\\")\\n\\n return KV[:,:,:d], KV[:,:,d:] # key, value\",\n \"def __init__(self):\\n Transform.__init__(self)\",\n \"def forward(self, x):\\n # encode\\n encode_block1 = self.conv_encode1(x)\\n encode_pool1 = self.conv_maxpool1(encode_block1)\\n encode_block2 = self.conv_encode2(encode_pool1)\\n encode_pool2 = self.conv_maxpool2(encode_block2)\\n encode_block3 = self.conv_encode3(encode_pool2)\\n encode_pool3 = self.conv_maxpool3(encode_block3)\\n # Bottleneck\\n bottleneck1 = self.bottleneck(encode_pool3)\\n # Decode\\n decode_block3 = crop_and_concat(\\n bottleneck1, encode_block3, crop=True)\\n cat_layer2 = self.conv_decode3(decode_block3)\\n decode_block2 = crop_and_concat(\\n cat_layer2, encode_block2, crop=True)\\n cat_layer1 = self.conv_decode2(decode_block2)\\n decode_block1 = crop_and_concat(\\n cat_layer1, encode_block1, crop=True)\\n final_layer = self.final_layer(decode_block1)\\n return final_layer\",\n \"def _encode(self):\\n with tf.variable_scope('encoding'):\\n self.sep_p_encodes, _ = bilstm_layer(self.p_emb, self.p_length, self.hidden_size)\\n tf.get_variable_scope().reuse_variables()\\n self.sep_q_encodes, _ = bilstm_layer(self.q_emb, self.q_length, self.hidden_size)\\n if self.use_dropout:\\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, 1-self.dropout)\\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, 1-self.dropout)\",\n \"def __init__(self, mbart_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\\n self.linear1_weight = mbart_layer.fc1.weight\\n self.linear1_bias = mbart_layer.fc1.bias\\n self.linear2_weight = mbart_layer.fc2.weight\\n self.linear2_bias = mbart_layer.fc2.bias\\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\\n self.norm2_eps = mbart_layer.final_layer_norm.eps\\n self.norm2_weight = mbart_layer.final_layer_norm.weight\\n self.norm2_bias = mbart_layer.final_layer_norm.bias\\n self.num_heads = mbart_layer.self_attn.num_heads\\n self.embed_dim = mbart_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def convert_mish(g, op, block):\\n\\n x = g.get_node(op.input(\\\"X\\\")[0])\\n dtype = infer_type(x).checked_type.dtype\\n exp = _op.exp(x)\\n add = _op.add(exp, _expr.const(1.0, dtype))\\n log = _op.log(add)\\n tanh = _op.tanh(log)\\n out = _op.multiply(x, tanh)\\n g.add_node(op.output(\\\"Out\\\")[0], out)\",\n \"def lnlike(self, block):\\n ##### Extract from datablock\\n self.cosmology = {'Omega_m': block.get_double('cosmological_parameters', 'Omega_m'),\\n 'Omega_l': block.get_double('cosmological_parameters', 'Omega_lambda'),\\n 'Omega_b': block.get_double('cosmological_parameters', 'Omega_b'),\\n 'h': block.get_double('cosmological_parameters', 'hubble')/100,\\n 'ns': block.get_double('cosmological_parameters', 'n_s'),\\n 'w0': block.get_double('cosmological_parameters', 'w'),\\n 'wa': block.get_double('cosmological_parameters', 'wa'),\\n 'sigma8': block.get_double('cosmological_parameters', 'sigma_8')}\\n self.scaling = {\\n # SZ\\n 'Asz': block.get_double('mor_parameters', 'Asz'),\\n 'Bsz': block.get_double('mor_parameters', 'Bsz'),\\n 'Csz': block.get_double('mor_parameters', 'Csz'),\\n 'Dsz': block.get_double('mor_parameters', 'Dsz'),\\n # X-ray\\n 'Ax': block.get_double('mor_parameters', 'Ax'),\\n 'Bx': block.get_double('mor_parameters', 'Bx'),\\n 'Cx': block.get_double('mor_parameters', 'Cx'),\\n 'Dx': block.get_double('mor_parameters', 'Dx'),\\n 'dlnMg_dlnr': block.get_double('mor_parameters', 'dlnMg_dlnr'),\\n # WL\\n 'WLbias': block.get_double('mor_parameters', 'WLbias'),\\n 'WLscatter': block.get_double('mor_parameters', 'WLscatter'),\\n 'HSTbias': block.get_double('mor_parameters', 'HSTbias'),\\n 'HSTscatterLSS': block.get_double('mor_parameters', 'HSTscatterLSS'),\\n 'MegacamBias': block.get_double('mor_parameters', 'MegacamBias'),\\n 'MegacamScatterLSS': block.get_double('mor_parameters', 'MegacamScatterLSS'),\\n # Correlation coefficients\\n 'rhoSZWL': block.get_double('mor_parameters', 'rhoSZWL'),\\n 'rhoWLX': block.get_double('mor_parameters', 'rhoWLX'),\\n 'rhoSZX': block.get_double('mor_parameters', 'rhoSZX'),\\n }\\n # Halo mass function\\n self.HMF = {'M_arr': block.get_double_array_1d('HMF', 'M_arr'),\\n 'z_arr': block.get_double_array_1d('HMF', 'z_arr'),\\n 'dNdlnM': block.get_double_array_nd('HMF', 'dNdlnM')}\\n self.HMF['len_z'] = len(self.HMF['z_arr'])\\n\\n ##### Setup stuff for WL\\n if self.todo['WL']:\\n # Set bias and scatter for Megacam from sim calibration\\n # and nuissance parameters\\n self.WL.set_scaling(self.scaling)\\n # Precompute array of angular diameter distances\\n self.WL.get_dAs(self.cosmology)\\n\\n ##### Populate and check observable covariance matrices\\n self.covmat = {'invertible': True}\\n if not observablecovmat.set_covmats(self.todo, self.scaling, self.covmat):\\n self.covmat['invertible'] = False\\n return -np.inf\\n\\n ##### Set up interpolation for HMF\\n HMF_in = self.HMF['dNdlnM'][1:,:]\\n if np.any(HMF_in==0):\\n HMF_in[np.where(HMF_in==0)] = np.nextafter(0, 1)\\n self.HMF_interp = RectBivariateSpline(np.log(self.HMF['z_arr'][1:]), np.log(self.HMF['M_arr']), np.log(HMF_in))\\n\\n ##### Initialize mass-concentration relation class (for WL)\\n if self.todo['WL']:\\n self.MCrel = Mconversion_concentration.ConcentrationConversion(self.mcType, self.cosmology)\\n\\n ##### Compute interpolation table for M500-M200\\n if self.todo['WL']:\\n z_arr = np.linspace(.1, 2, 20)\\n M500 = np.logspace(np.log10(self.HMF['M_arr'][0]), np.log10(self.HMF['M_arr'][-1]), 20)\\n M200 = np.array([np.array([self.MCrel.MDelta_to_M200(m, 500., z) for m in M500]) for z in z_arr])\\n self.lnM500_to_lnM200 = RectBivariateSpline(z_arr, np.log(M500), np.log(M200))\\n\\n ##### Evaluate the individual likelihoods\\n len_data = len(self.catalog['SPT_ID'])\\n\\n if self.NPROC==0:\\n # Iterate through cluster list\\n likelihoods = np.array([self.clusterlike(i) for i in range(len_data)])\\n else:\\n # Launch a multiprocessing pool and get the likelihoods\\n pool = Pool(processes=self.NPROC)\\n argin = zip([self]*len_data, range(len_data))\\n likelihoods = pool.map(unwrap_self_f, argin)\\n pool.close()\\n\\n # If likelihood computation failed it returned 0\\n if np.count_nonzero(likelihoods)<len_data:\\n return -np.inf\\n\\n lnlike = np.sum(np.log(likelihoods))\\n\\n return lnlike\",\n \"def forward(self, *args):\\r\\n enc_src, _, _ = self.unified_encoder(*args)\\r\\n enc_src = enc_src.view(enc_src.shape[0], -1)\\r\\n y_pred = self.mlp(enc_src)\\r\\n return y_pred\",\n \"def make_mlp_model():\\n return snt.Sequential([\\n snt.nets.MLP([LATENT_SIZE] * NUM_LAYERS, activate_final=True),\\n snt.LayerNorm()\\n ])\",\n \"def fit_transform(self):\\n if self.enc_types == \\\"label\\\":\\n return self._label_encoding()\\n elif self.enc_types == \\\"ohe\\\":\\n return self._one_hot_encoder()\\n elif self.enc_types == \\\"binary\\\":\\n return self._binarization()\\n else:\\n raise Exception(\\\"Encoding type not understood\\\")\",\n \"def __transition_block(self, x, stage, nb_filter, compression=1.0, dropout_rate=None, weight_decay=1E-4):\\n\\n\\t\\teps = 1.1e-5\\n\\t\\tconv_name_base = \\\"conv\\\" + str(stage) + \\\"_blk\\\"\\n\\t\\trelu_name_base = \\\"relu\\\" + str(stage) + \\\"_blk\\\"\\n\\t\\tpool_name_base = \\\"pool\\\" + str(stage) \\n\\n\\t\\tx = BatchNormalization(epsilon=eps, axis=self.concat_axis, name=conv_name_base+\\\"_bn\\\")(x)\\n\\t\\tx = Scale(axis=self.concat_axis, name=conv_name_base+\\\"_scale\\\")(x)\\n\\t\\tx = Activation(\\\"relu\\\", name=relu_name_base)(x)\\n\\t\\tx = Conv2D(int(nb_filter * compression), (1, 1), name=conv_name_base, use_bias=False)(x)\\n\\n\\t\\tif dropout_rate:\\n\\t\\t\\tx = Dropout(dropout_rate)(x)\\n\\n\\t\\tx = AveragePooling2D((2, 2), strides=(2, 2), name=pool_name_base)(x)\\n\\n\\t\\treturn x\",\n \"def __init__(self, in_channel_list, out_channel_list, block, n_blocks,\\n use_transform=False):\\n super(HighResolutionBlock, self).__init__()\\n self.out_chs_list = out_channel_list\\n self.branches = nn.ModuleList()\\n\\n for in_chs, out_chs in zip(in_channel_list, out_channel_list):\\n branch = Branch(in_chs, out_chs, block, n_blocks)\\n self.branches.append(branch)\\n if len(out_channel_list) > 1:\\n self.mul_resolution_layer = MultiResolutionLayer(out_channel_list)\\n else:\\n self.mul_resolution_layer = None\\n pass\",\n \"def forward(self, input):\\n x = self.emb(input)\\n output = self.attention(x)\\n # print(\\\"Idggfghrf \\\", output.size())\\n return output\",\n \"def beam_init(self, encoder_outputs, K):\\n encoder_outputs = TransformerEncoderOutput(\\n repeat_batch(encoder_outputs.output, K),\\n repeat_batch(encoder_outputs.src_mask, K)\\n )\\n return encoder_outputs\",\n \"def __init__(\\n self,\\n num_branches,\\n blocks,\\n num_blocks,\\n num_inchannels,\\n num_channels,\\n num_heads,\\n num_window_sizes,\\n num_mlp_ratios,\\n multi_scale_output=True,\\n drop_path=0.0,\\n ):\\n super(HighResolutionTransformerModule, self).__init__()\\n self._check_branches(\\n num_branches, blocks, num_blocks, num_inchannels, num_channels\\n )\\n\\n self.num_inchannels = num_inchannels\\n self.num_branches = num_branches\\n\\n self.multi_scale_output = multi_scale_output\\n self.branches = self._make_branches(\\n num_branches,\\n blocks,\\n num_blocks,\\n num_channels,\\n num_heads,\\n num_window_sizes,\\n num_mlp_ratios,\\n drop_path,\\n )\\n self.fuse_layers = self._make_fuse_layers()\\n self.relu = nn.ReLU(inplace=True)\\n\\n self.num_heads = num_heads\\n self.num_window_sizes = num_window_sizes\\n self.num_mlp_ratios = num_mlp_ratios\",\n \"def __call__(self, **kwargs):\\n segname = 'block_{}_expand_relu'\\n blocks = [13, 6, 3, 1]\\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\\n backbone_out = self._backbone.get_layer('block_16_project')\\n\\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\\n p4 = self._fpn_block(p5, skips[1].output)\\n p3 = self._fpn_block(p4, skips[2].output)\\n p2 = self._fpn_block(p3, skips[3].output)\\n\\n s5 = self._conv_block(p5, 128)\\n s4 = self._conv_block(p4, 128)\\n s3 = self._conv_block(p3, 128)\\n s2 = self._conv_block(p2, 128)\\n\\n s5 = tf.keras.layers.UpSampling2D(\\n size=(8, 8),\\n interpolation='nearest'\\n )(s5)\\n\\n s4 = tf.keras.layers.UpSampling2D(\\n size=(4, 4),\\n interpolation='nearest'\\n )(s4)\\n\\n s3 = tf.keras.layers.UpSampling2D(\\n size=(2, 2),\\n interpolation='nearest'\\n )(s3)\\n\\n concat = [s5, s4, s3, s2]\\n x = tf.keras.layers.Concatenate()(concat)\\n x = tf.keras.layers.Conv2D(\\n 64,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n x = tf.keras.layers.BatchNormalization()(x)\\n x = tf.keras.layers.Activation('relu')(x)\\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\\n\\n x = tf.keras.layers.Conv2D(\\n 1,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n out = tf.keras.layers.Activation('sigmoid')(x)\\n model = tf.keras.models.Model(\\n inputs=self._backbone.input,\\n outputs=out\\n )\\n\\n return model\",\n \"def __init__(self, distance_function=euclidean_distance, depth=64, pool_last_two=True):\\n super().__init__()\\n self.depth = depth\\n self.encoder = nn.Sequential(\\n conv_block(1, depth, pool_size = 2),\\n conv_block(depth, depth, pool_size = 2),\\n conv_block(depth, depth, pool_size = 2),\\n conv_block(depth, depth, pool_size = 2)\\n )\\n self.distance = distance_function\",\n \"def _define_encoder(self):\\n self.encoder = nn.Sequential(View((-1, 64 * 64 * 3)),\\n nn.Linear(64 * 64 * 3, 5120, bias=False), nn.SELU(),\\n nn.BatchNorm1d(5120),\\n nn.Linear(5120, 2560, bias=False), nn.SELU(),\\n nn.BatchNorm1d(2560),\\n nn.Linear(2560, 512, bias=False), nn.SELU(),\\n nn.BatchNorm1d(512),\\n nn.Linear(512, 128, bias=False), nn.SELU(),\\n nn.BatchNorm1d(128),\\n nn.Linear(128, self.encoding_shape, bias=False), nn.SELU(),\\n )\",\n \"def free_lb_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n\\n (x2, len2), (x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\\n\\n if len(params.ft_lgs) > 0:\\n lang1_id = params.lang2id[params.ft_lgs[0]]\\n langs = x2.clone().fill_(lang1_id)\\n else:\\n lang1_id = params.lang2id['en']\\n langs = x2.clone().fill_(lang1_id)\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n #\\n len1 = x1_mask.sum(dim=1)\\n x1 = x1.transpose(0, 1)\\n img_loc = img_loc.transpose(0, 1)\\n if len(params.ft_lgs) > 0:\\n lang1_id = params.lang2id[params.ft_lgs[0]]\\n langs_img = x1_mask.transpose(0, 1).clone().fill_(lang1_id)\\n else:\\n lang1_id = params.lang2id['en']\\n langs_img = x1_mask.transpose(0, 1).clone().fill_(lang1_id)\\n\\n # cuda\\n\\n x1, len1, img_loc, x2, len2, y, x1_mask, langs, langs_img = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,\\n langs, langs_img)\\n\\n # convert fp16\\n #assign lang ids\\n\\n #free lb\\n if params.free_text:\\n embeds_init, delta = self.deal_freelb_delta(model,x2.transpose(0,1),len2)\\n if params.free_img:\\n image_delta = self.deal_image_freelb_delta(x1)\\n\\n # the main loop\\n dp_masks = None\\n adv_steps=3\\n tr_loss, tb_loss = 0.0, 0.0\\n for astep in range(adv_steps):\\n # (0) forward\\n text_imb = None\\n if params.free_text:\\n delta.requires_grad_()\\n text_imb = delta + embeds_init\\n img_imb = x1\\n if params.free_img:\\n image_delta.requires_grad_()\\n img_imb = x1 + image_delta\\n\\n output = self.get_ic_output(params,model,x2,len2,img_imb,len1,img_loc,langs,langs_img,pred_mask,y,text_imb)\\n loss, dec_out = output[0], output[1]\\n\\n\\n # (1) backward\\n\\n # if params.n_gpu_per_node > 1:\\n # loss = loss.mean() # mean() to average on multi-gpu parallel training\\n # if params.accumulate_gradients > 1:\\n # loss = loss / params.accumulate_gradients\\n\\n loss = loss /(1.0*adv_steps)\\n\\n self.free_optimize(loss)\\n\\n tb_loss+=loss.item()\\n #\\n # self.free_optimize(loss,adv_steps)\\n\\n if astep == adv_steps - 1:\\n # further updates on delta\\n break\\n\\n # update delta\\n if params.free_text:\\n embeds_init, delta = self.update_freelb_delta(model, delta, embeds_init, x2.transpose(0,1))\\n if params.free_img:\\n image_delta = self.update_image_freelb_delta(x1, image_delta)\\n\\n # loss = tr_loss+tb_loss\\n\\n self.stats[('FRLB-IC-%s-%s' % (dataset, input_stream))].append(tb_loss)\\n loss = lambda_coeff * loss\\n\\n # optimize\\n # self.free_optimize(None,True)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def init_encoder(self):\\n\\n vgg = models.vgg16(pretrained=True)\\n\\n blocks = [self.layer_1,\\n self.layer_2,\\n self.layer_3,\\n self.layer_4,\\n self.layer_5]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\\n\\n blocks = [self.layer_11,\\n self.layer_12,\\n self.layer_13,\\n self.layer_14,\\n self.layer_15]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\",\n \"def __init__(self, idim, odim, args, ignore_id=-1):\\n torch.nn.Module.__init__(self)\\n if args.transformer_attn_dropout_rate is None:\\n args.transformer_attn_dropout_rate = args.dropout_rate\\n\\n self.mtlalpha = args.mtlalpha\\n if not hasattr(args, \\\"beta\\\"):\\n self.beta = 0.0\\n else:\\n self.beta = args.beta\\n self.num_steps = args.num_steps\\n self.tau = args.error_temperature\\n self.idx_blank = 0\\n self.masktoken = odim - 1\\n self.sos = odim - 1\\n self.eos = odim - 1\\n self.odim = odim\\n self.ignore_id = ignore_id\\n self.lsm_weight = args.lsm_weight\\n self.reporter = None\\n\\n self.subsample = [1]\\n self.encoder = Encoder(\\n idim=idim,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.eunits,\\n num_blocks=args.elayers,\\n input_layer=args.transformer_input_layer,\\n dropout_rate=args.dropout_rate,\\n death_rate=args.edeath_rate,\\n positional_dropout_rate=args.dropout_rate,\\n attention_dropout_rate=args.transformer_attn_dropout_rate\\n )\\n self.enc_odim = args.adim\\n\\n # Define alignment embedding module.\\n self.embed = torch.nn.Embedding(odim, self.enc_odim)\\n self.decoder = Decoder(\\n odim=odim,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n dropout_rate=args.dropout_rate,\\n death_rate=args.ddeath_rate,\\n positional_dropout_rate=args.dropout_rate,\\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n input_layer=\\\"embed_linear\\\",\\n pos_enc_class=ScaledPositionalEncoding\\n )\\n\\n # self.verbose = args.verbose\\n self.reset_parameters(args)\\n self.adim = args.adim\\n\\n # CTC directly works on top of encoder outputs.\\n self.ctc = CTC(odim, self.enc_odim, args.dropout_rate, ctc_type=args.ctc_type, reduce=True)\\n self.rnnlm = None\\n\\n # Decoder masking params\\n self.decoder_num_mask = args.decoder_num_mask\\n self.decoder_mask_width = args.decoder_mask_width\",\n \"def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\\n latent_dimension):\\n super(VAEEncoder, self).__init__()\\n self.latent_dimension = latent_dimension\\n\\n # Reduce dimension up to second last layer of Encoder\\n self.encode_nn = nn.Sequential(\\n nn.Linear(in_dimension, layer_1d),\\n nn.ReLU(),\\n nn.Linear(layer_1d, layer_2d),\\n nn.ReLU(),\\n nn.Linear(layer_2d, layer_3d),\\n nn.ReLU()\\n )\\n\\n # Latent space mean\\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\\n\\n # Latent space variance\\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)\",\n \"def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\\n latent_dimension):\\n super(VAEEncoder, self).__init__()\\n self.latent_dimension = latent_dimension\\n\\n # Reduce dimension up to second last layer of Encoder\\n self.encode_nn = nn.Sequential(\\n nn.Linear(in_dimension, layer_1d),\\n nn.ReLU(),\\n nn.Linear(layer_1d, layer_2d),\\n nn.ReLU(),\\n nn.Linear(layer_2d, layer_3d),\\n nn.ReLU()\\n )\\n\\n # Latent space mean\\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\\n\\n # Latent space variance\\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)\",\n \"def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\\n ptm=\\\"bert\\\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\\n super(PTMClassifier_PNN, self).__init__()\\n \\n if old_cols is None:\\n old_cols = []\\n \\n self.old_cols = []\\n \\n self.output_size = output_size\\n self.hidden_size = hidden_size # default\\n self.feature_size = feature_size\\n \\n self.fc1 = nn.Linear(self.hidden_size, 100)\\n self.classifier = nn.Linear(100, self.output_size)\\n \\n if len(old_cols) > 0:\\n self.old_fc1s = ListModule()\\n self.base_1 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n self.base_2 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n \\n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_1)\\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_2)\\n \\n for old_col in old_cols:\\n self.old_fc1s.append(\\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\\n \\n # ptm\\n self.ptm = ptm.lower()\\n assert self.ptm in supported_ptm\\n ptm_ = import_from(\\\"transformers\\\", supported_ptm[self.ptm][0] + \\\"Model\\\")\\n \\n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\\n if tokenizer is not None:\\n self.encoder.resize_token_embeddings(len(tokenizer))\\n self.prob_l = prob_l\\n \\n self.require_proto = require_proto\\n \\n # todo : modify net into one network\\n if self.require_proto:\\n # prototype-based classification\\n self.net = nn.Sequential(\\n self.fc1\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.fc1,\\n self.encoder_adaptor,\\n nn.ReLU(),\\n )\\n else:\\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\\n self.net = nn.Sequential(\\n self.classifier\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.classifier\\n )\\n \\n self.reset_parameters()\",\n \"def __init__(\\n self,\\n encoder,\\n num_classes=20,\\n shape_size=16,\\n mask_conv=True,\\n min_confidence=0.2,\\n min_remain=0.5,\\n mask_threshold=0.4,\\n ):\\n super(PaPs, self).__init__()\\n self.encoder = encoder\\n self.shape_size = shape_size\\n self.num_classes = num_classes\\n self.min_scale = 1 / shape_size\\n self.register_buffer(\\\"min_confidence\\\", torch.tensor([min_confidence]))\\n self.min_remain = min_remain\\n self.mask_threshold = mask_threshold\\n self.center_extractor = CenterExtractor()\\n\\n enc_dim = encoder.enc_dim\\n stack_dim = encoder.stack_dim\\n self.heatmap_conv = nn.Sequential(\\n ConvLayer(nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\\n padding_mode=\\\"reflect\\\"),\\n nn.Sigmoid(),\\n )\\n\\n self.saliency_conv = ConvLayer(\\n nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\\n padding_mode=\\\"reflect\\\"\\n )\\n\\n self.shape_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, shape_size ** 2),\\n )\\n\\n self.size_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, stack_dim // 4),\\n nn.BatchNorm1d(stack_dim // 4),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 4, 2),\\n nn.Softplus(),\\n )\\n\\n self.class_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, stack_dim // 4),\\n nn.Linear(stack_dim // 4, num_classes),\\n )\\n\\n if mask_conv:\\n self.mask_cnn = nn.Sequential(\\n nn.Conv2d(in_channels=1, out_channels=16, kernel_size=3, padding=1),\\n nn.GroupNorm(num_channels=16, num_groups=1),\\n nn.ReLU(),\\n nn.Conv2d(in_channels=16, out_channels=16, kernel_size=3, padding=1),\\n nn.ReLU(),\\n nn.Conv2d(in_channels=16, out_channels=1, kernel_size=3, padding=1),\\n )\\n else:\\n self.mask_cnn = None\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\",\n \"def _define_encoder(self):\\n raise NotImplementedError\",\n \"def create_pipelines_lingspam():\\n stop = ('stop', StopWordRemovalTransformer())\\n lemma = ('lemma', LemmatizeTransformer())\\n binz = ('binarizer', CountVectorizer())\\n we = ('document embedding', DocEmbeddingVectorizer())\\n sel = ('fsel', SelectKBest(score_func=mutual_info_classif, k=100))\\n clf = ('cls', BernoulliNB()) # Binary features in the original paper. \\n return Pipeline([binz, sel, clf]), \\\\\\n Pipeline([stop, binz, sel, clf]), \\\\\\n Pipeline([lemma, binz, sel, clf]), \\\\\\n Pipeline([stop, lemma, binz, sel, clf]), \\\\\\n Pipeline([stop, lemma, we, sel, clf])\",\n \"def __init__(self, transformer):\\n self.transformer = transformer\",\n \"def __init__(\\n self,\\n in_channels=128,\\n aux_channels=80,\\n channels=64,\\n out_channels=1,\\n kernel_size=9,\\n dilation=2,\\n bias=True,\\n noise_upsample_scales=[11, 2, 2, 2],\\n noise_upsample_activation=\\\"LeakyReLU\\\",\\n noise_upsample_activation_params={\\\"negative_slope\\\": 0.2},\\n upsample_scales=[2, 2, 2, 2, 2, 2, 2, 2, 1],\\n upsample_mode=\\\"nearest\\\",\\n gated_function=\\\"softmax\\\",\\n use_weight_norm=True,\\n ):\\n super().__init__()\\n\\n self.in_channels = in_channels\\n\\n noise_upsample = []\\n in_chs = in_channels\\n for noise_upsample_scale in noise_upsample_scales:\\n # NOTE(kan-bayashi): How should we design noise upsampling part?\\n noise_upsample += [\\n torch.nn.ConvTranspose1d(\\n in_chs,\\n channels,\\n noise_upsample_scale * 2,\\n stride=noise_upsample_scale,\\n padding=noise_upsample_scale // 2 + noise_upsample_scale % 2,\\n output_padding=noise_upsample_scale % 2,\\n bias=bias,\\n )\\n ]\\n noise_upsample += [\\n getattr(torch.nn, noise_upsample_activation)(\\n **noise_upsample_activation_params\\n )\\n ]\\n in_chs = channels\\n self.noise_upsample = torch.nn.Sequential(*noise_upsample)\\n self.noise_upsample_factor = np.prod(noise_upsample_scales)\\n\\n self.blocks = torch.nn.ModuleList()\\n aux_chs = aux_channels\\n for upsample_scale in upsample_scales:\\n self.blocks += [\\n TADEResBlock(\\n in_channels=channels,\\n aux_channels=aux_chs,\\n kernel_size=kernel_size,\\n dilation=dilation,\\n bias=bias,\\n upsample_factor=upsample_scale,\\n upsample_mode=upsample_mode,\\n gated_function=gated_function,\\n ),\\n ]\\n aux_chs = channels\\n self.upsample_factor = np.prod(upsample_scales)\\n\\n self.output_conv = torch.nn.Sequential(\\n torch.nn.Conv1d(\\n channels,\\n out_channels,\\n kernel_size,\\n 1,\\n bias=bias,\\n padding=(kernel_size - 1) // 2,\\n ),\\n torch.nn.Tanh(),\\n )\\n\\n # apply weight norm\\n if use_weight_norm:\\n self.apply_weight_norm()\\n\\n # reset parameters\\n self.reset_parameters()\",\n \"def __init__(self, c):\\n super(DeepConv, self).__init__(self)\\n\\n self.block_1 = BaseBlock(c, 16)\\n self.block_2 = BaseBlock(16, 32)\\n self.block_3 = BaseBlock(32, 32)\\n\\n self._body = nn.Sequential(self.block_1, self.block_2, self.block_3)\",\n \"def preprocess(item):\\n item = feature_engineering(item)\\n item = encode_features(item)\\n return item\",\n \"def mtf_transformer_lm_moe():\\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\\n hparams.decoder_layers = [\\\"att\\\", \\\"moe\\\"] * 4\\n moe.set_default_moe_hparams(hparams)\\n hparams.mesh_shape = \\\"all:8\\\"\\n hparams.layout = \\\"batch:all;experts:all\\\"\\n return hparams\",\n \"def forward(self, x):\\r\\n x = self.conv1(x)\\r\\n x = self.conv1_BN(x)\\r\\n x = F.relu(x)\\r\\n x = self.conv1_dp(x)\\r\\n x = self.Block2_1(x)\\r\\n x = self.Block2_2(x)\\r\\n x = self.Block3_1(x)\\r\\n x = self.Block3_2(x)\\r\\n x = self.Block3_3(x)\\r\\n x = self.Block3_4(x)\\r\\n x = self.Block4_1(x)\\r\\n x = self.Block4_2(x)\\r\\n x = self.Block4_3(x)\\r\\n x = self.Block4_4(x)\\r\\n x = self.Block5_1(x)\\r\\n x = self.Block5_2(x)\\r\\n x = self.MP(x)\\r\\n x = x.view(x.size(0),-1)\\r\\n x = self.fc(x)\\r\\n \\r\\n return x\",\n \"def __init__(self):\\n super(Generator_Model, self).__init__()\\n\\n # audio encoder\\n self.audio_encoder = nn.Sequential(\\n conv_block(1, 32, kernel_size=1, stride=1, padding=1),\\n conv_block(32, 32, kernel_size=1, stride=1, padding=1, residual=True),\\n \\n conv_block(32, 64, kernel_size=3, stride=(3, 1), padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(64, 128, kernel_size=3, stride=3, padding=1),\\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(128, 256, kernel_size=3, stride=(3, 2), padding=1),\\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(256, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 512, kernel_size=1, stride=1, padding=0)\\n )\\n\\n # face encoder\\n self.face_encoder = nn.ModuleList([\\n nn.Sequential(conv_block(6, 16, kernel_size=7, stride=1, padding=3)),\\n\\n nn.Sequential(conv_block(16, 32, kernel_size=3, stride=2, padding=1),\\n conv_block(32, 32, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(32, 64, kernel_size=3, stride=2, padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(64, 128, kernel_size=3, stride=2, padding=1),\\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(128, 256, kernel_size=3, stride=2, padding=1),\\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(256, 512, kernel_size=3, stride=2, padding=1),\\n conv_block(512, 512, kernel_size=3, stride=1, padding=1, residual=True)),\\n \\n nn.Sequential(conv_block(512, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 512, kernel_size=1, stride=1, padding=0))\\n ])\\n\\n # face decoder\\n self.face_decoder = nn.ModuleList([\\n nn.Sequential(conv_block(512, 512, kernel_size=1, stride=1, padding=0),),\\n\\n nn.Sequential(conv_t_block(512, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 1024, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(1024, 512, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(512, 768, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(768, 384, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(384, 512, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(512, 256, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(256, 320, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(320, 128, kernel_size=3, stride=2, padding=1, output_padding=1), \\n conv_block(128, 160, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(160, 64, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True))\\n ])\\n\\n # output blocks for face prediction\\n self.pred_output = nn.Sequential(conv_block(64, 32, kernel_size=3, stride=1, padding=1),\\n conv_block(32, 3, kernel_size=1, stride=1, padding=0),\\n nn.Sigmoid())\\n\\n # optimizer\\n self.optimizer = torch.optim.Adam(self.parameters(), lr=0.001, betas=(0.5, 0.999))\",\n \"def forward(self, x):\\n if self.training:\\n x = self.input_pert(x)\\n x = self.encoder(x)\\n x = self.decoder(x)\\n return x\",\n \"def transform(self, x):\",\n \"def prepare_emb(self):\\n with tf.variable_scope(\\\"PrepEmb\\\", reuse=tf.AUTO_REUSE):\\n self.src_ten = tf.cast(tf.convert_to_tensor(self.src_ten), tf.float32)\\n self.tgt_ten = tf.cast(tf.convert_to_tensor(self.tgt_ten), tf.float32)\\n # Mapping\\n self.src_ten = tf.matmul(self.src_ten, self.W)\\n # Normalization\\n self.src_ten = tf.nn.l2_normalize(self.src_ten, axis=1)\\n self.tgt_ten = tf.nn.l2_normalize(self.tgt_ten, axis=1)\",\n \"def create_bem_flow(name='bem', out_format='stl'):\\n \\\"\\\"\\\"\\n Initialize the workflow\\n \\\"\\\"\\\"\\n\\n bemflow = pe.Workflow(name=name)\\n \\\"\\\"\\\"\\n Define the inputs to the workflow.\\n \\\"\\\"\\\"\\n\\n inputnode = pe.Node(\\n niu.IdentityInterface(fields=['subject_id', 'subjects_dir']),\\n name='inputspec')\\n \\\"\\\"\\\"\\n Define all the nodes of the workflow:\\n\\n fssource: used to retrieve aseg.mgz\\n mri_convert : converts aseg.mgz to aseg.nii\\n tessellate : tessellates regions in aseg.mgz\\n surfconvert : converts regions to stereolithographic (.stl) format\\n\\n \\\"\\\"\\\"\\n\\n watershed_bem = pe.Node(interface=mne.WatershedBEM(), name='WatershedBEM')\\n\\n surfconvert = pe.MapNode(\\n fs.MRIsConvert(out_datatype=out_format),\\n iterfield=['in_file'],\\n name='surfconvert')\\n \\\"\\\"\\\"\\n Connect the nodes\\n \\\"\\\"\\\"\\n\\n bemflow.connect([\\n (inputnode, watershed_bem, [('subject_id', 'subject_id'),\\n ('subjects_dir', 'subjects_dir')]),\\n (watershed_bem, surfconvert, [('mesh_files', 'in_file')]),\\n ])\\n \\\"\\\"\\\"\\n Setup an outputnode that defines relevant inputs of the workflow.\\n \\\"\\\"\\\"\\n\\n outputnode = pe.Node(\\n niu.IdentityInterface(fields=[\\\"meshes\\\"]), name=\\\"outputspec\\\")\\n bemflow.connect([\\n (surfconvert, outputnode, [(\\\"converted\\\", \\\"meshes\\\")]),\\n ])\\n return bemflow\",\n \"def __init__(self, inputs, outputs):\\n super(ELM, self).__init__(inputs, outputs)\",\n \"def _encode(self):\\n with tf.variable_scope('passage_encoding'):\\n self.sep_p_encodes, _ = rnn('bi-lstm', self.p_emb, self.p_length, self.hidden_size)\\n with tf.variable_scope('question_encoding'):\\n self.sep_q_encodes, _ = rnn('bi-lstm', self.q_emb, self.q_length, self.hidden_size)\\n if self.use_dropout:\\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, self.dropout_keep_prob)\\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, self.dropout_keep_prob)\",\n \"def mtf_transformer2_base():\\n hparams = common_hparams.basic_params1()\\n\\n hparams.add_hparam(\\\"d_model\\\", 1024)\\n hparams.batch_size = 4\\n hparams.max_length = 1024\\n hparams.label_smoothing = 0.0\\n # a small positive value - this seems important for stability when training\\n # with bfloat16 activations.\\n hparams.add_hparam(\\\"z_loss\\\", 1e-4)\\n\\n # hparams applying to both encoder and decoder layer stacks.\\n hparams.add_hparam(\\\"d_ff\\\", 2048)\\n hparams.add_hparam(\\\"d_kv\\\", 128)\\n hparams.add_hparam(\\\"attention_dropout\\\", 0.0)\\n hparams.add_hparam(\\\"relu_dropout\\\", 0.0)\\n hparams.del_hparam(\\\"num_heads\\\")\\n hparams.del_hparam(\\\"num_hidden_layers\\\")\\n hparams.layer_prepostprocess_dropout = 0.0\\n hparams.add_hparam(\\\"extra_logit\\\", False)\\n # number of experts for moe_1d\\n hparams.moe_num_experts = 32\\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\\n hparams.add_hparam(\\\"moe_expert_x\\\", 8)\\n hparams.add_hparam(\\\"moe_expert_y\\\", 4)\\n hparams.add_hparam(\\\"moe_hidden_size\\\", 32768)\\n\\n # round up vocab sizes to be a multiple of this value\\n hparams.vocab_divisor = 128\\n\\n hparams.optimizer = \\\"Adafactor\\\"\\n hparams.learning_rate_schedule = \\\"rsqrt_decay*linear_decay\\\"\\n hparams.learning_rate_warmup_steps = 10000\\n hparams.add_hparam(\\\"master_dtype\\\", \\\"bfloat16\\\")\\n hparams.add_hparam(\\\"slice_dtype\\\", \\\"float32\\\")\\n hparams.activation_dtype = \\\"bfloat16\\\"\\n\\n # 8-way model-parallelism\\n hparams.add_hparam(\\\"mesh_shape\\\", \\\"model:8\\\")\\n hparams.add_hparam(\\\"layout\\\", \\\"batch:batch;vocab:model;d_ff:model;heads:model\\\")\\n\\n # If nonzero, we split the batch across two tensor-dimensions named\\n # \\\"outer_batch\\\" and \\\"inner_batch\\\", allowing for splitting across two mesh\\n # dimensions. This is necessary for hierarchical mixture of experts.\\n # The two tensor dimensions have sizes hparams.outer_batch_size and\\n # hparams.batch_size // hparams.outer_batch_size.\\n hparams.add_hparam(\\\"outer_batch_size\\\", 0)\\n\\n hparams.shared_embedding_and_softmax_weights = False\\n # length for training or decoding - defaults to max_length\\n hparams.add_hparam(\\\"length\\\", 0)\\n\\n # These parameters make Transformer model compatible with mtf\\n # Do not override these.\\n hparams.no_data_parallelism = True\\n hparams.use_fixed_batch_size = True\\n hparams.add_hparam(\\\"mtf_mode\\\", True)\\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\\n hparams.bottom = {\\n \\\"inputs\\\": modalities.identity_bottom,\\n \\\"targets\\\": modalities.identity_bottom,\\n }\\n hparams.top = {\\n \\\"targets\\\": modalities.identity_top,\\n }\\n hparams.add_hparam(\\\"beam_size\\\", 1)\\n\\n # If this is True, then in a packed dataset (where exaples are concatenated\\n # to form longer examples) we use the global position (within the concatenated\\n # sequence) to compute the positional embedding, instead of the position\\n # within the individual sequence. This is counterintuitive, but for some\\n # reason, it keeps the model from diverging.\\n hparams.add_hparam(\\\"use_global_position_in_packed_sequence\\\", True)\\n\\n return hparams\",\n \"def build(self, lang, linearInMLP=False):\\n inputLayers, concLayers = [], []\\n inputToken = Input((3 + configuration['embedding']['useB1'] + configuration['embedding']['useB-1'],))\\n inputLayers.append(inputToken)\\n tokenEmb = Embedding(len(self.vocabulary.tokenIndices), configuration['mlp']['tokenEmb'],\\n trainable=configuration['mlp']['trainable'],\\n weights=self.getWeightMatrix(self.vocabulary.tokenIndices, lang))(inputToken)\\n tokenFlatten = Flatten()(tokenEmb)\\n concLayers.append(tokenFlatten)\\n inputPos = Input((3 + configuration['embedding']['useB1'] + configuration['embedding']['useB-1'],))\\n inputLayers.append(inputPos)\\n posEmb = Embedding(len(self.vocabulary.posIndices), configuration['mlp']['posEmb'],\\n trainable=configuration['mlp']['trainable'])(inputPos)\\n posFlatten = Flatten()(posEmb)\\n concLayers.append(posFlatten)\\n if linearInMLP:\\n linearPredInput = Input(shape=(8,))\\n inputLayers.append(linearPredInput)\\n concLayers.append(linearPredInput)\\n\\n conc = keras.layers.concatenate(concLayers) if len(concLayers) > 1 else concLayers[0]\\n dense1Layer = Dense(configuration['mlp']['dense1UnitNumber'],\\n activation=configuration['nn']['dense1Activation'])(conc)\\n lastLayer = Dropout(configuration['mlp']['dense1Dropout'])(dense1Layer)\\n if configuration['mlp2']['dense2']:\\n dense2Layer = Dense(configuration['mlp2']['dense2UnitNumber'],\\n activation=configuration['mlp2']['dense2Activation'])(lastLayer)\\n lastLayer = Dropout(configuration['mlp2']['dense2Dropout'])(dense2Layer)\\n softmaxLayer = Dense(8 if enableCategorization else 4, activation='softmax')(lastLayer)\\n return inputLayers, softmaxLayer\",\n \"def init_encoder(self):\\n\\n vgg = models.vgg16(pretrained=True)\\n\\n blocks = [self.layer_1,\\n self.layer_2,\\n self.layer_3,\\n self.layer_4,\\n self.layer_5]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\",\n \"def mobilenetv2_block(x_in, ch_in, ch_out, expansion_factor, regularizer, strides=1):\\n x = conv_2d_with_bn_relu(ch_out=expansion_factor*ch_in, kernel_size=1, regularizer=regularizer)(x_in)\\n x = depthwise_conv_2d_with_bn_relu(strides, regularizer=regularizer)(x)\\n x = BatchNormalization(beta_regularizer=regularizer, gamma_regularizer=regularizer)(\\n Conv2D(ch_out, kernel_size=1, strides=1, padding=\\\"same\\\", use_bias=False, kernel_regularizer=regularizer)(x))\\n return x if strides == 2 or ch_in != ch_out else Add()([x, x_in])\",\n \"def forward(self, *args): # noqa: R0914\\r\\n encoder_out, (hn, cn) = self.unified_encoder(*args)\\r\\n device = hn.device\\r\\n non_sequential_cont_decoded = self.mlp_non_seq_cont(hn)\\r\\n non_sequential_cat_decoded = []\\r\\n for mlp_non_seq_cat in self.mlp_non_seq_cat_list:\\r\\n non_sequential_cat_decoded.append(mlp_non_seq_cat(hn))\\r\\n\\r\\n hn = torch.unsqueeze(hn, 0)\\r\\n cn = torch.unsqueeze(cn, 0)\\r\\n # decoded is the output prediction of timestep i-1 of the decoder\\r\\n decoded = torch.zeros(encoder_out.shape[0], int(\\r\\n self.unified_encoder.seq_cont_count + self.unified_encoder.no_of_embs_seq), device=device)\\r\\n seq_cont_decoded = torch.Tensor(device=device)\\r\\n seq_cat_decoded = []\\r\\n for _ in range(self.unified_encoder.seq_cat_count):\\r\\n seq_cat_decoded.append(torch.Tensor(device=device))\\r\\n\\r\\n for _ in range(encoder_out.shape[1]):\\r\\n decoded, (hn, cn), out_cont, out_cat = self.decoder(decoded, (hn, cn))\\r\\n # Predict all categorical columns\\r\\n out_cat_onehot = []\\r\\n if self.unified_encoder.seq_cat_count != 0:\\r\\n for idx, out in enumerate(out_cat):\\r\\n out_cat_onehot.append(torch.argmax(out, dim=1).unsqueeze(-1))\\r\\n seq_cat_decoded[idx] = torch.cat(\\r\\n [seq_cat_decoded[idx], out.view(out.shape[0], 1, -1)], dim=1)\\r\\n out_cat_onehot = torch.cat(out_cat_onehot, -1)\\r\\n out_cat_embedding = self.unified_encoder.seq_emb_layers(out_cat_onehot)\\r\\n decoded = torch.cat([out_cat_embedding, out_cont], dim=-1)\\r\\n else:\\r\\n decoded = out_cont\\r\\n seq_cont_decoded = torch.cat(\\r\\n [seq_cont_decoded, out_cont.view(out_cont.shape[0], 1, -1)], dim=1)\\r\\n\\r\\n return non_sequential_cont_decoded, non_sequential_cat_decoded, seq_cont_decoded, seq_cat_decoded\",\n \"def slide_step(self, dataset='slide', input_stream='img', lambda_coeff=1):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n\\n (x2, len2), (x1, x1_mask, img_loc, img_id), labels= self.get_batch('slide2img', dataset, input_stream)\\n # convert fp16\\n #assign lang ids\\n # follow uvp\\n\\n img_len = x1_mask.sum(dim=1)\\n x_img = x1.transpose(0, 1)\\n img_loc = img_loc.transpose(0, 1)\\n\\n x2, len2, x_img, img_loc, img_len = to_cuda(x2, len2, x_img, img_loc, img_len)\\n\\n encoder_outputs = model('jointfwd', x=x2, lengths=len2, x_img=x_img, lengths_img=img_len, causal=False,\\n langs=None,\\n image_loc=img_loc, refine_image=params.refine_image)\\n\\n encoder_outputs = encoder_outputs.transpose(0, 1)\\n\\n relation_scores = model('predict', tensor=encoder_outputs, is_relation=True)\\n\\n # loss\\n # the last word not apply prediction logic\\n s2 = torch.from_numpy(np.array(labels, dtype='float32')).view(-1)\\n loss = F.binary_cross_entropy_with_logits(relation_scores.view(-1).cpu(),\\n s2)\\n\\n\\n self.stats[('SLIDE-%s' % (input_stream))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def fit_transform(self, X, y, blk_start_loc=None):\\n # normalize y\\n self.scale = y.std()\\n y = y / self.scale\\n M, N = X.shape\\n # automatically set block partition\\n if blk_start_loc is None:\\n blkLen = int(N / 16.0)\\n blk_start_loc = np.arange(0, N, blkLen)\\n self.blk_len_list, self.is_equal_block = block_parse(blk_start_loc, N)\\n # init variables\\n self.init(X, y, blk_start_loc)\\n # bootstrap ADD one basis\\n ml, A, theta = self.logobj_mapping()\\n idx = ml.argmin(0)\\n Sig, w, Xu = self.bootup(A, idx)\\n # loops\\n ML = np.zeros(self.max_iters)\\n ML[0] = ml[idx]\\n for count in range(1, self.max_iters):\\n ml, A, theta = self.logobj_mapping()\\n idx = ml.argmin(0)\\n\\n # check convergence now\\n ML[count] = ml[idx]\\n if ML[count] >= 0:\\n break\\n if count > 1:\\n ml_ratio = np.abs(ML[count] - ML[count - 1]) / np.abs(ML[count] - ML[0])\\n if ml_ratio < self.epsilon:\\n break\\n\\n # operation on basis\\n if self.index[idx] is True:\\n if theta[idx] > self.prune_gamma:\\n proc = self.estimate\\n else:\\n proc = self.delete\\n else:\\n proc = self.add\\n # process Sig, w, Xu\\n Sig, w, Xu = proc(Sig, w, Xu, A, idx)\\n\\n # exit\\n self.count = count\\n return self.w_format(w)\",\n \"def __init__(self, params, model, name=\\\"ds2_encoder\\\", mode='train'):\\n super(DeepSpeech2Encoder, self).__init__(params, model, name, mode)\",\n \"def test_transform_prep_proba():\\n args = get_layer('transform', 'manual', 'temporal', True, True, window=2, step_size=3)\\n run_layer(*args)\",\n \"def transform(self, inputs: list, stage: str) -> datapack.DataPack:\",\n \"def __build_mol_to_latent_model(self):\\n\\n # Input tensor (MANDATORY)\\n encoder_inputs = Input(shape=self.input_shape, name=\\\"Encoder_Inputs\\\")\\n\\n x = encoder_inputs\\n\\n # The two encoder layers, number of cells are halved as Bidirectional\\n encoder = Bidirectional(\\n LSTM(\\n self.lstm_dim // 2,\\n return_sequences=True,\\n return_state=True, # Return the states at end of the batch\\n name=\\\"Encoder_LSTM_1\\\",\\n )\\n )\\n\\n x, state_h, state_c, state_h_reverse, state_c_reverse = encoder(x)\\n\\n if self.bn:\\n x = BatchNormalization(momentum=self.bn_momentum, name=\\\"BN_1\\\")(x)\\n\\n encoder2 = Bidirectional(\\n LSTM(\\n self.lstm_dim // 2,\\n return_state=True, # Return the states at end of the batch\\n name=\\\"Encoder_LSTM_2\\\",\\n )\\n )\\n\\n _, state_h2, state_c2, state_h2_reverse, state_c2_reverse = encoder2(x)\\n\\n # Concatenate all states of the forward and the backward LSTM layers\\n states = Concatenate(axis=-1, name=\\\"Concatenate_1\\\")(\\n [\\n state_h,\\n state_c,\\n state_h2,\\n state_c2,\\n state_h_reverse,\\n state_c_reverse,\\n state_h2_reverse,\\n state_c2_reverse,\\n ]\\n )\\n\\n if self.bn:\\n states = BatchNormalization(momentum=self.bn_momentum, name=\\\"BN_2\\\")(states)\\n\\n # A non-linear recombination\\n neck_relu = Dense(\\n self.codelayer_dim, activation=self.h_activation, name=\\\"Codelayer_Relu\\\"\\n )\\n neck_outputs = neck_relu(states)\\n\\n if self.bn:\\n neck_outputs = BatchNormalization(\\n momentum=self.bn_momentum, name=\\\"BN_Codelayer\\\"\\n )(neck_outputs)\\n\\n # Add Gaussian noise to \\\"spread\\\" the distribution of the latent variables during training\\n neck_outputs = GaussianNoise(self.noise_std, name=\\\"Gaussian_Noise\\\")(\\n neck_outputs\\n )\\n\\n # Define the model\\n self.__mol_to_latent_model = Model(encoder_inputs, neck_outputs)\\n\\n # Name it!\\n self.mol_to_latent_model.name = \\\"mol_to_latent_model\\\"\",\n \"def TransformerEncoderLayer(self, name, stride=1, first_n=None,\\n ff_hidden_dim=None, num_heads=None,\\n use_moe=False):\\n p = self.params\\n if ff_hidden_dim is None:\\n ff_hidden_dim = p.ff_hidden_dim\\n if num_heads is None:\\n num_heads = p.num_heads\\n if use_moe:\\n ffw_p = self.MoE('moe', ff_hidden_dim=ff_hidden_dim)\\n else:\\n ffw_p = self.Feedforward('ff', ff_hidden_dim=ff_hidden_dim)\\n return self._Seq(name, self._Seq(\\n 'block',\\n self._StridedAttention('self_atten', stride=stride,\\n first_n=first_n, num_heads=num_heads),\\n ffw_p))\",\n \"def DontuseThis():\\n BCM_outputs = ['phi','rho','theta',\\n 'r_probabilityMaps','l_probabilityMaps',\\n 'models']\\n BCM_Models = pe.Node(interface=nio.DataGrabber(input_names=['structures'],\\n outfields=BCM_outputs),\\n name='10_BCM_Models')\\n BCM_Models.inputs.base_directory = atlas_fname_wpath\\n BCM_Models.inputs.template_args['phi'] = [['spatialImages','phi','nii.gz']]\\n BCM_Models.inputs.template_args['rho'] = [['spatialImages','rho','nii.gz']]\\n BCM_Models.inputs.template_args['theta'] = [['spatialImages','theta','nii.gz']]\\n BCM_Models.inputs.template_args['r_probabilityMaps'] = [['structures']]\\n BCM_Models.inputs.template_args['l_probabilityMaps'] = [['structures']]\\n BCM_Models.inputs.template_args['models'] = [['structures']]\\n\\n BRAINSCut_structures = ['caudate','thalamus','putamen','hippocampus']\\n #BRAINSCut_structures = ['caudate','thalamus']\\n BCM_Models.iterables = ( 'structures', BRAINSCut_structures )\\n BCM_Models.inputs.template = '%s/%s.%s'\\n BCM_Models.inputs.field_template = dict(\\n r_probabilityMaps='probabilityMaps/r_%s_ProbabilityMap.nii.gz',\\n l_probabilityMaps='probabilityMaps/l_%s_ProbabilityMap.nii.gz',\\n models='modelFiles/%sModel*',\\n )\\n\\n \\\"\\\"\\\"\\n The xml creation and BRAINSCut need to be their own mini-pipeline that gets\\n executed once for each of the structures in BRAINSCut_structures. This can be\\n accomplished with a map node and a new pipeline.\\n \\\"\\\"\\\"\\n \\\"\\\"\\\"\\n Create xml file for BRAINSCut\\n \\\"\\\"\\\"\\n\\n\\n BFitAtlasToSubject = pe.Node(interface=BRAINSFit(),name=\\\"BFitAtlasToSubject\\\")\\n BFitAtlasToSubject.inputs.costMetric=\\\"MMI\\\"\\n BFitAtlasToSubject.inputs.maskProcessingMode=\\\"ROI\\\"\\n BFitAtlasToSubject.inputs.numberOfSamples=100000\\n BFitAtlasToSubject.inputs.numberOfIterations=[1500,1500]\\n BFitAtlasToSubject.inputs.numberOfHistogramBins=50\\n BFitAtlasToSubject.inputs.maximumStepLength=0.2\\n BFitAtlasToSubject.inputs.minimumStepLength=[0.005,0.005]\\n BFitAtlasToSubject.inputs.transformType= [\\\"Affine\\\",\\\"BSpline\\\"]\\n BFitAtlasToSubject.inputs.maxBSplineDisplacement= 7\\n BFitAtlasToSubject.inputs.maskInferiorCutOffFromCenter=65\\n BFitAtlasToSubject.inputs.splineGridSize=[28,20,24]\\n BFitAtlasToSubject.inputs.outputVolume=\\\"Trial_Initializer_Output.nii.gz\\\"\\n BFitAtlasToSubject.inputs.outputTransform=\\\"Trial_Initializer_Output.mat\\\"\\n cutWF.connect(SplitAvgBABC,'avgBABCT1',BFitAtlasToSubject,'fixedVolume')\\n cutWF.connect(BABC,'outputLabels',BFitAtlasToSubject,'fixedBinaryVolume')\\n cutWF.connect(BAtlas,'template_t1',BFitAtlasToSubject,'movingVolume')\\n cutWF.connect(BAtlas,'template_brain',BFitAtlasToSubject,'movingBinaryVolume')\\n cutWF.connect(BLI,'outputTransformFilename',BFitAtlasToSubject,'initialTransform')\\n\\n CreateBRAINSCutXML = pe.Node(Function(input_names=['rho','phi','theta',\\n 'model',\\n 'r_probabilityMap',\\n 'l_probabilityMap',\\n 'atlasT1','atlasBrain',\\n 'subjT1','subjT2',\\n 'subjT1GAD','subjT2GAD',\\n 'subjSGGAD','subjBrain',\\n 'atlasToSubj','output_dir'],\\n output_names=['xml_filename','rl_structure_filename_list'],\\n function = create_BRAINSCut_XML),\\n overwrite = True,\\n name=\\\"CreateBRAINSCutXML\\\")\\n\\n ## HACK Makde better directory\\n CreateBRAINSCutXML.inputs.output_dir = \\\".\\\" #os.path.join(cutWF.base_dir, \\\"BRAINSCut_output\\\")\\n cutWF.connect(BCM_Models,'models',CreateBRAINSCutXML,'model')\\n cutWF.connect(BCM_Models,'rho',CreateBRAINSCutXML,'rho')\\n cutWF.connect(BCM_Models,'phi',CreateBRAINSCutXML,'phi')\\n cutWF.connect(BCM_Models,'theta',CreateBRAINSCutXML,'theta')\\n cutWF.connect(BCM_Models,'r_probabilityMaps',CreateBRAINSCutXML,'r_probabilityMap')\\n cutWF.connect(BCM_Models,'l_probabilityMaps',CreateBRAINSCutXML,'l_probabilityMap')\\n cutWF.connect(BAtlas,'template_t1',CreateBRAINSCutXML,'atlasT1')\\n cutWF.connect(BAtlas,'template_brain',CreateBRAINSCutXML,'atlasBrain')\\n cutWF.connect(SplitAvgBABC,'avgBABCT1',CreateBRAINSCutXML,'subjT1')\\n cutWF.connect(SplitAvgBABC,'avgBABCT2',CreateBRAINSCutXML,'subjT2')\\n cutWF.connect(GADT1,'outputVolume',CreateBRAINSCutXML,'subjT1GAD')\\n cutWF.connect(GADT2,'outputVolume',CreateBRAINSCutXML,'subjT2GAD')\\n cutWF.connect(SGI,'outputFileName',CreateBRAINSCutXML,'subjSGGAD')\\n cutWF.connect(BABC,'outputLabels',CreateBRAINSCutXML,'subjBrain')\\n cutWF.connect(BFitAtlasToSubject,'outputTransform',CreateBRAINSCutXML,'atlasToSubj')\\n #CreateBRAINSCutXML.inputs.atlasToSubj = \\\"INTERNAL_REGISTER.mat\\\"\\n #cutWF.connect(BABC,'atlasToSubjectTransform',CreateBRAINSCutXML,'atlasToSubj')\\n\\n \\\"\\\"\\\"\\n ResampleNACLabels\\n \\\"\\\"\\\"\\n ResampleAtlasNACLabels=pe.Node(interface=BRAINSResample(),name=\\\"ResampleAtlasNACLabels\\\")\\n ResampleAtlasNACLabels.inputs.interpolationMode = \\\"NearestNeighbor\\\"\\n ResampleAtlasNACLabels.inputs.outputVolume = \\\"atlasToSubjectNACLabels.nii.gz\\\"\\n\\n cutWF.connect(cutWF,'OutputSpec.atlasToSubjectTransform',ResampleAtlasNACLabels,'warpTransform')\\n cutWF.connect(cutWF,'OutputSpec.t1_corrected',ResampleAtlasNACLabels,'referenceVolume')\\n cutWF.connect(BAtlas,'template_nac_lables',ResampleAtlasNACLabels,'inputVolume')\\n\\n \\\"\\\"\\\"\\n BRAINSMush\\n \\\"\\\"\\\"\\n BMUSH=pe.Node(interface=BRAINSMush(),name=\\\"BMUSH\\\")\\n BMUSH.inputs.outputVolume = \\\"MushImage.nii.gz\\\"\\n BMUSH.inputs.outputMask = \\\"MushMask.nii.gz\\\"\\n BMUSH.inputs.lowerThresholdFactor = 1.2\\n BMUSH.inputs.upperThresholdFactor = 0.55\\n\\n cutWF.connect(myLocalTCWF,'OutputSpec.t1_corrected',BMUSH,'inputFirstVolume')\\n cutWF.connect(myLocalTCWF,'OutputSpec.t2_corrected',BMUSH,'inputSecondVolume')\\n cutWF.connect(myLocalTCWF,'OutputSpec.outputLabels',BMUSH,'inputMaskVolume')\\n\\n \\\"\\\"\\\"\\n BRAINSROIAuto\\n \\\"\\\"\\\"\\n BROI = pe.Node(interface=BRAINSROIAuto(), name=\\\"BRAINSROIAuto\\\")\\n BROI.inputs.closingSize=12\\n BROI.inputs.otsuPercentileThreshold=0.01\\n BROI.inputs.thresholdCorrectionFactor=1.0\\n BROI.inputs.outputROIMaskVolume = \\\"temproiAuto_t1_ACPC_corrected_BRAINSABC.nii.gz\\\"\\n cutWF.connect(myLocalTCWF,'OutputSpec.t1_corrected',BROI,'inputVolume')\\n\\n \\\"\\\"\\\"\\n Split the implicit outputs of BABCext\\n \\\"\\\"\\\"\\n SplitAvgBABC = pe.Node(Function(input_names=['in_files','T1_count'], output_names=['avgBABCT1','avgBABCT2'],\\n function = get_first_T1_and_T2), run_without_submitting=True, name=\\\"99_SplitAvgBABC\\\")\\n SplitAvgBABC.inputs.T1_count = 1 ## There is only 1 average T1 image.\\n\\n cutWF.connect(myLocalTCWF,'OutputSpec.outputAverageImages',SplitAvgBABC,'in_files')\\n\\n\\n\\n def printFullPath(outFileFullPath):\\n print(\\\"=\\\"*80)\\n print(\\\"=\\\"*80)\\n print(\\\"=\\\"*80)\\n print(\\\"=\\\"*80)\\n print(\\\"{0}\\\".format(outFileFullPath))\\n return outFileFullPath\\n printOutImage = pe.Node( Function(function=printFullPath, input_names = ['outFileFullPath'], output_names = ['genoutFileFullPath']), run_without_submitting=True, name=\\\"99_printOutImage\\\")\\n cutWF.connect( GADT2, 'outputVolume', printOutImage, 'outFileFullPath' )\",\n \"def preprocessing_transform(self, x):\\n if isinstance(self.column_text, int) and self.column_text not in x.columns:\\n col = self.column_text\\n else:\\n col = list(x.columns).index(self.column_text)\\n\\n ct = x.shape[0]\\n # INPUTS\\n if self.method_embedding.lower() in ['roberta', \\\"camembert\\\", \\\"xlm-roberta\\\"]:\\n ids = np.ones((ct, self.maxlen), dtype='int32')\\n else:\\n ids = np.zeros((ct, self.maxlen), dtype='int32')\\n att = np.zeros((ct, self.maxlen), dtype='int32')\\n tok = np.zeros((ct, self.maxlen), dtype='int32')\\n\\n for k in range(ct):\\n text = \\\" \\\" + \\\" \\\".join(x.iloc[k, col].split())\\n\\n if self.method_embedding == 'RoBERTa':\\n enc = self.tokenizer.encode(text)\\n else:\\n enc = self.tokenizer.encode(text, max_length=self.maxlen, truncation=True)\\n\\n # CREATE BERT INPUTS\\n if self.method_embedding == 'RoBERTa':\\n ids[k, :len(enc.ids)] = enc.ids[:self.maxlen]\\n att[k, :len(enc.ids)] = 1\\n else:\\n ids[k, :len(enc)] = enc\\n att[k, :len(enc)] = 1\\n\\n x_preprocessed = [ids, att, tok]\\n if self.dimension_embedding == 'word_embedding':\\n return x_preprocessed\\n else:\\n model_extractor = self.model_extract_document_embedding()\\n document_embedding = model_extractor.predict(x_preprocessed)\\n return document_embedding\",\n \"def __init__(self, vocab_size, embedding_size, output_size, lat_dim):\\n super(Encoder, self).__init__()\\n\\n self.vocab_size = vocab_size\\n self.embedding = nn.Embedding(vocab_size, embedding_size)\\n self.lstm = nn.LSTM(embedding_size, output_size)\\n \\n self.hidden2mean = nn.Linear(output_size, lat_dim)\\n self.hidden2logv = nn.Linear(output_size, lat_dim)\\n \\n self.cell2mean = nn.Linear(output_size, lat_dim)\\n self.cell2logv = nn.Linear(output_size, lat_dim)\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n TransformerDecoderBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n multi_channel_cross_attention=self.multi_channel_cross_attention,\\n name=(\\\"layer_%d\\\" % i)))\\n super(TransformerDecoder, self).build(unused_input_shapes)\",\n \"def __build_ml_pipeline(self, clf: MultiOutputClassifier) -> Pipeline:\\r\\n pipeline = Pipeline([\\r\\n ('features', FeatureUnion(\\r\\n [('text', Pipeline(\\r\\n [('text_field_extractor', \\r\\n basic_utils.TextFieldExtractor('message')), \\r\\n #('vect', CountVectorizer(tokenizer=clean_text)),\\r\\n #('tfidf', TfidfTransformer())\\r\\n ('tfidf', TfidfVectorizer(tokenizer=basic_utils.tokenize, \\r\\n min_df=.0025, max_df=0.25, \\r\\n ngram_range=(1,2)))\\r\\n ])),\\r\\n \\r\\n ('numerics', FeatureUnion(\\r\\n [('text_len', \\r\\n Pipeline([('text_len_extractor', \\r\\n basic_utils.NumericFieldExtractor('text_len')), \\r\\n ('text_len_scaler', StandardScaler())\\r\\n ])),\\r\\n ('punt_perc', \\r\\n Pipeline([('punt_perc_extractor', \\r\\n basic_utils.NumericFieldExtractor('punt_perc')), \\r\\n ('punt_perc_scaler', StandardScaler())\\r\\n ]))\\r\\n ])),\\r\\n ('starting_verb', basic_utils.PosFieldExtractor('starting_verb_flag'))\\r\\n ])),\\r\\n \\r\\n ('clf', clf)\\r\\n ])\\r\\n \\r\\n return pipeline\",\n \"def make_encoder(self, input_size: int, latent_size: int) -> nn.Module:\\n pass\",\n \"def _create_encoder(self):\\n\\n def _init_weights(layer):\\n \\\"\\\"\\\"Initializes the weights of a layer based on type.\\\"\\\"\\\"\\n if isinstance(layer, (nn.Conv2d, nn.Linear)):\\n torch.nn.init.xavier_uniform_(layer.weight)\\n try:\\n # Some layers may not have biases, so catch the exception and pass.\\n layer.bias.data.fill_(0.0)\\n except AttributeError:\\n pass\\n\\n kernel_size = 5\\n pad = 2\\n input_channels = 1\\n first_conv_channels = 6\\n second_conv_channels = 16\\n max_pool_kernel = 2\\n linear_size = 120\\n n_pixels = 7\\n\\n encoder = nn.Sequential(\\n nn.Conv2d(\\n input_channels, first_conv_channels, kernel_size, padding=pad),\\n nn.BatchNorm2d(first_conv_channels),\\n nn.ReLU(),\\n nn.MaxPool2d(max_pool_kernel),\\n nn.Conv2d(\\n first_conv_channels, second_conv_channels, kernel_size,\\n padding=pad),\\n nn.BatchNorm2d(second_conv_channels),\\n nn.ReLU(),\\n nn.MaxPool2d(max_pool_kernel),\\n utils.Flatten(),\\n nn.Linear(n_pixels * n_pixels * self.n_digits * second_conv_channels,\\n linear_size),\\n nn.BatchNorm1d(linear_size),\\n nn.ReLU(),\\n nn.Linear(linear_size, self.embedding_dim),\\n nn.Linear(self.embedding_dim, self.n_classes, bias=False),\\n )\\n\\n encoder.apply(_init_weights)\\n\\n # This is the empirical approximation for initialization the vMF\\n # distributions for each class in the final layer.\\n if self.use_vmf:\\n utils.vmf_class_weight_init(encoder[-1].weight, self.kappa_confidence,\\n self.embedding_dim)\\n\\n return encoder\",\n \"def __init__(self,\\n input_feature: int,\\n output_feature: int = 19,\\n n_permutation_list: int = 6,\\n dropout: float = 0.2,\\n UseBN: bool = True):\\n super(LCNNBlock, self).__init__()\\n\\n self.conv_weights = nn.Linear(input_feature, output_feature)\\n self.batch_norm = nn.BatchNorm1d(output_feature)\\n self.UseBN = UseBN\\n self.activation = Shifted_softplus()\\n self.dropout = Custom_dropout(dropout, n_permutation_list)\\n self.permutation = n_permutation_list\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.63641286","0.61162806","0.6006187","0.59183264","0.56983024","0.5687006","0.5635837","0.56081927","0.55380064","0.5521205","0.54553306","0.5424783","0.5420473","0.54017746","0.53908736","0.5378384","0.5374217","0.5369382","0.5336188","0.5329734","0.5307692","0.52916044","0.5276007","0.5261386","0.5248024","0.52405596","0.523499","0.52210975","0.52181715","0.51951045","0.51901853","0.5187375","0.5180332","0.5169214","0.5166798","0.51507974","0.51466614","0.5142622","0.51340526","0.5130908","0.51202035","0.51197064","0.51108176","0.5104072","0.5103426","0.5101616","0.51014185","0.5097147","0.50933033","0.508618","0.5085971","0.5082887","0.5076999","0.5076419","0.50697815","0.50697815","0.5060799","0.5055889","0.5054564","0.5054564","0.5054564","0.5054564","0.5054564","0.5054564","0.5054564","0.50503474","0.50442654","0.5043586","0.50404817","0.50394833","0.50389206","0.50324315","0.50293016","0.50232995","0.5022257","0.5021825","0.5020447","0.50182474","0.50136125","0.5013126","0.50075006","0.50048584","0.50008893","0.49938014","0.4993562","0.49922997","0.4992163","0.4989229","0.49889767","0.49867058","0.49823126","0.49696803","0.4968982","0.49689698","0.49665368","0.4965924","0.49569455","0.49509102","0.49497998","0.49470967"],"string":"[\n \"0.63641286\",\n \"0.61162806\",\n \"0.6006187\",\n \"0.59183264\",\n \"0.56983024\",\n \"0.5687006\",\n \"0.5635837\",\n \"0.56081927\",\n \"0.55380064\",\n \"0.5521205\",\n \"0.54553306\",\n \"0.5424783\",\n \"0.5420473\",\n \"0.54017746\",\n \"0.53908736\",\n \"0.5378384\",\n \"0.5374217\",\n \"0.5369382\",\n \"0.5336188\",\n \"0.5329734\",\n \"0.5307692\",\n \"0.52916044\",\n \"0.5276007\",\n \"0.5261386\",\n \"0.5248024\",\n \"0.52405596\",\n \"0.523499\",\n \"0.52210975\",\n \"0.52181715\",\n \"0.51951045\",\n \"0.51901853\",\n \"0.5187375\",\n \"0.5180332\",\n \"0.5169214\",\n \"0.5166798\",\n \"0.51507974\",\n \"0.51466614\",\n \"0.5142622\",\n \"0.51340526\",\n \"0.5130908\",\n \"0.51202035\",\n \"0.51197064\",\n \"0.51108176\",\n \"0.5104072\",\n \"0.5103426\",\n \"0.5101616\",\n \"0.51014185\",\n \"0.5097147\",\n \"0.50933033\",\n \"0.508618\",\n \"0.5085971\",\n \"0.5082887\",\n \"0.5076999\",\n \"0.5076419\",\n \"0.50697815\",\n \"0.50697815\",\n \"0.5060799\",\n \"0.5055889\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.5054564\",\n \"0.50503474\",\n \"0.50442654\",\n \"0.5043586\",\n \"0.50404817\",\n \"0.50394833\",\n \"0.50389206\",\n \"0.50324315\",\n \"0.50293016\",\n \"0.50232995\",\n \"0.5022257\",\n \"0.5021825\",\n \"0.5020447\",\n \"0.50182474\",\n \"0.50136125\",\n \"0.5013126\",\n \"0.50075006\",\n \"0.50048584\",\n \"0.50008893\",\n \"0.49938014\",\n \"0.4993562\",\n \"0.49922997\",\n \"0.4992163\",\n \"0.4989229\",\n \"0.49889767\",\n \"0.49867058\",\n \"0.49823126\",\n \"0.49696803\",\n \"0.4968982\",\n \"0.49689698\",\n \"0.49665368\",\n \"0.4965924\",\n \"0.49569455\",\n \"0.49509102\",\n \"0.49497998\",\n \"0.49470967\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.5075871"},"document_rank":{"kind":"string","value":"54"}}},{"rowIdx":90878,"cells":{"query":{"kind":"string","value":"Pass output of multihead attention to mlp block"},"document":{"kind":"string","value":"def call(self, inputs, *args, **kwargs):\n\n outputs = self.mlp(inputs, *args, **kwargs)\n return outputs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def attention_module_multi_head(roi_feat, position_embedding,\n nongt_dim, fc_dim, feat_dim,\n dim=(1024, 1024, 1024),\n group=16, index=1):\n dim_group = (dim[0] / group, dim[1] / group, dim[2] / group)\n nongt_roi_feat = mx.symbol.slice_axis(data=roi_feat, axis=0, begin=0, end=nongt_dim)\n # [num_rois * nongt_dim, emb_dim]\n position_embedding_reshape = mx.sym.Reshape(position_embedding, shape=(-3, -2))\n # position_feat_1, [num_rois * nongt_dim, fc_dim]\n position_feat_1 = mx.sym.FullyConnected(name='pair_pos_fc1_' + str(index),\n data=position_embedding_reshape,\n num_hidden=fc_dim)\n position_feat_1_relu = mx.sym.Activation(data=position_feat_1, act_type='relu')\n # aff_weight, [num_rois, nongt_dim, fc_dim]\n aff_weight = mx.sym.Reshape(position_feat_1_relu, shape=(-1, nongt_dim, fc_dim))\n # aff_weight, [num_rois, fc_dim, nongt_dim]\n aff_weight = mx.sym.transpose(aff_weight, axes=(0, 2, 1))\n\n # multi head\n assert dim[0] == dim[1], 'Matrix multiply requires same dimensions!'\n q_data = mx.sym.FullyConnected(name='query_' + str(index),\n data=roi_feat,\n num_hidden=dim[0])\n q_data_batch = mx.sym.Reshape(q_data, shape=(-1, group, dim_group[0]))\n q_data_batch = mx.sym.transpose(q_data_batch, axes=(1, 0, 2))\n k_data = mx.symbol.FullyConnected(name='key_' + str(index),\n data=nongt_roi_feat,\n num_hidden=dim[1])\n k_data_batch = mx.sym.Reshape(k_data, shape=(-1, group, dim_group[1]))\n k_data_batch = mx.sym.transpose(k_data_batch, axes=(1, 0, 2))\n v_data = nongt_roi_feat\n # v_data = mx.symbol.FullyConnected(name='value_'+str(index)+'_'+str(gid), data=roi_feat, num_hidden=dim_group[2])\n aff = mx.symbol.batch_dot(lhs=q_data_batch, rhs=k_data_batch, transpose_a=False, transpose_b=True)\n # aff_scale, [group, num_rois, nongt_dim]\n aff_scale = (1.0 / math.sqrt(float(dim_group[1]))) * aff\n aff_scale = mx.sym.transpose(aff_scale, axes=(1, 0, 2))\n\n assert fc_dim == group, 'fc_dim != group'\n # weighted_aff, [num_rois, fc_dim, nongt_dim]\n weighted_aff = mx.sym.log(mx.sym.maximum(left=aff_weight, right=1e-6)) + aff_scale\n aff_softmax = mx.symbol.softmax(data=weighted_aff, axis=2, name='softmax_' + str(index))\n # [num_rois * fc_dim, nongt_dim]\n aff_softmax_reshape = mx.sym.Reshape(aff_softmax, shape=(-3, -2))\n # output_t, [num_rois * fc_dim, feat_dim]\n output_t = mx.symbol.dot(lhs=aff_softmax_reshape, rhs=v_data)\n # output_t, [num_rois, fc_dim * feat_dim, 1, 1]\n output_t = mx.sym.Reshape(output_t, shape=(-1, fc_dim * feat_dim, 1, 1))\n # linear_out, [num_rois, dim[2], 1, 1]\n linear_out = mx.symbol.Convolution(name='linear_out_' + str(index), data=output_t,\n kernel=(1, 1), num_filter=dim[2], num_group=fc_dim)\n output = mx.sym.Reshape(linear_out, shape=(0, 0))\n return output","def build_model_multihead_attention_multiscaleCNN4_covermore(self,\n dim_attention,headnum,\n embedding_vec,\n load_weights=False, weight_dir=None,\n nb_filters=32,filters_length1=1,\n filters_length2=5,\n filters_length3=10,\n pooling_size=3,\n drop_input=0,\n drop_cnn=0.2,\n drop_flat=0,\n W1_regularizer=0.005,\n W2_regularizer=0.005,\n Att_regularizer_weight=0.0005,\n BatchNorm=False,\n fc_dim = 50,\n fcnum=0,\n posembed=False,\n pos_dmodel=40,\n pos_nwaves = 20,\n posmod = 'concat',\n regularfun=1,\n huber_delta=1,\n activation='gelu',\n activationlast='gelu',\n add_avgpooling = False,\n poolingmod=1,\n normalizeatt=False,\n regressionmodel=False,\n attmod = \"softmax\",\n sharp_beta=1,\n lr = 0.001 \n ):\n ###print('Advanced Masking')\n def mask_func(x):\n return x[0] * x[1]\n \n ###print(posembed)\n ###print(posmod)\n input = Input(shape=(self.max_len,), dtype='int8')\n input_mask = Input(shape=([int(self.max_len/pooling_size), 1]), dtype='float32')\n embedding_layer = Embedding(len(embedding_vec), len(embedding_vec[0]), weights=[embedding_vec],\n input_length=self.max_len,\n trainable=False)\n embedding_output = Dropout(drop_input)(embedding_layer(input)) #layer2\n if 'gelu' in activation:\n activationfun=gelu\n else:\n activationfun = 'relu'\n \n if 'gelu' in activationlast:\n activationlastfun = gelu\n else:\n activationlastfun='relu'\n \n ###print(activationfun)\n ###print(activationlastfun)\n with tf.name_scope('first_cnn'):\n first_cnn = Convolution1D(nb_filters, filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN1')(embedding_output) #layer3\n first_cnn2 = Convolution1D(int(nb_filters/2), filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(first_cnn) #layer5\n second_cnn = Convolution1D(nb_filters, filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN2')(embedding_output) #layer4\n second_cnn2 = Convolution1D(int(nb_filters/2), filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(second_cnn)\n third_cnn = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN3')(embedding_output)\n \n third_cnn2 = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(third_cnn)\n if BatchNorm:\n first_cnn2 = BatchNormalization()(first_cnn2)\n second_cnn2 = BatchNormalization()(second_cnn2)\n third_cnn2 = BatchNormalization()(third_cnn2)\n \n if not add_avgpooling:\n if poolingmod == 1:\n pooling_layer = MaxPooling1D(pool_length=pooling_size, stride=pooling_size)\n else:\n pooling_layer = AveragePooling1D(pool_length=pooling_size, stride=pooling_size)\n \n cnn_output1 = Dropout(drop_cnn)(pooling_layer(first_cnn2))\n cnn_output2 = Dropout(drop_cnn)(pooling_layer(second_cnn2))\n cnn_output3 = Dropout(drop_cnn)(pooling_layer(third_cnn2))\n else:\n first_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\n first_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\n cnn_output1 = Dropout(drop_cnn)(concatenate([first_cnn2_max,first_cnn2_avg],axis=-1))\n second_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\n second_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\n cnn_output2 = Dropout(drop_cnn)(concatenate([second_cnn2_max,second_cnn2_avg],axis=-1))\n third_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\n third_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\n cnn_output3 = Dropout(drop_cnn)(concatenate([third_cnn2_max,third_cnn2_avg],axis=-1))\n \n \n \n if posembed:\n ##print(posmod)\n from position_embedding import PositionEmbedding\n if posmod == 'concat':\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb1')(cnn_output1)\n cnn_output1 = concatenate([cnn_output1, pos_emb1], axis=-1)\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb2')(cnn_output2)\n cnn_output2 = concatenate([cnn_output2, pos_emb2], axis=-1)\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb3')(cnn_output3)\n cnn_output3 = concatenate([cnn_output3, pos_emb3], axis=-1)\n else:\n ##print(\"yes add posmod\")\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output1)[-1]/2), d_model=int_shape(cnn_output1)[-1],name='pos_emb1')(cnn_output1)\n cnn_output1 = Add()([cnn_output1, pos_emb1])\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output2)[-1]/2), d_model=int_shape(cnn_output2)[-1],name='pos_emb2')(cnn_output2)\n cnn_output2 = Add()([cnn_output2, pos_emb2])\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output3)[-1]/2), d_model=int_shape(cnn_output3)[-1],name='pos_emb3')(cnn_output3)\n cnn_output3 = Add()([cnn_output3, pos_emb3])\n \n mask_input1 = []\n mask_input1.append(cnn_output1)\n mask_input1.append(input_mask)\n cnn_mask_output1 = Lambda(mask_func)(mask_input1)\n del mask_input1\n mask_input2 = []\n mask_input2.append(cnn_output2)\n mask_input2.append(input_mask)\n cnn_mask_output2 = Lambda(mask_func)(mask_input2)\n del mask_input2\n mask_input3 = []\n mask_input3.append(cnn_output3)\n mask_input3.append(input_mask)\n cnn_mask_output3 = Lambda(mask_func)(mask_input3)\n del mask_input3\n \n if regularfun==1:\n regularizerfunction_W1 = regularizers.l1(W1_regularizer)\n regularizerfunction_W2 = regularizers.l1(W2_regularizer)\n elif regularfun==2:\n regularizerfunction_W1 = regularizers.l2(W1_regularizer)\n regularizerfunction_W2 = regularizers.l2(W2_regularizer)\n elif regularfun ==3:\n regularizerfunction_W1 = smoothL1(W1_regularizer,huber_delta)\n regularizerfunction_W2 = smoothL1(W2_regularizer,huber_delta)\n \n with tf.name_scope('multihead_attention'):\n att1,att1_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att1\")(concatenate([cnn_mask_output1, input_mask]))#-5 layer\n \n att2,att2_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att2\")(concatenate([cnn_mask_output2, input_mask])) #-4 layer\n \n att3,att3_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att3\")(concatenate([cnn_mask_output3, input_mask])) #-3 layer\n \n if BatchNorm:\n att1 = BatchNormalization()(att1)\n att2 = BatchNormalization()(att2)\n att3 = BatchNormalization()(att3)\n \n \n output = Dropout(drop_flat)(Flatten()(concatenate([att1,att2,att3]))) #-2 layer\n \n fc = output\n for _ in range(fcnum):\n fc = Dense(fc_dim,activation='relu')(fc)\n fc = Dropout(drop_flat)(fc)\n \n with tf.name_scope(''):\n if regressionmodel:\n preds = Dense(self.nb_classes,activation='softmax')(fc) #-1 layer\n else:\n preds = Dense(self.nb_classes,activation='sigmoid')(fc) #-1 layer\n \n self.model = Model(inputs=[input,input_mask], outputs=preds)\n from keras import optimizers\n # optim = optimizers.RMSprop()\n optim = optimizers.Adam(lr=lr, decay=5e-5) #The paper uses a decay rate alpha = alpha/sqrt(t) updted each epoch (t) for the logistic regression demonstration.\n #optim = optimizers.nadam()\n #optim = RAdam()\n if regressionmodel:\n self.model.compile(loss='kld',optimizer=optim,metrics=['acc'])\n else:\n self.model.compile(loss='binary_crossentropy',optimizer=optim,metrics=['binary_accuracy','categorical_accuracy'])\n \n \n \n if load_weights:\n self.model.load_weights(weight_dir)\n \n self.is_built = True\n self.bn = False\n self.model.summary()","def block(inp, scope, past, config):\n with tf.variable_scope(scope):\n nx = inp.shape[-1].value\n norm = normalise_tensor(inp, 'ln_1')\n attn, present = attention(norm, 'attn', nx, past=past, config=config)\n out = attn + inp\n norm = normalise_tensor(out, 'ln_2')\n mlp_out = multilayer_perceptron(norm, 'mlp', nx * 4) # note that hidden dim is 4x\n out = out + mlp_out\n return out, present","def forward(self, input):\n x = self.emb(input)\n output = self.attention(x)\n # print(\"Idggfghrf \", output.size())\n return output","def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):\n super(TransformerBlock, self).__init__()\n # Attention\n self.attention = MultiHeadAttention(\n num_heads=num_heads, key_dim=D, dropout=dropout\n )\n self.norm_attention = LayerNormalization(epsilon=norm_eps)\n\n # MLP\n self.mlp = MLPBlock(hidden_layers, dropout)\n self.norm_mlp = LayerNormalization(epsilon=norm_eps)","def __init__(self, nheads, d_model):\n super(MultiheadAttention, self).__init__()\n assert d_model % nheads == 0\n self.d_head = d_model // nheads\n self.nheads = nheads\n self.Q_fc = nn.Linear(d_model, d_model, bias=False)\n self.K_fc = nn.Linear(d_model, d_model, bias=False)\n self.V_fc = nn.Linear(d_model, d_model, bias=False)\n self.output_fc = nn.Linear(d_model, d_model, bias=False)\n self.attn = None","def __init__(\n self,\n embed_dim,\n num_heads,\n dropout=0.,\n bias=True,\n kdim=None,\n vdim=None,\n ):\n super(MultiheadAttention, self).__init__()\n self.embed_dim = embed_dim\n self.kdim = kdim if kdim is not None else embed_dim\n self.vdim = vdim if vdim is not None else embed_dim\n self._qkv_same_embed_dim = self.kdim == embed_dim and self.vdim == embed_dim\n self.num_heads = num_heads\n self.dropout = dropout\n self.head_dim = embed_dim // num_heads\n if self.head_dim * num_heads != self.embed_dim:\n raise ValueError('<embed_dim> must be divisible by <num_heads>.')\n if not self._qkv_same_embed_dim:\n self.q_proj_weight = Parameter(Tensor(embed_dim, embed_dim))\n self.k_proj_weight = Parameter(Tensor(embed_dim, self.kdim))\n self.v_proj_weight = Parameter(Tensor(embed_dim, self.vdim))\n self.register_parameter('in_proj_weight', None)\n else:\n self.in_proj_weight = Parameter(Tensor(3 * embed_dim, embed_dim))\n self.register_parameter('q_proj_weight', None)\n self.register_parameter('k_proj_weight', None)\n self.register_parameter('v_proj_weight', None)\n if bias:\n self.in_proj_bias = Parameter(Tensor(3 * embed_dim))\n else:\n self.register_parameter('in_proj_bias', None)\n self.out_proj = Linear(embed_dim, embed_dim, bias=bias)\n self.reset_parameters()","def __init__(self, dm, h, hidden, drop_rate=0.1):\r\n super(EncoderBlock, self).__init__()\r\n self.mha = MultiHeadAttention(dm, h)\r\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\r\n activation='relu')\r\n self.dense_output = tf.keras.layers.Dense(units=dm)\r\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\r\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)","def mlp(input_dim,variance=1., weight_variance=None,bias_variance=100.,ARD=False):\r\n part = parts.mlp.MLP(input_dim,variance,weight_variance,bias_variance,ARD)\r\n return kern(input_dim, [part])","def forward(self, h_t, mem):\n\n # ==\n # Compute non-sparse attention\n cur_mem_size = mem.size(0)\n batch_size = mem.size(1)\n\n # Repeat h_t from (batch, hid_dim) to (mem_size, batch, hid_dim)\n # one for each memory slot\n h_repeated = (h_t.clone().unsqueeze(0)\n .repeat(cur_mem_size, 1, 1))\n\n # Concat to [h_t, memory], shape (mem_size, batch, hid_dim*2)\n mlp_h_attn = torch.cat((h_repeated, mem), dim=2)\n\n # (Optional) block past attention gradient\n if self.block_attn_grad_past:\n mlp_h_attn = mlp_h_attn.detach()\n\n # Compute attention via tanh then matmul with weight matrix\n # Weight matrix has shape (hid_dim * 2 , 1)\n # Output (attn_w) has shape (mem_size, batch, 1)\n mlp_h_attn = self.tanh(mlp_h_attn)\n attn_w = torch.matmul(mlp_h_attn, self.w_t)\n\n # ==\n # Sparsify attention: set top-k to non-zero and normalize\n attn_w = attn_w.view(cur_mem_size, batch_size)\n attn_w = self.sparse_attention(attn_w)\n attn_w = attn_w.view(cur_mem_size, batch_size, 1)\n\n # ==\n # Extract from memory using attention\n attn_w_rep = attn_w.repeat(1, 1, self.hidden_size)\n h_mem_w = attn_w_rep * mem # (mem_size, batch, hid_dim)\n\n # Attention-extracted memory information\n m_t = torch.sum(h_mem_w, dim=0) # (batch, hid_dim)\n\n return m_t","def forward(self, input):\n if self.dataset_name == 'mnist':\n if self.temp_attn:\n y, attn_weight_list = self.tcanet(input)\n o = self.decoder(y[:, :, -1])\n return F.log_softmax(o, dim=1).contiguous()\n else:\n y = self.tcanet(input)\n o = self.decoder(y[:, :, -1])\n return F.log_softmax(o, dim=1).contiguous()\n emb = self.drop(self.word_encoder(input))\n if self.temp_attn:\n y, attn_weight_list = self.tcanet(emb.transpose(1, 2))\n y = self.decoder(y.transpose(1, 2))\n return y.contiguous(), [attn_weight_list[0], attn_weight_list[self.num_levels // 2], attn_weight_list[-1]]\n else:\n y = self.tcanet(emb.transpose(1, 2))\n y = self.decoder(y.transpose(1, 2))\n return y.contiguous()","def __init__(self, cfg):\n super(MHBCoAtt, self).__init__()\n self.cfg = cfg\n # word embedding: q_vocab_size, 1024\n self.word_embedding = nn.Embedding(cfg.q_vocab_size, cfg.emb_dim)\n # LSTM\n if cfg.glove:\n self.lstm = nn.LSTM(input_size=cfg.emb_dim*2,\n hidden_size=cfg.hidden_dim,\n num_layers=cfg.num_layers,\n batch_first=True)\n else:\n self.lstm = nn.LSTM(input_size=cfg.emb_dim,\n hidden_size=cfg.hidden_dim,\n num_layers=cfg.num_layers,\n batch_first=True)\n\n self.dropout_l = nn.Dropout(p = 0.3)\n # question attention\n self.ques_att_conv1 = nn.Conv2d(cfg.hidden_dim, 512, [1,1])\n self.ques_att_conv2 = nn.Conv2d(512, 2, [1,1])\n\n # question attentive feature fuse with image feature, according to paper: k * o = 5000, k = 5\n self.ques_proj1 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.img_conv1d = nn.Conv2d(cfg.img_feature_channel, 5000, [1, 1])\n self.dropout_m = nn.Dropout(p = 0.1)\n\n # co-attention conv layers\n self.co_att_conv1 = nn.Conv2d(1000, 512, [1,1])\n self.co_att_conv2 = nn.Conv2d(512, 2, [1,1])\n\n # co_attentive feature fuse with question attentive feature\n self.ques_proj2 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.ques_proj3 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.img_proj2 = nn.Linear(2*cfg.img_feature_channel, 5000)\n self.img_proj3 = nn.Linear(2*cfg.img_feature_channel, 5000)\n\n # prediction fully connected layer\n self.linear_pred = nn.Linear(2000, cfg.a_vocab_size)","def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\n ptm=\"bert\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\n super(PTMClassifier_PNN, self).__init__()\n \n if old_cols is None:\n old_cols = []\n \n self.old_cols = []\n \n self.output_size = output_size\n self.hidden_size = hidden_size # default\n self.feature_size = feature_size\n \n self.fc1 = nn.Linear(self.hidden_size, 100)\n self.classifier = nn.Linear(100, self.output_size)\n \n if len(old_cols) > 0:\n self.old_fc1s = ListModule()\n self.base_1 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n self.base_2 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n \n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_1)\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_2)\n \n for old_col in old_cols:\n self.old_fc1s.append(\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\n \n # ptm\n self.ptm = ptm.lower()\n assert self.ptm in supported_ptm\n ptm_ = import_from(\"transformers\", supported_ptm[self.ptm][0] + \"Model\")\n \n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\n if tokenizer is not None:\n self.encoder.resize_token_embeddings(len(tokenizer))\n self.prob_l = prob_l\n \n self.require_proto = require_proto\n \n # todo : modify net into one network\n if self.require_proto:\n # prototype-based classification\n self.net = nn.Sequential(\n self.fc1\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.fc1,\n self.encoder_adaptor,\n nn.ReLU(),\n )\n else:\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\n self.net = nn.Sequential(\n self.classifier\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.classifier\n )\n \n self.reset_parameters()","def __init__(self, config: BertConfig):\r\n super().__init__(config)\r\n ### YOUR CODE HERE\r\n self.num_labels = config.num_labels # [0, 1] (start or end)\r\n self.bert = BertModel(config)\r\n self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels) # TODO: Not a separate FFN ? (For Start_FFN and End_FFN)\r\n\r\n ### END YOUR CODE\r\n\r\n # Don't forget initializing the weights\r\n self.init_weights()","def call(self, inputs):\n # Feed attention\n norm_attention = self.norm_attention(inputs)\n\n attention = self.attention(query=norm_attention, value=norm_attention)\n\n # Skip Connection\n attention += inputs\n\n # Feed MLP\n outputs = self.mlp(self.norm_mlp(attention))\n\n # Skip Connection\n outputs += attention\n\n return outputs","def mlpfwd(self,inputs):\n\n self.hidden = np.dot(inputs,self.weights1);\n self.hidden = 1.0/(1.0+np.exp(-self.beta*self.hidden))\n self.hidden = np.concatenate((self.hidden,-np.ones((np.shape(inputs)[0],1))),axis=1)\n\n outputs = np.dot(self.hidden,self.weights2);\n\n # Different types of output neurons\n if self.outtype == 'linear':\n \treturn outputs\n elif self.outtype == 'logistic':\n return 1.0/(1.0+np.exp(-self.beta*outputs))\n elif self.outtype == 'softmax':\n normalisers = np.sum(np.exp(outputs),axis=1)*np.ones((1,np.shape(outputs)[0]))\n return np.transpose(np.transpose(np.exp(outputs))/normalisers)\n else:\n print \"error\"","def __init__(self, num_heads: int, size: int, dropout: float = 0.1):\n super(MultiHeadedAttention, self).__init__()\n\n assert size % num_heads == 0\n\n self.head_size = head_size = size // num_heads\n self.model_size = size\n self.num_heads = num_heads\n\n self.k_layer = nn.Linear(size, num_heads * head_size)\n self.v_layer = nn.Linear(size, num_heads * head_size)\n self.q_layer = nn.Linear(size, num_heads * head_size)\n\n self.output_layer = nn.Linear(size, size)\n self.softmax = nn.Softmax(dim=-1)\n self.dropout = nn.Dropout(dropout)","def get_attention_multiscale_batch(self, X,X_mask):\n layer = -3\n attmodel1 = Model(self.model.inputs,self.model.get_layer('att1').output[1])\n attmodel2 = Model(self.model.inputs,self.model.get_layer('att2').output[1])\n attmodel3 = Model(self.model.inputs,self.model.get_layer('att3').output[1])\n return attmodel1.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100),attmodel2.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100),attmodel3.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100)","def mtf_bitransformer_base():\n hparams = mtf_transformer2_base()\n hparams.max_length = 256\n hparams.shared_embedding = True\n # HYPERPARAMETERS FOR THE LAYER STACKS\n hparams.add_hparam(\"encoder_layers\", [\"self_att\", \"drd\"] * 6)\n hparams.add_hparam(\"decoder_layers\", [\"self_att\", \"enc_att\", \"drd\"] * 6)\n hparams.add_hparam(\"encoder_num_layers\", 6)\n hparams.add_hparam(\"decoder_num_layers\", 6)\n # number of heads in multihead attention\n hparams.add_hparam(\"encoder_num_heads\", 8)\n hparams.add_hparam(\"decoder_num_heads\", 8)\n hparams.add_hparam(\"local_attention_radius\", 128)\n\n # default of 0 for standard transformer behavior\n # 1 means a single set of keys and values that are read by all query heads\n hparams.add_hparam(\"encoder_num_memory_heads\", 0)\n hparams.add_hparam(\"decoder_num_memory_heads\", 0)\n # share attention keys and values\n hparams.add_hparam(\"encoder_shared_kv\", False)\n hparams.add_hparam(\"decoder_shared_kv\", False)\n\n # Parameters for computing the maximum decode length in beam search.\n # Maximum decode length is:\n # min(max_length,\n # decode_length_multiplier * input_length + decode_length_constant)\n hparams.add_hparam(\"decode_length_multiplier\", 1.5)\n hparams.add_hparam(\"decode_length_constant\", 10.0)\n # used during decoding\n hparams.add_hparam(\"alpha\", 0.6)\n hparams.sampling_temp = 0.0\n return hparams","def __init__(self, num_heads: int, size: int, size_v: int, dropout: float = 0.1):\n super(ContMultiHeadedAttention, self).__init__()\n\n assert size % num_heads == 0\n\n self.head_size = head_size = size // num_heads\n self.model_size = size\n self.num_heads = num_heads\n\n self.k_layer = nn.Linear(size, num_heads * head_size)\n self.v_layer = nn.Linear(size_v, num_heads * head_size)\n self.q_layer = nn.Linear(size, num_heads * head_size)\n\n self.output_layer = nn.Linear(size, size_v)\n self.softmax = nn.Softmax(dim=-1)\n self.dropout = nn.Dropout(dropout)","def mtf_unitransformer_base():\n hparams = mtf_transformer2_base()\n hparams.add_hparam(\"autoregressive\", True)\n # HYPERPARAMETERS FOR THE SINGLE LAYER STACK\n hparams.add_hparam(\"layers\", [\"self_att\", \"drd\"] * 6)\n # number of heads in multihead attention\n hparams.add_hparam(\"num_heads\", 8)\n # default of 0 for standard transformer behavior\n # 1 means a single set of keys and values that are read by all query heads\n hparams.add_hparam(\"num_memory_heads\", 0)\n # share attention keys and values\n hparams.add_hparam(\"shared_kv\", False)\n # if nonzero then use local attention\n hparams.add_hparam(\"local_attention_radius\", 128)\n return hparams","def __init__(self, emb_size: int = 768, num_heads: int = 12, dropout: float = 0.1):\n super().__init__()\n self.emb_size = emb_size\n self.num_heads = num_heads\n\n # Check \"head dim\" = number of features per head (d_k).\n self.head_dim = emb_size // num_heads\n assert self.head_dim * num_heads == self.emb_size, \"emb_size must be divisible by num_heads\"\n # Calculate scaling factor.\n self.scale = 1 / (self.head_dim ** 0.5)\n\n # V1: in vanilla self-attention Q,K,V are square matrices.\n # self.keys = nn.Linear(emb_size, emb_size)\n # self.queries = nn.Linear(emb_size, emb_size)\n # self.values = nn.Linear(emb_size, emb_size)\n \n # V2: single layer with emb_size, split into num (heads * head_dim) * 3 (Q,K,V).\n self.qkv = nn.Linear(emb_size, emb_size * 3)\n\n # Attention dropout.\n self.att_drop = nn.Dropout(dropout)\n self.projection = nn.Linear(emb_size, emb_size)","def forward(\r\n self,\r\n input_ids,\r\n attention_mask: torch.Tensor,\r\n token_type_ids: torch.Tensor\r\n ):\r\n ### YOUR CODE HERE\r\n output = self.bert(\r\n input_ids=input_ids,\r\n attention_mask=attention_mask,\r\n token_type_ids=token_type_ids,\r\n )\r\n\r\n sequence_output = output[0] # the last hidden state (batch, sequence_length, hidden_size)\r\n logits = self.qa_outputs(sequence_output)\r\n start_logits, end_logits = logits.split(1, dim=-1)\r\n start_logits = start_logits.squeeze(-1)\r\n end_logits = end_logits.squeeze(-1)\r\n\r\n outputs = (start_logits, end_logits) # + output[2:]\r\n\r\n return outputs\r\n ### END YOUR CODE","def mtf_transformer_lm_moe():\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\n hparams.decoder_layers = [\"att\", \"moe\"] * 4\n moe.set_default_moe_hparams(hparams)\n hparams.mesh_shape = \"all:8\"\n hparams.layout = \"batch:all;experts:all\"\n return hparams","def exec_attention(self,curr_step): \n\n assert(self.curr_step_idx > 0 and self.dlist is not None), \"Step Error: Must call init before combine\" \n \n detectType = curr_step[\"detectionNetwork\"]\n paramsFile = curr_step[\"paramsFile\"]\n funclist = curr_step[\"funclist\"]\n\n #verify raw data & dlist\n # self.B_VER(self.sess_path, self.dlist)\n raw_datadir = self.sess_path\n dest_datadir = self.sess_path \n\n model_dict = {\"detectType\": detectType, \"paramsFile\" : paramsFile} \n\n for i, folder in enumerate(self.dlist):\n flist = funclist[i]\n self.data_utils.DETECT(raw_datadir, folder, dest_datadir, model_dict, flist=[], preview=False)\n self.default_vis(curr_step)","def build_mlp(input_, config):\n current_input = input_\n print(current_input)\n for i in range(len(config['fcc_layers']) - 1):\n current_input = tf.keras.layers.Dense(\n units=config['fcc_layers'][i], activation='tanh',\n name='fcc_layer_{}'.format(i + 1))(current_input)\n current_input = tf.keras.layers.Dropout(\n rate=config['dropout'], name='dropout_{}'.format(i + 1))(current_input)\n cascade_embedding_layer = tf.keras.layers.Dense(\n units=config['fcc_layers'][-1], activation='tanh',\n name='cascade_embedding_layer')(current_input)\n cascade_embedding_layer_do = tf.keras.layers.Dropout(\n rate=config['dropout'],\n name='cascade_embedding_dropout')(cascade_embedding_layer)\n prediction_layer = tf.keras.layers.Dense(\n units=1, activation='sigmoid',\n name='prediction_layer')(cascade_embedding_layer_do)\n return cascade_embedding_layer, prediction_layer","def pre_forward(self, *args, **kwargs):\n batch_size = args[0].shape[0]\n if not self.instantiated:\n self.hidden_dim = args[0].shape[-1]\n self.instantiate(hidden_dim=self.hidden_dim)\n if self.past_key_reparam is None:\n past_key = self.past_key\n else:\n past_key = self.past_key_reparam\n if self.past_value_reparam is None:\n past_value = self.past_value\n else:\n past_value = self.past_value_reparam\n\n\n def expand_batchsize(x):\n x = x.reshape(self.prefix_token_num, self.num_heads, -1).transpose(0,1)\n x = x.unsqueeze(0).expand(batch_size, *x.shape)\n return x\n # from IPython import embe\n\n if 'past_key_value' not in kwargs or kwargs['past_key_value'] is None:\n kwargs['past_key_value'] = (expand_batchsize(past_key), expand_batchsize(past_value))\n\n if 'attention_mask' in kwargs and kwargs['attention_mask'] is not None:\n am = kwargs['attention_mask'] # Should check the format of the attention_mask when moving to a new plm.\n kwargs['attention_mask'] = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\n elif len(args) >1: # attention mask is passed via positional argument\n am = args[1]\n am = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\n args = (args[0], am) + args[2:]\n # from IPython import embed\n # embed(header = \"Herein prefixroberta\")\n return args, kwargs","def forward(self, context, question, c_masks):\n\n # context_hidden_proj = self.attn_proj(context) # shape (batch_size, context_len, hidden_size)\n logits_1 = torch.bmm(self.attn_proj_1(context), torch.unsqueeze(question, 2)) # shape (batch_size, context_len, 1)\n logits_1 = torch.squeeze(logits_1, -1)\n\n logits_2 = torch.bmm(self.attn_proj_2(context), torch.unsqueeze(question, 2)) # shape (batch_size, context_len, 1)\n logits_2 = torch.squeeze(logits_2, -1)\n\n log_p1 = masked_softmax(logits_1, c_masks, dim=1, log_softmax=True)\n log_p2 = masked_softmax(logits_2, c_masks, dim=1, log_softmax=True)\n #\n # alpha_t = masked_softmax(scores, c_masks, dim=1) # shape (batch_size, context_len)\n # output_t = torch.mul(context, alpha_t.unsqueeze(2))\n #\n # output_t = torch.bmm(torch.unsqueeze(alpha_t, 1), context) # shape (batch_size, 1, hidden_size)\n # output_t = output_t.squeeze(1)\n return log_p1, log_p2","def create_quantizable_multihead_attention(module: torch.nn.MultiheadAttention) -> QuantizableMultiheadAttention:\n # inspect MHA if bias is required.\n bias = module.in_proj_bias is not None\n\n # if bias k/v parameter exist set quantizable MHA to create 3 separate bias tensors as expected.\n add_bias_kv = module.bias_k is not None and module.bias_v is not None\n\n q_MHA = QuantizableMultiheadAttention(embed_dim=module.embed_dim, num_heads=module.num_heads,\n dropout=module.dropout, bias=bias, add_bias_kv=add_bias_kv,\n add_zero_attn=module.add_zero_attn, kdim=module.kdim, vdim=module.vdim,\n batch_first=module.batch_first)\n\n # copy over weight and bias tensors\n with torch.no_grad():\n if module.in_proj_weight is not None:\n weights_q, weights_k, weights_v = torch.chunk(module.in_proj_weight.data, 3, dim=0)\n else:\n weights_q = module.q_proj_weight.data\n weights_k = module.k_proj_weight.data\n weights_v = module.v_proj_weight.data\n q_MHA.linear_Q.weight.copy_(weights_q)\n q_MHA.linear_K.weight.copy_(weights_k)\n q_MHA.linear_V.weight.copy_(weights_v)\n\n q_MHA.out_proj.weight.copy_(module.out_proj.weight.data)\n\n if bias:\n bias_q, bias_k, bias_v = torch.chunk(module.in_proj_bias.data, 3, dim=0)\n if add_bias_kv:\n bias_k = q_MHA.linear_K.bias.copy_(module.bias_k.data)\n bias_v = q_MHA.linear_V.bias.copy_(module.bias_v.data)\n q_MHA.linear_K.bias.copy_(bias_k)\n q_MHA.linear_V.bias.copy_(bias_v)\n q_MHA.linear_Q.bias.copy_(bias_q)\n\n q_MHA.out_proj.bias.copy_(module.out_proj.bias.data)\n\n return q_MHA","def forward(self, inputs):\n location_bias, context_bias, pos_emb, q, k, v, mask = inputs\n\n d_feature = q.shape[-1]\n n_heads = self._n_heads\n if d_feature % n_heads != 0:\n raise ValueError(\n f'Dimensionality of feature embedding ({d_feature}) is not a '\n f'multiple of the requested number of attention heads ({n_heads}).')\n\n if self._mode == 'predict':\n self._fast_inference_update_state((k, v), self.state)\n (k, v, _) = self.state\n\n per_head_results, dots = DotProductAttention(\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(q),\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(k),\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(v),\n pos_emb.reshape((-1, n_heads, d_feature // n_heads)),\n context_bias,\n location_bias,\n mask,\n dropout=self._dropout,\n mode=self._mode,\n rng=self.rng,\n chunk_len=self._chunk_len,\n chunk_offset=self._chunk_offset)\n if self._mode == 'viz':\n self.state = dots\n merged_results = MergeHeads(\n n_heads, merged_batch_and_head=False).forward(per_head_results)\n return merged_results, mask","def forward(\n self, inp: torch.Tensor, ucb_alpha: Optional[float] = None\n ) -> Dict[str, torch.Tensor]:\n\n mlp_out = self.deep_represent_layers(inp)\n # preprocess by DeepRepresent module before fed to LinUCB layer\n # trainer needs pred_label and mlp_out to update parameters\n\n # append a column of ones to features for LinUCB intercept/bias\n mlp_out_with_ones = torch.cat(\n (\n torch.ones(mlp_out.shape[:-1], device=mlp_out.device).unsqueeze(-1),\n mlp_out,\n ),\n -1,\n )\n\n if ucb_alpha is None:\n ucb_alpha = self.ucb_alpha\n\n if not self.nn_e2e:\n pred_label = torch.matmul(mlp_out_with_ones, self.coefs)\n else:\n pred_label = self.linear_layer(mlp_out_with_ones).squeeze(-1)\n\n if ucb_alpha != 0:\n pred_sigma = torch.sqrt(\n batch_quadratic_form(mlp_out_with_ones, self.inv_avg_A)\n / torch.clamp(self.sum_weight, min=0.00001)\n )\n ucb = pred_label + ucb_alpha * pred_sigma\n else:\n pred_sigma = torch.zeros_like(pred_label)\n ucb = pred_label\n return {\n \"pred_label\": pred_label,\n \"pred_sigma\": pred_sigma,\n \"ucb\": ucb,\n \"mlp_out_with_ones\": mlp_out_with_ones,\n }","def __init__(\n self,\n d_model: int,\n nheads: int,\n ff_dim: int,\n bias: bool = False,\n dropout: float = 0.2,\n ):\n super().__init__()\n self.d_model = d_model\n self.nheads = nheads\n self.ff_dim = ff_dim\n self.norm1 = nn.LayerNorm(d_model)\n self.norm2 = nn.LayerNorm(d_model)\n self.self_attention = MultiHeadAttention(d_model, nheads, bias, dropout)\n self.feedforward = PositionwiseFeedforward(d_model, ff_dim)","def forward(self, input):\n label=np.dot(input,self.w)+self.b\n return label","def update_output(self, latent_mat, weight_mat, y_list):","def forward(self, src, src_t, seq, seq_t, seq_e, mask, vars_dict):\n\n src_ext = torch.unsqueeze(src, dim=1) # src [B, 1, D]\n src_e_ph = torch.zeros_like(src_ext)\n q = torch.cat([src_ext, src_e_ph, src_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\n k = torch.cat([seq, seq_e, seq_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\n\n mask = torch.unsqueeze(mask, dim=2) # mask [B, N, 1]\n mask = mask.permute([0, 2, 1]) # mask [B, 1, N]\n\n # # target-attention\n output, attn = self.multi_head_target(q=q, k=k, v=k, vars_dict=vars_dict, mask=mask) # output: [B, 1, D + Dt], attn: [B, 1, N]\n # print('src.shape', src.shape)\n output = output.squeeze(1)\n # print('output.shape', output.shape)\n # print('output', output)\n # print('output.squeeze().shape', output.shape)\n attn = attn.squeeze()\n\n # output = self.merger(output, src)\n x = torch.cat([output, src], dim=1)\n # x = self.layer_norm(x)\n x = F.relu(F.linear(x, vars_dict['w1_agg_fc']))\n output = F.linear(x, vars_dict['w2_agg_fc'])\n\n return output, attn","def pre_forward(self, *args, **kwargs):\n batch_size = args[0].shape[0]\n if not self.instantiated:\n self.hidden_dim = args[0].shape[-1]\n self.instantiate(hidden_dim=self.hidden_dim)\n if self.past_key_reparam is None:\n past_key = self.past_key\n else:\n past_key = self.past_key_reparam\n if self.past_value_reparam is None:\n past_value = self.past_value\n else:\n past_value = self.past_value_reparam\n\n # from IPython import embed\n # embed()\n def expand_batchsize(x):\n x = x.reshape(self.prefix_token_num, self.num_heads, -1).transpose(0,1)\n x = x.unsqueeze(0).expand(batch_size, *x.shape)\n return x\n # from IPython import embe\n\n if 'past_key_value' not in kwargs or kwargs['past_key_value'] is None:\n kwargs['past_key_value'] = (expand_batchsize(past_key), expand_batchsize(past_value))\n\n if 'attention_mask' in kwargs and kwargs['attention_mask'] is not None:\n am = kwargs['attention_mask'] # Should check the format of the attention_mask when moving to a new plm.\n kwargs['attention_mask'] = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\n elif len(args) >1: # attention mask is passed via positional argument\n am = args[1]\n am = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\n args = (args[0], am) + args[2:]\n # from IPython import embed\n # embed(header = \"Herein prefixroberta\")\n return args, kwargs","def tied_featurize(batch, device, chain_dict, fixed_position_dict=None, omit_AA_dict=None, tied_positions_dict=None, pssm_dict=None, bias_by_res_dict=None, ca_only=False):\n alphabet = 'ACDEFGHIKLMNPQRSTVWYX'\n B = len(batch)\n lengths = np.array([len(b['seq']) for b in batch], dtype=np.int32) #sum of chain seq lengths\n L_max = max([len(b['seq']) for b in batch])\n if ca_only:\n X = np.zeros([B, L_max, 1, 3])\n else:\n X = np.zeros([B, L_max, 4, 3])\n residue_idx = -100*np.ones([B, L_max], dtype=np.int32)\n chain_M = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\n pssm_coef_all = np.zeros([B, L_max], dtype=np.float32) #1.0 for the bits that need to be predicted\n pssm_bias_all = np.zeros([B, L_max, 21], dtype=np.float32) #1.0 for the bits that need to be predicted\n pssm_log_odds_all = 10000.0*np.ones([B, L_max, 21], dtype=np.float32) #1.0 for the bits that need to be predicted\n chain_M_pos = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\n bias_by_res_all = np.zeros([B, L_max, 21], dtype=np.float32)\n chain_encoding_all = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\n S = np.zeros([B, L_max], dtype=np.int32)\n omit_AA_mask = np.zeros([B, L_max, len(alphabet)], dtype=np.int32)\n # Build the batch\n letter_list_list = []\n visible_list_list = []\n masked_list_list = []\n masked_chain_length_list_list = []\n tied_pos_list_of_lists_list = []\n for i, b in enumerate(batch):\n if chain_dict != None:\n masked_chains, visible_chains = chain_dict[b['name']] #masked_chains a list of chain letters to predict [A, D, F]\n else:\n masked_chains = [item[-1:] for item in list(b) if item[:10]=='seq_chain_']\n visible_chains = []\n masked_chains.sort() #sort masked_chains \n visible_chains.sort() #sort visible_chains \n all_chains = masked_chains + visible_chains\n for i, b in enumerate(batch):\n mask_dict = {}\n a = 0\n x_chain_list = []\n chain_mask_list = []\n chain_seq_list = []\n chain_encoding_list = []\n c = 1\n letter_list = []\n global_idx_start_list = [0]\n visible_list = []\n masked_list = []\n masked_chain_length_list = []\n fixed_position_mask_list = []\n omit_AA_mask_list = []\n pssm_coef_list = []\n pssm_bias_list = []\n pssm_log_odds_list = []\n bias_by_res_list = []\n l0 = 0\n l1 = 0\n for step, letter in enumerate(all_chains):\n if letter in visible_chains:\n letter_list.append(letter)\n visible_list.append(letter)\n chain_seq = b[f'seq_chain_{letter}']\n chain_seq = ''.join([a if a!='-' else 'X' for a in chain_seq])\n chain_length = len(chain_seq)\n global_idx_start_list.append(global_idx_start_list[-1]+chain_length)\n chain_coords = b[f'coords_chain_{letter}'] #this is a dictionary\n chain_mask = np.zeros(chain_length) #0.0 for visible chains\n if ca_only:\n x_chain = np.array(chain_coords[f'CA_chain_{letter}']) #[chain_lenght,1,3] #CA_diff\n if len(x_chain.shape) == 2:\n x_chain = x_chain[:,None,:]\n else:\n x_chain = np.stack([chain_coords[c] for c in [f'N_chain_{letter}', f'CA_chain_{letter}', f'C_chain_{letter}', f'O_chain_{letter}']], 1) #[chain_lenght,4,3]\n x_chain_list.append(x_chain)\n chain_mask_list.append(chain_mask)\n chain_seq_list.append(chain_seq)\n chain_encoding_list.append(c*np.ones(np.array(chain_mask).shape[0]))\n l1 += chain_length\n residue_idx[i, l0:l1] = 100*(c-1)+np.arange(l0, l1)\n l0 += chain_length\n c+=1\n fixed_position_mask = np.ones(chain_length)\n fixed_position_mask_list.append(fixed_position_mask)\n omit_AA_mask_temp = np.zeros([chain_length, len(alphabet)], np.int32)\n omit_AA_mask_list.append(omit_AA_mask_temp)\n pssm_coef = np.zeros(chain_length)\n pssm_bias = np.zeros([chain_length, 21])\n pssm_log_odds = 10000.0*np.ones([chain_length, 21])\n pssm_coef_list.append(pssm_coef)\n pssm_bias_list.append(pssm_bias)\n pssm_log_odds_list.append(pssm_log_odds)\n bias_by_res_list.append(np.zeros([chain_length, 21]))\n if letter in masked_chains:\n masked_list.append(letter)\n letter_list.append(letter)\n chain_seq = b[f'seq_chain_{letter}']\n chain_seq = ''.join([a if a!='-' else 'X' for a in chain_seq])\n chain_length = len(chain_seq)\n global_idx_start_list.append(global_idx_start_list[-1]+chain_length)\n masked_chain_length_list.append(chain_length)\n chain_coords = b[f'coords_chain_{letter}'] #this is a dictionary\n chain_mask = np.ones(chain_length) #1.0 for masked\n if ca_only:\n x_chain = np.array(chain_coords[f'CA_chain_{letter}']) #[chain_lenght,1,3] #CA_diff\n if len(x_chain.shape) == 2:\n x_chain = x_chain[:,None,:]\n else:\n x_chain = np.stack([chain_coords[c] for c in [f'N_chain_{letter}', f'CA_chain_{letter}', f'C_chain_{letter}', f'O_chain_{letter}']], 1) #[chain_lenght,4,3] \n x_chain_list.append(x_chain)\n chain_mask_list.append(chain_mask)\n chain_seq_list.append(chain_seq)\n chain_encoding_list.append(c*np.ones(np.array(chain_mask).shape[0]))\n l1 += chain_length\n residue_idx[i, l0:l1] = 100*(c-1)+np.arange(l0, l1)\n l0 += chain_length\n c+=1\n fixed_position_mask = np.ones(chain_length)\n if fixed_position_dict!=None:\n fixed_pos_list = fixed_position_dict[b['name']][letter]\n if fixed_pos_list:\n fixed_position_mask[np.array(fixed_pos_list)-1] = 0.0\n fixed_position_mask_list.append(fixed_position_mask)\n omit_AA_mask_temp = np.zeros([chain_length, len(alphabet)], np.int32)\n if omit_AA_dict!=None:\n for item in omit_AA_dict[b['name']][letter]:\n idx_AA = np.array(item[0])-1\n AA_idx = np.array([np.argwhere(np.array(list(alphabet))== AA)[0][0] for AA in item[1]]).repeat(idx_AA.shape[0])\n idx_ = np.array([[a, b] for a in idx_AA for b in AA_idx])\n omit_AA_mask_temp[idx_[:,0], idx_[:,1]] = 1\n omit_AA_mask_list.append(omit_AA_mask_temp)\n pssm_coef = np.zeros(chain_length)\n pssm_bias = np.zeros([chain_length, 21])\n pssm_log_odds = 10000.0*np.ones([chain_length, 21])\n if pssm_dict:\n if pssm_dict[b['name']][letter]:\n pssm_coef = pssm_dict[b['name']][letter]['pssm_coef']\n pssm_bias = pssm_dict[b['name']][letter]['pssm_bias']\n pssm_log_odds = pssm_dict[b['name']][letter]['pssm_log_odds']\n pssm_coef_list.append(pssm_coef)\n pssm_bias_list.append(pssm_bias)\n pssm_log_odds_list.append(pssm_log_odds)\n if bias_by_res_dict:\n bias_by_res_list.append(bias_by_res_dict[b['name']][letter])\n else:\n bias_by_res_list.append(np.zeros([chain_length, 21]))\n\n \n letter_list_np = np.array(letter_list)\n tied_pos_list_of_lists = []\n tied_beta = np.ones(L_max)\n if tied_positions_dict!=None:\n tied_pos_list = tied_positions_dict[b['name']]\n if tied_pos_list:\n set_chains_tied = set(list(itertools.chain(*[list(item) for item in tied_pos_list])))\n for tied_item in tied_pos_list:\n one_list = []\n for k, v in tied_item.items():\n start_idx = global_idx_start_list[np.argwhere(letter_list_np == k)[0][0]]\n if isinstance(v[0], list):\n for v_count in range(len(v[0])):\n one_list.append(start_idx+v[0][v_count]-1)#make 0 to be the first\n tied_beta[start_idx+v[0][v_count]-1] = v[1][v_count]\n else:\n for v_ in v:\n one_list.append(start_idx+v_-1)#make 0 to be the first\n tied_pos_list_of_lists.append(one_list)\n tied_pos_list_of_lists_list.append(tied_pos_list_of_lists)\n\n\n \n x = np.concatenate(x_chain_list,0) #[L, 4, 3]\n all_sequence = \"\".join(chain_seq_list)\n m = np.concatenate(chain_mask_list,0) #[L,], 1.0 for places that need to be predicted\n chain_encoding = np.concatenate(chain_encoding_list,0)\n m_pos = np.concatenate(fixed_position_mask_list,0) #[L,], 1.0 for places that need to be predicted\n\n pssm_coef_ = np.concatenate(pssm_coef_list,0) #[L,], 1.0 for places that need to be predicted\n pssm_bias_ = np.concatenate(pssm_bias_list,0) #[L,], 1.0 for places that need to be predicted\n pssm_log_odds_ = np.concatenate(pssm_log_odds_list,0) #[L,], 1.0 for places that need to be predicted\n\n bias_by_res_ = np.concatenate(bias_by_res_list, 0) #[L,21], 0.0 for places where AA frequencies don't need to be tweaked\n\n l = len(all_sequence)\n x_pad = np.pad(x, [[0,L_max-l], [0,0], [0,0]], 'constant', constant_values=(np.nan, ))\n X[i,:,:,:] = x_pad\n\n m_pad = np.pad(m, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\n m_pos_pad = np.pad(m_pos, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\n omit_AA_mask_pad = np.pad(np.concatenate(omit_AA_mask_list,0), [[0,L_max-l]], 'constant', constant_values=(0.0, ))\n chain_M[i,:] = m_pad\n chain_M_pos[i,:] = m_pos_pad\n omit_AA_mask[i,] = omit_AA_mask_pad\n\n chain_encoding_pad = np.pad(chain_encoding, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\n chain_encoding_all[i,:] = chain_encoding_pad\n\n pssm_coef_pad = np.pad(pssm_coef_, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\n pssm_bias_pad = np.pad(pssm_bias_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\n pssm_log_odds_pad = np.pad(pssm_log_odds_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\n\n pssm_coef_all[i,:] = pssm_coef_pad\n pssm_bias_all[i,:] = pssm_bias_pad\n pssm_log_odds_all[i,:] = pssm_log_odds_pad\n\n bias_by_res_pad = np.pad(bias_by_res_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\n bias_by_res_all[i,:] = bias_by_res_pad\n\n # Convert to labels\n indices = np.asarray([alphabet.index(a) for a in all_sequence], dtype=np.int32)\n S[i, :l] = indices\n letter_list_list.append(letter_list)\n visible_list_list.append(visible_list)\n masked_list_list.append(masked_list)\n masked_chain_length_list_list.append(masked_chain_length_list)\n\n\n isnan = np.isnan(X)\n mask = np.isfinite(np.sum(X,(2,3))).astype(np.float32)\n X[isnan] = 0.\n\n # Conversion\n pssm_coef_all = torch.from_numpy(pssm_coef_all).to(dtype=torch.float32, device=device)\n pssm_bias_all = torch.from_numpy(pssm_bias_all).to(dtype=torch.float32, device=device)\n pssm_log_odds_all = torch.from_numpy(pssm_log_odds_all).to(dtype=torch.float32, device=device)\n\n tied_beta = torch.from_numpy(tied_beta).to(dtype=torch.float32, device=device)\n\n jumps = ((residue_idx[:,1:]-residue_idx[:,:-1])==1).astype(np.float32)\n bias_by_res_all = torch.from_numpy(bias_by_res_all).to(dtype=torch.float32, device=device)\n phi_mask = np.pad(jumps, [[0,0],[1,0]])\n psi_mask = np.pad(jumps, [[0,0],[0,1]])\n omega_mask = np.pad(jumps, [[0,0],[0,1]])\n dihedral_mask = np.concatenate([phi_mask[:,:,None], psi_mask[:,:,None], omega_mask[:,:,None]], -1) #[B,L,3]\n dihedral_mask = torch.from_numpy(dihedral_mask).to(dtype=torch.float32, device=device)\n residue_idx = torch.from_numpy(residue_idx).to(dtype=torch.long,device=device)\n S = torch.from_numpy(S).to(dtype=torch.long,device=device)\n X = torch.from_numpy(X).to(dtype=torch.float32, device=device)\n mask = torch.from_numpy(mask).to(dtype=torch.float32, device=device)\n chain_M = torch.from_numpy(chain_M).to(dtype=torch.float32, device=device)\n chain_M_pos = torch.from_numpy(chain_M_pos).to(dtype=torch.float32, device=device)\n omit_AA_mask = torch.from_numpy(omit_AA_mask).to(dtype=torch.float32, device=device)\n chain_encoding_all = torch.from_numpy(chain_encoding_all).to(dtype=torch.long, device=device)\n if ca_only:\n X_out = X[:,:,0]\n else:\n X_out = X\n return X_out, S, mask, lengths, chain_M, chain_encoding_all, letter_list_list, visible_list_list, masked_list_list, masked_chain_length_list_list, chain_M_pos, omit_AA_mask, residue_idx, dihedral_mask, tied_pos_list_of_lists_list, pssm_coef_all, pssm_bias_all, pssm_log_odds_all, bias_by_res_all, tied_beta","def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, \n mlm_label=None, dvae_imgs=None, v_token_mask=None, hog_features=None, img_metas=None, **kwargs): \n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \n if self.from_scratch:\n imgs = imgs / 255.0\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\n text_input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\n if mlm_label is not None:\n mlm_label = mlm_label.reshape((-1, ) + mlm_label.shape[2:])\n\n\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\n\n B, D, T, H, W = visual_token.shape\n losses = dict()\n # -------------- nce loss ------------------- #\n if hasattr(self, 'ssl_head'):\n input_ssl_ids = torch.where(mlm_label == -100, token_ids.clone(), mlm_label.clone())\n input_ssl_mask = text_input_mask.clone()\n text_only_out = self.text_backbone(input_ssl_ids, input_ssl_mask)\n # ------------ complete T -------------- #\n text_out_no_mask = text_only_out['last_hidden_state']\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_no_mask, input_ssl_mask, input_ssl_ids)\n\n\n # ------------ complete V ---------------- #\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\n\n\n # ---------- foward mask text input ---------- # \n text_out_with_mask = self.text_backbone(token_ids, text_input_mask)\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\n\n # ---------- forward mask v input ------------ #\n visual_token_with_mask, v_mask = self.extract_visual_feat(imgs.clone(), v_token_mask) # b, d, T, h, w\n visual_token_mask = visual_token_with_mask.view(B, D, T, -1).permute(0, 2, 3, 1)\n \n v_fusion_output = self.multimodal_backbone(visual_token=visual_token_mask, text_input_mask=text_input_mask.clone(), text_input_embeds=text_out_no_mask.clone())\n \n t_fusion_output = self.multimodal_backbone(visual_token=visual_token, text_input_mask=text_input_mask, text_input_embeds=text_out_last_hidden_state)\n # for mlm #\n t_last_hidden_state = t_fusion_output['t_last_hidden_state']\n\n\n\n\n\n # ------------ MLM loss ------------ #\n\n if mlm_label is not None and self.mlm_head is not None:\n # we use mask text for MLM\n # because we doubt there will be miss interaction between wrong img-text pair \n # and the model not learn good relationship between vision and language\n # -------- forward masked text ----------- #\n mlm_prediction_score = self.mlm_head(t_last_hidden_state)\n \n if self.mlm_loss_func is not None:\n mlm_label_idx = torch.where(mlm_label.view(-1) != -100)\n mlm_prediction_mask_score = mlm_prediction_score.view(-1, self.text_vocab_size)[mlm_label_idx[0], :]\n mlm_label_mask = mlm_label.view(-1)[mlm_label_idx]\n mlm_loss = self.mlm_loss_func(mlm_prediction_mask_score, mlm_label_mask)\n else:\n mlm_loss = self.loss_func(mlm_prediction_score.view(-1, self.text_vocab_size), mlm_label.view(-1))\n losses['mlm_loss'] = mlm_loss\n\n\n # ------- Tri-modal alignment with mask sample and ranking --------- #\n if self.mlm_ssl_V_head is not None:\n mlm_visual_feat = v_fusion_output['t_last_hidden_state'][:, 0]\n mask_visual_recon_emb = self.mlm_ssl_V_head(mlm_visual_feat)\n mask_word_emb = self.ssl_head.forward_text(text_out_last_hidden_state) if self.use_Cmask else None\n loss_cvt_rank = self.ssl_loss(visual_emb, text_emb, mask_word_emb, mask_visual_recon_emb)\n losses.update(loss_cvt_rank)\n\n\n if self.symmetry_rank:\n mlm_word_feat = t_last_hidden_state[:, 0]\n mask_word_recon_emb = self.mlm_ssl_T_head(mlm_word_feat)\n\n mask_visual_emb = self.ssl_head.forward_vision(visual_token_with_mask) if self.use_Cmask else None\n \n loss_ctv_rank = self.ssl_loss(text_emb, visual_emb, mask_visual_emb, mask_word_recon_emb)\n loss_ctv_rank['v_nce_loss'] = loss_ctv_rank.pop('nce_loss')\n \n if self.ssl_loss.use_rank:\n loss_ctv_rank['rank_v_vm_loss'] = loss_ctv_rank.pop('rank_t_tm_loss')\n\n \n\n losses.update(loss_ctv_rank)\n\n\n\n return losses","def __init__(self, dm, h, hidden, drop_rate=0.1):\n super().__init__()\n self.mha = MultiHeadAttention(dm, h)\n self.ffn = point_wise_feed_forward_network(dm, hidden)\n\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)","def classfier_head_multi_proposal(self, statement, statement_mask, targets, ts_labels, ts_labels_mask, max_num_proposal=1, ce_prob_thd=0.01, iou_thd=0.5, extra_span_length=3):\n bsz, num_a, num_img, num_words = statement_mask.shape\n statement = statement.view(bsz * num_a * num_img, num_words, -1)\n statement_mask = statement_mask.view(bsz * num_a * num_img, num_words)\n statement = self.cls_encoder(statement, statement_mask)\n max_statement = torch.max(mask_logits(statement, statement_mask.unsqueeze(2)), 1)[0]\n max_statement_mask = (statement_mask.sum(1) != 0).float().view(bsz, num_a, num_img, 1)\n max_statement = max_statement.view(bsz * num_a, num_img, -1)\n t_score_container = []\n encoded_max_statement_container = []\n encoded_max_statement = max_statement\n for layer_idx in range(self.t_iter + 1):\n encoded_max_statement, prev_t_score = self.residual_temporal_predictor(layer_idx, encoded_max_statement)\n t_score_container.append(prev_t_score.view(bsz, num_a, num_img, 2))\n encoded_max_statement_container.append(encoded_max_statement)\n if self.t_iter > 0:\n temporal_scores_st_ed = 0.5 * (t_score_container[0] + torch.stack(t_score_container[:1]).mean(0))\n else:\n temporal_scores_st_ed = t_score_container[0]\n temporal_scores_st_ed = mask_logits(temporal_scores_st_ed, ts_labels_mask.view(bsz, 1, num_img, 1))\n stacked_max_statement = encoded_max_statement_container[0].view(bsz, num_a, num_img, -1)\n if self.add_local:\n max_max_statement, targets = self.get_proposals(stacked_max_statement, max_statement_mask, temporal_scores_st_ed, targets, ts_labels, max_num_proposal=max_num_proposal, iou_thd=iou_thd, ce_prob_thd=ce_prob_thd, extra_span_length=extra_span_length)\n else:\n max_max_statement = torch.max(mask_logits(stacked_max_statement, max_statement_mask), 2)[0]\n answer_scores = self.classifier(max_max_statement).squeeze(2)\n return answer_scores, targets, temporal_scores_st_ed","def __init__(self,\n num_units,\n line_memory,\n word_memory=None,\n soft_weight=None,\n hierarchy=True,\n line_memory_sequence_length=None,\n word_memory_sequence_length=None,\n scale=False,\n probability_fn=None,\n score_mask_value=float(\"-inf\"),\n name=\"CustomAttention\"):\n # For LuongAttention, we only transform the memory layer; thus\n # num_units **must** match expected the query depth.\n if probability_fn is None:\n probability_fn = nn_ops.softmax\n wrapped_probability_fn = lambda score: probability_fn(score)\n super(CustomAttention, self).__init__(\n query_layer=None,\n line_memory_layer=layers_core.Dense(\n num_units, name=\"line_memory_layer\", use_bias=False),\n line_memory=line_memory,\n word_memory=word_memory,\n probability_fn=wrapped_probability_fn,\n line_memory_sequence_length=line_memory_sequence_length,\n word_memory_sequence_length=word_memory_sequence_length,\n score_mask_value=score_mask_value,\n name=name)\n self._num_units = num_units\n self._scale = scale\n self._name = name\n self._hierarchy = hierarchy\n self._soft_weight = soft_weight","def _InitAttentionParams(self, atten_tpl):\n p = self.params\n source_atten_tpls = []\n # Set up each source attention.\n for i in range(p.num_source):\n src_key = 'source_%d' % i\n src_atten = atten_tpl.Copy()\n src_atten = super()._InitAttentionParams(src_atten)\n if isinstance(src_atten, list):\n raise ValueError(\n 'TransformerMultiSourceAttentionLayer does not support '\n 'num_heads > 1.')\n src_atten.name = 'multihead_atten_%s' % src_key\n source_atten_tpls.append((src_key, src_atten))\n\n # Initialize multi-source attention.\n msa = p.multi_source_atten.Copy()\n msa.name = 'multi_source_atten'\n msa.input_dim = p.input_dim\n msa.hidden_dim = p.hidden_dim\n msa.source_atten_tpls = source_atten_tpls\n msa.primary_source_key = 'source_%d' % p.primary_source_index\n return msa","def multi_label_cls_head__post_process(ctx, self, pred, **kwargs):\n return pred","def __init__(self, bert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.q_lin.weight, bert_layer.attention.k_lin.weight, bert_layer.attention.v_lin.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.q_lin.bias, bert_layer.attention.k_lin.bias, bert_layer.attention.v_lin.bias]))\n self.out_proj_weight = bert_layer.attention.out_lin.weight\n self.out_proj_bias = bert_layer.attention.out_lin.bias\n self.linear1_weight = bert_layer.ffn.lin1.weight\n self.linear1_bias = bert_layer.ffn.lin1.bias\n self.linear2_weight = bert_layer.ffn.lin2.weight\n self.linear2_bias = bert_layer.ffn.lin2.bias\n self.norm1_eps = bert_layer.sa_layer_norm.eps\n self.norm1_weight = bert_layer.sa_layer_norm.weight\n self.norm1_bias = bert_layer.sa_layer_norm.bias\n self.norm2_eps = bert_layer.output_layer_norm.eps\n self.norm2_weight = bert_layer.output_layer_norm.weight\n self.norm2_bias = bert_layer.output_layer_norm.bias\n self.num_heads = bert_layer.attention.n_heads\n self.embed_dim = bert_layer.attention.dim\n self.is_last_layer = False\n self.validate_bettertransformer()","def forward(self, h_text, mask=None):\n x = h_text\n for conv_layer in self.conv_stack:\n x = conv_layer(x, mask=mask)\n\n # Autoregressive Prediction\n p_0 = self.init_state(x)\n\n outputs = [p_0]\n for i in range(x.shape[1]):\n p_input = torch.cat((x[:, i], outputs[-1]), dim=-1) # [B, 3 * d_model]\n self.gru_hidden = self.gru_cell(p_input, self.gru_hidden) # [B, 2 * d_model]\n outputs.append(self.gru_hidden)\n outputs = torch.stack(outputs[1:], dim=1) # [B, src_len, 2 * d_model]\n\n # GMM-MDN\n w, sigma, mu = self.gmm_mdn(outputs)\n if mask is not None:\n w = w.masked_fill(mask.unsqueeze(-1), 0 if self.training else 1e-9) # 1e-9 for categorical sampling\n sigma = sigma.masked_fill(mask.unsqueeze(-1).unsqueeze(-1), 0)\n mu = mu.masked_fill(mask.unsqueeze(-1).unsqueeze(-1), 0)\n\n return w, sigma, mu","def forward(self, *args): # noqa: R0914\r\n encoder_out, (hn, cn) = self.unified_encoder(*args)\r\n device = hn.device\r\n non_sequential_cont_decoded = self.mlp_non_seq_cont(hn)\r\n non_sequential_cat_decoded = []\r\n for mlp_non_seq_cat in self.mlp_non_seq_cat_list:\r\n non_sequential_cat_decoded.append(mlp_non_seq_cat(hn))\r\n\r\n hn = torch.unsqueeze(hn, 0)\r\n cn = torch.unsqueeze(cn, 0)\r\n # decoded is the output prediction of timestep i-1 of the decoder\r\n decoded = torch.zeros(encoder_out.shape[0], int(\r\n self.unified_encoder.seq_cont_count + self.unified_encoder.no_of_embs_seq), device=device)\r\n seq_cont_decoded = torch.Tensor(device=device)\r\n seq_cat_decoded = []\r\n for _ in range(self.unified_encoder.seq_cat_count):\r\n seq_cat_decoded.append(torch.Tensor(device=device))\r\n\r\n for _ in range(encoder_out.shape[1]):\r\n decoded, (hn, cn), out_cont, out_cat = self.decoder(decoded, (hn, cn))\r\n # Predict all categorical columns\r\n out_cat_onehot = []\r\n if self.unified_encoder.seq_cat_count != 0:\r\n for idx, out in enumerate(out_cat):\r\n out_cat_onehot.append(torch.argmax(out, dim=1).unsqueeze(-1))\r\n seq_cat_decoded[idx] = torch.cat(\r\n [seq_cat_decoded[idx], out.view(out.shape[0], 1, -1)], dim=1)\r\n out_cat_onehot = torch.cat(out_cat_onehot, -1)\r\n out_cat_embedding = self.unified_encoder.seq_emb_layers(out_cat_onehot)\r\n decoded = torch.cat([out_cat_embedding, out_cont], dim=-1)\r\n else:\r\n decoded = out_cont\r\n seq_cont_decoded = torch.cat(\r\n [seq_cont_decoded, out_cont.view(out_cont.shape[0], 1, -1)], dim=1)\r\n\r\n return non_sequential_cont_decoded, non_sequential_cat_decoded, seq_cont_decoded, seq_cat_decoded","def __init__(self,\n num_heads,\n seq_len,\n block=16,\n different_layout_per_head=False,\n num_local_blocks=4,\n num_global_blocks=1,\n attention='bidirectional',\n horizental_global_attention=False,\n num_differnt_global_patterns=1):\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\n\n if (seq_len % (block * num_local_blocks) != 0):\n raise ValueError(\n f'Sequence Length, {seq_len}, must be dividable by local window, {block*num_local_blocks}!'\n )\n self.num_local_blocks = num_local_blocks\n\n if (num_local_blocks % num_global_blocks != 0):\n raise ValueError(\n f'Number of blocks in a local window, {num_local_blocks}, must be dividable by number of global blocks, {num_global_blocks}!'\n )\n self.num_global_blocks = num_global_blocks\n\n if (attention != 'unidirectional' and attention != 'bidirectional'):\n raise NotImplementedError(\n 'only \\\"uni/bi-directional\\\" attentions are supported for now!')\n self.attention = attention\n\n if (attention != 'bidirectional' and horizental_global_attention):\n raise ValueError(\n 'only \\\"bi-directional\\\" attentions can support horizental global attention!'\n )\n self.horizental_global_attention = horizental_global_attention\n\n if (num_differnt_global_patterns > 1 and not different_layout_per_head):\n raise ValueError(\n f'Number of different layouts cannot be more than one when you have set a single layout for all heads! Set different_layout_per_head to True.'\n )\n if (num_differnt_global_patterns > (num_local_blocks // num_global_blocks)):\n raise ValueError(\n f'Number of layout versions (num_differnt_global_patterns), {num_differnt_global_patterns}, cannot be larger than number of local window blocks divided by number of global blocks, {num_local_blocks} / {num_global_blocks} = {num_local_blocks//num_global_blocks}!'\n )\n self.num_differnt_global_patterns = num_differnt_global_patterns\n self.make_layout()","def create_model(self, model_input, vocab_size, num_frames, **unused_params):\n num_frames_t=num_frames\n num_frames = tf.cast(tf.expand_dims(num_frames, 1), tf.float32)\n feature_size = model_input.get_shape().as_list()[2]\n iterations=5#150\n attention_size=8\n if FLAGS.is_train: \n iterations=120\n model_input = utils.SampleRandomFrames(model_input[:,15:,:], num_frames-15-15,\n iterations)\n model_input=model_input+tf.random_normal(shape=tf.shape(model_input), mean=0.0, stddev=1e-3, dtype=tf.float32)\n\n aggregated_model = getattr(video_level_models,\n FLAGS.video_level_classifier_model)\n\n video_attention = MultiAttentionLayers(1024,iterations,256,attention_size)#256\n audio_attention = MultiAttentionLayers(128,iterations,256/4,attention_size)#256/4\n\n model_input = slim.batch_norm(\n model_input,\n center=True,\n scale=True,\n is_training=True,\n scope=\"model_input_bn\")\n\n with tf.variable_scope(\"video_Attention\"):\n attention_video = video_attention.forward(model_input[:,:,0:1024]) \n with tf.variable_scope(\"audio_Attention\"):\n attention_audio = audio_attention.forward(model_input[:,:,1024:])\n\n pooled=tf.concat([attention_video,attention_audio],axis=1)\n #instance_att#tf.reduce_mean(pooledi,axis=1)\n\n print('pooled is',pooled)\n pooled=tf.reshape(tf.transpose(pooled,perm=[0,2,1]),[-1,1152])\n dr2 = tf.get_variable(\"dr2\",\n [feature_size,1024],\n initializer = tf.random_normal_initializer(stddev=1 / math.sqrt(feature_size)))\n pooled=tf.matmul(pooled,dr2)\n\n pooled = slim.batch_norm(\n pooled,\n center=True,\n scale=True,\n is_training=True,\n scope=\"pooled_bn\")\n\n gating_weights = tf.get_variable(\"gating_weights_2\",\n [1024, 1024],\n initializer = tf.random_normal_initializer(stddev=1 / math.sqrt(1024))) \n gates = tf.matmul(pooled, gating_weights) \n gates = slim.batch_norm(\n gates,\n center=True,\n scale=True,\n is_training=True,\n scope=\"gating_bn\")\n gates = tf.sigmoid(gates)\n pooled = tf.multiply(pooled,gates)\n\n results_temp=aggregated_model().create_model(\n model_input=pooled, vocab_size=vocab_size, **unused_params)\n results_temp['predictions']=tf.reduce_max(tf.reshape(results_temp['predictions'],[-1,attention_size,vocab_size]),axis=1)\n print(results_temp)\n return results_temp","def __init__(self, args, number_of_labels, number_of_features,adj):\n super(SpGAT, self).__init__()\n self.args=args\n \n self.number_of_labels = number_of_labels\n self.number_of_features = number_of_features\n self.device = args.device\n self.adj= sparse_mx_to_torch_sparse_tensor(adj).to(self.device).to_dense()\n self.attentions = [SpGraphAttentionLayer(number_of_features, \n args.hidden, \n dropout=args.dropout, \n alpha=args.alpha, \n concat=True) for _ in range(args.nheads)]\n for i, attention in enumerate(self.attentions):\n self.add_module('attention_{}'.format(i), attention)\n\n self.out_att = SpGraphAttentionLayer(args.hidden * args.nheads, \n args.Q, \n dropout=args.dropout, \n alpha=args.alpha, \n concat=False)","def __init__(self, # noqa: R0913\n hid_dim,\n n_heads,\n pf_dim,\n dropout):\n\n super().__init__()\n\n self.self_attn_layer_norm = nn.LayerNorm(hid_dim)\n self.ff_layer_norm = nn.LayerNorm(hid_dim)\n self.self_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\n self.positionwise_feedforward = PositionwiseFeedforwardLayer(hid_dim,\n pf_dim,\n dropout)\n self.dropout = nn.Dropout(dropout)","def mtf_transformer2_base():\n hparams = common_hparams.basic_params1()\n\n hparams.add_hparam(\"d_model\", 1024)\n hparams.batch_size = 4\n hparams.max_length = 1024\n hparams.label_smoothing = 0.0\n # a small positive value - this seems important for stability when training\n # with bfloat16 activations.\n hparams.add_hparam(\"z_loss\", 1e-4)\n\n # hparams applying to both encoder and decoder layer stacks.\n hparams.add_hparam(\"d_ff\", 2048)\n hparams.add_hparam(\"d_kv\", 128)\n hparams.add_hparam(\"attention_dropout\", 0.0)\n hparams.add_hparam(\"relu_dropout\", 0.0)\n hparams.del_hparam(\"num_heads\")\n hparams.del_hparam(\"num_hidden_layers\")\n hparams.layer_prepostprocess_dropout = 0.0\n hparams.add_hparam(\"extra_logit\", False)\n # number of experts for moe_1d\n hparams.moe_num_experts = 32\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\n hparams.add_hparam(\"moe_expert_x\", 8)\n hparams.add_hparam(\"moe_expert_y\", 4)\n hparams.add_hparam(\"moe_hidden_size\", 32768)\n\n # round up vocab sizes to be a multiple of this value\n hparams.vocab_divisor = 128\n\n hparams.optimizer = \"Adafactor\"\n hparams.learning_rate_schedule = \"rsqrt_decay*linear_decay\"\n hparams.learning_rate_warmup_steps = 10000\n hparams.add_hparam(\"master_dtype\", \"bfloat16\")\n hparams.add_hparam(\"slice_dtype\", \"float32\")\n hparams.activation_dtype = \"bfloat16\"\n\n # 8-way model-parallelism\n hparams.add_hparam(\"mesh_shape\", \"model:8\")\n hparams.add_hparam(\"layout\", \"batch:batch;vocab:model;d_ff:model;heads:model\")\n\n # If nonzero, we split the batch across two tensor-dimensions named\n # \"outer_batch\" and \"inner_batch\", allowing for splitting across two mesh\n # dimensions. This is necessary for hierarchical mixture of experts.\n # The two tensor dimensions have sizes hparams.outer_batch_size and\n # hparams.batch_size // hparams.outer_batch_size.\n hparams.add_hparam(\"outer_batch_size\", 0)\n\n hparams.shared_embedding_and_softmax_weights = False\n # length for training or decoding - defaults to max_length\n hparams.add_hparam(\"length\", 0)\n\n # These parameters make Transformer model compatible with mtf\n # Do not override these.\n hparams.no_data_parallelism = True\n hparams.use_fixed_batch_size = True\n hparams.add_hparam(\"mtf_mode\", True)\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\n hparams.bottom = {\n \"inputs\": modalities.identity_bottom,\n \"targets\": modalities.identity_bottom,\n }\n hparams.top = {\n \"targets\": modalities.identity_top,\n }\n hparams.add_hparam(\"beam_size\", 1)\n\n # If this is True, then in a packed dataset (where exaples are concatenated\n # to form longer examples) we use the global position (within the concatenated\n # sequence) to compute the positional embedding, instead of the position\n # within the individual sequence. This is counterintuitive, but for some\n # reason, it keeps the model from diverging.\n hparams.add_hparam(\"use_global_position_in_packed_sequence\", True)\n\n return hparams","def forward(self, input, targets=None, get_head_out=False):\n fpn_out = self.fpn(input)\n\n img_height, img_width = input.shape[2:]\n strides = self.fpn.strides\n hws = [level_out.shape[2:] for level_out in fpn_out]\n max_box_sides = [256, 128, 64, 32, 16]\n if self.levels is not None:\n max_box_sides = [max_box_sides[l] for l in self.levels]\n self.pyramid_shape = [\n (s, m, h, w) for s, m, (h, w) in zip(strides, max_box_sides, hws)]\n\n head_out = []\n for i, level_out in enumerate(fpn_out):\n head_out.append(self.head(level_out, self.scale_params[i]))\n\n if targets is None:\n out = decode_batch_output(head_out, self.pyramid_shape, img_height, img_width)\n if get_head_out:\n return out, head_out\n return out\n\n loss_dict = fcos_batch_loss(\n head_out, targets, self.pyramid_shape, self.num_labels)\n if get_head_out:\n return loss_dict, head_out\n return loss_dict","def __init__(self, nfeat, nhid, nclass, dropout, alpha, nheads, direct=False, supervised=True):\n super(SpGAT, self).__init__()\n self.dropout = dropout\n\n self.attentions = [SpGraphAttentionLayer(nfeat,\n nhid,\n dropout=dropout,\n alpha=alpha,\n concat=True) for _ in range(nheads)]\n for i, attention in enumerate(self.attentions):\n self.add_module('attention_{}'.format(i), attention)\n\n self.out_att = SpGraphAttentionLayer(nhid * nheads,\n nhid,\n dropout=dropout,\n alpha=alpha,\n concat=False)\n self.fm = FMLayer(nfeat, nhid)\n self.supervised = supervised\n if self.supervised:\n n_output = nclass\n else:\n n_output = nhid\n\n self.final_linear = nn.Linear(2 * nhid, n_output)\n self.final_linear_single = nn.Linear(nhid, n_output)\n # self.final_linear = nn.Linear(nhid, nclass)\n self.pa_attn = AttentionalFactorizationMachine(nhid, nhid, dropouts=[0.2, 0.2])\n self.direct = direct","def forward(self, text, prev_mel):\n # forward pass through text embedding and get k and v\n kv = self.t_encoder(text)\n k = kv[:,:self.hp.d,:]\n v = kv[:,self.hp.d:,:]\n # forward pass through audio encoding and get Q\n q = self.a_encoder(prev_mel)\n \n # compute attention\n a = (k.transpose(2,1)).matmul(q)/np.sqrt(self.hp.d)\n a = F.softmax(a, dim=1)\n r = v.matmul(a)\n \n # create R' and forward pass through decoder\n # note that the decoder does not have sigmoid transform at the end, so we are actually getting \n # ylogit\n rprime = torch.cat((r, q), dim=1)\n ylogit = self.decoder(rprime)\n y = F.sigmoid(ylogit)\n return y, ylogit, a","def mt_step(self, lang1, lang2, lambda_coeff):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n lang1_id = params.lang2id[lang1]\n lang2_id = params.lang2id[lang2]\n\n # generate batch\n if lang1 == lang2:\n (x1, len1) = self.get_batch('ae', lang1)\n (x2, len2) = (x1, len1)\n (x1, len1) = self.add_noise(x1, len1)\n else:\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\n langs1 = x1.clone().fill_(lang1_id)\n langs2 = x2.clone().fill_(lang2_id)\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n # key point generate label\n\n # cuda\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\n\n # encode source sentence\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\n enc1 = enc1.transpose(0, 1)\n\n # decode target sentence\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\n src_len=len1)\n\n # loss\n # the last word not apply prediction logic\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def __init__(self, bert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.self.query.weight, bert_layer.attention.self.key.weight, bert_layer.attention.self.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.self.query.bias, bert_layer.attention.self.key.bias, bert_layer.attention.self.value.bias]))\n self.out_proj_weight = bert_layer.attention.output.dense.weight\n self.out_proj_bias = bert_layer.attention.output.dense.bias\n self.linear1_weight = bert_layer.intermediate.dense.weight\n self.linear1_bias = bert_layer.intermediate.dense.bias\n self.linear2_weight = bert_layer.output.dense.weight\n self.linear2_bias = bert_layer.output.dense.bias\n self.norm1_eps = bert_layer.attention.output.LayerNorm.eps\n self.norm1_weight = bert_layer.attention.output.LayerNorm.weight\n self.norm1_bias = bert_layer.attention.output.LayerNorm.bias\n self.norm2_eps = bert_layer.output.LayerNorm.eps\n self.norm2_weight = bert_layer.output.LayerNorm.weight\n self.norm2_bias = bert_layer.output.LayerNorm.bias\n self.num_heads = bert_layer.attention.self.num_attention_heads\n self.embed_dim = bert_layer.attention.self.all_head_size\n self.is_last_layer = False\n self.validate_bettertransformer()","def get_output_for(self, input, **kwargs):\n\t\tbatch_size = input.shape[0]\n\t\tnum_states = self.num_states\n\t\timg_dim = self.img_dim\n\t\tN = self.N\n\t\tattender = self.attender\n\t\ttheano.gradient.grad_clip\n\n\t\tdef step(c_tm1, h_tm1, att_acc_tm1, input, W, Wg):\n\t\t\tcenter_y, center_x, delta, sigma, gamma = gp_from_hidden(h_tm1, Wg, img_dim, N)\n\t\t\tg, att = attender.read(input, center_y, center_x, delta, sigma, gamma) # (batch_size, N, N) and (batch_size, img_dim, img_dim)\n\t\t\n\t\t\tatt_acc_t = T.clip(att_acc_tm1 + att, 0.0, 1.0)\t# (batch_size, img_dim, img_dim)\n\t\t\tr = input[:, :, :img_dim] * (1.0 - att_acc_t) # (batch_size, img_dim, img_dim)\n\t\t\tR , _ = attender.read(r, *gp_from_hidden(T.zeros((batch_size, 5)), T.eye(5), img_dim, N)) # (batch_size, N, N)\n\t\t\t\n\t\t\tflat_g = g.reshape((batch_size, N * N)) # (batch_size, N * N)\n\t\t\tflat_R = R.reshape((batch_size, N * N)) # (batch_size, N * N)\n\t\t\t\n\t\t\t# concatenate gA, gB and h_tm1 to form a single matrix # (batch_size, N * N + N * N + num_states + 1)\n\t\t\tlstm_inp = T.concatenate([flat_g, flat_R, h_tm1, T.ones((batch_size, 1))], axis=1)\n\n\t\t\t# multiply by LSTM weights\n\t\t\t# (num_states * 4, num_input + num_states + 1) dot (batch_size, N * N + N * N + num_states + 1).T\n\t\t\tpre_act = T.dot(W, lstm_inp.T) \t# (4 * num_states, batch_size)\n\n\t\t\t# split up to get individual gates\n\t\t\tz = T.tanh(pre_act[0*num_states:1*num_states]) # (num_states, batch_size)\n\t\t\ti = T.nnet.sigmoid(pre_act[1*num_states:2*num_states])\n\t\t\tf = T.nnet.sigmoid(pre_act[2*num_states:3*num_states])\n\t\t\to = T.nnet.sigmoid(pre_act[3*num_states:4*num_states])\n\n\t\t\t# do LSTM update\n\t\t\tc_t = f * c_tm1.T + i * z\n\t\t\th_t = o * T.tanh(c_t)\n\n\t\t\treturn c_t.T, h_t.T, att_acc_t\t# 1, 2: (batch_size, num_states); 3, 4: (batch_size, img_dim, img_dim)\n\n\t\tc0 = T.zeros((batch_size, num_states))\n\t\th0 = T.zeros((batch_size, num_states))\n\t\tatt_acc0 = T.zeros((batch_size, img_dim, img_dim))\n\t\t\n\t\tcells, hiddens, att_acc_T = theano.scan(fn=step, non_sequences=[input, self.W, self.Wg], outputs_info=[c0, h0, att_acc0], \n\t\t\t\t\t\t\t\t\t\tn_steps=self.num_glimpses, strict=True)[0]\n\t\tif self.final_state_only:\n\t\t\treturn hiddens[-1]\n\t\telse:\n\t\t\treturn hiddens","def apply_to_batch(self, batch_dict):\r\n \r\n self._last_batch = batch_dict\r\n \r\n if isinstance(self.model,NMTModelWithMLTM):\r\n y_pred = self.model(x_source=batch_dict['x_source'], \r\n x_mltm=batch_dict['x_source_mltm_vector'],\r\n x_source_lengths=batch_dict['x_source_length'], \r\n target_sequence=batch_dict['x_target'])\r\n else:\r\n y_pred = self.model(x_source=batch_dict['x_source'], \r\n x_source_lengths=batch_dict['x_source_length'], \r\n target_sequence=batch_dict['x_target'])\r\n self._last_batch['y_pred'] = y_pred\r\n \r\n attention_batched = np.stack(self.model.decoder._cached_p_attn).transpose(1, 0, 2)\r\n self._last_batch['attention'] = attention_batched","def sample_beam(model, input_embedding, char2idx, idx2char, k=5, maxlen=30,\n start='START', use_head=True):\n with torch.no_grad():\n device = input_embedding.device\n softmax = nn.Softmax(dim=1)\n if use_head:\n input_embedding = input_embedding.view(1, -1)\n\n inp = [torch.LongTensor([char2idx[start]]).to(device)]\n inp = nn.utils.rnn.pack_sequence(inp)\n out, hidden = model(input_embedding, inp, use_head=use_head)\n\n out = softmax(out.data).view(-1).cpu().numpy()\n max_k = np.argsort(out)[-k:][::-1]\n oldprobs = out[max_k]\n words = [[i] for i in max_k]\n inp = pack([torch.LongTensor([j]).to(device) for j in max_k])\n\n if model.mode == 'LSTM':\n hidden0 = torch.cat([hidden[0] for i in range(k)], dim=1)\n hidden1 = torch.cat([hidden[1] for i in range(k)], dim=1)\n hidden = hidden0, hidden1\n else:\n hidden = torch.cat([hidden for i in range(k)], dim=1)\n WORDS = []\n for c in range(maxlen):\n out, hidden = model(hidden, inp, use_head=False)\n out = softmax(out.data).cpu().numpy()\n\n #print(out.shape)\n inpnp = inp.data.detach().cpu().numpy()\n done = np.where(inpnp == char2idx['END'])\n out[done] = 0\n if len(out[done]) != 0:\n #print(out[done].shape)\n for d in done[0]:\n out[d][char2idx['END']] = 1\n #print(done)\n\n #print(out)\n #print(out[done])\n out = (oldprobs.reshape(-1, 1)*out)\n max_k = np.argsort(out)[:, -k:][:, ::-1]\n\n #print(max_k)\n probs = np.array([out[i][max_k[i]] for i in range(k)])\n #print(probs)\n flat = probs.reshape(-1)\n max_k2 = np.argsort(flat)[::-1][:k]\n word_inds = max_k2//k\n next_chars_inds = max_k2%k\n\n oldprobs = flat[max_k2]\n #print(oldprobs)\n\n new_words = []\n new_inp = []\n for i, word_ind in enumerate(word_inds):\n next_char = max_k[word_ind][next_chars_inds[i]]\n if next_char == char2idx['END']:\n #print(\"HIT AN END at word {}\".format(word_ind))\n WORDS.append((words[word_ind], oldprobs[i]))\n #the_word = words[word_ind]\n #return ''.join([idx2char[i] for i in the_word])\n new_inp.append(torch.LongTensor([next_char]).to(device))\n word = words[word_ind][:]\n word = word + [next_char]\n new_words.append(word)\n words = new_words[:]\n\n if model.mode == 'LSTM':\n h1, h2 = hidden\n h1, h2 = h1[0][word_inds].view(1, k, -1), h2[0][word_inds].view(1, k, -1)\n hidden = h1, h2\n else:\n hidden = hidden[0][word_inds].view(1, k, -1)\n\n\n inp = pack(new_inp)\n\n return [''.join([idx2char[i] for i in word if i != char2idx['END']]) for word in words], oldprobs","def forward(self, *args):\r\n _, (hn, _) = self.enc(*args)\r\n\r\n y_pred = self.mlp(hn)\r\n return y_pred","def mtr_lm_v1():\n hparams = mtr_lm_dense(0)\n hparams.layers = ([\"local_self_att\", \"local_self_att\", \"drd\",\n \"self_att\", \"drd\", \"local_self_att\",\n \"local_self_att\", \"moe_2d\"] * 4)[:-1]\n hparams.d_kv = 128\n hparams.moe_expert_x = 8\n hparams.moe_expert_y = 4\n hparams.moe_hidden_size = 32768\n hparams.d_ff = 2048\n hparams.num_memory_heads = 0\n hparams.mesh_shape = \"b0:4;b1:8\"\n hparams.layout = \"outer_batch:b0;inner_batch:b1,expert_x:b1,expert_y:b0\"\n hparams.outer_batch_size = 4\n return hparams","def generate_plots(self, input_data, input_labels=None):\n super(MlpModel, self).generate_plots(input_data, input_labels)\n feed_dict = self.get_feed_dict(input_data, input_labels)\n eval_list = [self.global_step, self.get_encodings(), self.mlp_module.weight_list]\n train_on_adversarial = feed_dict[self.train_on_adversarial]\n if(train_on_adversarial):\n eval_list += [self.adv_module.get_adv_input()]\n eval_out = tf.compat.v1.get_default_session().run(eval_list, feed_dict)\n current_step = str(eval_out[0])\n filename_suffix = \"_v\"+self.params.version+\"_\"+current_step.zfill(5)+\".png\"\n activity = eval_out[1]\n fig = pf.plot_activity_hist(activity, title=\"Logit Histogram\",\n save_filename=self.params.disp_dir+\"act_hist\"+filename_suffix)\n #First layer weights\n mlp_weights = eval_out[2]\n w_enc = mlp_weights[0]\n if self.params.mlp_layer_types[0] == \"fc\":\n w_enc_norm = np.linalg.norm(w_enc, axis=0, keepdims=False)\n # Don't plot weights as images if input is not square\n w_input_sqrt = np.sqrt(w_enc.shape[0])\n if(np.floor(w_input_sqrt) == np.ceil(w_input_sqrt)):\n w_enc = dp.reshape_data(w_enc.T, flatten=False)[0] # [num_neurons, height, width]\n fig = pf.plot_data_tiled(w_enc, normalize=False,\n title=\"Weights at step \"+current_step, vmin=None, vmax=None,\n save_filename=self.params.disp_dir+\"w_enc\"+filename_suffix)\n else: # conv\n w_enc = np.transpose(dp.rescale_data_to_one(w_enc.T)[0].T, axes=(3,0,1,2))\n if(w_enc.shape[-1] == 1 or w_enc.shape[-1] == 3):\n pf.plot_data_tiled(w_enc, normalize=False, title=\"Weights at step \"+current_step,\n save_filename=self.params.disp_dir+\"w_enc\"+filename_suffix)\n for (w, tf_w) in zip(mlp_weights, self.mlp_module.weight_list):\n #simplify tensorflow node name to only be the last one\n w_name = tf_w.name.split(\"/\")[-1].split(\":\")[0]\n num_f = w.shape[-1]\n w_reshape = np.reshape(w, [-1, num_f])\n w_norm = np.linalg.norm(w_reshape, axis=0, keepdims=False)\n fig = pf.plot_bar(w_norm, num_xticks=5,\n title=w_name+\" l2 norm\", xlabel=\"w index\", ylabel=\"L2 Norm\",\n save_filename=self.params.disp_dir+\"w_norm_\"+w_name+filename_suffix)\n if(train_on_adversarial):\n adv_input = eval_out[-1]\n adv_input = dp.reshape_data(adv_input, flatten=False)[0]\n fig = pf.plot_data_tiled(adv_input, normalize=False,\n title=\"Adv inputs at \"+current_step,\n save_filename=self.params.disp_dir+\"adv_input\"+filename_suffix)","def Attention_block(up_in,down_in,nf):\n# def MyTile(alpha, channels):\n# alpha = tf.tile(alpha, [1, 1, 1, 1, channels])\n# return alpha\n# def MyTile_output_shape(input_shape):\n# shape = list(input_shape)\n# assert len(shape) == 5 # only valid for 2D tensors\n# shape[-1] *= channels\n# return tuple(shape)\n ndims = len(up_in.get_shape()) - 2\n assert ndims in [1, 2, 3], \"ndims should be one of 1, 2, or 3. found: %d\" % ndims\n input_channels = up_in.get_shape().as_list()[-1]\n# batch_size1 = tf.shape(down_in)[0]\n# nf = tf.min(batch_size0,batch_size1)\n Conv = getattr(KL, 'Conv%dD' % ndims)\n up = Conv(nf, kernel_size=1, padding='same',\n kernel_initializer='he_normal', strides=2)(up_in)\n down = Conv(nf, kernel_size=1, padding='same',\n kernel_initializer='he_normal', strides=1)(down_in)\n\n# x = NCCLayer(channels=nf)([up,down])\n x = Add()([up,down])\n x = Activation('relu')(x)\n x = Conv(1, kernel_size=1, padding='same',\n kernel_initializer='he_normal', use_bias = True, bias_initializer='zeros',strides=1,activation='sigmoid')(x)\n# x = Activation('sigmoid')(x)\n upsample_layer = getattr(KL, 'UpSampling%dD' % ndims)\n alpha = upsample_layer()(x)\n# alpha = Lambda(MyTile)(alpha, input_channels)\n alpha = MyTile(channels=input_channels)(alpha)\n up_out = Multiply()([alpha,up_in])\n return up_out","def _esim(self, head_output_steps, body_output_steps, head_sent_sizes, body_sent_sizes):\n with tf.variable_scope(\"ESIM_single_sentence_pair\", reuse=True):\n # head_sent_sizes = tf.Print(head_sent_sizes, [head_sent_sizes, tf.shape(head_sent_sizes)],\n # \"head_sent_sizes:\\n\", summarize=32)\n # body_sent_sizes = tf.Print(body_sent_sizes, [body_sent_sizes, tf.shape(body_sent_sizes)],\n # \"body_sent_sizes:\\n\", summarize=32)\n # batch_size * sents * head_words * body_words\n attention_matrix = tf.matmul(head_output_steps, body_output_steps, transpose_b=True)\n self.attention_matrix = attention_matrix\n # batch_size * sents * head_words * body_words\n soft_attention_matrix_for_head = tf.nn.softmax(attention_matrix, 3)\n self.softmax_matrix_head = soft_attention_matrix_for_head\n # batch_size * sents * body_words * head_words\n soft_attention_matrix_for_body = tf.nn.softmax(tf.transpose(attention_matrix, [0, 1, 3, 2]), 3)\n self.softmax_matrix_body = soft_attention_matrix_for_body\n # batch_size * sents * head_words * embed\n head_attended = tf.matmul(soft_attention_matrix_for_head, body_output_steps, name=\"attend_head\")\n # head_attended = tf.Print(head_attended, [head_attended, tf.shape(head_attended)], \"head_attended:\\n\",\n # summarize=15000)\n\n # batch_size * sents * body_words * embed\n body_attended = tf.matmul(soft_attention_matrix_for_body, head_output_steps, name=\"attend_body\")\n # body_attended = tf.Print(body_attended, [body_attended, tf.shape(body_attended)], \"body_attended:\\n\",\n # summarize=15000)\n\n batch_size, num_sents, max_head_words, embed_size = tf.unstack(tf.shape(head_attended))\n flat_head_sent_sizes = tf.reshape(head_sent_sizes, [batch_size * num_sents])\n flat_head_seq_masks = tf.sequence_mask(flat_head_sent_sizes, max_head_words)\n head_seq_masks = tf.reshape(flat_head_seq_masks, [batch_size, num_sents, max_head_words])\n # batch_size * sents * head_words\n head_masks_per_word = tf.where(head_seq_masks, tf.ones_like(head_seq_masks, dtype=tf.float32),\n tf.zeros_like(head_seq_masks, dtype=tf.float32))\n encode_size = 2 * self.num_neurons[0]\n # now batch_size * sents * head_words * embed\n head_masks = tf.tile(tf.expand_dims(head_masks_per_word, 3), [1, 1, 1, encode_size])\n masked_head_attended = head_attended * head_masks\n\n _, _, max_body_words, _ = tf.unstack(tf.shape(body_attended))\n flat_body_sent_sizes = tf.reshape(body_sent_sizes, [batch_size * num_sents])\n flat_body_seq_masks = tf.sequence_mask(flat_body_sent_sizes, max_body_words)\n body_seq_masks = tf.reshape(flat_body_seq_masks, [batch_size, num_sents, max_body_words])\n # batch_size * sents * body_words\n body_masks_per_word = tf.where(body_seq_masks, tf.ones_like(body_seq_masks, dtype=tf.float32),\n tf.zeros_like(body_seq_masks, dtype=tf.float32))\n # now batch_size * sents * body_words * embed\n body_masks = tf.tile(tf.expand_dims(body_masks_per_word, 3), [1, 1, 1, encode_size])\n masked_body_attended = body_attended * body_masks\n\n # batch_size * sents * head_words * (4 * embed)\n head_concat = tf.concat(\n [head_output_steps, masked_head_attended, tf.abs(tf.subtract(masked_head_attended, head_output_steps)),\n tf.multiply(head_output_steps, masked_head_attended)], 3, name=\"concat_head_attended\")\n # batch_size * sents * body_words * (4 * embed)\n body_concat = tf.concat(\n [body_output_steps, masked_body_attended, tf.abs(tf.subtract(masked_body_attended, body_output_steps)),\n tf.multiply(body_output_steps, masked_body_attended)], 3, name=\"concat_body_attended\")\n return head_concat, body_concat, head_masks_per_word, body_masks_per_word","def task_metalearn(inp, reuse=True):\n inputa1,inputa2,inputa3, inputb1,inputb2,inputb3, labela, labelb = inp\n print(\"Input a: \" , inputa1)\n task_outputbs, task_lossesb, auto_losses = [], [], []\n auto_loss = None\n\n #One for each. \n\n # This takes in the input and passes out the latent variables.\n temp_in_a_1 = self.encoder(inputa1,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_a_1 = self.decoder(temp_in_a_1,self.auto_weights, reuse=reuse)\n auto_out_a_1 = temp_out_a_1\n\n temp_in_a_2 = self.encoder(inputa2,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_a_2 = self.decoder(temp_in_a_2,self.auto_weights, reuse=reuse)\n auto_out_a_2 = temp_out_a_2\n\n temp_in_a_3 = self.encoder(inputa3,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_a_3 = self.decoder(temp_in_a_3,self.auto_weights, reuse=reuse)\n auto_out_a_3 = temp_out_a_3\n\n # This takes in the input and passes out the latent variables.\n temp_in_b_1 = self.encoder(inputb1,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_b_1 = self.decoder(temp_in_b_1,self.auto_weights, reuse=reuse)\n auto_out_b_1 = temp_out_b_1\n\n temp_in_b_2 = self.encoder(inputb2,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_b_2 = self.decoder(temp_in_b_2,self.auto_weights, reuse=reuse)\n auto_out_b_2 = temp_out_b_2\n\n temp_in_b_3 = self.encoder(inputb3,self.auto_weights, reuse=reuse)\n #Then transform it back, and take the loss\n temp_out_b_3 = self.decoder(temp_in_b_3,self.auto_weights, reuse=reuse)\n auto_out_b_3 = temp_out_b_3\n\n\n #print(\"temp out a: \" , temp_out_a)\n la_1 = self.loss_func(temp_out_a_1,inputa1)\n la_2 = self.loss_func(temp_out_a_2,inputa2)\n la_3 = self.loss_func(temp_out_a_3,inputa3)\n\n lb_1 = self.loss_func(temp_out_b_1,inputb1)\n lb_2 = self.loss_func(temp_out_b_2,inputb2)\n lb_3 = self.loss_func(temp_out_b_3,inputb3)\n\n auto_loss = la_1+la_2+la_3\n\n print(\"Input a1: \" , temp_in_a_1)\n inputa=tf.concat([temp_in_a_1, temp_in_a_2,temp_in_a_3],1)\n inputb=tf.concat([temp_in_b_1, temp_in_b_2,temp_in_b_3],1)\n print(\"Inputa: \" , inputa)\n task_outputa = self.forward(inputa, weights, reuse=reuse) # only reuse on the first iter\n temp_outputa = self.decoder(task_outputa,self.auto_weights,reuse=reuse)\n \n print(\"Task outputa: \" , temp_outputa)\n print(\"Label a: \" , labela)\n task_lossa = self.loss_func(temp_outputa, labela)\n\n grads = tf.gradients(task_lossa, list(weights.values()))\n if FLAGS.stop_grad:\n grads = [tf.stop_gradient(grad) for grad in grads]\n gradients = dict(zip(weights.keys(), grads))\n fast_weights = dict(zip(weights.keys(), [weights[key] - self.update_lr*gradients[key] for key in weights.keys()]))\n output = self.forward(inputb, fast_weights, reuse=True)\n \n temp_outputb = self.decoder(output,self.auto_weights, reuse=True)\n output = temp_outputb\n\n task_outputbs.append(output)\n print(\"Output: \" , output)\n print(\"Labels: \" , labelb)\n task_lossesb.append(self.loss_func(output, labelb))\n print(\"Num updates is: \" , num_updates-1)\n\n for j in range(num_updates - 1):\n loss = self.loss_func(self.decoder(self.forward(inputa, fast_weights, reuse=True),self.auto_weights), labela)\n grads = tf.gradients(loss, list(fast_weights.values()))\n if FLAGS.stop_grad:\n grads = [tf.stop_gradient(grad) for grad in grads]\n gradients = dict(zip(fast_weights.keys(), grads))\n fast_weights = dict(zip(fast_weights.keys(), [fast_weights[key] - self.update_lr*gradients[key] for key in fast_weights.keys()]))\n \n output = self.forward(inputb, fast_weights, reuse=True)\n output = self.decoder(output,self.auto_weights)\n \n task_outputbs.append(output)\n task_lossesb.append(self.loss_func(output, labelb))\n # Task Outputa Direct output, non-decoded (only first step). This is the location (so it is in hyperparameter space)\n # Task_outputbs Output from complete system, for each step (first is after 1st step, 2nd is after next, etc.)\n # task_lossa Loss of difference in first step (decoded)\n # task_lossesb Losses from complete ssystem from each step \n # auto_loss Sum of all auto-encoding loss of a's\n # auto_out_a_1 Auto-encoded a1 output\n # auto_out_a_2 Auto-encoded a2 output\n\n task_output = [temp_outputa, task_outputbs, task_lossa, task_lossesb, auto_loss,auto_out_a_1,auto_out_a_2]\n return task_output","def forward(self, x, y, x_mask, candidate_aggre):\n x_ans_mask = candidate_aggre.sum(dim=2).ge(0).float()\n Wy = self.linear(y) if self.linear is not None else y # batch * h1\n p = torch.bmm(x,Wy.unsqueeze(2)).squeeze(2) # batch * len\n p.data.masked_fill_(x_mask.data, -float('inf')) \n pm = F.softmax(p, dim=1) * x_ans_mask # batch * len\n unnormalized_probs=torch.bmm(pm.unsqueeze(1), candidate_aggre).squeeze(1) # batch * c\n\n normalized_probs=unnormalized_probs/unnormalized_probs.sum(dim=1, keepdim=True)+very_small_number\n if self.training:\n return torch.log(normalized_probs)\n else:\n return normalized_probs","def forward(self, *input):\n \n item_text, item_img, uidx = input\n # compute latent codes\n user_codes = self.user_embdding(uidx)\n item_img_codes = self.ilatent_img_codes(item_img)\n item_text_codes = self.ilatent_text_codes(item_text)\n #ubianry = self.hash(user_codes)\n #ibinary = [self.hash(h) for h in item_codes]\n uscore_img = self.uscores(user_codes, item_img_codes)\n iscore_img = self.iscores(item_img_codes)\n uscore_text = self.uscores(user_codes, item_text_codes)\n iscore_text = self.iscores(item_text_codes)\n score = uscore_img + uscore_text + iscore_img + iscore_text\n\n binary = torch.zeros_like(score)#self.scores(ubianry, ibinary).view(-1, 1)\n return score, binary","def init_boldpostprocess_wf(\n lower_bpf,\n upper_bpf,\n contigvol,\n bpf_order,\n motion_filter_order,\n motion_filter_type,\n band_stop_min,\n band_stop_max,\n smoothing,\n bold_file,\n head_radius,\n params,\n custom_conf,\n omp_nthreads,\n dummytime,\n output_dir,\n fd_thresh,\n num_bold,\n mni_to_t1w,\n despike,\n brain_template='MNI152NLin2009cAsym',\n layout=None,\n name='bold_postprocess_wf',\n ):\n\n\n TR = layout.get_tr(bold_file)\n file_base = os.path.basename(str(bold_file))\n workflow = Workflow(name=name)\n\n workflow.__desc__ = \"\"\"\nFor each of the {num_bold} BOLD series found per subject (across all\ntasks and sessions), the following post-processing was performed:\n\"\"\".format(num_bold=num2words(num_bold))\n\n if dummytime > 0:\n nvolx = str(np.floor(dummytime / TR))\n workflow.__desc__ = workflow.__desc__ + \"\"\" \\\nbefore nuisance regression and filtering of the data, the first {nvol} were discarded,\n.Furthermore,volumes with framewise-displacement greater than \n{fd_thresh} mm [@power_fd_dvars;@satterthwaite_2013] were flagged as outliers\n and excluded from nuisance regression.\n\"\"\".format(nvol=num2words(nvolx),fd_thresh=fd_thresh)\n\n else:\n workflow.__desc__ = workflow.__desc__ + \"\"\" \\\nbefore nuisance regression and filtering of the data, volumes with framewise-displacement greater than \n{fd_thresh} mm [@power_fd_dvars;@satterthwaite_2013] were flagged as outliers\n and excluded from nuisance regression.\n\"\"\".format(fd_thresh=fd_thresh)\n\n workflow.__desc__ = workflow.__desc__ + \"\"\" \\\n{regressors} [@benchmarkp;@satterthwaite_2013]. These nuisance regressors were \nregressed from the BOLD data using linear regression - as implemented in Scikit-Learn {sclver} [@scikit-learn].\nResidual timeseries from this regression were then band-pass filtered to retain signals within the {highpass}-{lowpass} Hz frequency band. \n \"\"\".format(regressors=stringforparams(params=params),sclver=sklearn.__version__,\n lowpass=upper_bpf,highpass=lower_bpf)\n\n\n # get reference and mask\n mask_file,ref_file = _get_ref_mask(fname=bold_file)\n\n inputnode = pe.Node(niu.IdentityInterface(\n fields=['bold_file','ref_file','bold_mask','cutstom_conf','mni_to_t1w','t1w','t1seg']),\n name='inputnode')\n\n inputnode.inputs.bold_file = str(bold_file)\n inputnode.inputs.ref_file = str(ref_file)\n inputnode.inputs.bold_mask = str(mask_file)\n inputnode.inputs.custom_conf = str(custom_conf)\n\n\n outputnode = pe.Node(niu.IdentityInterface(\n fields=['processed_bold', 'smoothed_bold','alff_out','smoothed_alff',\n 'reho_out','sc217_ts', 'sc217_fc','sc417_ts','sc417_fc','ts50_ts','ts50_fc',\n 'gs360_ts', 'gs360_fc','gd333_ts', 'gd333_fc','qc_file','fd']),\n name='outputnode')\n\n mem_gbx = _create_mem_gb(bold_file)\n\n\n fcon_ts_wf = init_fcon_ts_wf(mem_gb=mem_gbx['timeseries'],mni_to_t1w=mni_to_t1w,\n t1w_to_native=_t12native(bold_file),bold_file=bold_file,\n brain_template=brain_template,name=\"fcons_ts_wf\")\n\n alff_compute_wf = init_compute_alff_wf(mem_gb=mem_gbx['timeseries'], TR=TR,\n lowpass=upper_bpf,highpass=lower_bpf,smoothing=smoothing, cifti=False,\n name=\"compute_alff_wf\" )\n\n reho_compute_wf = init_3d_reho_wf(mem_gb=mem_gbx['timeseries'],smoothing=smoothing,\n name=\"afni_reho_wf\")\n\n write_derivative_wf = init_writederivatives_wf(smoothing=smoothing,bold_file=bold_file,\n params=params,cifti=None,output_dir=output_dir,dummytime=dummytime,\n lowpass=upper_bpf,highpass=lower_bpf,TR=TR,omp_nthreads=omp_nthreads,\n name=\"write_derivative_wf\")\n\n confoundmat_wf = pe.Node(ConfoundMatrix(head_radius=head_radius, params=params,\n filtertype=motion_filter_type,cutoff=band_stop_max,\n low_freq=band_stop_max,high_freq=band_stop_min,TR=TR,\n filterorder=motion_filter_order),\n name=\"ConfoundMatrix_wf\", mem_gb=mem_gbx['resampled'])\n\n censorscrub_wf = init_censoring_wf(mem_gb=mem_gbx['timeseries'],TR=TR,custom_conf=custom_conf,head_radius=head_radius,\n contigvol=contigvol,dummytime=dummytime,fd_thresh=fd_thresh,name='censoring')\n \n resdsmoothing_wf = init_resd_smoohthing(mem_gb=mem_gbx['timeseries'],smoothing=smoothing,cifti=False,\n name=\"resd_smoothing_wf\")\n \n filtering_wf = pe.Node(FilteringData(tr=TR,lowpass=upper_bpf,highpass=lower_bpf,\n filter_order=bpf_order),\n name=\"filtering_wf\", mem_gb=mem_gbx['timeseries'])\n\n regression_wf = pe.Node(regress(tr=TR),\n name=\"regression_wf\",mem_gb = mem_gbx['timeseries'])\n\n interpolate_wf = pe.Node(interpolate(TR=TR),\n name=\"interpolation_wf\",mem_gb = mem_gbx['timeseries'])\n\n \n executivesummary_wf =init_execsummary_wf(tr=TR,bold_file=bold_file,layout=layout,\n output_dir=output_dir,mni_to_t1w=mni_to_t1w,omp_nthreads=2)\n \n\n # get transform file for resampling and fcon\n \n \n \n transformfile = get_transformfile(bold_file=bold_file,\n mni_to_t1w=mni_to_t1w,t1w_to_native=_t12native(bold_file))\n t1w_mask = get_maskfiles(bold_file=bold_file,mni_to_t1w=mni_to_t1w)[1]\n\n bold2MNI_trans,bold2T1w_trans = get_transformfilex(bold_file=bold_file,\n mni_to_t1w=mni_to_t1w,t1w_to_native=_t12native(bold_file)) \n\n \n resample_parc = pe.Node(ApplyTransforms(\n dimension=3,\n input_image=str(get_template(\n 'MNI152NLin2009cAsym', resolution=1, desc='carpet',\n suffix='dseg', extension=['.nii', '.nii.gz'])),\n interpolation='MultiLabel',transforms=transformfile),\n name='resample_parc')\n \n resample_bold2T1w = pe.Node(ApplyTransforms(\n dimension=3,\n input_image=mask_file,reference_image=t1w_mask,\n interpolation='NearestNeighbor',transforms=bold2T1w_trans),\n name='bold2t1_trans')\n \n resample_bold2MNI = pe.Node(ApplyTransforms(\n dimension=3,\n input_image=mask_file,reference_image=str(get_template(\n 'MNI152NLin2009cAsym', resolution=2, desc='brain',\n suffix='mask', extension=['.nii', '.nii.gz'])),\n interpolation='NearestNeighbor',transforms=bold2MNI_trans),\n name='bold2mni_trans')\n\n qcreport = pe.Node(computeqcplot(TR=TR,bold_file=bold_file,dummytime=dummytime,t1w_mask=t1w_mask,\n template_mask = str(get_template('MNI152NLin2009cAsym', resolution=2, desc='brain',\n suffix='mask', extension=['.nii', '.nii.gz'])),\n head_radius=head_radius), name=\"qc_report\",mem_gb = mem_gbx['resampled'])\n \n\n workflow.connect([\n # connect bold confound matrix to extract confound matrix \n (inputnode, confoundmat_wf, [('bold_file', 'in_file'),]),\n ])\n \n # if there is despiking\n if despike:\n despike_wf = pe.Node(Despike(outputtype='NIFTI_GZ',args='-NEW'),name=\"despike_wf\",mem_gb=mem_gbx['resampled'])\n\n workflow.connect([\n (inputnode,despike_wf,[('bold_file','in_file')]),\n (despike_wf,censorscrub_wf,[('out_file','inputnode.bold')])\n ])\n else:\n workflow.connect([\n (inputnode,censorscrub_wf,[('bold_file','inputnode.bold')]),\n ])\n \n # add neccessary input for censoring if there is one\n workflow.connect([\n\t (inputnode,censorscrub_wf,[('bold_file','inputnode.bold_file'),\n\t ('bold_mask','inputnode.bold_mask')]),\n\t (confoundmat_wf,censorscrub_wf,[('confound_file','inputnode.confound_file')])\n ])\n\n # regression workflow \n workflow.connect([\n\t (inputnode,regression_wf,[('bold_mask','mask')]),\n\t (censorscrub_wf,regression_wf,[('outputnode.bold_censored','in_file'),\n\t ('outputnode.fmriprepconf_censored','confounds'), \n\t\t ('outputnode.customconf_censored','custom_conf')])\n ])\n # interpolation workflow\n workflow.connect([\n\t (inputnode,interpolate_wf,[('bold_file','bold_file'),('bold_mask','mask_file')]),\n\t (censorscrub_wf,interpolate_wf,[('outputnode.tmask','tmask')]),\n\t (regression_wf,interpolate_wf,[('res_file','in_file')]), \n\t])\n # add filtering workflow \n workflow.connect([\n (inputnode,filtering_wf,[('bold_mask','mask')]),\n\t (interpolate_wf,filtering_wf,[('bold_interpolated','in_file')]),\n\n ])\n \n # residual smoothing \n workflow.connect([\n\t (filtering_wf,resdsmoothing_wf,[('filt_file','inputnode.bold_file')]) \n ])\n\n #functional connect workflow\n workflow.connect([\n (inputnode,fcon_ts_wf,[('ref_file','inputnode.ref_file'),]),\n (filtering_wf,fcon_ts_wf,[('filt_file','inputnode.clean_bold'),]),\n ])\n # reho and alff\n workflow.connect([ \n\t (inputnode,alff_compute_wf,[('bold_mask','inputnode.bold_mask')]),\n\t (inputnode,reho_compute_wf,[('bold_mask','inputnode.bold_mask')]),\n\t (filtering_wf, alff_compute_wf,[('filt_file','inputnode.clean_bold')]),\n\t (filtering_wf, reho_compute_wf,[('filt_file','inputnode.clean_bold')]),\n ])\n\n # qc report\n workflow.connect([\n (inputnode,qcreport,[('bold_mask','mask_file')]),\n (filtering_wf,qcreport,[('filt_file','cleaned_file')]),\n (censorscrub_wf,qcreport,[('outputnode.tmask','tmask')]),\n (inputnode,resample_parc,[('ref_file','reference_image')]),\n (resample_parc,qcreport,[('output_image','seg_file')]),\n (resample_bold2T1w,qcreport,[('output_image','bold2T1w_mask')]),\n (resample_bold2MNI,qcreport,[('output_image','bold2temp_mask')]),\n (qcreport,outputnode,[('qc_file','qc_file')]),\n ])\n\n \n\n # write to the outputnode, may be use in future\n workflow.connect([\n\t(filtering_wf,outputnode,[('filt_file','processed_bold')]),\n\t(censorscrub_wf,outputnode,[('outputnode.fd','fd')]),\n\t(resdsmoothing_wf,outputnode,[('outputnode.smoothed_bold','smoothed_bold')]),\n\t(alff_compute_wf,outputnode,[('outputnode.alff_out','alff_out'),\n ('outputnode.smoothed_alff','smoothed_alff')]),\n (reho_compute_wf,outputnode,[('outputnode.reho_out','reho_out')]),\n\t (fcon_ts_wf,outputnode,[('outputnode.sc217_ts','sc217_ts' ),('outputnode.sc217_fc','sc217_fc'),\n ('outputnode.sc417_ts','sc417_ts'),('outputnode.sc417_fc','sc417_fc'),\n ('outputnode.gs360_ts','gs360_ts'),('outputnode.gs360_fc','gs360_fc'),\n ('outputnode.gd333_ts','gd333_ts'),('outputnode.gd333_fc','gd333_fc'),\n ('outputnode.ts50_ts','ts50_ts'),('outputnode.ts50_fc','ts50_fc')]),\n\n ])\n \n # write derivatives \n workflow.connect([\n (filtering_wf,write_derivative_wf,[('filt_file','inputnode.processed_bold')]),\n\t (resdsmoothing_wf,write_derivative_wf,[('outputnode.smoothed_bold','inputnode.smoothed_bold')]),\n (censorscrub_wf,write_derivative_wf,[('outputnode.fd','inputnode.fd')]),\n (alff_compute_wf,write_derivative_wf,[('outputnode.alff_out','inputnode.alff_out'),\n ('outputnode.smoothed_alff','inputnode.smoothed_alff')]),\n (reho_compute_wf,write_derivative_wf,[('outputnode.reho_out','inputnode.reho_out')]),\n (fcon_ts_wf,write_derivative_wf,[('outputnode.sc217_ts','inputnode.sc217_ts' ),\n ('outputnode.sc217_fc','inputnode.sc217_fc'),\n ('outputnode.sc417_ts','inputnode.sc417_ts'),\n ('outputnode.sc417_fc','inputnode.sc417_fc'),\n ('outputnode.gs360_ts','inputnode.gs360_ts'),\n ('outputnode.gs360_fc','inputnode.gs360_fc'),\n ('outputnode.gd333_ts','inputnode.gd333_ts'),\n ('outputnode.gd333_fc','inputnode.gd333_fc'),\n ('outputnode.ts50_ts','inputnode.ts50_ts'),\n ('outputnode.ts50_fc','inputnode.ts50_fc')]),\n (qcreport,write_derivative_wf,[('qc_file','inputnode.qc_file')]),\n\n\n\n ])\n functional_qc = pe.Node(FunctionalSummary(bold_file=bold_file,tr=TR),\n name='qcsummary', run_without_submitting=True)\n\n ds_report_qualitycontrol = pe.Node(\n DerivativesDataSink(base_directory=output_dir, desc='qualitycontrol',source_file=bold_file, datatype=\"figures\"),\n name='ds_report_qualitycontrol', run_without_submitting=True)\n\n ds_report_preprocessing = pe.Node(\n DerivativesDataSink(base_directory=output_dir, desc='preprocessing',source_file=bold_file, datatype=\"figures\"),\n name='ds_report_preprocessing', run_without_submitting=True)\n ds_report_postprocessing = pe.Node(\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='postprocessing', datatype=\"figures\"),\n name='ds_report_postprocessing', run_without_submitting=True)\n\n ds_report_connectivity = pe.Node(\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='connectvityplot', datatype=\"figures\"),\n name='ds_report_connectivity', run_without_submitting=True)\n\n ds_report_rehoplot = pe.Node(\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='rehoplot', datatype=\"figures\"),\n name='ds_report_rehoplot', run_without_submitting=True)\n\n ds_report_afniplot = pe.Node(\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='afniplot', datatype=\"figures\"),\n name='ds_report_afniplot', run_without_submitting=True)\n\n workflow.connect([\n (qcreport,ds_report_preprocessing,[('raw_qcplot','in_file')]),\n (qcreport,ds_report_postprocessing ,[('clean_qcplot','in_file')]),\n (qcreport,functional_qc,[('qc_file','qc_file')]),\n (functional_qc,ds_report_qualitycontrol,[('out_report','in_file')]),\n (fcon_ts_wf,ds_report_connectivity,[('outputnode.connectplot','in_file')]),\n (reho_compute_wf,ds_report_rehoplot,[('outputnode.rehohtml','in_file')]),\n (alff_compute_wf,ds_report_afniplot ,[('outputnode.alffhtml','in_file')]),\n ])\n\n\n ## exexetive summary workflow\n workflow.connect([\n (inputnode,executivesummary_wf,[('t1w','inputnode.t1w'),('t1seg','inputnode.t1seg'),\n ('bold_file','inputnode.bold_file'),('bold_mask','inputnode.mask')]),\n\n (regression_wf,executivesummary_wf,[('res_file','inputnode.regdata'),]),\n (filtering_wf,executivesummary_wf,[('filt_file','inputnode.resddata')]),\n (censorscrub_wf,executivesummary_wf,[('outputnode.fd','inputnode.fd')]),\n ]),\n\n return workflow","def p(self, evidence, history):\n raise NotImplementedError(\"The p method for the MLE model has not been implemented yet.\")","def forward(self, inputs, token_types, valid_length=None): # pylint: disable=arguments-differ\n bert_output = self.bert(inputs, token_types, valid_length)\n if self.add_query:\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\n mask = 1 - token_types\n avg_q = mx.nd.sum(mx.nd.multiply(mask, o), axis=2) / mx.nd.sum(mask, axis=1)\n o = mx.nd.add(o, mx.nd.multiply(avg_q.expand_dims(axis=2), token_types))\n attended_output = mx.ndarray.transpose(o, axes=(1,2,0))\n if self.apply_coattention:\n #############################\n '''\n # option 3: use exactly the QANet way\n # not a good idea, this will cause index shift and thus cause bug\n # but if we are going to use QANet's method for modeling and extracting the output, it'll be fine to do so\n # so I just kept these lines here\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\n context_mask = token_types\n query_mask = 1 - context_mask\n context_max_len = bert_output.shape[1] # int(context_mask.sum(axis=1).max().asscalar())\n query_max_len = bert_output.shape[1] # int(query_mask.sum(axis=1).max().asscalar())\n context_raw = mx.nd.multiply(context_mask, o)\n context_raw = mx.ndarray.expand_dims(context_raw, 0)\n # print(context_raw[0,0,0,:])\n # to get the offset to shift using gridgenerator and bilinear-sampler\n raw_offset = query_mask.sum(axis=1).reshape(len(query_mask),1).tile(bert_output.shape[1])\n warp_matrix = mx.ndarray.expand_dims(mx.ndarray.stack(raw_offset, \n mx.nd.zeros(raw_offset.shape).as_in_context(raw_offset.context)), 0)\n grid = GridGenerator(data=warp_matrix, transform_type='warp')\n warpped_out = BilinearSampler(context_raw, grid)\n # the context mask also needs to be shifted\n context_mask = mx.ndarray.expand_dims(context_mask, 0)\n context_mask = mx.ndarray.expand_dims(context_mask, 0)\n context_mask = BilinearSampler(context_mask, grid)\n context_mask = mx.ndarray.squeeze(context_mask, axis=(0, 1))\n # get the two encodings separated\n context_emb_encoded = mx.ndarray.transpose(mx.ndarray.squeeze(warpped_out, axis=0), axes=(1,2,0))\n '''\n #################################\n # option 2: get the two encodings separated\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\n context_mask = token_types\n query_mask = 1 - context_mask\n context_max_len = bert_output.shape[1] # int(context_mask.sum(axis=1).max().asscalar())\n query_max_len = bert_output.shape[1] # int(query_mask.sum(axis=1).max().asscalar())\n context_emb_encoded = mx.ndarray.transpose(mx.nd.multiply(context_mask, o), axes=(1,2,0))\n query_emb_encoded = mx.ndarray.transpose(mx.nd.multiply(query_mask, o), axes=(1,2,0))\n '''\n # option 1: context and query differently masked but use the same values\n # problem: almost the same with simply self-attention\n context_mask = token_types\n query_mask = 1 - context_mask\n context_max_len = bert_output.shape[1]\n query_max_len = bert_output.shape[1]\n context_emb_encoded = bert_output\n query_emb_encoded = bert_output\n '''\n # context_mask = context_mask[:,:context_max_len]\n # query_mask = query_mask[:,:query_max_len]\n attended_output = self.co_attention(context_emb_encoded, query_emb_encoded, \n context_mask, query_mask, \n context_max_len, query_max_len)\n if self.apply_self_attention:\n attended_output, att_weights = self.multi_head_attention(bert_output, bert_output) \n if self.add_query or self.apply_self_attention or self.apply_coattention:\n output = self.span_classifier(attended_output)\n else:\n output = self.span_classifier(bert_output)\n return output","def __init__(self, h, d_model, dropout=0.1):\n super(MultiHeadedAttention, self).__init__()\n assert d_model % h == 0\n # We assume d_v always equals d_k\n self.d_k = d_model // h\n self.h = h\n self.linears = clones(nn.Linear(d_model, d_model), 4)\n self.attn = None\n self.dropout = nn.Dropout(p=dropout)","def generate(self, text, prev_mel):\n # forward pass through text embedding and get k and v\n kv = self.t_encoder(text)\n k = kv[:,:self.hp.d,:]\n v = kv[:,self.hp.d:,:]\n # forward pass through audio encoding and get Q\n q = self.a_encoder(prev_mel)\n \n # compute attention and use forcible incremental attention (section 4.2)\n a = (k.transpose(2,1)).matmul(q)/np.sqrt(self.hp.d)\n a = F.softmax(a, dim=1)\n \"\"\"\n # get argmax\n argmax = a[0].argmax(dim=0) # argmax on the N dimension\n # forcibly incremental attention\n preva = -1\n for i in range(a.shape[-1]):\n if argmax[i] < preva -1 or preva + 3 < argmax[i]:\n # force the ith column to be zero\n a[:,:,i] = 0\n # find correct position\n position = min(a.shape[1]-1, preva + 1)\n a[:,position,i] = 1.0\n # update preva\n preva = a[0,:,i].argmax()\"\"\"\n\n # finish computing y and a\n r = r = v.matmul(a)\n\n rprime = torch.cat((r, q), dim=1)\n ylogit = self.decoder(rprime)\n y = F.sigmoid(ylogit)\n return y, ylogit, a","def __init__(self, mbart_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\n self.linear1_weight = mbart_layer.fc1.weight\n self.linear1_bias = mbart_layer.fc1.bias\n self.linear2_weight = mbart_layer.fc2.weight\n self.linear2_bias = mbart_layer.fc2.bias\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\n self.norm2_eps = mbart_layer.final_layer_norm.eps\n self.norm2_weight = mbart_layer.final_layer_norm.weight\n self.norm2_bias = mbart_layer.final_layer_norm.bias\n self.num_heads = mbart_layer.self_attn.num_heads\n self.embed_dim = mbart_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def on_train_batch_end(\n self,\n trainer: 'pl.Trainer',\n pl_module: 'pl.LightningModule',\n outputs: Optional[STEP_OUTPUT],\n batch: Any,\n batch_idx: int,\n dataloader_idx: int,\n ) -> None:\n if trainer.global_step % self.every_n_steps == 0: \n text, x = batch\n sample_text = text[:1]\n token_list = sample_text.masked_select(sample_text != 0).tolist()\n decoded_text = self.tokenizer.decode(token_list) \n text = text.to(pl_module.device)\n x = x.to(pl_module.device) \n with torch.no_grad():\n pl_module.eval()\n #generate sample with image provided\n x_rec = pl_module.generate_images(text[:1], img = x[:1], filter_thres=0.9) # topk sampling at 0.9\n\n #generate sample without image\n x_gen = pl_module.generate_images(text[:1], filter_thres=0.9) # topk sampling at 0.9\n\n pl_module.train() \n\n\n x_grid = torchvision.utils.make_grid(\n tensor=x,\n nrow=self.nrow,\n padding=self.padding,\n normalize=self.normalize,\n value_range=self.norm_range,\n scale_each=self.scale_each,\n pad_value=self.pad_value,\n ) \n xrec_grid = torchvision.utils.make_grid(\n tensor=x_rec,\n nrow=self.nrow,\n padding=self.padding,\n normalize=self.normalize,\n value_range=self.norm_range,\n scale_each=self.scale_each,\n pad_value=self.pad_value,\n ) \n xgen_grid = torchvision.utils.make_grid(\n tensor=x_gen,\n nrow=self.nrow,\n padding=self.padding,\n normalize=self.normalize,\n value_range=self.norm_range,\n scale_each=self.scale_each,\n pad_value=self.pad_value,\n ) \n text_title = \"train/text\"\n trainer.logger.experiment.add_text(text_title, decoded_text, global_step=trainer.global_step)\n x_title = \"train/input\"\n trainer.logger.experiment.add_image(x_title, x_grid, global_step=trainer.global_step)\n xrec_title = \"train/half_reconstruction\"\n trainer.logger.experiment.add_image(xrec_title, xrec_grid, global_step=trainer.global_step)\n xgen_title = \"train/generation\"\n trainer.logger.experiment.add_image(xgen_title, xgen_grid, global_step=trainer.global_step)","def attention(decoder_state, coverage=None, num_words_section=None, step=None):\n with variable_scope.variable_scope(\"Attention\"):\n # Pass the decoder state through a linear layer (this is W_s s_t + b_attn in the paper)\n # (W_s s_t) + b_att is decoder_features; s_t = decoder_state\n decoder_features = linear(decoder_state, attention_vec_size, True) # shape (batch_size, attention_vec_size)\n decoder_features = tf.expand_dims(tf.expand_dims(decoder_features, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\n\n def masked_attention(e, enc_padding_mask):\n if enc_section_padding_mask is not None:\n enc_padding_mask = tf.reshape(enc_section_padding_mask, [batch_size, -1])\n enc_padding_mask = tf.cast(enc_padding_mask, tf.float32)\n \"\"\"Take softmax of e then apply enc_padding_mask and re-normalize\"\"\"\n attn_dist = nn_ops.softmax(e) # take softmax. shape (batch_size, attn_length)\n attn_dist *= enc_padding_mask # apply mask\n masked_sums = tf.reduce_sum(attn_dist, axis=1) # shape (batch_size)\n return attn_dist / tf.reshape(masked_sums, [-1, 1]) # re-normalize\n\n if use_coverage and coverage is not None: # non-first step of coverage\n if not hier:\n # Multiply coverage vector by w_c to get coverage_features.\n coverage_features = nn_ops.conv2d(coverage, w_c, [1, 1, 1, 1], \"SAME\") # c has shape (batch_size, seq_len, 1, attention_vec_size)\n \n # Calculate v^T tanh(W_h h_i + W_s s_t + w_c c_i^t + b_attn)\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features + coverage_features), [2, 3]) # shape (batch_size,seq_len)\n \n # Take softmax of e to get the attention distribution\n # attn_dist = nn_ops.softmax(e) # shape (batch_size, seq_len)\n attn_dist = masked_attention(e, enc_padding_mask)\n \n # Update coverage vector\n coverage += array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) # shape=(batch_size, seq_len,1,1)\n else:\n with tf.variable_scope(\"attention_sections\"):\n if FLAGS.fixed_attn:\n tf.logging.debug('running with fixed attn', '\\r')\n decoder_features_sec = linear(decoder_state, attention_vec_size, True, scope='Linear--Section-Features') # shape (batch_size, attention_vec_size)\n decoder_features_sec = tf.expand_dims(tf.expand_dims(decoder_features_sec, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\n e_sec = math_ops.reduce_sum(v_sec * math_ops.tanh(encoder_section_features + decoder_features_sec), [2, 3]) # [batch_size x seq_len_sections]\n attn_dist_sec = nn_ops.softmax(e_sec)\n else:\n e_sec = math_ops.reduce_sum(v_sec * math_ops.tanh(encoder_section_features + decoder_features), [2, 3]) # [batch_size x seq_len_sections]\n attn_dist_sec = nn_ops.softmax(e_sec)\n with tf.variable_scope(\"attention_words\"):\n coverage_features = nn_ops.conv2d(coverage, w_c, [1, 1, 1, 1], \"SAME\") # c has shape (batch_size, seq_len, 1, attention_vec_size)\n \n # Calculate v^T tanh(W_h h_i + W_s s_t + w_c c_i^t + b_attn)\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features + coverage_features), [2, 3]) # shape (batch_size,seq_len)\n\n # Multiply by section weights\n \n e = tf.reshape(e, [batch_size, -1, num_words_section[0][0]])\n e = tf.multiply(e, attn_dist_sec[:,:,tf.newaxis])\n e = tf.reshape(e, [batch_size,-1])\n\n\n# --- Some hack for reweighting attention (similar to temp for softmax)\n if temperature > 0.0:\n e = e * temperature\n \n attn_dist = masked_attention(e, enc_padding_mask)\n coverage += array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) # shape=(batch_size, seq_len,1,1)\n \n else:\n # Calculate v^T tanh(W_h h_i + W_s s_t + b_attn)\n if hier:\n with tf.variable_scope(\"attention_sections\"):\n if FLAGS.fixed_attn:\n decoder_features_sec = linear(decoder_state, attention_vec_size, True, scope='Linear--Section-Features') # shape (batch_size, attention_vec_size)\n decoder_features_sec = tf.expand_dims(tf.expand_dims(decoder_features_sec, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\n e_sec = math_ops.reduce_sum(\n v_sec * math_ops.tanh(encoder_section_features + decoder_features_sec), [2, 3]) # [batch_size x seq_len_sections]\n attn_dist_sec = nn_ops.softmax(e_sec)\n else:\n e_sec = math_ops.reduce_sum(\n v_sec * math_ops.tanh(encoder_section_features + decoder_features), [2, 3]) # [batch_size x seq_len_sections]\n attn_dist_sec = nn_ops.softmax(e_sec)\n\n with tf.variable_scope(\"attention_words\"):\n\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features), [2, 3]) #[batch_size x seq_len]\n\n e = tf.reshape(e, [batch_size, -1, num_words_section[0][0]])\n e = tf.multiply(e, attn_dist_sec[:,:,tf.newaxis])\n e = tf.reshape(e, [batch_size,-1])\n\n if temperature > 0.0:\n e = e * temperature\n \n attn_dist = masked_attention(e, enc_padding_mask)\n \n else:\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features), [2, 3]) # calculate e\n # Take softmax of e to get the attention distribution\n if enc_padding_mask is not None:\n attn_dist = masked_attention(e, enc_padding_mask)\n else:\n attn_dist = nn_ops.softmax(e) # shape (batch_size, seq_len)\n\n if use_coverage: # first step of training\n coverage = tf.expand_dims(tf.expand_dims(attn_dist,2),2) # initialize coverage\n\n # TODO: coverage for hier\n\n # Calculate the context vector from attn_dist and encoder_states\n # ecnoder_sates = [batch , seq_len , 1 , encoder_output_size], attn_dist = [batch, seq_len, 1, 1]\n context_vector = math_ops.reduce_sum(array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) * encoder_states, [1, 2]) # shape (batch_size, enc_output_size).\n context_vector = array_ops.reshape(context_vector, [-1, enc_output_size])\n\n if hier:\n return context_vector, attn_dist, coverage, attn_dist_sec\n else:\n return context_vector, attn_dist, coverage","def mlp_policy_net(x, output_sizes):\n net = hk.Sequential([hk.nets.MLP(output_sizes), jnp.tanh])\n return net(x)","def __init__(self, dim_hv, dim_hw, msg_dim):\n super(PairMessageGenerator, self).__init__()\n self.dim_hv, self.dim_hw, self.msg_dim = dim_hv, dim_hw, msg_dim\n self.in_dim = dim_hv + dim_hw # row * feature_dim, 2048\n self.mlp = nn.Sequential(\n nn.LayerNorm(self.in_dim), # this layer norm is important to create diversity\n nn.Linear(self.in_dim, self.msg_dim),\n nn.LeakyReLU(0.2)\n )","def self_attention_model(self):\n inputs = self.prepare_inputs()\n \n features = self.prepare_features(inputs)\n support_history_context_concat_emb = features['support_history_context_concat_emb']\n support_link_info_concat_emb = features['support_link_info_concat_emb']\n support_start_node_info_concat_emb = features['support_start_node_info_concat_emb']\n support_end_node_info_concat_emb = features['support_end_node_info_concat_emb']\n support_future_context_concat_emb = features['support_future_context_concat_emb']\n support_neighbor_link_info_concat_emb = features['support_neighbor_link_info_concat_emb']\n support_neighbor_start_node_info_concat_emb = features['support_neighbor_start_node_info_concat_emb']\n support_neighbor_end_node_info_concat_emb = features['support_neighbor_end_node_info_concat_emb']\n support_neighbor_link_state_concat_emb = features['support_neighbor_link_state_concat_emb']\n support_y_embed = features['support_duration_delta_concat_emb']\n\n query_history_context_concat_emb = features['query_history_context_concat_emb']\n query_link_info_concat_emb = features['query_link_info_concat_emb']\n query_start_node_info_concat_emb = features['query_start_node_info_concat_emb']\n query_end_node_info_concat_emb = features['query_end_node_info_concat_emb']\n query_future_context_concat_emb = features['query_future_context_concat_emb']\n query_neighbor_link_info_concat_emb = features['query_neighbor_link_info_concat_emb']\n query_neighbor_start_node_info_concat_emb = features['query_neighbor_start_node_info_concat_emb']\n query_neighbor_end_node_info_concat_emb = features['query_neighbor_end_node_info_concat_emb']\n query_neighbor_link_state_concat_emb = features['query_neighbor_link_state_concat_emb']\n \n support_x = self.constgat({\n 'history_context': support_history_context_concat_emb,\n 'link_info': support_link_info_concat_emb,\n 'start_node_info': support_start_node_info_concat_emb,\n 'end_node_info': support_end_node_info_concat_emb,\n 'future_context': support_future_context_concat_emb,\n 'neighbor_link_info': support_neighbor_link_info_concat_emb,\n 'neighbor_start_node_info': support_neighbor_start_node_info_concat_emb,\n 'neighbor_end_node_info': support_neighbor_end_node_info_concat_emb,\n 'neighbor_link_state': support_neighbor_link_state_concat_emb})\n\n query_x = self.constgat({\n 'history_context': query_history_context_concat_emb,\n 'link_info': query_link_info_concat_emb,\n 'start_node_info': query_start_node_info_concat_emb,\n 'end_node_info': query_end_node_info_concat_emb,\n 'future_context': query_future_context_concat_emb,\n 'neighbor_link_info': query_neighbor_link_info_concat_emb,\n 'neighbor_start_node_info': query_neighbor_start_node_info_concat_emb,\n 'neighbor_end_node_info': query_neighbor_end_node_info_concat_emb,\n 'neighbor_link_state': query_neighbor_link_state_concat_emb})\n query_x = query_x + layers.reduce_sum(inputs['query_mask']) * 0.0 + layers.reduce_sum(inputs['support_mask']) * 0.0 # IMPORTANT: for save_inference_model\n\n def forward_attention(indicator, support_x, support_y_embed, support_mask, query_x, query_y, query_mask):\n \"\"\"\n support_indicator: length = support_len\n if attention(support, query), indicator = 0\n if attention(support, support), indicator = 1\n \"\"\"\n support_y_embed = support_y_embed * support_mask\n support_xy = layers.concat([support_x, support_y_embed, indicator], axis=1)\n\n pad_value = layers.assign(input=numpy.array([0.0], dtype=numpy.float32))\n support_pad, support_len = layers.sequence_pad(support_xy, pad_value=pad_value)\n query_pad, query_len = layers.sequence_pad(query_x, pad_value=pad_value)\n\n attention = self.attention(query_pad, support_pad, support_pad, self.hidden_dim, 'meta')\n attention = layers.sequence_unpad(attention, length=query_len)\n pred_input = layers.concat([attention, query_x], axis=1)\n\n pred = self.prepare_preds_with_name(pred_input, 'out_pred')\n label = layers.cast(query_y, dtype='float32')\n label = layers.scale(label, scale=0.01)\n\n loss = layers.huber_loss(pred, label, 1.0) * query_mask\n loss = layers.mean(loss)\n return pred, label, loss\n\n indicator = support_y_embed * 0.0\n pred, label, loss1 = forward_attention(\n indicator, support_x, support_y_embed, inputs['support_mask'], \n query_x, inputs['query_duration_delta'], 1.0)\n indicator = support_y_embed * 1.0\n _, _, loss2 = forward_attention(\n indicator, support_x, support_y_embed, inputs['support_mask'], \n support_x, inputs['support_duration_delta'] - 120, (inputs['support_mask'] * (-1.0) + 1))\n loss = loss1 + loss2\n return pred, label, loss","def forward(self, input):\n ### YOUR CODE HERE for part 1h\n # sentence_length x batch_size x max_word_length\n orig_size = input.size()\n merged_input = input.reshape(-1, orig_size[-1])\n # s*b x max_word_length\n x_char_embed = self.char_embed(merged_input)\n # s*b x max_word_length x char_embed_size\n x_reshaped = x_char_embed.transpose(-1, -2)\n # s*b x char_embed_size x max_word_length\n x_conv = self.conv(x_reshaped)\n # out s*b x word_embed_size\n x_emb = self.highway(x_conv)\n x_drop = self.dropout(x_emb)\n split_output = x_drop.reshape(list(orig_size[:-1]) + [self.word_embed_size])\n return split_output\n ### END YOUR CODE","def predict(self, first_preprocessed_inputs,second_preprocessed_inputs,third_preprocessed_inputs):\r\n pass","def prelu(input, weight):\n return FunctionLib.apply('PRelu', input.device, [input, weight])","def forward_step(self, batch):\n input_ids = torch.as_tensor(batch.input_ids).to(self.device).reshape((1, -1)) # batch.get('input_ids').to(self.device)\n attention_mask = torch.as_tensor(batch.attention_mask).to(self.device).reshape((1, -1)) # batch.get('attention_mask').to(self.device)\n outputs = self.model(input_ids=input_ids, attention_mask=attention_mask)[0]\n\n _, _, num_label = outputs.shape\n \"\"\"\n outputs : (batch, seq_length, feat_dim) => (seq_length, feat_dim)\n labels : (batch, seq_length) => (seq_length,)\n \"\"\"\n outputs = outputs.view(-1, num_label)\n labels = torch.argmax(outputs, dim=1) # torch.argmax(outputs, dim=1)\n batch_losses = self.criterion(outputs, labels)\n loss = torch.mean(batch_losses) # mean average\n self.batch_output = [input_ids, outputs]\n return loss","def __build_ml_pipeline(self, clf: MultiOutputClassifier) -> Pipeline:\r\n pipeline = Pipeline([\r\n ('features', FeatureUnion(\r\n [('text', Pipeline(\r\n [('text_field_extractor', \r\n basic_utils.TextFieldExtractor('message')), \r\n #('vect', CountVectorizer(tokenizer=clean_text)),\r\n #('tfidf', TfidfTransformer())\r\n ('tfidf', TfidfVectorizer(tokenizer=basic_utils.tokenize, \r\n min_df=.0025, max_df=0.25, \r\n ngram_range=(1,2)))\r\n ])),\r\n \r\n ('numerics', FeatureUnion(\r\n [('text_len', \r\n Pipeline([('text_len_extractor', \r\n basic_utils.NumericFieldExtractor('text_len')), \r\n ('text_len_scaler', StandardScaler())\r\n ])),\r\n ('punt_perc', \r\n Pipeline([('punt_perc_extractor', \r\n basic_utils.NumericFieldExtractor('punt_perc')), \r\n ('punt_perc_scaler', StandardScaler())\r\n ]))\r\n ])),\r\n ('starting_verb', basic_utils.PosFieldExtractor('starting_verb_flag'))\r\n ])),\r\n \r\n ('clf', clf)\r\n ])\r\n \r\n return pipeline","def bias_prior(self):","def on_predict_batch_begin(self, batch, logs=None):","def forward(self, encoder_state, context, context_lens):\n attn = self.attention(context, encoder_state.squeeze(0), context_lens)\n return attn","def attention(inp, scope, e_dim, past, config):\n assert inp.shape.ndims == 3 # input should be of shape [batch, seqlen, embeddings] # [batch, sequence, features]\n assert e_dim % config.num_heads == 0 # embedding can be split in heads\n\n if past is not None:\n assert past.shape.ndims == 5 # [batch, 2, heads, seqlen, emebeddings]\n\n def split_heads(x):\n out = split_into_n_states(x, config.num_heads)\n out = tf.transpose(out, [0, 2, 1, 3])\n return out\n\n def merge_heads(x):\n out = merge_n_states(tf.transpose(x, [0, 2, 1, 3]))\n return out\n\n def mask_attention_weights(w):\n # w should have shape [batches, heads, dst_seq, src_seq], where information flows from scr to dst\n _, _, nd, ns = shapes_list(w)\n b = attention_mask(nd, ns, w.dtype)\n b = tf.reshape(b, [1, 1, nd, ns])\n w = w * b - tf.cast(1e10, w.dtype) * (1 - b)\n return w\n\n def multihead_attention(q, k, v):\n w = tf.matmul(q, k, transpose_b=True)\n w *= tf.rsqrt(tf.cast(v.shape[-1].value, w.dtype))\n\n # mask attention weights\n w = mask_attention_weights(w)\n w = softmax_with_reduce_max(w)\n out = tf.matmul(w, v)\n return out\n\n with tf.variable_scope(scope):\n c = conv1d(inp, 'convolutional_attention', e_dim * 3)\n q, k, v = map(split_heads, tf.split(c, 3, axis=2))\n present = tf.stack([k, v], axis=1)\n if past is not None:\n # there is a stack below it\n pk, pv = tf.unstack(past, axis=1)\n k = tf.concat([pk, k], axis=2)\n v = tf.concat([pv, v], axis=2)\n\n attn = multihead_attention(q, k, v)\n attn = merge_heads(attn)\n\n out = conv1d(attn, 'convolutional_projection', e_dim)\n return out, present","def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, ans_ids=None, ans_mask=None, **kwargs):\n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \n if self.from_scratch:\n imgs = imgs / 255.0\n B_text = token_ids.shape[0]\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\n segment_ids = segment_ids.reshape((-1, ) + segment_ids.shape[2:])\n input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\n losses = dict()\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\n B, D, T, H, W = visual_token.shape\n if B_text != B:\n visual_token = visual_token.view(B_text, -1, D, T, H, W)\n visual_token = visual_token.mean(dim=1)\n \n # text feature #\n text_out_with_mask = self.text_backbone(token_ids, input_mask)\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\n\n # contrastive type finetuning retrieval #\n if self.task == 'retrieval':\n # text_only_out = self.text_backbone(token_ids, input_mask)\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_last_hidden_state, input_mask, token_ids)\n nce_loss = self.loss_func(visual_emb, text_emb)\n losses['retrieval_nce_loss'] = nce_loss \n elif self.task == 'video_qa' or self.task == 'FIB':\n B, D, T, H, W = visual_token.shape\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\n if hasattr(self.qa_head, 'num_labels'):\n num_choices = self.qa_head.num_labels\n visual_token_all = visual_token\n else:\n num_choices = int(token_ids.shape[0] / B)\n visual_token_all = visual_token.unsqueeze(1).expand(-1, num_choices, -1, -1, -1).flatten(0,1)\n\n output = self.multimodal_backbone(visual_token=visual_token_all, text_input_mask=input_mask, text_input_embeds=text_out_last_hidden_state)\n \n if self.answer_mask:\n mask_idx = torch.where(token_ids == 103)\n itm_output = output['t_last_hidden_state'][mask_idx]\n elif self.answer_cls:\n if 'cls_last_hidden_state' in output:\n itm_output = output['cls_last_hidden_state'].squeeze()\n else:\n itm_output = output['t_last_hidden_state'][:, 0]\n if self.itm_head is not None:\n itm_output = self.itm_head(itm_output)\n\n else:\n all_cls_emb = output['last_hidden_state'][:, 0]\n itm_output = self.itm_head(all_cls_emb)\n \n if self.qa_head is not None:\n final_output = self.qa_head(itm_output).view(-1, num_choices)\n final_label = label\n else:\n final_output = itm_output[:, 1]\n final_label = label\n\n\n qa_loss = self.loss_func(final_output, final_label.view(-1))\n losses['qa_loss'] = qa_loss\n\n\n\n return losses","def model(mtf_features, other_features, params, mesh, variable_dtype, context=None):\n\n x, batch_dim, sequence_dim, embd_dim, vocab_dim, embed_sequence_dim = parse_inputs(mtf_features, other_features)\n\n if is_incremental_inference(context):\n # reshape inputs if in inference mode\n x = mtf.gather(x, context.position - 1, sequence_dim)\n x = mtf.reshape(x, [batch_dim])\n\n use_axial_pos_emb = params[\"axial_pos_emb\"] is not None\n\n if not use_axial_pos_emb:\n # Use standard position encoding\n wpe = mtf.get_variable(mesh, \"wpe\", mtf.Shape([embed_sequence_dim, embd_dim]),\n initializer=tf.random_normal_initializer(stddev=0.01),\n master_dtype=variable_dtype.master_dtype,\n slice_dtype=variable_dtype.slice_dtype,\n activation_dtype=variable_dtype.activation_dtype)\n else:\n wpe = axial_positional_emb(embd_dim, mesh, params, variable_dtype)\n\n # Text encoding\n wte = mtf.get_variable(mesh, \"wte\", mtf.Shape([vocab_dim, embd_dim]),\n initializer=tf.random_normal_initializer(stddev=0.02),\n master_dtype=variable_dtype.master_dtype,\n slice_dtype=variable_dtype.slice_dtype,\n activation_dtype=variable_dtype.activation_dtype)\n\n with tf.variable_scope(\"token_embd\"):\n # Text embedding\n h = mtf.gather(wte, x, vocab_dim)\n if params[\"embed_dropout\"] > 0 and params[\"mode\"] == \"train\":\n h = mtf.dropout(h, rate=params[\"embed_dropout\"], name=\"wte_dropout\")\n\n with tf.variable_scope(\"pos_embd\"):\n # Positional embedding\n position_indices = mtf.range(mesh, sequence_dim, tf.int64) if not is_incremental_inference(context) else (\n context.position - 1)\n pos_emb = mtf.gather(wpe, position_indices, wpe.shape[0])\n if params[\"embed_dropout\"] > 0 and params[\"mode\"] == \"train\":\n pos_emb = mtf.dropout(pos_emb, rate=params[\"embed_dropout\"], name=\"wte_dropout\")\n h += pos_emb\n\n aux_losses = 0 # instantiate auxiliary losses (for MOE models)\n\n for layer in range(params[\"n_layer\"]):\n # attn blocks\n share_parameters = exists(params[\"share_parameters\"]) and params[\"share_parameters\"] == True\n block_scope = f\"h{layer}\" if not share_parameters else \"\"\n\n block_fn = block(params=params, scope=block_scope, layer_num=layer,\n bias=other_features[\"attn_bias\"],\n sequence_dim=sequence_dim,\n memory_length_dim=other_features[\"memory_length_dim\"],\n variable_dtype=variable_dtype,\n context=context)\n\n # If true and in train mode, enable gradient checkpointing\n recompute_grad = params[\"recompute_grad\"] and (params[\"mode\"] == \"train\") == True\n h, loss = block_fn(h) if not recompute_grad else mtf.recompute_grad(block_fn, [h])\n aux_losses += loss\n\n no_weight_tie_emb = params[\"no_weight_tie\"] == True\n if no_weight_tie_emb:\n with tf.variable_scope(\"wte_final_linear\"):\n logits = linear(h, \"linear_out\", vocab_dim, variable_dtype=variable_dtype, params=params)\n else:\n # Layer normalize & affine transform\n h = layer_norm(h, \"ln_f\", variable_dtype=variable_dtype)\n seq_dim = sequence_dim if not is_incremental_inference(context) else mtf.Dimension(\"sequence\", 1)\n with tf.variable_scope(\"wte_final_einsum\"):\n # Equivalent to tf.matmul\n logits = mtf.einsum([h, wte], output_shape=[batch_dim, seq_dim, vocab_dim])\n\n if params[\"mode\"] in [\"train\", \"eval\"]:\n labels = mtf_features[\"labels\"]\n z_loss = params.get(\"z_loss\", 1e-4) # an auxiliary loss used to stabilize mtf xentropy\n\n # Go to full precision for the logits \n logits = mtf.cast(logits, tf.float32)\n\n with tf.variable_scope(\"xentropy_final\"):\n loss_batch = mtf.layers.softmax_cross_entropy_with_logits(logits=logits, targets=labels,\n vocab_dim=logits.shape[-1], z_loss=z_loss)\n\n # For non-autoregressive models (masked language modeling training)\n # Make sure labels with padding tokens are not counted in the loss\n if not params[\"causal\"]:\n padding_id = params.get(\"padding_id\", 0)\n loss_batch = mtf.where(mtf.not_equal(labels, padding_id), loss_batch, mtf.zeros_like(loss_batch))\n\n with tf.variable_scope(\"reduce_mean_final\"):\n loss = mtf.reduce_mean(loss_batch)\n\n loss += aux_losses # Add on auxiliary losses (currently only used for MoE)\n loss /= params[\"num_microbatches\"]\n # Convert to train dtype\n loss = mtf.cast(loss, variable_dtype.slice_dtype)\n else:\n loss = None\n loss_batch = None\n\n # Cast back to checkpoint dtype\n logits = mtf.cast(logits, variable_dtype.master_dtype)\n return logits, loss, loss_batch","def forward(self, hg, samp_bias1=None, samp_bias2=None):\n h_1_all = [];h_2_all = [];c_all = [];logits = []\n result = {}\n # process features\n features = hg.srcdata['h']\n feats = self.normal_feat(features, self.meta_paths)\n # shuffled features\n shuf_feats = self.shuf_feats(feats)\n\n for idx, meta_path in enumerate(self.meta_paths):\n new_g = dgl.metapath_reachable_graph(hg, meta_path)\n for i in range(self.sc):\n new_g = dgl.add_self_loop(new_g)\n\n feats[idx] = F.dropout(feats[idx], self.dropout, training=self.training)\n shuf_feats[idx] = F.dropout(shuf_feats[idx], self.dropout, training=self.training)\n\n h_1 = self.gcn[idx](new_g, feats[idx])\n c = self.readout(h_1)\n c = self.readout_act_func(c)\n h_2 = self.gcn[idx](new_g, shuf_feats[idx])\n\n\n logit = self.disc(c, h_1, h_2, samp_bias1, samp_bias2)\n h_1_all.append(h_1.unsqueeze(0))\n h_2_all.append(h_2.unsqueeze(0))\n c_all.append(c)\n logits.append(logit)\n result['logits'] = logits\n\n # Attention or not\n if self.isAttn:\n r\"\"\"\n .. math::\n \\begin{equation}\n \\mathbf{h}_{i}=\\mathcal{Q}\\left(\\left\\{\\mathbf{h}^{(r)} \\mid r \\in \\mathcal{R}\\right\\}\\right)=\\sum_{r \\in \\mathcal{R}} a_{i}^{(r)} \\mathbf{h}^{(r)}\n \\end{equation}\n\n where :math:`a_{i}^{(r)}` denotes the importance of relationr in generating the final embedding of node videfined as:\n\n .. math::\n \\begin{equation}\n a_{i}^{(r)}=\\frac{\\exp \\left(\\mathbf{q}^{(r)} \\cdot \\mathbf{h}_{i}^{(r)}\\right)}{\\sum_{r^{\\prime} \\in \\mathcal{R}} \\exp \\left(\\mathbf{q}^{\\left(r^{\\prime}\\right)} \\cdot \\mathbf{h}_{i}^{r^{\\prime}}\\right)}\n \\end{equation}\n \"\"\"\n\n h_1_all_lst = [];h_2_all_lst = [];c_all_lst = []\n for h_idx in range(self.nheads):\n h_1_all_, h_2_all_, c_all_ = self.attn[h_idx](h_1_all, h_2_all, c_all)\n h_1_all_lst.append(h_1_all_);h_2_all_lst.append(h_2_all_); c_all_lst.append(c_all_)\n\n h_1_all = torch.mean(torch.cat(h_1_all_lst, 0), 0).unsqueeze(0)\n h_2_all = torch.mean(torch.cat(h_2_all_lst, 0), 0).unsqueeze(0)\n\n else:\n h_1_all = torch.mean(torch.cat(h_1_all, 0), 0).unsqueeze(0)\n h_2_all = torch.mean(torch.cat(h_2_all, 0), 0).unsqueeze(0)\n\n # Lcs = [Z − AVG { H(r)|r∈ R }]^2 - [Z − AVG { ~H(r)|r∈ R }]^2\n pos_reg_loss = ((self.H - h_1_all) ** 2).sum()\n neg_reg_loss = ((self.H - h_2_all) ** 2).sum()\n reg_loss = pos_reg_loss - neg_reg_loss\n result['reg_loss'] = reg_loss\n\n # semi-supervised module\n if self.isSemi:\n r\"\"\"\n Extension to Semi-Supervised Learning\n\n .. math::\n \\begin{equation}\n \\ell_{\\text {sup }}=-\\frac{1}{\\left|\\mathcal{Y}_{L}\\right|} \\sum_{l \\in \\mathcal{Y}_{L}} \\sum_{i=1}^{c} Y_{l i} \\ln \\hat{Y}_{l i}\n \\end{equation}\n\n Where :math:`mathcal{Y}_{L}` is the set of node indices with labels\n \"\"\"\n semi = self.logistic(self.H).squeeze(0)\n result['semi'] = semi\n\n # result: ['logits','reg_loss','semi']\n return result","def __init__(self, # noqa: R0913\n hid_dim,\n n_heads,\n pf_dim,\n dropout):\n super().__init__()\n\n self.self_attn_layer_norm = nn.LayerNorm(hid_dim)\n self.enc_attn_layer_norm = nn.LayerNorm(hid_dim)\n self.ff_layer_norm = nn.LayerNorm(hid_dim)\n self.self_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\n self.encoder_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\n self.positionwise_feedforward = PositionwiseFeedforwardLayer(hid_dim,\n pf_dim,\n dropout)\n self.dropout = nn.Dropout(dropout)","def _add_bert_self_attention_layer(model_1, model_2):\n\n class Config:\n pass\n\n config = Config()\n config.hidden_size = model_1.all_head_size + model_2.all_head_size\n config.num_attention_heads = (\n model_1.num_attention_heads + model_2.num_attention_heads\n )\n config.output_attentions = model_1.output_attentions\n config.attention_probs_dropout_prob = model_1.dropout.p\n\n return BertSelfAttention(config)","def forward(self,output,target,Pinv,pts_2d,device):\n \n # concatenate a third row of ones with pts_2d\n ones = torch.ones((pts_2d.shape[0],1,pts_2d.shape[-1])).to(device)\n pts_2d = torch.cat((pts_2d,ones),1)\n \n # multiply pts_2d by output and target, respectively\n output = output.unsqueeze(1)\n target = target.unsqueeze(1)\n out_scaled = torch.mul(pts_2d,output)\n targ_scaled = torch.mul(pts_2d,target)\n \n # multiply each result by Pinv\n out_3d = torch.bmm(Pinv,out_scaled)\n targ_3d = torch.bmm(Pinv,targ_scaled)\n loss = nn.SmoothL1Loss()#MSELoss()\n return loss(out_3d,targ_3d)","def forward(self, inp):\n out = self.features(inp)\n out = out.view(out.size(0), -1) # linearized the output of the module 'features'\n out = self.classifier(out)\n return out","def forward(self, enc_states, h_prev):\n\t\talpha_t = torch.bmm(h_prev.transpose(0,1), enc_states.transpose(1,2)) # [batch, 1, seq_len]\n\t\talpha_t = F.softmax(alpha_t, dim=-1)\n\t\tc_t = torch.bmm(alpha_t, enc_states) # [batch, 1, dim]\n\t\treturn c_t","def interAct(model, encoded_query, dataloader_tester, params_runtime , args,mark='Interactive', verbose=False):\n # set model to evaluation mode\n model.eval()\n predictions = []\n for batch in dataloader_tester:\n \n batch = tuple(b.to(params_runtime.device) for b in batch)\n \n inputs = {'input_ids': batch[0],\n 'attention_mask': batch[1],\n }\n \n outputs = model(**inputs)\n logits = outputs[0]\n #loss_val_total += loss.item()\n sm = nn.Softmax(dim=1)\n logits = sm(logits)\n \n logits = logits.detach().cpu().numpy()\n predictions.append(logits)\n predictions = np.concatenate(predictions, axis=0)\n \n with open(args.pretrained_dir+'/params.json', 'r') as fp:\n label_dict = json.load(fp)\n pred_label = get_label(predictions, label_dict)\n \n \n return(pred_label)","def make_mlp_model():\n return snt.Sequential([\n snt.nets.MLP([LATENT_SIZE] * NUM_LAYERS, activate_final=True),\n snt.LayerNorm()\n ])","def __init__(self, is_training, transductive_batch_norm,\n backprop_through_moments, ema_object, embedding_fn, reader,\n weight_decay, num_update_steps, additional_test_update_steps,\n first_order, alpha, train_batch_norm, depth_multiplier, debug,\n zero_fc_layer, proto_maml_fc_layer_on_query_set,\n proto_maml_fc_layer_on_support_set, proto_maml_fc_layer_init):\n super(MAMLLearner, self).__init__(is_training, transductive_batch_norm,\n backprop_through_moments, ema_object,\n embedding_fn, reader)\n\n maml_embedding_fns = [\n four_layer_convnet_maml,\n resnet_maml,\n ]\n if not any(self.embedding_fn is maml_fn for maml_fn in maml_embedding_fns):\n raise ValueError('MAML requires a specific architecture to work.')\n\n # For aggregating statistics later.\n self.test_targets = self.data.test_labels\n self.way = compute_way(self.data)\n\n self.weight_decay = weight_decay\n self.alpha = alpha\n self.num_update_steps = num_update_steps\n self.additional_test_update_steps = additional_test_update_steps\n self.first_order = first_order\n self.train_batch_norm = train_batch_norm\n self.debug_log = debug\n self.depth_multiplier = depth_multiplier\n self.zero_fc_layer = zero_fc_layer\n self.proto_maml_fc_layer_on_query_set = proto_maml_fc_layer_on_query_set\n self.proto_maml_fc_layer_on_support_set = proto_maml_fc_layer_on_support_set\n self.proto_maml_fc_layer_init = proto_maml_fc_layer_init\n\n tf.logging.info('alpha: {}, num_update_steps: {}'.format(\n self.alpha, self.num_update_steps))\n self.forward_pass()","def __init__(self, albert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([albert_layer.attention.query.weight, albert_layer.attention.key.weight, albert_layer.attention.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([albert_layer.attention.query.bias, albert_layer.attention.key.bias, albert_layer.attention.value.bias]))\n self.out_proj_weight = albert_layer.attention.dense.weight\n self.out_proj_bias = albert_layer.attention.dense.bias\n self.linear1_weight = albert_layer.ffn.weight\n self.linear1_bias = albert_layer.ffn.bias\n self.linear2_weight = albert_layer.ffn_output.weight\n self.linear2_bias = albert_layer.ffn_output.bias\n self.norm1_eps = albert_layer.attention.LayerNorm.eps\n self.norm1_weight = albert_layer.attention.LayerNorm.weight\n self.norm1_bias = albert_layer.attention.LayerNorm.bias\n self.norm2_eps = albert_layer.full_layer_layer_norm.eps\n self.norm2_weight = albert_layer.full_layer_layer_norm.weight\n self.norm2_bias = albert_layer.full_layer_layer_norm.bias\n self.num_heads = albert_layer.attention.num_attention_heads\n self.embed_dim = albert_layer.attention.all_head_size\n self.is_last_layer = False\n self.validate_bettertransformer()","def forward(self, output):\n \n hidden_states = self.extract_hidden_states(output)\n \n # Obtaining the attention weights\n weighted_states = self.w1(hidden_states)\n activated_states = self.tanh(weighted_states)\n score_weights = self.w2(activated_states)\n attention_weights = self.softmax(score_weights)\n \n # Applying attention to the matrix with hidden states\n attentional_vector = torch.bmm(torch.transpose(attention_weights,2,1),hidden_states) \n attentional_vector = self.fc(torch.transpose(attentional_vector,2,1)).squeeze(2)\n attentional_vector = self._activation_fn(attentional_vector)\n \n return attentional_vector","def forward(\n self,\n input_ids=None,\n attention_mask=None,\n token_type_ids=None,\n masked_lm_labels=None,\n ent_masked_lm_labels=None,\n prot_masked_lm_labels=None,\n return_dict=None,\n head_mask=None,\n ):\n # No backpropagation is needed for getting the initial embeddings from the backbones\n with torch.no_grad():\n # 1. Use the LM backbone to get the pre-trained token embeddings\n # batch x number_text_tokens x hidden_size\n # The first element of the returned tuple from the LM backbone forward() pass is the sequence of hidden\n # states\n text_embeddings = torch.cat(\n [\n self.lm_backbone(\n input_ids[\n :, i * (self.kg_start_idx // 3) : (i + 1) * (self.kg_start_idx // 3)\n ]\n )[0]\n for i in range(3)\n ],\n dim=1,\n )\n\n # 2. Use the KG backbone to obtain the pre-trained entity embeddings\n # batch x number_kg_tokens x hidden_size\n ent_embeddings = torch.stack(\n [\n # for each numeric index in the random walks sequence: get the embedding vector from the KG backbone\n torch.stack([self.kg_backbone[i.item()] for i in j])\n # for each example in the batch: get the random walks sequence\n for j in input_ids[:, self.kg_start_idx : self.prot_start_idx]\n ],\n )\n # 3. Use the Prot backbone to obtain the pre-trained entity embeddings\n # batch x number_prot_tokens x prot_hidden_size (prot_hidden_size != hidden_size)\n prot_embeddings_original_dim = self.prot_backbone(input_ids[:, self.prot_start_idx :])[\n 0\n ]\n\n # Additional layer to project prot_hidden_size onto hidden_size\n prot_embeddings = self.prot_to_lm_hidden_linear(prot_embeddings_original_dim)\n\n # Concatenate token, KG and prot embeddings obtained from the LM, KG and prot backbones and cast to float\n # batch x seq_len x hidden_size\n inputs_embeds = (\n torch.cat(\n [\n text_embeddings,\n ent_embeddings.to(text_embeddings.device),\n prot_embeddings.to(text_embeddings.device),\n ],\n dim=1,\n )\n .type(torch.FloatTensor)\n .to(self.device)\n )\n\n # Get the hidden states from the basic ProtSTonKGs Transformer layers\n # batch x seq_len x hidden_size\n outputs = self.bert(\n inputs_embeds=inputs_embeds,\n attention_mask=attention_mask,\n return_dict=None,\n )\n # batch x seq_len x hidden_size\n # sequence_output, pooled_output = outputs.last_hidden_state, outputs.pooler_output\n sequence_output, pooled_output = outputs[:2]\n\n # Generate the prediction scores (mapping to text and entity vocab sizes + NSP) for the training objectives\n # prediction_scores = Text MLM, entity \"MLM\" and protein \"MLM\" scores\n prediction_scores, _ = self.cls(sequence_output, pooled_output)\n # The custom STonKGsELMPredictionHead returns a triple of prediction scores for tokens, entities,\n # and protein sequences, respectively\n (\n token_prediction_scores,\n entity_predictions_scores,\n prot_predictions_scores,\n ) = prediction_scores\n\n # Calculate the loss\n total_loss = None\n if (\n masked_lm_labels is not None\n and ent_masked_lm_labels is not None\n and prot_masked_lm_labels is not None\n ):\n loss_fct = nn.CrossEntropyLoss()\n # 1. Text-based MLM\n masked_lm_loss = loss_fct(\n token_prediction_scores.view(-1, self.config.lm_vocab_size),\n masked_lm_labels.view(-1),\n )\n # 2. Entity-based masked \"language\" (entity) modeling\n ent_masked_lm_loss = loss_fct(\n entity_predictions_scores.view(-1, self.config.kg_vocab_size),\n ent_masked_lm_labels.view(-1),\n )\n # 3. Protein-based masked \"language\" (entity) modeling\n prot_masked_lm_loss = loss_fct(\n prot_predictions_scores.view(-1, self.config.prot_vocab_size),\n prot_masked_lm_labels.view(-1),\n )\n # Total loss = the sum of the individual training objective losses\n total_loss = masked_lm_loss + ent_masked_lm_loss + prot_masked_lm_loss\n\n if not return_dict:\n output = prediction_scores + outputs[2:]\n return ((total_loss,) + output) if total_loss is not None else output\n\n return BigBirdForPreTrainingOutputWithPooling(\n loss=total_loss,\n prediction_logits=prediction_scores,\n hidden_states=sequence_output,\n attentions=outputs.attentions,\n pooler_output=pooled_output,\n )"],"string":"[\n \"def attention_module_multi_head(roi_feat, position_embedding,\\n nongt_dim, fc_dim, feat_dim,\\n dim=(1024, 1024, 1024),\\n group=16, index=1):\\n dim_group = (dim[0] / group, dim[1] / group, dim[2] / group)\\n nongt_roi_feat = mx.symbol.slice_axis(data=roi_feat, axis=0, begin=0, end=nongt_dim)\\n # [num_rois * nongt_dim, emb_dim]\\n position_embedding_reshape = mx.sym.Reshape(position_embedding, shape=(-3, -2))\\n # position_feat_1, [num_rois * nongt_dim, fc_dim]\\n position_feat_1 = mx.sym.FullyConnected(name='pair_pos_fc1_' + str(index),\\n data=position_embedding_reshape,\\n num_hidden=fc_dim)\\n position_feat_1_relu = mx.sym.Activation(data=position_feat_1, act_type='relu')\\n # aff_weight, [num_rois, nongt_dim, fc_dim]\\n aff_weight = mx.sym.Reshape(position_feat_1_relu, shape=(-1, nongt_dim, fc_dim))\\n # aff_weight, [num_rois, fc_dim, nongt_dim]\\n aff_weight = mx.sym.transpose(aff_weight, axes=(0, 2, 1))\\n\\n # multi head\\n assert dim[0] == dim[1], 'Matrix multiply requires same dimensions!'\\n q_data = mx.sym.FullyConnected(name='query_' + str(index),\\n data=roi_feat,\\n num_hidden=dim[0])\\n q_data_batch = mx.sym.Reshape(q_data, shape=(-1, group, dim_group[0]))\\n q_data_batch = mx.sym.transpose(q_data_batch, axes=(1, 0, 2))\\n k_data = mx.symbol.FullyConnected(name='key_' + str(index),\\n data=nongt_roi_feat,\\n num_hidden=dim[1])\\n k_data_batch = mx.sym.Reshape(k_data, shape=(-1, group, dim_group[1]))\\n k_data_batch = mx.sym.transpose(k_data_batch, axes=(1, 0, 2))\\n v_data = nongt_roi_feat\\n # v_data = mx.symbol.FullyConnected(name='value_'+str(index)+'_'+str(gid), data=roi_feat, num_hidden=dim_group[2])\\n aff = mx.symbol.batch_dot(lhs=q_data_batch, rhs=k_data_batch, transpose_a=False, transpose_b=True)\\n # aff_scale, [group, num_rois, nongt_dim]\\n aff_scale = (1.0 / math.sqrt(float(dim_group[1]))) * aff\\n aff_scale = mx.sym.transpose(aff_scale, axes=(1, 0, 2))\\n\\n assert fc_dim == group, 'fc_dim != group'\\n # weighted_aff, [num_rois, fc_dim, nongt_dim]\\n weighted_aff = mx.sym.log(mx.sym.maximum(left=aff_weight, right=1e-6)) + aff_scale\\n aff_softmax = mx.symbol.softmax(data=weighted_aff, axis=2, name='softmax_' + str(index))\\n # [num_rois * fc_dim, nongt_dim]\\n aff_softmax_reshape = mx.sym.Reshape(aff_softmax, shape=(-3, -2))\\n # output_t, [num_rois * fc_dim, feat_dim]\\n output_t = mx.symbol.dot(lhs=aff_softmax_reshape, rhs=v_data)\\n # output_t, [num_rois, fc_dim * feat_dim, 1, 1]\\n output_t = mx.sym.Reshape(output_t, shape=(-1, fc_dim * feat_dim, 1, 1))\\n # linear_out, [num_rois, dim[2], 1, 1]\\n linear_out = mx.symbol.Convolution(name='linear_out_' + str(index), data=output_t,\\n kernel=(1, 1), num_filter=dim[2], num_group=fc_dim)\\n output = mx.sym.Reshape(linear_out, shape=(0, 0))\\n return output\",\n \"def build_model_multihead_attention_multiscaleCNN4_covermore(self,\\n dim_attention,headnum,\\n embedding_vec,\\n load_weights=False, weight_dir=None,\\n nb_filters=32,filters_length1=1,\\n filters_length2=5,\\n filters_length3=10,\\n pooling_size=3,\\n drop_input=0,\\n drop_cnn=0.2,\\n drop_flat=0,\\n W1_regularizer=0.005,\\n W2_regularizer=0.005,\\n Att_regularizer_weight=0.0005,\\n BatchNorm=False,\\n fc_dim = 50,\\n fcnum=0,\\n posembed=False,\\n pos_dmodel=40,\\n pos_nwaves = 20,\\n posmod = 'concat',\\n regularfun=1,\\n huber_delta=1,\\n activation='gelu',\\n activationlast='gelu',\\n add_avgpooling = False,\\n poolingmod=1,\\n normalizeatt=False,\\n regressionmodel=False,\\n attmod = \\\"softmax\\\",\\n sharp_beta=1,\\n lr = 0.001 \\n ):\\n ###print('Advanced Masking')\\n def mask_func(x):\\n return x[0] * x[1]\\n \\n ###print(posembed)\\n ###print(posmod)\\n input = Input(shape=(self.max_len,), dtype='int8')\\n input_mask = Input(shape=([int(self.max_len/pooling_size), 1]), dtype='float32')\\n embedding_layer = Embedding(len(embedding_vec), len(embedding_vec[0]), weights=[embedding_vec],\\n input_length=self.max_len,\\n trainable=False)\\n embedding_output = Dropout(drop_input)(embedding_layer(input)) #layer2\\n if 'gelu' in activation:\\n activationfun=gelu\\n else:\\n activationfun = 'relu'\\n \\n if 'gelu' in activationlast:\\n activationlastfun = gelu\\n else:\\n activationlastfun='relu'\\n \\n ###print(activationfun)\\n ###print(activationlastfun)\\n with tf.name_scope('first_cnn'):\\n first_cnn = Convolution1D(nb_filters, filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN1')(embedding_output) #layer3\\n first_cnn2 = Convolution1D(int(nb_filters/2), filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(first_cnn) #layer5\\n second_cnn = Convolution1D(nb_filters, filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN2')(embedding_output) #layer4\\n second_cnn2 = Convolution1D(int(nb_filters/2), filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(second_cnn)\\n third_cnn = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN3')(embedding_output)\\n \\n third_cnn2 = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(third_cnn)\\n if BatchNorm:\\n first_cnn2 = BatchNormalization()(first_cnn2)\\n second_cnn2 = BatchNormalization()(second_cnn2)\\n third_cnn2 = BatchNormalization()(third_cnn2)\\n \\n if not add_avgpooling:\\n if poolingmod == 1:\\n pooling_layer = MaxPooling1D(pool_length=pooling_size, stride=pooling_size)\\n else:\\n pooling_layer = AveragePooling1D(pool_length=pooling_size, stride=pooling_size)\\n \\n cnn_output1 = Dropout(drop_cnn)(pooling_layer(first_cnn2))\\n cnn_output2 = Dropout(drop_cnn)(pooling_layer(second_cnn2))\\n cnn_output3 = Dropout(drop_cnn)(pooling_layer(third_cnn2))\\n else:\\n first_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\\n first_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\\n cnn_output1 = Dropout(drop_cnn)(concatenate([first_cnn2_max,first_cnn2_avg],axis=-1))\\n second_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\\n second_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\\n cnn_output2 = Dropout(drop_cnn)(concatenate([second_cnn2_max,second_cnn2_avg],axis=-1))\\n third_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\\n third_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\\n cnn_output3 = Dropout(drop_cnn)(concatenate([third_cnn2_max,third_cnn2_avg],axis=-1))\\n \\n \\n \\n if posembed:\\n ##print(posmod)\\n from position_embedding import PositionEmbedding\\n if posmod == 'concat':\\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb1')(cnn_output1)\\n cnn_output1 = concatenate([cnn_output1, pos_emb1], axis=-1)\\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb2')(cnn_output2)\\n cnn_output2 = concatenate([cnn_output2, pos_emb2], axis=-1)\\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb3')(cnn_output3)\\n cnn_output3 = concatenate([cnn_output3, pos_emb3], axis=-1)\\n else:\\n ##print(\\\"yes add posmod\\\")\\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output1)[-1]/2), d_model=int_shape(cnn_output1)[-1],name='pos_emb1')(cnn_output1)\\n cnn_output1 = Add()([cnn_output1, pos_emb1])\\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output2)[-1]/2), d_model=int_shape(cnn_output2)[-1],name='pos_emb2')(cnn_output2)\\n cnn_output2 = Add()([cnn_output2, pos_emb2])\\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output3)[-1]/2), d_model=int_shape(cnn_output3)[-1],name='pos_emb3')(cnn_output3)\\n cnn_output3 = Add()([cnn_output3, pos_emb3])\\n \\n mask_input1 = []\\n mask_input1.append(cnn_output1)\\n mask_input1.append(input_mask)\\n cnn_mask_output1 = Lambda(mask_func)(mask_input1)\\n del mask_input1\\n mask_input2 = []\\n mask_input2.append(cnn_output2)\\n mask_input2.append(input_mask)\\n cnn_mask_output2 = Lambda(mask_func)(mask_input2)\\n del mask_input2\\n mask_input3 = []\\n mask_input3.append(cnn_output3)\\n mask_input3.append(input_mask)\\n cnn_mask_output3 = Lambda(mask_func)(mask_input3)\\n del mask_input3\\n \\n if regularfun==1:\\n regularizerfunction_W1 = regularizers.l1(W1_regularizer)\\n regularizerfunction_W2 = regularizers.l1(W2_regularizer)\\n elif regularfun==2:\\n regularizerfunction_W1 = regularizers.l2(W1_regularizer)\\n regularizerfunction_W2 = regularizers.l2(W2_regularizer)\\n elif regularfun ==3:\\n regularizerfunction_W1 = smoothL1(W1_regularizer,huber_delta)\\n regularizerfunction_W2 = smoothL1(W2_regularizer,huber_delta)\\n \\n with tf.name_scope('multihead_attention'):\\n att1,att1_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att1\\\")(concatenate([cnn_mask_output1, input_mask]))#-5 layer\\n \\n att2,att2_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att2\\\")(concatenate([cnn_mask_output2, input_mask])) #-4 layer\\n \\n att3,att3_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att3\\\")(concatenate([cnn_mask_output3, input_mask])) #-3 layer\\n \\n if BatchNorm:\\n att1 = BatchNormalization()(att1)\\n att2 = BatchNormalization()(att2)\\n att3 = BatchNormalization()(att3)\\n \\n \\n output = Dropout(drop_flat)(Flatten()(concatenate([att1,att2,att3]))) #-2 layer\\n \\n fc = output\\n for _ in range(fcnum):\\n fc = Dense(fc_dim,activation='relu')(fc)\\n fc = Dropout(drop_flat)(fc)\\n \\n with tf.name_scope(''):\\n if regressionmodel:\\n preds = Dense(self.nb_classes,activation='softmax')(fc) #-1 layer\\n else:\\n preds = Dense(self.nb_classes,activation='sigmoid')(fc) #-1 layer\\n \\n self.model = Model(inputs=[input,input_mask], outputs=preds)\\n from keras import optimizers\\n # optim = optimizers.RMSprop()\\n optim = optimizers.Adam(lr=lr, decay=5e-5) #The paper uses a decay rate alpha = alpha/sqrt(t) updted each epoch (t) for the logistic regression demonstration.\\n #optim = optimizers.nadam()\\n #optim = RAdam()\\n if regressionmodel:\\n self.model.compile(loss='kld',optimizer=optim,metrics=['acc'])\\n else:\\n self.model.compile(loss='binary_crossentropy',optimizer=optim,metrics=['binary_accuracy','categorical_accuracy'])\\n \\n \\n \\n if load_weights:\\n self.model.load_weights(weight_dir)\\n \\n self.is_built = True\\n self.bn = False\\n self.model.summary()\",\n \"def block(inp, scope, past, config):\\n with tf.variable_scope(scope):\\n nx = inp.shape[-1].value\\n norm = normalise_tensor(inp, 'ln_1')\\n attn, present = attention(norm, 'attn', nx, past=past, config=config)\\n out = attn + inp\\n norm = normalise_tensor(out, 'ln_2')\\n mlp_out = multilayer_perceptron(norm, 'mlp', nx * 4) # note that hidden dim is 4x\\n out = out + mlp_out\\n return out, present\",\n \"def forward(self, input):\\n x = self.emb(input)\\n output = self.attention(x)\\n # print(\\\"Idggfghrf \\\", output.size())\\n return output\",\n \"def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):\\n super(TransformerBlock, self).__init__()\\n # Attention\\n self.attention = MultiHeadAttention(\\n num_heads=num_heads, key_dim=D, dropout=dropout\\n )\\n self.norm_attention = LayerNormalization(epsilon=norm_eps)\\n\\n # MLP\\n self.mlp = MLPBlock(hidden_layers, dropout)\\n self.norm_mlp = LayerNormalization(epsilon=norm_eps)\",\n \"def __init__(self, nheads, d_model):\\n super(MultiheadAttention, self).__init__()\\n assert d_model % nheads == 0\\n self.d_head = d_model // nheads\\n self.nheads = nheads\\n self.Q_fc = nn.Linear(d_model, d_model, bias=False)\\n self.K_fc = nn.Linear(d_model, d_model, bias=False)\\n self.V_fc = nn.Linear(d_model, d_model, bias=False)\\n self.output_fc = nn.Linear(d_model, d_model, bias=False)\\n self.attn = None\",\n \"def __init__(\\n self,\\n embed_dim,\\n num_heads,\\n dropout=0.,\\n bias=True,\\n kdim=None,\\n vdim=None,\\n ):\\n super(MultiheadAttention, self).__init__()\\n self.embed_dim = embed_dim\\n self.kdim = kdim if kdim is not None else embed_dim\\n self.vdim = vdim if vdim is not None else embed_dim\\n self._qkv_same_embed_dim = self.kdim == embed_dim and self.vdim == embed_dim\\n self.num_heads = num_heads\\n self.dropout = dropout\\n self.head_dim = embed_dim // num_heads\\n if self.head_dim * num_heads != self.embed_dim:\\n raise ValueError('<embed_dim> must be divisible by <num_heads>.')\\n if not self._qkv_same_embed_dim:\\n self.q_proj_weight = Parameter(Tensor(embed_dim, embed_dim))\\n self.k_proj_weight = Parameter(Tensor(embed_dim, self.kdim))\\n self.v_proj_weight = Parameter(Tensor(embed_dim, self.vdim))\\n self.register_parameter('in_proj_weight', None)\\n else:\\n self.in_proj_weight = Parameter(Tensor(3 * embed_dim, embed_dim))\\n self.register_parameter('q_proj_weight', None)\\n self.register_parameter('k_proj_weight', None)\\n self.register_parameter('v_proj_weight', None)\\n if bias:\\n self.in_proj_bias = Parameter(Tensor(3 * embed_dim))\\n else:\\n self.register_parameter('in_proj_bias', None)\\n self.out_proj = Linear(embed_dim, embed_dim, bias=bias)\\n self.reset_parameters()\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\r\\n super(EncoderBlock, self).__init__()\\r\\n self.mha = MultiHeadAttention(dm, h)\\r\\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\\r\\n activation='relu')\\r\\n self.dense_output = tf.keras.layers.Dense(units=dm)\\r\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\r\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\",\n \"def mlp(input_dim,variance=1., weight_variance=None,bias_variance=100.,ARD=False):\\r\\n part = parts.mlp.MLP(input_dim,variance,weight_variance,bias_variance,ARD)\\r\\n return kern(input_dim, [part])\",\n \"def forward(self, h_t, mem):\\n\\n # ==\\n # Compute non-sparse attention\\n cur_mem_size = mem.size(0)\\n batch_size = mem.size(1)\\n\\n # Repeat h_t from (batch, hid_dim) to (mem_size, batch, hid_dim)\\n # one for each memory slot\\n h_repeated = (h_t.clone().unsqueeze(0)\\n .repeat(cur_mem_size, 1, 1))\\n\\n # Concat to [h_t, memory], shape (mem_size, batch, hid_dim*2)\\n mlp_h_attn = torch.cat((h_repeated, mem), dim=2)\\n\\n # (Optional) block past attention gradient\\n if self.block_attn_grad_past:\\n mlp_h_attn = mlp_h_attn.detach()\\n\\n # Compute attention via tanh then matmul with weight matrix\\n # Weight matrix has shape (hid_dim * 2 , 1)\\n # Output (attn_w) has shape (mem_size, batch, 1)\\n mlp_h_attn = self.tanh(mlp_h_attn)\\n attn_w = torch.matmul(mlp_h_attn, self.w_t)\\n\\n # ==\\n # Sparsify attention: set top-k to non-zero and normalize\\n attn_w = attn_w.view(cur_mem_size, batch_size)\\n attn_w = self.sparse_attention(attn_w)\\n attn_w = attn_w.view(cur_mem_size, batch_size, 1)\\n\\n # ==\\n # Extract from memory using attention\\n attn_w_rep = attn_w.repeat(1, 1, self.hidden_size)\\n h_mem_w = attn_w_rep * mem # (mem_size, batch, hid_dim)\\n\\n # Attention-extracted memory information\\n m_t = torch.sum(h_mem_w, dim=0) # (batch, hid_dim)\\n\\n return m_t\",\n \"def forward(self, input):\\n if self.dataset_name == 'mnist':\\n if self.temp_attn:\\n y, attn_weight_list = self.tcanet(input)\\n o = self.decoder(y[:, :, -1])\\n return F.log_softmax(o, dim=1).contiguous()\\n else:\\n y = self.tcanet(input)\\n o = self.decoder(y[:, :, -1])\\n return F.log_softmax(o, dim=1).contiguous()\\n emb = self.drop(self.word_encoder(input))\\n if self.temp_attn:\\n y, attn_weight_list = self.tcanet(emb.transpose(1, 2))\\n y = self.decoder(y.transpose(1, 2))\\n return y.contiguous(), [attn_weight_list[0], attn_weight_list[self.num_levels // 2], attn_weight_list[-1]]\\n else:\\n y = self.tcanet(emb.transpose(1, 2))\\n y = self.decoder(y.transpose(1, 2))\\n return y.contiguous()\",\n \"def __init__(self, cfg):\\n super(MHBCoAtt, self).__init__()\\n self.cfg = cfg\\n # word embedding: q_vocab_size, 1024\\n self.word_embedding = nn.Embedding(cfg.q_vocab_size, cfg.emb_dim)\\n # LSTM\\n if cfg.glove:\\n self.lstm = nn.LSTM(input_size=cfg.emb_dim*2,\\n hidden_size=cfg.hidden_dim,\\n num_layers=cfg.num_layers,\\n batch_first=True)\\n else:\\n self.lstm = nn.LSTM(input_size=cfg.emb_dim,\\n hidden_size=cfg.hidden_dim,\\n num_layers=cfg.num_layers,\\n batch_first=True)\\n\\n self.dropout_l = nn.Dropout(p = 0.3)\\n # question attention\\n self.ques_att_conv1 = nn.Conv2d(cfg.hidden_dim, 512, [1,1])\\n self.ques_att_conv2 = nn.Conv2d(512, 2, [1,1])\\n\\n # question attentive feature fuse with image feature, according to paper: k * o = 5000, k = 5\\n self.ques_proj1 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.img_conv1d = nn.Conv2d(cfg.img_feature_channel, 5000, [1, 1])\\n self.dropout_m = nn.Dropout(p = 0.1)\\n\\n # co-attention conv layers\\n self.co_att_conv1 = nn.Conv2d(1000, 512, [1,1])\\n self.co_att_conv2 = nn.Conv2d(512, 2, [1,1])\\n\\n # co_attentive feature fuse with question attentive feature\\n self.ques_proj2 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.ques_proj3 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.img_proj2 = nn.Linear(2*cfg.img_feature_channel, 5000)\\n self.img_proj3 = nn.Linear(2*cfg.img_feature_channel, 5000)\\n\\n # prediction fully connected layer\\n self.linear_pred = nn.Linear(2000, cfg.a_vocab_size)\",\n \"def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\\n ptm=\\\"bert\\\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\\n super(PTMClassifier_PNN, self).__init__()\\n \\n if old_cols is None:\\n old_cols = []\\n \\n self.old_cols = []\\n \\n self.output_size = output_size\\n self.hidden_size = hidden_size # default\\n self.feature_size = feature_size\\n \\n self.fc1 = nn.Linear(self.hidden_size, 100)\\n self.classifier = nn.Linear(100, self.output_size)\\n \\n if len(old_cols) > 0:\\n self.old_fc1s = ListModule()\\n self.base_1 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n self.base_2 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n \\n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_1)\\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_2)\\n \\n for old_col in old_cols:\\n self.old_fc1s.append(\\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\\n \\n # ptm\\n self.ptm = ptm.lower()\\n assert self.ptm in supported_ptm\\n ptm_ = import_from(\\\"transformers\\\", supported_ptm[self.ptm][0] + \\\"Model\\\")\\n \\n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\\n if tokenizer is not None:\\n self.encoder.resize_token_embeddings(len(tokenizer))\\n self.prob_l = prob_l\\n \\n self.require_proto = require_proto\\n \\n # todo : modify net into one network\\n if self.require_proto:\\n # prototype-based classification\\n self.net = nn.Sequential(\\n self.fc1\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.fc1,\\n self.encoder_adaptor,\\n nn.ReLU(),\\n )\\n else:\\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\\n self.net = nn.Sequential(\\n self.classifier\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.classifier\\n )\\n \\n self.reset_parameters()\",\n \"def __init__(self, config: BertConfig):\\r\\n super().__init__(config)\\r\\n ### YOUR CODE HERE\\r\\n self.num_labels = config.num_labels # [0, 1] (start or end)\\r\\n self.bert = BertModel(config)\\r\\n self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels) # TODO: Not a separate FFN ? (For Start_FFN and End_FFN)\\r\\n\\r\\n ### END YOUR CODE\\r\\n\\r\\n # Don't forget initializing the weights\\r\\n self.init_weights()\",\n \"def call(self, inputs):\\n # Feed attention\\n norm_attention = self.norm_attention(inputs)\\n\\n attention = self.attention(query=norm_attention, value=norm_attention)\\n\\n # Skip Connection\\n attention += inputs\\n\\n # Feed MLP\\n outputs = self.mlp(self.norm_mlp(attention))\\n\\n # Skip Connection\\n outputs += attention\\n\\n return outputs\",\n \"def mlpfwd(self,inputs):\\n\\n self.hidden = np.dot(inputs,self.weights1);\\n self.hidden = 1.0/(1.0+np.exp(-self.beta*self.hidden))\\n self.hidden = np.concatenate((self.hidden,-np.ones((np.shape(inputs)[0],1))),axis=1)\\n\\n outputs = np.dot(self.hidden,self.weights2);\\n\\n # Different types of output neurons\\n if self.outtype == 'linear':\\n \\treturn outputs\\n elif self.outtype == 'logistic':\\n return 1.0/(1.0+np.exp(-self.beta*outputs))\\n elif self.outtype == 'softmax':\\n normalisers = np.sum(np.exp(outputs),axis=1)*np.ones((1,np.shape(outputs)[0]))\\n return np.transpose(np.transpose(np.exp(outputs))/normalisers)\\n else:\\n print \\\"error\\\"\",\n \"def __init__(self, num_heads: int, size: int, dropout: float = 0.1):\\n super(MultiHeadedAttention, self).__init__()\\n\\n assert size % num_heads == 0\\n\\n self.head_size = head_size = size // num_heads\\n self.model_size = size\\n self.num_heads = num_heads\\n\\n self.k_layer = nn.Linear(size, num_heads * head_size)\\n self.v_layer = nn.Linear(size, num_heads * head_size)\\n self.q_layer = nn.Linear(size, num_heads * head_size)\\n\\n self.output_layer = nn.Linear(size, size)\\n self.softmax = nn.Softmax(dim=-1)\\n self.dropout = nn.Dropout(dropout)\",\n \"def get_attention_multiscale_batch(self, X,X_mask):\\n layer = -3\\n attmodel1 = Model(self.model.inputs,self.model.get_layer('att1').output[1])\\n attmodel2 = Model(self.model.inputs,self.model.get_layer('att2').output[1])\\n attmodel3 = Model(self.model.inputs,self.model.get_layer('att3').output[1])\\n return attmodel1.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100),attmodel2.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100),attmodel3.predict([X,X_mask.reshape(-1,X_mask.shape[1],1)],batch_size=100)\",\n \"def mtf_bitransformer_base():\\n hparams = mtf_transformer2_base()\\n hparams.max_length = 256\\n hparams.shared_embedding = True\\n # HYPERPARAMETERS FOR THE LAYER STACKS\\n hparams.add_hparam(\\\"encoder_layers\\\", [\\\"self_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"decoder_layers\\\", [\\\"self_att\\\", \\\"enc_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"encoder_num_layers\\\", 6)\\n hparams.add_hparam(\\\"decoder_num_layers\\\", 6)\\n # number of heads in multihead attention\\n hparams.add_hparam(\\\"encoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"decoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"local_attention_radius\\\", 128)\\n\\n # default of 0 for standard transformer behavior\\n # 1 means a single set of keys and values that are read by all query heads\\n hparams.add_hparam(\\\"encoder_num_memory_heads\\\", 0)\\n hparams.add_hparam(\\\"decoder_num_memory_heads\\\", 0)\\n # share attention keys and values\\n hparams.add_hparam(\\\"encoder_shared_kv\\\", False)\\n hparams.add_hparam(\\\"decoder_shared_kv\\\", False)\\n\\n # Parameters for computing the maximum decode length in beam search.\\n # Maximum decode length is:\\n # min(max_length,\\n # decode_length_multiplier * input_length + decode_length_constant)\\n hparams.add_hparam(\\\"decode_length_multiplier\\\", 1.5)\\n hparams.add_hparam(\\\"decode_length_constant\\\", 10.0)\\n # used during decoding\\n hparams.add_hparam(\\\"alpha\\\", 0.6)\\n hparams.sampling_temp = 0.0\\n return hparams\",\n \"def __init__(self, num_heads: int, size: int, size_v: int, dropout: float = 0.1):\\n super(ContMultiHeadedAttention, self).__init__()\\n\\n assert size % num_heads == 0\\n\\n self.head_size = head_size = size // num_heads\\n self.model_size = size\\n self.num_heads = num_heads\\n\\n self.k_layer = nn.Linear(size, num_heads * head_size)\\n self.v_layer = nn.Linear(size_v, num_heads * head_size)\\n self.q_layer = nn.Linear(size, num_heads * head_size)\\n\\n self.output_layer = nn.Linear(size, size_v)\\n self.softmax = nn.Softmax(dim=-1)\\n self.dropout = nn.Dropout(dropout)\",\n \"def mtf_unitransformer_base():\\n hparams = mtf_transformer2_base()\\n hparams.add_hparam(\\\"autoregressive\\\", True)\\n # HYPERPARAMETERS FOR THE SINGLE LAYER STACK\\n hparams.add_hparam(\\\"layers\\\", [\\\"self_att\\\", \\\"drd\\\"] * 6)\\n # number of heads in multihead attention\\n hparams.add_hparam(\\\"num_heads\\\", 8)\\n # default of 0 for standard transformer behavior\\n # 1 means a single set of keys and values that are read by all query heads\\n hparams.add_hparam(\\\"num_memory_heads\\\", 0)\\n # share attention keys and values\\n hparams.add_hparam(\\\"shared_kv\\\", False)\\n # if nonzero then use local attention\\n hparams.add_hparam(\\\"local_attention_radius\\\", 128)\\n return hparams\",\n \"def __init__(self, emb_size: int = 768, num_heads: int = 12, dropout: float = 0.1):\\n super().__init__()\\n self.emb_size = emb_size\\n self.num_heads = num_heads\\n\\n # Check \\\"head dim\\\" = number of features per head (d_k).\\n self.head_dim = emb_size // num_heads\\n assert self.head_dim * num_heads == self.emb_size, \\\"emb_size must be divisible by num_heads\\\"\\n # Calculate scaling factor.\\n self.scale = 1 / (self.head_dim ** 0.5)\\n\\n # V1: in vanilla self-attention Q,K,V are square matrices.\\n # self.keys = nn.Linear(emb_size, emb_size)\\n # self.queries = nn.Linear(emb_size, emb_size)\\n # self.values = nn.Linear(emb_size, emb_size)\\n \\n # V2: single layer with emb_size, split into num (heads * head_dim) * 3 (Q,K,V).\\n self.qkv = nn.Linear(emb_size, emb_size * 3)\\n\\n # Attention dropout.\\n self.att_drop = nn.Dropout(dropout)\\n self.projection = nn.Linear(emb_size, emb_size)\",\n \"def forward(\\r\\n self,\\r\\n input_ids,\\r\\n attention_mask: torch.Tensor,\\r\\n token_type_ids: torch.Tensor\\r\\n ):\\r\\n ### YOUR CODE HERE\\r\\n output = self.bert(\\r\\n input_ids=input_ids,\\r\\n attention_mask=attention_mask,\\r\\n token_type_ids=token_type_ids,\\r\\n )\\r\\n\\r\\n sequence_output = output[0] # the last hidden state (batch, sequence_length, hidden_size)\\r\\n logits = self.qa_outputs(sequence_output)\\r\\n start_logits, end_logits = logits.split(1, dim=-1)\\r\\n start_logits = start_logits.squeeze(-1)\\r\\n end_logits = end_logits.squeeze(-1)\\r\\n\\r\\n outputs = (start_logits, end_logits) # + output[2:]\\r\\n\\r\\n return outputs\\r\\n ### END YOUR CODE\",\n \"def mtf_transformer_lm_moe():\\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\\n hparams.decoder_layers = [\\\"att\\\", \\\"moe\\\"] * 4\\n moe.set_default_moe_hparams(hparams)\\n hparams.mesh_shape = \\\"all:8\\\"\\n hparams.layout = \\\"batch:all;experts:all\\\"\\n return hparams\",\n \"def exec_attention(self,curr_step): \\n\\n assert(self.curr_step_idx > 0 and self.dlist is not None), \\\"Step Error: Must call init before combine\\\" \\n \\n detectType = curr_step[\\\"detectionNetwork\\\"]\\n paramsFile = curr_step[\\\"paramsFile\\\"]\\n funclist = curr_step[\\\"funclist\\\"]\\n\\n #verify raw data & dlist\\n # self.B_VER(self.sess_path, self.dlist)\\n raw_datadir = self.sess_path\\n dest_datadir = self.sess_path \\n\\n model_dict = {\\\"detectType\\\": detectType, \\\"paramsFile\\\" : paramsFile} \\n\\n for i, folder in enumerate(self.dlist):\\n flist = funclist[i]\\n self.data_utils.DETECT(raw_datadir, folder, dest_datadir, model_dict, flist=[], preview=False)\\n self.default_vis(curr_step)\",\n \"def build_mlp(input_, config):\\n current_input = input_\\n print(current_input)\\n for i in range(len(config['fcc_layers']) - 1):\\n current_input = tf.keras.layers.Dense(\\n units=config['fcc_layers'][i], activation='tanh',\\n name='fcc_layer_{}'.format(i + 1))(current_input)\\n current_input = tf.keras.layers.Dropout(\\n rate=config['dropout'], name='dropout_{}'.format(i + 1))(current_input)\\n cascade_embedding_layer = tf.keras.layers.Dense(\\n units=config['fcc_layers'][-1], activation='tanh',\\n name='cascade_embedding_layer')(current_input)\\n cascade_embedding_layer_do = tf.keras.layers.Dropout(\\n rate=config['dropout'],\\n name='cascade_embedding_dropout')(cascade_embedding_layer)\\n prediction_layer = tf.keras.layers.Dense(\\n units=1, activation='sigmoid',\\n name='prediction_layer')(cascade_embedding_layer_do)\\n return cascade_embedding_layer, prediction_layer\",\n \"def pre_forward(self, *args, **kwargs):\\n batch_size = args[0].shape[0]\\n if not self.instantiated:\\n self.hidden_dim = args[0].shape[-1]\\n self.instantiate(hidden_dim=self.hidden_dim)\\n if self.past_key_reparam is None:\\n past_key = self.past_key\\n else:\\n past_key = self.past_key_reparam\\n if self.past_value_reparam is None:\\n past_value = self.past_value\\n else:\\n past_value = self.past_value_reparam\\n\\n\\n def expand_batchsize(x):\\n x = x.reshape(self.prefix_token_num, self.num_heads, -1).transpose(0,1)\\n x = x.unsqueeze(0).expand(batch_size, *x.shape)\\n return x\\n # from IPython import embe\\n\\n if 'past_key_value' not in kwargs or kwargs['past_key_value'] is None:\\n kwargs['past_key_value'] = (expand_batchsize(past_key), expand_batchsize(past_value))\\n\\n if 'attention_mask' in kwargs and kwargs['attention_mask'] is not None:\\n am = kwargs['attention_mask'] # Should check the format of the attention_mask when moving to a new plm.\\n kwargs['attention_mask'] = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\\n elif len(args) >1: # attention mask is passed via positional argument\\n am = args[1]\\n am = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\\n args = (args[0], am) + args[2:]\\n # from IPython import embed\\n # embed(header = \\\"Herein prefixroberta\\\")\\n return args, kwargs\",\n \"def forward(self, context, question, c_masks):\\n\\n # context_hidden_proj = self.attn_proj(context) # shape (batch_size, context_len, hidden_size)\\n logits_1 = torch.bmm(self.attn_proj_1(context), torch.unsqueeze(question, 2)) # shape (batch_size, context_len, 1)\\n logits_1 = torch.squeeze(logits_1, -1)\\n\\n logits_2 = torch.bmm(self.attn_proj_2(context), torch.unsqueeze(question, 2)) # shape (batch_size, context_len, 1)\\n logits_2 = torch.squeeze(logits_2, -1)\\n\\n log_p1 = masked_softmax(logits_1, c_masks, dim=1, log_softmax=True)\\n log_p2 = masked_softmax(logits_2, c_masks, dim=1, log_softmax=True)\\n #\\n # alpha_t = masked_softmax(scores, c_masks, dim=1) # shape (batch_size, context_len)\\n # output_t = torch.mul(context, alpha_t.unsqueeze(2))\\n #\\n # output_t = torch.bmm(torch.unsqueeze(alpha_t, 1), context) # shape (batch_size, 1, hidden_size)\\n # output_t = output_t.squeeze(1)\\n return log_p1, log_p2\",\n \"def create_quantizable_multihead_attention(module: torch.nn.MultiheadAttention) -> QuantizableMultiheadAttention:\\n # inspect MHA if bias is required.\\n bias = module.in_proj_bias is not None\\n\\n # if bias k/v parameter exist set quantizable MHA to create 3 separate bias tensors as expected.\\n add_bias_kv = module.bias_k is not None and module.bias_v is not None\\n\\n q_MHA = QuantizableMultiheadAttention(embed_dim=module.embed_dim, num_heads=module.num_heads,\\n dropout=module.dropout, bias=bias, add_bias_kv=add_bias_kv,\\n add_zero_attn=module.add_zero_attn, kdim=module.kdim, vdim=module.vdim,\\n batch_first=module.batch_first)\\n\\n # copy over weight and bias tensors\\n with torch.no_grad():\\n if module.in_proj_weight is not None:\\n weights_q, weights_k, weights_v = torch.chunk(module.in_proj_weight.data, 3, dim=0)\\n else:\\n weights_q = module.q_proj_weight.data\\n weights_k = module.k_proj_weight.data\\n weights_v = module.v_proj_weight.data\\n q_MHA.linear_Q.weight.copy_(weights_q)\\n q_MHA.linear_K.weight.copy_(weights_k)\\n q_MHA.linear_V.weight.copy_(weights_v)\\n\\n q_MHA.out_proj.weight.copy_(module.out_proj.weight.data)\\n\\n if bias:\\n bias_q, bias_k, bias_v = torch.chunk(module.in_proj_bias.data, 3, dim=0)\\n if add_bias_kv:\\n bias_k = q_MHA.linear_K.bias.copy_(module.bias_k.data)\\n bias_v = q_MHA.linear_V.bias.copy_(module.bias_v.data)\\n q_MHA.linear_K.bias.copy_(bias_k)\\n q_MHA.linear_V.bias.copy_(bias_v)\\n q_MHA.linear_Q.bias.copy_(bias_q)\\n\\n q_MHA.out_proj.bias.copy_(module.out_proj.bias.data)\\n\\n return q_MHA\",\n \"def forward(self, inputs):\\n location_bias, context_bias, pos_emb, q, k, v, mask = inputs\\n\\n d_feature = q.shape[-1]\\n n_heads = self._n_heads\\n if d_feature % n_heads != 0:\\n raise ValueError(\\n f'Dimensionality of feature embedding ({d_feature}) is not a '\\n f'multiple of the requested number of attention heads ({n_heads}).')\\n\\n if self._mode == 'predict':\\n self._fast_inference_update_state((k, v), self.state)\\n (k, v, _) = self.state\\n\\n per_head_results, dots = DotProductAttention(\\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(q),\\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(k),\\n SplitIntoHeads(n_heads, merged_batch_and_head=False).forward(v),\\n pos_emb.reshape((-1, n_heads, d_feature // n_heads)),\\n context_bias,\\n location_bias,\\n mask,\\n dropout=self._dropout,\\n mode=self._mode,\\n rng=self.rng,\\n chunk_len=self._chunk_len,\\n chunk_offset=self._chunk_offset)\\n if self._mode == 'viz':\\n self.state = dots\\n merged_results = MergeHeads(\\n n_heads, merged_batch_and_head=False).forward(per_head_results)\\n return merged_results, mask\",\n \"def forward(\\n self, inp: torch.Tensor, ucb_alpha: Optional[float] = None\\n ) -> Dict[str, torch.Tensor]:\\n\\n mlp_out = self.deep_represent_layers(inp)\\n # preprocess by DeepRepresent module before fed to LinUCB layer\\n # trainer needs pred_label and mlp_out to update parameters\\n\\n # append a column of ones to features for LinUCB intercept/bias\\n mlp_out_with_ones = torch.cat(\\n (\\n torch.ones(mlp_out.shape[:-1], device=mlp_out.device).unsqueeze(-1),\\n mlp_out,\\n ),\\n -1,\\n )\\n\\n if ucb_alpha is None:\\n ucb_alpha = self.ucb_alpha\\n\\n if not self.nn_e2e:\\n pred_label = torch.matmul(mlp_out_with_ones, self.coefs)\\n else:\\n pred_label = self.linear_layer(mlp_out_with_ones).squeeze(-1)\\n\\n if ucb_alpha != 0:\\n pred_sigma = torch.sqrt(\\n batch_quadratic_form(mlp_out_with_ones, self.inv_avg_A)\\n / torch.clamp(self.sum_weight, min=0.00001)\\n )\\n ucb = pred_label + ucb_alpha * pred_sigma\\n else:\\n pred_sigma = torch.zeros_like(pred_label)\\n ucb = pred_label\\n return {\\n \\\"pred_label\\\": pred_label,\\n \\\"pred_sigma\\\": pred_sigma,\\n \\\"ucb\\\": ucb,\\n \\\"mlp_out_with_ones\\\": mlp_out_with_ones,\\n }\",\n \"def __init__(\\n self,\\n d_model: int,\\n nheads: int,\\n ff_dim: int,\\n bias: bool = False,\\n dropout: float = 0.2,\\n ):\\n super().__init__()\\n self.d_model = d_model\\n self.nheads = nheads\\n self.ff_dim = ff_dim\\n self.norm1 = nn.LayerNorm(d_model)\\n self.norm2 = nn.LayerNorm(d_model)\\n self.self_attention = MultiHeadAttention(d_model, nheads, bias, dropout)\\n self.feedforward = PositionwiseFeedforward(d_model, ff_dim)\",\n \"def forward(self, input):\\n label=np.dot(input,self.w)+self.b\\n return label\",\n \"def update_output(self, latent_mat, weight_mat, y_list):\",\n \"def forward(self, src, src_t, seq, seq_t, seq_e, mask, vars_dict):\\n\\n src_ext = torch.unsqueeze(src, dim=1) # src [B, 1, D]\\n src_e_ph = torch.zeros_like(src_ext)\\n q = torch.cat([src_ext, src_e_ph, src_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\\n k = torch.cat([seq, seq_e, seq_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\\n\\n mask = torch.unsqueeze(mask, dim=2) # mask [B, N, 1]\\n mask = mask.permute([0, 2, 1]) # mask [B, 1, N]\\n\\n # # target-attention\\n output, attn = self.multi_head_target(q=q, k=k, v=k, vars_dict=vars_dict, mask=mask) # output: [B, 1, D + Dt], attn: [B, 1, N]\\n # print('src.shape', src.shape)\\n output = output.squeeze(1)\\n # print('output.shape', output.shape)\\n # print('output', output)\\n # print('output.squeeze().shape', output.shape)\\n attn = attn.squeeze()\\n\\n # output = self.merger(output, src)\\n x = torch.cat([output, src], dim=1)\\n # x = self.layer_norm(x)\\n x = F.relu(F.linear(x, vars_dict['w1_agg_fc']))\\n output = F.linear(x, vars_dict['w2_agg_fc'])\\n\\n return output, attn\",\n \"def pre_forward(self, *args, **kwargs):\\n batch_size = args[0].shape[0]\\n if not self.instantiated:\\n self.hidden_dim = args[0].shape[-1]\\n self.instantiate(hidden_dim=self.hidden_dim)\\n if self.past_key_reparam is None:\\n past_key = self.past_key\\n else:\\n past_key = self.past_key_reparam\\n if self.past_value_reparam is None:\\n past_value = self.past_value\\n else:\\n past_value = self.past_value_reparam\\n\\n # from IPython import embed\\n # embed()\\n def expand_batchsize(x):\\n x = x.reshape(self.prefix_token_num, self.num_heads, -1).transpose(0,1)\\n x = x.unsqueeze(0).expand(batch_size, *x.shape)\\n return x\\n # from IPython import embe\\n\\n if 'past_key_value' not in kwargs or kwargs['past_key_value'] is None:\\n kwargs['past_key_value'] = (expand_batchsize(past_key), expand_batchsize(past_value))\\n\\n if 'attention_mask' in kwargs and kwargs['attention_mask'] is not None:\\n am = kwargs['attention_mask'] # Should check the format of the attention_mask when moving to a new plm.\\n kwargs['attention_mask'] = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\\n elif len(args) >1: # attention mask is passed via positional argument\\n am = args[1]\\n am = torch.cat([-torch.zeros((*am.shape[:-1],self.prefix_token_num), dtype = am.dtype,device=am.device), am], dim=-1)\\n args = (args[0], am) + args[2:]\\n # from IPython import embed\\n # embed(header = \\\"Herein prefixroberta\\\")\\n return args, kwargs\",\n \"def tied_featurize(batch, device, chain_dict, fixed_position_dict=None, omit_AA_dict=None, tied_positions_dict=None, pssm_dict=None, bias_by_res_dict=None, ca_only=False):\\n alphabet = 'ACDEFGHIKLMNPQRSTVWYX'\\n B = len(batch)\\n lengths = np.array([len(b['seq']) for b in batch], dtype=np.int32) #sum of chain seq lengths\\n L_max = max([len(b['seq']) for b in batch])\\n if ca_only:\\n X = np.zeros([B, L_max, 1, 3])\\n else:\\n X = np.zeros([B, L_max, 4, 3])\\n residue_idx = -100*np.ones([B, L_max], dtype=np.int32)\\n chain_M = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\\n pssm_coef_all = np.zeros([B, L_max], dtype=np.float32) #1.0 for the bits that need to be predicted\\n pssm_bias_all = np.zeros([B, L_max, 21], dtype=np.float32) #1.0 for the bits that need to be predicted\\n pssm_log_odds_all = 10000.0*np.ones([B, L_max, 21], dtype=np.float32) #1.0 for the bits that need to be predicted\\n chain_M_pos = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\\n bias_by_res_all = np.zeros([B, L_max, 21], dtype=np.float32)\\n chain_encoding_all = np.zeros([B, L_max], dtype=np.int32) #1.0 for the bits that need to be predicted\\n S = np.zeros([B, L_max], dtype=np.int32)\\n omit_AA_mask = np.zeros([B, L_max, len(alphabet)], dtype=np.int32)\\n # Build the batch\\n letter_list_list = []\\n visible_list_list = []\\n masked_list_list = []\\n masked_chain_length_list_list = []\\n tied_pos_list_of_lists_list = []\\n for i, b in enumerate(batch):\\n if chain_dict != None:\\n masked_chains, visible_chains = chain_dict[b['name']] #masked_chains a list of chain letters to predict [A, D, F]\\n else:\\n masked_chains = [item[-1:] for item in list(b) if item[:10]=='seq_chain_']\\n visible_chains = []\\n masked_chains.sort() #sort masked_chains \\n visible_chains.sort() #sort visible_chains \\n all_chains = masked_chains + visible_chains\\n for i, b in enumerate(batch):\\n mask_dict = {}\\n a = 0\\n x_chain_list = []\\n chain_mask_list = []\\n chain_seq_list = []\\n chain_encoding_list = []\\n c = 1\\n letter_list = []\\n global_idx_start_list = [0]\\n visible_list = []\\n masked_list = []\\n masked_chain_length_list = []\\n fixed_position_mask_list = []\\n omit_AA_mask_list = []\\n pssm_coef_list = []\\n pssm_bias_list = []\\n pssm_log_odds_list = []\\n bias_by_res_list = []\\n l0 = 0\\n l1 = 0\\n for step, letter in enumerate(all_chains):\\n if letter in visible_chains:\\n letter_list.append(letter)\\n visible_list.append(letter)\\n chain_seq = b[f'seq_chain_{letter}']\\n chain_seq = ''.join([a if a!='-' else 'X' for a in chain_seq])\\n chain_length = len(chain_seq)\\n global_idx_start_list.append(global_idx_start_list[-1]+chain_length)\\n chain_coords = b[f'coords_chain_{letter}'] #this is a dictionary\\n chain_mask = np.zeros(chain_length) #0.0 for visible chains\\n if ca_only:\\n x_chain = np.array(chain_coords[f'CA_chain_{letter}']) #[chain_lenght,1,3] #CA_diff\\n if len(x_chain.shape) == 2:\\n x_chain = x_chain[:,None,:]\\n else:\\n x_chain = np.stack([chain_coords[c] for c in [f'N_chain_{letter}', f'CA_chain_{letter}', f'C_chain_{letter}', f'O_chain_{letter}']], 1) #[chain_lenght,4,3]\\n x_chain_list.append(x_chain)\\n chain_mask_list.append(chain_mask)\\n chain_seq_list.append(chain_seq)\\n chain_encoding_list.append(c*np.ones(np.array(chain_mask).shape[0]))\\n l1 += chain_length\\n residue_idx[i, l0:l1] = 100*(c-1)+np.arange(l0, l1)\\n l0 += chain_length\\n c+=1\\n fixed_position_mask = np.ones(chain_length)\\n fixed_position_mask_list.append(fixed_position_mask)\\n omit_AA_mask_temp = np.zeros([chain_length, len(alphabet)], np.int32)\\n omit_AA_mask_list.append(omit_AA_mask_temp)\\n pssm_coef = np.zeros(chain_length)\\n pssm_bias = np.zeros([chain_length, 21])\\n pssm_log_odds = 10000.0*np.ones([chain_length, 21])\\n pssm_coef_list.append(pssm_coef)\\n pssm_bias_list.append(pssm_bias)\\n pssm_log_odds_list.append(pssm_log_odds)\\n bias_by_res_list.append(np.zeros([chain_length, 21]))\\n if letter in masked_chains:\\n masked_list.append(letter)\\n letter_list.append(letter)\\n chain_seq = b[f'seq_chain_{letter}']\\n chain_seq = ''.join([a if a!='-' else 'X' for a in chain_seq])\\n chain_length = len(chain_seq)\\n global_idx_start_list.append(global_idx_start_list[-1]+chain_length)\\n masked_chain_length_list.append(chain_length)\\n chain_coords = b[f'coords_chain_{letter}'] #this is a dictionary\\n chain_mask = np.ones(chain_length) #1.0 for masked\\n if ca_only:\\n x_chain = np.array(chain_coords[f'CA_chain_{letter}']) #[chain_lenght,1,3] #CA_diff\\n if len(x_chain.shape) == 2:\\n x_chain = x_chain[:,None,:]\\n else:\\n x_chain = np.stack([chain_coords[c] for c in [f'N_chain_{letter}', f'CA_chain_{letter}', f'C_chain_{letter}', f'O_chain_{letter}']], 1) #[chain_lenght,4,3] \\n x_chain_list.append(x_chain)\\n chain_mask_list.append(chain_mask)\\n chain_seq_list.append(chain_seq)\\n chain_encoding_list.append(c*np.ones(np.array(chain_mask).shape[0]))\\n l1 += chain_length\\n residue_idx[i, l0:l1] = 100*(c-1)+np.arange(l0, l1)\\n l0 += chain_length\\n c+=1\\n fixed_position_mask = np.ones(chain_length)\\n if fixed_position_dict!=None:\\n fixed_pos_list = fixed_position_dict[b['name']][letter]\\n if fixed_pos_list:\\n fixed_position_mask[np.array(fixed_pos_list)-1] = 0.0\\n fixed_position_mask_list.append(fixed_position_mask)\\n omit_AA_mask_temp = np.zeros([chain_length, len(alphabet)], np.int32)\\n if omit_AA_dict!=None:\\n for item in omit_AA_dict[b['name']][letter]:\\n idx_AA = np.array(item[0])-1\\n AA_idx = np.array([np.argwhere(np.array(list(alphabet))== AA)[0][0] for AA in item[1]]).repeat(idx_AA.shape[0])\\n idx_ = np.array([[a, b] for a in idx_AA for b in AA_idx])\\n omit_AA_mask_temp[idx_[:,0], idx_[:,1]] = 1\\n omit_AA_mask_list.append(omit_AA_mask_temp)\\n pssm_coef = np.zeros(chain_length)\\n pssm_bias = np.zeros([chain_length, 21])\\n pssm_log_odds = 10000.0*np.ones([chain_length, 21])\\n if pssm_dict:\\n if pssm_dict[b['name']][letter]:\\n pssm_coef = pssm_dict[b['name']][letter]['pssm_coef']\\n pssm_bias = pssm_dict[b['name']][letter]['pssm_bias']\\n pssm_log_odds = pssm_dict[b['name']][letter]['pssm_log_odds']\\n pssm_coef_list.append(pssm_coef)\\n pssm_bias_list.append(pssm_bias)\\n pssm_log_odds_list.append(pssm_log_odds)\\n if bias_by_res_dict:\\n bias_by_res_list.append(bias_by_res_dict[b['name']][letter])\\n else:\\n bias_by_res_list.append(np.zeros([chain_length, 21]))\\n\\n \\n letter_list_np = np.array(letter_list)\\n tied_pos_list_of_lists = []\\n tied_beta = np.ones(L_max)\\n if tied_positions_dict!=None:\\n tied_pos_list = tied_positions_dict[b['name']]\\n if tied_pos_list:\\n set_chains_tied = set(list(itertools.chain(*[list(item) for item in tied_pos_list])))\\n for tied_item in tied_pos_list:\\n one_list = []\\n for k, v in tied_item.items():\\n start_idx = global_idx_start_list[np.argwhere(letter_list_np == k)[0][0]]\\n if isinstance(v[0], list):\\n for v_count in range(len(v[0])):\\n one_list.append(start_idx+v[0][v_count]-1)#make 0 to be the first\\n tied_beta[start_idx+v[0][v_count]-1] = v[1][v_count]\\n else:\\n for v_ in v:\\n one_list.append(start_idx+v_-1)#make 0 to be the first\\n tied_pos_list_of_lists.append(one_list)\\n tied_pos_list_of_lists_list.append(tied_pos_list_of_lists)\\n\\n\\n \\n x = np.concatenate(x_chain_list,0) #[L, 4, 3]\\n all_sequence = \\\"\\\".join(chain_seq_list)\\n m = np.concatenate(chain_mask_list,0) #[L,], 1.0 for places that need to be predicted\\n chain_encoding = np.concatenate(chain_encoding_list,0)\\n m_pos = np.concatenate(fixed_position_mask_list,0) #[L,], 1.0 for places that need to be predicted\\n\\n pssm_coef_ = np.concatenate(pssm_coef_list,0) #[L,], 1.0 for places that need to be predicted\\n pssm_bias_ = np.concatenate(pssm_bias_list,0) #[L,], 1.0 for places that need to be predicted\\n pssm_log_odds_ = np.concatenate(pssm_log_odds_list,0) #[L,], 1.0 for places that need to be predicted\\n\\n bias_by_res_ = np.concatenate(bias_by_res_list, 0) #[L,21], 0.0 for places where AA frequencies don't need to be tweaked\\n\\n l = len(all_sequence)\\n x_pad = np.pad(x, [[0,L_max-l], [0,0], [0,0]], 'constant', constant_values=(np.nan, ))\\n X[i,:,:,:] = x_pad\\n\\n m_pad = np.pad(m, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\\n m_pos_pad = np.pad(m_pos, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\\n omit_AA_mask_pad = np.pad(np.concatenate(omit_AA_mask_list,0), [[0,L_max-l]], 'constant', constant_values=(0.0, ))\\n chain_M[i,:] = m_pad\\n chain_M_pos[i,:] = m_pos_pad\\n omit_AA_mask[i,] = omit_AA_mask_pad\\n\\n chain_encoding_pad = np.pad(chain_encoding, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\\n chain_encoding_all[i,:] = chain_encoding_pad\\n\\n pssm_coef_pad = np.pad(pssm_coef_, [[0,L_max-l]], 'constant', constant_values=(0.0, ))\\n pssm_bias_pad = np.pad(pssm_bias_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\\n pssm_log_odds_pad = np.pad(pssm_log_odds_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\\n\\n pssm_coef_all[i,:] = pssm_coef_pad\\n pssm_bias_all[i,:] = pssm_bias_pad\\n pssm_log_odds_all[i,:] = pssm_log_odds_pad\\n\\n bias_by_res_pad = np.pad(bias_by_res_, [[0,L_max-l], [0,0]], 'constant', constant_values=(0.0, ))\\n bias_by_res_all[i,:] = bias_by_res_pad\\n\\n # Convert to labels\\n indices = np.asarray([alphabet.index(a) for a in all_sequence], dtype=np.int32)\\n S[i, :l] = indices\\n letter_list_list.append(letter_list)\\n visible_list_list.append(visible_list)\\n masked_list_list.append(masked_list)\\n masked_chain_length_list_list.append(masked_chain_length_list)\\n\\n\\n isnan = np.isnan(X)\\n mask = np.isfinite(np.sum(X,(2,3))).astype(np.float32)\\n X[isnan] = 0.\\n\\n # Conversion\\n pssm_coef_all = torch.from_numpy(pssm_coef_all).to(dtype=torch.float32, device=device)\\n pssm_bias_all = torch.from_numpy(pssm_bias_all).to(dtype=torch.float32, device=device)\\n pssm_log_odds_all = torch.from_numpy(pssm_log_odds_all).to(dtype=torch.float32, device=device)\\n\\n tied_beta = torch.from_numpy(tied_beta).to(dtype=torch.float32, device=device)\\n\\n jumps = ((residue_idx[:,1:]-residue_idx[:,:-1])==1).astype(np.float32)\\n bias_by_res_all = torch.from_numpy(bias_by_res_all).to(dtype=torch.float32, device=device)\\n phi_mask = np.pad(jumps, [[0,0],[1,0]])\\n psi_mask = np.pad(jumps, [[0,0],[0,1]])\\n omega_mask = np.pad(jumps, [[0,0],[0,1]])\\n dihedral_mask = np.concatenate([phi_mask[:,:,None], psi_mask[:,:,None], omega_mask[:,:,None]], -1) #[B,L,3]\\n dihedral_mask = torch.from_numpy(dihedral_mask).to(dtype=torch.float32, device=device)\\n residue_idx = torch.from_numpy(residue_idx).to(dtype=torch.long,device=device)\\n S = torch.from_numpy(S).to(dtype=torch.long,device=device)\\n X = torch.from_numpy(X).to(dtype=torch.float32, device=device)\\n mask = torch.from_numpy(mask).to(dtype=torch.float32, device=device)\\n chain_M = torch.from_numpy(chain_M).to(dtype=torch.float32, device=device)\\n chain_M_pos = torch.from_numpy(chain_M_pos).to(dtype=torch.float32, device=device)\\n omit_AA_mask = torch.from_numpy(omit_AA_mask).to(dtype=torch.float32, device=device)\\n chain_encoding_all = torch.from_numpy(chain_encoding_all).to(dtype=torch.long, device=device)\\n if ca_only:\\n X_out = X[:,:,0]\\n else:\\n X_out = X\\n return X_out, S, mask, lengths, chain_M, chain_encoding_all, letter_list_list, visible_list_list, masked_list_list, masked_chain_length_list_list, chain_M_pos, omit_AA_mask, residue_idx, dihedral_mask, tied_pos_list_of_lists_list, pssm_coef_all, pssm_bias_all, pssm_log_odds_all, bias_by_res_all, tied_beta\",\n \"def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, \\n mlm_label=None, dvae_imgs=None, v_token_mask=None, hog_features=None, img_metas=None, **kwargs): \\n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \\n if self.from_scratch:\\n imgs = imgs / 255.0\\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\\n text_input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\\n if mlm_label is not None:\\n mlm_label = mlm_label.reshape((-1, ) + mlm_label.shape[2:])\\n\\n\\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\\n\\n B, D, T, H, W = visual_token.shape\\n losses = dict()\\n # -------------- nce loss ------------------- #\\n if hasattr(self, 'ssl_head'):\\n input_ssl_ids = torch.where(mlm_label == -100, token_ids.clone(), mlm_label.clone())\\n input_ssl_mask = text_input_mask.clone()\\n text_only_out = self.text_backbone(input_ssl_ids, input_ssl_mask)\\n # ------------ complete T -------------- #\\n text_out_no_mask = text_only_out['last_hidden_state']\\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_no_mask, input_ssl_mask, input_ssl_ids)\\n\\n\\n # ------------ complete V ---------------- #\\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\\n\\n\\n # ---------- foward mask text input ---------- # \\n text_out_with_mask = self.text_backbone(token_ids, text_input_mask)\\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\\n\\n # ---------- forward mask v input ------------ #\\n visual_token_with_mask, v_mask = self.extract_visual_feat(imgs.clone(), v_token_mask) # b, d, T, h, w\\n visual_token_mask = visual_token_with_mask.view(B, D, T, -1).permute(0, 2, 3, 1)\\n \\n v_fusion_output = self.multimodal_backbone(visual_token=visual_token_mask, text_input_mask=text_input_mask.clone(), text_input_embeds=text_out_no_mask.clone())\\n \\n t_fusion_output = self.multimodal_backbone(visual_token=visual_token, text_input_mask=text_input_mask, text_input_embeds=text_out_last_hidden_state)\\n # for mlm #\\n t_last_hidden_state = t_fusion_output['t_last_hidden_state']\\n\\n\\n\\n\\n\\n # ------------ MLM loss ------------ #\\n\\n if mlm_label is not None and self.mlm_head is not None:\\n # we use mask text for MLM\\n # because we doubt there will be miss interaction between wrong img-text pair \\n # and the model not learn good relationship between vision and language\\n # -------- forward masked text ----------- #\\n mlm_prediction_score = self.mlm_head(t_last_hidden_state)\\n \\n if self.mlm_loss_func is not None:\\n mlm_label_idx = torch.where(mlm_label.view(-1) != -100)\\n mlm_prediction_mask_score = mlm_prediction_score.view(-1, self.text_vocab_size)[mlm_label_idx[0], :]\\n mlm_label_mask = mlm_label.view(-1)[mlm_label_idx]\\n mlm_loss = self.mlm_loss_func(mlm_prediction_mask_score, mlm_label_mask)\\n else:\\n mlm_loss = self.loss_func(mlm_prediction_score.view(-1, self.text_vocab_size), mlm_label.view(-1))\\n losses['mlm_loss'] = mlm_loss\\n\\n\\n # ------- Tri-modal alignment with mask sample and ranking --------- #\\n if self.mlm_ssl_V_head is not None:\\n mlm_visual_feat = v_fusion_output['t_last_hidden_state'][:, 0]\\n mask_visual_recon_emb = self.mlm_ssl_V_head(mlm_visual_feat)\\n mask_word_emb = self.ssl_head.forward_text(text_out_last_hidden_state) if self.use_Cmask else None\\n loss_cvt_rank = self.ssl_loss(visual_emb, text_emb, mask_word_emb, mask_visual_recon_emb)\\n losses.update(loss_cvt_rank)\\n\\n\\n if self.symmetry_rank:\\n mlm_word_feat = t_last_hidden_state[:, 0]\\n mask_word_recon_emb = self.mlm_ssl_T_head(mlm_word_feat)\\n\\n mask_visual_emb = self.ssl_head.forward_vision(visual_token_with_mask) if self.use_Cmask else None\\n \\n loss_ctv_rank = self.ssl_loss(text_emb, visual_emb, mask_visual_emb, mask_word_recon_emb)\\n loss_ctv_rank['v_nce_loss'] = loss_ctv_rank.pop('nce_loss')\\n \\n if self.ssl_loss.use_rank:\\n loss_ctv_rank['rank_v_vm_loss'] = loss_ctv_rank.pop('rank_t_tm_loss')\\n\\n \\n\\n losses.update(loss_ctv_rank)\\n\\n\\n\\n return losses\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\n super().__init__()\\n self.mha = MultiHeadAttention(dm, h)\\n self.ffn = point_wise_feed_forward_network(dm, hidden)\\n\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\",\n \"def classfier_head_multi_proposal(self, statement, statement_mask, targets, ts_labels, ts_labels_mask, max_num_proposal=1, ce_prob_thd=0.01, iou_thd=0.5, extra_span_length=3):\\n bsz, num_a, num_img, num_words = statement_mask.shape\\n statement = statement.view(bsz * num_a * num_img, num_words, -1)\\n statement_mask = statement_mask.view(bsz * num_a * num_img, num_words)\\n statement = self.cls_encoder(statement, statement_mask)\\n max_statement = torch.max(mask_logits(statement, statement_mask.unsqueeze(2)), 1)[0]\\n max_statement_mask = (statement_mask.sum(1) != 0).float().view(bsz, num_a, num_img, 1)\\n max_statement = max_statement.view(bsz * num_a, num_img, -1)\\n t_score_container = []\\n encoded_max_statement_container = []\\n encoded_max_statement = max_statement\\n for layer_idx in range(self.t_iter + 1):\\n encoded_max_statement, prev_t_score = self.residual_temporal_predictor(layer_idx, encoded_max_statement)\\n t_score_container.append(prev_t_score.view(bsz, num_a, num_img, 2))\\n encoded_max_statement_container.append(encoded_max_statement)\\n if self.t_iter > 0:\\n temporal_scores_st_ed = 0.5 * (t_score_container[0] + torch.stack(t_score_container[:1]).mean(0))\\n else:\\n temporal_scores_st_ed = t_score_container[0]\\n temporal_scores_st_ed = mask_logits(temporal_scores_st_ed, ts_labels_mask.view(bsz, 1, num_img, 1))\\n stacked_max_statement = encoded_max_statement_container[0].view(bsz, num_a, num_img, -1)\\n if self.add_local:\\n max_max_statement, targets = self.get_proposals(stacked_max_statement, max_statement_mask, temporal_scores_st_ed, targets, ts_labels, max_num_proposal=max_num_proposal, iou_thd=iou_thd, ce_prob_thd=ce_prob_thd, extra_span_length=extra_span_length)\\n else:\\n max_max_statement = torch.max(mask_logits(stacked_max_statement, max_statement_mask), 2)[0]\\n answer_scores = self.classifier(max_max_statement).squeeze(2)\\n return answer_scores, targets, temporal_scores_st_ed\",\n \"def __init__(self,\\n num_units,\\n line_memory,\\n word_memory=None,\\n soft_weight=None,\\n hierarchy=True,\\n line_memory_sequence_length=None,\\n word_memory_sequence_length=None,\\n scale=False,\\n probability_fn=None,\\n score_mask_value=float(\\\"-inf\\\"),\\n name=\\\"CustomAttention\\\"):\\n # For LuongAttention, we only transform the memory layer; thus\\n # num_units **must** match expected the query depth.\\n if probability_fn is None:\\n probability_fn = nn_ops.softmax\\n wrapped_probability_fn = lambda score: probability_fn(score)\\n super(CustomAttention, self).__init__(\\n query_layer=None,\\n line_memory_layer=layers_core.Dense(\\n num_units, name=\\\"line_memory_layer\\\", use_bias=False),\\n line_memory=line_memory,\\n word_memory=word_memory,\\n probability_fn=wrapped_probability_fn,\\n line_memory_sequence_length=line_memory_sequence_length,\\n word_memory_sequence_length=word_memory_sequence_length,\\n score_mask_value=score_mask_value,\\n name=name)\\n self._num_units = num_units\\n self._scale = scale\\n self._name = name\\n self._hierarchy = hierarchy\\n self._soft_weight = soft_weight\",\n \"def _InitAttentionParams(self, atten_tpl):\\n p = self.params\\n source_atten_tpls = []\\n # Set up each source attention.\\n for i in range(p.num_source):\\n src_key = 'source_%d' % i\\n src_atten = atten_tpl.Copy()\\n src_atten = super()._InitAttentionParams(src_atten)\\n if isinstance(src_atten, list):\\n raise ValueError(\\n 'TransformerMultiSourceAttentionLayer does not support '\\n 'num_heads > 1.')\\n src_atten.name = 'multihead_atten_%s' % src_key\\n source_atten_tpls.append((src_key, src_atten))\\n\\n # Initialize multi-source attention.\\n msa = p.multi_source_atten.Copy()\\n msa.name = 'multi_source_atten'\\n msa.input_dim = p.input_dim\\n msa.hidden_dim = p.hidden_dim\\n msa.source_atten_tpls = source_atten_tpls\\n msa.primary_source_key = 'source_%d' % p.primary_source_index\\n return msa\",\n \"def multi_label_cls_head__post_process(ctx, self, pred, **kwargs):\\n return pred\",\n \"def __init__(self, bert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.q_lin.weight, bert_layer.attention.k_lin.weight, bert_layer.attention.v_lin.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.q_lin.bias, bert_layer.attention.k_lin.bias, bert_layer.attention.v_lin.bias]))\\n self.out_proj_weight = bert_layer.attention.out_lin.weight\\n self.out_proj_bias = bert_layer.attention.out_lin.bias\\n self.linear1_weight = bert_layer.ffn.lin1.weight\\n self.linear1_bias = bert_layer.ffn.lin1.bias\\n self.linear2_weight = bert_layer.ffn.lin2.weight\\n self.linear2_bias = bert_layer.ffn.lin2.bias\\n self.norm1_eps = bert_layer.sa_layer_norm.eps\\n self.norm1_weight = bert_layer.sa_layer_norm.weight\\n self.norm1_bias = bert_layer.sa_layer_norm.bias\\n self.norm2_eps = bert_layer.output_layer_norm.eps\\n self.norm2_weight = bert_layer.output_layer_norm.weight\\n self.norm2_bias = bert_layer.output_layer_norm.bias\\n self.num_heads = bert_layer.attention.n_heads\\n self.embed_dim = bert_layer.attention.dim\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def forward(self, h_text, mask=None):\\n x = h_text\\n for conv_layer in self.conv_stack:\\n x = conv_layer(x, mask=mask)\\n\\n # Autoregressive Prediction\\n p_0 = self.init_state(x)\\n\\n outputs = [p_0]\\n for i in range(x.shape[1]):\\n p_input = torch.cat((x[:, i], outputs[-1]), dim=-1) # [B, 3 * d_model]\\n self.gru_hidden = self.gru_cell(p_input, self.gru_hidden) # [B, 2 * d_model]\\n outputs.append(self.gru_hidden)\\n outputs = torch.stack(outputs[1:], dim=1) # [B, src_len, 2 * d_model]\\n\\n # GMM-MDN\\n w, sigma, mu = self.gmm_mdn(outputs)\\n if mask is not None:\\n w = w.masked_fill(mask.unsqueeze(-1), 0 if self.training else 1e-9) # 1e-9 for categorical sampling\\n sigma = sigma.masked_fill(mask.unsqueeze(-1).unsqueeze(-1), 0)\\n mu = mu.masked_fill(mask.unsqueeze(-1).unsqueeze(-1), 0)\\n\\n return w, sigma, mu\",\n \"def forward(self, *args): # noqa: R0914\\r\\n encoder_out, (hn, cn) = self.unified_encoder(*args)\\r\\n device = hn.device\\r\\n non_sequential_cont_decoded = self.mlp_non_seq_cont(hn)\\r\\n non_sequential_cat_decoded = []\\r\\n for mlp_non_seq_cat in self.mlp_non_seq_cat_list:\\r\\n non_sequential_cat_decoded.append(mlp_non_seq_cat(hn))\\r\\n\\r\\n hn = torch.unsqueeze(hn, 0)\\r\\n cn = torch.unsqueeze(cn, 0)\\r\\n # decoded is the output prediction of timestep i-1 of the decoder\\r\\n decoded = torch.zeros(encoder_out.shape[0], int(\\r\\n self.unified_encoder.seq_cont_count + self.unified_encoder.no_of_embs_seq), device=device)\\r\\n seq_cont_decoded = torch.Tensor(device=device)\\r\\n seq_cat_decoded = []\\r\\n for _ in range(self.unified_encoder.seq_cat_count):\\r\\n seq_cat_decoded.append(torch.Tensor(device=device))\\r\\n\\r\\n for _ in range(encoder_out.shape[1]):\\r\\n decoded, (hn, cn), out_cont, out_cat = self.decoder(decoded, (hn, cn))\\r\\n # Predict all categorical columns\\r\\n out_cat_onehot = []\\r\\n if self.unified_encoder.seq_cat_count != 0:\\r\\n for idx, out in enumerate(out_cat):\\r\\n out_cat_onehot.append(torch.argmax(out, dim=1).unsqueeze(-1))\\r\\n seq_cat_decoded[idx] = torch.cat(\\r\\n [seq_cat_decoded[idx], out.view(out.shape[0], 1, -1)], dim=1)\\r\\n out_cat_onehot = torch.cat(out_cat_onehot, -1)\\r\\n out_cat_embedding = self.unified_encoder.seq_emb_layers(out_cat_onehot)\\r\\n decoded = torch.cat([out_cat_embedding, out_cont], dim=-1)\\r\\n else:\\r\\n decoded = out_cont\\r\\n seq_cont_decoded = torch.cat(\\r\\n [seq_cont_decoded, out_cont.view(out_cont.shape[0], 1, -1)], dim=1)\\r\\n\\r\\n return non_sequential_cont_decoded, non_sequential_cat_decoded, seq_cont_decoded, seq_cat_decoded\",\n \"def __init__(self,\\n num_heads,\\n seq_len,\\n block=16,\\n different_layout_per_head=False,\\n num_local_blocks=4,\\n num_global_blocks=1,\\n attention='bidirectional',\\n horizental_global_attention=False,\\n num_differnt_global_patterns=1):\\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\\n\\n if (seq_len % (block * num_local_blocks) != 0):\\n raise ValueError(\\n f'Sequence Length, {seq_len}, must be dividable by local window, {block*num_local_blocks}!'\\n )\\n self.num_local_blocks = num_local_blocks\\n\\n if (num_local_blocks % num_global_blocks != 0):\\n raise ValueError(\\n f'Number of blocks in a local window, {num_local_blocks}, must be dividable by number of global blocks, {num_global_blocks}!'\\n )\\n self.num_global_blocks = num_global_blocks\\n\\n if (attention != 'unidirectional' and attention != 'bidirectional'):\\n raise NotImplementedError(\\n 'only \\\\\\\"uni/bi-directional\\\\\\\" attentions are supported for now!')\\n self.attention = attention\\n\\n if (attention != 'bidirectional' and horizental_global_attention):\\n raise ValueError(\\n 'only \\\\\\\"bi-directional\\\\\\\" attentions can support horizental global attention!'\\n )\\n self.horizental_global_attention = horizental_global_attention\\n\\n if (num_differnt_global_patterns > 1 and not different_layout_per_head):\\n raise ValueError(\\n f'Number of different layouts cannot be more than one when you have set a single layout for all heads! Set different_layout_per_head to True.'\\n )\\n if (num_differnt_global_patterns > (num_local_blocks // num_global_blocks)):\\n raise ValueError(\\n f'Number of layout versions (num_differnt_global_patterns), {num_differnt_global_patterns}, cannot be larger than number of local window blocks divided by number of global blocks, {num_local_blocks} / {num_global_blocks} = {num_local_blocks//num_global_blocks}!'\\n )\\n self.num_differnt_global_patterns = num_differnt_global_patterns\\n self.make_layout()\",\n \"def create_model(self, model_input, vocab_size, num_frames, **unused_params):\\n num_frames_t=num_frames\\n num_frames = tf.cast(tf.expand_dims(num_frames, 1), tf.float32)\\n feature_size = model_input.get_shape().as_list()[2]\\n iterations=5#150\\n attention_size=8\\n if FLAGS.is_train: \\n iterations=120\\n model_input = utils.SampleRandomFrames(model_input[:,15:,:], num_frames-15-15,\\n iterations)\\n model_input=model_input+tf.random_normal(shape=tf.shape(model_input), mean=0.0, stddev=1e-3, dtype=tf.float32)\\n\\n aggregated_model = getattr(video_level_models,\\n FLAGS.video_level_classifier_model)\\n\\n video_attention = MultiAttentionLayers(1024,iterations,256,attention_size)#256\\n audio_attention = MultiAttentionLayers(128,iterations,256/4,attention_size)#256/4\\n\\n model_input = slim.batch_norm(\\n model_input,\\n center=True,\\n scale=True,\\n is_training=True,\\n scope=\\\"model_input_bn\\\")\\n\\n with tf.variable_scope(\\\"video_Attention\\\"):\\n attention_video = video_attention.forward(model_input[:,:,0:1024]) \\n with tf.variable_scope(\\\"audio_Attention\\\"):\\n attention_audio = audio_attention.forward(model_input[:,:,1024:])\\n\\n pooled=tf.concat([attention_video,attention_audio],axis=1)\\n #instance_att#tf.reduce_mean(pooledi,axis=1)\\n\\n print('pooled is',pooled)\\n pooled=tf.reshape(tf.transpose(pooled,perm=[0,2,1]),[-1,1152])\\n dr2 = tf.get_variable(\\\"dr2\\\",\\n [feature_size,1024],\\n initializer = tf.random_normal_initializer(stddev=1 / math.sqrt(feature_size)))\\n pooled=tf.matmul(pooled,dr2)\\n\\n pooled = slim.batch_norm(\\n pooled,\\n center=True,\\n scale=True,\\n is_training=True,\\n scope=\\\"pooled_bn\\\")\\n\\n gating_weights = tf.get_variable(\\\"gating_weights_2\\\",\\n [1024, 1024],\\n initializer = tf.random_normal_initializer(stddev=1 / math.sqrt(1024))) \\n gates = tf.matmul(pooled, gating_weights) \\n gates = slim.batch_norm(\\n gates,\\n center=True,\\n scale=True,\\n is_training=True,\\n scope=\\\"gating_bn\\\")\\n gates = tf.sigmoid(gates)\\n pooled = tf.multiply(pooled,gates)\\n\\n results_temp=aggregated_model().create_model(\\n model_input=pooled, vocab_size=vocab_size, **unused_params)\\n results_temp['predictions']=tf.reduce_max(tf.reshape(results_temp['predictions'],[-1,attention_size,vocab_size]),axis=1)\\n print(results_temp)\\n return results_temp\",\n \"def __init__(self, args, number_of_labels, number_of_features,adj):\\n super(SpGAT, self).__init__()\\n self.args=args\\n \\n self.number_of_labels = number_of_labels\\n self.number_of_features = number_of_features\\n self.device = args.device\\n self.adj= sparse_mx_to_torch_sparse_tensor(adj).to(self.device).to_dense()\\n self.attentions = [SpGraphAttentionLayer(number_of_features, \\n args.hidden, \\n dropout=args.dropout, \\n alpha=args.alpha, \\n concat=True) for _ in range(args.nheads)]\\n for i, attention in enumerate(self.attentions):\\n self.add_module('attention_{}'.format(i), attention)\\n\\n self.out_att = SpGraphAttentionLayer(args.hidden * args.nheads, \\n args.Q, \\n dropout=args.dropout, \\n alpha=args.alpha, \\n concat=False)\",\n \"def __init__(self, # noqa: R0913\\n hid_dim,\\n n_heads,\\n pf_dim,\\n dropout):\\n\\n super().__init__()\\n\\n self.self_attn_layer_norm = nn.LayerNorm(hid_dim)\\n self.ff_layer_norm = nn.LayerNorm(hid_dim)\\n self.self_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\\n self.positionwise_feedforward = PositionwiseFeedforwardLayer(hid_dim,\\n pf_dim,\\n dropout)\\n self.dropout = nn.Dropout(dropout)\",\n \"def mtf_transformer2_base():\\n hparams = common_hparams.basic_params1()\\n\\n hparams.add_hparam(\\\"d_model\\\", 1024)\\n hparams.batch_size = 4\\n hparams.max_length = 1024\\n hparams.label_smoothing = 0.0\\n # a small positive value - this seems important for stability when training\\n # with bfloat16 activations.\\n hparams.add_hparam(\\\"z_loss\\\", 1e-4)\\n\\n # hparams applying to both encoder and decoder layer stacks.\\n hparams.add_hparam(\\\"d_ff\\\", 2048)\\n hparams.add_hparam(\\\"d_kv\\\", 128)\\n hparams.add_hparam(\\\"attention_dropout\\\", 0.0)\\n hparams.add_hparam(\\\"relu_dropout\\\", 0.0)\\n hparams.del_hparam(\\\"num_heads\\\")\\n hparams.del_hparam(\\\"num_hidden_layers\\\")\\n hparams.layer_prepostprocess_dropout = 0.0\\n hparams.add_hparam(\\\"extra_logit\\\", False)\\n # number of experts for moe_1d\\n hparams.moe_num_experts = 32\\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\\n hparams.add_hparam(\\\"moe_expert_x\\\", 8)\\n hparams.add_hparam(\\\"moe_expert_y\\\", 4)\\n hparams.add_hparam(\\\"moe_hidden_size\\\", 32768)\\n\\n # round up vocab sizes to be a multiple of this value\\n hparams.vocab_divisor = 128\\n\\n hparams.optimizer = \\\"Adafactor\\\"\\n hparams.learning_rate_schedule = \\\"rsqrt_decay*linear_decay\\\"\\n hparams.learning_rate_warmup_steps = 10000\\n hparams.add_hparam(\\\"master_dtype\\\", \\\"bfloat16\\\")\\n hparams.add_hparam(\\\"slice_dtype\\\", \\\"float32\\\")\\n hparams.activation_dtype = \\\"bfloat16\\\"\\n\\n # 8-way model-parallelism\\n hparams.add_hparam(\\\"mesh_shape\\\", \\\"model:8\\\")\\n hparams.add_hparam(\\\"layout\\\", \\\"batch:batch;vocab:model;d_ff:model;heads:model\\\")\\n\\n # If nonzero, we split the batch across two tensor-dimensions named\\n # \\\"outer_batch\\\" and \\\"inner_batch\\\", allowing for splitting across two mesh\\n # dimensions. This is necessary for hierarchical mixture of experts.\\n # The two tensor dimensions have sizes hparams.outer_batch_size and\\n # hparams.batch_size // hparams.outer_batch_size.\\n hparams.add_hparam(\\\"outer_batch_size\\\", 0)\\n\\n hparams.shared_embedding_and_softmax_weights = False\\n # length for training or decoding - defaults to max_length\\n hparams.add_hparam(\\\"length\\\", 0)\\n\\n # These parameters make Transformer model compatible with mtf\\n # Do not override these.\\n hparams.no_data_parallelism = True\\n hparams.use_fixed_batch_size = True\\n hparams.add_hparam(\\\"mtf_mode\\\", True)\\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\\n hparams.bottom = {\\n \\\"inputs\\\": modalities.identity_bottom,\\n \\\"targets\\\": modalities.identity_bottom,\\n }\\n hparams.top = {\\n \\\"targets\\\": modalities.identity_top,\\n }\\n hparams.add_hparam(\\\"beam_size\\\", 1)\\n\\n # If this is True, then in a packed dataset (where exaples are concatenated\\n # to form longer examples) we use the global position (within the concatenated\\n # sequence) to compute the positional embedding, instead of the position\\n # within the individual sequence. This is counterintuitive, but for some\\n # reason, it keeps the model from diverging.\\n hparams.add_hparam(\\\"use_global_position_in_packed_sequence\\\", True)\\n\\n return hparams\",\n \"def forward(self, input, targets=None, get_head_out=False):\\n fpn_out = self.fpn(input)\\n\\n img_height, img_width = input.shape[2:]\\n strides = self.fpn.strides\\n hws = [level_out.shape[2:] for level_out in fpn_out]\\n max_box_sides = [256, 128, 64, 32, 16]\\n if self.levels is not None:\\n max_box_sides = [max_box_sides[l] for l in self.levels]\\n self.pyramid_shape = [\\n (s, m, h, w) for s, m, (h, w) in zip(strides, max_box_sides, hws)]\\n\\n head_out = []\\n for i, level_out in enumerate(fpn_out):\\n head_out.append(self.head(level_out, self.scale_params[i]))\\n\\n if targets is None:\\n out = decode_batch_output(head_out, self.pyramid_shape, img_height, img_width)\\n if get_head_out:\\n return out, head_out\\n return out\\n\\n loss_dict = fcos_batch_loss(\\n head_out, targets, self.pyramid_shape, self.num_labels)\\n if get_head_out:\\n return loss_dict, head_out\\n return loss_dict\",\n \"def __init__(self, nfeat, nhid, nclass, dropout, alpha, nheads, direct=False, supervised=True):\\n super(SpGAT, self).__init__()\\n self.dropout = dropout\\n\\n self.attentions = [SpGraphAttentionLayer(nfeat,\\n nhid,\\n dropout=dropout,\\n alpha=alpha,\\n concat=True) for _ in range(nheads)]\\n for i, attention in enumerate(self.attentions):\\n self.add_module('attention_{}'.format(i), attention)\\n\\n self.out_att = SpGraphAttentionLayer(nhid * nheads,\\n nhid,\\n dropout=dropout,\\n alpha=alpha,\\n concat=False)\\n self.fm = FMLayer(nfeat, nhid)\\n self.supervised = supervised\\n if self.supervised:\\n n_output = nclass\\n else:\\n n_output = nhid\\n\\n self.final_linear = nn.Linear(2 * nhid, n_output)\\n self.final_linear_single = nn.Linear(nhid, n_output)\\n # self.final_linear = nn.Linear(nhid, nclass)\\n self.pa_attn = AttentionalFactorizationMachine(nhid, nhid, dropouts=[0.2, 0.2])\\n self.direct = direct\",\n \"def forward(self, text, prev_mel):\\n # forward pass through text embedding and get k and v\\n kv = self.t_encoder(text)\\n k = kv[:,:self.hp.d,:]\\n v = kv[:,self.hp.d:,:]\\n # forward pass through audio encoding and get Q\\n q = self.a_encoder(prev_mel)\\n \\n # compute attention\\n a = (k.transpose(2,1)).matmul(q)/np.sqrt(self.hp.d)\\n a = F.softmax(a, dim=1)\\n r = v.matmul(a)\\n \\n # create R' and forward pass through decoder\\n # note that the decoder does not have sigmoid transform at the end, so we are actually getting \\n # ylogit\\n rprime = torch.cat((r, q), dim=1)\\n ylogit = self.decoder(rprime)\\n y = F.sigmoid(ylogit)\\n return y, ylogit, a\",\n \"def mt_step(self, lang1, lang2, lambda_coeff):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n lang1_id = params.lang2id[lang1]\\n lang2_id = params.lang2id[lang2]\\n\\n # generate batch\\n if lang1 == lang2:\\n (x1, len1) = self.get_batch('ae', lang1)\\n (x2, len2) = (x1, len1)\\n (x1, len1) = self.add_noise(x1, len1)\\n else:\\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\\n langs1 = x1.clone().fill_(lang1_id)\\n langs2 = x2.clone().fill_(lang2_id)\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n # key point generate label\\n\\n # cuda\\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\\n\\n # encode source sentence\\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\\n enc1 = enc1.transpose(0, 1)\\n\\n # decode target sentence\\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\\n src_len=len1)\\n\\n # loss\\n # the last word not apply prediction logic\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def __init__(self, bert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.self.query.weight, bert_layer.attention.self.key.weight, bert_layer.attention.self.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.self.query.bias, bert_layer.attention.self.key.bias, bert_layer.attention.self.value.bias]))\\n self.out_proj_weight = bert_layer.attention.output.dense.weight\\n self.out_proj_bias = bert_layer.attention.output.dense.bias\\n self.linear1_weight = bert_layer.intermediate.dense.weight\\n self.linear1_bias = bert_layer.intermediate.dense.bias\\n self.linear2_weight = bert_layer.output.dense.weight\\n self.linear2_bias = bert_layer.output.dense.bias\\n self.norm1_eps = bert_layer.attention.output.LayerNorm.eps\\n self.norm1_weight = bert_layer.attention.output.LayerNorm.weight\\n self.norm1_bias = bert_layer.attention.output.LayerNorm.bias\\n self.norm2_eps = bert_layer.output.LayerNorm.eps\\n self.norm2_weight = bert_layer.output.LayerNorm.weight\\n self.norm2_bias = bert_layer.output.LayerNorm.bias\\n self.num_heads = bert_layer.attention.self.num_attention_heads\\n self.embed_dim = bert_layer.attention.self.all_head_size\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def get_output_for(self, input, **kwargs):\\n\\t\\tbatch_size = input.shape[0]\\n\\t\\tnum_states = self.num_states\\n\\t\\timg_dim = self.img_dim\\n\\t\\tN = self.N\\n\\t\\tattender = self.attender\\n\\t\\ttheano.gradient.grad_clip\\n\\n\\t\\tdef step(c_tm1, h_tm1, att_acc_tm1, input, W, Wg):\\n\\t\\t\\tcenter_y, center_x, delta, sigma, gamma = gp_from_hidden(h_tm1, Wg, img_dim, N)\\n\\t\\t\\tg, att = attender.read(input, center_y, center_x, delta, sigma, gamma) # (batch_size, N, N) and (batch_size, img_dim, img_dim)\\n\\t\\t\\n\\t\\t\\tatt_acc_t = T.clip(att_acc_tm1 + att, 0.0, 1.0)\\t# (batch_size, img_dim, img_dim)\\n\\t\\t\\tr = input[:, :, :img_dim] * (1.0 - att_acc_t) # (batch_size, img_dim, img_dim)\\n\\t\\t\\tR , _ = attender.read(r, *gp_from_hidden(T.zeros((batch_size, 5)), T.eye(5), img_dim, N)) # (batch_size, N, N)\\n\\t\\t\\t\\n\\t\\t\\tflat_g = g.reshape((batch_size, N * N)) # (batch_size, N * N)\\n\\t\\t\\tflat_R = R.reshape((batch_size, N * N)) # (batch_size, N * N)\\n\\t\\t\\t\\n\\t\\t\\t# concatenate gA, gB and h_tm1 to form a single matrix # (batch_size, N * N + N * N + num_states + 1)\\n\\t\\t\\tlstm_inp = T.concatenate([flat_g, flat_R, h_tm1, T.ones((batch_size, 1))], axis=1)\\n\\n\\t\\t\\t# multiply by LSTM weights\\n\\t\\t\\t# (num_states * 4, num_input + num_states + 1) dot (batch_size, N * N + N * N + num_states + 1).T\\n\\t\\t\\tpre_act = T.dot(W, lstm_inp.T) \\t# (4 * num_states, batch_size)\\n\\n\\t\\t\\t# split up to get individual gates\\n\\t\\t\\tz = T.tanh(pre_act[0*num_states:1*num_states]) # (num_states, batch_size)\\n\\t\\t\\ti = T.nnet.sigmoid(pre_act[1*num_states:2*num_states])\\n\\t\\t\\tf = T.nnet.sigmoid(pre_act[2*num_states:3*num_states])\\n\\t\\t\\to = T.nnet.sigmoid(pre_act[3*num_states:4*num_states])\\n\\n\\t\\t\\t# do LSTM update\\n\\t\\t\\tc_t = f * c_tm1.T + i * z\\n\\t\\t\\th_t = o * T.tanh(c_t)\\n\\n\\t\\t\\treturn c_t.T, h_t.T, att_acc_t\\t# 1, 2: (batch_size, num_states); 3, 4: (batch_size, img_dim, img_dim)\\n\\n\\t\\tc0 = T.zeros((batch_size, num_states))\\n\\t\\th0 = T.zeros((batch_size, num_states))\\n\\t\\tatt_acc0 = T.zeros((batch_size, img_dim, img_dim))\\n\\t\\t\\n\\t\\tcells, hiddens, att_acc_T = theano.scan(fn=step, non_sequences=[input, self.W, self.Wg], outputs_info=[c0, h0, att_acc0], \\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\tn_steps=self.num_glimpses, strict=True)[0]\\n\\t\\tif self.final_state_only:\\n\\t\\t\\treturn hiddens[-1]\\n\\t\\telse:\\n\\t\\t\\treturn hiddens\",\n \"def apply_to_batch(self, batch_dict):\\r\\n \\r\\n self._last_batch = batch_dict\\r\\n \\r\\n if isinstance(self.model,NMTModelWithMLTM):\\r\\n y_pred = self.model(x_source=batch_dict['x_source'], \\r\\n x_mltm=batch_dict['x_source_mltm_vector'],\\r\\n x_source_lengths=batch_dict['x_source_length'], \\r\\n target_sequence=batch_dict['x_target'])\\r\\n else:\\r\\n y_pred = self.model(x_source=batch_dict['x_source'], \\r\\n x_source_lengths=batch_dict['x_source_length'], \\r\\n target_sequence=batch_dict['x_target'])\\r\\n self._last_batch['y_pred'] = y_pred\\r\\n \\r\\n attention_batched = np.stack(self.model.decoder._cached_p_attn).transpose(1, 0, 2)\\r\\n self._last_batch['attention'] = attention_batched\",\n \"def sample_beam(model, input_embedding, char2idx, idx2char, k=5, maxlen=30,\\n start='START', use_head=True):\\n with torch.no_grad():\\n device = input_embedding.device\\n softmax = nn.Softmax(dim=1)\\n if use_head:\\n input_embedding = input_embedding.view(1, -1)\\n\\n inp = [torch.LongTensor([char2idx[start]]).to(device)]\\n inp = nn.utils.rnn.pack_sequence(inp)\\n out, hidden = model(input_embedding, inp, use_head=use_head)\\n\\n out = softmax(out.data).view(-1).cpu().numpy()\\n max_k = np.argsort(out)[-k:][::-1]\\n oldprobs = out[max_k]\\n words = [[i] for i in max_k]\\n inp = pack([torch.LongTensor([j]).to(device) for j in max_k])\\n\\n if model.mode == 'LSTM':\\n hidden0 = torch.cat([hidden[0] for i in range(k)], dim=1)\\n hidden1 = torch.cat([hidden[1] for i in range(k)], dim=1)\\n hidden = hidden0, hidden1\\n else:\\n hidden = torch.cat([hidden for i in range(k)], dim=1)\\n WORDS = []\\n for c in range(maxlen):\\n out, hidden = model(hidden, inp, use_head=False)\\n out = softmax(out.data).cpu().numpy()\\n\\n #print(out.shape)\\n inpnp = inp.data.detach().cpu().numpy()\\n done = np.where(inpnp == char2idx['END'])\\n out[done] = 0\\n if len(out[done]) != 0:\\n #print(out[done].shape)\\n for d in done[0]:\\n out[d][char2idx['END']] = 1\\n #print(done)\\n\\n #print(out)\\n #print(out[done])\\n out = (oldprobs.reshape(-1, 1)*out)\\n max_k = np.argsort(out)[:, -k:][:, ::-1]\\n\\n #print(max_k)\\n probs = np.array([out[i][max_k[i]] for i in range(k)])\\n #print(probs)\\n flat = probs.reshape(-1)\\n max_k2 = np.argsort(flat)[::-1][:k]\\n word_inds = max_k2//k\\n next_chars_inds = max_k2%k\\n\\n oldprobs = flat[max_k2]\\n #print(oldprobs)\\n\\n new_words = []\\n new_inp = []\\n for i, word_ind in enumerate(word_inds):\\n next_char = max_k[word_ind][next_chars_inds[i]]\\n if next_char == char2idx['END']:\\n #print(\\\"HIT AN END at word {}\\\".format(word_ind))\\n WORDS.append((words[word_ind], oldprobs[i]))\\n #the_word = words[word_ind]\\n #return ''.join([idx2char[i] for i in the_word])\\n new_inp.append(torch.LongTensor([next_char]).to(device))\\n word = words[word_ind][:]\\n word = word + [next_char]\\n new_words.append(word)\\n words = new_words[:]\\n\\n if model.mode == 'LSTM':\\n h1, h2 = hidden\\n h1, h2 = h1[0][word_inds].view(1, k, -1), h2[0][word_inds].view(1, k, -1)\\n hidden = h1, h2\\n else:\\n hidden = hidden[0][word_inds].view(1, k, -1)\\n\\n\\n inp = pack(new_inp)\\n\\n return [''.join([idx2char[i] for i in word if i != char2idx['END']]) for word in words], oldprobs\",\n \"def forward(self, *args):\\r\\n _, (hn, _) = self.enc(*args)\\r\\n\\r\\n y_pred = self.mlp(hn)\\r\\n return y_pred\",\n \"def mtr_lm_v1():\\n hparams = mtr_lm_dense(0)\\n hparams.layers = ([\\\"local_self_att\\\", \\\"local_self_att\\\", \\\"drd\\\",\\n \\\"self_att\\\", \\\"drd\\\", \\\"local_self_att\\\",\\n \\\"local_self_att\\\", \\\"moe_2d\\\"] * 4)[:-1]\\n hparams.d_kv = 128\\n hparams.moe_expert_x = 8\\n hparams.moe_expert_y = 4\\n hparams.moe_hidden_size = 32768\\n hparams.d_ff = 2048\\n hparams.num_memory_heads = 0\\n hparams.mesh_shape = \\\"b0:4;b1:8\\\"\\n hparams.layout = \\\"outer_batch:b0;inner_batch:b1,expert_x:b1,expert_y:b0\\\"\\n hparams.outer_batch_size = 4\\n return hparams\",\n \"def generate_plots(self, input_data, input_labels=None):\\n super(MlpModel, self).generate_plots(input_data, input_labels)\\n feed_dict = self.get_feed_dict(input_data, input_labels)\\n eval_list = [self.global_step, self.get_encodings(), self.mlp_module.weight_list]\\n train_on_adversarial = feed_dict[self.train_on_adversarial]\\n if(train_on_adversarial):\\n eval_list += [self.adv_module.get_adv_input()]\\n eval_out = tf.compat.v1.get_default_session().run(eval_list, feed_dict)\\n current_step = str(eval_out[0])\\n filename_suffix = \\\"_v\\\"+self.params.version+\\\"_\\\"+current_step.zfill(5)+\\\".png\\\"\\n activity = eval_out[1]\\n fig = pf.plot_activity_hist(activity, title=\\\"Logit Histogram\\\",\\n save_filename=self.params.disp_dir+\\\"act_hist\\\"+filename_suffix)\\n #First layer weights\\n mlp_weights = eval_out[2]\\n w_enc = mlp_weights[0]\\n if self.params.mlp_layer_types[0] == \\\"fc\\\":\\n w_enc_norm = np.linalg.norm(w_enc, axis=0, keepdims=False)\\n # Don't plot weights as images if input is not square\\n w_input_sqrt = np.sqrt(w_enc.shape[0])\\n if(np.floor(w_input_sqrt) == np.ceil(w_input_sqrt)):\\n w_enc = dp.reshape_data(w_enc.T, flatten=False)[0] # [num_neurons, height, width]\\n fig = pf.plot_data_tiled(w_enc, normalize=False,\\n title=\\\"Weights at step \\\"+current_step, vmin=None, vmax=None,\\n save_filename=self.params.disp_dir+\\\"w_enc\\\"+filename_suffix)\\n else: # conv\\n w_enc = np.transpose(dp.rescale_data_to_one(w_enc.T)[0].T, axes=(3,0,1,2))\\n if(w_enc.shape[-1] == 1 or w_enc.shape[-1] == 3):\\n pf.plot_data_tiled(w_enc, normalize=False, title=\\\"Weights at step \\\"+current_step,\\n save_filename=self.params.disp_dir+\\\"w_enc\\\"+filename_suffix)\\n for (w, tf_w) in zip(mlp_weights, self.mlp_module.weight_list):\\n #simplify tensorflow node name to only be the last one\\n w_name = tf_w.name.split(\\\"/\\\")[-1].split(\\\":\\\")[0]\\n num_f = w.shape[-1]\\n w_reshape = np.reshape(w, [-1, num_f])\\n w_norm = np.linalg.norm(w_reshape, axis=0, keepdims=False)\\n fig = pf.plot_bar(w_norm, num_xticks=5,\\n title=w_name+\\\" l2 norm\\\", xlabel=\\\"w index\\\", ylabel=\\\"L2 Norm\\\",\\n save_filename=self.params.disp_dir+\\\"w_norm_\\\"+w_name+filename_suffix)\\n if(train_on_adversarial):\\n adv_input = eval_out[-1]\\n adv_input = dp.reshape_data(adv_input, flatten=False)[0]\\n fig = pf.plot_data_tiled(adv_input, normalize=False,\\n title=\\\"Adv inputs at \\\"+current_step,\\n save_filename=self.params.disp_dir+\\\"adv_input\\\"+filename_suffix)\",\n \"def Attention_block(up_in,down_in,nf):\\n# def MyTile(alpha, channels):\\n# alpha = tf.tile(alpha, [1, 1, 1, 1, channels])\\n# return alpha\\n# def MyTile_output_shape(input_shape):\\n# shape = list(input_shape)\\n# assert len(shape) == 5 # only valid for 2D tensors\\n# shape[-1] *= channels\\n# return tuple(shape)\\n ndims = len(up_in.get_shape()) - 2\\n assert ndims in [1, 2, 3], \\\"ndims should be one of 1, 2, or 3. found: %d\\\" % ndims\\n input_channels = up_in.get_shape().as_list()[-1]\\n# batch_size1 = tf.shape(down_in)[0]\\n# nf = tf.min(batch_size0,batch_size1)\\n Conv = getattr(KL, 'Conv%dD' % ndims)\\n up = Conv(nf, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', strides=2)(up_in)\\n down = Conv(nf, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', strides=1)(down_in)\\n\\n# x = NCCLayer(channels=nf)([up,down])\\n x = Add()([up,down])\\n x = Activation('relu')(x)\\n x = Conv(1, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', use_bias = True, bias_initializer='zeros',strides=1,activation='sigmoid')(x)\\n# x = Activation('sigmoid')(x)\\n upsample_layer = getattr(KL, 'UpSampling%dD' % ndims)\\n alpha = upsample_layer()(x)\\n# alpha = Lambda(MyTile)(alpha, input_channels)\\n alpha = MyTile(channels=input_channels)(alpha)\\n up_out = Multiply()([alpha,up_in])\\n return up_out\",\n \"def _esim(self, head_output_steps, body_output_steps, head_sent_sizes, body_sent_sizes):\\n with tf.variable_scope(\\\"ESIM_single_sentence_pair\\\", reuse=True):\\n # head_sent_sizes = tf.Print(head_sent_sizes, [head_sent_sizes, tf.shape(head_sent_sizes)],\\n # \\\"head_sent_sizes:\\\\n\\\", summarize=32)\\n # body_sent_sizes = tf.Print(body_sent_sizes, [body_sent_sizes, tf.shape(body_sent_sizes)],\\n # \\\"body_sent_sizes:\\\\n\\\", summarize=32)\\n # batch_size * sents * head_words * body_words\\n attention_matrix = tf.matmul(head_output_steps, body_output_steps, transpose_b=True)\\n self.attention_matrix = attention_matrix\\n # batch_size * sents * head_words * body_words\\n soft_attention_matrix_for_head = tf.nn.softmax(attention_matrix, 3)\\n self.softmax_matrix_head = soft_attention_matrix_for_head\\n # batch_size * sents * body_words * head_words\\n soft_attention_matrix_for_body = tf.nn.softmax(tf.transpose(attention_matrix, [0, 1, 3, 2]), 3)\\n self.softmax_matrix_body = soft_attention_matrix_for_body\\n # batch_size * sents * head_words * embed\\n head_attended = tf.matmul(soft_attention_matrix_for_head, body_output_steps, name=\\\"attend_head\\\")\\n # head_attended = tf.Print(head_attended, [head_attended, tf.shape(head_attended)], \\\"head_attended:\\\\n\\\",\\n # summarize=15000)\\n\\n # batch_size * sents * body_words * embed\\n body_attended = tf.matmul(soft_attention_matrix_for_body, head_output_steps, name=\\\"attend_body\\\")\\n # body_attended = tf.Print(body_attended, [body_attended, tf.shape(body_attended)], \\\"body_attended:\\\\n\\\",\\n # summarize=15000)\\n\\n batch_size, num_sents, max_head_words, embed_size = tf.unstack(tf.shape(head_attended))\\n flat_head_sent_sizes = tf.reshape(head_sent_sizes, [batch_size * num_sents])\\n flat_head_seq_masks = tf.sequence_mask(flat_head_sent_sizes, max_head_words)\\n head_seq_masks = tf.reshape(flat_head_seq_masks, [batch_size, num_sents, max_head_words])\\n # batch_size * sents * head_words\\n head_masks_per_word = tf.where(head_seq_masks, tf.ones_like(head_seq_masks, dtype=tf.float32),\\n tf.zeros_like(head_seq_masks, dtype=tf.float32))\\n encode_size = 2 * self.num_neurons[0]\\n # now batch_size * sents * head_words * embed\\n head_masks = tf.tile(tf.expand_dims(head_masks_per_word, 3), [1, 1, 1, encode_size])\\n masked_head_attended = head_attended * head_masks\\n\\n _, _, max_body_words, _ = tf.unstack(tf.shape(body_attended))\\n flat_body_sent_sizes = tf.reshape(body_sent_sizes, [batch_size * num_sents])\\n flat_body_seq_masks = tf.sequence_mask(flat_body_sent_sizes, max_body_words)\\n body_seq_masks = tf.reshape(flat_body_seq_masks, [batch_size, num_sents, max_body_words])\\n # batch_size * sents * body_words\\n body_masks_per_word = tf.where(body_seq_masks, tf.ones_like(body_seq_masks, dtype=tf.float32),\\n tf.zeros_like(body_seq_masks, dtype=tf.float32))\\n # now batch_size * sents * body_words * embed\\n body_masks = tf.tile(tf.expand_dims(body_masks_per_word, 3), [1, 1, 1, encode_size])\\n masked_body_attended = body_attended * body_masks\\n\\n # batch_size * sents * head_words * (4 * embed)\\n head_concat = tf.concat(\\n [head_output_steps, masked_head_attended, tf.abs(tf.subtract(masked_head_attended, head_output_steps)),\\n tf.multiply(head_output_steps, masked_head_attended)], 3, name=\\\"concat_head_attended\\\")\\n # batch_size * sents * body_words * (4 * embed)\\n body_concat = tf.concat(\\n [body_output_steps, masked_body_attended, tf.abs(tf.subtract(masked_body_attended, body_output_steps)),\\n tf.multiply(body_output_steps, masked_body_attended)], 3, name=\\\"concat_body_attended\\\")\\n return head_concat, body_concat, head_masks_per_word, body_masks_per_word\",\n \"def task_metalearn(inp, reuse=True):\\n inputa1,inputa2,inputa3, inputb1,inputb2,inputb3, labela, labelb = inp\\n print(\\\"Input a: \\\" , inputa1)\\n task_outputbs, task_lossesb, auto_losses = [], [], []\\n auto_loss = None\\n\\n #One for each. \\n\\n # This takes in the input and passes out the latent variables.\\n temp_in_a_1 = self.encoder(inputa1,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_a_1 = self.decoder(temp_in_a_1,self.auto_weights, reuse=reuse)\\n auto_out_a_1 = temp_out_a_1\\n\\n temp_in_a_2 = self.encoder(inputa2,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_a_2 = self.decoder(temp_in_a_2,self.auto_weights, reuse=reuse)\\n auto_out_a_2 = temp_out_a_2\\n\\n temp_in_a_3 = self.encoder(inputa3,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_a_3 = self.decoder(temp_in_a_3,self.auto_weights, reuse=reuse)\\n auto_out_a_3 = temp_out_a_3\\n\\n # This takes in the input and passes out the latent variables.\\n temp_in_b_1 = self.encoder(inputb1,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_b_1 = self.decoder(temp_in_b_1,self.auto_weights, reuse=reuse)\\n auto_out_b_1 = temp_out_b_1\\n\\n temp_in_b_2 = self.encoder(inputb2,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_b_2 = self.decoder(temp_in_b_2,self.auto_weights, reuse=reuse)\\n auto_out_b_2 = temp_out_b_2\\n\\n temp_in_b_3 = self.encoder(inputb3,self.auto_weights, reuse=reuse)\\n #Then transform it back, and take the loss\\n temp_out_b_3 = self.decoder(temp_in_b_3,self.auto_weights, reuse=reuse)\\n auto_out_b_3 = temp_out_b_3\\n\\n\\n #print(\\\"temp out a: \\\" , temp_out_a)\\n la_1 = self.loss_func(temp_out_a_1,inputa1)\\n la_2 = self.loss_func(temp_out_a_2,inputa2)\\n la_3 = self.loss_func(temp_out_a_3,inputa3)\\n\\n lb_1 = self.loss_func(temp_out_b_1,inputb1)\\n lb_2 = self.loss_func(temp_out_b_2,inputb2)\\n lb_3 = self.loss_func(temp_out_b_3,inputb3)\\n\\n auto_loss = la_1+la_2+la_3\\n\\n print(\\\"Input a1: \\\" , temp_in_a_1)\\n inputa=tf.concat([temp_in_a_1, temp_in_a_2,temp_in_a_3],1)\\n inputb=tf.concat([temp_in_b_1, temp_in_b_2,temp_in_b_3],1)\\n print(\\\"Inputa: \\\" , inputa)\\n task_outputa = self.forward(inputa, weights, reuse=reuse) # only reuse on the first iter\\n temp_outputa = self.decoder(task_outputa,self.auto_weights,reuse=reuse)\\n \\n print(\\\"Task outputa: \\\" , temp_outputa)\\n print(\\\"Label a: \\\" , labela)\\n task_lossa = self.loss_func(temp_outputa, labela)\\n\\n grads = tf.gradients(task_lossa, list(weights.values()))\\n if FLAGS.stop_grad:\\n grads = [tf.stop_gradient(grad) for grad in grads]\\n gradients = dict(zip(weights.keys(), grads))\\n fast_weights = dict(zip(weights.keys(), [weights[key] - self.update_lr*gradients[key] for key in weights.keys()]))\\n output = self.forward(inputb, fast_weights, reuse=True)\\n \\n temp_outputb = self.decoder(output,self.auto_weights, reuse=True)\\n output = temp_outputb\\n\\n task_outputbs.append(output)\\n print(\\\"Output: \\\" , output)\\n print(\\\"Labels: \\\" , labelb)\\n task_lossesb.append(self.loss_func(output, labelb))\\n print(\\\"Num updates is: \\\" , num_updates-1)\\n\\n for j in range(num_updates - 1):\\n loss = self.loss_func(self.decoder(self.forward(inputa, fast_weights, reuse=True),self.auto_weights), labela)\\n grads = tf.gradients(loss, list(fast_weights.values()))\\n if FLAGS.stop_grad:\\n grads = [tf.stop_gradient(grad) for grad in grads]\\n gradients = dict(zip(fast_weights.keys(), grads))\\n fast_weights = dict(zip(fast_weights.keys(), [fast_weights[key] - self.update_lr*gradients[key] for key in fast_weights.keys()]))\\n \\n output = self.forward(inputb, fast_weights, reuse=True)\\n output = self.decoder(output,self.auto_weights)\\n \\n task_outputbs.append(output)\\n task_lossesb.append(self.loss_func(output, labelb))\\n # Task Outputa Direct output, non-decoded (only first step). This is the location (so it is in hyperparameter space)\\n # Task_outputbs Output from complete system, for each step (first is after 1st step, 2nd is after next, etc.)\\n # task_lossa Loss of difference in first step (decoded)\\n # task_lossesb Losses from complete ssystem from each step \\n # auto_loss Sum of all auto-encoding loss of a's\\n # auto_out_a_1 Auto-encoded a1 output\\n # auto_out_a_2 Auto-encoded a2 output\\n\\n task_output = [temp_outputa, task_outputbs, task_lossa, task_lossesb, auto_loss,auto_out_a_1,auto_out_a_2]\\n return task_output\",\n \"def forward(self, x, y, x_mask, candidate_aggre):\\n x_ans_mask = candidate_aggre.sum(dim=2).ge(0).float()\\n Wy = self.linear(y) if self.linear is not None else y # batch * h1\\n p = torch.bmm(x,Wy.unsqueeze(2)).squeeze(2) # batch * len\\n p.data.masked_fill_(x_mask.data, -float('inf')) \\n pm = F.softmax(p, dim=1) * x_ans_mask # batch * len\\n unnormalized_probs=torch.bmm(pm.unsqueeze(1), candidate_aggre).squeeze(1) # batch * c\\n\\n normalized_probs=unnormalized_probs/unnormalized_probs.sum(dim=1, keepdim=True)+very_small_number\\n if self.training:\\n return torch.log(normalized_probs)\\n else:\\n return normalized_probs\",\n \"def forward(self, *input):\\n \\n item_text, item_img, uidx = input\\n # compute latent codes\\n user_codes = self.user_embdding(uidx)\\n item_img_codes = self.ilatent_img_codes(item_img)\\n item_text_codes = self.ilatent_text_codes(item_text)\\n #ubianry = self.hash(user_codes)\\n #ibinary = [self.hash(h) for h in item_codes]\\n uscore_img = self.uscores(user_codes, item_img_codes)\\n iscore_img = self.iscores(item_img_codes)\\n uscore_text = self.uscores(user_codes, item_text_codes)\\n iscore_text = self.iscores(item_text_codes)\\n score = uscore_img + uscore_text + iscore_img + iscore_text\\n\\n binary = torch.zeros_like(score)#self.scores(ubianry, ibinary).view(-1, 1)\\n return score, binary\",\n \"def init_boldpostprocess_wf(\\n lower_bpf,\\n upper_bpf,\\n contigvol,\\n bpf_order,\\n motion_filter_order,\\n motion_filter_type,\\n band_stop_min,\\n band_stop_max,\\n smoothing,\\n bold_file,\\n head_radius,\\n params,\\n custom_conf,\\n omp_nthreads,\\n dummytime,\\n output_dir,\\n fd_thresh,\\n num_bold,\\n mni_to_t1w,\\n despike,\\n brain_template='MNI152NLin2009cAsym',\\n layout=None,\\n name='bold_postprocess_wf',\\n ):\\n\\n\\n TR = layout.get_tr(bold_file)\\n file_base = os.path.basename(str(bold_file))\\n workflow = Workflow(name=name)\\n\\n workflow.__desc__ = \\\"\\\"\\\"\\nFor each of the {num_bold} BOLD series found per subject (across all\\ntasks and sessions), the following post-processing was performed:\\n\\\"\\\"\\\".format(num_bold=num2words(num_bold))\\n\\n if dummytime > 0:\\n nvolx = str(np.floor(dummytime / TR))\\n workflow.__desc__ = workflow.__desc__ + \\\"\\\"\\\" \\\\\\nbefore nuisance regression and filtering of the data, the first {nvol} were discarded,\\n.Furthermore,volumes with framewise-displacement greater than \\n{fd_thresh} mm [@power_fd_dvars;@satterthwaite_2013] were flagged as outliers\\n and excluded from nuisance regression.\\n\\\"\\\"\\\".format(nvol=num2words(nvolx),fd_thresh=fd_thresh)\\n\\n else:\\n workflow.__desc__ = workflow.__desc__ + \\\"\\\"\\\" \\\\\\nbefore nuisance regression and filtering of the data, volumes with framewise-displacement greater than \\n{fd_thresh} mm [@power_fd_dvars;@satterthwaite_2013] were flagged as outliers\\n and excluded from nuisance regression.\\n\\\"\\\"\\\".format(fd_thresh=fd_thresh)\\n\\n workflow.__desc__ = workflow.__desc__ + \\\"\\\"\\\" \\\\\\n{regressors} [@benchmarkp;@satterthwaite_2013]. These nuisance regressors were \\nregressed from the BOLD data using linear regression - as implemented in Scikit-Learn {sclver} [@scikit-learn].\\nResidual timeseries from this regression were then band-pass filtered to retain signals within the {highpass}-{lowpass} Hz frequency band. \\n \\\"\\\"\\\".format(regressors=stringforparams(params=params),sclver=sklearn.__version__,\\n lowpass=upper_bpf,highpass=lower_bpf)\\n\\n\\n # get reference and mask\\n mask_file,ref_file = _get_ref_mask(fname=bold_file)\\n\\n inputnode = pe.Node(niu.IdentityInterface(\\n fields=['bold_file','ref_file','bold_mask','cutstom_conf','mni_to_t1w','t1w','t1seg']),\\n name='inputnode')\\n\\n inputnode.inputs.bold_file = str(bold_file)\\n inputnode.inputs.ref_file = str(ref_file)\\n inputnode.inputs.bold_mask = str(mask_file)\\n inputnode.inputs.custom_conf = str(custom_conf)\\n\\n\\n outputnode = pe.Node(niu.IdentityInterface(\\n fields=['processed_bold', 'smoothed_bold','alff_out','smoothed_alff',\\n 'reho_out','sc217_ts', 'sc217_fc','sc417_ts','sc417_fc','ts50_ts','ts50_fc',\\n 'gs360_ts', 'gs360_fc','gd333_ts', 'gd333_fc','qc_file','fd']),\\n name='outputnode')\\n\\n mem_gbx = _create_mem_gb(bold_file)\\n\\n\\n fcon_ts_wf = init_fcon_ts_wf(mem_gb=mem_gbx['timeseries'],mni_to_t1w=mni_to_t1w,\\n t1w_to_native=_t12native(bold_file),bold_file=bold_file,\\n brain_template=brain_template,name=\\\"fcons_ts_wf\\\")\\n\\n alff_compute_wf = init_compute_alff_wf(mem_gb=mem_gbx['timeseries'], TR=TR,\\n lowpass=upper_bpf,highpass=lower_bpf,smoothing=smoothing, cifti=False,\\n name=\\\"compute_alff_wf\\\" )\\n\\n reho_compute_wf = init_3d_reho_wf(mem_gb=mem_gbx['timeseries'],smoothing=smoothing,\\n name=\\\"afni_reho_wf\\\")\\n\\n write_derivative_wf = init_writederivatives_wf(smoothing=smoothing,bold_file=bold_file,\\n params=params,cifti=None,output_dir=output_dir,dummytime=dummytime,\\n lowpass=upper_bpf,highpass=lower_bpf,TR=TR,omp_nthreads=omp_nthreads,\\n name=\\\"write_derivative_wf\\\")\\n\\n confoundmat_wf = pe.Node(ConfoundMatrix(head_radius=head_radius, params=params,\\n filtertype=motion_filter_type,cutoff=band_stop_max,\\n low_freq=band_stop_max,high_freq=band_stop_min,TR=TR,\\n filterorder=motion_filter_order),\\n name=\\\"ConfoundMatrix_wf\\\", mem_gb=mem_gbx['resampled'])\\n\\n censorscrub_wf = init_censoring_wf(mem_gb=mem_gbx['timeseries'],TR=TR,custom_conf=custom_conf,head_radius=head_radius,\\n contigvol=contigvol,dummytime=dummytime,fd_thresh=fd_thresh,name='censoring')\\n \\n resdsmoothing_wf = init_resd_smoohthing(mem_gb=mem_gbx['timeseries'],smoothing=smoothing,cifti=False,\\n name=\\\"resd_smoothing_wf\\\")\\n \\n filtering_wf = pe.Node(FilteringData(tr=TR,lowpass=upper_bpf,highpass=lower_bpf,\\n filter_order=bpf_order),\\n name=\\\"filtering_wf\\\", mem_gb=mem_gbx['timeseries'])\\n\\n regression_wf = pe.Node(regress(tr=TR),\\n name=\\\"regression_wf\\\",mem_gb = mem_gbx['timeseries'])\\n\\n interpolate_wf = pe.Node(interpolate(TR=TR),\\n name=\\\"interpolation_wf\\\",mem_gb = mem_gbx['timeseries'])\\n\\n \\n executivesummary_wf =init_execsummary_wf(tr=TR,bold_file=bold_file,layout=layout,\\n output_dir=output_dir,mni_to_t1w=mni_to_t1w,omp_nthreads=2)\\n \\n\\n # get transform file for resampling and fcon\\n \\n \\n \\n transformfile = get_transformfile(bold_file=bold_file,\\n mni_to_t1w=mni_to_t1w,t1w_to_native=_t12native(bold_file))\\n t1w_mask = get_maskfiles(bold_file=bold_file,mni_to_t1w=mni_to_t1w)[1]\\n\\n bold2MNI_trans,bold2T1w_trans = get_transformfilex(bold_file=bold_file,\\n mni_to_t1w=mni_to_t1w,t1w_to_native=_t12native(bold_file)) \\n\\n \\n resample_parc = pe.Node(ApplyTransforms(\\n dimension=3,\\n input_image=str(get_template(\\n 'MNI152NLin2009cAsym', resolution=1, desc='carpet',\\n suffix='dseg', extension=['.nii', '.nii.gz'])),\\n interpolation='MultiLabel',transforms=transformfile),\\n name='resample_parc')\\n \\n resample_bold2T1w = pe.Node(ApplyTransforms(\\n dimension=3,\\n input_image=mask_file,reference_image=t1w_mask,\\n interpolation='NearestNeighbor',transforms=bold2T1w_trans),\\n name='bold2t1_trans')\\n \\n resample_bold2MNI = pe.Node(ApplyTransforms(\\n dimension=3,\\n input_image=mask_file,reference_image=str(get_template(\\n 'MNI152NLin2009cAsym', resolution=2, desc='brain',\\n suffix='mask', extension=['.nii', '.nii.gz'])),\\n interpolation='NearestNeighbor',transforms=bold2MNI_trans),\\n name='bold2mni_trans')\\n\\n qcreport = pe.Node(computeqcplot(TR=TR,bold_file=bold_file,dummytime=dummytime,t1w_mask=t1w_mask,\\n template_mask = str(get_template('MNI152NLin2009cAsym', resolution=2, desc='brain',\\n suffix='mask', extension=['.nii', '.nii.gz'])),\\n head_radius=head_radius), name=\\\"qc_report\\\",mem_gb = mem_gbx['resampled'])\\n \\n\\n workflow.connect([\\n # connect bold confound matrix to extract confound matrix \\n (inputnode, confoundmat_wf, [('bold_file', 'in_file'),]),\\n ])\\n \\n # if there is despiking\\n if despike:\\n despike_wf = pe.Node(Despike(outputtype='NIFTI_GZ',args='-NEW'),name=\\\"despike_wf\\\",mem_gb=mem_gbx['resampled'])\\n\\n workflow.connect([\\n (inputnode,despike_wf,[('bold_file','in_file')]),\\n (despike_wf,censorscrub_wf,[('out_file','inputnode.bold')])\\n ])\\n else:\\n workflow.connect([\\n (inputnode,censorscrub_wf,[('bold_file','inputnode.bold')]),\\n ])\\n \\n # add neccessary input for censoring if there is one\\n workflow.connect([\\n\\t (inputnode,censorscrub_wf,[('bold_file','inputnode.bold_file'),\\n\\t ('bold_mask','inputnode.bold_mask')]),\\n\\t (confoundmat_wf,censorscrub_wf,[('confound_file','inputnode.confound_file')])\\n ])\\n\\n # regression workflow \\n workflow.connect([\\n\\t (inputnode,regression_wf,[('bold_mask','mask')]),\\n\\t (censorscrub_wf,regression_wf,[('outputnode.bold_censored','in_file'),\\n\\t ('outputnode.fmriprepconf_censored','confounds'), \\n\\t\\t ('outputnode.customconf_censored','custom_conf')])\\n ])\\n # interpolation workflow\\n workflow.connect([\\n\\t (inputnode,interpolate_wf,[('bold_file','bold_file'),('bold_mask','mask_file')]),\\n\\t (censorscrub_wf,interpolate_wf,[('outputnode.tmask','tmask')]),\\n\\t (regression_wf,interpolate_wf,[('res_file','in_file')]), \\n\\t])\\n # add filtering workflow \\n workflow.connect([\\n (inputnode,filtering_wf,[('bold_mask','mask')]),\\n\\t (interpolate_wf,filtering_wf,[('bold_interpolated','in_file')]),\\n\\n ])\\n \\n # residual smoothing \\n workflow.connect([\\n\\t (filtering_wf,resdsmoothing_wf,[('filt_file','inputnode.bold_file')]) \\n ])\\n\\n #functional connect workflow\\n workflow.connect([\\n (inputnode,fcon_ts_wf,[('ref_file','inputnode.ref_file'),]),\\n (filtering_wf,fcon_ts_wf,[('filt_file','inputnode.clean_bold'),]),\\n ])\\n # reho and alff\\n workflow.connect([ \\n\\t (inputnode,alff_compute_wf,[('bold_mask','inputnode.bold_mask')]),\\n\\t (inputnode,reho_compute_wf,[('bold_mask','inputnode.bold_mask')]),\\n\\t (filtering_wf, alff_compute_wf,[('filt_file','inputnode.clean_bold')]),\\n\\t (filtering_wf, reho_compute_wf,[('filt_file','inputnode.clean_bold')]),\\n ])\\n\\n # qc report\\n workflow.connect([\\n (inputnode,qcreport,[('bold_mask','mask_file')]),\\n (filtering_wf,qcreport,[('filt_file','cleaned_file')]),\\n (censorscrub_wf,qcreport,[('outputnode.tmask','tmask')]),\\n (inputnode,resample_parc,[('ref_file','reference_image')]),\\n (resample_parc,qcreport,[('output_image','seg_file')]),\\n (resample_bold2T1w,qcreport,[('output_image','bold2T1w_mask')]),\\n (resample_bold2MNI,qcreport,[('output_image','bold2temp_mask')]),\\n (qcreport,outputnode,[('qc_file','qc_file')]),\\n ])\\n\\n \\n\\n # write to the outputnode, may be use in future\\n workflow.connect([\\n\\t(filtering_wf,outputnode,[('filt_file','processed_bold')]),\\n\\t(censorscrub_wf,outputnode,[('outputnode.fd','fd')]),\\n\\t(resdsmoothing_wf,outputnode,[('outputnode.smoothed_bold','smoothed_bold')]),\\n\\t(alff_compute_wf,outputnode,[('outputnode.alff_out','alff_out'),\\n ('outputnode.smoothed_alff','smoothed_alff')]),\\n (reho_compute_wf,outputnode,[('outputnode.reho_out','reho_out')]),\\n\\t (fcon_ts_wf,outputnode,[('outputnode.sc217_ts','sc217_ts' ),('outputnode.sc217_fc','sc217_fc'),\\n ('outputnode.sc417_ts','sc417_ts'),('outputnode.sc417_fc','sc417_fc'),\\n ('outputnode.gs360_ts','gs360_ts'),('outputnode.gs360_fc','gs360_fc'),\\n ('outputnode.gd333_ts','gd333_ts'),('outputnode.gd333_fc','gd333_fc'),\\n ('outputnode.ts50_ts','ts50_ts'),('outputnode.ts50_fc','ts50_fc')]),\\n\\n ])\\n \\n # write derivatives \\n workflow.connect([\\n (filtering_wf,write_derivative_wf,[('filt_file','inputnode.processed_bold')]),\\n\\t (resdsmoothing_wf,write_derivative_wf,[('outputnode.smoothed_bold','inputnode.smoothed_bold')]),\\n (censorscrub_wf,write_derivative_wf,[('outputnode.fd','inputnode.fd')]),\\n (alff_compute_wf,write_derivative_wf,[('outputnode.alff_out','inputnode.alff_out'),\\n ('outputnode.smoothed_alff','inputnode.smoothed_alff')]),\\n (reho_compute_wf,write_derivative_wf,[('outputnode.reho_out','inputnode.reho_out')]),\\n (fcon_ts_wf,write_derivative_wf,[('outputnode.sc217_ts','inputnode.sc217_ts' ),\\n ('outputnode.sc217_fc','inputnode.sc217_fc'),\\n ('outputnode.sc417_ts','inputnode.sc417_ts'),\\n ('outputnode.sc417_fc','inputnode.sc417_fc'),\\n ('outputnode.gs360_ts','inputnode.gs360_ts'),\\n ('outputnode.gs360_fc','inputnode.gs360_fc'),\\n ('outputnode.gd333_ts','inputnode.gd333_ts'),\\n ('outputnode.gd333_fc','inputnode.gd333_fc'),\\n ('outputnode.ts50_ts','inputnode.ts50_ts'),\\n ('outputnode.ts50_fc','inputnode.ts50_fc')]),\\n (qcreport,write_derivative_wf,[('qc_file','inputnode.qc_file')]),\\n\\n\\n\\n ])\\n functional_qc = pe.Node(FunctionalSummary(bold_file=bold_file,tr=TR),\\n name='qcsummary', run_without_submitting=True)\\n\\n ds_report_qualitycontrol = pe.Node(\\n DerivativesDataSink(base_directory=output_dir, desc='qualitycontrol',source_file=bold_file, datatype=\\\"figures\\\"),\\n name='ds_report_qualitycontrol', run_without_submitting=True)\\n\\n ds_report_preprocessing = pe.Node(\\n DerivativesDataSink(base_directory=output_dir, desc='preprocessing',source_file=bold_file, datatype=\\\"figures\\\"),\\n name='ds_report_preprocessing', run_without_submitting=True)\\n ds_report_postprocessing = pe.Node(\\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='postprocessing', datatype=\\\"figures\\\"),\\n name='ds_report_postprocessing', run_without_submitting=True)\\n\\n ds_report_connectivity = pe.Node(\\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='connectvityplot', datatype=\\\"figures\\\"),\\n name='ds_report_connectivity', run_without_submitting=True)\\n\\n ds_report_rehoplot = pe.Node(\\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='rehoplot', datatype=\\\"figures\\\"),\\n name='ds_report_rehoplot', run_without_submitting=True)\\n\\n ds_report_afniplot = pe.Node(\\n DerivativesDataSink(base_directory=output_dir,source_file=bold_file, desc='afniplot', datatype=\\\"figures\\\"),\\n name='ds_report_afniplot', run_without_submitting=True)\\n\\n workflow.connect([\\n (qcreport,ds_report_preprocessing,[('raw_qcplot','in_file')]),\\n (qcreport,ds_report_postprocessing ,[('clean_qcplot','in_file')]),\\n (qcreport,functional_qc,[('qc_file','qc_file')]),\\n (functional_qc,ds_report_qualitycontrol,[('out_report','in_file')]),\\n (fcon_ts_wf,ds_report_connectivity,[('outputnode.connectplot','in_file')]),\\n (reho_compute_wf,ds_report_rehoplot,[('outputnode.rehohtml','in_file')]),\\n (alff_compute_wf,ds_report_afniplot ,[('outputnode.alffhtml','in_file')]),\\n ])\\n\\n\\n ## exexetive summary workflow\\n workflow.connect([\\n (inputnode,executivesummary_wf,[('t1w','inputnode.t1w'),('t1seg','inputnode.t1seg'),\\n ('bold_file','inputnode.bold_file'),('bold_mask','inputnode.mask')]),\\n\\n (regression_wf,executivesummary_wf,[('res_file','inputnode.regdata'),]),\\n (filtering_wf,executivesummary_wf,[('filt_file','inputnode.resddata')]),\\n (censorscrub_wf,executivesummary_wf,[('outputnode.fd','inputnode.fd')]),\\n ]),\\n\\n return workflow\",\n \"def p(self, evidence, history):\\n raise NotImplementedError(\\\"The p method for the MLE model has not been implemented yet.\\\")\",\n \"def forward(self, inputs, token_types, valid_length=None): # pylint: disable=arguments-differ\\n bert_output = self.bert(inputs, token_types, valid_length)\\n if self.add_query:\\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\\n mask = 1 - token_types\\n avg_q = mx.nd.sum(mx.nd.multiply(mask, o), axis=2) / mx.nd.sum(mask, axis=1)\\n o = mx.nd.add(o, mx.nd.multiply(avg_q.expand_dims(axis=2), token_types))\\n attended_output = mx.ndarray.transpose(o, axes=(1,2,0))\\n if self.apply_coattention:\\n #############################\\n '''\\n # option 3: use exactly the QANet way\\n # not a good idea, this will cause index shift and thus cause bug\\n # but if we are going to use QANet's method for modeling and extracting the output, it'll be fine to do so\\n # so I just kept these lines here\\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\\n context_mask = token_types\\n query_mask = 1 - context_mask\\n context_max_len = bert_output.shape[1] # int(context_mask.sum(axis=1).max().asscalar())\\n query_max_len = bert_output.shape[1] # int(query_mask.sum(axis=1).max().asscalar())\\n context_raw = mx.nd.multiply(context_mask, o)\\n context_raw = mx.ndarray.expand_dims(context_raw, 0)\\n # print(context_raw[0,0,0,:])\\n # to get the offset to shift using gridgenerator and bilinear-sampler\\n raw_offset = query_mask.sum(axis=1).reshape(len(query_mask),1).tile(bert_output.shape[1])\\n warp_matrix = mx.ndarray.expand_dims(mx.ndarray.stack(raw_offset, \\n mx.nd.zeros(raw_offset.shape).as_in_context(raw_offset.context)), 0)\\n grid = GridGenerator(data=warp_matrix, transform_type='warp')\\n warpped_out = BilinearSampler(context_raw, grid)\\n # the context mask also needs to be shifted\\n context_mask = mx.ndarray.expand_dims(context_mask, 0)\\n context_mask = mx.ndarray.expand_dims(context_mask, 0)\\n context_mask = BilinearSampler(context_mask, grid)\\n context_mask = mx.ndarray.squeeze(context_mask, axis=(0, 1))\\n # get the two encodings separated\\n context_emb_encoded = mx.ndarray.transpose(mx.ndarray.squeeze(warpped_out, axis=0), axes=(1,2,0))\\n '''\\n #################################\\n # option 2: get the two encodings separated\\n o = mx.ndarray.transpose(bert_output, axes=(2,0,1))\\n context_mask = token_types\\n query_mask = 1 - context_mask\\n context_max_len = bert_output.shape[1] # int(context_mask.sum(axis=1).max().asscalar())\\n query_max_len = bert_output.shape[1] # int(query_mask.sum(axis=1).max().asscalar())\\n context_emb_encoded = mx.ndarray.transpose(mx.nd.multiply(context_mask, o), axes=(1,2,0))\\n query_emb_encoded = mx.ndarray.transpose(mx.nd.multiply(query_mask, o), axes=(1,2,0))\\n '''\\n # option 1: context and query differently masked but use the same values\\n # problem: almost the same with simply self-attention\\n context_mask = token_types\\n query_mask = 1 - context_mask\\n context_max_len = bert_output.shape[1]\\n query_max_len = bert_output.shape[1]\\n context_emb_encoded = bert_output\\n query_emb_encoded = bert_output\\n '''\\n # context_mask = context_mask[:,:context_max_len]\\n # query_mask = query_mask[:,:query_max_len]\\n attended_output = self.co_attention(context_emb_encoded, query_emb_encoded, \\n context_mask, query_mask, \\n context_max_len, query_max_len)\\n if self.apply_self_attention:\\n attended_output, att_weights = self.multi_head_attention(bert_output, bert_output) \\n if self.add_query or self.apply_self_attention or self.apply_coattention:\\n output = self.span_classifier(attended_output)\\n else:\\n output = self.span_classifier(bert_output)\\n return output\",\n \"def __init__(self, h, d_model, dropout=0.1):\\n super(MultiHeadedAttention, self).__init__()\\n assert d_model % h == 0\\n # We assume d_v always equals d_k\\n self.d_k = d_model // h\\n self.h = h\\n self.linears = clones(nn.Linear(d_model, d_model), 4)\\n self.attn = None\\n self.dropout = nn.Dropout(p=dropout)\",\n \"def generate(self, text, prev_mel):\\n # forward pass through text embedding and get k and v\\n kv = self.t_encoder(text)\\n k = kv[:,:self.hp.d,:]\\n v = kv[:,self.hp.d:,:]\\n # forward pass through audio encoding and get Q\\n q = self.a_encoder(prev_mel)\\n \\n # compute attention and use forcible incremental attention (section 4.2)\\n a = (k.transpose(2,1)).matmul(q)/np.sqrt(self.hp.d)\\n a = F.softmax(a, dim=1)\\n \\\"\\\"\\\"\\n # get argmax\\n argmax = a[0].argmax(dim=0) # argmax on the N dimension\\n # forcibly incremental attention\\n preva = -1\\n for i in range(a.shape[-1]):\\n if argmax[i] < preva -1 or preva + 3 < argmax[i]:\\n # force the ith column to be zero\\n a[:,:,i] = 0\\n # find correct position\\n position = min(a.shape[1]-1, preva + 1)\\n a[:,position,i] = 1.0\\n # update preva\\n preva = a[0,:,i].argmax()\\\"\\\"\\\"\\n\\n # finish computing y and a\\n r = r = v.matmul(a)\\n\\n rprime = torch.cat((r, q), dim=1)\\n ylogit = self.decoder(rprime)\\n y = F.sigmoid(ylogit)\\n return y, ylogit, a\",\n \"def __init__(self, mbart_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\\n self.linear1_weight = mbart_layer.fc1.weight\\n self.linear1_bias = mbart_layer.fc1.bias\\n self.linear2_weight = mbart_layer.fc2.weight\\n self.linear2_bias = mbart_layer.fc2.bias\\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\\n self.norm2_eps = mbart_layer.final_layer_norm.eps\\n self.norm2_weight = mbart_layer.final_layer_norm.weight\\n self.norm2_bias = mbart_layer.final_layer_norm.bias\\n self.num_heads = mbart_layer.self_attn.num_heads\\n self.embed_dim = mbart_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def on_train_batch_end(\\n self,\\n trainer: 'pl.Trainer',\\n pl_module: 'pl.LightningModule',\\n outputs: Optional[STEP_OUTPUT],\\n batch: Any,\\n batch_idx: int,\\n dataloader_idx: int,\\n ) -> None:\\n if trainer.global_step % self.every_n_steps == 0: \\n text, x = batch\\n sample_text = text[:1]\\n token_list = sample_text.masked_select(sample_text != 0).tolist()\\n decoded_text = self.tokenizer.decode(token_list) \\n text = text.to(pl_module.device)\\n x = x.to(pl_module.device) \\n with torch.no_grad():\\n pl_module.eval()\\n #generate sample with image provided\\n x_rec = pl_module.generate_images(text[:1], img = x[:1], filter_thres=0.9) # topk sampling at 0.9\\n\\n #generate sample without image\\n x_gen = pl_module.generate_images(text[:1], filter_thres=0.9) # topk sampling at 0.9\\n\\n pl_module.train() \\n\\n\\n x_grid = torchvision.utils.make_grid(\\n tensor=x,\\n nrow=self.nrow,\\n padding=self.padding,\\n normalize=self.normalize,\\n value_range=self.norm_range,\\n scale_each=self.scale_each,\\n pad_value=self.pad_value,\\n ) \\n xrec_grid = torchvision.utils.make_grid(\\n tensor=x_rec,\\n nrow=self.nrow,\\n padding=self.padding,\\n normalize=self.normalize,\\n value_range=self.norm_range,\\n scale_each=self.scale_each,\\n pad_value=self.pad_value,\\n ) \\n xgen_grid = torchvision.utils.make_grid(\\n tensor=x_gen,\\n nrow=self.nrow,\\n padding=self.padding,\\n normalize=self.normalize,\\n value_range=self.norm_range,\\n scale_each=self.scale_each,\\n pad_value=self.pad_value,\\n ) \\n text_title = \\\"train/text\\\"\\n trainer.logger.experiment.add_text(text_title, decoded_text, global_step=trainer.global_step)\\n x_title = \\\"train/input\\\"\\n trainer.logger.experiment.add_image(x_title, x_grid, global_step=trainer.global_step)\\n xrec_title = \\\"train/half_reconstruction\\\"\\n trainer.logger.experiment.add_image(xrec_title, xrec_grid, global_step=trainer.global_step)\\n xgen_title = \\\"train/generation\\\"\\n trainer.logger.experiment.add_image(xgen_title, xgen_grid, global_step=trainer.global_step)\",\n \"def attention(decoder_state, coverage=None, num_words_section=None, step=None):\\n with variable_scope.variable_scope(\\\"Attention\\\"):\\n # Pass the decoder state through a linear layer (this is W_s s_t + b_attn in the paper)\\n # (W_s s_t) + b_att is decoder_features; s_t = decoder_state\\n decoder_features = linear(decoder_state, attention_vec_size, True) # shape (batch_size, attention_vec_size)\\n decoder_features = tf.expand_dims(tf.expand_dims(decoder_features, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\\n\\n def masked_attention(e, enc_padding_mask):\\n if enc_section_padding_mask is not None:\\n enc_padding_mask = tf.reshape(enc_section_padding_mask, [batch_size, -1])\\n enc_padding_mask = tf.cast(enc_padding_mask, tf.float32)\\n \\\"\\\"\\\"Take softmax of e then apply enc_padding_mask and re-normalize\\\"\\\"\\\"\\n attn_dist = nn_ops.softmax(e) # take softmax. shape (batch_size, attn_length)\\n attn_dist *= enc_padding_mask # apply mask\\n masked_sums = tf.reduce_sum(attn_dist, axis=1) # shape (batch_size)\\n return attn_dist / tf.reshape(masked_sums, [-1, 1]) # re-normalize\\n\\n if use_coverage and coverage is not None: # non-first step of coverage\\n if not hier:\\n # Multiply coverage vector by w_c to get coverage_features.\\n coverage_features = nn_ops.conv2d(coverage, w_c, [1, 1, 1, 1], \\\"SAME\\\") # c has shape (batch_size, seq_len, 1, attention_vec_size)\\n \\n # Calculate v^T tanh(W_h h_i + W_s s_t + w_c c_i^t + b_attn)\\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features + coverage_features), [2, 3]) # shape (batch_size,seq_len)\\n \\n # Take softmax of e to get the attention distribution\\n # attn_dist = nn_ops.softmax(e) # shape (batch_size, seq_len)\\n attn_dist = masked_attention(e, enc_padding_mask)\\n \\n # Update coverage vector\\n coverage += array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) # shape=(batch_size, seq_len,1,1)\\n else:\\n with tf.variable_scope(\\\"attention_sections\\\"):\\n if FLAGS.fixed_attn:\\n tf.logging.debug('running with fixed attn', '\\\\r')\\n decoder_features_sec = linear(decoder_state, attention_vec_size, True, scope='Linear--Section-Features') # shape (batch_size, attention_vec_size)\\n decoder_features_sec = tf.expand_dims(tf.expand_dims(decoder_features_sec, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\\n e_sec = math_ops.reduce_sum(v_sec * math_ops.tanh(encoder_section_features + decoder_features_sec), [2, 3]) # [batch_size x seq_len_sections]\\n attn_dist_sec = nn_ops.softmax(e_sec)\\n else:\\n e_sec = math_ops.reduce_sum(v_sec * math_ops.tanh(encoder_section_features + decoder_features), [2, 3]) # [batch_size x seq_len_sections]\\n attn_dist_sec = nn_ops.softmax(e_sec)\\n with tf.variable_scope(\\\"attention_words\\\"):\\n coverage_features = nn_ops.conv2d(coverage, w_c, [1, 1, 1, 1], \\\"SAME\\\") # c has shape (batch_size, seq_len, 1, attention_vec_size)\\n \\n # Calculate v^T tanh(W_h h_i + W_s s_t + w_c c_i^t + b_attn)\\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features + coverage_features), [2, 3]) # shape (batch_size,seq_len)\\n\\n # Multiply by section weights\\n \\n e = tf.reshape(e, [batch_size, -1, num_words_section[0][0]])\\n e = tf.multiply(e, attn_dist_sec[:,:,tf.newaxis])\\n e = tf.reshape(e, [batch_size,-1])\\n\\n\\n# --- Some hack for reweighting attention (similar to temp for softmax)\\n if temperature > 0.0:\\n e = e * temperature\\n \\n attn_dist = masked_attention(e, enc_padding_mask)\\n coverage += array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) # shape=(batch_size, seq_len,1,1)\\n \\n else:\\n # Calculate v^T tanh(W_h h_i + W_s s_t + b_attn)\\n if hier:\\n with tf.variable_scope(\\\"attention_sections\\\"):\\n if FLAGS.fixed_attn:\\n decoder_features_sec = linear(decoder_state, attention_vec_size, True, scope='Linear--Section-Features') # shape (batch_size, attention_vec_size)\\n decoder_features_sec = tf.expand_dims(tf.expand_dims(decoder_features_sec, 1), 1) # reshape to (batch_size, 1, 1, attention_vec_size)\\n e_sec = math_ops.reduce_sum(\\n v_sec * math_ops.tanh(encoder_section_features + decoder_features_sec), [2, 3]) # [batch_size x seq_len_sections]\\n attn_dist_sec = nn_ops.softmax(e_sec)\\n else:\\n e_sec = math_ops.reduce_sum(\\n v_sec * math_ops.tanh(encoder_section_features + decoder_features), [2, 3]) # [batch_size x seq_len_sections]\\n attn_dist_sec = nn_ops.softmax(e_sec)\\n\\n with tf.variable_scope(\\\"attention_words\\\"):\\n\\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features), [2, 3]) #[batch_size x seq_len]\\n\\n e = tf.reshape(e, [batch_size, -1, num_words_section[0][0]])\\n e = tf.multiply(e, attn_dist_sec[:,:,tf.newaxis])\\n e = tf.reshape(e, [batch_size,-1])\\n\\n if temperature > 0.0:\\n e = e * temperature\\n \\n attn_dist = masked_attention(e, enc_padding_mask)\\n \\n else:\\n e = math_ops.reduce_sum(v * math_ops.tanh(encoder_features + decoder_features), [2, 3]) # calculate e\\n # Take softmax of e to get the attention distribution\\n if enc_padding_mask is not None:\\n attn_dist = masked_attention(e, enc_padding_mask)\\n else:\\n attn_dist = nn_ops.softmax(e) # shape (batch_size, seq_len)\\n\\n if use_coverage: # first step of training\\n coverage = tf.expand_dims(tf.expand_dims(attn_dist,2),2) # initialize coverage\\n\\n # TODO: coverage for hier\\n\\n # Calculate the context vector from attn_dist and encoder_states\\n # ecnoder_sates = [batch , seq_len , 1 , encoder_output_size], attn_dist = [batch, seq_len, 1, 1]\\n context_vector = math_ops.reduce_sum(array_ops.reshape(attn_dist, [batch_size, -1, 1, 1]) * encoder_states, [1, 2]) # shape (batch_size, enc_output_size).\\n context_vector = array_ops.reshape(context_vector, [-1, enc_output_size])\\n\\n if hier:\\n return context_vector, attn_dist, coverage, attn_dist_sec\\n else:\\n return context_vector, attn_dist, coverage\",\n \"def mlp_policy_net(x, output_sizes):\\n net = hk.Sequential([hk.nets.MLP(output_sizes), jnp.tanh])\\n return net(x)\",\n \"def __init__(self, dim_hv, dim_hw, msg_dim):\\n super(PairMessageGenerator, self).__init__()\\n self.dim_hv, self.dim_hw, self.msg_dim = dim_hv, dim_hw, msg_dim\\n self.in_dim = dim_hv + dim_hw # row * feature_dim, 2048\\n self.mlp = nn.Sequential(\\n nn.LayerNorm(self.in_dim), # this layer norm is important to create diversity\\n nn.Linear(self.in_dim, self.msg_dim),\\n nn.LeakyReLU(0.2)\\n )\",\n \"def self_attention_model(self):\\n inputs = self.prepare_inputs()\\n \\n features = self.prepare_features(inputs)\\n support_history_context_concat_emb = features['support_history_context_concat_emb']\\n support_link_info_concat_emb = features['support_link_info_concat_emb']\\n support_start_node_info_concat_emb = features['support_start_node_info_concat_emb']\\n support_end_node_info_concat_emb = features['support_end_node_info_concat_emb']\\n support_future_context_concat_emb = features['support_future_context_concat_emb']\\n support_neighbor_link_info_concat_emb = features['support_neighbor_link_info_concat_emb']\\n support_neighbor_start_node_info_concat_emb = features['support_neighbor_start_node_info_concat_emb']\\n support_neighbor_end_node_info_concat_emb = features['support_neighbor_end_node_info_concat_emb']\\n support_neighbor_link_state_concat_emb = features['support_neighbor_link_state_concat_emb']\\n support_y_embed = features['support_duration_delta_concat_emb']\\n\\n query_history_context_concat_emb = features['query_history_context_concat_emb']\\n query_link_info_concat_emb = features['query_link_info_concat_emb']\\n query_start_node_info_concat_emb = features['query_start_node_info_concat_emb']\\n query_end_node_info_concat_emb = features['query_end_node_info_concat_emb']\\n query_future_context_concat_emb = features['query_future_context_concat_emb']\\n query_neighbor_link_info_concat_emb = features['query_neighbor_link_info_concat_emb']\\n query_neighbor_start_node_info_concat_emb = features['query_neighbor_start_node_info_concat_emb']\\n query_neighbor_end_node_info_concat_emb = features['query_neighbor_end_node_info_concat_emb']\\n query_neighbor_link_state_concat_emb = features['query_neighbor_link_state_concat_emb']\\n \\n support_x = self.constgat({\\n 'history_context': support_history_context_concat_emb,\\n 'link_info': support_link_info_concat_emb,\\n 'start_node_info': support_start_node_info_concat_emb,\\n 'end_node_info': support_end_node_info_concat_emb,\\n 'future_context': support_future_context_concat_emb,\\n 'neighbor_link_info': support_neighbor_link_info_concat_emb,\\n 'neighbor_start_node_info': support_neighbor_start_node_info_concat_emb,\\n 'neighbor_end_node_info': support_neighbor_end_node_info_concat_emb,\\n 'neighbor_link_state': support_neighbor_link_state_concat_emb})\\n\\n query_x = self.constgat({\\n 'history_context': query_history_context_concat_emb,\\n 'link_info': query_link_info_concat_emb,\\n 'start_node_info': query_start_node_info_concat_emb,\\n 'end_node_info': query_end_node_info_concat_emb,\\n 'future_context': query_future_context_concat_emb,\\n 'neighbor_link_info': query_neighbor_link_info_concat_emb,\\n 'neighbor_start_node_info': query_neighbor_start_node_info_concat_emb,\\n 'neighbor_end_node_info': query_neighbor_end_node_info_concat_emb,\\n 'neighbor_link_state': query_neighbor_link_state_concat_emb})\\n query_x = query_x + layers.reduce_sum(inputs['query_mask']) * 0.0 + layers.reduce_sum(inputs['support_mask']) * 0.0 # IMPORTANT: for save_inference_model\\n\\n def forward_attention(indicator, support_x, support_y_embed, support_mask, query_x, query_y, query_mask):\\n \\\"\\\"\\\"\\n support_indicator: length = support_len\\n if attention(support, query), indicator = 0\\n if attention(support, support), indicator = 1\\n \\\"\\\"\\\"\\n support_y_embed = support_y_embed * support_mask\\n support_xy = layers.concat([support_x, support_y_embed, indicator], axis=1)\\n\\n pad_value = layers.assign(input=numpy.array([0.0], dtype=numpy.float32))\\n support_pad, support_len = layers.sequence_pad(support_xy, pad_value=pad_value)\\n query_pad, query_len = layers.sequence_pad(query_x, pad_value=pad_value)\\n\\n attention = self.attention(query_pad, support_pad, support_pad, self.hidden_dim, 'meta')\\n attention = layers.sequence_unpad(attention, length=query_len)\\n pred_input = layers.concat([attention, query_x], axis=1)\\n\\n pred = self.prepare_preds_with_name(pred_input, 'out_pred')\\n label = layers.cast(query_y, dtype='float32')\\n label = layers.scale(label, scale=0.01)\\n\\n loss = layers.huber_loss(pred, label, 1.0) * query_mask\\n loss = layers.mean(loss)\\n return pred, label, loss\\n\\n indicator = support_y_embed * 0.0\\n pred, label, loss1 = forward_attention(\\n indicator, support_x, support_y_embed, inputs['support_mask'], \\n query_x, inputs['query_duration_delta'], 1.0)\\n indicator = support_y_embed * 1.0\\n _, _, loss2 = forward_attention(\\n indicator, support_x, support_y_embed, inputs['support_mask'], \\n support_x, inputs['support_duration_delta'] - 120, (inputs['support_mask'] * (-1.0) + 1))\\n loss = loss1 + loss2\\n return pred, label, loss\",\n \"def forward(self, input):\\n ### YOUR CODE HERE for part 1h\\n # sentence_length x batch_size x max_word_length\\n orig_size = input.size()\\n merged_input = input.reshape(-1, orig_size[-1])\\n # s*b x max_word_length\\n x_char_embed = self.char_embed(merged_input)\\n # s*b x max_word_length x char_embed_size\\n x_reshaped = x_char_embed.transpose(-1, -2)\\n # s*b x char_embed_size x max_word_length\\n x_conv = self.conv(x_reshaped)\\n # out s*b x word_embed_size\\n x_emb = self.highway(x_conv)\\n x_drop = self.dropout(x_emb)\\n split_output = x_drop.reshape(list(orig_size[:-1]) + [self.word_embed_size])\\n return split_output\\n ### END YOUR CODE\",\n \"def predict(self, first_preprocessed_inputs,second_preprocessed_inputs,third_preprocessed_inputs):\\r\\n pass\",\n \"def prelu(input, weight):\\n return FunctionLib.apply('PRelu', input.device, [input, weight])\",\n \"def forward_step(self, batch):\\n input_ids = torch.as_tensor(batch.input_ids).to(self.device).reshape((1, -1)) # batch.get('input_ids').to(self.device)\\n attention_mask = torch.as_tensor(batch.attention_mask).to(self.device).reshape((1, -1)) # batch.get('attention_mask').to(self.device)\\n outputs = self.model(input_ids=input_ids, attention_mask=attention_mask)[0]\\n\\n _, _, num_label = outputs.shape\\n \\\"\\\"\\\"\\n outputs : (batch, seq_length, feat_dim) => (seq_length, feat_dim)\\n labels : (batch, seq_length) => (seq_length,)\\n \\\"\\\"\\\"\\n outputs = outputs.view(-1, num_label)\\n labels = torch.argmax(outputs, dim=1) # torch.argmax(outputs, dim=1)\\n batch_losses = self.criterion(outputs, labels)\\n loss = torch.mean(batch_losses) # mean average\\n self.batch_output = [input_ids, outputs]\\n return loss\",\n \"def __build_ml_pipeline(self, clf: MultiOutputClassifier) -> Pipeline:\\r\\n pipeline = Pipeline([\\r\\n ('features', FeatureUnion(\\r\\n [('text', Pipeline(\\r\\n [('text_field_extractor', \\r\\n basic_utils.TextFieldExtractor('message')), \\r\\n #('vect', CountVectorizer(tokenizer=clean_text)),\\r\\n #('tfidf', TfidfTransformer())\\r\\n ('tfidf', TfidfVectorizer(tokenizer=basic_utils.tokenize, \\r\\n min_df=.0025, max_df=0.25, \\r\\n ngram_range=(1,2)))\\r\\n ])),\\r\\n \\r\\n ('numerics', FeatureUnion(\\r\\n [('text_len', \\r\\n Pipeline([('text_len_extractor', \\r\\n basic_utils.NumericFieldExtractor('text_len')), \\r\\n ('text_len_scaler', StandardScaler())\\r\\n ])),\\r\\n ('punt_perc', \\r\\n Pipeline([('punt_perc_extractor', \\r\\n basic_utils.NumericFieldExtractor('punt_perc')), \\r\\n ('punt_perc_scaler', StandardScaler())\\r\\n ]))\\r\\n ])),\\r\\n ('starting_verb', basic_utils.PosFieldExtractor('starting_verb_flag'))\\r\\n ])),\\r\\n \\r\\n ('clf', clf)\\r\\n ])\\r\\n \\r\\n return pipeline\",\n \"def bias_prior(self):\",\n \"def on_predict_batch_begin(self, batch, logs=None):\",\n \"def forward(self, encoder_state, context, context_lens):\\n attn = self.attention(context, encoder_state.squeeze(0), context_lens)\\n return attn\",\n \"def attention(inp, scope, e_dim, past, config):\\n assert inp.shape.ndims == 3 # input should be of shape [batch, seqlen, embeddings] # [batch, sequence, features]\\n assert e_dim % config.num_heads == 0 # embedding can be split in heads\\n\\n if past is not None:\\n assert past.shape.ndims == 5 # [batch, 2, heads, seqlen, emebeddings]\\n\\n def split_heads(x):\\n out = split_into_n_states(x, config.num_heads)\\n out = tf.transpose(out, [0, 2, 1, 3])\\n return out\\n\\n def merge_heads(x):\\n out = merge_n_states(tf.transpose(x, [0, 2, 1, 3]))\\n return out\\n\\n def mask_attention_weights(w):\\n # w should have shape [batches, heads, dst_seq, src_seq], where information flows from scr to dst\\n _, _, nd, ns = shapes_list(w)\\n b = attention_mask(nd, ns, w.dtype)\\n b = tf.reshape(b, [1, 1, nd, ns])\\n w = w * b - tf.cast(1e10, w.dtype) * (1 - b)\\n return w\\n\\n def multihead_attention(q, k, v):\\n w = tf.matmul(q, k, transpose_b=True)\\n w *= tf.rsqrt(tf.cast(v.shape[-1].value, w.dtype))\\n\\n # mask attention weights\\n w = mask_attention_weights(w)\\n w = softmax_with_reduce_max(w)\\n out = tf.matmul(w, v)\\n return out\\n\\n with tf.variable_scope(scope):\\n c = conv1d(inp, 'convolutional_attention', e_dim * 3)\\n q, k, v = map(split_heads, tf.split(c, 3, axis=2))\\n present = tf.stack([k, v], axis=1)\\n if past is not None:\\n # there is a stack below it\\n pk, pv = tf.unstack(past, axis=1)\\n k = tf.concat([pk, k], axis=2)\\n v = tf.concat([pv, v], axis=2)\\n\\n attn = multihead_attention(q, k, v)\\n attn = merge_heads(attn)\\n\\n out = conv1d(attn, 'convolutional_projection', e_dim)\\n return out, present\",\n \"def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, ans_ids=None, ans_mask=None, **kwargs):\\n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \\n if self.from_scratch:\\n imgs = imgs / 255.0\\n B_text = token_ids.shape[0]\\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\\n segment_ids = segment_ids.reshape((-1, ) + segment_ids.shape[2:])\\n input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\\n losses = dict()\\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\\n B, D, T, H, W = visual_token.shape\\n if B_text != B:\\n visual_token = visual_token.view(B_text, -1, D, T, H, W)\\n visual_token = visual_token.mean(dim=1)\\n \\n # text feature #\\n text_out_with_mask = self.text_backbone(token_ids, input_mask)\\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\\n\\n # contrastive type finetuning retrieval #\\n if self.task == 'retrieval':\\n # text_only_out = self.text_backbone(token_ids, input_mask)\\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_last_hidden_state, input_mask, token_ids)\\n nce_loss = self.loss_func(visual_emb, text_emb)\\n losses['retrieval_nce_loss'] = nce_loss \\n elif self.task == 'video_qa' or self.task == 'FIB':\\n B, D, T, H, W = visual_token.shape\\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\\n if hasattr(self.qa_head, 'num_labels'):\\n num_choices = self.qa_head.num_labels\\n visual_token_all = visual_token\\n else:\\n num_choices = int(token_ids.shape[0] / B)\\n visual_token_all = visual_token.unsqueeze(1).expand(-1, num_choices, -1, -1, -1).flatten(0,1)\\n\\n output = self.multimodal_backbone(visual_token=visual_token_all, text_input_mask=input_mask, text_input_embeds=text_out_last_hidden_state)\\n \\n if self.answer_mask:\\n mask_idx = torch.where(token_ids == 103)\\n itm_output = output['t_last_hidden_state'][mask_idx]\\n elif self.answer_cls:\\n if 'cls_last_hidden_state' in output:\\n itm_output = output['cls_last_hidden_state'].squeeze()\\n else:\\n itm_output = output['t_last_hidden_state'][:, 0]\\n if self.itm_head is not None:\\n itm_output = self.itm_head(itm_output)\\n\\n else:\\n all_cls_emb = output['last_hidden_state'][:, 0]\\n itm_output = self.itm_head(all_cls_emb)\\n \\n if self.qa_head is not None:\\n final_output = self.qa_head(itm_output).view(-1, num_choices)\\n final_label = label\\n else:\\n final_output = itm_output[:, 1]\\n final_label = label\\n\\n\\n qa_loss = self.loss_func(final_output, final_label.view(-1))\\n losses['qa_loss'] = qa_loss\\n\\n\\n\\n return losses\",\n \"def model(mtf_features, other_features, params, mesh, variable_dtype, context=None):\\n\\n x, batch_dim, sequence_dim, embd_dim, vocab_dim, embed_sequence_dim = parse_inputs(mtf_features, other_features)\\n\\n if is_incremental_inference(context):\\n # reshape inputs if in inference mode\\n x = mtf.gather(x, context.position - 1, sequence_dim)\\n x = mtf.reshape(x, [batch_dim])\\n\\n use_axial_pos_emb = params[\\\"axial_pos_emb\\\"] is not None\\n\\n if not use_axial_pos_emb:\\n # Use standard position encoding\\n wpe = mtf.get_variable(mesh, \\\"wpe\\\", mtf.Shape([embed_sequence_dim, embd_dim]),\\n initializer=tf.random_normal_initializer(stddev=0.01),\\n master_dtype=variable_dtype.master_dtype,\\n slice_dtype=variable_dtype.slice_dtype,\\n activation_dtype=variable_dtype.activation_dtype)\\n else:\\n wpe = axial_positional_emb(embd_dim, mesh, params, variable_dtype)\\n\\n # Text encoding\\n wte = mtf.get_variable(mesh, \\\"wte\\\", mtf.Shape([vocab_dim, embd_dim]),\\n initializer=tf.random_normal_initializer(stddev=0.02),\\n master_dtype=variable_dtype.master_dtype,\\n slice_dtype=variable_dtype.slice_dtype,\\n activation_dtype=variable_dtype.activation_dtype)\\n\\n with tf.variable_scope(\\\"token_embd\\\"):\\n # Text embedding\\n h = mtf.gather(wte, x, vocab_dim)\\n if params[\\\"embed_dropout\\\"] > 0 and params[\\\"mode\\\"] == \\\"train\\\":\\n h = mtf.dropout(h, rate=params[\\\"embed_dropout\\\"], name=\\\"wte_dropout\\\")\\n\\n with tf.variable_scope(\\\"pos_embd\\\"):\\n # Positional embedding\\n position_indices = mtf.range(mesh, sequence_dim, tf.int64) if not is_incremental_inference(context) else (\\n context.position - 1)\\n pos_emb = mtf.gather(wpe, position_indices, wpe.shape[0])\\n if params[\\\"embed_dropout\\\"] > 0 and params[\\\"mode\\\"] == \\\"train\\\":\\n pos_emb = mtf.dropout(pos_emb, rate=params[\\\"embed_dropout\\\"], name=\\\"wte_dropout\\\")\\n h += pos_emb\\n\\n aux_losses = 0 # instantiate auxiliary losses (for MOE models)\\n\\n for layer in range(params[\\\"n_layer\\\"]):\\n # attn blocks\\n share_parameters = exists(params[\\\"share_parameters\\\"]) and params[\\\"share_parameters\\\"] == True\\n block_scope = f\\\"h{layer}\\\" if not share_parameters else \\\"\\\"\\n\\n block_fn = block(params=params, scope=block_scope, layer_num=layer,\\n bias=other_features[\\\"attn_bias\\\"],\\n sequence_dim=sequence_dim,\\n memory_length_dim=other_features[\\\"memory_length_dim\\\"],\\n variable_dtype=variable_dtype,\\n context=context)\\n\\n # If true and in train mode, enable gradient checkpointing\\n recompute_grad = params[\\\"recompute_grad\\\"] and (params[\\\"mode\\\"] == \\\"train\\\") == True\\n h, loss = block_fn(h) if not recompute_grad else mtf.recompute_grad(block_fn, [h])\\n aux_losses += loss\\n\\n no_weight_tie_emb = params[\\\"no_weight_tie\\\"] == True\\n if no_weight_tie_emb:\\n with tf.variable_scope(\\\"wte_final_linear\\\"):\\n logits = linear(h, \\\"linear_out\\\", vocab_dim, variable_dtype=variable_dtype, params=params)\\n else:\\n # Layer normalize & affine transform\\n h = layer_norm(h, \\\"ln_f\\\", variable_dtype=variable_dtype)\\n seq_dim = sequence_dim if not is_incremental_inference(context) else mtf.Dimension(\\\"sequence\\\", 1)\\n with tf.variable_scope(\\\"wte_final_einsum\\\"):\\n # Equivalent to tf.matmul\\n logits = mtf.einsum([h, wte], output_shape=[batch_dim, seq_dim, vocab_dim])\\n\\n if params[\\\"mode\\\"] in [\\\"train\\\", \\\"eval\\\"]:\\n labels = mtf_features[\\\"labels\\\"]\\n z_loss = params.get(\\\"z_loss\\\", 1e-4) # an auxiliary loss used to stabilize mtf xentropy\\n\\n # Go to full precision for the logits \\n logits = mtf.cast(logits, tf.float32)\\n\\n with tf.variable_scope(\\\"xentropy_final\\\"):\\n loss_batch = mtf.layers.softmax_cross_entropy_with_logits(logits=logits, targets=labels,\\n vocab_dim=logits.shape[-1], z_loss=z_loss)\\n\\n # For non-autoregressive models (masked language modeling training)\\n # Make sure labels with padding tokens are not counted in the loss\\n if not params[\\\"causal\\\"]:\\n padding_id = params.get(\\\"padding_id\\\", 0)\\n loss_batch = mtf.where(mtf.not_equal(labels, padding_id), loss_batch, mtf.zeros_like(loss_batch))\\n\\n with tf.variable_scope(\\\"reduce_mean_final\\\"):\\n loss = mtf.reduce_mean(loss_batch)\\n\\n loss += aux_losses # Add on auxiliary losses (currently only used for MoE)\\n loss /= params[\\\"num_microbatches\\\"]\\n # Convert to train dtype\\n loss = mtf.cast(loss, variable_dtype.slice_dtype)\\n else:\\n loss = None\\n loss_batch = None\\n\\n # Cast back to checkpoint dtype\\n logits = mtf.cast(logits, variable_dtype.master_dtype)\\n return logits, loss, loss_batch\",\n \"def forward(self, hg, samp_bias1=None, samp_bias2=None):\\n h_1_all = [];h_2_all = [];c_all = [];logits = []\\n result = {}\\n # process features\\n features = hg.srcdata['h']\\n feats = self.normal_feat(features, self.meta_paths)\\n # shuffled features\\n shuf_feats = self.shuf_feats(feats)\\n\\n for idx, meta_path in enumerate(self.meta_paths):\\n new_g = dgl.metapath_reachable_graph(hg, meta_path)\\n for i in range(self.sc):\\n new_g = dgl.add_self_loop(new_g)\\n\\n feats[idx] = F.dropout(feats[idx], self.dropout, training=self.training)\\n shuf_feats[idx] = F.dropout(shuf_feats[idx], self.dropout, training=self.training)\\n\\n h_1 = self.gcn[idx](new_g, feats[idx])\\n c = self.readout(h_1)\\n c = self.readout_act_func(c)\\n h_2 = self.gcn[idx](new_g, shuf_feats[idx])\\n\\n\\n logit = self.disc(c, h_1, h_2, samp_bias1, samp_bias2)\\n h_1_all.append(h_1.unsqueeze(0))\\n h_2_all.append(h_2.unsqueeze(0))\\n c_all.append(c)\\n logits.append(logit)\\n result['logits'] = logits\\n\\n # Attention or not\\n if self.isAttn:\\n r\\\"\\\"\\\"\\n .. math::\\n \\\\begin{equation}\\n \\\\mathbf{h}_{i}=\\\\mathcal{Q}\\\\left(\\\\left\\\\{\\\\mathbf{h}^{(r)} \\\\mid r \\\\in \\\\mathcal{R}\\\\right\\\\}\\\\right)=\\\\sum_{r \\\\in \\\\mathcal{R}} a_{i}^{(r)} \\\\mathbf{h}^{(r)}\\n \\\\end{equation}\\n\\n where :math:`a_{i}^{(r)}` denotes the importance of relationr in generating the final embedding of node videfined as:\\n\\n .. math::\\n \\\\begin{equation}\\n a_{i}^{(r)}=\\\\frac{\\\\exp \\\\left(\\\\mathbf{q}^{(r)} \\\\cdot \\\\mathbf{h}_{i}^{(r)}\\\\right)}{\\\\sum_{r^{\\\\prime} \\\\in \\\\mathcal{R}} \\\\exp \\\\left(\\\\mathbf{q}^{\\\\left(r^{\\\\prime}\\\\right)} \\\\cdot \\\\mathbf{h}_{i}^{r^{\\\\prime}}\\\\right)}\\n \\\\end{equation}\\n \\\"\\\"\\\"\\n\\n h_1_all_lst = [];h_2_all_lst = [];c_all_lst = []\\n for h_idx in range(self.nheads):\\n h_1_all_, h_2_all_, c_all_ = self.attn[h_idx](h_1_all, h_2_all, c_all)\\n h_1_all_lst.append(h_1_all_);h_2_all_lst.append(h_2_all_); c_all_lst.append(c_all_)\\n\\n h_1_all = torch.mean(torch.cat(h_1_all_lst, 0), 0).unsqueeze(0)\\n h_2_all = torch.mean(torch.cat(h_2_all_lst, 0), 0).unsqueeze(0)\\n\\n else:\\n h_1_all = torch.mean(torch.cat(h_1_all, 0), 0).unsqueeze(0)\\n h_2_all = torch.mean(torch.cat(h_2_all, 0), 0).unsqueeze(0)\\n\\n # Lcs = [Z − AVG { H(r)|r∈ R }]^2 - [Z − AVG { ~H(r)|r∈ R }]^2\\n pos_reg_loss = ((self.H - h_1_all) ** 2).sum()\\n neg_reg_loss = ((self.H - h_2_all) ** 2).sum()\\n reg_loss = pos_reg_loss - neg_reg_loss\\n result['reg_loss'] = reg_loss\\n\\n # semi-supervised module\\n if self.isSemi:\\n r\\\"\\\"\\\"\\n Extension to Semi-Supervised Learning\\n\\n .. math::\\n \\\\begin{equation}\\n \\\\ell_{\\\\text {sup }}=-\\\\frac{1}{\\\\left|\\\\mathcal{Y}_{L}\\\\right|} \\\\sum_{l \\\\in \\\\mathcal{Y}_{L}} \\\\sum_{i=1}^{c} Y_{l i} \\\\ln \\\\hat{Y}_{l i}\\n \\\\end{equation}\\n\\n Where :math:`mathcal{Y}_{L}` is the set of node indices with labels\\n \\\"\\\"\\\"\\n semi = self.logistic(self.H).squeeze(0)\\n result['semi'] = semi\\n\\n # result: ['logits','reg_loss','semi']\\n return result\",\n \"def __init__(self, # noqa: R0913\\n hid_dim,\\n n_heads,\\n pf_dim,\\n dropout):\\n super().__init__()\\n\\n self.self_attn_layer_norm = nn.LayerNorm(hid_dim)\\n self.enc_attn_layer_norm = nn.LayerNorm(hid_dim)\\n self.ff_layer_norm = nn.LayerNorm(hid_dim)\\n self.self_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\\n self.encoder_attention = MultiHeadAttentionLayer(hid_dim, n_heads, dropout)\\n self.positionwise_feedforward = PositionwiseFeedforwardLayer(hid_dim,\\n pf_dim,\\n dropout)\\n self.dropout = nn.Dropout(dropout)\",\n \"def _add_bert_self_attention_layer(model_1, model_2):\\n\\n class Config:\\n pass\\n\\n config = Config()\\n config.hidden_size = model_1.all_head_size + model_2.all_head_size\\n config.num_attention_heads = (\\n model_1.num_attention_heads + model_2.num_attention_heads\\n )\\n config.output_attentions = model_1.output_attentions\\n config.attention_probs_dropout_prob = model_1.dropout.p\\n\\n return BertSelfAttention(config)\",\n \"def forward(self,output,target,Pinv,pts_2d,device):\\n \\n # concatenate a third row of ones with pts_2d\\n ones = torch.ones((pts_2d.shape[0],1,pts_2d.shape[-1])).to(device)\\n pts_2d = torch.cat((pts_2d,ones),1)\\n \\n # multiply pts_2d by output and target, respectively\\n output = output.unsqueeze(1)\\n target = target.unsqueeze(1)\\n out_scaled = torch.mul(pts_2d,output)\\n targ_scaled = torch.mul(pts_2d,target)\\n \\n # multiply each result by Pinv\\n out_3d = torch.bmm(Pinv,out_scaled)\\n targ_3d = torch.bmm(Pinv,targ_scaled)\\n loss = nn.SmoothL1Loss()#MSELoss()\\n return loss(out_3d,targ_3d)\",\n \"def forward(self, inp):\\n out = self.features(inp)\\n out = out.view(out.size(0), -1) # linearized the output of the module 'features'\\n out = self.classifier(out)\\n return out\",\n \"def forward(self, enc_states, h_prev):\\n\\t\\talpha_t = torch.bmm(h_prev.transpose(0,1), enc_states.transpose(1,2)) # [batch, 1, seq_len]\\n\\t\\talpha_t = F.softmax(alpha_t, dim=-1)\\n\\t\\tc_t = torch.bmm(alpha_t, enc_states) # [batch, 1, dim]\\n\\t\\treturn c_t\",\n \"def interAct(model, encoded_query, dataloader_tester, params_runtime , args,mark='Interactive', verbose=False):\\n # set model to evaluation mode\\n model.eval()\\n predictions = []\\n for batch in dataloader_tester:\\n \\n batch = tuple(b.to(params_runtime.device) for b in batch)\\n \\n inputs = {'input_ids': batch[0],\\n 'attention_mask': batch[1],\\n }\\n \\n outputs = model(**inputs)\\n logits = outputs[0]\\n #loss_val_total += loss.item()\\n sm = nn.Softmax(dim=1)\\n logits = sm(logits)\\n \\n logits = logits.detach().cpu().numpy()\\n predictions.append(logits)\\n predictions = np.concatenate(predictions, axis=0)\\n \\n with open(args.pretrained_dir+'/params.json', 'r') as fp:\\n label_dict = json.load(fp)\\n pred_label = get_label(predictions, label_dict)\\n \\n \\n return(pred_label)\",\n \"def make_mlp_model():\\n return snt.Sequential([\\n snt.nets.MLP([LATENT_SIZE] * NUM_LAYERS, activate_final=True),\\n snt.LayerNorm()\\n ])\",\n \"def __init__(self, is_training, transductive_batch_norm,\\n backprop_through_moments, ema_object, embedding_fn, reader,\\n weight_decay, num_update_steps, additional_test_update_steps,\\n first_order, alpha, train_batch_norm, depth_multiplier, debug,\\n zero_fc_layer, proto_maml_fc_layer_on_query_set,\\n proto_maml_fc_layer_on_support_set, proto_maml_fc_layer_init):\\n super(MAMLLearner, self).__init__(is_training, transductive_batch_norm,\\n backprop_through_moments, ema_object,\\n embedding_fn, reader)\\n\\n maml_embedding_fns = [\\n four_layer_convnet_maml,\\n resnet_maml,\\n ]\\n if not any(self.embedding_fn is maml_fn for maml_fn in maml_embedding_fns):\\n raise ValueError('MAML requires a specific architecture to work.')\\n\\n # For aggregating statistics later.\\n self.test_targets = self.data.test_labels\\n self.way = compute_way(self.data)\\n\\n self.weight_decay = weight_decay\\n self.alpha = alpha\\n self.num_update_steps = num_update_steps\\n self.additional_test_update_steps = additional_test_update_steps\\n self.first_order = first_order\\n self.train_batch_norm = train_batch_norm\\n self.debug_log = debug\\n self.depth_multiplier = depth_multiplier\\n self.zero_fc_layer = zero_fc_layer\\n self.proto_maml_fc_layer_on_query_set = proto_maml_fc_layer_on_query_set\\n self.proto_maml_fc_layer_on_support_set = proto_maml_fc_layer_on_support_set\\n self.proto_maml_fc_layer_init = proto_maml_fc_layer_init\\n\\n tf.logging.info('alpha: {}, num_update_steps: {}'.format(\\n self.alpha, self.num_update_steps))\\n self.forward_pass()\",\n \"def __init__(self, albert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([albert_layer.attention.query.weight, albert_layer.attention.key.weight, albert_layer.attention.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([albert_layer.attention.query.bias, albert_layer.attention.key.bias, albert_layer.attention.value.bias]))\\n self.out_proj_weight = albert_layer.attention.dense.weight\\n self.out_proj_bias = albert_layer.attention.dense.bias\\n self.linear1_weight = albert_layer.ffn.weight\\n self.linear1_bias = albert_layer.ffn.bias\\n self.linear2_weight = albert_layer.ffn_output.weight\\n self.linear2_bias = albert_layer.ffn_output.bias\\n self.norm1_eps = albert_layer.attention.LayerNorm.eps\\n self.norm1_weight = albert_layer.attention.LayerNorm.weight\\n self.norm1_bias = albert_layer.attention.LayerNorm.bias\\n self.norm2_eps = albert_layer.full_layer_layer_norm.eps\\n self.norm2_weight = albert_layer.full_layer_layer_norm.weight\\n self.norm2_bias = albert_layer.full_layer_layer_norm.bias\\n self.num_heads = albert_layer.attention.num_attention_heads\\n self.embed_dim = albert_layer.attention.all_head_size\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def forward(self, output):\\n \\n hidden_states = self.extract_hidden_states(output)\\n \\n # Obtaining the attention weights\\n weighted_states = self.w1(hidden_states)\\n activated_states = self.tanh(weighted_states)\\n score_weights = self.w2(activated_states)\\n attention_weights = self.softmax(score_weights)\\n \\n # Applying attention to the matrix with hidden states\\n attentional_vector = torch.bmm(torch.transpose(attention_weights,2,1),hidden_states) \\n attentional_vector = self.fc(torch.transpose(attentional_vector,2,1)).squeeze(2)\\n attentional_vector = self._activation_fn(attentional_vector)\\n \\n return attentional_vector\",\n \"def forward(\\n self,\\n input_ids=None,\\n attention_mask=None,\\n token_type_ids=None,\\n masked_lm_labels=None,\\n ent_masked_lm_labels=None,\\n prot_masked_lm_labels=None,\\n return_dict=None,\\n head_mask=None,\\n ):\\n # No backpropagation is needed for getting the initial embeddings from the backbones\\n with torch.no_grad():\\n # 1. Use the LM backbone to get the pre-trained token embeddings\\n # batch x number_text_tokens x hidden_size\\n # The first element of the returned tuple from the LM backbone forward() pass is the sequence of hidden\\n # states\\n text_embeddings = torch.cat(\\n [\\n self.lm_backbone(\\n input_ids[\\n :, i * (self.kg_start_idx // 3) : (i + 1) * (self.kg_start_idx // 3)\\n ]\\n )[0]\\n for i in range(3)\\n ],\\n dim=1,\\n )\\n\\n # 2. Use the KG backbone to obtain the pre-trained entity embeddings\\n # batch x number_kg_tokens x hidden_size\\n ent_embeddings = torch.stack(\\n [\\n # for each numeric index in the random walks sequence: get the embedding vector from the KG backbone\\n torch.stack([self.kg_backbone[i.item()] for i in j])\\n # for each example in the batch: get the random walks sequence\\n for j in input_ids[:, self.kg_start_idx : self.prot_start_idx]\\n ],\\n )\\n # 3. Use the Prot backbone to obtain the pre-trained entity embeddings\\n # batch x number_prot_tokens x prot_hidden_size (prot_hidden_size != hidden_size)\\n prot_embeddings_original_dim = self.prot_backbone(input_ids[:, self.prot_start_idx :])[\\n 0\\n ]\\n\\n # Additional layer to project prot_hidden_size onto hidden_size\\n prot_embeddings = self.prot_to_lm_hidden_linear(prot_embeddings_original_dim)\\n\\n # Concatenate token, KG and prot embeddings obtained from the LM, KG and prot backbones and cast to float\\n # batch x seq_len x hidden_size\\n inputs_embeds = (\\n torch.cat(\\n [\\n text_embeddings,\\n ent_embeddings.to(text_embeddings.device),\\n prot_embeddings.to(text_embeddings.device),\\n ],\\n dim=1,\\n )\\n .type(torch.FloatTensor)\\n .to(self.device)\\n )\\n\\n # Get the hidden states from the basic ProtSTonKGs Transformer layers\\n # batch x seq_len x hidden_size\\n outputs = self.bert(\\n inputs_embeds=inputs_embeds,\\n attention_mask=attention_mask,\\n return_dict=None,\\n )\\n # batch x seq_len x hidden_size\\n # sequence_output, pooled_output = outputs.last_hidden_state, outputs.pooler_output\\n sequence_output, pooled_output = outputs[:2]\\n\\n # Generate the prediction scores (mapping to text and entity vocab sizes + NSP) for the training objectives\\n # prediction_scores = Text MLM, entity \\\"MLM\\\" and protein \\\"MLM\\\" scores\\n prediction_scores, _ = self.cls(sequence_output, pooled_output)\\n # The custom STonKGsELMPredictionHead returns a triple of prediction scores for tokens, entities,\\n # and protein sequences, respectively\\n (\\n token_prediction_scores,\\n entity_predictions_scores,\\n prot_predictions_scores,\\n ) = prediction_scores\\n\\n # Calculate the loss\\n total_loss = None\\n if (\\n masked_lm_labels is not None\\n and ent_masked_lm_labels is not None\\n and prot_masked_lm_labels is not None\\n ):\\n loss_fct = nn.CrossEntropyLoss()\\n # 1. Text-based MLM\\n masked_lm_loss = loss_fct(\\n token_prediction_scores.view(-1, self.config.lm_vocab_size),\\n masked_lm_labels.view(-1),\\n )\\n # 2. Entity-based masked \\\"language\\\" (entity) modeling\\n ent_masked_lm_loss = loss_fct(\\n entity_predictions_scores.view(-1, self.config.kg_vocab_size),\\n ent_masked_lm_labels.view(-1),\\n )\\n # 3. Protein-based masked \\\"language\\\" (entity) modeling\\n prot_masked_lm_loss = loss_fct(\\n prot_predictions_scores.view(-1, self.config.prot_vocab_size),\\n prot_masked_lm_labels.view(-1),\\n )\\n # Total loss = the sum of the individual training objective losses\\n total_loss = masked_lm_loss + ent_masked_lm_loss + prot_masked_lm_loss\\n\\n if not return_dict:\\n output = prediction_scores + outputs[2:]\\n return ((total_loss,) + output) if total_loss is not None else output\\n\\n return BigBirdForPreTrainingOutputWithPooling(\\n loss=total_loss,\\n prediction_logits=prediction_scores,\\n hidden_states=sequence_output,\\n attentions=outputs.attentions,\\n pooler_output=pooled_output,\\n )\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.63806444","0.62879974","0.61243266","0.6001117","0.59858584","0.58377385","0.57497156","0.57036","0.56529164","0.56093395","0.5535092","0.55335355","0.55295086","0.54910344","0.54799116","0.5470056","0.5456178","0.54472595","0.53975135","0.53934085","0.53866965","0.5383578","0.5338398","0.53125423","0.53017414","0.5287938","0.52775156","0.5275427","0.5269375","0.5254605","0.52182466","0.5211929","0.52111804","0.520527","0.5201256","0.5198322","0.51935077","0.51902175","0.518374","0.5175201","0.5173995","0.5173468","0.5172982","0.5172037","0.51649934","0.5158009","0.515795","0.5155442","0.5149546","0.51391757","0.5135067","0.51347834","0.51333356","0.5133168","0.5131806","0.5125652","0.51208246","0.51184094","0.5105967","0.5105732","0.51044583","0.5103918","0.51024073","0.50960886","0.5094233","0.508038","0.50797194","0.50775003","0.50774187","0.50717956","0.5070488","0.507044","0.5066686","0.5062407","0.5060366","0.505741","0.5056092","0.50516534","0.5046982","0.504585","0.50442785","0.504391","0.50387406","0.50376993","0.50354034","0.50331444","0.5028975","0.5023699","0.5022261","0.5019299","0.50113976","0.50092846","0.50077516","0.5007704","0.5003211","0.49992922","0.49987355","0.49948937","0.49939734","0.4989889","0.4986584"],"string":"[\n \"0.63806444\",\n \"0.62879974\",\n \"0.61243266\",\n \"0.6001117\",\n \"0.59858584\",\n \"0.58377385\",\n \"0.57497156\",\n \"0.57036\",\n \"0.56529164\",\n \"0.56093395\",\n \"0.5535092\",\n \"0.55335355\",\n \"0.55295086\",\n \"0.54910344\",\n \"0.54799116\",\n \"0.5470056\",\n \"0.5456178\",\n \"0.54472595\",\n \"0.53975135\",\n \"0.53934085\",\n \"0.53866965\",\n \"0.5383578\",\n \"0.5338398\",\n \"0.53125423\",\n \"0.53017414\",\n \"0.5287938\",\n \"0.52775156\",\n \"0.5275427\",\n \"0.5269375\",\n \"0.5254605\",\n \"0.52182466\",\n \"0.5211929\",\n \"0.52111804\",\n \"0.520527\",\n \"0.5201256\",\n \"0.5198322\",\n \"0.51935077\",\n \"0.51902175\",\n \"0.518374\",\n \"0.5175201\",\n \"0.5173995\",\n \"0.5173468\",\n \"0.5172982\",\n \"0.5172037\",\n \"0.51649934\",\n \"0.5158009\",\n \"0.515795\",\n \"0.5155442\",\n \"0.5149546\",\n \"0.51391757\",\n \"0.5135067\",\n \"0.51347834\",\n \"0.51333356\",\n \"0.5133168\",\n \"0.5131806\",\n \"0.5125652\",\n \"0.51208246\",\n \"0.51184094\",\n \"0.5105967\",\n \"0.5105732\",\n \"0.51044583\",\n \"0.5103918\",\n \"0.51024073\",\n \"0.50960886\",\n \"0.5094233\",\n \"0.508038\",\n \"0.50797194\",\n \"0.50775003\",\n \"0.50774187\",\n \"0.50717956\",\n \"0.5070488\",\n \"0.507044\",\n \"0.5066686\",\n \"0.5062407\",\n \"0.5060366\",\n \"0.505741\",\n \"0.5056092\",\n \"0.50516534\",\n \"0.5046982\",\n \"0.504585\",\n \"0.50442785\",\n \"0.504391\",\n \"0.50387406\",\n \"0.50376993\",\n \"0.50354034\",\n \"0.50331444\",\n \"0.5028975\",\n \"0.5023699\",\n \"0.5022261\",\n \"0.5019299\",\n \"0.50113976\",\n \"0.50092846\",\n \"0.50077516\",\n \"0.5007704\",\n \"0.5003211\",\n \"0.49992922\",\n \"0.49987355\",\n \"0.49948937\",\n \"0.49939734\",\n \"0.4989889\",\n \"0.4986584\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90879,"cells":{"query":{"kind":"string","value":"Transformer blocks which includes multihead attention layer and mlp block"},"document":{"kind":"string","value":"def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):\n super(TransformerBlock, self).__init__()\n # Attention\n self.attention = MultiHeadAttention(\n num_heads=num_heads, key_dim=D, dropout=dropout\n )\n self.norm_attention = LayerNormalization(epsilon=norm_eps)\n\n # MLP\n self.mlp = MLPBlock(hidden_layers, dropout)\n self.norm_mlp = LayerNormalization(epsilon=norm_eps)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def build_model_multihead_attention_multiscaleCNN4_covermore(self,\n dim_attention,headnum,\n embedding_vec,\n load_weights=False, weight_dir=None,\n nb_filters=32,filters_length1=1,\n filters_length2=5,\n filters_length3=10,\n pooling_size=3,\n drop_input=0,\n drop_cnn=0.2,\n drop_flat=0,\n W1_regularizer=0.005,\n W2_regularizer=0.005,\n Att_regularizer_weight=0.0005,\n BatchNorm=False,\n fc_dim = 50,\n fcnum=0,\n posembed=False,\n pos_dmodel=40,\n pos_nwaves = 20,\n posmod = 'concat',\n regularfun=1,\n huber_delta=1,\n activation='gelu',\n activationlast='gelu',\n add_avgpooling = False,\n poolingmod=1,\n normalizeatt=False,\n regressionmodel=False,\n attmod = \"softmax\",\n sharp_beta=1,\n lr = 0.001 \n ):\n ###print('Advanced Masking')\n def mask_func(x):\n return x[0] * x[1]\n \n ###print(posembed)\n ###print(posmod)\n input = Input(shape=(self.max_len,), dtype='int8')\n input_mask = Input(shape=([int(self.max_len/pooling_size), 1]), dtype='float32')\n embedding_layer = Embedding(len(embedding_vec), len(embedding_vec[0]), weights=[embedding_vec],\n input_length=self.max_len,\n trainable=False)\n embedding_output = Dropout(drop_input)(embedding_layer(input)) #layer2\n if 'gelu' in activation:\n activationfun=gelu\n else:\n activationfun = 'relu'\n \n if 'gelu' in activationlast:\n activationlastfun = gelu\n else:\n activationlastfun='relu'\n \n ###print(activationfun)\n ###print(activationlastfun)\n with tf.name_scope('first_cnn'):\n first_cnn = Convolution1D(nb_filters, filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN1')(embedding_output) #layer3\n first_cnn2 = Convolution1D(int(nb_filters/2), filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(first_cnn) #layer5\n second_cnn = Convolution1D(nb_filters, filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN2')(embedding_output) #layer4\n second_cnn2 = Convolution1D(int(nb_filters/2), filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(second_cnn)\n third_cnn = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationfun, use_bias=False,name='CNN3')(embedding_output)\n \n third_cnn2 = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\n border_mode='same', activation=activationlastfun, use_bias=False)(third_cnn)\n if BatchNorm:\n first_cnn2 = BatchNormalization()(first_cnn2)\n second_cnn2 = BatchNormalization()(second_cnn2)\n third_cnn2 = BatchNormalization()(third_cnn2)\n \n if not add_avgpooling:\n if poolingmod == 1:\n pooling_layer = MaxPooling1D(pool_length=pooling_size, stride=pooling_size)\n else:\n pooling_layer = AveragePooling1D(pool_length=pooling_size, stride=pooling_size)\n \n cnn_output1 = Dropout(drop_cnn)(pooling_layer(first_cnn2))\n cnn_output2 = Dropout(drop_cnn)(pooling_layer(second_cnn2))\n cnn_output3 = Dropout(drop_cnn)(pooling_layer(third_cnn2))\n else:\n first_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\n first_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\n cnn_output1 = Dropout(drop_cnn)(concatenate([first_cnn2_max,first_cnn2_avg],axis=-1))\n second_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\n second_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\n cnn_output2 = Dropout(drop_cnn)(concatenate([second_cnn2_max,second_cnn2_avg],axis=-1))\n third_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\n third_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\n cnn_output3 = Dropout(drop_cnn)(concatenate([third_cnn2_max,third_cnn2_avg],axis=-1))\n \n \n \n if posembed:\n ##print(posmod)\n from position_embedding import PositionEmbedding\n if posmod == 'concat':\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb1')(cnn_output1)\n cnn_output1 = concatenate([cnn_output1, pos_emb1], axis=-1)\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb2')(cnn_output2)\n cnn_output2 = concatenate([cnn_output2, pos_emb2], axis=-1)\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb3')(cnn_output3)\n cnn_output3 = concatenate([cnn_output3, pos_emb3], axis=-1)\n else:\n ##print(\"yes add posmod\")\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output1)[-1]/2), d_model=int_shape(cnn_output1)[-1],name='pos_emb1')(cnn_output1)\n cnn_output1 = Add()([cnn_output1, pos_emb1])\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output2)[-1]/2), d_model=int_shape(cnn_output2)[-1],name='pos_emb2')(cnn_output2)\n cnn_output2 = Add()([cnn_output2, pos_emb2])\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output3)[-1]/2), d_model=int_shape(cnn_output3)[-1],name='pos_emb3')(cnn_output3)\n cnn_output3 = Add()([cnn_output3, pos_emb3])\n \n mask_input1 = []\n mask_input1.append(cnn_output1)\n mask_input1.append(input_mask)\n cnn_mask_output1 = Lambda(mask_func)(mask_input1)\n del mask_input1\n mask_input2 = []\n mask_input2.append(cnn_output2)\n mask_input2.append(input_mask)\n cnn_mask_output2 = Lambda(mask_func)(mask_input2)\n del mask_input2\n mask_input3 = []\n mask_input3.append(cnn_output3)\n mask_input3.append(input_mask)\n cnn_mask_output3 = Lambda(mask_func)(mask_input3)\n del mask_input3\n \n if regularfun==1:\n regularizerfunction_W1 = regularizers.l1(W1_regularizer)\n regularizerfunction_W2 = regularizers.l1(W2_regularizer)\n elif regularfun==2:\n regularizerfunction_W1 = regularizers.l2(W1_regularizer)\n regularizerfunction_W2 = regularizers.l2(W2_regularizer)\n elif regularfun ==3:\n regularizerfunction_W1 = smoothL1(W1_regularizer,huber_delta)\n regularizerfunction_W2 = smoothL1(W2_regularizer,huber_delta)\n \n with tf.name_scope('multihead_attention'):\n att1,att1_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att1\")(concatenate([cnn_mask_output1, input_mask]))#-5 layer\n \n att2,att2_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att2\")(concatenate([cnn_mask_output2, input_mask])) #-4 layer\n \n att3,att3_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\n W1_regularizer=regularizerfunction_W1,\n W2_regularizer=regularizerfunction_W2,\n W1_constraint=None, W2_constraint=None, return_attention=True,\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\"att3\")(concatenate([cnn_mask_output3, input_mask])) #-3 layer\n \n if BatchNorm:\n att1 = BatchNormalization()(att1)\n att2 = BatchNormalization()(att2)\n att3 = BatchNormalization()(att3)\n \n \n output = Dropout(drop_flat)(Flatten()(concatenate([att1,att2,att3]))) #-2 layer\n \n fc = output\n for _ in range(fcnum):\n fc = Dense(fc_dim,activation='relu')(fc)\n fc = Dropout(drop_flat)(fc)\n \n with tf.name_scope(''):\n if regressionmodel:\n preds = Dense(self.nb_classes,activation='softmax')(fc) #-1 layer\n else:\n preds = Dense(self.nb_classes,activation='sigmoid')(fc) #-1 layer\n \n self.model = Model(inputs=[input,input_mask], outputs=preds)\n from keras import optimizers\n # optim = optimizers.RMSprop()\n optim = optimizers.Adam(lr=lr, decay=5e-5) #The paper uses a decay rate alpha = alpha/sqrt(t) updted each epoch (t) for the logistic regression demonstration.\n #optim = optimizers.nadam()\n #optim = RAdam()\n if regressionmodel:\n self.model.compile(loss='kld',optimizer=optim,metrics=['acc'])\n else:\n self.model.compile(loss='binary_crossentropy',optimizer=optim,metrics=['binary_accuracy','categorical_accuracy'])\n \n \n \n if load_weights:\n self.model.load_weights(weight_dir)\n \n self.is_built = True\n self.bn = False\n self.model.summary()","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n CustomTransformerBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n backward_compatible=self.backward_compatible,\n float_type=self.float_type,\n name=(\"layer_%d\" % i)))\n super(CustomTransformer, self).build(unused_input_shapes)","def __init__(self, dm, h, hidden, drop_rate=0.1):\r\n super(EncoderBlock, self).__init__()\r\n self.mha = MultiHeadAttention(dm, h)\r\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\r\n activation='relu')\r\n self.dense_output = tf.keras.layers.Dense(units=dm)\r\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\r\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\r\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)","def mtf_transformer2_base():\n hparams = common_hparams.basic_params1()\n\n hparams.add_hparam(\"d_model\", 1024)\n hparams.batch_size = 4\n hparams.max_length = 1024\n hparams.label_smoothing = 0.0\n # a small positive value - this seems important for stability when training\n # with bfloat16 activations.\n hparams.add_hparam(\"z_loss\", 1e-4)\n\n # hparams applying to both encoder and decoder layer stacks.\n hparams.add_hparam(\"d_ff\", 2048)\n hparams.add_hparam(\"d_kv\", 128)\n hparams.add_hparam(\"attention_dropout\", 0.0)\n hparams.add_hparam(\"relu_dropout\", 0.0)\n hparams.del_hparam(\"num_heads\")\n hparams.del_hparam(\"num_hidden_layers\")\n hparams.layer_prepostprocess_dropout = 0.0\n hparams.add_hparam(\"extra_logit\", False)\n # number of experts for moe_1d\n hparams.moe_num_experts = 32\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\n hparams.add_hparam(\"moe_expert_x\", 8)\n hparams.add_hparam(\"moe_expert_y\", 4)\n hparams.add_hparam(\"moe_hidden_size\", 32768)\n\n # round up vocab sizes to be a multiple of this value\n hparams.vocab_divisor = 128\n\n hparams.optimizer = \"Adafactor\"\n hparams.learning_rate_schedule = \"rsqrt_decay*linear_decay\"\n hparams.learning_rate_warmup_steps = 10000\n hparams.add_hparam(\"master_dtype\", \"bfloat16\")\n hparams.add_hparam(\"slice_dtype\", \"float32\")\n hparams.activation_dtype = \"bfloat16\"\n\n # 8-way model-parallelism\n hparams.add_hparam(\"mesh_shape\", \"model:8\")\n hparams.add_hparam(\"layout\", \"batch:batch;vocab:model;d_ff:model;heads:model\")\n\n # If nonzero, we split the batch across two tensor-dimensions named\n # \"outer_batch\" and \"inner_batch\", allowing for splitting across two mesh\n # dimensions. This is necessary for hierarchical mixture of experts.\n # The two tensor dimensions have sizes hparams.outer_batch_size and\n # hparams.batch_size // hparams.outer_batch_size.\n hparams.add_hparam(\"outer_batch_size\", 0)\n\n hparams.shared_embedding_and_softmax_weights = False\n # length for training or decoding - defaults to max_length\n hparams.add_hparam(\"length\", 0)\n\n # These parameters make Transformer model compatible with mtf\n # Do not override these.\n hparams.no_data_parallelism = True\n hparams.use_fixed_batch_size = True\n hparams.add_hparam(\"mtf_mode\", True)\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\n hparams.bottom = {\n \"inputs\": modalities.identity_bottom,\n \"targets\": modalities.identity_bottom,\n }\n hparams.top = {\n \"targets\": modalities.identity_top,\n }\n hparams.add_hparam(\"beam_size\", 1)\n\n # If this is True, then in a packed dataset (where exaples are concatenated\n # to form longer examples) we use the global position (within the concatenated\n # sequence) to compute the positional embedding, instead of the position\n # within the individual sequence. This is counterintuitive, but for some\n # reason, it keeps the model from diverging.\n hparams.add_hparam(\"use_global_position_in_packed_sequence\", True)\n\n return hparams","def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\n super(TransformerEncoder, self).__init__()\n\n # Create num_layers of TransformerBlock\n self.encoder = Sequential(\n [\n TransformerBlock(num_heads=num_heads,\n D=D,\n hidden_layers=[mlp_dim, D],\n dropout=dropout,\n norm_eps=norm_eps)\n for _ in range(num_layers)\n ]\n )","def attention_module_multi_head(roi_feat, position_embedding,\n nongt_dim, fc_dim, feat_dim,\n dim=(1024, 1024, 1024),\n group=16, index=1):\n dim_group = (dim[0] / group, dim[1] / group, dim[2] / group)\n nongt_roi_feat = mx.symbol.slice_axis(data=roi_feat, axis=0, begin=0, end=nongt_dim)\n # [num_rois * nongt_dim, emb_dim]\n position_embedding_reshape = mx.sym.Reshape(position_embedding, shape=(-3, -2))\n # position_feat_1, [num_rois * nongt_dim, fc_dim]\n position_feat_1 = mx.sym.FullyConnected(name='pair_pos_fc1_' + str(index),\n data=position_embedding_reshape,\n num_hidden=fc_dim)\n position_feat_1_relu = mx.sym.Activation(data=position_feat_1, act_type='relu')\n # aff_weight, [num_rois, nongt_dim, fc_dim]\n aff_weight = mx.sym.Reshape(position_feat_1_relu, shape=(-1, nongt_dim, fc_dim))\n # aff_weight, [num_rois, fc_dim, nongt_dim]\n aff_weight = mx.sym.transpose(aff_weight, axes=(0, 2, 1))\n\n # multi head\n assert dim[0] == dim[1], 'Matrix multiply requires same dimensions!'\n q_data = mx.sym.FullyConnected(name='query_' + str(index),\n data=roi_feat,\n num_hidden=dim[0])\n q_data_batch = mx.sym.Reshape(q_data, shape=(-1, group, dim_group[0]))\n q_data_batch = mx.sym.transpose(q_data_batch, axes=(1, 0, 2))\n k_data = mx.symbol.FullyConnected(name='key_' + str(index),\n data=nongt_roi_feat,\n num_hidden=dim[1])\n k_data_batch = mx.sym.Reshape(k_data, shape=(-1, group, dim_group[1]))\n k_data_batch = mx.sym.transpose(k_data_batch, axes=(1, 0, 2))\n v_data = nongt_roi_feat\n # v_data = mx.symbol.FullyConnected(name='value_'+str(index)+'_'+str(gid), data=roi_feat, num_hidden=dim_group[2])\n aff = mx.symbol.batch_dot(lhs=q_data_batch, rhs=k_data_batch, transpose_a=False, transpose_b=True)\n # aff_scale, [group, num_rois, nongt_dim]\n aff_scale = (1.0 / math.sqrt(float(dim_group[1]))) * aff\n aff_scale = mx.sym.transpose(aff_scale, axes=(1, 0, 2))\n\n assert fc_dim == group, 'fc_dim != group'\n # weighted_aff, [num_rois, fc_dim, nongt_dim]\n weighted_aff = mx.sym.log(mx.sym.maximum(left=aff_weight, right=1e-6)) + aff_scale\n aff_softmax = mx.symbol.softmax(data=weighted_aff, axis=2, name='softmax_' + str(index))\n # [num_rois * fc_dim, nongt_dim]\n aff_softmax_reshape = mx.sym.Reshape(aff_softmax, shape=(-3, -2))\n # output_t, [num_rois * fc_dim, feat_dim]\n output_t = mx.symbol.dot(lhs=aff_softmax_reshape, rhs=v_data)\n # output_t, [num_rois, fc_dim * feat_dim, 1, 1]\n output_t = mx.sym.Reshape(output_t, shape=(-1, fc_dim * feat_dim, 1, 1))\n # linear_out, [num_rois, dim[2], 1, 1]\n linear_out = mx.symbol.Convolution(name='linear_out_' + str(index), data=output_t,\n kernel=(1, 1), num_filter=dim[2], num_group=fc_dim)\n output = mx.sym.Reshape(linear_out, shape=(0, 0))\n return output","def mtf_bitransformer_base():\n hparams = mtf_transformer2_base()\n hparams.max_length = 256\n hparams.shared_embedding = True\n # HYPERPARAMETERS FOR THE LAYER STACKS\n hparams.add_hparam(\"encoder_layers\", [\"self_att\", \"drd\"] * 6)\n hparams.add_hparam(\"decoder_layers\", [\"self_att\", \"enc_att\", \"drd\"] * 6)\n hparams.add_hparam(\"encoder_num_layers\", 6)\n hparams.add_hparam(\"decoder_num_layers\", 6)\n # number of heads in multihead attention\n hparams.add_hparam(\"encoder_num_heads\", 8)\n hparams.add_hparam(\"decoder_num_heads\", 8)\n hparams.add_hparam(\"local_attention_radius\", 128)\n\n # default of 0 for standard transformer behavior\n # 1 means a single set of keys and values that are read by all query heads\n hparams.add_hparam(\"encoder_num_memory_heads\", 0)\n hparams.add_hparam(\"decoder_num_memory_heads\", 0)\n # share attention keys and values\n hparams.add_hparam(\"encoder_shared_kv\", False)\n hparams.add_hparam(\"decoder_shared_kv\", False)\n\n # Parameters for computing the maximum decode length in beam search.\n # Maximum decode length is:\n # min(max_length,\n # decode_length_multiplier * input_length + decode_length_constant)\n hparams.add_hparam(\"decode_length_multiplier\", 1.5)\n hparams.add_hparam(\"decode_length_constant\", 10.0)\n # used during decoding\n hparams.add_hparam(\"alpha\", 0.6)\n hparams.sampling_temp = 0.0\n return hparams","def mtf_unitransformer_base():\n hparams = mtf_transformer2_base()\n hparams.add_hparam(\"autoregressive\", True)\n # HYPERPARAMETERS FOR THE SINGLE LAYER STACK\n hparams.add_hparam(\"layers\", [\"self_att\", \"drd\"] * 6)\n # number of heads in multihead attention\n hparams.add_hparam(\"num_heads\", 8)\n # default of 0 for standard transformer behavior\n # 1 means a single set of keys and values that are read by all query heads\n hparams.add_hparam(\"num_memory_heads\", 0)\n # share attention keys and values\n hparams.add_hparam(\"shared_kv\", False)\n # if nonzero then use local attention\n hparams.add_hparam(\"local_attention_radius\", 128)\n return hparams","def forward(self, x):\n # Encoder1 --block1\n encode_block1 = self.conv_encoder1(x)\n if self.residus[0] == 1:\n encode_block1 += self.residual_shortcut1(x)\n encode_pool1 = self.max_pool_encoder1(encode_block1)\n\n # Encoder2 --block2\n encode_block2 = self.conv_encoder2(encode_pool1)\n if self.residus[1] == 1:\n encode_block2 += self.residual_shortcut2(encode_pool1)\n encode_pool2 = self.max_pool_encoder2(encode_block2)\n \n # Encoder3 --block3\n encode_block3 = self.conv_encoder3(encode_pool2)\n if self.residus[2] == 1:\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\n encode_pool3 = self.max_pool_encoder3(encode_block3)\n \n # Encoder4 --block4\n encode_block4 = self.conv_encoder4(encode_pool3) \n if self.residus[3] == 1:\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\n encode_pool4 = self.max_pool_encoder4(encode_block4)\n\n # Transitional block \n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\n \n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\n \n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\n\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\n \n convTranspose_transitional = self.convTranspose_transitional(middle_block) \n # Decoder4 --block5\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \n if self.residus[4] == 1:\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\n\n #--block6\n cat_layer3 = self.conv_decoder4(decode_block4) \n if self.residus[5] == 1:\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\n \n \n # Decoder3 --block7\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\n if self.residus[6] == 1:\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\n \n #--block8\n cat_layer2 = self.conv_decoder3(decode_block3) \n if self.residus[7] == 1:\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\n \n # Decoder2 --block9\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \n if self.residus[8] == 1:\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\n \n #--block10\n cat_layer1 = self.conv_decoder2(decode_block2)\n if self.residus[9] == 1:\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\n \n # Decoder1 --block11\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \n if self.residus[10] == 1:\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\n \n #--block12\n final_layer = self.final_layer(decode_block1)\n if self.residus[11] == 1:\n final_layer += self.residual_shortcut_final_layer(decode_block1)\n \n \n return final_layer","def __init__(self, ):\n super().__init__()\n channels = 3\n\n # Initial convolution block\n out_features = 64\n # encoder\n self.input = nn.Sequential(\n nn.ReflectionPad2d(channels),\n nn.Conv2d(3, out_features, (7, 7)),\n nn.InstanceNorm2d(out_features),\n nn.MaxPool2d(2),\n nn.LeakyReLU(0.2, inplace=True),\n nn.Conv2d(out_features, out_features * 2, (3, 3), stride=(1, 1), padding=(1, 1)),\n nn.LeakyReLU(0.2, inplace=True),\n nn.MaxPool2d(2),\n nn.Conv2d(out_features * 2, out_features * 4, (3, 3), stride=(1, 1), padding=(1, 1)),\n nn.LeakyReLU(0.2, inplace=True),\n )\n\n self.in0 = nn.InstanceNorm2d(256)\n self.block0 = blocks()\n self.block1 = blocks()\n self.block2 = blocks()\n self.block3 = blocks()\n self.block4 = blocks()\n self.block5 = blocks()\n self.block6 = blocks()\n self.block7 = blocks()\n\n self.out = nn.Sequential(\n\n nn.Upsample(scale_factor=2),\n nn.Conv2d(out_features * 4, out_features * 2, 3, stride=1, padding=1),\n nn.InstanceNorm2d(out_features * 2),\n nn.LeakyReLU(0.2, inplace=True),\n\n nn.Upsample(scale_factor=2),\n nn.Conv2d(out_features * 2, out_features, 3, stride=1, padding=1),\n nn.InstanceNorm2d(out_features),\n nn.LeakyReLU(0.2, inplace=True),\n\n nn.ReflectionPad2d(channels),\n nn.Conv2d(out_features, channels, 7),\n nn.Tanh(),\n )","def __init__(self):\n super(Encoder3, self).__init__()\n self.lblocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n ]\n )\n\n self.blocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n ]\n )","def __init__(self):\n super(Encoder2, self).__init__()\n self.lblocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n ),\n ]\n )\n\n self.blocks = nn.ModuleList(\n [\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n ),\n ]\n )","def build(self, unused_input_shapes):\n\n dense3d_impl = Dense3D\n dense2dprojection_impl = Dense2DProjection\n\n self.attention_layer = CustomAttention(\n num_attention_heads=self.num_attention_heads,\n size_per_head=self.attention_head_size,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n backward_compatible=self.backward_compatible,\n name=\"self\")\n\n self.attention_output_dense = dense3d_impl(\n num_attention_heads=self.num_attention_heads,\n size_per_head=int(self.hidden_size / self.num_attention_heads),\n kernel_initializer=get_initializer(self.initializer_range),\n output_projection=True,\n backward_compatible=self.backward_compatible,\n name=\"dense\")\n self.attention_dropout = tf.keras.layers.Dropout(\n rate=self.hidden_dropout_prob)\n self.attention_layer_norm = (\n tf.keras.layers.LayerNormalization(\n name=LAYER_NORM_NAME, axis=-1, epsilon=1e-12,\n # We do layer norm in float32 for numeric stability.\n dtype=tf.float32))\n\n self.intermediate_dense = dense2dprojection_impl(\n output_size=self.intermediate_size,\n kernel_initializer=get_initializer(self.initializer_range),\n activation=self.intermediate_activation,\n # Uses float32 so that gelu activation is done in float32.\n fp32_activation=True, name='dense')\n self.output_dense = dense2dprojection_impl(\n output_size=self.hidden_size,\n kernel_initializer=get_initializer(self.initializer_range),\n name='dense')\n self.output_dropout = tf.keras.layers.Dropout(rate=self.hidden_dropout_prob)\n self.output_layer_norm = tf.keras.layers.LayerNormalization(axis=-1,\n epsilon=1e-12, dtype=tf.float32, name=LAYER_NORM_NAME)\n super(CustomTransformerBlock, self).build(unused_input_shapes)","def __init__(self, in_features, out_features, activation, batch_norm=False, combine_with_1x1=False):\n super(InceptionBlock, self).__init__()\n\n self.activation = activation\n\n if not combine_with_1x1:\n if out_features % 4 != 0:\n raise ValueError(\"If not combining with 1x1, out_features must be divisible by 4\")\n conv_out_features = int(out_features/4)\n else:\n conv_out_features = out_features\n\n if batch_norm:\n self.batch_norm = nn.BatchNorm2d(conv_out_features)\n self.batch_norm1 = nn.BatchNorm2d(conv_out_features)\n self.batch_norm3 = nn.BatchNorm2d(conv_out_features)\n self.batch_norm5 = nn.BatchNorm2d(conv_out_features)\n self.batch_norm7 = nn.BatchNorm2d(conv_out_features)\n else:\n self.batch_norm = None\n\n self.conv1x1 = nn.Conv2d(in_features, conv_out_features, kernel_size=1, padding=0)\n self.conv3x3 = nn.Conv2d(in_features, conv_out_features, kernel_size=3, padding=1)\n self.conv5x5 = nn.Conv2d(in_features, conv_out_features, kernel_size=5, padding=2)\n self.conv7x7 = nn.Conv2d(in_features, conv_out_features, kernel_size=7, padding=3)\n\n self.combine_with_1x1 = combine_with_1x1\n if combine_with_1x1:\n self.combiner = nn.Conv2d(out_features*4, out_features, kernel_size=1)\n self.batch_norm_comb = nn.BatchNorm2d(out_features)","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n TransformerDecoderBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n multi_channel_cross_attention=self.multi_channel_cross_attention,\n name=(\"layer_%d\" % i)))\n super(TransformerDecoder, self).build(unused_input_shapes)","def __init__(\n self,\n in_channels=128,\n aux_channels=80,\n channels=64,\n out_channels=1,\n kernel_size=9,\n dilation=2,\n bias=True,\n noise_upsample_scales=[11, 2, 2, 2],\n noise_upsample_activation=\"LeakyReLU\",\n noise_upsample_activation_params={\"negative_slope\": 0.2},\n upsample_scales=[2, 2, 2, 2, 2, 2, 2, 2, 1],\n upsample_mode=\"nearest\",\n gated_function=\"softmax\",\n use_weight_norm=True,\n ):\n super().__init__()\n\n self.in_channels = in_channels\n\n noise_upsample = []\n in_chs = in_channels\n for noise_upsample_scale in noise_upsample_scales:\n # NOTE(kan-bayashi): How should we design noise upsampling part?\n noise_upsample += [\n torch.nn.ConvTranspose1d(\n in_chs,\n channels,\n noise_upsample_scale * 2,\n stride=noise_upsample_scale,\n padding=noise_upsample_scale // 2 + noise_upsample_scale % 2,\n output_padding=noise_upsample_scale % 2,\n bias=bias,\n )\n ]\n noise_upsample += [\n getattr(torch.nn, noise_upsample_activation)(\n **noise_upsample_activation_params\n )\n ]\n in_chs = channels\n self.noise_upsample = torch.nn.Sequential(*noise_upsample)\n self.noise_upsample_factor = np.prod(noise_upsample_scales)\n\n self.blocks = torch.nn.ModuleList()\n aux_chs = aux_channels\n for upsample_scale in upsample_scales:\n self.blocks += [\n TADEResBlock(\n in_channels=channels,\n aux_channels=aux_chs,\n kernel_size=kernel_size,\n dilation=dilation,\n bias=bias,\n upsample_factor=upsample_scale,\n upsample_mode=upsample_mode,\n gated_function=gated_function,\n ),\n ]\n aux_chs = channels\n self.upsample_factor = np.prod(upsample_scales)\n\n self.output_conv = torch.nn.Sequential(\n torch.nn.Conv1d(\n channels,\n out_channels,\n kernel_size,\n 1,\n bias=bias,\n padding=(kernel_size - 1) // 2,\n ),\n torch.nn.Tanh(),\n )\n\n # apply weight norm\n if use_weight_norm:\n self.apply_weight_norm()\n\n # reset parameters\n self.reset_parameters()","def mobilenetv2 (inputs, k, alpha = 1.0, train_bn = False):\n\n x = conv_block(inputs, 32, alpha, (3, 3), strides=(2, 2), block_id=0, train_bn=train_bn) # Input Res: 1\n\n x = inverted_residual_block(x, 16, (3, 3), t=1, strides=1, n=1, alpha=1.0, block_id=1, train_bn=train_bn)\t# Input Res: 1/2\n x = inverted_residual_block(x, 24, (3, 3), t=6, strides=2, n=2, alpha=1.0, block_id=2, train_bn=train_bn)\t# Input Res: 1/2\n x = inverted_residual_block(x, 32, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=4, train_bn=train_bn)\t# Input Res: 1/4\n x = inverted_residual_block(x, 64, (3, 3), t=6, strides=2, n=4, alpha=1.0, block_id=7, train_bn=train_bn)\t# Input Res: 1/8\n x = inverted_residual_block(x, 96, (3, 3), t=6, strides=1, n=3, alpha=1.0, block_id=11, train_bn=train_bn)\t# Input Res: 1/8\n x = inverted_residual_block(x, 160, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=14, train_bn=train_bn)\t# Input Res: 1/16\n x = inverted_residual_block(x, 320, (3, 3), t=6, strides=1, n=1, alpha=1.0, block_id=17, train_bn=train_bn)\t# Input Res: 1/32\n\n x = conv_block(x, 1280, alpha, (1, 1), strides=(1, 1), block_id=18, train_bn=train_bn) # Input Res: 1/32\n\n x = KL.GlobalAveragePooling2D()(x)\n x = KL.Reshape((1, 1, 1280))(x)\n x = KL.Dropout(0.3, name='Dropout')(x)\n x = KL.Conv2D(k, (1, 1), padding='same')(x)\n\n x = KL.Activation('softmax', name='softmax')(x)\n output = KL.Reshape((k,))(x)\n\n model = KM.Model(inputs, output)\n plot_model(model, to_file='MobileNetv2.png', show_shapes=True)\n\n return model","def __init__(self, cfg):\n super(MHBCoAtt, self).__init__()\n self.cfg = cfg\n # word embedding: q_vocab_size, 1024\n self.word_embedding = nn.Embedding(cfg.q_vocab_size, cfg.emb_dim)\n # LSTM\n if cfg.glove:\n self.lstm = nn.LSTM(input_size=cfg.emb_dim*2,\n hidden_size=cfg.hidden_dim,\n num_layers=cfg.num_layers,\n batch_first=True)\n else:\n self.lstm = nn.LSTM(input_size=cfg.emb_dim,\n hidden_size=cfg.hidden_dim,\n num_layers=cfg.num_layers,\n batch_first=True)\n\n self.dropout_l = nn.Dropout(p = 0.3)\n # question attention\n self.ques_att_conv1 = nn.Conv2d(cfg.hidden_dim, 512, [1,1])\n self.ques_att_conv2 = nn.Conv2d(512, 2, [1,1])\n\n # question attentive feature fuse with image feature, according to paper: k * o = 5000, k = 5\n self.ques_proj1 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.img_conv1d = nn.Conv2d(cfg.img_feature_channel, 5000, [1, 1])\n self.dropout_m = nn.Dropout(p = 0.1)\n\n # co-attention conv layers\n self.co_att_conv1 = nn.Conv2d(1000, 512, [1,1])\n self.co_att_conv2 = nn.Conv2d(512, 2, [1,1])\n\n # co_attentive feature fuse with question attentive feature\n self.ques_proj2 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.ques_proj3 = nn.Linear(2*cfg.hidden_dim, 5000)\n self.img_proj2 = nn.Linear(2*cfg.img_feature_channel, 5000)\n self.img_proj3 = nn.Linear(2*cfg.img_feature_channel, 5000)\n\n # prediction fully connected layer\n self.linear_pred = nn.Linear(2000, cfg.a_vocab_size)","def __call__(self, **kwargs):\n segname = 'block_{}_expand_relu'\n blocks = [13, 6, 3, 1]\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\n backbone_out = self._backbone.get_layer('block_16_project')\n\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\n p4 = self._fpn_block(p5, skips[1].output)\n p3 = self._fpn_block(p4, skips[2].output)\n p2 = self._fpn_block(p3, skips[3].output)\n\n s5 = self._conv_block(p5, 128)\n s4 = self._conv_block(p4, 128)\n s3 = self._conv_block(p3, 128)\n s2 = self._conv_block(p2, 128)\n\n s5 = tf.keras.layers.UpSampling2D(\n size=(8, 8),\n interpolation='nearest'\n )(s5)\n\n s4 = tf.keras.layers.UpSampling2D(\n size=(4, 4),\n interpolation='nearest'\n )(s4)\n\n s3 = tf.keras.layers.UpSampling2D(\n size=(2, 2),\n interpolation='nearest'\n )(s3)\n\n concat = [s5, s4, s3, s2]\n x = tf.keras.layers.Concatenate()(concat)\n x = tf.keras.layers.Conv2D(\n 64,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n x = tf.keras.layers.BatchNormalization()(x)\n x = tf.keras.layers.Activation('relu')(x)\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\n\n x = tf.keras.layers.Conv2D(\n 1,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n out = tf.keras.layers.Activation('sigmoid')(x)\n model = tf.keras.models.Model(\n inputs=self._backbone.input,\n outputs=out\n )\n\n return model","def __init__(self, bert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.q_lin.weight, bert_layer.attention.k_lin.weight, bert_layer.attention.v_lin.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.q_lin.bias, bert_layer.attention.k_lin.bias, bert_layer.attention.v_lin.bias]))\n self.out_proj_weight = bert_layer.attention.out_lin.weight\n self.out_proj_bias = bert_layer.attention.out_lin.bias\n self.linear1_weight = bert_layer.ffn.lin1.weight\n self.linear1_bias = bert_layer.ffn.lin1.bias\n self.linear2_weight = bert_layer.ffn.lin2.weight\n self.linear2_bias = bert_layer.ffn.lin2.bias\n self.norm1_eps = bert_layer.sa_layer_norm.eps\n self.norm1_weight = bert_layer.sa_layer_norm.weight\n self.norm1_bias = bert_layer.sa_layer_norm.bias\n self.norm2_eps = bert_layer.output_layer_norm.eps\n self.norm2_weight = bert_layer.output_layer_norm.weight\n self.norm2_bias = bert_layer.output_layer_norm.bias\n self.num_heads = bert_layer.attention.n_heads\n self.embed_dim = bert_layer.attention.dim\n self.is_last_layer = False\n self.validate_bettertransformer()","def __init__(self, conv_features_sizes, linear_layer_sizes, connector_shape):\n super().__init__()\n \n self.conv = nn.Sequential()\n self.mlp = nn.Sequential()\n self.flat = nn.Flatten()\n\n self.conv.add_module(name=f\"e-fconv{0}\", module=_conv2d_block(1, conv_features_sizes[0], kernel_size=3, padding=1))\n self.conv.add_module(name=f\"e-max{0}\", module=nn.MaxPool2d(2, 2))\n for i, (in_size, out_size) in enumerate(zip(conv_features_sizes[:-1], conv_features_sizes[1:]), 1):\n self.conv.add_module(name=f\"e-fconv{i}\", module=_conv2d_block(in_size, out_size, kernel_size=3, padding=1))\n self.conv.add_module(name=f\"e-max{i}\", module=nn.MaxPool2d(2, 2))\n\n mlp_input_shape = int(reduce((lambda x,y: x * y), connector_shape))\n self.mlp.add_module(name=f\"e-linear{0}\", module=nn.Linear(mlp_input_shape, linear_layer_sizes[0]))\n self.mlp.add_module(name=f\"e-batchnorm{0}\", module=nn.BatchNorm1d(linear_layer_sizes[0]))\n self.mlp.add_module(name=f\"e-relu{0}\", module=nn.ReLU())\n for i, (in_size, out_size) in enumerate(zip(linear_layer_sizes[:-1], linear_layer_sizes[1:]), 1):\n self.mlp.add_module(name=f\"e-linear{i}\", module=nn.Linear(in_size, out_size))\n self.mlp.add_module(name=f\"e-batchnorm{i}\", module=nn.BatchNorm1d(out_size))\n self.mlp.add_module(name=f\"e-relu{i}\", module=nn.ReLU())","def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\n ptm=\"bert\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\n super(PTMClassifier_PNN, self).__init__()\n \n if old_cols is None:\n old_cols = []\n \n self.old_cols = []\n \n self.output_size = output_size\n self.hidden_size = hidden_size # default\n self.feature_size = feature_size\n \n self.fc1 = nn.Linear(self.hidden_size, 100)\n self.classifier = nn.Linear(100, self.output_size)\n \n if len(old_cols) > 0:\n self.old_fc1s = ListModule()\n self.base_1 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n self.base_2 = nn.Sequential(\n nn.Linear(100 * len(old_cols), 100),\n nn.ReLU(),\n )\n \n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_1)\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\n self.base_2)\n \n for old_col in old_cols:\n self.old_fc1s.append(\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\n \n # ptm\n self.ptm = ptm.lower()\n assert self.ptm in supported_ptm\n ptm_ = import_from(\"transformers\", supported_ptm[self.ptm][0] + \"Model\")\n \n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\n if tokenizer is not None:\n self.encoder.resize_token_embeddings(len(tokenizer))\n self.prob_l = prob_l\n \n self.require_proto = require_proto\n \n # todo : modify net into one network\n if self.require_proto:\n # prototype-based classification\n self.net = nn.Sequential(\n self.fc1\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.fc1,\n self.encoder_adaptor,\n nn.ReLU(),\n )\n else:\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\n self.net = nn.Sequential(\n self.classifier\n )\n self.net_ = nn.Sequential(\n self.encoder,\n self.classifier\n )\n \n self.reset_parameters()","def __init__(self):\n super(Decoder_1m, self).__init__()\n self.lconvtwos = nn.ModuleList(\n [\n nn.Sequential(\n nn.Dropout(p=0.1),\n nn.Conv2d(128, 32, kernel_size=(3, 3), padding=1),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=1),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n ),\n ]\n )\n\n self.convtwos = nn.ModuleList(\n [\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=1),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=1),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n nn.Sequential(\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(32),\n nn.ReLU(inplace=True),\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\n nn.BatchNorm2d(64),\n nn.ReLU(inplace=True),\n ),\n ]\n )\n self.final = nn.Sequential(\n nn.Conv2d(64, 5, kernel_size=(1, 1), padding=0),\n nn.BatchNorm2d(5),\n nn.ReLU(inplace=True),\n nn.Conv2d(5, 1, kernel_size=(1, 1), padding=0),\n )","def transformer_model(input_tensor,\n attention_mask=None,\n hidden_size=768,\n num_hidden_layers=12,\n num_attention_heads=12,\n intermediate_size=3072,\n intermediate_act_fn=gelu,\n hidden_dropout_prob=0.1,\n attention_probs_dropout_prob=0.1,\n initializer_range=0.02,\n do_return_all_layers=False,\n conv_kernel_size=3,\n head_ratio=2,\n conv_type=\"noconv\",\n **kargs):\n if hidden_size % num_attention_heads != 0:\n raise ValueError(\n \"The hidden size (%d) is not a multiple of the number of attention \"\n \"heads (%d)\" % (hidden_size, num_attention_heads))\n\n attention_head_size = int(hidden_size / num_attention_heads)\n input_shape = bert_utils.get_shape_list(input_tensor, expected_rank=3)\n batch_size = input_shape[0]\n seq_length = input_shape[1]\n input_width = input_shape[2]\n\n # The Transformer performs sum residuals on all layers so the input needs\n # to be the same as the hidden size.\n if input_width != hidden_size:\n raise ValueError(\"The width of the input tensor (%d) != hidden size (%d)\" %\n (input_width, hidden_size))\n\n # We keep the representation as a 2D tensor to avoid re-shaping it back and\n # forth from a 3D tensor to a 2D tensor. Re-shapes are normally free on\n # the GPU/CPU but may not be free on the TPU, so we want to minimize them to\n # help the optimizer.\n prev_output = bert_utils.reshape_to_matrix(input_tensor)\n\n attn_maps = []\n all_layer_outputs = []\n all_value_outputs = []\n for layer_idx in range(num_hidden_layers):\n with tf.variable_scope(\"layer_%d\" % layer_idx):\n with tf.variable_scope(\"attention\"):\n attention_heads = []\n with tf.variable_scope(\"self\"):\n attention_head, probs, value_layer = attention_layer(\n from_tensor=prev_output,\n to_tensor=prev_output,\n attention_mask=attention_mask,\n num_attention_heads=num_attention_heads,\n size_per_head=attention_head_size,\n attention_probs_dropout_prob=attention_probs_dropout_prob,\n initializer_range=initializer_range,\n do_return_2d_tensor=True,\n batch_size=batch_size,\n from_seq_length=seq_length,\n to_seq_length=seq_length,\n conv_kernel_size=conv_kernel_size,\n head_ratio=head_ratio,\n conv_type=conv_type,\n from_tensor_mask=kargs.get('from_tensor_mask', None),\n to_tensor_mask=kargs.get('to_tensor_mask', None),\n conv_method=kargs.get('conv_method', \"dot\"))\n attention_heads.append(attention_head)\n attn_maps.append(probs)\n all_value_outputs.append(value_layer)\n\n attention_output = None\n if len(attention_heads) == 1:\n attention_output = attention_heads[0]\n else:\n # In the case where we have other sequences, we just concatenate\n # them to the self-attention head before the projection.\n attention_output = tf.concat(attention_heads, axis=-1)\n\n with tf.variable_scope(\"output\"):\n # Run a linear projection of `hidden_size` then add a residual\n # with `layer_input`.\n attention_output = tf.layers.dense(\n attention_output,\n hidden_size,\n kernel_initializer=create_initializer(initializer_range))\n \n\n attention_output = dropout(attention_output, hidden_dropout_prob)\n attention_output = layer_norm(attention_output + prev_output)\n\n # The activation is only applied to the \"intermediate\" hidden layer.\n with tf.variable_scope(\"intermediate\"):\n intermediate_output = tf.layers.dense(\n attention_output,\n intermediate_size,\n activation=intermediate_act_fn,\n kernel_initializer=create_initializer(initializer_range))\n # Down-project back to `hidden_size` then add the residual.\n with tf.variable_scope(\"output\"):\n prev_output = tf.layers.dense(\n intermediate_output,\n hidden_size,\n kernel_initializer=create_initializer(initializer_range))\n\n prev_output = dropout(prev_output, hidden_dropout_prob)\n prev_output = layer_norm(prev_output + attention_output)\n all_layer_outputs.append(prev_output)\n\n attn_maps = tf.stack(attn_maps, 0)\n if do_return_all_layers:\n final_outputs = []\n for layer_output in all_layer_outputs:\n final_output = bert_utils.reshape_from_matrix(layer_output, input_shape)\n final_outputs.append(final_output)\n return final_outputs, attn_maps, all_value_outputs\n else:\n final_output = bert_utils.reshape_from_matrix(prev_output, input_shape)\n return final_output, attn_maps, all_value_outputs","def Attention_block(up_in,down_in,nf):\n# def MyTile(alpha, channels):\n# alpha = tf.tile(alpha, [1, 1, 1, 1, channels])\n# return alpha\n# def MyTile_output_shape(input_shape):\n# shape = list(input_shape)\n# assert len(shape) == 5 # only valid for 2D tensors\n# shape[-1] *= channels\n# return tuple(shape)\n ndims = len(up_in.get_shape()) - 2\n assert ndims in [1, 2, 3], \"ndims should be one of 1, 2, or 3. found: %d\" % ndims\n input_channels = up_in.get_shape().as_list()[-1]\n# batch_size1 = tf.shape(down_in)[0]\n# nf = tf.min(batch_size0,batch_size1)\n Conv = getattr(KL, 'Conv%dD' % ndims)\n up = Conv(nf, kernel_size=1, padding='same',\n kernel_initializer='he_normal', strides=2)(up_in)\n down = Conv(nf, kernel_size=1, padding='same',\n kernel_initializer='he_normal', strides=1)(down_in)\n\n# x = NCCLayer(channels=nf)([up,down])\n x = Add()([up,down])\n x = Activation('relu')(x)\n x = Conv(1, kernel_size=1, padding='same',\n kernel_initializer='he_normal', use_bias = True, bias_initializer='zeros',strides=1,activation='sigmoid')(x)\n# x = Activation('sigmoid')(x)\n upsample_layer = getattr(KL, 'UpSampling%dD' % ndims)\n alpha = upsample_layer()(x)\n# alpha = Lambda(MyTile)(alpha, input_channels)\n alpha = MyTile(channels=input_channels)(alpha)\n up_out = Multiply()([alpha,up_in])\n return up_out","def mtf_bitransformer_all_layers_tiny():\n hparams = mtf_bitransformer_tiny()\n hparams.moe_num_experts = 4\n hparams.moe_expert_x = 4\n hparams.moe_expert_y = 4\n hparams.moe_hidden_size = 512\n hparams.encoder_layers = [\n \"self_att\", \"local_self_att\", \"moe_1d\", \"moe_2d\", \"drd\"]\n hparams.decoder_layers = [\n \"self_att\", \"local_self_att\", \"enc_att\", \"moe_1d\", \"moe_2d\", \"drd\"]\n return hparams","def __init__(self,\n size: int = 0,\n bb_size: int = 64,\n ff_size: int = 0,\n num_heads: int = 0,\n dropout: float = 0.1):\n super(CustomTransformerDecoderLayer, self).__init__()\n self.size = size\n\n self.trg_trg_att = ContMultiHeadedAttention(num_heads, bb_size, bb_size,\n dropout=dropout)\n self.src_trg_att = ContMultiHeadedAttention(num_heads, size, size,\n dropout=dropout)\n\n self.feed_forward_h1 = PositionwiseFeedForward(bb_size, ff_size=ff_size)\n self.feed_forward_h2 = PositionwiseFeedForward(size, ff_size=ff_size)\n\n self.x_layer_norm = nn.LayerNorm(size, eps=1e-6)\n self.spa_layer_norm = nn.LayerNorm(bb_size, eps=1e-6)\n\n self.dropout = nn.Dropout(dropout)","def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, \n mlm_label=None, dvae_imgs=None, v_token_mask=None, hog_features=None, img_metas=None, **kwargs): \n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \n if self.from_scratch:\n imgs = imgs / 255.0\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\n text_input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\n if mlm_label is not None:\n mlm_label = mlm_label.reshape((-1, ) + mlm_label.shape[2:])\n\n\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\n\n B, D, T, H, W = visual_token.shape\n losses = dict()\n # -------------- nce loss ------------------- #\n if hasattr(self, 'ssl_head'):\n input_ssl_ids = torch.where(mlm_label == -100, token_ids.clone(), mlm_label.clone())\n input_ssl_mask = text_input_mask.clone()\n text_only_out = self.text_backbone(input_ssl_ids, input_ssl_mask)\n # ------------ complete T -------------- #\n text_out_no_mask = text_only_out['last_hidden_state']\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_no_mask, input_ssl_mask, input_ssl_ids)\n\n\n # ------------ complete V ---------------- #\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\n\n\n # ---------- foward mask text input ---------- # \n text_out_with_mask = self.text_backbone(token_ids, text_input_mask)\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\n\n # ---------- forward mask v input ------------ #\n visual_token_with_mask, v_mask = self.extract_visual_feat(imgs.clone(), v_token_mask) # b, d, T, h, w\n visual_token_mask = visual_token_with_mask.view(B, D, T, -1).permute(0, 2, 3, 1)\n \n v_fusion_output = self.multimodal_backbone(visual_token=visual_token_mask, text_input_mask=text_input_mask.clone(), text_input_embeds=text_out_no_mask.clone())\n \n t_fusion_output = self.multimodal_backbone(visual_token=visual_token, text_input_mask=text_input_mask, text_input_embeds=text_out_last_hidden_state)\n # for mlm #\n t_last_hidden_state = t_fusion_output['t_last_hidden_state']\n\n\n\n\n\n # ------------ MLM loss ------------ #\n\n if mlm_label is not None and self.mlm_head is not None:\n # we use mask text for MLM\n # because we doubt there will be miss interaction between wrong img-text pair \n # and the model not learn good relationship between vision and language\n # -------- forward masked text ----------- #\n mlm_prediction_score = self.mlm_head(t_last_hidden_state)\n \n if self.mlm_loss_func is not None:\n mlm_label_idx = torch.where(mlm_label.view(-1) != -100)\n mlm_prediction_mask_score = mlm_prediction_score.view(-1, self.text_vocab_size)[mlm_label_idx[0], :]\n mlm_label_mask = mlm_label.view(-1)[mlm_label_idx]\n mlm_loss = self.mlm_loss_func(mlm_prediction_mask_score, mlm_label_mask)\n else:\n mlm_loss = self.loss_func(mlm_prediction_score.view(-1, self.text_vocab_size), mlm_label.view(-1))\n losses['mlm_loss'] = mlm_loss\n\n\n # ------- Tri-modal alignment with mask sample and ranking --------- #\n if self.mlm_ssl_V_head is not None:\n mlm_visual_feat = v_fusion_output['t_last_hidden_state'][:, 0]\n mask_visual_recon_emb = self.mlm_ssl_V_head(mlm_visual_feat)\n mask_word_emb = self.ssl_head.forward_text(text_out_last_hidden_state) if self.use_Cmask else None\n loss_cvt_rank = self.ssl_loss(visual_emb, text_emb, mask_word_emb, mask_visual_recon_emb)\n losses.update(loss_cvt_rank)\n\n\n if self.symmetry_rank:\n mlm_word_feat = t_last_hidden_state[:, 0]\n mask_word_recon_emb = self.mlm_ssl_T_head(mlm_word_feat)\n\n mask_visual_emb = self.ssl_head.forward_vision(visual_token_with_mask) if self.use_Cmask else None\n \n loss_ctv_rank = self.ssl_loss(text_emb, visual_emb, mask_visual_emb, mask_word_recon_emb)\n loss_ctv_rank['v_nce_loss'] = loss_ctv_rank.pop('nce_loss')\n \n if self.ssl_loss.use_rank:\n loss_ctv_rank['rank_v_vm_loss'] = loss_ctv_rank.pop('rank_t_tm_loss')\n\n \n\n losses.update(loss_ctv_rank)\n\n\n\n return losses","def identity_block(self, input_tensor, kernel_size, filters, stage, block):\r\n\t\tfilters1, filters2, filters3 = filters # this indicates the number of filter (here, assume that all filter number are same)\r\n\r\n\t\tbn_axis = self.channel_last # number of classes being classified\r\n \r\n\t\t# names for the layers\r\n\t\tconv_name_base = 'res' + str(stage) + block + '_branch'\r\n\t\tbn_name_base = 'bn' + str(stage) + block + '_branch'\r\n\r\n\t\t# model definition\r\n\t\tx = Conv2D(filters1, (1, 1),\r\n\t\t\t\t\t\t kernel_initializer='he_normal',\r\n\t\t\t\t\t\t name=conv_name_base + '2a')(input_tensor)\r\n\t\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')(x)\r\n\t\tx = Activation('relu')(x)\r\n\r\n\t\tx = Conv2D(filters2, kernel_size,\r\n\t\t\t\t\t\t padding='same',\r\n\t\t\t\t\t\t kernel_initializer='he_normal',\r\n\t\t\t\t\t\t name=conv_name_base + '2b')(x)\r\n \r\n\t\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')(x)\r\n\t\tx = Activation('relu')(x)\r\n\r\n\t\tx = Conv2D(filters3, (1, 1),\r\n\t\t\t\t\t\t kernel_initializer='he_normal',\r\n\t\t\t\t\t\t name=conv_name_base + '2c')(x)\r\n\t\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')(x)\r\n\r\n\t\t# Here, we add the outcome tensor(layer) x to original input tensor\r\n\t\tx = add([x, input_tensor])\r\n\t\tx = Activation('relu')(x)\r\n \r\n\t\treturn x","def forward(self, src, src_t, seq, seq_t, seq_e, mask, vars_dict):\n\n src_ext = torch.unsqueeze(src, dim=1) # src [B, 1, D]\n src_e_ph = torch.zeros_like(src_ext)\n q = torch.cat([src_ext, src_e_ph, src_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\n k = torch.cat([seq, seq_e, seq_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\n\n mask = torch.unsqueeze(mask, dim=2) # mask [B, N, 1]\n mask = mask.permute([0, 2, 1]) # mask [B, 1, N]\n\n # # target-attention\n output, attn = self.multi_head_target(q=q, k=k, v=k, vars_dict=vars_dict, mask=mask) # output: [B, 1, D + Dt], attn: [B, 1, N]\n # print('src.shape', src.shape)\n output = output.squeeze(1)\n # print('output.shape', output.shape)\n # print('output', output)\n # print('output.squeeze().shape', output.shape)\n attn = attn.squeeze()\n\n # output = self.merger(output, src)\n x = torch.cat([output, src], dim=1)\n # x = self.layer_norm(x)\n x = F.relu(F.linear(x, vars_dict['w1_agg_fc']))\n output = F.linear(x, vars_dict['w2_agg_fc'])\n\n return output, attn","def __init__(self):\n super(Generator_Model, self).__init__()\n\n # audio encoder\n self.audio_encoder = nn.Sequential(\n conv_block(1, 32, kernel_size=1, stride=1, padding=1),\n conv_block(32, 32, kernel_size=1, stride=1, padding=1, residual=True),\n \n conv_block(32, 64, kernel_size=3, stride=(3, 1), padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(64, 128, kernel_size=3, stride=3, padding=1),\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(128, 256, kernel_size=3, stride=(3, 2), padding=1),\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True),\n\n conv_block(256, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 512, kernel_size=1, stride=1, padding=0)\n )\n\n # face encoder\n self.face_encoder = nn.ModuleList([\n nn.Sequential(conv_block(6, 16, kernel_size=7, stride=1, padding=3)),\n\n nn.Sequential(conv_block(16, 32, kernel_size=3, stride=2, padding=1),\n conv_block(32, 32, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(32, 64, kernel_size=3, stride=2, padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(64, 128, kernel_size=3, stride=2, padding=1),\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(128, 256, kernel_size=3, stride=2, padding=1),\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_block(256, 512, kernel_size=3, stride=2, padding=1),\n conv_block(512, 512, kernel_size=3, stride=1, padding=1, residual=True)),\n \n nn.Sequential(conv_block(512, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 512, kernel_size=1, stride=1, padding=0))\n ])\n\n # face decoder\n self.face_decoder = nn.ModuleList([\n nn.Sequential(conv_block(512, 512, kernel_size=1, stride=1, padding=0),),\n\n nn.Sequential(conv_t_block(512, 512, kernel_size=3, stride=1, padding=0),\n conv_block(512, 1024, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(1024, 512, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(512, 768, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(768, 384, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(384, 512, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(512, 256, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(256, 320, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(320, 128, kernel_size=3, stride=2, padding=1, output_padding=1), \n conv_block(128, 160, kernel_size=3, stride=1, padding=1, residual=True)),\n\n nn.Sequential(conv_t_block(160, 64, kernel_size=3, stride=2, padding=1, output_padding=1),\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True))\n ])\n\n # output blocks for face prediction\n self.pred_output = nn.Sequential(conv_block(64, 32, kernel_size=3, stride=1, padding=1),\n conv_block(32, 3, kernel_size=1, stride=1, padding=0),\n nn.Sigmoid())\n\n # optimizer\n self.optimizer = torch.optim.Adam(self.parameters(), lr=0.001, betas=(0.5, 0.999))","def __init__(self, mbart_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\n self.linear1_weight = mbart_layer.fc1.weight\n self.linear1_bias = mbart_layer.fc1.bias\n self.linear2_weight = mbart_layer.fc2.weight\n self.linear2_bias = mbart_layer.fc2.bias\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\n self.norm2_eps = mbart_layer.final_layer_norm.eps\n self.norm2_weight = mbart_layer.final_layer_norm.weight\n self.norm2_bias = mbart_layer.final_layer_norm.bias\n self.num_heads = mbart_layer.self_attn.num_heads\n self.embed_dim = mbart_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def _init_layers(self) -> None:\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\n self.embed_dims = self.self_attn.embed_dims\n self.ffn = FFN(**self.ffn_cfg)\n norms_list = [\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\n for _ in range(3)\n ]\n self.norms = ModuleList(norms_list)","def __init__(self, bert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.self.query.weight, bert_layer.attention.self.key.weight, bert_layer.attention.self.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.self.query.bias, bert_layer.attention.self.key.bias, bert_layer.attention.self.value.bias]))\n self.out_proj_weight = bert_layer.attention.output.dense.weight\n self.out_proj_bias = bert_layer.attention.output.dense.bias\n self.linear1_weight = bert_layer.intermediate.dense.weight\n self.linear1_bias = bert_layer.intermediate.dense.bias\n self.linear2_weight = bert_layer.output.dense.weight\n self.linear2_bias = bert_layer.output.dense.bias\n self.norm1_eps = bert_layer.attention.output.LayerNorm.eps\n self.norm1_weight = bert_layer.attention.output.LayerNorm.weight\n self.norm1_bias = bert_layer.attention.output.LayerNorm.bias\n self.norm2_eps = bert_layer.output.LayerNorm.eps\n self.norm2_weight = bert_layer.output.LayerNorm.weight\n self.norm2_bias = bert_layer.output.LayerNorm.bias\n self.num_heads = bert_layer.attention.self.num_attention_heads\n self.embed_dim = bert_layer.attention.self.all_head_size\n self.is_last_layer = False\n self.validate_bettertransformer()","def mt_step(self, lang1, lang2, lambda_coeff):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n lang1_id = params.lang2id[lang1]\n lang2_id = params.lang2id[lang2]\n\n # generate batch\n if lang1 == lang2:\n (x1, len1) = self.get_batch('ae', lang1)\n (x2, len2) = (x1, len1)\n (x1, len1) = self.add_noise(x1, len1)\n else:\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\n langs1 = x1.clone().fill_(lang1_id)\n langs2 = x2.clone().fill_(lang2_id)\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n # key point generate label\n\n # cuda\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\n\n # encode source sentence\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\n enc1 = enc1.transpose(0, 1)\n\n # decode target sentence\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\n src_len=len1)\n\n # loss\n # the last word not apply prediction logic\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def __init__(\n self,\n num_branches,\n blocks,\n num_blocks,\n num_inchannels,\n num_channels,\n num_heads,\n num_window_sizes,\n num_mlp_ratios,\n multi_scale_output=True,\n drop_path=0.0,\n ):\n super(HighResolutionTransformerModule, self).__init__()\n self._check_branches(\n num_branches, blocks, num_blocks, num_inchannels, num_channels\n )\n\n self.num_inchannels = num_inchannels\n self.num_branches = num_branches\n\n self.multi_scale_output = multi_scale_output\n self.branches = self._make_branches(\n num_branches,\n blocks,\n num_blocks,\n num_channels,\n num_heads,\n num_window_sizes,\n num_mlp_ratios,\n drop_path,\n )\n self.fuse_layers = self._make_fuse_layers()\n self.relu = nn.ReLU(inplace=True)\n\n self.num_heads = num_heads\n self.num_window_sizes = num_window_sizes\n self.num_mlp_ratios = num_mlp_ratios","def build_bisenet(inputs, num_classes):\n\n ### The spatial path\n ### The number of feature maps for each convolution is not specified in the paper\n ### It was chosen here to be equal to the number of feature maps of a classification\n ### model at each corresponding stage\n # spatial_net = fluid.layers.resize_bilinear(inputs, [Image_Height/8, Image_Width/8])\n # print('spatial_net_1',spatial_net)\n\n ## spatial path\n spatial_net = ConvBlock(inputs, num_filters=64, kernel_size=3, stride=2)\n spatial_net = ConvBlock(spatial_net, num_filters=128, kernel_size=3, stride=2)\n spatial_net = ConvBlock(spatial_net, num_filters=256, kernel_size=3, stride=2)\n # print(\"spatial_net:\", spatial_net)\n\n # spatial_net = fluid.layers.resize_bilinear(spatial_net, [Image_Height/8, Image_Width/8])\n # print('spatial_net_2',spatial_net)\n ### Context path\n model = ResNet(is_test=False)\n # spatial_net = model.bottleneck_block1(inputs)\n end_points_16, end_points_32 = model.net(inputs)\n net_4 = AttentionRefinementModule(end_points_16, num_filters=512)\n net_5 = AttentionRefinementModule(end_points_32, num_filters=1024)\n global_channels = fluid.layers.reduce_mean(net_5, [2, 3], keep_dim=True)\n net_5_scaled = fluid.layers.elementwise_mul(net_5, global_channels, axis=0)\n\n ### Combining the paths\n net_4 = Upsample(net_4, scale=2)\n net_5_scaled = Upsample(net_5_scaled, scale=4)\n # print('net_4, net_5:', [net_4, net_5_scaled])\n # layers_concat = list()\n # layers_concat.append(spatial_net)\n ## layers_concat.append(net_4)\n # layers_concat.append(net_5_scaled)\n context_net = fluid.layers.concat([spatial_net, net_4, net_5_scaled], axis=1) #\n # context_net = fluid.layers.concat(input=layers_concat,axis=1)\n # print('context_net', context_net)\n # context_net = fluid.layers.concat([net_4, net_5_scaled], axis=1)\n # print('context_net', context_net)\n # context_net = fluid.layers.concat([spatial_net,context_net], axis=1)\n # print('context_net2',context_net)\n\n ### FFM\n # net = FeatureFusionModule(input_1=spatial_net, input_2=context_net, num_filters=num_classes)\n net = FeatureFusionModule(inputs=context_net, num_filters=num_classes)\n\n # print('net', net)\n\n ## [batch_zize, num_filters, 128, 64]\n\n ### Final upscaling and finish\n # net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[256, 128])\n # print('conv2d_transpose', net)\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn1')\n net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[128, 256])\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn2')\n net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[256, 512])\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn3')\n #net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[512, 1024])\n #net = batch_normalization(net, relu=True, name='conv2d_transpose_bn4')\n # print('net',net)\n net = fluid.layers.image_resize(net, out_shape=[512, 1024], resample='BILINEAR')\n\n net = fluid.layers.conv2d(net, num_classes, 1)\n return net","def __init__(\n self,\n fc_layers: Tuple[int] = (1000,),\n inverted_residual_setting: Optional[List[List[int]]] = None,\n block: Optional[Callable[..., nn.Module]] = None,\n norm_layer: Optional[Callable[..., nn.Module]] = None,\n c = 3,\n h = 218,\n w = 178\n ) -> None:\n super(MobileNetV2, self).__init__()\n if block is None:\n block = InvertedResidual\n\n if norm_layer is None:\n norm_layer = nn.BatchNorm2d\n\n input_channel = c\n new_h, new_w = h, w\n\n if inverted_residual_setting is None:\n inverted_residual_setting = [\n # t, c, n, s\n [1, 16, 1, 1],\n [6, 24, 2, 2],\n [6, 32, 3, 2],\n [6, 64, 4, 2],\n [6, 96, 3, 1],\n [6, 160, 3, 2],\n [6, 320, 1, 1],\n ]\n\n # only check the first element, assuming user knows t,c,n,s are required\n if len(inverted_residual_setting) == 0 or len(inverted_residual_setting[0]) != 4:\n raise ValueError(\"inverted_residual_setting should be non-empty \"\n \"or a 4-element list, got {}\".format(inverted_residual_setting))\n\n # building first layer\n features: List[nn.Module] = []\n # building inverted residual blocks\n for t, c, n, s in inverted_residual_setting:\n output_channel = c\n new_h //= s\n new_w //= s\n for i in range(n):\n stride = s if i == 0 else 1\n features.append(block(input_channel, output_channel, stride, expand_ratio=t, norm_layer=norm_layer))\n input_channel = output_channel\n # make it nn.Sequential\n self.features = nn.Sequential(*features)\n\n # building classifier\n fc_layers = (new_h*new_w*output_channel,) + fc_layers\n self.classifier = MLP(fc_layers, end_fn=nn.Identity)\n\n # weight initialization\n init_params(self.modules())","def __init__(self, **kwargs):\n FeatureDefinition.__init__(self)\n\n nbTypes = self._getTypeNumber(kwargs)\n \n print(\"BETTER FEATURES\")\n \n \n block_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\n (\"xywh\", Pipeline([\n ('selector', NodeTransformerXYWH_v2()),\n #v1 ('xywh', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('xywh', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n , (\"neighbors\", Pipeline([\n ('selector', NodeTransformerNeighbors()),\n #v1 ('neighbors', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('neighbors', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n , (\"1hot\", Pipeline([\n ('1hot', Node1HotFeatures()) #does the 1-hot encoding directly\n ])\n )\n ])\n grid_line_transformer = GridLine_NodeTransformer_v2()\n \n self._node_transformer = TransformerListByType([block_transformer, grid_line_transformer]) \n \n edge_BB_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\n (\"1hot\", Pipeline([\n ('1hot', Edge1HotFeatures(PageNumberSimpleSequenciality()))\n ])\n )\n , (\"boolean\", Pipeline([\n ('boolean', EdgeBooleanFeatures_v2())\n ])\n )\n , (\"numerical\", Pipeline([\n ('selector', EdgeNumericalSelector()),\n #v1 ('numerical', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\n ('numerical', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\n ])\n )\n ] )\n edge_BL_transformer = DU_ABPTableRG.Block2GridLine_EdgeTransformer()\n edge_LL_transformer = DU_ABPTableRG.GridLine2GridLine_EdgeTransformer()\n self._edge_transformer = TransformerListByType([edge_BB_transformer,\n edge_BL_transformer,\n edge_BL_transformer, # useless but required\n edge_LL_transformer \n ])\n \n self.tfidfNodeTextVectorizer = None #tdifNodeTextVectorizer","def _InitAttentionParams(self, atten_tpl):\n p = self.params\n source_atten_tpls = []\n # Set up each source attention.\n for i in range(p.num_source):\n src_key = 'source_%d' % i\n src_atten = atten_tpl.Copy()\n src_atten = super()._InitAttentionParams(src_atten)\n if isinstance(src_atten, list):\n raise ValueError(\n 'TransformerMultiSourceAttentionLayer does not support '\n 'num_heads > 1.')\n src_atten.name = 'multihead_atten_%s' % src_key\n source_atten_tpls.append((src_key, src_atten))\n\n # Initialize multi-source attention.\n msa = p.multi_source_atten.Copy()\n msa.name = 'multi_source_atten'\n msa.input_dim = p.input_dim\n msa.hidden_dim = p.hidden_dim\n msa.source_atten_tpls = source_atten_tpls\n msa.primary_source_key = 'source_%d' % p.primary_source_index\n return msa","def __init__(self, block, layers, groups, reduction, dropout_p=0.2, in_channels=3,\n inplanes=128, input_3x3=True, downsample_kernel_size=3,\n downsample_padding=1, num_classes=1000):\n super(SENet, self).__init__()\n \n self.in_channels = in_channels\n self.inplanes = inplanes\n\n if input_3x3:\n layer0_modules = [\n ('conv1', nn.Conv2d(in_channels, 64, 3, stride=2, padding=1,\n bias=False)),\n ('bn1', nn.BatchNorm2d(64)),\n ('relu1', nn.ReLU(inplace=True)),\n ('conv2', nn.Conv2d(64, 64, 3, stride=1, padding=1,\n bias=False)),\n ('bn2', nn.BatchNorm2d(64)),\n ('relu2', nn.ReLU(inplace=True)),\n ('conv3', nn.Conv2d(64, inplanes, 3, stride=1, padding=1,\n bias=False)),\n ('bn3', nn.BatchNorm2d(inplanes)),\n ('relu3', nn.ReLU(inplace=True)),\n ]\n else:\n layer0_modules = [\n ('conv1', nn.Conv2d(in_channels, inplanes, kernel_size=7, stride=2,\n padding=3, bias=False)),\n ('bn1', nn.BatchNorm2d(inplanes)),\n ('relu1', nn.ReLU(inplace=True)),\n ]\n\n layer0_modules.append(('pool', nn.MaxPool2d(3, stride=2, ceil_mode=True)))\n \n self.layer0 = nn.Sequential(OrderedDict(layer0_modules))\n \n self.layer1 = self._make_layer(\n block,\n planes=64,\n blocks=layers[0],\n groups=groups,\n reduction=reduction,\n downsample_kernel_size=1,\n downsample_padding=0\n )\n \n self.layer2 = self._make_layer(\n block,\n planes=128,\n blocks=layers[1],\n stride=2,\n groups=groups,\n reduction=reduction,\n downsample_kernel_size=downsample_kernel_size,\n downsample_padding=downsample_padding\n )\n \n self.layer3 = self._make_layer(\n block,\n planes=256,\n blocks=layers[2],\n stride=2,\n groups=groups,\n reduction=reduction,\n downsample_kernel_size=downsample_kernel_size,\n downsample_padding=downsample_padding\n )\n \n self.layer4 = self._make_layer(\n block,\n planes=512,\n blocks=layers[3],\n stride=2,\n groups=groups,\n reduction=reduction,\n downsample_kernel_size=downsample_kernel_size,\n downsample_padding=downsample_padding\n )\n \n self.avg_pool = nn.AvgPool2d(7, stride=1)\n self.dropout = nn.Dropout2d(dropout_p) if dropout_p is not None else None\n self.last_linear = nn.Linear(512 * block.expansion, num_classes)","def __init__(self, nheads, d_model):\n super(MultiheadAttention, self).__init__()\n assert d_model % nheads == 0\n self.d_head = d_model // nheads\n self.nheads = nheads\n self.Q_fc = nn.Linear(d_model, d_model, bias=False)\n self.K_fc = nn.Linear(d_model, d_model, bias=False)\n self.V_fc = nn.Linear(d_model, d_model, bias=False)\n self.output_fc = nn.Linear(d_model, d_model, bias=False)\n self.attn = None","def block(inp, scope, past, config):\n with tf.variable_scope(scope):\n nx = inp.shape[-1].value\n norm = normalise_tensor(inp, 'ln_1')\n attn, present = attention(norm, 'attn', nx, past=past, config=config)\n out = attn + inp\n norm = normalise_tensor(out, 'ln_2')\n mlp_out = multilayer_perceptron(norm, 'mlp', nx * 4) # note that hidden dim is 4x\n out = out + mlp_out\n return out, present","def call(self, inputs, training = False):\n \"\"\" includes training argument as batch_norm functions differently during training and during inference \"\"\"\n \n # input --> block_1 --> block_2 --> block_3 \n x = self.block_1(inputs, training = training)\n x = self.block_2(x, training = training)\n x = self.block_3(x, training = training)\n\n # output for MAA2C or DDPG \n # block_3 --> outputs \n x = self.outputs(x)\n\n return x","def identity_block(input_tensor, kernel_size, filters, stage, block):\n filters0, filters1, filters2 = filters\n conv_name_base = 'res' + str(stage) + block\n bn_name_base = 'bn' + str(stage) + block\n add_name = 'add' + str(stage) + \"_\" + block\n relu_name = 'relu' + str(stage) + \"_\" + block\n\n # Tensors\n input_tensor_chans = input_tensor.dims(\n 3) if input_tensor.shape.layout == sg.NHWC else input_tensor.dims(1)\n conv0_tensor = sg.Tensor(\n data_layout=sg.NHWC, tensor_data=generate_random_data(\n (filters0, 1, 1, input_tensor_chans)))\n bn0_mean_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\n bn0_var_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\n bn0_gamma_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\n bn0_beta_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\n conv1_tensor = sg.Tensor(\n data_layout=sg.NHWC, tensor_data=generate_random_data(\n (filters1, kernel_size, kernel_size, filters0)))\n bn1_mean_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\n bn1_var_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\n bn1_gamma_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\n bn1_beta_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\n conv2_tensor = sg.Tensor(\n data_layout=sg.NHWC, tensor_data=generate_random_data(\n (filters2, 1, 1, filters1)))\n bn2_mean_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\n bn2_var_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\n bn2_gamma_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\n bn2_beta_tensor = sg.Tensor(\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\n\n x = sg.nn.convolution(\n input_tensor, conv0_tensor, stride=[1, 1], padding=\"same\",\n name=conv_name_base + '_2a')\n x = sg.nn.batch_norm(\n x, bn0_mean_tensor, bn0_var_tensor, bn0_gamma_tensor, bn0_beta_tensor,\n activation=\"relu\", name=bn_name_base + '_2a')\n x = sg.nn.convolution(\n x, conv1_tensor, stride=[1, 1], padding=\"same\",\n name=conv_name_base + '_2b')\n x = sg.nn.batch_norm(\n x, bn1_mean_tensor, bn1_var_tensor, bn1_gamma_tensor, bn1_beta_tensor,\n activation=\"relu\", name=bn_name_base + '_2b')\n x = sg.nn.convolution(\n x, conv2_tensor, stride=[1, 1], padding=\"same\",\n name=conv_name_base + '_2c')\n x = sg.nn.batch_norm(\n x, bn2_mean_tensor, bn2_var_tensor, bn2_gamma_tensor, bn2_beta_tensor,\n name=bn_name_base + '_2c')\n x = sg.math.add(x, input_tensor, name=add_name)\n x = sg.nn.relu(x, name=relu_name)\n return x","def __init__(self,\n input_dim,\n dec_seq_len,\n out_seq_len,\n d_model=512,\n nhead=8,\n num_encoder_layers=6,\n num_decoder_layers=6,\n dim_feedforward=2048,\n dropout=0.1,\n activation='relu',\n custom_encoder=None,\n custom_decoder=None):\n super(TransformerTS, self).__init__()\n self.transform = nn.Transformer(\n d_model=d_model,\n nhead=nhead,\n num_encoder_layers=num_encoder_layers,\n num_decoder_layers=num_decoder_layers,\n dim_feedforward=dim_feedforward,\n dropout=dropout,\n activation=activation,\n custom_encoder=custom_encoder,\n custom_decoder=custom_decoder\n )\n self.pos = PositionalEncoding(d_model)\n self.enc_input_fc = nn.Linear(input_dim, d_model)\n self.dec_input_fc = nn.Linear(input_dim, d_model)\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\n self.dec_seq_len = dec_seq_len","def __init__(self, emb_size: int = 768, num_heads: int = 12, dropout: float = 0.1):\n super().__init__()\n self.emb_size = emb_size\n self.num_heads = num_heads\n\n # Check \"head dim\" = number of features per head (d_k).\n self.head_dim = emb_size // num_heads\n assert self.head_dim * num_heads == self.emb_size, \"emb_size must be divisible by num_heads\"\n # Calculate scaling factor.\n self.scale = 1 / (self.head_dim ** 0.5)\n\n # V1: in vanilla self-attention Q,K,V are square matrices.\n # self.keys = nn.Linear(emb_size, emb_size)\n # self.queries = nn.Linear(emb_size, emb_size)\n # self.values = nn.Linear(emb_size, emb_size)\n \n # V2: single layer with emb_size, split into num (heads * head_dim) * 3 (Q,K,V).\n self.qkv = nn.Linear(emb_size, emb_size * 3)\n\n # Attention dropout.\n self.att_drop = nn.Dropout(dropout)\n self.projection = nn.Linear(emb_size, emb_size)","def _fp32_mte_process_lt_one_block(axis_1_lp_index, sub_axis_1):\n\n def _fp32_inner_last_dim_lt_one_block(axis_0_lp_index, sub_axis_0):\n \"\"\"\n inner process of last dim less than one block\n \"\"\"\n\n # move data in\n in_offset = (block_idx * per_core_col_size +\n axis_1_lp_index * max_core_axis_size +\n axis_0_lp_index * max_no_core_axis_size * axis_1) * axis_2\n data_pos_info = (sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, in_offset)\n _data_move_in_last_dim_lt_one_block(tik_inst, ub_input, data_in, data_pos_info)\n\n # do transpose\n with tik_inst.new_stmt_scope():\n temp_sub_axis_1 = tik_inst.Scalar(\"int64\")\n temp_sub_axis_0 = tik_inst.Scalar(\"int64\")\n data_size_one_block = _get_elment_cnt_one_block(data_in.dtype)\n axis_1_0_2_size = axis_0 * axis_1 * axis_2\n sub_axis_1_0_2_size = sub_axis_1 * sub_axis_0 * axis_2\n\n # to avoid multiple core dirty data\n with tik_inst.if_scope(tik.all(sub_axis_1_0_2_size < data_size_one_block,\n axis_1_0_2_size > data_size_one_block)):\n with tik_inst.if_scope(sub_axis_1 == 1):\n temp_sub_axis_0.set_as(_ceil_div(data_size_one_block, axis_2))\n temp_sub_axis_1.set_as(sub_axis_1)\n with tik_inst.else_scope():\n temp_sub_axis_0.set_as(sub_axis_0)\n temp_sub_axis_1.set_as(_ceil_div(data_size_one_block,\n axis_0 * axis_2))\n with tik_inst.else_scope():\n temp_sub_axis_1.set_as(sub_axis_1)\n temp_sub_axis_0.set_as(sub_axis_0)\n\n sub_dim_size = (temp_sub_axis_1, temp_sub_axis_0, axis_2)\n _transpose_by_2_vnchwconv_not_last_dim(tik_inst, ub_input[ub_offset],\n ub_input, sub_dim_size)\n\n # move data out\n out_offset = ((block_idx * per_core_col_size +\n axis_1_lp_index * max_core_axis_size) * axis_0 +\n axis_0_lp_index * max_no_core_axis_size) * axis_2\n data_pos_info = (sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, out_offset)\n _data_move_out_last_dim_lt_one_block(tik_inst, data_out, ub_input[ub_offset],\n data_pos_info)\n\n with tik_inst.for_range(0, no_core_loop_cnt) as axis_0_lp_idx:\n _fp32_inner_last_dim_lt_one_block(axis_0_lp_idx, max_no_core_axis_size)\n with tik_inst.if_scope(no_core_left > 0):\n _fp32_inner_last_dim_lt_one_block(no_core_loop_cnt, no_core_left)","def __init__(self, config: BertConfig):\r\n super().__init__(config)\r\n ### YOUR CODE HERE\r\n self.num_labels = config.num_labels # [0, 1] (start or end)\r\n self.bert = BertModel(config)\r\n self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels) # TODO: Not a separate FFN ? (For Start_FFN and End_FFN)\r\n\r\n ### END YOUR CODE\r\n\r\n # Don't forget initializing the weights\r\n self.init_weights()","def __init__(\n self,\n num_frames = 2,\n dim = 64,\n channels = 3,\n stages = 4,\n num_blocks = 2,\n dim_head = 64,\n window_size = 16,\n heads = 8,\n ff_mult = 4,\n input_channels = None,\n output_channels = None,\n device = None,\n skip_type = 'ConcatCross'\n ):\n super().__init__()\n input_channels = channels if input_channels is None else input_channels\n output_channels = channels if output_channels is None else output_channels\n\n self.project_in = nn.Sequential(\n nn.Conv3d(input_channels, dim, kernel_size = (1, 3, 3), padding = (0, 1, 1)),\n nn.GELU()\n )\n\n self.project_out = nn.Sequential(\n nn.Conv3d(dim, output_channels, kernel_size = (1, 3, 3), padding = (0, 1, 1))\n )\n\n self.downs = nn.ModuleList([])\n self.ups = nn.ModuleList([])\n\n # render each hyperparam a tuple so that we can iterate thru them\n heads, window_size, dim_head, num_blocks = map(partial(cast_tuple, depth = stages), (heads, window_size, dim_head, num_blocks))\n\n for ind, heads, window_size, dim_head, num_blocks in zip(range(stages), heads, window_size, dim_head, num_blocks):\n is_last = ind == (stages - 1)\n\n self.downs.append(nn.ModuleList([ # need 3d version of this\n Block(dim, depth = num_blocks, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size, device = device),\n nn.Conv3d(dim, dim * 2, kernel_size = (1, 4, 4), stride = (1, 2, 2), padding = (0, 1, 1))\n ]))\n\n self.ups.append(nn.ModuleList([\n nn.ConvTranspose3d(dim * 2, dim, (1, 2, 2), stride = (1, 2, 2)),\n Block(dim, depth = num_blocks, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size, device = device, skip_type = skip_type)\n ]))\n\n dim *= 2\n if is_last:\n self.mid = nn.Sequential(\n Rearrange('b c f h w -> b (c f) 1 h w'),\n Block(dim = dim*num_frames, depth = 1, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size),\n nn.Conv3d(dim*num_frames, dim, kernel_size = (1, 3, 3), padding = (0, 1, 1)),\n Block(dim = dim, depth = 2, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size),\n )","def __init__(self,\n num_heads,\n seq_len,\n block=16,\n different_layout_per_head=False,\n num_random_blocks=0,\n local_window_blocks=[4],\n global_block_indices=[0],\n global_block_end_indices=None,\n attention='bidirectional',\n horizental_global_attention=False):\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\n self.num_random_blocks = num_random_blocks\n all_local_blocks = 0\n for local_blocks in local_window_blocks:\n all_local_blocks += local_blocks\n if (self.num_blocks < all_local_blocks):\n raise ValueError(\n f'Number of all local window blocks, {all_local_blocks}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\n )\n self.local_window_blocks = local_window_blocks\n if (self.num_blocks < len(global_block_indices)):\n raise ValueError(\n f'Number of global blocks indices, {global_block_indices}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\n )\n for idx in global_block_indices:\n if idx >= self.num_blocks:\n raise ValueError(\n f'Global block index, {global_block_indices[idx]}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\n )\n self.global_block_indices = global_block_indices\n\n if (global_block_end_indices is not None):\n if (len(global_block_indices) != len(global_block_end_indices)):\n raise ValueError(\n f'Global block start indices length, {len(global_block_indices)}, must be same as global block end indices length, {len(global_block_end_indices)}!'\n )\n for _, (start_idx, end_idx) in enumerate(zip(global_block_indices, global_block_end_indices)):\n if end_idx > self.num_blocks:\n raise ValueError(\n f'Global block end index, {global_block_end_indices[idx]}, must be smaller (equal) than overal number of blocks in a row, {self.num_blocks}!'\n )\n if start_idx >= end_idx:\n raise ValueError(\n f'Global block start index, {start_idx}, must be smaller than global block end index, {end_idx}!'\n )\n self.global_block_end_indices = global_block_end_indices\n\n if (attention != 'unidirectional' and attention != 'bidirectional'):\n raise NotImplementedError(\n 'only \\\"uni/bi-directional\\\" attentions are supported for now!')\n self.attention = attention\n\n if (attention != 'bidirectional' and horizental_global_attention):\n raise ValueError(\n 'only \\\"bi-directional\\\" attentions can support horizental global attention!'\n )\n self.horizental_global_attention = horizental_global_attention\n self.make_layout()","def inception_block_1a(X):\n\n X_3x3 = Conv2D(96, (1, 1), data_format='channels_first', name='inception_3a_3x3_conv1')(X)\n X_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn1')(X_3x3)\n X_3x3 = Activation('relu')(X_3x3)\n X_3x3 = ZeroPadding2D(padding=(1, 1), data_format='channels_first')(X_3x3)\n X_3x3 = Conv2D(128, (3, 3), data_format='channels_first', name='inception_3a_3x3_conv2')(X_3x3)\n X_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn2')(X_3x3)\n X_3x3 = Activation('relu')(X_3x3)\n\n X_5x5 = Conv2D(16, (1, 1), data_format='channels_first', name='inception_3a_5x5_conv1')(X)\n X_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn1')(X_5x5)\n X_5x5 = Activation('relu')(X_5x5)\n X_5x5 = ZeroPadding2D(padding=(2, 2), data_format='channels_first')(X_5x5)\n X_5x5 = Conv2D(32, (5, 5), data_format='channels_first', name='inception_3a_5x5_conv2')(X_5x5)\n X_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn2')(X_5x5)\n X_5x5 = Activation('relu')(X_5x5)\n\n X_pool = MaxPooling2D(pool_size=3, strides=2, data_format='channels_first')(X)\n X_pool = Conv2D(32, (1, 1), data_format='channels_first', name='inception_3a_pool_conv')(X_pool)\n X_pool = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_pool_bn')(X_pool)\n X_pool = Activation('relu')(X_pool)\n X_pool = ZeroPadding2D(padding=((3, 4), (3, 4)), data_format='channels_first')(X_pool)\n\n X_1x1 = Conv2D(64, (1, 1), data_format='channels_first', name='inception_3a_1x1_conv')(X)\n X_1x1 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_1x1_bn')(X_1x1)\n X_1x1 = Activation('relu')(X_1x1)\n\n # CONCAT\n inception = concatenate([X_3x3, X_5x5, X_pool, X_1x1], axis=1)\n\n return inception","def __init__(\n self,\n include_top=True,\n weights='imagenet',\n input_tensor=None,\n input_shape=None,\n pooling=None,\n classes=1000,\n **kwargs):\n backend, layers, models, keras_utils = get_submodules_from_kwargs(kwargs)\n self.include_top = include_top\n self.pooling = pooling\n self.weights = weights\n self.backend = backend\n self.layers = layers\n self.classes = classes\n\n if not (weights in {'imagenet', None} or os.path.exists(weights)):\n raise ValueError('The `weights` argument should be either '\n '`None` (random initialization), `imagenet` '\n '(pre-training on ImageNet), '\n 'or the path to the weights file to be loaded.')\n\n if weights == 'imagenet' and include_top and classes != 1000:\n raise ValueError('If using `weights` as `\"imagenet\"` with `include_top`'\n ' as true, `classes` should be 1000')\n self.block1_conv1 = []\n self.block1_conv2 = []\n self.block1_pool = []\n\n self.block2_conv1 = []\n self.block2_conv2 = []\n self.block2_pool = []\n\n self.block3_conv1 = []\n self.block3_conv2 = []\n self.block3_conv3 = []\n self.block3_conv4 = []\n self.block3_pool = []\n\n self.block4_conv1 = []\n self.block4_conv2 = []\n self.block4_conv3 = []\n self.block4_conv4 = []\n self.block4_pool = []\n\n self.block5_conv1 = []\n self.block5_conv2 = []\n self.block5_conv3 = []\n self.block5_conv4 = []\n self.block5_pool = []\n\n for i in xrange(FLAGS.num_replica):\n # Block 1\n self.block1_conv1.append(layers.Conv2D(64, (3, 3),\n activation='relu',\n padding='same',\n name='block1_conv1'))\n self.block1_conv2.append(layers.Conv2D(64, (3, 3),\n activation='relu',\n padding='same',\n name='block1_conv2'))\n self.block1_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block1_pool'))\n\n # Block 2\n self.block2_conv1.append(layers.Conv2D(128, (3, 3),\n activation='relu',\n padding='same',\n name='block2_conv1'))\n self.block2_conv2.append(layers.Conv2D(128, (3, 3),\n activation='relu',\n padding='same',\n name='block2_conv2'))\n self.block2_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block2_pool'))\n\n # Block 3\n self.block3_conv1.append(layers.Conv2D(256, (3, 3),\n activation='relu',\n padding='same',\n name='block3_conv1'))\n self.block3_conv2.append(layers.Conv2D(256, (3, 3),\n activation='relu',\n padding='same',\n name='block3_conv2'))\n self.block3_conv3.append(layers.Conv2D(256, (3, 3),\n activation='relu',\n padding='same',\n name='block3_conv3'))\n self.block3_conv4.append(layers.Conv2D(256, (3, 3),\n activation='relu',\n padding='same',\n name='block3_conv4'))\n self.block3_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block3_pool'))\n\n # Block 4\n self.block4_conv1.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block4_conv1'))\n self.block4_conv2.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block4_conv2'))\n self.block4_conv3.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block4_conv3'))\n self.block4_conv4.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block4_conv4'))\n self.block4_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block4_pool'))\n\n # Block 5\n self.block5_conv1.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block5_conv1'))\n self.block5_conv2.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block5_conv2'))\n self.block5_conv3.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block5_conv3'))\n self.block5_conv4.append(layers.Conv2D(512, (3, 3),\n activation='relu',\n padding='same',\n name='block5_conv4'))\n self.block5_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block5_pool'))\n\n if include_top:\n # Classification block\n self.flatten = layers.Flatten(name='flatten')\n self.fc1 = layers.Dense(4096, activation='relu', name='fc1')\n self.fc2 = layers.Dense(4096, activation='relu', name='fc2')\n self.predict = layers.Dense(classes, activation='softmax', name='predictions')\n else:\n if pooling == 'avg':\n self.pool = layers.GlobalAveragePooling2D()\n elif pooling == 'max':\n self.pool = layers.GlobalMaxPooling2D()","def _init_layers(self) -> None:\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\n self.embed_dims = self.self_attn.embed_dims\n self.ffn = FFN(**self.ffn_cfg)\n norms_list = [\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\n for _ in range(2)\n ]\n self.norms = ModuleList(norms_list)","def mtf_transformer_lm_moe():\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\n hparams.decoder_layers = [\"att\", \"moe\"] * 4\n moe.set_default_moe_hparams(hparams)\n hparams.mesh_shape = \"all:8\"\n hparams.layout = \"batch:all;experts:all\"\n return hparams","def init_layers(self):\n self.user_emb = nn.Embedding(self.n_users, self.emb_dim)\n self.item_emb = nn.Embedding(self.n_items, self.emb_dim)\n init_range = 0.1 * (self.emb_dim) ** (-1 / 2)\n self.item_emb.weight.data.uniform_(-init_range, init_range)\n self.fc_u_1_mu = nn.Linear(self.user_fea_dim, self.late_dim)\n self.fc_u_2_mu = nn.Linear(self.late_dim, self.emb_dim * 2)\n self.fc_i_1_mu = nn.Linear(self.item_fea_dim, self.late_dim)\n self.fc_i_2_mu = nn.Linear(self.late_dim, self.emb_dim * 2)","def _indentity_block(self, X, filters, f, stage, block):\n\t\tconv_layer_name = 'res' + str(stage) + block + '_branch'\n\t\tbn_layer_name = 'bm' + str(stage) + block + '_branch'\n\n\t\tX_shortcut = X\n\n\t\tF1, F2, F3 = filters\n\n\t\t# First component of main path\n\t\tX = Conv2D(filters = F1, kernel_size = (1, 1), strides = (1, 1), padding = 'valid',\n\t\t\tname = conv_layer_name + '2a', kernel_initializer = glorot_uniform(seed=0))(X)\n\t\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2a')(X)\n\t\tX = Activation('relu')(X)\n\n\t\t# Second component of main path\n\t\tX = Conv2D(filters = F2, kernel_size = (1, 1), strides = (1, 1), padding = 'same',\n\t\t\tname = conv_layer_name + '2b', kernel_initializer = glorot_uniform(seed=0))(X)\n\t\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2b')(X)\n\t\tX = Activation('relu')(X)\n\n\t\t# Third component of main path\n\t\tX = Conv2D(filters = F3, kernel_size = (1, 1), strides = (1, 1), padding = 'valid',\n\t\t\tname = conv_layer_name + '2c', kernel_initializer = glorot_uniform(seed=0))(X)\n\t\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2c')(X)\n\n\t\t# Final step : adding the shortcut componet to X and applying 'relu' activation on the combination \n\t\tX = Add()([X_shortcut,X])\n\t\tX = Activation('relu')(X)\n\n\t\treturn X","def identity_block(filters, stage, block):\n\n def layer(input_tensor):\n conv_params = get_conv_params()\n bn_params = get_bn_params()\n conv_name, bn_name, relu_name, sc_name = handle_block_names(stage, block)\n\n x = BatchNormalization(name=bn_name + '0', **bn_params)(input_tensor)\n x = Activation('relu', name=relu_name + '0')(x)\n x = Conv2D(filters // 2, (1, 1), name=conv_name + '1', **conv_params)(x)\n x = BatchNormalization(name=bn_name + '1', **bn_params)(x)\n x = Activation('relu', name=relu_name + '1')(x)\n\n x = ZeroPadding2D(padding=(1, 1))(x)\n x = GroupConv2D(filters // 2, (3, 3), conv_params, conv_name + '2')(x)\n x = BatchNormalization(name=bn_name + '2', **bn_params)(x)\n x = Activation('relu', name=relu_name + '2')(x)\n\n # x = Conv2D(filters * 2, (1, 1), name=conv_name + '3', **conv_params)(x)\n x = Conv2D(filters, (1, 1), name=conv_name + '3', **conv_params)(x)\n x = BatchNormalization(name=bn_name + '3', **bn_params)(x)\n x = Add()([x, input_tensor])\n\n return x\n\n return layer","def __init__(self, embed_dim: int, hidden_dim: int, num_max_positions: int, num_heads: int, num_layers: int, dropout: float, causal: bool,\n initializer_range: float):\n super().__init__()\n tokenizer = BertTokenizer.from_pretrained('bert-base-cased', do_lower_case=False)\n num_embeddings = len(tokenizer.vocab)\n self.initializer_range = initializer_range\n self.transformer = Transformer(embed_dim, hidden_dim, num_embeddings,\n num_max_positions, num_heads, num_layers,\n dropout, causal=causal)\n\n self.lm_head = torch.nn.Linear(embed_dim, num_embeddings, bias=False)\n self.apply(self.init_weights)\n self.tie_weights()","def blocks2modules(blocks, net_info, device=\"cpu\"):\n modules = torch.nn.ModuleList()\n\n # Track number of channels (filters) in the output of each layer; this\n # is necessary to determine layer input/output shapes for various layers.\n curr_out_channels = None\n prev_layer_out_channels = net_info[\"channels\"]\n out_channels_list = []\n\n for i, block in enumerate(blocks):\n module = torch.nn.Sequential()\n\n if block[\"type\"] == \"convolutional\":\n batch_normalize = \"batch_normalize\" in block\n bias = not batch_normalize\n kernel_size = block[\"size\"]\n padding = (kernel_size - 1) // 2 if \"pad\" in block else 0\n in_channels = prev_layer_out_channels\n out_channels = block[\"filters\"]\n\n conv = torch.nn.Conv2d(\n in_channels=in_channels, out_channels=out_channels,\n kernel_size=kernel_size, stride=block[\"stride\"],\n padding=padding, bias=bias\n )\n module.add_module(\"conv_{}\".format(i), conv)\n\n if batch_normalize:\n bn = torch.nn.BatchNorm2d(num_features=out_channels)\n module.add_module(\"batch_norm_{}\".format(i), bn)\n\n if block[\"activation\"] == \"leaky\":\n acti = torch.nn.LeakyReLU(negative_slope=0.1, inplace=True)\n module.add_module(\"leaky_{}\".format(i), acti)\n elif block[\"activation\"] == \"linear\":\n # NOTE: Darknet src files call out \"linear\" vs \"relu\" but we\n # use ReLU here.\n acti = torch.nn.ReLU(inplace=True)\n\n # Update the number of current (output) channels.\n curr_out_channels = out_channels\n\n elif block[\"type\"] == \"maxpool\":\n stride = block[\"stride\"]\n maxpool = MaxPool2d(\n kernel_size=block[\"size\"], stride=stride\n )\n module.add_module(\"maxpool_{}\".format(i), maxpool)\n\n elif block[\"type\"] == \"route\":\n # Route layer concatenates outputs along channel dim; add dummy\n # layer and handle the actual logic in Darknet.forward().\n module.add_module(\"route_{}\".format(i), DummyLayer())\n\n out_channels = sum(\n out_channels_list[layer_idx] for layer_idx in block[\"layers\"]\n )\n\n curr_out_channels = out_channels\n\n elif block[\"type\"] == \"shortcut\":\n # Shortcut layer sums outputs from previous layers; add dummy\n # layer and handle the actual logic in Darknet.forward().\n module.add_module(\"shortcut_{}\".format(i), DummyLayer())\n\n if \"activation\" in block:\n if block[\"activation\"] == \"leaky\":\n acti = torch.nn.LeakyReLU(negative_slope=0.1, inplace=True)\n module.add_module(\"leaky_{}\".format(i), acti)\n elif block[\"activation\"] == \"linear\":\n acti = torch.nn.ReLU(inplace=True)\n\n assert out_channels == out_channels_list[i + block[\"from\"]]\n curr_out_channels = out_channels\n\n elif block[\"type\"] == \"upsample\":\n # NOTE: torch.nn.Upsample is deprecated in favor of Interpolate;\n # consider using this and/or other interpolation methods?\n upsample = torch.nn.Upsample(\n scale_factor=block[\"stride\"], mode=\"nearest\"\n )\n module.add_module(\"upsample_{}\".format(i), upsample)\n\n elif block[\"type\"] == \"yolo\":\n yolo = YOLOLayer(block[\"anchors\"], block[\"mask\"], device=device)\n module.add_module(\"yolo_{}\".format(i), yolo)\n\n modules.append(module)\n prev_layer_out_channels = curr_out_channels\n out_channels_list.append(curr_out_channels)\n\n return modules","def __init__(self, dm, h, hidden, drop_rate=0.1):\n super().__init__()\n self.mha = MultiHeadAttention(dm, h)\n self.ffn = point_wise_feed_forward_network(dm, hidden)\n\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)","def __init__(self, hidden_layers, dropout=0.1, activation='gelu'):\n super(MLPBlock, self).__init__()\n\n layers = []\n for num_units in hidden_layers:\n layers.extend([\n Dense(num_units, activation=activation),\n Dropout(dropout)\n ])\n\n self.mlp = Sequential(layers)","def __init__(self, cfg, momentum=False):\n\n super().__init__()\n self.ent_output_size = cfg.ent_output_size\n self.span_batch_size = cfg.span_batch_size\n self.position_embedding_dims = cfg.position_embedding_dims\n self.att_size = cfg.att_size\n self.momentum = momentum\n self.activation = nn.GELU()\n self.device = cfg.device\n\n self.bert_encoder = BertEncoder(bert_model_name=cfg.bert_model_name,\n trainable=cfg.fine_tune,\n output_size=cfg.bert_output_size,\n activation=self.activation)\n\n self.entity_span_extractor = CNNSpanExtractor(\n input_size=self.bert_encoder.get_output_dims(),\n num_filters=cfg.entity_cnn_output_channels,\n ngram_filter_sizes=cfg.entity_cnn_kernel_sizes,\n dropout=cfg.dropout)\n\n if self.ent_output_size > 0:\n self.ent2hidden = BertLinear(input_size=self.entity_span_extractor.get_output_dims(),\n output_size=self.ent_output_size,\n activation=self.activation,\n dropout=cfg.dropout)\n else:\n self.ent_output_size = self.entity_span_extractor.get_output_dims()\n self.ent2hidden = lambda x: x\n\n self.entity_span_mlp = BertLinear(input_size=self.ent_output_size,\n output_size=self.ent_output_size,\n activation=self.activation,\n dropout=cfg.dropout)\n self.entity_span_decoder = VanillaSoftmaxDecoder(hidden_size=self.ent_output_size,\n label_size=6)\n\n self.global_position_embedding = nn.Embedding(150, 200)\n self.global_position_embedding.weight.data.normal_(mean=0.0, std=0.02)\n self.masked_token_mlp = BertLinear(input_size=self.bert_encoder.get_output_dims() + 200,\n output_size=self.bert_encoder.get_output_dims(),\n activation=self.activation,\n dropout=cfg.dropout)\n self.masked_token_decoder = nn.Linear(self.bert_encoder.get_output_dims(),\n 28996,\n bias=False)\n self.masked_token_decoder.weight.data.normal_(mean=0.0, std=0.02)\n self.masked_token_decoder_bias = nn.Parameter(torch.zeros(28996))\n\n self.position_embedding = nn.Embedding(7, self.position_embedding_dims)\n self.position_embedding.weight.data.normal_(mean=0.0, std=0.02)\n self.attention_encoder = PosAwareAttEncoder(self.ent_output_size,\n self.bert_encoder.get_output_dims(),\n 2 * self.position_embedding_dims,\n self.att_size,\n activation=self.activation,\n dropout=cfg.dropout)\n\n self.mlp_head1 = BertLinear(self.ent_output_size,\n self.bert_encoder.get_output_dims(),\n activation=self.activation,\n dropout=cfg.dropout)\n self.mlp_head2 = BertLinear(self.bert_encoder.get_output_dims(),\n self.bert_encoder.get_output_dims(),\n activation=self.activation,\n dropout=cfg.dropout)\n\n self.masked_token_loss = nn.CrossEntropyLoss()","def __init__(self, input_dim=6, hidden_dim=128, T=10, in_features=2*128 + 16, num_layers=1, out_features=1):\n super(InputAttentionEncoder, self).__init__()\n self.input_dim = input_dim\n self.hidden_dim = hidden_dim\n self.T = T\n\n in_features = hidden_dim * 2 + T - 1\n\n self.lstm_unit = nn.LSTM(input_size=input_dim, hidden_size=hidden_dim, num_layers=num_layers)\n self.attn_linear = nn.Linear(in_features=in_features, out_features=out_features)\n self.bn1 = nn.BatchNorm1d(out_features)\n self.fc2 = nn.Linear(out_features, out_features)","def __init__(self, wav2vec2_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([wav2vec2_layer.attention.q_proj.weight, wav2vec2_layer.attention.k_proj.weight, wav2vec2_layer.attention.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([wav2vec2_layer.attention.q_proj.bias, wav2vec2_layer.attention.k_proj.bias, wav2vec2_layer.attention.v_proj.bias]))\n self.out_proj_weight = wav2vec2_layer.attention.out_proj.weight\n self.out_proj_bias = wav2vec2_layer.attention.out_proj.bias\n self.linear1_weight = wav2vec2_layer.feed_forward.intermediate_dense.weight\n self.linear1_bias = wav2vec2_layer.feed_forward.intermediate_dense.bias\n self.linear2_weight = wav2vec2_layer.feed_forward.output_dense.weight\n self.linear2_bias = wav2vec2_layer.feed_forward.output_dense.bias\n self.norm1_eps = wav2vec2_layer.layer_norm.eps\n self.norm1_weight = wav2vec2_layer.layer_norm.weight\n self.norm1_bias = wav2vec2_layer.layer_norm.bias\n self.norm2_eps = wav2vec2_layer.final_layer_norm.eps\n self.norm2_weight = wav2vec2_layer.final_layer_norm.weight\n self.norm2_bias = wav2vec2_layer.final_layer_norm.bias\n self.num_heads = wav2vec2_layer.attention.num_heads\n self.embed_dim = wav2vec2_layer.attention.embed_dim\n self.is_last_layer = False\n self.validate_bettertransformer()","def __init__(self,\r\n in_channels_1, in_channels_2, out_channels,\r\n kernel_size_1, kernel_size_2,\r\n stride_1, up_stride_2,\r\n padding_1, up_padding_2, output_padding=0,\r\n activation_in='relu', activation_out='lrelu',\r\n norm_in='bn', norm_out='none'):\r\n\r\n super(RefinementBlock, self).__init__()\r\n\r\n self.conv_1 = conv_block(\r\n in_channels=in_channels_1,\r\n out_channels=out_channels,\r\n kernel_size=kernel_size_1,\r\n stride=stride_1,\r\n padding=padding_1,\r\n norm='none',\r\n activation=activation_in\r\n )\r\n\r\n self.upconv_2 = upconv_block(\r\n in_channels=in_channels_2,\r\n out_channels=out_channels,\r\n kernel_size=kernel_size_2,\r\n stride=up_stride_2,\r\n padding=up_padding_2,\r\n output_padding=output_padding,\r\n norm=norm_in,\r\n activation='none'\r\n )\r\n\r\n self.conv_3 = conv_block(\r\n in_channels=out_channels,\r\n out_channels=out_channels,\r\n kernel_size=3,\r\n stride=1,\r\n padding=1,\r\n norm='none',\r\n activation='none'\r\n )\r\n\r\n self.conv_4 = conv_block(\r\n in_channels=out_channels,\r\n out_channels=out_channels,\r\n kernel_size=3,\r\n stride=1,\r\n padding=1,\r\n norm='none',\r\n activation='none'\r\n )\r\n\r\n self.out_act = _activation(act_type=activation_out)\r\n self.out_norm = _norm(norm_type=norm_out, channels=out_channels)","def __init__(self, albert_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([albert_layer.attention.query.weight, albert_layer.attention.key.weight, albert_layer.attention.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([albert_layer.attention.query.bias, albert_layer.attention.key.bias, albert_layer.attention.value.bias]))\n self.out_proj_weight = albert_layer.attention.dense.weight\n self.out_proj_bias = albert_layer.attention.dense.bias\n self.linear1_weight = albert_layer.ffn.weight\n self.linear1_bias = albert_layer.ffn.bias\n self.linear2_weight = albert_layer.ffn_output.weight\n self.linear2_bias = albert_layer.ffn_output.bias\n self.norm1_eps = albert_layer.attention.LayerNorm.eps\n self.norm1_weight = albert_layer.attention.LayerNorm.weight\n self.norm1_bias = albert_layer.attention.LayerNorm.bias\n self.norm2_eps = albert_layer.full_layer_layer_norm.eps\n self.norm2_weight = albert_layer.full_layer_layer_norm.weight\n self.norm2_bias = albert_layer.full_layer_layer_norm.bias\n self.num_heads = albert_layer.attention.num_attention_heads\n self.embed_dim = albert_layer.attention.all_head_size\n self.is_last_layer = False\n self.validate_bettertransformer()","def forward(self, x):\n\n if x.dim() == 3:\n x = x.unsqueeze(1)\n x = x.transpose(1, 3)\n x = self.norm0(x)\n x = x.transpose(1, 3)\n\n x = self.conv_block1(x, pool_size=(2, 2), pool_type=\"avg\")\n x = F.dropout(x, p=0.2, training=self.training)\n x = self.conv_block2(x, pool_size=(2, 2), pool_type=\"avg\")\n x = F.dropout(x, p=0.2, training=self.training)\n x = self.conv_block3(x, pool_size=(2, 2), pool_type=\"avg\")\n x = F.dropout(x, p=0.2, training=self.training)\n x3_out = self.conv_block4(x, pool_size=(2, 2), pool_type=\"avg\")\n x = F.dropout(x3_out, p=0.2, training=self.training)\n x2_out = self.conv_block5(x, pool_size=(2, 2), pool_type=\"avg\")\n x = F.dropout(x2_out, p=0.2, training=self.training)\n x1_out = self.conv_block6(x, pool_size=(1, 1), pool_type=\"avg\")\n x = F.dropout(x1_out, p=0.2, training=self.training)\n x = torch.mean(x, dim=3)\n\n (x1, _) = torch.max(x, dim=2)\n x2 = torch.mean(x, dim=2)\n x = x1 + x2\n\n # [B x 1 x emb_dim]\n if not self.return_reps:\n return x.unsqueeze(1)\n\n return x.unsqueeze(1), (x1_out, x2_out, x3_out)","def __init__(self, whisper_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([whisper_layer.self_attn.q_proj.weight, whisper_layer.self_attn.k_proj.weight, whisper_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([whisper_layer.self_attn.q_proj.bias, torch.zeros_like(whisper_layer.self_attn.q_proj.bias), whisper_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = whisper_layer.self_attn.out_proj.weight\n self.out_proj_bias = whisper_layer.self_attn.out_proj.bias\n self.linear1_weight = whisper_layer.fc1.weight\n self.linear1_bias = whisper_layer.fc1.bias\n self.linear2_weight = whisper_layer.fc2.weight\n self.linear2_bias = whisper_layer.fc2.bias\n self.norm1_eps = whisper_layer.self_attn_layer_norm.eps\n self.norm1_weight = whisper_layer.self_attn_layer_norm.weight\n self.norm1_bias = whisper_layer.self_attn_layer_norm.bias\n self.norm2_eps = whisper_layer.final_layer_norm.eps\n self.norm2_weight = whisper_layer.final_layer_norm.weight\n self.norm2_bias = whisper_layer.final_layer_norm.bias\n self.num_heads = whisper_layer.self_attn.num_heads\n self.embed_dim = whisper_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def forward_test(self, imgs, token_ids=None, segment_ids=None, input_mask=None, **kwargs):\n imgs = imgs.reshape((-1,) + imgs.shape[2:])\n if self.from_scratch:\n imgs = imgs / 255.0\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\n B, D, T, H, W = visual_token.shape\n B_text = token_ids.shape[0]\n if B_text != B:\n visual_token = visual_token.view(B_text, -1, D, T, H, W)\n visual_token = visual_token.mean(dim=1)\n B = B_text\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\n segment_ids = segment_ids.reshape((-1, ) + segment_ids.shape[2:])\n input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\n # only use the multimodal transformer for \n if self.separate_test:\n # text_only_out = self.multimodal_backbone.forward_text(token_ids, input_mask)\n # visual_emb, text_emb = self.ssl_head(visual_token, text_only_out['last_hidden_state'], input_mask, token_ids)\n text_only_out = self.text_backbone(token_ids, input_mask)\n text_out_last_hidden_state = text_only_out['last_hidden_state']\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_last_hidden_state, input_mask, token_ids)\n\n return visual_emb, text_emb\n \n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\n output = self.multimodal_backbone(visual_token, token_ids, segment_ids, input_mask)\n all_cls_emb = output['last_hidden_state'][:, 0]\n itm_output = self.itm_head(all_cls_emb)\n itm_output = torch.softmax(itm_output, dim=-1)\n return itm_output","def identity_block(input_tensor, kernel_size, filters, stage, block_id, data_format='channels_last', train_bn=True):\n\n filter1, filter2, filter3 = filters\n conv_name_base = 'res' + str(stage) + block_id + '_branch'\n bn_name_base = 'bn' + str(stage) + block_id + '_branch'\n\n shortcut = input_tensor\n\n x = tl.layers.Conv2d(input_tensor, filter1, (1, 1), padding='VALID', data_format=data_format,\n W_init=W_init, b_init=b_init, name=conv_name_base + '2a')\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2a')\n\n x = tl.layers.Conv2d(x, filter2, (kernel_size, kernel_size), padding='SAME', data_format=data_format,\n W_init=W_init, b_init=b_init, name=conv_name_base + '2b')\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2b')\n\n x = tl.layers.Conv2d(x, filter3, (1, 1), padding='VALID', data_format=data_format,\n W_init=W_init, b_init=b_init, name=conv_name_base + '2c')\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2c')\n\n x = tl.layers.ElementwiseLayer([x, shortcut], combine_fn=tf.add, act=tf.nn.relu, name=\"add\")\n return x","def inception_block_1a(X):\n\tX_3x3 = Conv2D(96, (1, 1), data_format='channels_first', name ='inception_3a_3x3_conv1')(X)\n\tX_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name = 'inception_3a_3x3_bn1')(X_3x3)\n\tX_3x3 = Activation('relu')(X_3x3)\n\tX_3x3 = ZeroPadding2D(padding=(1, 1), data_format='channels_first')(X_3x3)\n\tX_3x3 = Conv2D(128, (3, 3), data_format='channels_first', name='inception_3a_3x3_conv2')(X_3x3)\n\tX_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn2')(X_3x3)\n\tX_3x3 = Activation('relu')(X_3x3)\n\tX_5x5 = Conv2D(16, (1, 1), data_format='channels_first', name='inception_3a_5x5_conv1')(X)\n\tX_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn1')(X_5x5)\n\tX_5x5 = Activation('relu')(X_5x5)\n\tX_5x5 = ZeroPadding2D(padding=(2, 2), data_format='channels_first')(X_5x5)\n\tX_5x5 = Conv2D(32, (5, 5), data_format='channels_first', name='inception_3a_5x5_conv2')(X_5x5)\n\tX_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn2')(X_5x5)\n\tX_5x5 = Activation('relu')(X_5x5)\n\tX_pool = MaxPooling2D(pool_size=3, strides=2, data_format='channels_first')(X)\n\tX_pool = Conv2D(32, (1, 1), data_format='channels_first', name='inception_3a_pool_conv')(X_pool)\n\tX_pool = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_pool_bn')(X_pool)\n\tX_pool = Activation('relu')(X_pool)\n\tX_pool = ZeroPadding2D(padding=((3, 4), (3, 4)), data_format='channels_first')(X_pool)\n\tX_1x1 = Conv2D(64, (1, 1), data_format='channels_first', name='inception_3a_1x1_conv')(X)\n\tX_1x1 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_1x1_bn')(X_1x1)\n\tX_1x1 = Activation('relu')(X_1x1)\n\t# CONCAT\n\tinception = concatenate([X_3x3, X_5x5, X_pool, X_1x1], axis=1)\n\treturn inception","def __init__(self, num_heads: int, size: int, dropout: float = 0.1):\n super(MultiHeadedAttention, self).__init__()\n\n assert size % num_heads == 0\n\n self.head_size = head_size = size // num_heads\n self.model_size = size\n self.num_heads = num_heads\n\n self.k_layer = nn.Linear(size, num_heads * head_size)\n self.v_layer = nn.Linear(size, num_heads * head_size)\n self.q_layer = nn.Linear(size, num_heads * head_size)\n\n self.output_layer = nn.Linear(size, size)\n self.softmax = nn.Softmax(dim=-1)\n self.dropout = nn.Dropout(dropout)","def __init__(self, idim, odim_tgt, odim_src, args, ignore_id=-1):\n torch.nn.Module.__init__(self)\n if args.transformer_attn_dropout_rate is None:\n args.transformer_attn_dropout_rate = args.dropout_rate\n\n # special tokens and model dimensions\n self.pad = 0\n self.sos_tgt = odim_tgt - 1\n self.eos_tgt = odim_tgt - 1\n self.sos_src = odim_src - 1\n self.eos_src = odim_src - 1\n self.odim_tgt = odim_tgt\n self.odim_src = odim_src\n self.idim = idim\n self.adim = args.adim\n self.ignore_id = ignore_id\n\n # submodule\n self.mtlalpha = getattr(args, \"mtlalpha\", 0.0)\n self.asr_weight = getattr(args, \"asr_weight\", 0.0)\n self.mt_weight = getattr(args, \"mt_weight\", 0.0)\n self.num_decoders = getattr(args, \"num_decoders\", 2)\n self.do_st = getattr(args, \"do_st\", True)\n self.do_mt = getattr(args, \"do_mt\", self.mt_weight > 0.0)\n self.do_asr = self.asr_weight > 0 and self.mtlalpha < 1\n\n # cross-attention parameters\n self.cross_weight = getattr(args, \"cross_weight\", 0.0)\n self.cross_self = getattr(args, \"cross_self\", False)\n self.cross_src = getattr(args, \"cross_src\", False)\n self.cross_operator = getattr(args, \"cross_operator\", None)\n self.cross_to_asr = getattr(args, \"cross_to_asr\", False)\n self.cross_to_st = getattr(args, \"cross_to_st\", False)\n self.wait_k_asr = getattr(args, \"wait_k_asr\", 0)\n self.wait_k_st = getattr(args, \"wait_k_st\", 0)\n self.cross_src_from = getattr(args, \"cross_src_from\", \"embedding\")\n self.cross_self_from = getattr(args, \"cross_self_from\", \"embedding\")\n self.cross_shared = getattr(args, \"cross_shared\", False)\n self.cross_weight_learnable = getattr(args, \"cross_weight_learnable\", False)\n\n # one-to-many models parameters\n self.use_joint_dict = getattr(args, \"use_joint_dict\", True)\n self.one_to_many = getattr(args, \"one_to_many\", False)\n self.use_lid = getattr(args, \"use_lid\", False)\n if self.use_joint_dict:\n self.langs_dict = getattr(args, \"langs_dict_tgt\", None)\n self.lang_tok = getattr(args, \"lang_tok\", None)\n self.lang_tok_mt = getattr(args, \"lang_tok_mt\", None)\n\n self.subsample = get_subsample(args, \n mode='mt' if self.do_mt else 'st', \n arch='transformer')\n self.reporter = MTReporter() if self.do_mt else Reporter() \n self.normalize_before = getattr(args, \"normalize_before\", True)\n\n # Backward compatability\n if self.cross_operator in [\"sum\", \"concat\"]:\n if self.cross_self and self.cross_src:\n self.cross_operator = \"self_src\" + self.cross_operator\n elif self.cross_self:\n self.cross_operator = \"self_\" + self.cross_operator\n elif self.cross_src:\n self.cross_operator = \"src_\" + self.cross_operator\n if self.cross_operator:\n assert self.cross_operator in ['self_sum', 'self_concat', 'src_sum', \n 'src_concat', 'self_src_sum', 'self_src_concat']\n\n # Check parameters\n if self.one_to_many:\n self.use_lid = True\n if not self.do_st:\n assert (not self.cross_to_asr) and (not self.cross_to_st)\n if self.cross_operator and 'sum' in self.cross_operator and self.cross_weight <= 0:\n assert (not self.cross_to_asr) and (not self.cross_to_st)\n if self.cross_to_asr or self.cross_to_st:\n assert self.do_st and self.do_asr\n assert self.cross_self or self.cross_src\n assert bool(self.cross_operator) == (self.do_asr and (self.cross_to_asr or self.cross_to_st))\n if self.cross_src_from != \"embedding\" or self.cross_self_from != \"embedding\":\n assert self.normalize_before\n if self.wait_k_asr > 0:\n assert self.wait_k_st == 0\n elif self.wait_k_st > 0:\n assert self.wait_k_asr == 0\n else:\n assert self.wait_k_asr == 0\n assert self.wait_k_st == 0\n\n logging.info(\"*** Cross attention parameters ***\")\n if self.cross_to_asr:\n logging.info(\"| Cross to ASR\")\n if self.cross_to_st:\n logging.info(\"| Cross to ST\")\n if self.cross_self:\n logging.info(\"| Cross at Self\")\n if self.cross_src:\n logging.info(\"| Cross at Source\")\n if self.cross_to_asr or self.cross_to_st:\n logging.info(f'| Cross operator: {self.cross_operator}')\n logging.info(f'| Cross sum weight: {self.cross_weight}')\n if self.cross_src:\n logging.info(f'| Cross source from: {self.cross_src_from}')\n if self.cross_self:\n logging.info(f'| Cross self from: {self.cross_self_from}')\n logging.info(f\"Use joint dictionary: {self.use_joint_dict}\")\n \n if (self.cross_src_from != \"embedding\" and self.cross_src) \\\n and (not self.normalize_before):\n logging.warning(f'WARNING: Resort to using \\\n self.cross_src_from == embedding for cross at source attention.')\n if (self.cross_self_from != \"embedding\" and self.cross_self) \\\n and (not self.normalize_before):\n logging.warning(f'WARNING: Resort to using \\\n self.cross_self_from == embedding for cross at self attention.')\n\n # Adapters\n self.use_adapters = getattr(args, \"use_adapters\", False)\n self.use_adapters_in_enc = getattr(args, \"use_adapters_in_enc\", False)\n adapter_names = getattr(args, \"adapters\", None)\n adapter_reduction_factor = getattr(args, \"adapter_reduction_factor\", None)\n adapter_reduction_factor_enc = getattr(args, \"adapter_reduction_factor_enc\", adapter_reduction_factor)\n use_adapters_for_asr = getattr(args, \"use_adapters_for_asr\", True)\n adapter_before_src_attn = getattr(args, \"adapter_before_src_attn\", False)\n adapter_after_mha = getattr(args, \"adapter_after_mha\", False)\n use_shared_adapters = getattr(args, \"use_shared_adapters\", False)\n use_shared_adapters_enc = getattr(args, \"use_shared_adapters_enc\", False)\n # if self.use_adapters and not use_adapters_for_asr:\n # assert not self.do_asr or \\\n # (self.do_asr and self.num_decoders != 1) or \\\n # (self.do_asr and not self.do_st) # for backward compatibility\n\n if adapter_names:\n if self.do_asr and not self.do_st:\n adapter_names = [str(args.char_list_src.index(f'<2{l}>')) for l in adapter_names]\n else:\n adapter_names = [str(args.char_list_tgt.index(f'<2{l}>')) for l in adapter_names]\n logging.info(f'| adapters = {adapter_names}')\n\n if self.do_st or self.do_asr:\n logging.info(f'Speech encoder')\n self.encoder = Encoder(\n idim=idim,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.eunits,\n num_blocks=args.elayers,\n input_layer=getattr(args, \"transformer_input_layer\", \"conv2d\"),\n dropout_rate=args.dropout_rate,\n positional_dropout_rate=args.dropout_rate,\n attention_dropout_rate=args.transformer_attn_dropout_rate,\n adapter_names=adapter_names if self.use_adapters_in_enc else None,\n reduction_factor=adapter_reduction_factor_enc,\n adapter_after_mha=adapter_after_mha,\n shared_adapters=use_shared_adapters_enc,\n )\n if self.do_st:\n logging.info('ST decoder')\n self.decoder = Decoder(\n odim=odim_tgt,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n dropout_rate=args.dropout_rate,\n positional_dropout_rate=args.dropout_rate,\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\n normalize_before=self.normalize_before,\n cross_operator=self.cross_operator if self.cross_to_st else None,\n cross_shared=self.cross_shared,\n cross_weight_learnable=self.cross_weight_learnable,\n cross_weight=self.cross_weight,\n use_output_layer=True if (self.use_joint_dict or \\\n (self.do_st and not self.do_asr)) else False,\n adapter_names=adapter_names,\n reduction_factor=adapter_reduction_factor,\n adapter_before_src_attn=adapter_before_src_attn,\n adapter_after_mha=adapter_after_mha,\n shared_adapters=use_shared_adapters,\n )\n if self.do_asr:\n logging.info('ASR decoder')\n self.decoder_asr = Decoder(\n odim=odim_src,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n dropout_rate=args.dropout_rate,\n positional_dropout_rate=args.dropout_rate,\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\n normalize_before=self.normalize_before,\n cross_operator=self.cross_operator if self.cross_to_asr else None,\n cross_shared=self.cross_shared,\n cross_weight_learnable=self.cross_weight_learnable,\n cross_weight=self.cross_weight,\n use_output_layer=True if (self.use_joint_dict or \\\n (self.do_asr and not self.do_st)) else False,\n adapter_names=adapter_names,\n reduction_factor=adapter_reduction_factor,\n adapter_before_src_attn=adapter_before_src_attn,\n adapter_after_mha=adapter_after_mha,\n shared_adapters=use_shared_adapters,\n )\n if self.num_decoders == 1 and self.do_st:\n logging.info('*** Use shared decoders *** ')\n self.decoder_asr = self.decoder\n\n if not self.use_joint_dict and (self.do_st and self.do_asr):\n self.output_layer = torch.nn.Linear(args.adim, odim_tgt)\n self.output_layer_asr = torch.nn.Linear(args.adim, odim_src)\n\n # submodule for MT task\n if self.do_mt:\n logging.info('MT encoder')\n self.encoder_mt = Encoder(\n idim=odim_src,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n input_layer='embed',\n dropout_rate=args.dropout_rate,\n positional_dropout_rate=args.dropout_rate,\n attention_dropout_rate=args.transformer_attn_dropout_rate,\n padding_idx=0\n )\n if not self.do_st:\n logging.info('MT decoder')\n self.decoder_mt = Decoder(\n odim=odim_tgt,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n dropout_rate=args.dropout_rate,\n positional_dropout_rate=args.dropout_rate,\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\n normalize_before=self.normalize_before,\n use_output_layer=True,\n )\n self.reset_parameters(args) # place after the submodule initialization\n if self.mtlalpha > 0.0:\n self.ctc = CTC(odim_src, args.adim, args.dropout_rate, \n ctc_type=args.ctc_type, reduce=True,\n zero_infinity=True)\n else:\n self.ctc = None\n\n if self.asr_weight > 0 and (args.report_cer or args.report_wer):\n from espnet.nets.e2e_asr_common import ErrorCalculator\n self.error_calculator = ErrorCalculator(args.char_list_src,\n args.sym_space, args.sym_blank,\n args.report_cer, args.report_wer)\n elif self.do_mt and getattr(args, \"report_bleu\", False):\n from espnet.nets.e2e_mt_common import ErrorCalculator\n self.error_calculator = ErrorCalculator(args.char_list_tgt,\n args.sym_space,\n args.report_bleu)\n else:\n self.error_calculator = None\n self.rnnlm = None\n\n # criterion\n if self.do_st:\n self.criterion_st = LabelSmoothingLoss(self.odim_tgt, self.ignore_id, args.lsm_weight,\n args.transformer_length_normalized_loss)\n if self.do_asr:\n self.criterion_asr = LabelSmoothingLoss(self.odim_src, self.ignore_id, args.lsm_weight,\n args.transformer_length_normalized_loss)\n if self.do_mt:\n self.criterion_mt = LabelSmoothingLoss(self.odim_tgt, self.ignore_id, args.lsm_weight,\n args.transformer_length_normalized_loss)\n self.normalize_length = args.transformer_length_normalized_loss # for PPL\n\n # Language embedding layer\n if self.lang_tok == \"encoder-pre-sum\":\n self.language_embeddings = build_embedding(self.langs_dict, self.idim, \n padding_idx=self.pad)\n logging.info(f'language_embeddings: {self.language_embeddings}')\n\n # Backward compatability\n if self.cross_operator:\n if \"sum\" in self.cross_operator:\n self.cross_operator = \"sum\"\n if \"concat\" in self.cross_operator: \n self.cross_operator = \"concat\"","def forward(self, data_batch):\n\n x = data_batch[0]\n im_info = data_batch[1]\n gt_boxes = data_batch[2]\n num_boxes = data_batch[3]\n rel_mat = data_batch[4]\n\n if self.training:\n self.iter_counter += 1\n\n input_imgs = x.clone()\n\n sources = list()\n loc = list()\n conf = list()\n\n self.batch_size = x.size(0)\n\n # apply vgg up to conv4_3 relu\n if isinstance(self.base, nn.ModuleList):\n for k,v in enumerate(self.base):\n x = v(x)\n else:\n x = self.base(x)\n\n s = self.L2Norm(x)\n sources.append(s)\n base_feat = s\n\n # apply vgg up to fc7\n if isinstance(self.conv5, nn.ModuleList):\n for k,v in enumerate(self.conv5):\n x = v(x)\n else:\n x = self.conv5(x)\n sources.append(x)\n\n # apply extra layers and cache source layer outputs\n for k, v in enumerate(self.extras):\n x = F.relu(v(x), inplace=True)\n if k % 2 == 1:\n sources.append(x)\n\n # apply multibox head to source layers\n for (x, l, c) in zip(sources, self.loc, self.conf):\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\n\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\n\n loc = loc.view(loc.size(0), -1, 4)\n conf = conf.view(conf.size(0), -1, self.num_classes)\n\n SSD_loss_cls = 0\n SSD_loss_bbox = 0\n if self.training:\n predictions = (\n loc,\n conf,\n self.priors.type_as(loc)\n )\n # targets = torch.cat([gt_boxes[:,:,:4] / self.size, gt_boxes[:,:,4:5]],dim=2)\n targets = gt_boxes\n SSD_loss_bbox, SSD_loss_cls = self.criterion(predictions, targets, num_boxes)\n\n conf = self.softmax(conf)\n\n # online data\n if self.training:\n if self.iter_counter > cfg.TRAIN.VMRN.ONLINEDATA_BEGIN_ITER:\n obj_rois, obj_num = self._obj_det(conf, loc, self.batch_size, im_info)\n obj_rois = obj_rois.type_as(gt_boxes)\n obj_num = obj_num.type_as(num_boxes)\n else:\n obj_rois = torch.FloatTensor([]).type_as(gt_boxes)\n obj_num = torch.LongTensor([]).type_as(num_boxes)\n obj_labels = None\n else:\n # when testing, this is object detection results\n # TODO: SUPPORT MULTI-IMAGE BATCH\n obj_rois, obj_num = self._obj_det(conf, loc, self.batch_size, im_info)\n if obj_rois.numel() > 0:\n obj_labels = obj_rois[:, 5]\n obj_rois = obj_rois[:, :5]\n obj_rois = obj_rois.type_as(gt_boxes)\n obj_num = obj_num.type_as(num_boxes)\n else:\n # there is no object detected\n obj_labels = torch.Tensor([]).type_as(gt_boxes).long()\n obj_rois = obj_rois.type_as(gt_boxes)\n obj_num = obj_num.type_as(num_boxes)\n\n if self.training:\n # offline data\n for i in range(self.batch_size):\n obj_rois = torch.cat([obj_rois,\n torch.cat([(i * torch.ones(num_boxes[i].item(), 1)).type_as(gt_boxes),\n (gt_boxes[i][:num_boxes[i]][:, 0:4])], 1)\n ])\n obj_num = torch.cat([obj_num, torch.Tensor([num_boxes[i]]).type_as(obj_num)])\n\n\n obj_rois = Variable(obj_rois)\n\n VMRN_rel_loss_cls = 0\n rel_cls_prob = torch.Tensor([]).type_as(obj_rois)\n if (obj_num > 1).sum().item() > 0:\n\n obj_pair_feat = self.VMRN_obj_pair_feat_extractor(input_imgs, obj_rois, self.batch_size, obj_num)\n # obj_pair_feat = obj_pair_feat.detach()\n rel_cls_score = self.VMRN_rel_cls_score(obj_pair_feat)\n\n rel_cls_prob = F.softmax(rel_cls_score)\n\n self.rel_batch_size = obj_pair_feat.size(0)\n\n if self.training:\n obj_pair_rel_label = self._generate_rel_labels(obj_rois, gt_boxes, obj_num, rel_mat)\n obj_pair_rel_label = obj_pair_rel_label.type_as(gt_boxes).long()\n\n rel_not_keep = (obj_pair_rel_label == 0)\n # no relationship is kept\n if (rel_not_keep == 0).sum().item() > 0:\n rel_keep = torch.nonzero(rel_not_keep == 0).view(-1)\n\n rel_cls_score = rel_cls_score[rel_keep]\n\n obj_pair_rel_label = obj_pair_rel_label[rel_keep]\n obj_pair_rel_label -= 1\n VMRN_rel_loss_cls = F.cross_entropy(rel_cls_score, obj_pair_rel_label)\n else:\n if (not cfg.TEST.VMRN.ISEX) and cfg.TRAIN.VMRN.ISEX:\n rel_cls_prob = rel_cls_prob[::2, :]\n\n rel_result = None\n if not self.training:\n if obj_rois.numel() > 0:\n pred_boxes = obj_rois.data[:,1:5]\n pred_boxes[:, 0::2] /= im_info[0][3].item()\n pred_boxes[:, 1::2] /= im_info[0][2].item()\n rel_result = (pred_boxes, obj_labels, rel_cls_prob.data)\n else:\n rel_result = (obj_rois.data, obj_labels, rel_cls_prob.data)\n\n return loc, conf, rel_result, SSD_loss_bbox, SSD_loss_cls, VMRN_rel_loss_cls","def __init__(self, in_channels, num_classes):\n super(InceptionAux, self).__init__()\n self.relu = nn.LeakyReLU()\n self.dropout1 = nn.Dropout(p=0.5)\n self.pool = nn.AvgPool2d(kernel_size=5, stride=3)\n self.conv = conv_block(in_channels, 128, kernel_size=1)\n self.fc1 = nn.Linear(12800, 1024)\n self.fc2 = nn.Linear(1024, 420)\n self.dropout2 = nn.Dropout(p=0.4)\n self.fc3 = nn.Linear(420,128)\n self.dropout3 = nn.Dropout(p=0.4)\n self.fc4 = nn.Linear(128,num_classes)","def __init__(self, num_heads: int, size: int, size_v: int, dropout: float = 0.1):\n super(ContMultiHeadedAttention, self).__init__()\n\n assert size % num_heads == 0\n\n self.head_size = head_size = size // num_heads\n self.model_size = size\n self.num_heads = num_heads\n\n self.k_layer = nn.Linear(size, num_heads * head_size)\n self.v_layer = nn.Linear(size_v, num_heads * head_size)\n self.q_layer = nn.Linear(size, num_heads * head_size)\n\n self.output_layer = nn.Linear(size, size_v)\n self.softmax = nn.Softmax(dim=-1)\n self.dropout = nn.Dropout(dropout)","def forward(self, x):\n sources = list()\n loc = list()\n conf = list()\n\n # apply vgg up to conv4_3 relu\n for k in range(23):\n x = self.vgg[k](x)\n\n s = self.L2Norm(x)\n sources.append(s)\n\n # apply vgg up to fc7\n for k in range(23, len(self.vgg)):\n x = self.vgg[k](x)\n sources.append(x)\n\n # apply extra layers and cache source layer outputs\n for k, v in enumerate(self.extras):\n x = F.relu(v(x), inplace=True)\n if k % 2 == 1:\n sources.append(x)\n\n # apply multibox head to source layers\n for (x, l, c) in zip(sources, self.loc, self.conf):\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\n\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\n if self.phase == \"test\":\n output = self.detect(\n loc.view(loc.size(0), -1, 4), # loc preds\n self.softmax(conf.view(conf.size(0), -1,\n self.num_classes)), # conf preds\n self.priors.type(type(x.data)) # default boxes\n )\n else:\n output = (\n loc.view(loc.size(0), -1, 4),\n conf.view(conf.size(0), -1, self.num_classes),\n self.priors\n )\n return output","def forward(self, enc_states, enc_len, dec_states):\n if self.precomputed_enc_h is None:\n\n self.precomputed_enc_h = self.mlp_enc(enc_states)\n self.mask = length_to_mask(\n enc_len, max_len=enc_states.size(1), device=enc_states.device\n )\n\n # multiply mask by 1/Ln for each row\n self.prev_attn = self.mask * (1 / enc_len.float()).unsqueeze(1)\n\n # compute location-aware features\n # [B, 1, L] -> [B, C, L]\n attn_conv = self.conv_loc(self.prev_attn.unsqueeze(1))\n # [B, C, L] -> [B, L, C] -> [B, L, F]\n attn_conv = self.mlp_loc(attn_conv.transpose(1, 2))\n\n dec_h = self.mlp_dec(dec_states.unsqueeze(1))\n attn = self.mlp_attn(\n torch.tanh(self.precomputed_enc_h + dec_h + attn_conv)\n ).squeeze(-1)\n\n # mask the padded frames\n attn = attn.masked_fill(self.mask == 0, -np.inf)\n attn = self.softmax(attn * self.scaling)\n\n # set prev_attn to current attn for the next timestep\n self.prev_attn = attn.detach()\n\n # compute context vectors\n # [B, 1, L] X [B, L, F]\n context = torch.bmm(attn.unsqueeze(1), enc_states).squeeze(1)\n context = self.mlp_out(context)\n\n return context, attn","def __init__(self,\n num_heads,\n seq_len,\n block=16,\n different_layout_per_head=False,\n num_local_blocks=4,\n num_global_blocks=1,\n attention='bidirectional',\n horizental_global_attention=False,\n num_differnt_global_patterns=1):\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\n\n if (seq_len % (block * num_local_blocks) != 0):\n raise ValueError(\n f'Sequence Length, {seq_len}, must be dividable by local window, {block*num_local_blocks}!'\n )\n self.num_local_blocks = num_local_blocks\n\n if (num_local_blocks % num_global_blocks != 0):\n raise ValueError(\n f'Number of blocks in a local window, {num_local_blocks}, must be dividable by number of global blocks, {num_global_blocks}!'\n )\n self.num_global_blocks = num_global_blocks\n\n if (attention != 'unidirectional' and attention != 'bidirectional'):\n raise NotImplementedError(\n 'only \\\"uni/bi-directional\\\" attentions are supported for now!')\n self.attention = attention\n\n if (attention != 'bidirectional' and horizental_global_attention):\n raise ValueError(\n 'only \\\"bi-directional\\\" attentions can support horizental global attention!'\n )\n self.horizental_global_attention = horizental_global_attention\n\n if (num_differnt_global_patterns > 1 and not different_layout_per_head):\n raise ValueError(\n f'Number of different layouts cannot be more than one when you have set a single layout for all heads! Set different_layout_per_head to True.'\n )\n if (num_differnt_global_patterns > (num_local_blocks // num_global_blocks)):\n raise ValueError(\n f'Number of layout versions (num_differnt_global_patterns), {num_differnt_global_patterns}, cannot be larger than number of local window blocks divided by number of global blocks, {num_local_blocks} / {num_global_blocks} = {num_local_blocks//num_global_blocks}!'\n )\n self.num_differnt_global_patterns = num_differnt_global_patterns\n self.make_layout()","def mt_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\n assert lambda_coeff >= 0\n if lambda_coeff == 0:\n return\n params = self.params\n name = 'model' if params.encoder_only else 'encoder'\n model = getattr(self, name)\n model.train()\n\n # _lang1 = _lang2 = 'en'\n # lang2_id = params.lang2id[_lang2]\n\n # generate batch\n\n (x_src,len_src),(x2, len2),(x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\n # convert fp16\n #assign lang ids\n\n\n lang0_id = params.lang2id[params.ft_lgs[0]]\n lang_src = x_src.clone().fill_(lang0_id)\n lang1_id = params.lang2id[params.ft_lgs[1]]\n langs = x2.clone().fill_(lang1_id)\n\n\n # target words to predict\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\n assert len(y) == (len2 - 1).sum().item()\n\n #\n len1 = x1_mask.sum(dim=1)\n x1 = x1.transpose(0, 1)\n img_loc = img_loc.transpose(0, 1)\n\n # cuda\n\n x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src)\n\n if params.mt_only_text:\n encoder_outputs = model('crossfwd', stream_='text', x=x_src, lengths=len_src, langs=lang_src, causal=False,refine_image=params.refine_image)\n len_all = len_src\n else:\n encoder_outputs = model('jointfwd', x=x_src, lengths=len_src, x_img=x1, lengths_img=len1, causal=False,\n langs=None,\n image_loc=img_loc, refine_image=params.refine_image)\n len_all = len_src + len1\n\n # encode source sentence\n # enc1 = model('jo', stream_='img', x=x1, lengths=len1, langs=None, causal=False, cross_modal=True,\n # image_loc=img_loc, refine_image=params.refine_image, refine_encoder=params.refine_encoder,\n # image_dist=None)\n enc1 = encoder_outputs.transpose(0, 1)\n\n\n # decode target sentence\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs, causal=True, src_enc=enc1,\n src_len=len_all)\n\n # loss\n # the last word not apply prediction logic\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\n self.stats[('IC-%s-%s' % (dataset, input_stream))].append(loss.item())\n loss = lambda_coeff * loss\n\n # optimize\n self.optimize(loss)\n\n # number of processed sentences / words\n self.n_sentences += params.batch_size\n self.stats['processed_s'] += len2.size(0)\n self.stats['processed_w'] += (len2 - 1).sum().item()","def __init__(self, idim, odim, args, ignore_id=-1):\n torch.nn.Module.__init__(self)\n if args.transformer_attn_dropout_rate is None:\n args.transformer_attn_dropout_rate = args.dropout_rate\n\n self.mtlalpha = args.mtlalpha\n if not hasattr(args, \"beta\"):\n self.beta = 0.0\n else:\n self.beta = args.beta\n self.num_steps = args.num_steps\n self.tau = args.error_temperature\n self.idx_blank = 0\n self.masktoken = odim - 1\n self.sos = odim - 1\n self.eos = odim - 1\n self.odim = odim\n self.ignore_id = ignore_id\n self.lsm_weight = args.lsm_weight\n self.reporter = None\n\n self.subsample = [1]\n self.encoder = Encoder(\n idim=idim,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.eunits,\n num_blocks=args.elayers,\n input_layer=args.transformer_input_layer,\n dropout_rate=args.dropout_rate,\n death_rate=args.edeath_rate,\n positional_dropout_rate=args.dropout_rate,\n attention_dropout_rate=args.transformer_attn_dropout_rate\n )\n self.enc_odim = args.adim\n\n # Define alignment embedding module.\n self.embed = torch.nn.Embedding(odim, self.enc_odim)\n self.decoder = Decoder(\n odim=odim,\n attention_dim=args.adim,\n attention_heads=args.aheads,\n linear_units=args.dunits,\n num_blocks=args.dlayers,\n dropout_rate=args.dropout_rate,\n death_rate=args.ddeath_rate,\n positional_dropout_rate=args.dropout_rate,\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\n input_layer=\"embed_linear\",\n pos_enc_class=ScaledPositionalEncoding\n )\n\n # self.verbose = args.verbose\n self.reset_parameters(args)\n self.adim = args.adim\n\n # CTC directly works on top of encoder outputs.\n self.ctc = CTC(odim, self.enc_odim, args.dropout_rate, ctc_type=args.ctc_type, reduce=True)\n self.rnnlm = None\n\n # Decoder masking params\n self.decoder_num_mask = args.decoder_num_mask\n self.decoder_mask_width = args.decoder_mask_width","def __init__(self, block, layers, num_classes=1000, width=64):\n\n torch.nn.Module.__init__(self) # Skip the parent constructor. This replaces it.\n self._norm_layer = torch.nn.BatchNorm2d\n self.inplanes = width\n self.dilation = 1\n self.groups = 1\n self.base_width = 64\n\n # The initial convolutional layer.\n self.conv1 = torch.nn.Conv2d(3, self.inplanes, kernel_size=7, stride=2, padding=3, bias=False)\n self.bn1 = self._norm_layer(self.inplanes)\n self.relu = torch.nn.ReLU(inplace=True)\n self.maxpool = torch.nn.MaxPool2d(kernel_size=3, stride=2, padding=1)\n\n # The subsequent blocks.\n self.layer1 = self._make_layer(block, width, layers[0])\n self.layer2 = self._make_layer(block, width*2, layers[1], stride=2, dilate=False)\n self.layer3 = self._make_layer(block, width*4, layers[2], stride=2, dilate=False)\n self.layer4 = self._make_layer(block, width*8, layers[3], stride=2, dilate=False)\n\n # The last layers.\n self.avgpool = torch.nn.AdaptiveAvgPool2d((1, 1))\n self.fc = torch.nn.Linear(width*8*block.expansion, num_classes)\n\n # Default init.\n for m in self.modules():\n if isinstance(m, torch.nn.Conv2d):\n torch.nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')\n elif isinstance(m, torch.nn.BatchNorm2d):\n torch.nn.init.constant_(m.weight, 1)\n torch.nn.init.constant_(m.bias, 0)","def __init__(self, id, node_type=NodeType.HIDDEN, activation=F.relu, layer_type=nn.Conv2d,\n conv_window_size=3, conv_stride=1, max_pool_size=2):\n\n super(ModuleNEATNode, self).__init__(id, node_type)\n\n batch_norm_chance = 0.65 # chance that a new node will start with batch norm\n use_batch_norm = random.random() < batch_norm_chance\n\n dropout_chance = 0.2 # chance that a new node will start with drop out\n use_dropout = random.random() < dropout_chance\n\n max_pool_chance = 0.3 # chance that a new node will start with drop out\n use_max_pool = random.random() < max_pool_chance\n\n self.activation = Mutagen(F.relu, F.leaky_relu, torch.sigmoid, F.relu6,\n discreet_value=activation, name=\"activation function\",\n mutation_chance=0.15) # TODO try add in Selu, Elu\n\n conv_out_features = 25 + random.randint(0, 25)\n linear_out_features = 100 + random.randint(0, 100)\n\n linear_submutagens = \\\n {\n \"regularisation\": Mutagen(None, nn.BatchNorm1d,\n discreet_value=nn.BatchNorm1d if use_batch_norm else None,\n mutation_chance=0.15),\n\n \"dropout\": Mutagen(None, nn.Dropout, discreet_value=nn.Dropout if use_dropout else None, sub_mutagens=\n {\n nn.Dropout: {\n \"dropout_factor\": Mutagen(value_type=ValueType.CONTINUOUS, current_value=0.15, start_range=0,\n end_range=0.75)}\n }, mutation_chance=0.08),\n\n \"out_features\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=linear_out_features,\n start_range=10,\n end_range=1024, name=\"num out features\", mutation_chance=0.22,\n distance_weighting=Props.LAYER_SIZE_COEFFICIENT if Config.allow_attribute_distance else 0)\n }\n\n conv_submutagens = {\n \"conv_window_size\": Mutagen(3, 5, 7, discreet_value=conv_window_size, mutation_chance=0.13),\n\n \"conv_stride\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=conv_stride, start_range=1,\n end_range=5),\n\n \"reduction\": Mutagen(None, nn.MaxPool2d, discreet_value=nn.MaxPool2d if use_max_pool else None,\n sub_mutagens=\n {\n nn.MaxPool2d: {\"pool_size\": Mutagen(\n value_type=ValueType.WHOLE_NUMBERS, current_value=max_pool_size, start_range=2,\n end_range=5)}\n }, mutation_chance=0.15),\n\n \"regularisation\": Mutagen(None, nn.BatchNorm2d, discreet_value=nn.BatchNorm2d if use_batch_norm else None,\n mutation_chance=0.15),\n\n \"dropout\": Mutagen(None, nn.Dropout2d, discreet_value=nn.Dropout2d if use_dropout else None, sub_mutagens=\n {\n nn.Dropout2d: {\n \"dropout_factor\": Mutagen(value_type=ValueType.CONTINUOUS, current_value=0.1,\n start_range=0, end_range=0.75)}\n }, mutation_chance=0.08),\n\n \"out_features\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=conv_out_features, start_range=1,\n end_range=100, name=\"num out features\", mutation_chance=0.22,\n distance_weighting=Props.LAYER_SIZE_COEFFICIENT if Config.allow_attribute_distance else 0)\n }\n\n if use_linears and not use_convs:\n self.layer_type = Mutagen(nn.Linear, discreet_value=nn.Linear,\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\n sub_mutagens={nn.Linear: linear_submutagens}\n )\n if use_convs and not use_linears:\n self.layer_type = Mutagen(nn.Conv2d, discreet_value=nn.Conv2d,\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\n sub_mutagens={nn.Conv2d: conv_submutagens})\n if use_convs and use_linears:\n self.layer_type = Mutagen(nn.Conv2d, nn.Linear, discreet_value=layer_type,\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\n sub_mutagens={\n nn.Conv2d: conv_submutagens,\n nn.Linear: linear_submutagens\n }, name=\"deep layer type\", mutation_chance=0.08)","def __init__(self, fsmt_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([fsmt_layer.self_attn.q_proj.weight, fsmt_layer.self_attn.k_proj.weight, fsmt_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([fsmt_layer.self_attn.q_proj.bias, fsmt_layer.self_attn.k_proj.bias, fsmt_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = fsmt_layer.self_attn.out_proj.weight\n self.out_proj_bias = fsmt_layer.self_attn.out_proj.bias\n self.linear1_weight = fsmt_layer.fc1.weight\n self.linear1_bias = fsmt_layer.fc1.bias\n self.linear2_weight = fsmt_layer.fc2.weight\n self.linear2_bias = fsmt_layer.fc2.bias\n self.norm1_eps = fsmt_layer.self_attn_layer_norm.eps\n self.norm1_weight = fsmt_layer.self_attn_layer_norm.weight\n self.norm1_bias = fsmt_layer.self_attn_layer_norm.bias\n self.norm2_eps = fsmt_layer.final_layer_norm.eps\n self.norm2_weight = fsmt_layer.final_layer_norm.weight\n self.norm2_bias = fsmt_layer.final_layer_norm.bias\n self.num_heads = fsmt_layer.self_attn.num_heads\n self.embed_dim = fsmt_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.validate_bettertransformer()","def preprocess_module(mod):\n\n def alter_conv(attrs, inputs, tinfos, out_type):\n new_attrs = dict(attrs)\n data_info = tinfos[0]\n weight_info = tinfos[1]\n (desired_data_layout, desired_kernel_layout) = (\"NCHW\", \"OIHW\")\n new_attrs[\"data_layout\"] = desired_data_layout\n new_attrs[\"kernel_layout\"] = desired_kernel_layout\n\n if is_depthwise_conv2d(\n data_info.shape,\n attrs[\"data_layout\"],\n weight_info.shape,\n attrs[\"kernel_layout\"],\n attrs[\"groups\"],\n ):\n dkl = desired_kernel_layout\n new_attrs[\"kernel_layout\"] = dkl[1] + dkl[0] + dkl[2] + dkl[3]\n return relay.nn.conv2d(*inputs, **new_attrs)\n\n with OpAttrContext(\"nn.conv2d\", \"FTVMAlterOpLayout\", alter_conv):\n seq = tvm.transform.Sequential(\n [\n transform.ConvertLayout({\"nn.conv2d\": [\"NCHW\", \"OIHW\"]}),\n transform.ConvertLayout({\"nn.conv2d_transpose\": [\"NCHW\", \"OIHW\"]}),\n transform.AlterOpLayout(),\n transform.FoldConstant(),\n ]\n )\n with tvm.transform.PassContext(opt_level=3):\n preprocessed_mod = seq(mod)\n return preprocessed_mod","def __init__(self):\n super(Backbone, self).__init__()\n\n # input size: (128, 282, 282)\n # Block 1:\n # relu + 4 conv + bn\n self.conv1 = torch.nn.Conv2d(128, 64, kernel_size=3, stride=1, padding=0)\n self.conv2 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\n self.conv3 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\n self.conv4 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\n\n self.bn1 = torch.nn.BatchNorm2d(64)\n self.bn2 = torch.nn.BatchNorm2d(64)\n self.bn3 = torch.nn.BatchNorm2d(64)\n self.bn4 = torch.nn.BatchNorm2d(64)\n\n # Block 2:\n # relu + 6 conv + stride 2 + bn\n self.conv5 = torch.nn.Conv2d(64, 128, kernel_size=3, stride=2, padding=0)\n self.conv6 = torch.nn.Conv2d(128, 128, kernel_size=3, stride=2, padding=0)\n self.conv7 = torch.nn.Conv2d(128, 64, kernel_size=3, stride=1, padding=0)\n self.conv8 = torch.nn.Conv2d(64, 32, kernel_size=3, stride=1, padding=0)\n self.conv9 = torch.nn.Conv2d(32, 16, kernel_size=3, stride=1, padding=0)\n self.conv10 = torch.nn.Conv2d(16, 8, kernel_size=3, stride=1, padding=0)\n\n self.bn5 = torch.nn.BatchNorm2d(128)\n self.bn6 = torch.nn.BatchNorm2d(128)\n self.bn7 = torch.nn.BatchNorm2d(64)\n self.bn8 = torch.nn.BatchNorm2d(32)\n self.bn9 = torch.nn.BatchNorm2d(16)\n self.bn10 = torch.nn.BatchNorm2d(8)\n\n # Block 3:\n # 2 fully connected with drop out.\n\n self.fc1 = torch.nn.Linear( 8 * 59 * 59, 32)\n self.fc1_bn = torch.nn.BatchNorm1d(32)\n self.fc_out = torch.nn.Linear(32, 3)","def __init__(self, vit_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([vit_layer.attention.attention.query.weight, vit_layer.attention.attention.key.weight, vit_layer.attention.attention.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([vit_layer.attention.attention.query.bias, vit_layer.attention.attention.key.bias, vit_layer.attention.attention.value.bias]))\n self.out_proj_weight = vit_layer.attention.output.dense.weight\n self.out_proj_bias = vit_layer.attention.output.dense.bias\n self.linear1_weight = vit_layer.intermediate.dense.weight\n self.linear1_bias = vit_layer.intermediate.dense.bias\n self.linear2_weight = vit_layer.output.dense.weight\n self.linear2_bias = vit_layer.output.dense.bias\n self.norm1_eps = vit_layer.layernorm_before.eps\n self.norm1_weight = vit_layer.layernorm_before.weight\n self.norm1_bias = vit_layer.layernorm_before.bias\n self.norm2_eps = vit_layer.layernorm_after.eps\n self.norm2_weight = vit_layer.layernorm_after.weight\n self.norm2_bias = vit_layer.layernorm_after.bias\n self.num_heads = vit_layer.attention.attention.num_attention_heads\n self.embed_dim = int(vit_layer.attention.attention.attention_head_size * self.num_heads)\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def _Attention(self, name, is_causal=True):\n p = self.params\n tr_atten_p = TransformerAttentionLayer.Params().Set(\n name='transformer_atten',\n input_dim=p.model_dim,\n hidden_dim=p.attention_hidden_dim or p.model_dim,\n is_masked=is_causal,\n num_heads=p.num_heads,\n residual_dropout_prob=p.residual_dropout_prob,\n atten_dropout_prob=p.atten_dropout_prob,\n fprop_dtype=p.fprop_dtype,\n add_unnormalized_input=p.selfatten_add_unnormalized_input,\n )\n tr_atten_p.atten_tpl.use_bias = p.use_bias\n tr_atten_p.atten_tpl.enable_value_proj = p.selfatten_enable_value_proj\n tr_atten_p.atten_tpl.enable_query_scale = p.enable_query_scale\n tr_atten_p.atten_tpl.enable_per_dim_scale = p.enable_per_dim_scale\n tr_atten_p.atten_tpl.device_mesh = p.device_mesh\n tr_atten_p.atten_tpl.weight_split_dims_mapping = (\n p.weight_split_dims_mapping.dnh)\n tr_atten_p.atten_tpl.activation_split_dims_mapping.blnh = (\n p.activation_split_dims_mapping.blnh)\n tr_atten_p.atten_tpl.activation_split_dims_mapping.bld = (\n p.activation_split_dims_mapping.bld)\n if p.deterministic_dropout:\n tr_atten_p.dropout_tpl = layers.DeterministicDropoutLayer.Params()\n tr_atten_p.atten_p.dropout_tpl = layers.DeterministicDropoutLayer.Params()\n\n return self._Graph(\n name,\n ['i'], # input NestedMap with {vec, paddings}\n ['o'], # output NestedMap with {vec, paddings}\n ('i.vec->split_i',\n self.MeshSplit('input_split', p.activation_split_dims_mapping.bld)),\n ('split_i,split_i,i.paddings->o.vec,unused_prob', tr_atten_p),\n ('i.paddings->o.paddings', self._Id('id')))","def __init__(self, d_model=256, n_levels=4, n_heads=8, n_points=4):\n super().__init__()\n if d_model % n_heads != 0:\n raise ValueError(\"d_model must be divisible by n_heads, but got {} and {}\".format(d_model, n_heads))\n _d_per_head = d_model // n_heads\n # you'd better set _d_per_head to a power of 2 which is more efficient in our CUDA implementation\n if not _is_power_of_2(_d_per_head):\n warnings.warn(\n \"You'd better set d_model in MSDeformAttn to make the dimension of each attention head a power of 2 \"\n \"which is more efficient in our CUDA implementation.\"\n )\n self.im2col_step = 64\n ###\n self.d_model = d_model\n self.n_levels = n_levels\n self.n_heads = n_heads\n self.n_points = n_points\n self.sampling_offsets = nn.Dense(d_model, n_heads * n_levels * n_points * 2)\n self.attention_weights = nn.Dense(d_model, n_heads * n_levels * n_points)\n self.value_proj = nn.Dense(d_model, d_model)\n self.output_proj = nn.Dense(d_model, d_model)","def identity_block(inputs, kernel_size, filters, stage, block,\n training):\n nb_filter1, nb_filter2, nb_filter3 = filters\n\n scope_name = 'conv'+str(stage)+block+'_branch'\n scope_name = 'bn'+str(stage)+block+'_branch'\n\n x = slim.conv2d(inputs,nb_filter1,[1,1],stride=[1,1],padding='VALID',activation_fn=None,\n trainable=training,scope=scope_name+'2a')\n x = slim.batch_norm(x,scope=scope_name+'2a',is_training=training)\n x = tf.nn.relu(x)\n\n\n x = slim.conv2d(x,nb_filter2,[kernel_size,kernel_size],stride=[1,1],padding='SAME',\n activation_fn=None,trainable=training,scope=scope_name+'2b')\n x = slim.batch_norm(x,scope=scope_name+'2b',is_training=training)\n x = tf.nn.relu(x)\n\n\n x = slim.conv2d(x,nb_filter3,[1,1],stride=[1,1],padding='VALID',\n activation_fn=None,trainable=training,scope=scope_name+'2c')\n x = slim.batch_norm(x,scope=scope_name+'2c',is_training=training)\n\n \n x = tf.add(x,inputs)\n x = tf.nn.relu(x,name='res'+str(stage)+block+\"_out\")\n return x","def forward(self, x):\n sources = list()\n loc = list()\n conf = list()\n\n # apply vgg up to conv4_3 relu\n #print('Reached start of vgg')\n for k in self.vgg._modules.keys():\n if int(k) < 23:\n #print('Reached ' + k + ' ', x.size())\n x = self.vgg._modules[k].cuda()(x)\n #print('Reached L2Norm')\n s = self.L2Norm(x)\n sources.append(s)\n\n #print('Reached after L2Norm')\n # apply vgg up to fc7\n for k in self.vgg._modules.keys():\n if int(k) >= 23:\n #print('Reached ' + k + ' ', x.size())\n x = self.vgg._modules[k].cuda()(x)\n sources.append(x)\n #print('Reached end of VGG')\n\n # apply extra layers and cache source layer outputs\n for k, v in enumerate(self.extras):\n x = F.relu(v(x), inplace=True)\n if k % 2 == 1:\n sources.append(x)\n\n # apply multibox head to source layers\n for (x, l, c) in zip(sources, self.loc, self.conf):\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\n\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\n\n if self.phase == \"test\":\n output = self.detect(\n loc.view(loc.size(0), -1, 4), # loc preds\n self.softmax(conf.view(-1, self.num_classes)), # conf preds\n self.priors # default boxes\n )\n else:\n output = (\n loc.view(loc.size(0), -1, 4),\n conf.view(conf.size(0), -1, self.num_classes),\n self.priors\n )\n return output","def __init__(\n self,\n encoder,\n num_classes=20,\n shape_size=16,\n mask_conv=True,\n min_confidence=0.2,\n min_remain=0.5,\n mask_threshold=0.4,\n ):\n super(PaPs, self).__init__()\n self.encoder = encoder\n self.shape_size = shape_size\n self.num_classes = num_classes\n self.min_scale = 1 / shape_size\n self.register_buffer(\"min_confidence\", torch.tensor([min_confidence]))\n self.min_remain = min_remain\n self.mask_threshold = mask_threshold\n self.center_extractor = CenterExtractor()\n\n enc_dim = encoder.enc_dim\n stack_dim = encoder.stack_dim\n self.heatmap_conv = nn.Sequential(\n ConvLayer(nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\n padding_mode=\"reflect\"),\n nn.Sigmoid(),\n )\n\n self.saliency_conv = ConvLayer(\n nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\n padding_mode=\"reflect\"\n )\n\n self.shape_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, shape_size ** 2),\n )\n\n self.size_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, stack_dim // 4),\n nn.BatchNorm1d(stack_dim // 4),\n nn.ReLU(),\n nn.Linear(stack_dim // 4, 2),\n nn.Softplus(),\n )\n\n self.class_mlp = nn.Sequential(\n nn.Linear(stack_dim, stack_dim // 2),\n nn.BatchNorm1d(stack_dim // 2),\n nn.ReLU(),\n nn.Linear(stack_dim // 2, stack_dim // 4),\n nn.Linear(stack_dim // 4, num_classes),\n )\n\n if mask_conv:\n self.mask_cnn = nn.Sequential(\n nn.Conv2d(in_channels=1, out_channels=16, kernel_size=3, padding=1),\n nn.GroupNorm(num_channels=16, num_groups=1),\n nn.ReLU(),\n nn.Conv2d(in_channels=16, out_channels=16, kernel_size=3, padding=1),\n nn.ReLU(),\n nn.Conv2d(in_channels=16, out_channels=1, kernel_size=3, padding=1),\n )\n else:\n self.mask_cnn = None","def __init__(self, args):\n \n super(MicroNet, self).__init__()\n self.conv1 = nn.Conv2d(3, 1, kernel_size=1)\n self.conv2 = nn.Conv2d(1, 29, kernel_size=5)\n self.maxpool2 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\n self.conv3 = nn.Conv2d(29, 59, kernel_size=3)\n self.maxpool3 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\n self.conv4 = nn.Conv2d(59, 74, kernel_size=3)\n self.maxpool4 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\n self.conv2_drop = nn.Dropout2d()\n self.conv3_drop = nn.Dropout2d()\n self.fc1 = nn.Linear(1184, 300)\n self.fc2 = nn.Linear(300, args.num_classes)\n self.conv0_bn = nn.BatchNorm2d(3)\n self.conv1_bn = nn.BatchNorm2d(1)\n self.conv2_bn = nn.BatchNorm2d(29)\n self.conv3_bn = nn.BatchNorm2d(59)\n self.conv4_bn = nn.BatchNorm2d(74)\n self.dense1_bn = nn.BatchNorm1d(300)","def td_identity_block(input_tensor, kernel_size, filters, stage, block, use_conv_bias=True,\n weight_regularizer=None, bias_regularizer=None, bn_training=False, separate_scale=False):\n filters1, filters2, filters3 = filters\n bn_axis = 3\n conv_name_base = 'res' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n scale_name_base = 'scale' + str(stage) + block + '_branch'\n eps = 1e-5\n\n x = TimeDistributed(Conv2D(filters1, (1, 1), use_bias=use_conv_bias,\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer,\n ),\n name=conv_name_base + '2a')(input_tensor)\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\n name=bn_name_base + '2a',)(x, training=bn_training)\n if separate_scale:\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2a')(x, training=bn_training)\n x = Activation('relu')(x)\n\n x = TimeDistributed(Conv2D(filters2, kernel_size, padding='same', use_bias=use_conv_bias,\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer\n ),\n name=conv_name_base + '2b')(x)\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\n name=bn_name_base + '2b')(x, training=bn_training)\n if separate_scale:\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2b')(x, training=bn_training)\n x = Activation('relu')(x)\n\n x = TimeDistributed(Conv2D(filters3, (1, 1), use_bias=use_conv_bias,\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer\n ),\n name=conv_name_base + '2c')(x)\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\n name=bn_name_base + '2c')(x, training=bn_training)\n if separate_scale:\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2c')(x, training=bn_training)\n\n x = layers.add([x, input_tensor])\n x = TimeDistributed(Activation('relu'))(x)\n return x","def TextEncBlock(L,d,scope=\"TextEncBlock\"):\n e = L.shape.as_list()[2]\n\n with tf.variable_scope(scope):\n conv_params = {\"filters\":2*d,\"kernel_size\":1,\"padding\":\"same\"} \n with tf.variable_scope(\"C_block1\"):\n L1 = tf.nn.relu(conv1d(inputs=L,**conv_params)) # relu(conv)\n L2 = conv1d(inputs=L1,**conv_params) # conv\n # hc_dilation_blocks \n L3_HC1_1 = hc_dilation_block(L2,local_scope=\"HC_dilation_block1\")\n L3_HC1_2 = hc_dilation_block(L3_HC1_1,local_scope=\"HC_dilation_block2\")\n # highway_conv_block2\n with tf.variable_scope(\"HC_block2\"):\n L4_HC2_1 = highway_activation_conv(L3_HC1_2,kernel_size=3,scope=\"HC1\")\n L4_HC2_2 = highway_activation_conv(L4_HC2_1,kernel_size=3,scope=\"HC2\")\n # highway_conv_block3\n with tf.variable_scope(\"HC_block3\"):\n L5_HC3_1 = highway_activation_conv(L4_HC2_2,kernel_size=1,scope=\"HC1\")\n KV = highway_activation_conv(L5_HC3_1,kernel_size=1,scope=\"HC2\")\n\n return KV[:,:,:d], KV[:,:,d:] # key, value","def __init__(self, bart_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([bart_layer.self_attn.q_proj.weight, bart_layer.self_attn.k_proj.weight, bart_layer.self_attn.v_proj.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([bart_layer.self_attn.q_proj.bias, bart_layer.self_attn.k_proj.bias, bart_layer.self_attn.v_proj.bias]))\n self.out_proj_weight = bart_layer.self_attn.out_proj.weight\n self.out_proj_bias = bart_layer.self_attn.out_proj.bias\n self.linear1_weight = bart_layer.fc1.weight\n self.linear1_bias = bart_layer.fc1.bias\n self.linear2_weight = bart_layer.fc2.weight\n self.linear2_bias = bart_layer.fc2.bias\n self.norm1_eps = bart_layer.self_attn_layer_norm.eps\n self.norm1_weight = bart_layer.self_attn_layer_norm.weight\n self.norm1_bias = bart_layer.self_attn_layer_norm.bias\n self.norm2_eps = bart_layer.final_layer_norm.eps\n self.norm2_weight = bart_layer.final_layer_norm.weight\n self.norm2_bias = bart_layer.final_layer_norm.bias\n self.num_heads = bart_layer.self_attn.num_heads\n self.embed_dim = bart_layer.self_attn.embed_dim\n self.is_last_layer = False\n self.validate_bettertransformer()","def __init__(self, vilt_layer, config):\n super().__init__(config)\n self.in_proj_weight = nn.Parameter(torch.cat([vilt_layer.attention.attention.query.weight, vilt_layer.attention.attention.key.weight, vilt_layer.attention.attention.value.weight]))\n self.in_proj_bias = nn.Parameter(torch.cat([vilt_layer.attention.attention.query.bias, vilt_layer.attention.attention.key.bias, vilt_layer.attention.attention.value.bias]))\n self.out_proj_weight = vilt_layer.attention.output.dense.weight\n self.out_proj_bias = vilt_layer.attention.output.dense.bias\n self.linear1_weight = vilt_layer.intermediate.dense.weight\n self.linear1_bias = vilt_layer.intermediate.dense.bias\n self.linear2_weight = vilt_layer.output.dense.weight\n self.linear2_bias = vilt_layer.output.dense.bias\n self.norm1_eps = vilt_layer.layernorm_before.eps\n self.norm1_weight = vilt_layer.layernorm_before.weight\n self.norm1_bias = vilt_layer.layernorm_before.bias\n self.norm2_eps = vilt_layer.layernorm_after.eps\n self.norm2_weight = vilt_layer.layernorm_after.weight\n self.norm2_bias = vilt_layer.layernorm_after.bias\n self.num_heads = vilt_layer.attention.attention.num_attention_heads\n self.embed_dim = int(vilt_layer.attention.attention.attention_head_size * self.num_heads)\n self.is_last_layer = False\n self.norm_first = True\n self.validate_bettertransformer()","def __init__(self, dm, h, hidden, drop_rate=0.1):\n super().__init__()\n\n self.mha1 = MultiHeadAttention(dm, h)\n self.mha2 = MultiHeadAttention(dm, h)\n\n self.ffn = point_wise_feed_forward_network(dm, hidden)\n\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n self.layernorm3 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\n\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\n self.dropout3 = tf.keras.layers.Dropout(drop_rate)","def __init__(self):\n super(Generator, self).__init__()\n\n # output feature map will have the size of 8x8\n self.first_deconv = TransitionUp(in_channels=3, out_channels=256, stride=2, padding=1, kernel_size=4)\n self.first_batch_norm = nn.BatchNorm2d(256)\n\n # output feature map will have the size of 16x16\n self.second_deconv = TransitionUp(in_channels=256, out_channels=128, stride=2, padding=1, kernel_size=4)\n self.second_batch_norm = nn.BatchNorm2d(128)\n\n # output feature map will have the size of 32x32\n self.third_deconv = TransitionUp(in_channels=128, out_channels=32, stride=2, padding=1, kernel_size=4)\n self.third_batch_norm = nn.BatchNorm2d(32)\n\n # output feature map will have the size of 64x64\n self.fourth_deconv = TransitionUp(in_channels=32, out_channels=16, stride=2, padding=1, kernel_size=4)\n self.fourth_batch_norm = nn.BatchNorm2d(16)\n\n # output feature map will have the size of 128x128\n self.fifth_deconv = TransitionUp(in_channels=16, out_channels=16, stride=1, padding=1, kernel_size=4)\n self.fifth_batch_norm = nn.BatchNorm2d(16)\n\n # output feature map will have the size of 128x128\n self.sixth_deconv = TransitionUp(in_channels=16, out_channels=16, stride=1, padding=1, kernel_size=4)\n self.sixth_batch_norm = nn.BatchNorm2d(16)\n\n # output feature map will have the size of 128x128\n self.seventh_deconv = TransitionUp(in_channels=16, out_channels=3, stride=1, padding=1, kernel_size=4)"],"string":"[\n \"def build_model_multihead_attention_multiscaleCNN4_covermore(self,\\n dim_attention,headnum,\\n embedding_vec,\\n load_weights=False, weight_dir=None,\\n nb_filters=32,filters_length1=1,\\n filters_length2=5,\\n filters_length3=10,\\n pooling_size=3,\\n drop_input=0,\\n drop_cnn=0.2,\\n drop_flat=0,\\n W1_regularizer=0.005,\\n W2_regularizer=0.005,\\n Att_regularizer_weight=0.0005,\\n BatchNorm=False,\\n fc_dim = 50,\\n fcnum=0,\\n posembed=False,\\n pos_dmodel=40,\\n pos_nwaves = 20,\\n posmod = 'concat',\\n regularfun=1,\\n huber_delta=1,\\n activation='gelu',\\n activationlast='gelu',\\n add_avgpooling = False,\\n poolingmod=1,\\n normalizeatt=False,\\n regressionmodel=False,\\n attmod = \\\"softmax\\\",\\n sharp_beta=1,\\n lr = 0.001 \\n ):\\n ###print('Advanced Masking')\\n def mask_func(x):\\n return x[0] * x[1]\\n \\n ###print(posembed)\\n ###print(posmod)\\n input = Input(shape=(self.max_len,), dtype='int8')\\n input_mask = Input(shape=([int(self.max_len/pooling_size), 1]), dtype='float32')\\n embedding_layer = Embedding(len(embedding_vec), len(embedding_vec[0]), weights=[embedding_vec],\\n input_length=self.max_len,\\n trainable=False)\\n embedding_output = Dropout(drop_input)(embedding_layer(input)) #layer2\\n if 'gelu' in activation:\\n activationfun=gelu\\n else:\\n activationfun = 'relu'\\n \\n if 'gelu' in activationlast:\\n activationlastfun = gelu\\n else:\\n activationlastfun='relu'\\n \\n ###print(activationfun)\\n ###print(activationlastfun)\\n with tf.name_scope('first_cnn'):\\n first_cnn = Convolution1D(nb_filters, filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN1')(embedding_output) #layer3\\n first_cnn2 = Convolution1D(int(nb_filters/2), filters_length1, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(first_cnn) #layer5\\n second_cnn = Convolution1D(nb_filters, filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN2')(embedding_output) #layer4\\n second_cnn2 = Convolution1D(int(nb_filters/2), filters_length2, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(second_cnn)\\n third_cnn = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationfun, use_bias=False,name='CNN3')(embedding_output)\\n \\n third_cnn2 = Convolution1D(int(nb_filters/2), filters_length3, #kernel_regularizer=regularizers.l2(0.0001),\\n border_mode='same', activation=activationlastfun, use_bias=False)(third_cnn)\\n if BatchNorm:\\n first_cnn2 = BatchNormalization()(first_cnn2)\\n second_cnn2 = BatchNormalization()(second_cnn2)\\n third_cnn2 = BatchNormalization()(third_cnn2)\\n \\n if not add_avgpooling:\\n if poolingmod == 1:\\n pooling_layer = MaxPooling1D(pool_length=pooling_size, stride=pooling_size)\\n else:\\n pooling_layer = AveragePooling1D(pool_length=pooling_size, stride=pooling_size)\\n \\n cnn_output1 = Dropout(drop_cnn)(pooling_layer(first_cnn2))\\n cnn_output2 = Dropout(drop_cnn)(pooling_layer(second_cnn2))\\n cnn_output3 = Dropout(drop_cnn)(pooling_layer(third_cnn2))\\n else:\\n first_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\\n first_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(first_cnn2)\\n cnn_output1 = Dropout(drop_cnn)(concatenate([first_cnn2_max,first_cnn2_avg],axis=-1))\\n second_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\\n second_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(second_cnn2)\\n cnn_output2 = Dropout(drop_cnn)(concatenate([second_cnn2_max,second_cnn2_avg],axis=-1))\\n third_cnn2_max=MaxPooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\\n third_cnn2_avg=AveragePooling1D(pool_length=pooling_size, stride=pooling_size)(third_cnn2)\\n cnn_output3 = Dropout(drop_cnn)(concatenate([third_cnn2_max,third_cnn2_avg],axis=-1))\\n \\n \\n \\n if posembed:\\n ##print(posmod)\\n from position_embedding import PositionEmbedding\\n if posmod == 'concat':\\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb1')(cnn_output1)\\n cnn_output1 = concatenate([cnn_output1, pos_emb1], axis=-1)\\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb2')(cnn_output2)\\n cnn_output2 = concatenate([cnn_output2, pos_emb2], axis=-1)\\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=pos_nwaves, d_model=pos_dmodel,name='pos_emb3')(cnn_output3)\\n cnn_output3 = concatenate([cnn_output3, pos_emb3], axis=-1)\\n else:\\n ##print(\\\"yes add posmod\\\")\\n pos_emb1 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output1)[-1]/2), d_model=int_shape(cnn_output1)[-1],name='pos_emb1')(cnn_output1)\\n cnn_output1 = Add()([cnn_output1, pos_emb1])\\n pos_emb2 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output2)[-1]/2), d_model=int_shape(cnn_output2)[-1],name='pos_emb2')(cnn_output2)\\n cnn_output2 = Add()([cnn_output2, pos_emb2])\\n pos_emb3 = PositionEmbedding(max_time=int(self.max_len/pooling_size), n_waves=int(int_shape(cnn_output3)[-1]/2), d_model=int_shape(cnn_output3)[-1],name='pos_emb3')(cnn_output3)\\n cnn_output3 = Add()([cnn_output3, pos_emb3])\\n \\n mask_input1 = []\\n mask_input1.append(cnn_output1)\\n mask_input1.append(input_mask)\\n cnn_mask_output1 = Lambda(mask_func)(mask_input1)\\n del mask_input1\\n mask_input2 = []\\n mask_input2.append(cnn_output2)\\n mask_input2.append(input_mask)\\n cnn_mask_output2 = Lambda(mask_func)(mask_input2)\\n del mask_input2\\n mask_input3 = []\\n mask_input3.append(cnn_output3)\\n mask_input3.append(input_mask)\\n cnn_mask_output3 = Lambda(mask_func)(mask_input3)\\n del mask_input3\\n \\n if regularfun==1:\\n regularizerfunction_W1 = regularizers.l1(W1_regularizer)\\n regularizerfunction_W2 = regularizers.l1(W2_regularizer)\\n elif regularfun==2:\\n regularizerfunction_W1 = regularizers.l2(W1_regularizer)\\n regularizerfunction_W2 = regularizers.l2(W2_regularizer)\\n elif regularfun ==3:\\n regularizerfunction_W1 = smoothL1(W1_regularizer,huber_delta)\\n regularizerfunction_W2 = smoothL1(W2_regularizer,huber_delta)\\n \\n with tf.name_scope('multihead_attention'):\\n att1,att1_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att1\\\")(concatenate([cnn_mask_output1, input_mask]))#-5 layer\\n \\n att2,att2_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att2\\\")(concatenate([cnn_mask_output2, input_mask])) #-4 layer\\n \\n att3,att3_A = Attention_mask(hidden=cnn_output1.get_shape()[-1].value, da=dim_attention, r=headnum, init='glorot_uniform', activation='tanh',\\n W1_regularizer=regularizerfunction_W1,\\n W2_regularizer=regularizerfunction_W2,\\n W1_constraint=None, W2_constraint=None, return_attention=True,\\n attention_regularizer_weight=Att_regularizer_weight,normalize=normalizeatt,attmod=attmod,sharp_beta=sharp_beta,name=\\\"att3\\\")(concatenate([cnn_mask_output3, input_mask])) #-3 layer\\n \\n if BatchNorm:\\n att1 = BatchNormalization()(att1)\\n att2 = BatchNormalization()(att2)\\n att3 = BatchNormalization()(att3)\\n \\n \\n output = Dropout(drop_flat)(Flatten()(concatenate([att1,att2,att3]))) #-2 layer\\n \\n fc = output\\n for _ in range(fcnum):\\n fc = Dense(fc_dim,activation='relu')(fc)\\n fc = Dropout(drop_flat)(fc)\\n \\n with tf.name_scope(''):\\n if regressionmodel:\\n preds = Dense(self.nb_classes,activation='softmax')(fc) #-1 layer\\n else:\\n preds = Dense(self.nb_classes,activation='sigmoid')(fc) #-1 layer\\n \\n self.model = Model(inputs=[input,input_mask], outputs=preds)\\n from keras import optimizers\\n # optim = optimizers.RMSprop()\\n optim = optimizers.Adam(lr=lr, decay=5e-5) #The paper uses a decay rate alpha = alpha/sqrt(t) updted each epoch (t) for the logistic regression demonstration.\\n #optim = optimizers.nadam()\\n #optim = RAdam()\\n if regressionmodel:\\n self.model.compile(loss='kld',optimizer=optim,metrics=['acc'])\\n else:\\n self.model.compile(loss='binary_crossentropy',optimizer=optim,metrics=['binary_accuracy','categorical_accuracy'])\\n \\n \\n \\n if load_weights:\\n self.model.load_weights(weight_dir)\\n \\n self.is_built = True\\n self.bn = False\\n self.model.summary()\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n CustomTransformerBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n backward_compatible=self.backward_compatible,\\n float_type=self.float_type,\\n name=(\\\"layer_%d\\\" % i)))\\n super(CustomTransformer, self).build(unused_input_shapes)\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\r\\n super(EncoderBlock, self).__init__()\\r\\n self.mha = MultiHeadAttention(dm, h)\\r\\n self.dense_hidden = tf.keras.layers.Dense(units=hidden,\\r\\n activation='relu')\\r\\n self.dense_output = tf.keras.layers.Dense(units=dm)\\r\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\r\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\r\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\",\n \"def mtf_transformer2_base():\\n hparams = common_hparams.basic_params1()\\n\\n hparams.add_hparam(\\\"d_model\\\", 1024)\\n hparams.batch_size = 4\\n hparams.max_length = 1024\\n hparams.label_smoothing = 0.0\\n # a small positive value - this seems important for stability when training\\n # with bfloat16 activations.\\n hparams.add_hparam(\\\"z_loss\\\", 1e-4)\\n\\n # hparams applying to both encoder and decoder layer stacks.\\n hparams.add_hparam(\\\"d_ff\\\", 2048)\\n hparams.add_hparam(\\\"d_kv\\\", 128)\\n hparams.add_hparam(\\\"attention_dropout\\\", 0.0)\\n hparams.add_hparam(\\\"relu_dropout\\\", 0.0)\\n hparams.del_hparam(\\\"num_heads\\\")\\n hparams.del_hparam(\\\"num_hidden_layers\\\")\\n hparams.layer_prepostprocess_dropout = 0.0\\n hparams.add_hparam(\\\"extra_logit\\\", False)\\n # number of experts for moe_1d\\n hparams.moe_num_experts = 32\\n # number of experts for moe_2d = moe_expert_x * moe_expert_y\\n hparams.add_hparam(\\\"moe_expert_x\\\", 8)\\n hparams.add_hparam(\\\"moe_expert_y\\\", 4)\\n hparams.add_hparam(\\\"moe_hidden_size\\\", 32768)\\n\\n # round up vocab sizes to be a multiple of this value\\n hparams.vocab_divisor = 128\\n\\n hparams.optimizer = \\\"Adafactor\\\"\\n hparams.learning_rate_schedule = \\\"rsqrt_decay*linear_decay\\\"\\n hparams.learning_rate_warmup_steps = 10000\\n hparams.add_hparam(\\\"master_dtype\\\", \\\"bfloat16\\\")\\n hparams.add_hparam(\\\"slice_dtype\\\", \\\"float32\\\")\\n hparams.activation_dtype = \\\"bfloat16\\\"\\n\\n # 8-way model-parallelism\\n hparams.add_hparam(\\\"mesh_shape\\\", \\\"model:8\\\")\\n hparams.add_hparam(\\\"layout\\\", \\\"batch:batch;vocab:model;d_ff:model;heads:model\\\")\\n\\n # If nonzero, we split the batch across two tensor-dimensions named\\n # \\\"outer_batch\\\" and \\\"inner_batch\\\", allowing for splitting across two mesh\\n # dimensions. This is necessary for hierarchical mixture of experts.\\n # The two tensor dimensions have sizes hparams.outer_batch_size and\\n # hparams.batch_size // hparams.outer_batch_size.\\n hparams.add_hparam(\\\"outer_batch_size\\\", 0)\\n\\n hparams.shared_embedding_and_softmax_weights = False\\n # length for training or decoding - defaults to max_length\\n hparams.add_hparam(\\\"length\\\", 0)\\n\\n # These parameters make Transformer model compatible with mtf\\n # Do not override these.\\n hparams.no_data_parallelism = True\\n hparams.use_fixed_batch_size = True\\n hparams.add_hparam(\\\"mtf_mode\\\", True)\\n hparams.clip_grad_norm = 0. # i.e. no gradient clipping\\n hparams.bottom = {\\n \\\"inputs\\\": modalities.identity_bottom,\\n \\\"targets\\\": modalities.identity_bottom,\\n }\\n hparams.top = {\\n \\\"targets\\\": modalities.identity_top,\\n }\\n hparams.add_hparam(\\\"beam_size\\\", 1)\\n\\n # If this is True, then in a packed dataset (where exaples are concatenated\\n # to form longer examples) we use the global position (within the concatenated\\n # sequence) to compute the positional embedding, instead of the position\\n # within the individual sequence. This is counterintuitive, but for some\\n # reason, it keeps the model from diverging.\\n hparams.add_hparam(\\\"use_global_position_in_packed_sequence\\\", True)\\n\\n return hparams\",\n \"def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\\n super(TransformerEncoder, self).__init__()\\n\\n # Create num_layers of TransformerBlock\\n self.encoder = Sequential(\\n [\\n TransformerBlock(num_heads=num_heads,\\n D=D,\\n hidden_layers=[mlp_dim, D],\\n dropout=dropout,\\n norm_eps=norm_eps)\\n for _ in range(num_layers)\\n ]\\n )\",\n \"def attention_module_multi_head(roi_feat, position_embedding,\\n nongt_dim, fc_dim, feat_dim,\\n dim=(1024, 1024, 1024),\\n group=16, index=1):\\n dim_group = (dim[0] / group, dim[1] / group, dim[2] / group)\\n nongt_roi_feat = mx.symbol.slice_axis(data=roi_feat, axis=0, begin=0, end=nongt_dim)\\n # [num_rois * nongt_dim, emb_dim]\\n position_embedding_reshape = mx.sym.Reshape(position_embedding, shape=(-3, -2))\\n # position_feat_1, [num_rois * nongt_dim, fc_dim]\\n position_feat_1 = mx.sym.FullyConnected(name='pair_pos_fc1_' + str(index),\\n data=position_embedding_reshape,\\n num_hidden=fc_dim)\\n position_feat_1_relu = mx.sym.Activation(data=position_feat_1, act_type='relu')\\n # aff_weight, [num_rois, nongt_dim, fc_dim]\\n aff_weight = mx.sym.Reshape(position_feat_1_relu, shape=(-1, nongt_dim, fc_dim))\\n # aff_weight, [num_rois, fc_dim, nongt_dim]\\n aff_weight = mx.sym.transpose(aff_weight, axes=(0, 2, 1))\\n\\n # multi head\\n assert dim[0] == dim[1], 'Matrix multiply requires same dimensions!'\\n q_data = mx.sym.FullyConnected(name='query_' + str(index),\\n data=roi_feat,\\n num_hidden=dim[0])\\n q_data_batch = mx.sym.Reshape(q_data, shape=(-1, group, dim_group[0]))\\n q_data_batch = mx.sym.transpose(q_data_batch, axes=(1, 0, 2))\\n k_data = mx.symbol.FullyConnected(name='key_' + str(index),\\n data=nongt_roi_feat,\\n num_hidden=dim[1])\\n k_data_batch = mx.sym.Reshape(k_data, shape=(-1, group, dim_group[1]))\\n k_data_batch = mx.sym.transpose(k_data_batch, axes=(1, 0, 2))\\n v_data = nongt_roi_feat\\n # v_data = mx.symbol.FullyConnected(name='value_'+str(index)+'_'+str(gid), data=roi_feat, num_hidden=dim_group[2])\\n aff = mx.symbol.batch_dot(lhs=q_data_batch, rhs=k_data_batch, transpose_a=False, transpose_b=True)\\n # aff_scale, [group, num_rois, nongt_dim]\\n aff_scale = (1.0 / math.sqrt(float(dim_group[1]))) * aff\\n aff_scale = mx.sym.transpose(aff_scale, axes=(1, 0, 2))\\n\\n assert fc_dim == group, 'fc_dim != group'\\n # weighted_aff, [num_rois, fc_dim, nongt_dim]\\n weighted_aff = mx.sym.log(mx.sym.maximum(left=aff_weight, right=1e-6)) + aff_scale\\n aff_softmax = mx.symbol.softmax(data=weighted_aff, axis=2, name='softmax_' + str(index))\\n # [num_rois * fc_dim, nongt_dim]\\n aff_softmax_reshape = mx.sym.Reshape(aff_softmax, shape=(-3, -2))\\n # output_t, [num_rois * fc_dim, feat_dim]\\n output_t = mx.symbol.dot(lhs=aff_softmax_reshape, rhs=v_data)\\n # output_t, [num_rois, fc_dim * feat_dim, 1, 1]\\n output_t = mx.sym.Reshape(output_t, shape=(-1, fc_dim * feat_dim, 1, 1))\\n # linear_out, [num_rois, dim[2], 1, 1]\\n linear_out = mx.symbol.Convolution(name='linear_out_' + str(index), data=output_t,\\n kernel=(1, 1), num_filter=dim[2], num_group=fc_dim)\\n output = mx.sym.Reshape(linear_out, shape=(0, 0))\\n return output\",\n \"def mtf_bitransformer_base():\\n hparams = mtf_transformer2_base()\\n hparams.max_length = 256\\n hparams.shared_embedding = True\\n # HYPERPARAMETERS FOR THE LAYER STACKS\\n hparams.add_hparam(\\\"encoder_layers\\\", [\\\"self_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"decoder_layers\\\", [\\\"self_att\\\", \\\"enc_att\\\", \\\"drd\\\"] * 6)\\n hparams.add_hparam(\\\"encoder_num_layers\\\", 6)\\n hparams.add_hparam(\\\"decoder_num_layers\\\", 6)\\n # number of heads in multihead attention\\n hparams.add_hparam(\\\"encoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"decoder_num_heads\\\", 8)\\n hparams.add_hparam(\\\"local_attention_radius\\\", 128)\\n\\n # default of 0 for standard transformer behavior\\n # 1 means a single set of keys and values that are read by all query heads\\n hparams.add_hparam(\\\"encoder_num_memory_heads\\\", 0)\\n hparams.add_hparam(\\\"decoder_num_memory_heads\\\", 0)\\n # share attention keys and values\\n hparams.add_hparam(\\\"encoder_shared_kv\\\", False)\\n hparams.add_hparam(\\\"decoder_shared_kv\\\", False)\\n\\n # Parameters for computing the maximum decode length in beam search.\\n # Maximum decode length is:\\n # min(max_length,\\n # decode_length_multiplier * input_length + decode_length_constant)\\n hparams.add_hparam(\\\"decode_length_multiplier\\\", 1.5)\\n hparams.add_hparam(\\\"decode_length_constant\\\", 10.0)\\n # used during decoding\\n hparams.add_hparam(\\\"alpha\\\", 0.6)\\n hparams.sampling_temp = 0.0\\n return hparams\",\n \"def mtf_unitransformer_base():\\n hparams = mtf_transformer2_base()\\n hparams.add_hparam(\\\"autoregressive\\\", True)\\n # HYPERPARAMETERS FOR THE SINGLE LAYER STACK\\n hparams.add_hparam(\\\"layers\\\", [\\\"self_att\\\", \\\"drd\\\"] * 6)\\n # number of heads in multihead attention\\n hparams.add_hparam(\\\"num_heads\\\", 8)\\n # default of 0 for standard transformer behavior\\n # 1 means a single set of keys and values that are read by all query heads\\n hparams.add_hparam(\\\"num_memory_heads\\\", 0)\\n # share attention keys and values\\n hparams.add_hparam(\\\"shared_kv\\\", False)\\n # if nonzero then use local attention\\n hparams.add_hparam(\\\"local_attention_radius\\\", 128)\\n return hparams\",\n \"def forward(self, x):\\n # Encoder1 --block1\\n encode_block1 = self.conv_encoder1(x)\\n if self.residus[0] == 1:\\n encode_block1 += self.residual_shortcut1(x)\\n encode_pool1 = self.max_pool_encoder1(encode_block1)\\n\\n # Encoder2 --block2\\n encode_block2 = self.conv_encoder2(encode_pool1)\\n if self.residus[1] == 1:\\n encode_block2 += self.residual_shortcut2(encode_pool1)\\n encode_pool2 = self.max_pool_encoder2(encode_block2)\\n \\n # Encoder3 --block3\\n encode_block3 = self.conv_encoder3(encode_pool2)\\n if self.residus[2] == 1:\\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\\n encode_pool3 = self.max_pool_encoder3(encode_block3)\\n \\n # Encoder4 --block4\\n encode_block4 = self.conv_encoder4(encode_pool3) \\n if self.residus[3] == 1:\\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\\n encode_pool4 = self.max_pool_encoder4(encode_block4)\\n\\n # Transitional block \\n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\\n \\n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\\n \\n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\\n\\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\\n \\n convTranspose_transitional = self.convTranspose_transitional(middle_block) \\n # Decoder4 --block5\\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \\n if self.residus[4] == 1:\\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\\n\\n #--block6\\n cat_layer3 = self.conv_decoder4(decode_block4) \\n if self.residus[5] == 1:\\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\\n \\n \\n # Decoder3 --block7\\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\\n if self.residus[6] == 1:\\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\\n \\n #--block8\\n cat_layer2 = self.conv_decoder3(decode_block3) \\n if self.residus[7] == 1:\\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\\n \\n # Decoder2 --block9\\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \\n if self.residus[8] == 1:\\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\\n \\n #--block10\\n cat_layer1 = self.conv_decoder2(decode_block2)\\n if self.residus[9] == 1:\\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\\n \\n # Decoder1 --block11\\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \\n if self.residus[10] == 1:\\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\\n \\n #--block12\\n final_layer = self.final_layer(decode_block1)\\n if self.residus[11] == 1:\\n final_layer += self.residual_shortcut_final_layer(decode_block1)\\n \\n \\n return final_layer\",\n \"def __init__(self, ):\\n super().__init__()\\n channels = 3\\n\\n # Initial convolution block\\n out_features = 64\\n # encoder\\n self.input = nn.Sequential(\\n nn.ReflectionPad2d(channels),\\n nn.Conv2d(3, out_features, (7, 7)),\\n nn.InstanceNorm2d(out_features),\\n nn.MaxPool2d(2),\\n nn.LeakyReLU(0.2, inplace=True),\\n nn.Conv2d(out_features, out_features * 2, (3, 3), stride=(1, 1), padding=(1, 1)),\\n nn.LeakyReLU(0.2, inplace=True),\\n nn.MaxPool2d(2),\\n nn.Conv2d(out_features * 2, out_features * 4, (3, 3), stride=(1, 1), padding=(1, 1)),\\n nn.LeakyReLU(0.2, inplace=True),\\n )\\n\\n self.in0 = nn.InstanceNorm2d(256)\\n self.block0 = blocks()\\n self.block1 = blocks()\\n self.block2 = blocks()\\n self.block3 = blocks()\\n self.block4 = blocks()\\n self.block5 = blocks()\\n self.block6 = blocks()\\n self.block7 = blocks()\\n\\n self.out = nn.Sequential(\\n\\n nn.Upsample(scale_factor=2),\\n nn.Conv2d(out_features * 4, out_features * 2, 3, stride=1, padding=1),\\n nn.InstanceNorm2d(out_features * 2),\\n nn.LeakyReLU(0.2, inplace=True),\\n\\n nn.Upsample(scale_factor=2),\\n nn.Conv2d(out_features * 2, out_features, 3, stride=1, padding=1),\\n nn.InstanceNorm2d(out_features),\\n nn.LeakyReLU(0.2, inplace=True),\\n\\n nn.ReflectionPad2d(channels),\\n nn.Conv2d(out_features, channels, 7),\\n nn.Tanh(),\\n )\",\n \"def __init__(self):\\n super(Encoder3, self).__init__()\\n self.lblocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n ]\\n )\\n\\n self.blocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n ]\\n )\",\n \"def __init__(self):\\n super(Encoder2, self).__init__()\\n self.lblocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n ),\\n ]\\n )\\n\\n self.blocks = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n ),\\n ]\\n )\",\n \"def build(self, unused_input_shapes):\\n\\n dense3d_impl = Dense3D\\n dense2dprojection_impl = Dense2DProjection\\n\\n self.attention_layer = CustomAttention(\\n num_attention_heads=self.num_attention_heads,\\n size_per_head=self.attention_head_size,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n backward_compatible=self.backward_compatible,\\n name=\\\"self\\\")\\n\\n self.attention_output_dense = dense3d_impl(\\n num_attention_heads=self.num_attention_heads,\\n size_per_head=int(self.hidden_size / self.num_attention_heads),\\n kernel_initializer=get_initializer(self.initializer_range),\\n output_projection=True,\\n backward_compatible=self.backward_compatible,\\n name=\\\"dense\\\")\\n self.attention_dropout = tf.keras.layers.Dropout(\\n rate=self.hidden_dropout_prob)\\n self.attention_layer_norm = (\\n tf.keras.layers.LayerNormalization(\\n name=LAYER_NORM_NAME, axis=-1, epsilon=1e-12,\\n # We do layer norm in float32 for numeric stability.\\n dtype=tf.float32))\\n\\n self.intermediate_dense = dense2dprojection_impl(\\n output_size=self.intermediate_size,\\n kernel_initializer=get_initializer(self.initializer_range),\\n activation=self.intermediate_activation,\\n # Uses float32 so that gelu activation is done in float32.\\n fp32_activation=True, name='dense')\\n self.output_dense = dense2dprojection_impl(\\n output_size=self.hidden_size,\\n kernel_initializer=get_initializer(self.initializer_range),\\n name='dense')\\n self.output_dropout = tf.keras.layers.Dropout(rate=self.hidden_dropout_prob)\\n self.output_layer_norm = tf.keras.layers.LayerNormalization(axis=-1,\\n epsilon=1e-12, dtype=tf.float32, name=LAYER_NORM_NAME)\\n super(CustomTransformerBlock, self).build(unused_input_shapes)\",\n \"def __init__(self, in_features, out_features, activation, batch_norm=False, combine_with_1x1=False):\\n super(InceptionBlock, self).__init__()\\n\\n self.activation = activation\\n\\n if not combine_with_1x1:\\n if out_features % 4 != 0:\\n raise ValueError(\\\"If not combining with 1x1, out_features must be divisible by 4\\\")\\n conv_out_features = int(out_features/4)\\n else:\\n conv_out_features = out_features\\n\\n if batch_norm:\\n self.batch_norm = nn.BatchNorm2d(conv_out_features)\\n self.batch_norm1 = nn.BatchNorm2d(conv_out_features)\\n self.batch_norm3 = nn.BatchNorm2d(conv_out_features)\\n self.batch_norm5 = nn.BatchNorm2d(conv_out_features)\\n self.batch_norm7 = nn.BatchNorm2d(conv_out_features)\\n else:\\n self.batch_norm = None\\n\\n self.conv1x1 = nn.Conv2d(in_features, conv_out_features, kernel_size=1, padding=0)\\n self.conv3x3 = nn.Conv2d(in_features, conv_out_features, kernel_size=3, padding=1)\\n self.conv5x5 = nn.Conv2d(in_features, conv_out_features, kernel_size=5, padding=2)\\n self.conv7x7 = nn.Conv2d(in_features, conv_out_features, kernel_size=7, padding=3)\\n\\n self.combine_with_1x1 = combine_with_1x1\\n if combine_with_1x1:\\n self.combiner = nn.Conv2d(out_features*4, out_features, kernel_size=1)\\n self.batch_norm_comb = nn.BatchNorm2d(out_features)\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n TransformerDecoderBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n multi_channel_cross_attention=self.multi_channel_cross_attention,\\n name=(\\\"layer_%d\\\" % i)))\\n super(TransformerDecoder, self).build(unused_input_shapes)\",\n \"def __init__(\\n self,\\n in_channels=128,\\n aux_channels=80,\\n channels=64,\\n out_channels=1,\\n kernel_size=9,\\n dilation=2,\\n bias=True,\\n noise_upsample_scales=[11, 2, 2, 2],\\n noise_upsample_activation=\\\"LeakyReLU\\\",\\n noise_upsample_activation_params={\\\"negative_slope\\\": 0.2},\\n upsample_scales=[2, 2, 2, 2, 2, 2, 2, 2, 1],\\n upsample_mode=\\\"nearest\\\",\\n gated_function=\\\"softmax\\\",\\n use_weight_norm=True,\\n ):\\n super().__init__()\\n\\n self.in_channels = in_channels\\n\\n noise_upsample = []\\n in_chs = in_channels\\n for noise_upsample_scale in noise_upsample_scales:\\n # NOTE(kan-bayashi): How should we design noise upsampling part?\\n noise_upsample += [\\n torch.nn.ConvTranspose1d(\\n in_chs,\\n channels,\\n noise_upsample_scale * 2,\\n stride=noise_upsample_scale,\\n padding=noise_upsample_scale // 2 + noise_upsample_scale % 2,\\n output_padding=noise_upsample_scale % 2,\\n bias=bias,\\n )\\n ]\\n noise_upsample += [\\n getattr(torch.nn, noise_upsample_activation)(\\n **noise_upsample_activation_params\\n )\\n ]\\n in_chs = channels\\n self.noise_upsample = torch.nn.Sequential(*noise_upsample)\\n self.noise_upsample_factor = np.prod(noise_upsample_scales)\\n\\n self.blocks = torch.nn.ModuleList()\\n aux_chs = aux_channels\\n for upsample_scale in upsample_scales:\\n self.blocks += [\\n TADEResBlock(\\n in_channels=channels,\\n aux_channels=aux_chs,\\n kernel_size=kernel_size,\\n dilation=dilation,\\n bias=bias,\\n upsample_factor=upsample_scale,\\n upsample_mode=upsample_mode,\\n gated_function=gated_function,\\n ),\\n ]\\n aux_chs = channels\\n self.upsample_factor = np.prod(upsample_scales)\\n\\n self.output_conv = torch.nn.Sequential(\\n torch.nn.Conv1d(\\n channels,\\n out_channels,\\n kernel_size,\\n 1,\\n bias=bias,\\n padding=(kernel_size - 1) // 2,\\n ),\\n torch.nn.Tanh(),\\n )\\n\\n # apply weight norm\\n if use_weight_norm:\\n self.apply_weight_norm()\\n\\n # reset parameters\\n self.reset_parameters()\",\n \"def mobilenetv2 (inputs, k, alpha = 1.0, train_bn = False):\\n\\n x = conv_block(inputs, 32, alpha, (3, 3), strides=(2, 2), block_id=0, train_bn=train_bn) # Input Res: 1\\n\\n x = inverted_residual_block(x, 16, (3, 3), t=1, strides=1, n=1, alpha=1.0, block_id=1, train_bn=train_bn)\\t# Input Res: 1/2\\n x = inverted_residual_block(x, 24, (3, 3), t=6, strides=2, n=2, alpha=1.0, block_id=2, train_bn=train_bn)\\t# Input Res: 1/2\\n x = inverted_residual_block(x, 32, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=4, train_bn=train_bn)\\t# Input Res: 1/4\\n x = inverted_residual_block(x, 64, (3, 3), t=6, strides=2, n=4, alpha=1.0, block_id=7, train_bn=train_bn)\\t# Input Res: 1/8\\n x = inverted_residual_block(x, 96, (3, 3), t=6, strides=1, n=3, alpha=1.0, block_id=11, train_bn=train_bn)\\t# Input Res: 1/8\\n x = inverted_residual_block(x, 160, (3, 3), t=6, strides=2, n=3, alpha=1.0, block_id=14, train_bn=train_bn)\\t# Input Res: 1/16\\n x = inverted_residual_block(x, 320, (3, 3), t=6, strides=1, n=1, alpha=1.0, block_id=17, train_bn=train_bn)\\t# Input Res: 1/32\\n\\n x = conv_block(x, 1280, alpha, (1, 1), strides=(1, 1), block_id=18, train_bn=train_bn) # Input Res: 1/32\\n\\n x = KL.GlobalAveragePooling2D()(x)\\n x = KL.Reshape((1, 1, 1280))(x)\\n x = KL.Dropout(0.3, name='Dropout')(x)\\n x = KL.Conv2D(k, (1, 1), padding='same')(x)\\n\\n x = KL.Activation('softmax', name='softmax')(x)\\n output = KL.Reshape((k,))(x)\\n\\n model = KM.Model(inputs, output)\\n plot_model(model, to_file='MobileNetv2.png', show_shapes=True)\\n\\n return model\",\n \"def __init__(self, cfg):\\n super(MHBCoAtt, self).__init__()\\n self.cfg = cfg\\n # word embedding: q_vocab_size, 1024\\n self.word_embedding = nn.Embedding(cfg.q_vocab_size, cfg.emb_dim)\\n # LSTM\\n if cfg.glove:\\n self.lstm = nn.LSTM(input_size=cfg.emb_dim*2,\\n hidden_size=cfg.hidden_dim,\\n num_layers=cfg.num_layers,\\n batch_first=True)\\n else:\\n self.lstm = nn.LSTM(input_size=cfg.emb_dim,\\n hidden_size=cfg.hidden_dim,\\n num_layers=cfg.num_layers,\\n batch_first=True)\\n\\n self.dropout_l = nn.Dropout(p = 0.3)\\n # question attention\\n self.ques_att_conv1 = nn.Conv2d(cfg.hidden_dim, 512, [1,1])\\n self.ques_att_conv2 = nn.Conv2d(512, 2, [1,1])\\n\\n # question attentive feature fuse with image feature, according to paper: k * o = 5000, k = 5\\n self.ques_proj1 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.img_conv1d = nn.Conv2d(cfg.img_feature_channel, 5000, [1, 1])\\n self.dropout_m = nn.Dropout(p = 0.1)\\n\\n # co-attention conv layers\\n self.co_att_conv1 = nn.Conv2d(1000, 512, [1,1])\\n self.co_att_conv2 = nn.Conv2d(512, 2, [1,1])\\n\\n # co_attentive feature fuse with question attentive feature\\n self.ques_proj2 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.ques_proj3 = nn.Linear(2*cfg.hidden_dim, 5000)\\n self.img_proj2 = nn.Linear(2*cfg.img_feature_channel, 5000)\\n self.img_proj3 = nn.Linear(2*cfg.img_feature_channel, 5000)\\n\\n # prediction fully connected layer\\n self.linear_pred = nn.Linear(2000, cfg.a_vocab_size)\",\n \"def __call__(self, **kwargs):\\n segname = 'block_{}_expand_relu'\\n blocks = [13, 6, 3, 1]\\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\\n backbone_out = self._backbone.get_layer('block_16_project')\\n\\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\\n p4 = self._fpn_block(p5, skips[1].output)\\n p3 = self._fpn_block(p4, skips[2].output)\\n p2 = self._fpn_block(p3, skips[3].output)\\n\\n s5 = self._conv_block(p5, 128)\\n s4 = self._conv_block(p4, 128)\\n s3 = self._conv_block(p3, 128)\\n s2 = self._conv_block(p2, 128)\\n\\n s5 = tf.keras.layers.UpSampling2D(\\n size=(8, 8),\\n interpolation='nearest'\\n )(s5)\\n\\n s4 = tf.keras.layers.UpSampling2D(\\n size=(4, 4),\\n interpolation='nearest'\\n )(s4)\\n\\n s3 = tf.keras.layers.UpSampling2D(\\n size=(2, 2),\\n interpolation='nearest'\\n )(s3)\\n\\n concat = [s5, s4, s3, s2]\\n x = tf.keras.layers.Concatenate()(concat)\\n x = tf.keras.layers.Conv2D(\\n 64,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n x = tf.keras.layers.BatchNormalization()(x)\\n x = tf.keras.layers.Activation('relu')(x)\\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\\n\\n x = tf.keras.layers.Conv2D(\\n 1,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n out = tf.keras.layers.Activation('sigmoid')(x)\\n model = tf.keras.models.Model(\\n inputs=self._backbone.input,\\n outputs=out\\n )\\n\\n return model\",\n \"def __init__(self, bert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.q_lin.weight, bert_layer.attention.k_lin.weight, bert_layer.attention.v_lin.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.q_lin.bias, bert_layer.attention.k_lin.bias, bert_layer.attention.v_lin.bias]))\\n self.out_proj_weight = bert_layer.attention.out_lin.weight\\n self.out_proj_bias = bert_layer.attention.out_lin.bias\\n self.linear1_weight = bert_layer.ffn.lin1.weight\\n self.linear1_bias = bert_layer.ffn.lin1.bias\\n self.linear2_weight = bert_layer.ffn.lin2.weight\\n self.linear2_bias = bert_layer.ffn.lin2.bias\\n self.norm1_eps = bert_layer.sa_layer_norm.eps\\n self.norm1_weight = bert_layer.sa_layer_norm.weight\\n self.norm1_bias = bert_layer.sa_layer_norm.bias\\n self.norm2_eps = bert_layer.output_layer_norm.eps\\n self.norm2_weight = bert_layer.output_layer_norm.weight\\n self.norm2_bias = bert_layer.output_layer_norm.bias\\n self.num_heads = bert_layer.attention.n_heads\\n self.embed_dim = bert_layer.attention.dim\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def __init__(self, conv_features_sizes, linear_layer_sizes, connector_shape):\\n super().__init__()\\n \\n self.conv = nn.Sequential()\\n self.mlp = nn.Sequential()\\n self.flat = nn.Flatten()\\n\\n self.conv.add_module(name=f\\\"e-fconv{0}\\\", module=_conv2d_block(1, conv_features_sizes[0], kernel_size=3, padding=1))\\n self.conv.add_module(name=f\\\"e-max{0}\\\", module=nn.MaxPool2d(2, 2))\\n for i, (in_size, out_size) in enumerate(zip(conv_features_sizes[:-1], conv_features_sizes[1:]), 1):\\n self.conv.add_module(name=f\\\"e-fconv{i}\\\", module=_conv2d_block(in_size, out_size, kernel_size=3, padding=1))\\n self.conv.add_module(name=f\\\"e-max{i}\\\", module=nn.MaxPool2d(2, 2))\\n\\n mlp_input_shape = int(reduce((lambda x,y: x * y), connector_shape))\\n self.mlp.add_module(name=f\\\"e-linear{0}\\\", module=nn.Linear(mlp_input_shape, linear_layer_sizes[0]))\\n self.mlp.add_module(name=f\\\"e-batchnorm{0}\\\", module=nn.BatchNorm1d(linear_layer_sizes[0]))\\n self.mlp.add_module(name=f\\\"e-relu{0}\\\", module=nn.ReLU())\\n for i, (in_size, out_size) in enumerate(zip(linear_layer_sizes[:-1], linear_layer_sizes[1:]), 1):\\n self.mlp.add_module(name=f\\\"e-linear{i}\\\", module=nn.Linear(in_size, out_size))\\n self.mlp.add_module(name=f\\\"e-batchnorm{i}\\\", module=nn.BatchNorm1d(out_size))\\n self.mlp.add_module(name=f\\\"e-relu{i}\\\", module=nn.ReLU())\",\n \"def __init__(self, input_size: int, output_size: int, old_cols: List[AlphaModule] = None, hidden_size=768,\\n ptm=\\\"bert\\\", feature_size=768, require_proto=False, tokenizer=None, prob_l=-1) -> None:\\n super(PTMClassifier_PNN, self).__init__()\\n \\n if old_cols is None:\\n old_cols = []\\n \\n self.old_cols = []\\n \\n self.output_size = output_size\\n self.hidden_size = hidden_size # default\\n self.feature_size = feature_size\\n \\n self.fc1 = nn.Linear(self.hidden_size, 100)\\n self.classifier = nn.Linear(100, self.output_size)\\n \\n if len(old_cols) > 0:\\n self.old_fc1s = ListModule()\\n self.base_1 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n self.base_2 = nn.Sequential(\\n nn.Linear(100 * len(old_cols), 100),\\n nn.ReLU(),\\n )\\n \\n self.adaptor1 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_1)\\n self.adaptor2 = nn.Sequential(AlphaModule(100 * len(old_cols)),\\n self.base_2)\\n \\n for old_col in old_cols:\\n self.old_fc1s.append(\\n nn.Sequential(nn.Linear(self.hidden_size, 100), nn.ReLU()))\\n self.old_fc1s[-1][0].load_state_dict(old_col.fc1.state_dict())\\n \\n # ptm\\n self.ptm = ptm.lower()\\n assert self.ptm in supported_ptm\\n ptm_ = import_from(\\\"transformers\\\", supported_ptm[self.ptm][0] + \\\"Model\\\")\\n \\n self.encoder = ptm_.from_pretrained(supported_ptm[self.ptm][1], output_hidden_states=True)\\n if tokenizer is not None:\\n self.encoder.resize_token_embeddings(len(tokenizer))\\n self.prob_l = prob_l\\n \\n self.require_proto = require_proto\\n \\n # todo : modify net into one network\\n if self.require_proto:\\n # prototype-based classification\\n self.net = nn.Sequential(\\n self.fc1\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.fc1,\\n self.encoder_adaptor,\\n nn.ReLU(),\\n )\\n else:\\n self.classifier = nn.Linear(self.feature_size, self.output_size, bias=True)\\n self.net = nn.Sequential(\\n self.classifier\\n )\\n self.net_ = nn.Sequential(\\n self.encoder,\\n self.classifier\\n )\\n \\n self.reset_parameters()\",\n \"def __init__(self):\\n super(Decoder_1m, self).__init__()\\n self.lconvtwos = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Dropout(p=0.1),\\n nn.Conv2d(128, 32, kernel_size=(3, 3), padding=1),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=1),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n ),\\n ]\\n )\\n\\n self.convtwos = nn.ModuleList(\\n [\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=1),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=1),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=2, dilation=2),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=4, dilation=4),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=8, dilation=8),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=16, dilation=16),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=32, dilation=32),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n nn.Sequential(\\n nn.Conv2d(64, 32, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(32),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(32, 64, kernel_size=(3, 3), padding=64, dilation=64),\\n nn.BatchNorm2d(64),\\n nn.ReLU(inplace=True),\\n ),\\n ]\\n )\\n self.final = nn.Sequential(\\n nn.Conv2d(64, 5, kernel_size=(1, 1), padding=0),\\n nn.BatchNorm2d(5),\\n nn.ReLU(inplace=True),\\n nn.Conv2d(5, 1, kernel_size=(1, 1), padding=0),\\n )\",\n \"def transformer_model(input_tensor,\\n attention_mask=None,\\n hidden_size=768,\\n num_hidden_layers=12,\\n num_attention_heads=12,\\n intermediate_size=3072,\\n intermediate_act_fn=gelu,\\n hidden_dropout_prob=0.1,\\n attention_probs_dropout_prob=0.1,\\n initializer_range=0.02,\\n do_return_all_layers=False,\\n conv_kernel_size=3,\\n head_ratio=2,\\n conv_type=\\\"noconv\\\",\\n **kargs):\\n if hidden_size % num_attention_heads != 0:\\n raise ValueError(\\n \\\"The hidden size (%d) is not a multiple of the number of attention \\\"\\n \\\"heads (%d)\\\" % (hidden_size, num_attention_heads))\\n\\n attention_head_size = int(hidden_size / num_attention_heads)\\n input_shape = bert_utils.get_shape_list(input_tensor, expected_rank=3)\\n batch_size = input_shape[0]\\n seq_length = input_shape[1]\\n input_width = input_shape[2]\\n\\n # The Transformer performs sum residuals on all layers so the input needs\\n # to be the same as the hidden size.\\n if input_width != hidden_size:\\n raise ValueError(\\\"The width of the input tensor (%d) != hidden size (%d)\\\" %\\n (input_width, hidden_size))\\n\\n # We keep the representation as a 2D tensor to avoid re-shaping it back and\\n # forth from a 3D tensor to a 2D tensor. Re-shapes are normally free on\\n # the GPU/CPU but may not be free on the TPU, so we want to minimize them to\\n # help the optimizer.\\n prev_output = bert_utils.reshape_to_matrix(input_tensor)\\n\\n attn_maps = []\\n all_layer_outputs = []\\n all_value_outputs = []\\n for layer_idx in range(num_hidden_layers):\\n with tf.variable_scope(\\\"layer_%d\\\" % layer_idx):\\n with tf.variable_scope(\\\"attention\\\"):\\n attention_heads = []\\n with tf.variable_scope(\\\"self\\\"):\\n attention_head, probs, value_layer = attention_layer(\\n from_tensor=prev_output,\\n to_tensor=prev_output,\\n attention_mask=attention_mask,\\n num_attention_heads=num_attention_heads,\\n size_per_head=attention_head_size,\\n attention_probs_dropout_prob=attention_probs_dropout_prob,\\n initializer_range=initializer_range,\\n do_return_2d_tensor=True,\\n batch_size=batch_size,\\n from_seq_length=seq_length,\\n to_seq_length=seq_length,\\n conv_kernel_size=conv_kernel_size,\\n head_ratio=head_ratio,\\n conv_type=conv_type,\\n from_tensor_mask=kargs.get('from_tensor_mask', None),\\n to_tensor_mask=kargs.get('to_tensor_mask', None),\\n conv_method=kargs.get('conv_method', \\\"dot\\\"))\\n attention_heads.append(attention_head)\\n attn_maps.append(probs)\\n all_value_outputs.append(value_layer)\\n\\n attention_output = None\\n if len(attention_heads) == 1:\\n attention_output = attention_heads[0]\\n else:\\n # In the case where we have other sequences, we just concatenate\\n # them to the self-attention head before the projection.\\n attention_output = tf.concat(attention_heads, axis=-1)\\n\\n with tf.variable_scope(\\\"output\\\"):\\n # Run a linear projection of `hidden_size` then add a residual\\n # with `layer_input`.\\n attention_output = tf.layers.dense(\\n attention_output,\\n hidden_size,\\n kernel_initializer=create_initializer(initializer_range))\\n \\n\\n attention_output = dropout(attention_output, hidden_dropout_prob)\\n attention_output = layer_norm(attention_output + prev_output)\\n\\n # The activation is only applied to the \\\"intermediate\\\" hidden layer.\\n with tf.variable_scope(\\\"intermediate\\\"):\\n intermediate_output = tf.layers.dense(\\n attention_output,\\n intermediate_size,\\n activation=intermediate_act_fn,\\n kernel_initializer=create_initializer(initializer_range))\\n # Down-project back to `hidden_size` then add the residual.\\n with tf.variable_scope(\\\"output\\\"):\\n prev_output = tf.layers.dense(\\n intermediate_output,\\n hidden_size,\\n kernel_initializer=create_initializer(initializer_range))\\n\\n prev_output = dropout(prev_output, hidden_dropout_prob)\\n prev_output = layer_norm(prev_output + attention_output)\\n all_layer_outputs.append(prev_output)\\n\\n attn_maps = tf.stack(attn_maps, 0)\\n if do_return_all_layers:\\n final_outputs = []\\n for layer_output in all_layer_outputs:\\n final_output = bert_utils.reshape_from_matrix(layer_output, input_shape)\\n final_outputs.append(final_output)\\n return final_outputs, attn_maps, all_value_outputs\\n else:\\n final_output = bert_utils.reshape_from_matrix(prev_output, input_shape)\\n return final_output, attn_maps, all_value_outputs\",\n \"def Attention_block(up_in,down_in,nf):\\n# def MyTile(alpha, channels):\\n# alpha = tf.tile(alpha, [1, 1, 1, 1, channels])\\n# return alpha\\n# def MyTile_output_shape(input_shape):\\n# shape = list(input_shape)\\n# assert len(shape) == 5 # only valid for 2D tensors\\n# shape[-1] *= channels\\n# return tuple(shape)\\n ndims = len(up_in.get_shape()) - 2\\n assert ndims in [1, 2, 3], \\\"ndims should be one of 1, 2, or 3. found: %d\\\" % ndims\\n input_channels = up_in.get_shape().as_list()[-1]\\n# batch_size1 = tf.shape(down_in)[0]\\n# nf = tf.min(batch_size0,batch_size1)\\n Conv = getattr(KL, 'Conv%dD' % ndims)\\n up = Conv(nf, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', strides=2)(up_in)\\n down = Conv(nf, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', strides=1)(down_in)\\n\\n# x = NCCLayer(channels=nf)([up,down])\\n x = Add()([up,down])\\n x = Activation('relu')(x)\\n x = Conv(1, kernel_size=1, padding='same',\\n kernel_initializer='he_normal', use_bias = True, bias_initializer='zeros',strides=1,activation='sigmoid')(x)\\n# x = Activation('sigmoid')(x)\\n upsample_layer = getattr(KL, 'UpSampling%dD' % ndims)\\n alpha = upsample_layer()(x)\\n# alpha = Lambda(MyTile)(alpha, input_channels)\\n alpha = MyTile(channels=input_channels)(alpha)\\n up_out = Multiply()([alpha,up_in])\\n return up_out\",\n \"def mtf_bitransformer_all_layers_tiny():\\n hparams = mtf_bitransformer_tiny()\\n hparams.moe_num_experts = 4\\n hparams.moe_expert_x = 4\\n hparams.moe_expert_y = 4\\n hparams.moe_hidden_size = 512\\n hparams.encoder_layers = [\\n \\\"self_att\\\", \\\"local_self_att\\\", \\\"moe_1d\\\", \\\"moe_2d\\\", \\\"drd\\\"]\\n hparams.decoder_layers = [\\n \\\"self_att\\\", \\\"local_self_att\\\", \\\"enc_att\\\", \\\"moe_1d\\\", \\\"moe_2d\\\", \\\"drd\\\"]\\n return hparams\",\n \"def __init__(self,\\n size: int = 0,\\n bb_size: int = 64,\\n ff_size: int = 0,\\n num_heads: int = 0,\\n dropout: float = 0.1):\\n super(CustomTransformerDecoderLayer, self).__init__()\\n self.size = size\\n\\n self.trg_trg_att = ContMultiHeadedAttention(num_heads, bb_size, bb_size,\\n dropout=dropout)\\n self.src_trg_att = ContMultiHeadedAttention(num_heads, size, size,\\n dropout=dropout)\\n\\n self.feed_forward_h1 = PositionwiseFeedForward(bb_size, ff_size=ff_size)\\n self.feed_forward_h2 = PositionwiseFeedForward(size, ff_size=ff_size)\\n\\n self.x_layer_norm = nn.LayerNorm(size, eps=1e-6)\\n self.spa_layer_norm = nn.LayerNorm(bb_size, eps=1e-6)\\n\\n self.dropout = nn.Dropout(dropout)\",\n \"def forward_train(self, imgs, label, token_ids=None, segment_ids=None, input_mask=None, \\n mlm_label=None, dvae_imgs=None, v_token_mask=None, hog_features=None, img_metas=None, **kwargs): \\n # (batch_size, num_clips*num_crops, channel, num_segments, h, w) -> (batch_size*num_clips*num_crops, channel, num_segments, h, w)\\n imgs = imgs.reshape((-1, ) + imgs.shape[2:]) \\n if self.from_scratch:\\n imgs = imgs / 255.0\\n # text reshape: (batch_size, num_candidates, seq_length) -> (batch_size * num_candidates, seq_length)\\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\\n text_input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\\n if mlm_label is not None:\\n mlm_label = mlm_label.reshape((-1, ) + mlm_label.shape[2:])\\n\\n\\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\\n\\n B, D, T, H, W = visual_token.shape\\n losses = dict()\\n # -------------- nce loss ------------------- #\\n if hasattr(self, 'ssl_head'):\\n input_ssl_ids = torch.where(mlm_label == -100, token_ids.clone(), mlm_label.clone())\\n input_ssl_mask = text_input_mask.clone()\\n text_only_out = self.text_backbone(input_ssl_ids, input_ssl_mask)\\n # ------------ complete T -------------- #\\n text_out_no_mask = text_only_out['last_hidden_state']\\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_no_mask, input_ssl_mask, input_ssl_ids)\\n\\n\\n # ------------ complete V ---------------- #\\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\\n\\n\\n # ---------- foward mask text input ---------- # \\n text_out_with_mask = self.text_backbone(token_ids, text_input_mask)\\n text_out_last_hidden_state = text_out_with_mask['last_hidden_state']\\n\\n # ---------- forward mask v input ------------ #\\n visual_token_with_mask, v_mask = self.extract_visual_feat(imgs.clone(), v_token_mask) # b, d, T, h, w\\n visual_token_mask = visual_token_with_mask.view(B, D, T, -1).permute(0, 2, 3, 1)\\n \\n v_fusion_output = self.multimodal_backbone(visual_token=visual_token_mask, text_input_mask=text_input_mask.clone(), text_input_embeds=text_out_no_mask.clone())\\n \\n t_fusion_output = self.multimodal_backbone(visual_token=visual_token, text_input_mask=text_input_mask, text_input_embeds=text_out_last_hidden_state)\\n # for mlm #\\n t_last_hidden_state = t_fusion_output['t_last_hidden_state']\\n\\n\\n\\n\\n\\n # ------------ MLM loss ------------ #\\n\\n if mlm_label is not None and self.mlm_head is not None:\\n # we use mask text for MLM\\n # because we doubt there will be miss interaction between wrong img-text pair \\n # and the model not learn good relationship between vision and language\\n # -------- forward masked text ----------- #\\n mlm_prediction_score = self.mlm_head(t_last_hidden_state)\\n \\n if self.mlm_loss_func is not None:\\n mlm_label_idx = torch.where(mlm_label.view(-1) != -100)\\n mlm_prediction_mask_score = mlm_prediction_score.view(-1, self.text_vocab_size)[mlm_label_idx[0], :]\\n mlm_label_mask = mlm_label.view(-1)[mlm_label_idx]\\n mlm_loss = self.mlm_loss_func(mlm_prediction_mask_score, mlm_label_mask)\\n else:\\n mlm_loss = self.loss_func(mlm_prediction_score.view(-1, self.text_vocab_size), mlm_label.view(-1))\\n losses['mlm_loss'] = mlm_loss\\n\\n\\n # ------- Tri-modal alignment with mask sample and ranking --------- #\\n if self.mlm_ssl_V_head is not None:\\n mlm_visual_feat = v_fusion_output['t_last_hidden_state'][:, 0]\\n mask_visual_recon_emb = self.mlm_ssl_V_head(mlm_visual_feat)\\n mask_word_emb = self.ssl_head.forward_text(text_out_last_hidden_state) if self.use_Cmask else None\\n loss_cvt_rank = self.ssl_loss(visual_emb, text_emb, mask_word_emb, mask_visual_recon_emb)\\n losses.update(loss_cvt_rank)\\n\\n\\n if self.symmetry_rank:\\n mlm_word_feat = t_last_hidden_state[:, 0]\\n mask_word_recon_emb = self.mlm_ssl_T_head(mlm_word_feat)\\n\\n mask_visual_emb = self.ssl_head.forward_vision(visual_token_with_mask) if self.use_Cmask else None\\n \\n loss_ctv_rank = self.ssl_loss(text_emb, visual_emb, mask_visual_emb, mask_word_recon_emb)\\n loss_ctv_rank['v_nce_loss'] = loss_ctv_rank.pop('nce_loss')\\n \\n if self.ssl_loss.use_rank:\\n loss_ctv_rank['rank_v_vm_loss'] = loss_ctv_rank.pop('rank_t_tm_loss')\\n\\n \\n\\n losses.update(loss_ctv_rank)\\n\\n\\n\\n return losses\",\n \"def identity_block(self, input_tensor, kernel_size, filters, stage, block):\\r\\n\\t\\tfilters1, filters2, filters3 = filters # this indicates the number of filter (here, assume that all filter number are same)\\r\\n\\r\\n\\t\\tbn_axis = self.channel_last # number of classes being classified\\r\\n \\r\\n\\t\\t# names for the layers\\r\\n\\t\\tconv_name_base = 'res' + str(stage) + block + '_branch'\\r\\n\\t\\tbn_name_base = 'bn' + str(stage) + block + '_branch'\\r\\n\\r\\n\\t\\t# model definition\\r\\n\\t\\tx = Conv2D(filters1, (1, 1),\\r\\n\\t\\t\\t\\t\\t\\t kernel_initializer='he_normal',\\r\\n\\t\\t\\t\\t\\t\\t name=conv_name_base + '2a')(input_tensor)\\r\\n\\t\\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')(x)\\r\\n\\t\\tx = Activation('relu')(x)\\r\\n\\r\\n\\t\\tx = Conv2D(filters2, kernel_size,\\r\\n\\t\\t\\t\\t\\t\\t padding='same',\\r\\n\\t\\t\\t\\t\\t\\t kernel_initializer='he_normal',\\r\\n\\t\\t\\t\\t\\t\\t name=conv_name_base + '2b')(x)\\r\\n \\r\\n\\t\\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')(x)\\r\\n\\t\\tx = Activation('relu')(x)\\r\\n\\r\\n\\t\\tx = Conv2D(filters3, (1, 1),\\r\\n\\t\\t\\t\\t\\t\\t kernel_initializer='he_normal',\\r\\n\\t\\t\\t\\t\\t\\t name=conv_name_base + '2c')(x)\\r\\n\\t\\tx = BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')(x)\\r\\n\\r\\n\\t\\t# Here, we add the outcome tensor(layer) x to original input tensor\\r\\n\\t\\tx = add([x, input_tensor])\\r\\n\\t\\tx = Activation('relu')(x)\\r\\n \\r\\n\\t\\treturn x\",\n \"def forward(self, src, src_t, seq, seq_t, seq_e, mask, vars_dict):\\n\\n src_ext = torch.unsqueeze(src, dim=1) # src [B, 1, D]\\n src_e_ph = torch.zeros_like(src_ext)\\n q = torch.cat([src_ext, src_e_ph, src_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\\n k = torch.cat([seq, seq_e, seq_t], dim=2) # [B, 1, D + De + Dt] -> [B, 1, D]\\n\\n mask = torch.unsqueeze(mask, dim=2) # mask [B, N, 1]\\n mask = mask.permute([0, 2, 1]) # mask [B, 1, N]\\n\\n # # target-attention\\n output, attn = self.multi_head_target(q=q, k=k, v=k, vars_dict=vars_dict, mask=mask) # output: [B, 1, D + Dt], attn: [B, 1, N]\\n # print('src.shape', src.shape)\\n output = output.squeeze(1)\\n # print('output.shape', output.shape)\\n # print('output', output)\\n # print('output.squeeze().shape', output.shape)\\n attn = attn.squeeze()\\n\\n # output = self.merger(output, src)\\n x = torch.cat([output, src], dim=1)\\n # x = self.layer_norm(x)\\n x = F.relu(F.linear(x, vars_dict['w1_agg_fc']))\\n output = F.linear(x, vars_dict['w2_agg_fc'])\\n\\n return output, attn\",\n \"def __init__(self):\\n super(Generator_Model, self).__init__()\\n\\n # audio encoder\\n self.audio_encoder = nn.Sequential(\\n conv_block(1, 32, kernel_size=1, stride=1, padding=1),\\n conv_block(32, 32, kernel_size=1, stride=1, padding=1, residual=True),\\n \\n conv_block(32, 64, kernel_size=3, stride=(3, 1), padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(64, 128, kernel_size=3, stride=3, padding=1),\\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(128, 256, kernel_size=3, stride=(3, 2), padding=1),\\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True),\\n\\n conv_block(256, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 512, kernel_size=1, stride=1, padding=0)\\n )\\n\\n # face encoder\\n self.face_encoder = nn.ModuleList([\\n nn.Sequential(conv_block(6, 16, kernel_size=7, stride=1, padding=3)),\\n\\n nn.Sequential(conv_block(16, 32, kernel_size=3, stride=2, padding=1),\\n conv_block(32, 32, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(32, 64, kernel_size=3, stride=2, padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(64, 128, kernel_size=3, stride=2, padding=1),\\n conv_block(128, 128, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(128, 256, kernel_size=3, stride=2, padding=1),\\n conv_block(256, 256, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_block(256, 512, kernel_size=3, stride=2, padding=1),\\n conv_block(512, 512, kernel_size=3, stride=1, padding=1, residual=True)),\\n \\n nn.Sequential(conv_block(512, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 512, kernel_size=1, stride=1, padding=0))\\n ])\\n\\n # face decoder\\n self.face_decoder = nn.ModuleList([\\n nn.Sequential(conv_block(512, 512, kernel_size=1, stride=1, padding=0),),\\n\\n nn.Sequential(conv_t_block(512, 512, kernel_size=3, stride=1, padding=0),\\n conv_block(512, 1024, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(1024, 512, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(512, 768, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(768, 384, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(384, 512, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(512, 256, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(256, 320, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(320, 128, kernel_size=3, stride=2, padding=1, output_padding=1), \\n conv_block(128, 160, kernel_size=3, stride=1, padding=1, residual=True)),\\n\\n nn.Sequential(conv_t_block(160, 64, kernel_size=3, stride=2, padding=1, output_padding=1),\\n conv_block(64, 64, kernel_size=3, stride=1, padding=1, residual=True))\\n ])\\n\\n # output blocks for face prediction\\n self.pred_output = nn.Sequential(conv_block(64, 32, kernel_size=3, stride=1, padding=1),\\n conv_block(32, 3, kernel_size=1, stride=1, padding=0),\\n nn.Sigmoid())\\n\\n # optimizer\\n self.optimizer = torch.optim.Adam(self.parameters(), lr=0.001, betas=(0.5, 0.999))\",\n \"def __init__(self, mbart_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.weight, mbart_layer.self_attn.k_proj.weight, mbart_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([mbart_layer.self_attn.q_proj.bias, mbart_layer.self_attn.k_proj.bias, mbart_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = mbart_layer.self_attn.out_proj.weight\\n self.out_proj_bias = mbart_layer.self_attn.out_proj.bias\\n self.linear1_weight = mbart_layer.fc1.weight\\n self.linear1_bias = mbart_layer.fc1.bias\\n self.linear2_weight = mbart_layer.fc2.weight\\n self.linear2_bias = mbart_layer.fc2.bias\\n self.norm1_eps = mbart_layer.self_attn_layer_norm.eps\\n self.norm1_weight = mbart_layer.self_attn_layer_norm.weight\\n self.norm1_bias = mbart_layer.self_attn_layer_norm.bias\\n self.norm2_eps = mbart_layer.final_layer_norm.eps\\n self.norm2_weight = mbart_layer.final_layer_norm.weight\\n self.norm2_bias = mbart_layer.final_layer_norm.bias\\n self.num_heads = mbart_layer.self_attn.num_heads\\n self.embed_dim = mbart_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def _init_layers(self) -> None:\\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\\n self.embed_dims = self.self_attn.embed_dims\\n self.ffn = FFN(**self.ffn_cfg)\\n norms_list = [\\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\\n for _ in range(3)\\n ]\\n self.norms = ModuleList(norms_list)\",\n \"def __init__(self, bert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([bert_layer.attention.self.query.weight, bert_layer.attention.self.key.weight, bert_layer.attention.self.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([bert_layer.attention.self.query.bias, bert_layer.attention.self.key.bias, bert_layer.attention.self.value.bias]))\\n self.out_proj_weight = bert_layer.attention.output.dense.weight\\n self.out_proj_bias = bert_layer.attention.output.dense.bias\\n self.linear1_weight = bert_layer.intermediate.dense.weight\\n self.linear1_bias = bert_layer.intermediate.dense.bias\\n self.linear2_weight = bert_layer.output.dense.weight\\n self.linear2_bias = bert_layer.output.dense.bias\\n self.norm1_eps = bert_layer.attention.output.LayerNorm.eps\\n self.norm1_weight = bert_layer.attention.output.LayerNorm.weight\\n self.norm1_bias = bert_layer.attention.output.LayerNorm.bias\\n self.norm2_eps = bert_layer.output.LayerNorm.eps\\n self.norm2_weight = bert_layer.output.LayerNorm.weight\\n self.norm2_bias = bert_layer.output.LayerNorm.bias\\n self.num_heads = bert_layer.attention.self.num_attention_heads\\n self.embed_dim = bert_layer.attention.self.all_head_size\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def mt_step(self, lang1, lang2, lambda_coeff):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n lang1_id = params.lang2id[lang1]\\n lang2_id = params.lang2id[lang2]\\n\\n # generate batch\\n if lang1 == lang2:\\n (x1, len1) = self.get_batch('ae', lang1)\\n (x2, len2) = (x1, len1)\\n (x1, len1) = self.add_noise(x1, len1)\\n else:\\n (x1, len1), (x2, len2) = self.get_batch('mt', lang1, lang2)\\n langs1 = x1.clone().fill_(lang1_id)\\n langs2 = x2.clone().fill_(lang2_id)\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n # key point generate label\\n\\n # cuda\\n x1, len1, langs1, x2, len2, langs2, y = to_cuda(x1, len1, langs1, x2, len2, langs2, y)\\n\\n # encode source sentence\\n enc1 = model('crossfwd', stream_='text', x=x1, lengths=len1, langs=langs1, causal=False)\\n enc1 = enc1.transpose(0, 1)\\n\\n # decode target sentence\\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs2, causal=True, src_enc=enc1,\\n src_len=len1)\\n\\n # loss\\n # the last word not apply prediction logic\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n self.stats[('AE-%s' % lang1) if lang1 == lang2 else ('MT-%s-%s' % (lang1, lang2))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def __init__(\\n self,\\n num_branches,\\n blocks,\\n num_blocks,\\n num_inchannels,\\n num_channels,\\n num_heads,\\n num_window_sizes,\\n num_mlp_ratios,\\n multi_scale_output=True,\\n drop_path=0.0,\\n ):\\n super(HighResolutionTransformerModule, self).__init__()\\n self._check_branches(\\n num_branches, blocks, num_blocks, num_inchannels, num_channels\\n )\\n\\n self.num_inchannels = num_inchannels\\n self.num_branches = num_branches\\n\\n self.multi_scale_output = multi_scale_output\\n self.branches = self._make_branches(\\n num_branches,\\n blocks,\\n num_blocks,\\n num_channels,\\n num_heads,\\n num_window_sizes,\\n num_mlp_ratios,\\n drop_path,\\n )\\n self.fuse_layers = self._make_fuse_layers()\\n self.relu = nn.ReLU(inplace=True)\\n\\n self.num_heads = num_heads\\n self.num_window_sizes = num_window_sizes\\n self.num_mlp_ratios = num_mlp_ratios\",\n \"def build_bisenet(inputs, num_classes):\\n\\n ### The spatial path\\n ### The number of feature maps for each convolution is not specified in the paper\\n ### It was chosen here to be equal to the number of feature maps of a classification\\n ### model at each corresponding stage\\n # spatial_net = fluid.layers.resize_bilinear(inputs, [Image_Height/8, Image_Width/8])\\n # print('spatial_net_1',spatial_net)\\n\\n ## spatial path\\n spatial_net = ConvBlock(inputs, num_filters=64, kernel_size=3, stride=2)\\n spatial_net = ConvBlock(spatial_net, num_filters=128, kernel_size=3, stride=2)\\n spatial_net = ConvBlock(spatial_net, num_filters=256, kernel_size=3, stride=2)\\n # print(\\\"spatial_net:\\\", spatial_net)\\n\\n # spatial_net = fluid.layers.resize_bilinear(spatial_net, [Image_Height/8, Image_Width/8])\\n # print('spatial_net_2',spatial_net)\\n ### Context path\\n model = ResNet(is_test=False)\\n # spatial_net = model.bottleneck_block1(inputs)\\n end_points_16, end_points_32 = model.net(inputs)\\n net_4 = AttentionRefinementModule(end_points_16, num_filters=512)\\n net_5 = AttentionRefinementModule(end_points_32, num_filters=1024)\\n global_channels = fluid.layers.reduce_mean(net_5, [2, 3], keep_dim=True)\\n net_5_scaled = fluid.layers.elementwise_mul(net_5, global_channels, axis=0)\\n\\n ### Combining the paths\\n net_4 = Upsample(net_4, scale=2)\\n net_5_scaled = Upsample(net_5_scaled, scale=4)\\n # print('net_4, net_5:', [net_4, net_5_scaled])\\n # layers_concat = list()\\n # layers_concat.append(spatial_net)\\n ## layers_concat.append(net_4)\\n # layers_concat.append(net_5_scaled)\\n context_net = fluid.layers.concat([spatial_net, net_4, net_5_scaled], axis=1) #\\n # context_net = fluid.layers.concat(input=layers_concat,axis=1)\\n # print('context_net', context_net)\\n # context_net = fluid.layers.concat([net_4, net_5_scaled], axis=1)\\n # print('context_net', context_net)\\n # context_net = fluid.layers.concat([spatial_net,context_net], axis=1)\\n # print('context_net2',context_net)\\n\\n ### FFM\\n # net = FeatureFusionModule(input_1=spatial_net, input_2=context_net, num_filters=num_classes)\\n net = FeatureFusionModule(inputs=context_net, num_filters=num_classes)\\n\\n # print('net', net)\\n\\n ## [batch_zize, num_filters, 128, 64]\\n\\n ### Final upscaling and finish\\n # net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[256, 128])\\n # print('conv2d_transpose', net)\\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn1')\\n net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[128, 256])\\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn2')\\n net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[256, 512])\\n net = batch_normalization(net, relu=True, name='conv2d_transpose_bn3')\\n #net = fluid.layers.conv2d_transpose(input=net, num_filters=num_classes, output_size=[512, 1024])\\n #net = batch_normalization(net, relu=True, name='conv2d_transpose_bn4')\\n # print('net',net)\\n net = fluid.layers.image_resize(net, out_shape=[512, 1024], resample='BILINEAR')\\n\\n net = fluid.layers.conv2d(net, num_classes, 1)\\n return net\",\n \"def __init__(\\n self,\\n fc_layers: Tuple[int] = (1000,),\\n inverted_residual_setting: Optional[List[List[int]]] = None,\\n block: Optional[Callable[..., nn.Module]] = None,\\n norm_layer: Optional[Callable[..., nn.Module]] = None,\\n c = 3,\\n h = 218,\\n w = 178\\n ) -> None:\\n super(MobileNetV2, self).__init__()\\n if block is None:\\n block = InvertedResidual\\n\\n if norm_layer is None:\\n norm_layer = nn.BatchNorm2d\\n\\n input_channel = c\\n new_h, new_w = h, w\\n\\n if inverted_residual_setting is None:\\n inverted_residual_setting = [\\n # t, c, n, s\\n [1, 16, 1, 1],\\n [6, 24, 2, 2],\\n [6, 32, 3, 2],\\n [6, 64, 4, 2],\\n [6, 96, 3, 1],\\n [6, 160, 3, 2],\\n [6, 320, 1, 1],\\n ]\\n\\n # only check the first element, assuming user knows t,c,n,s are required\\n if len(inverted_residual_setting) == 0 or len(inverted_residual_setting[0]) != 4:\\n raise ValueError(\\\"inverted_residual_setting should be non-empty \\\"\\n \\\"or a 4-element list, got {}\\\".format(inverted_residual_setting))\\n\\n # building first layer\\n features: List[nn.Module] = []\\n # building inverted residual blocks\\n for t, c, n, s in inverted_residual_setting:\\n output_channel = c\\n new_h //= s\\n new_w //= s\\n for i in range(n):\\n stride = s if i == 0 else 1\\n features.append(block(input_channel, output_channel, stride, expand_ratio=t, norm_layer=norm_layer))\\n input_channel = output_channel\\n # make it nn.Sequential\\n self.features = nn.Sequential(*features)\\n\\n # building classifier\\n fc_layers = (new_h*new_w*output_channel,) + fc_layers\\n self.classifier = MLP(fc_layers, end_fn=nn.Identity)\\n\\n # weight initialization\\n init_params(self.modules())\",\n \"def __init__(self, **kwargs):\\n FeatureDefinition.__init__(self)\\n\\n nbTypes = self._getTypeNumber(kwargs)\\n \\n print(\\\"BETTER FEATURES\\\")\\n \\n \\n block_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\\n (\\\"xywh\\\", Pipeline([\\n ('selector', NodeTransformerXYWH_v2()),\\n #v1 ('xywh', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('xywh', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n , (\\\"neighbors\\\", Pipeline([\\n ('selector', NodeTransformerNeighbors()),\\n #v1 ('neighbors', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('neighbors', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n , (\\\"1hot\\\", Pipeline([\\n ('1hot', Node1HotFeatures()) #does the 1-hot encoding directly\\n ])\\n )\\n ])\\n grid_line_transformer = GridLine_NodeTransformer_v2()\\n \\n self._node_transformer = TransformerListByType([block_transformer, grid_line_transformer]) \\n \\n edge_BB_transformer = FeatureUnion( [ #CAREFUL IF YOU CHANGE THIS - see cleanTransformers method!!!!\\n (\\\"1hot\\\", Pipeline([\\n ('1hot', Edge1HotFeatures(PageNumberSimpleSequenciality()))\\n ])\\n )\\n , (\\\"boolean\\\", Pipeline([\\n ('boolean', EdgeBooleanFeatures_v2())\\n ])\\n )\\n , (\\\"numerical\\\", Pipeline([\\n ('selector', EdgeNumericalSelector()),\\n #v1 ('numerical', StandardScaler(copy=False, with_mean=True, with_std=True)) #use in-place scaling\\n ('numerical', QuantileTransformer(n_quantiles=self.n_QUANTILES, copy=False)) #use in-place scaling\\n ])\\n )\\n ] )\\n edge_BL_transformer = DU_ABPTableRG.Block2GridLine_EdgeTransformer()\\n edge_LL_transformer = DU_ABPTableRG.GridLine2GridLine_EdgeTransformer()\\n self._edge_transformer = TransformerListByType([edge_BB_transformer,\\n edge_BL_transformer,\\n edge_BL_transformer, # useless but required\\n edge_LL_transformer \\n ])\\n \\n self.tfidfNodeTextVectorizer = None #tdifNodeTextVectorizer\",\n \"def _InitAttentionParams(self, atten_tpl):\\n p = self.params\\n source_atten_tpls = []\\n # Set up each source attention.\\n for i in range(p.num_source):\\n src_key = 'source_%d' % i\\n src_atten = atten_tpl.Copy()\\n src_atten = super()._InitAttentionParams(src_atten)\\n if isinstance(src_atten, list):\\n raise ValueError(\\n 'TransformerMultiSourceAttentionLayer does not support '\\n 'num_heads > 1.')\\n src_atten.name = 'multihead_atten_%s' % src_key\\n source_atten_tpls.append((src_key, src_atten))\\n\\n # Initialize multi-source attention.\\n msa = p.multi_source_atten.Copy()\\n msa.name = 'multi_source_atten'\\n msa.input_dim = p.input_dim\\n msa.hidden_dim = p.hidden_dim\\n msa.source_atten_tpls = source_atten_tpls\\n msa.primary_source_key = 'source_%d' % p.primary_source_index\\n return msa\",\n \"def __init__(self, block, layers, groups, reduction, dropout_p=0.2, in_channels=3,\\n inplanes=128, input_3x3=True, downsample_kernel_size=3,\\n downsample_padding=1, num_classes=1000):\\n super(SENet, self).__init__()\\n \\n self.in_channels = in_channels\\n self.inplanes = inplanes\\n\\n if input_3x3:\\n layer0_modules = [\\n ('conv1', nn.Conv2d(in_channels, 64, 3, stride=2, padding=1,\\n bias=False)),\\n ('bn1', nn.BatchNorm2d(64)),\\n ('relu1', nn.ReLU(inplace=True)),\\n ('conv2', nn.Conv2d(64, 64, 3, stride=1, padding=1,\\n bias=False)),\\n ('bn2', nn.BatchNorm2d(64)),\\n ('relu2', nn.ReLU(inplace=True)),\\n ('conv3', nn.Conv2d(64, inplanes, 3, stride=1, padding=1,\\n bias=False)),\\n ('bn3', nn.BatchNorm2d(inplanes)),\\n ('relu3', nn.ReLU(inplace=True)),\\n ]\\n else:\\n layer0_modules = [\\n ('conv1', nn.Conv2d(in_channels, inplanes, kernel_size=7, stride=2,\\n padding=3, bias=False)),\\n ('bn1', nn.BatchNorm2d(inplanes)),\\n ('relu1', nn.ReLU(inplace=True)),\\n ]\\n\\n layer0_modules.append(('pool', nn.MaxPool2d(3, stride=2, ceil_mode=True)))\\n \\n self.layer0 = nn.Sequential(OrderedDict(layer0_modules))\\n \\n self.layer1 = self._make_layer(\\n block,\\n planes=64,\\n blocks=layers[0],\\n groups=groups,\\n reduction=reduction,\\n downsample_kernel_size=1,\\n downsample_padding=0\\n )\\n \\n self.layer2 = self._make_layer(\\n block,\\n planes=128,\\n blocks=layers[1],\\n stride=2,\\n groups=groups,\\n reduction=reduction,\\n downsample_kernel_size=downsample_kernel_size,\\n downsample_padding=downsample_padding\\n )\\n \\n self.layer3 = self._make_layer(\\n block,\\n planes=256,\\n blocks=layers[2],\\n stride=2,\\n groups=groups,\\n reduction=reduction,\\n downsample_kernel_size=downsample_kernel_size,\\n downsample_padding=downsample_padding\\n )\\n \\n self.layer4 = self._make_layer(\\n block,\\n planes=512,\\n blocks=layers[3],\\n stride=2,\\n groups=groups,\\n reduction=reduction,\\n downsample_kernel_size=downsample_kernel_size,\\n downsample_padding=downsample_padding\\n )\\n \\n self.avg_pool = nn.AvgPool2d(7, stride=1)\\n self.dropout = nn.Dropout2d(dropout_p) if dropout_p is not None else None\\n self.last_linear = nn.Linear(512 * block.expansion, num_classes)\",\n \"def __init__(self, nheads, d_model):\\n super(MultiheadAttention, self).__init__()\\n assert d_model % nheads == 0\\n self.d_head = d_model // nheads\\n self.nheads = nheads\\n self.Q_fc = nn.Linear(d_model, d_model, bias=False)\\n self.K_fc = nn.Linear(d_model, d_model, bias=False)\\n self.V_fc = nn.Linear(d_model, d_model, bias=False)\\n self.output_fc = nn.Linear(d_model, d_model, bias=False)\\n self.attn = None\",\n \"def block(inp, scope, past, config):\\n with tf.variable_scope(scope):\\n nx = inp.shape[-1].value\\n norm = normalise_tensor(inp, 'ln_1')\\n attn, present = attention(norm, 'attn', nx, past=past, config=config)\\n out = attn + inp\\n norm = normalise_tensor(out, 'ln_2')\\n mlp_out = multilayer_perceptron(norm, 'mlp', nx * 4) # note that hidden dim is 4x\\n out = out + mlp_out\\n return out, present\",\n \"def call(self, inputs, training = False):\\n \\\"\\\"\\\" includes training argument as batch_norm functions differently during training and during inference \\\"\\\"\\\"\\n \\n # input --> block_1 --> block_2 --> block_3 \\n x = self.block_1(inputs, training = training)\\n x = self.block_2(x, training = training)\\n x = self.block_3(x, training = training)\\n\\n # output for MAA2C or DDPG \\n # block_3 --> outputs \\n x = self.outputs(x)\\n\\n return x\",\n \"def identity_block(input_tensor, kernel_size, filters, stage, block):\\n filters0, filters1, filters2 = filters\\n conv_name_base = 'res' + str(stage) + block\\n bn_name_base = 'bn' + str(stage) + block\\n add_name = 'add' + str(stage) + \\\"_\\\" + block\\n relu_name = 'relu' + str(stage) + \\\"_\\\" + block\\n\\n # Tensors\\n input_tensor_chans = input_tensor.dims(\\n 3) if input_tensor.shape.layout == sg.NHWC else input_tensor.dims(1)\\n conv0_tensor = sg.Tensor(\\n data_layout=sg.NHWC, tensor_data=generate_random_data(\\n (filters0, 1, 1, input_tensor_chans)))\\n bn0_mean_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\\n bn0_var_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\\n bn0_gamma_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\\n bn0_beta_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters0)))\\n conv1_tensor = sg.Tensor(\\n data_layout=sg.NHWC, tensor_data=generate_random_data(\\n (filters1, kernel_size, kernel_size, filters0)))\\n bn1_mean_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\\n bn1_var_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\\n bn1_gamma_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\\n bn1_beta_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters1)))\\n conv2_tensor = sg.Tensor(\\n data_layout=sg.NHWC, tensor_data=generate_random_data(\\n (filters2, 1, 1, filters1)))\\n bn2_mean_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\\n bn2_var_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\\n bn2_gamma_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\\n bn2_beta_tensor = sg.Tensor(\\n data_layout=sg.NC, tensor_data=generate_random_data((1, filters2)))\\n\\n x = sg.nn.convolution(\\n input_tensor, conv0_tensor, stride=[1, 1], padding=\\\"same\\\",\\n name=conv_name_base + '_2a')\\n x = sg.nn.batch_norm(\\n x, bn0_mean_tensor, bn0_var_tensor, bn0_gamma_tensor, bn0_beta_tensor,\\n activation=\\\"relu\\\", name=bn_name_base + '_2a')\\n x = sg.nn.convolution(\\n x, conv1_tensor, stride=[1, 1], padding=\\\"same\\\",\\n name=conv_name_base + '_2b')\\n x = sg.nn.batch_norm(\\n x, bn1_mean_tensor, bn1_var_tensor, bn1_gamma_tensor, bn1_beta_tensor,\\n activation=\\\"relu\\\", name=bn_name_base + '_2b')\\n x = sg.nn.convolution(\\n x, conv2_tensor, stride=[1, 1], padding=\\\"same\\\",\\n name=conv_name_base + '_2c')\\n x = sg.nn.batch_norm(\\n x, bn2_mean_tensor, bn2_var_tensor, bn2_gamma_tensor, bn2_beta_tensor,\\n name=bn_name_base + '_2c')\\n x = sg.math.add(x, input_tensor, name=add_name)\\n x = sg.nn.relu(x, name=relu_name)\\n return x\",\n \"def __init__(self,\\n input_dim,\\n dec_seq_len,\\n out_seq_len,\\n d_model=512,\\n nhead=8,\\n num_encoder_layers=6,\\n num_decoder_layers=6,\\n dim_feedforward=2048,\\n dropout=0.1,\\n activation='relu',\\n custom_encoder=None,\\n custom_decoder=None):\\n super(TransformerTS, self).__init__()\\n self.transform = nn.Transformer(\\n d_model=d_model,\\n nhead=nhead,\\n num_encoder_layers=num_encoder_layers,\\n num_decoder_layers=num_decoder_layers,\\n dim_feedforward=dim_feedforward,\\n dropout=dropout,\\n activation=activation,\\n custom_encoder=custom_encoder,\\n custom_decoder=custom_decoder\\n )\\n self.pos = PositionalEncoding(d_model)\\n self.enc_input_fc = nn.Linear(input_dim, d_model)\\n self.dec_input_fc = nn.Linear(input_dim, d_model)\\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\\n self.dec_seq_len = dec_seq_len\",\n \"def __init__(self, emb_size: int = 768, num_heads: int = 12, dropout: float = 0.1):\\n super().__init__()\\n self.emb_size = emb_size\\n self.num_heads = num_heads\\n\\n # Check \\\"head dim\\\" = number of features per head (d_k).\\n self.head_dim = emb_size // num_heads\\n assert self.head_dim * num_heads == self.emb_size, \\\"emb_size must be divisible by num_heads\\\"\\n # Calculate scaling factor.\\n self.scale = 1 / (self.head_dim ** 0.5)\\n\\n # V1: in vanilla self-attention Q,K,V are square matrices.\\n # self.keys = nn.Linear(emb_size, emb_size)\\n # self.queries = nn.Linear(emb_size, emb_size)\\n # self.values = nn.Linear(emb_size, emb_size)\\n \\n # V2: single layer with emb_size, split into num (heads * head_dim) * 3 (Q,K,V).\\n self.qkv = nn.Linear(emb_size, emb_size * 3)\\n\\n # Attention dropout.\\n self.att_drop = nn.Dropout(dropout)\\n self.projection = nn.Linear(emb_size, emb_size)\",\n \"def _fp32_mte_process_lt_one_block(axis_1_lp_index, sub_axis_1):\\n\\n def _fp32_inner_last_dim_lt_one_block(axis_0_lp_index, sub_axis_0):\\n \\\"\\\"\\\"\\n inner process of last dim less than one block\\n \\\"\\\"\\\"\\n\\n # move data in\\n in_offset = (block_idx * per_core_col_size +\\n axis_1_lp_index * max_core_axis_size +\\n axis_0_lp_index * max_no_core_axis_size * axis_1) * axis_2\\n data_pos_info = (sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, in_offset)\\n _data_move_in_last_dim_lt_one_block(tik_inst, ub_input, data_in, data_pos_info)\\n\\n # do transpose\\n with tik_inst.new_stmt_scope():\\n temp_sub_axis_1 = tik_inst.Scalar(\\\"int64\\\")\\n temp_sub_axis_0 = tik_inst.Scalar(\\\"int64\\\")\\n data_size_one_block = _get_elment_cnt_one_block(data_in.dtype)\\n axis_1_0_2_size = axis_0 * axis_1 * axis_2\\n sub_axis_1_0_2_size = sub_axis_1 * sub_axis_0 * axis_2\\n\\n # to avoid multiple core dirty data\\n with tik_inst.if_scope(tik.all(sub_axis_1_0_2_size < data_size_one_block,\\n axis_1_0_2_size > data_size_one_block)):\\n with tik_inst.if_scope(sub_axis_1 == 1):\\n temp_sub_axis_0.set_as(_ceil_div(data_size_one_block, axis_2))\\n temp_sub_axis_1.set_as(sub_axis_1)\\n with tik_inst.else_scope():\\n temp_sub_axis_0.set_as(sub_axis_0)\\n temp_sub_axis_1.set_as(_ceil_div(data_size_one_block,\\n axis_0 * axis_2))\\n with tik_inst.else_scope():\\n temp_sub_axis_1.set_as(sub_axis_1)\\n temp_sub_axis_0.set_as(sub_axis_0)\\n\\n sub_dim_size = (temp_sub_axis_1, temp_sub_axis_0, axis_2)\\n _transpose_by_2_vnchwconv_not_last_dim(tik_inst, ub_input[ub_offset],\\n ub_input, sub_dim_size)\\n\\n # move data out\\n out_offset = ((block_idx * per_core_col_size +\\n axis_1_lp_index * max_core_axis_size) * axis_0 +\\n axis_0_lp_index * max_no_core_axis_size) * axis_2\\n data_pos_info = (sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, out_offset)\\n _data_move_out_last_dim_lt_one_block(tik_inst, data_out, ub_input[ub_offset],\\n data_pos_info)\\n\\n with tik_inst.for_range(0, no_core_loop_cnt) as axis_0_lp_idx:\\n _fp32_inner_last_dim_lt_one_block(axis_0_lp_idx, max_no_core_axis_size)\\n with tik_inst.if_scope(no_core_left > 0):\\n _fp32_inner_last_dim_lt_one_block(no_core_loop_cnt, no_core_left)\",\n \"def __init__(self, config: BertConfig):\\r\\n super().__init__(config)\\r\\n ### YOUR CODE HERE\\r\\n self.num_labels = config.num_labels # [0, 1] (start or end)\\r\\n self.bert = BertModel(config)\\r\\n self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels) # TODO: Not a separate FFN ? (For Start_FFN and End_FFN)\\r\\n\\r\\n ### END YOUR CODE\\r\\n\\r\\n # Don't forget initializing the weights\\r\\n self.init_weights()\",\n \"def __init__(\\n self,\\n num_frames = 2,\\n dim = 64,\\n channels = 3,\\n stages = 4,\\n num_blocks = 2,\\n dim_head = 64,\\n window_size = 16,\\n heads = 8,\\n ff_mult = 4,\\n input_channels = None,\\n output_channels = None,\\n device = None,\\n skip_type = 'ConcatCross'\\n ):\\n super().__init__()\\n input_channels = channels if input_channels is None else input_channels\\n output_channels = channels if output_channels is None else output_channels\\n\\n self.project_in = nn.Sequential(\\n nn.Conv3d(input_channels, dim, kernel_size = (1, 3, 3), padding = (0, 1, 1)),\\n nn.GELU()\\n )\\n\\n self.project_out = nn.Sequential(\\n nn.Conv3d(dim, output_channels, kernel_size = (1, 3, 3), padding = (0, 1, 1))\\n )\\n\\n self.downs = nn.ModuleList([])\\n self.ups = nn.ModuleList([])\\n\\n # render each hyperparam a tuple so that we can iterate thru them\\n heads, window_size, dim_head, num_blocks = map(partial(cast_tuple, depth = stages), (heads, window_size, dim_head, num_blocks))\\n\\n for ind, heads, window_size, dim_head, num_blocks in zip(range(stages), heads, window_size, dim_head, num_blocks):\\n is_last = ind == (stages - 1)\\n\\n self.downs.append(nn.ModuleList([ # need 3d version of this\\n Block(dim, depth = num_blocks, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size, device = device),\\n nn.Conv3d(dim, dim * 2, kernel_size = (1, 4, 4), stride = (1, 2, 2), padding = (0, 1, 1))\\n ]))\\n\\n self.ups.append(nn.ModuleList([\\n nn.ConvTranspose3d(dim * 2, dim, (1, 2, 2), stride = (1, 2, 2)),\\n Block(dim, depth = num_blocks, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size, device = device, skip_type = skip_type)\\n ]))\\n\\n dim *= 2\\n if is_last:\\n self.mid = nn.Sequential(\\n Rearrange('b c f h w -> b (c f) 1 h w'),\\n Block(dim = dim*num_frames, depth = 1, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size),\\n nn.Conv3d(dim*num_frames, dim, kernel_size = (1, 3, 3), padding = (0, 1, 1)),\\n Block(dim = dim, depth = 2, dim_head = dim_head, heads = heads, ff_mult = ff_mult, window_size = window_size),\\n )\",\n \"def __init__(self,\\n num_heads,\\n seq_len,\\n block=16,\\n different_layout_per_head=False,\\n num_random_blocks=0,\\n local_window_blocks=[4],\\n global_block_indices=[0],\\n global_block_end_indices=None,\\n attention='bidirectional',\\n horizental_global_attention=False):\\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\\n self.num_random_blocks = num_random_blocks\\n all_local_blocks = 0\\n for local_blocks in local_window_blocks:\\n all_local_blocks += local_blocks\\n if (self.num_blocks < all_local_blocks):\\n raise ValueError(\\n f'Number of all local window blocks, {all_local_blocks}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\\n )\\n self.local_window_blocks = local_window_blocks\\n if (self.num_blocks < len(global_block_indices)):\\n raise ValueError(\\n f'Number of global blocks indices, {global_block_indices}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\\n )\\n for idx in global_block_indices:\\n if idx >= self.num_blocks:\\n raise ValueError(\\n f'Global block index, {global_block_indices[idx]}, must be smaller than overal number of blocks in a row, {self.num_blocks}!'\\n )\\n self.global_block_indices = global_block_indices\\n\\n if (global_block_end_indices is not None):\\n if (len(global_block_indices) != len(global_block_end_indices)):\\n raise ValueError(\\n f'Global block start indices length, {len(global_block_indices)}, must be same as global block end indices length, {len(global_block_end_indices)}!'\\n )\\n for _, (start_idx, end_idx) in enumerate(zip(global_block_indices, global_block_end_indices)):\\n if end_idx > self.num_blocks:\\n raise ValueError(\\n f'Global block end index, {global_block_end_indices[idx]}, must be smaller (equal) than overal number of blocks in a row, {self.num_blocks}!'\\n )\\n if start_idx >= end_idx:\\n raise ValueError(\\n f'Global block start index, {start_idx}, must be smaller than global block end index, {end_idx}!'\\n )\\n self.global_block_end_indices = global_block_end_indices\\n\\n if (attention != 'unidirectional' and attention != 'bidirectional'):\\n raise NotImplementedError(\\n 'only \\\\\\\"uni/bi-directional\\\\\\\" attentions are supported for now!')\\n self.attention = attention\\n\\n if (attention != 'bidirectional' and horizental_global_attention):\\n raise ValueError(\\n 'only \\\\\\\"bi-directional\\\\\\\" attentions can support horizental global attention!'\\n )\\n self.horizental_global_attention = horizental_global_attention\\n self.make_layout()\",\n \"def inception_block_1a(X):\\n\\n X_3x3 = Conv2D(96, (1, 1), data_format='channels_first', name='inception_3a_3x3_conv1')(X)\\n X_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn1')(X_3x3)\\n X_3x3 = Activation('relu')(X_3x3)\\n X_3x3 = ZeroPadding2D(padding=(1, 1), data_format='channels_first')(X_3x3)\\n X_3x3 = Conv2D(128, (3, 3), data_format='channels_first', name='inception_3a_3x3_conv2')(X_3x3)\\n X_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn2')(X_3x3)\\n X_3x3 = Activation('relu')(X_3x3)\\n\\n X_5x5 = Conv2D(16, (1, 1), data_format='channels_first', name='inception_3a_5x5_conv1')(X)\\n X_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn1')(X_5x5)\\n X_5x5 = Activation('relu')(X_5x5)\\n X_5x5 = ZeroPadding2D(padding=(2, 2), data_format='channels_first')(X_5x5)\\n X_5x5 = Conv2D(32, (5, 5), data_format='channels_first', name='inception_3a_5x5_conv2')(X_5x5)\\n X_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn2')(X_5x5)\\n X_5x5 = Activation('relu')(X_5x5)\\n\\n X_pool = MaxPooling2D(pool_size=3, strides=2, data_format='channels_first')(X)\\n X_pool = Conv2D(32, (1, 1), data_format='channels_first', name='inception_3a_pool_conv')(X_pool)\\n X_pool = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_pool_bn')(X_pool)\\n X_pool = Activation('relu')(X_pool)\\n X_pool = ZeroPadding2D(padding=((3, 4), (3, 4)), data_format='channels_first')(X_pool)\\n\\n X_1x1 = Conv2D(64, (1, 1), data_format='channels_first', name='inception_3a_1x1_conv')(X)\\n X_1x1 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_1x1_bn')(X_1x1)\\n X_1x1 = Activation('relu')(X_1x1)\\n\\n # CONCAT\\n inception = concatenate([X_3x3, X_5x5, X_pool, X_1x1], axis=1)\\n\\n return inception\",\n \"def __init__(\\n self,\\n include_top=True,\\n weights='imagenet',\\n input_tensor=None,\\n input_shape=None,\\n pooling=None,\\n classes=1000,\\n **kwargs):\\n backend, layers, models, keras_utils = get_submodules_from_kwargs(kwargs)\\n self.include_top = include_top\\n self.pooling = pooling\\n self.weights = weights\\n self.backend = backend\\n self.layers = layers\\n self.classes = classes\\n\\n if not (weights in {'imagenet', None} or os.path.exists(weights)):\\n raise ValueError('The `weights` argument should be either '\\n '`None` (random initialization), `imagenet` '\\n '(pre-training on ImageNet), '\\n 'or the path to the weights file to be loaded.')\\n\\n if weights == 'imagenet' and include_top and classes != 1000:\\n raise ValueError('If using `weights` as `\\\"imagenet\\\"` with `include_top`'\\n ' as true, `classes` should be 1000')\\n self.block1_conv1 = []\\n self.block1_conv2 = []\\n self.block1_pool = []\\n\\n self.block2_conv1 = []\\n self.block2_conv2 = []\\n self.block2_pool = []\\n\\n self.block3_conv1 = []\\n self.block3_conv2 = []\\n self.block3_conv3 = []\\n self.block3_conv4 = []\\n self.block3_pool = []\\n\\n self.block4_conv1 = []\\n self.block4_conv2 = []\\n self.block4_conv3 = []\\n self.block4_conv4 = []\\n self.block4_pool = []\\n\\n self.block5_conv1 = []\\n self.block5_conv2 = []\\n self.block5_conv3 = []\\n self.block5_conv4 = []\\n self.block5_pool = []\\n\\n for i in xrange(FLAGS.num_replica):\\n # Block 1\\n self.block1_conv1.append(layers.Conv2D(64, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block1_conv1'))\\n self.block1_conv2.append(layers.Conv2D(64, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block1_conv2'))\\n self.block1_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block1_pool'))\\n\\n # Block 2\\n self.block2_conv1.append(layers.Conv2D(128, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block2_conv1'))\\n self.block2_conv2.append(layers.Conv2D(128, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block2_conv2'))\\n self.block2_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block2_pool'))\\n\\n # Block 3\\n self.block3_conv1.append(layers.Conv2D(256, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block3_conv1'))\\n self.block3_conv2.append(layers.Conv2D(256, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block3_conv2'))\\n self.block3_conv3.append(layers.Conv2D(256, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block3_conv3'))\\n self.block3_conv4.append(layers.Conv2D(256, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block3_conv4'))\\n self.block3_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block3_pool'))\\n\\n # Block 4\\n self.block4_conv1.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block4_conv1'))\\n self.block4_conv2.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block4_conv2'))\\n self.block4_conv3.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block4_conv3'))\\n self.block4_conv4.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block4_conv4'))\\n self.block4_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block4_pool'))\\n\\n # Block 5\\n self.block5_conv1.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block5_conv1'))\\n self.block5_conv2.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block5_conv2'))\\n self.block5_conv3.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block5_conv3'))\\n self.block5_conv4.append(layers.Conv2D(512, (3, 3),\\n activation='relu',\\n padding='same',\\n name='block5_conv4'))\\n self.block5_pool.append(layers.MaxPooling2D((2, 2), strides=(2, 2), name='block5_pool'))\\n\\n if include_top:\\n # Classification block\\n self.flatten = layers.Flatten(name='flatten')\\n self.fc1 = layers.Dense(4096, activation='relu', name='fc1')\\n self.fc2 = layers.Dense(4096, activation='relu', name='fc2')\\n self.predict = layers.Dense(classes, activation='softmax', name='predictions')\\n else:\\n if pooling == 'avg':\\n self.pool = layers.GlobalAveragePooling2D()\\n elif pooling == 'max':\\n self.pool = layers.GlobalMaxPooling2D()\",\n \"def _init_layers(self) -> None:\\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\\n self.embed_dims = self.self_attn.embed_dims\\n self.ffn = FFN(**self.ffn_cfg)\\n norms_list = [\\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\\n for _ in range(2)\\n ]\\n self.norms = ModuleList(norms_list)\",\n \"def mtf_transformer_lm_moe():\\n hparams = mtf_transformer.mtf_transformer_lm_baseline()\\n hparams.decoder_layers = [\\\"att\\\", \\\"moe\\\"] * 4\\n moe.set_default_moe_hparams(hparams)\\n hparams.mesh_shape = \\\"all:8\\\"\\n hparams.layout = \\\"batch:all;experts:all\\\"\\n return hparams\",\n \"def init_layers(self):\\n self.user_emb = nn.Embedding(self.n_users, self.emb_dim)\\n self.item_emb = nn.Embedding(self.n_items, self.emb_dim)\\n init_range = 0.1 * (self.emb_dim) ** (-1 / 2)\\n self.item_emb.weight.data.uniform_(-init_range, init_range)\\n self.fc_u_1_mu = nn.Linear(self.user_fea_dim, self.late_dim)\\n self.fc_u_2_mu = nn.Linear(self.late_dim, self.emb_dim * 2)\\n self.fc_i_1_mu = nn.Linear(self.item_fea_dim, self.late_dim)\\n self.fc_i_2_mu = nn.Linear(self.late_dim, self.emb_dim * 2)\",\n \"def _indentity_block(self, X, filters, f, stage, block):\\n\\t\\tconv_layer_name = 'res' + str(stage) + block + '_branch'\\n\\t\\tbn_layer_name = 'bm' + str(stage) + block + '_branch'\\n\\n\\t\\tX_shortcut = X\\n\\n\\t\\tF1, F2, F3 = filters\\n\\n\\t\\t# First component of main path\\n\\t\\tX = Conv2D(filters = F1, kernel_size = (1, 1), strides = (1, 1), padding = 'valid',\\n\\t\\t\\tname = conv_layer_name + '2a', kernel_initializer = glorot_uniform(seed=0))(X)\\n\\t\\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2a')(X)\\n\\t\\tX = Activation('relu')(X)\\n\\n\\t\\t# Second component of main path\\n\\t\\tX = Conv2D(filters = F2, kernel_size = (1, 1), strides = (1, 1), padding = 'same',\\n\\t\\t\\tname = conv_layer_name + '2b', kernel_initializer = glorot_uniform(seed=0))(X)\\n\\t\\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2b')(X)\\n\\t\\tX = Activation('relu')(X)\\n\\n\\t\\t# Third component of main path\\n\\t\\tX = Conv2D(filters = F3, kernel_size = (1, 1), strides = (1, 1), padding = 'valid',\\n\\t\\t\\tname = conv_layer_name + '2c', kernel_initializer = glorot_uniform(seed=0))(X)\\n\\t\\tX = BatchNormalization(axis = 3, name = bn_layer_name + '2c')(X)\\n\\n\\t\\t# Final step : adding the shortcut componet to X and applying 'relu' activation on the combination \\n\\t\\tX = Add()([X_shortcut,X])\\n\\t\\tX = Activation('relu')(X)\\n\\n\\t\\treturn X\",\n \"def identity_block(filters, stage, block):\\n\\n def layer(input_tensor):\\n conv_params = get_conv_params()\\n bn_params = get_bn_params()\\n conv_name, bn_name, relu_name, sc_name = handle_block_names(stage, block)\\n\\n x = BatchNormalization(name=bn_name + '0', **bn_params)(input_tensor)\\n x = Activation('relu', name=relu_name + '0')(x)\\n x = Conv2D(filters // 2, (1, 1), name=conv_name + '1', **conv_params)(x)\\n x = BatchNormalization(name=bn_name + '1', **bn_params)(x)\\n x = Activation('relu', name=relu_name + '1')(x)\\n\\n x = ZeroPadding2D(padding=(1, 1))(x)\\n x = GroupConv2D(filters // 2, (3, 3), conv_params, conv_name + '2')(x)\\n x = BatchNormalization(name=bn_name + '2', **bn_params)(x)\\n x = Activation('relu', name=relu_name + '2')(x)\\n\\n # x = Conv2D(filters * 2, (1, 1), name=conv_name + '3', **conv_params)(x)\\n x = Conv2D(filters, (1, 1), name=conv_name + '3', **conv_params)(x)\\n x = BatchNormalization(name=bn_name + '3', **bn_params)(x)\\n x = Add()([x, input_tensor])\\n\\n return x\\n\\n return layer\",\n \"def __init__(self, embed_dim: int, hidden_dim: int, num_max_positions: int, num_heads: int, num_layers: int, dropout: float, causal: bool,\\n initializer_range: float):\\n super().__init__()\\n tokenizer = BertTokenizer.from_pretrained('bert-base-cased', do_lower_case=False)\\n num_embeddings = len(tokenizer.vocab)\\n self.initializer_range = initializer_range\\n self.transformer = Transformer(embed_dim, hidden_dim, num_embeddings,\\n num_max_positions, num_heads, num_layers,\\n dropout, causal=causal)\\n\\n self.lm_head = torch.nn.Linear(embed_dim, num_embeddings, bias=False)\\n self.apply(self.init_weights)\\n self.tie_weights()\",\n \"def blocks2modules(blocks, net_info, device=\\\"cpu\\\"):\\n modules = torch.nn.ModuleList()\\n\\n # Track number of channels (filters) in the output of each layer; this\\n # is necessary to determine layer input/output shapes for various layers.\\n curr_out_channels = None\\n prev_layer_out_channels = net_info[\\\"channels\\\"]\\n out_channels_list = []\\n\\n for i, block in enumerate(blocks):\\n module = torch.nn.Sequential()\\n\\n if block[\\\"type\\\"] == \\\"convolutional\\\":\\n batch_normalize = \\\"batch_normalize\\\" in block\\n bias = not batch_normalize\\n kernel_size = block[\\\"size\\\"]\\n padding = (kernel_size - 1) // 2 if \\\"pad\\\" in block else 0\\n in_channels = prev_layer_out_channels\\n out_channels = block[\\\"filters\\\"]\\n\\n conv = torch.nn.Conv2d(\\n in_channels=in_channels, out_channels=out_channels,\\n kernel_size=kernel_size, stride=block[\\\"stride\\\"],\\n padding=padding, bias=bias\\n )\\n module.add_module(\\\"conv_{}\\\".format(i), conv)\\n\\n if batch_normalize:\\n bn = torch.nn.BatchNorm2d(num_features=out_channels)\\n module.add_module(\\\"batch_norm_{}\\\".format(i), bn)\\n\\n if block[\\\"activation\\\"] == \\\"leaky\\\":\\n acti = torch.nn.LeakyReLU(negative_slope=0.1, inplace=True)\\n module.add_module(\\\"leaky_{}\\\".format(i), acti)\\n elif block[\\\"activation\\\"] == \\\"linear\\\":\\n # NOTE: Darknet src files call out \\\"linear\\\" vs \\\"relu\\\" but we\\n # use ReLU here.\\n acti = torch.nn.ReLU(inplace=True)\\n\\n # Update the number of current (output) channels.\\n curr_out_channels = out_channels\\n\\n elif block[\\\"type\\\"] == \\\"maxpool\\\":\\n stride = block[\\\"stride\\\"]\\n maxpool = MaxPool2d(\\n kernel_size=block[\\\"size\\\"], stride=stride\\n )\\n module.add_module(\\\"maxpool_{}\\\".format(i), maxpool)\\n\\n elif block[\\\"type\\\"] == \\\"route\\\":\\n # Route layer concatenates outputs along channel dim; add dummy\\n # layer and handle the actual logic in Darknet.forward().\\n module.add_module(\\\"route_{}\\\".format(i), DummyLayer())\\n\\n out_channels = sum(\\n out_channels_list[layer_idx] for layer_idx in block[\\\"layers\\\"]\\n )\\n\\n curr_out_channels = out_channels\\n\\n elif block[\\\"type\\\"] == \\\"shortcut\\\":\\n # Shortcut layer sums outputs from previous layers; add dummy\\n # layer and handle the actual logic in Darknet.forward().\\n module.add_module(\\\"shortcut_{}\\\".format(i), DummyLayer())\\n\\n if \\\"activation\\\" in block:\\n if block[\\\"activation\\\"] == \\\"leaky\\\":\\n acti = torch.nn.LeakyReLU(negative_slope=0.1, inplace=True)\\n module.add_module(\\\"leaky_{}\\\".format(i), acti)\\n elif block[\\\"activation\\\"] == \\\"linear\\\":\\n acti = torch.nn.ReLU(inplace=True)\\n\\n assert out_channels == out_channels_list[i + block[\\\"from\\\"]]\\n curr_out_channels = out_channels\\n\\n elif block[\\\"type\\\"] == \\\"upsample\\\":\\n # NOTE: torch.nn.Upsample is deprecated in favor of Interpolate;\\n # consider using this and/or other interpolation methods?\\n upsample = torch.nn.Upsample(\\n scale_factor=block[\\\"stride\\\"], mode=\\\"nearest\\\"\\n )\\n module.add_module(\\\"upsample_{}\\\".format(i), upsample)\\n\\n elif block[\\\"type\\\"] == \\\"yolo\\\":\\n yolo = YOLOLayer(block[\\\"anchors\\\"], block[\\\"mask\\\"], device=device)\\n module.add_module(\\\"yolo_{}\\\".format(i), yolo)\\n\\n modules.append(module)\\n prev_layer_out_channels = curr_out_channels\\n out_channels_list.append(curr_out_channels)\\n\\n return modules\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\n super().__init__()\\n self.mha = MultiHeadAttention(dm, h)\\n self.ffn = point_wise_feed_forward_network(dm, hidden)\\n\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\",\n \"def __init__(self, hidden_layers, dropout=0.1, activation='gelu'):\\n super(MLPBlock, self).__init__()\\n\\n layers = []\\n for num_units in hidden_layers:\\n layers.extend([\\n Dense(num_units, activation=activation),\\n Dropout(dropout)\\n ])\\n\\n self.mlp = Sequential(layers)\",\n \"def __init__(self, cfg, momentum=False):\\n\\n super().__init__()\\n self.ent_output_size = cfg.ent_output_size\\n self.span_batch_size = cfg.span_batch_size\\n self.position_embedding_dims = cfg.position_embedding_dims\\n self.att_size = cfg.att_size\\n self.momentum = momentum\\n self.activation = nn.GELU()\\n self.device = cfg.device\\n\\n self.bert_encoder = BertEncoder(bert_model_name=cfg.bert_model_name,\\n trainable=cfg.fine_tune,\\n output_size=cfg.bert_output_size,\\n activation=self.activation)\\n\\n self.entity_span_extractor = CNNSpanExtractor(\\n input_size=self.bert_encoder.get_output_dims(),\\n num_filters=cfg.entity_cnn_output_channels,\\n ngram_filter_sizes=cfg.entity_cnn_kernel_sizes,\\n dropout=cfg.dropout)\\n\\n if self.ent_output_size > 0:\\n self.ent2hidden = BertLinear(input_size=self.entity_span_extractor.get_output_dims(),\\n output_size=self.ent_output_size,\\n activation=self.activation,\\n dropout=cfg.dropout)\\n else:\\n self.ent_output_size = self.entity_span_extractor.get_output_dims()\\n self.ent2hidden = lambda x: x\\n\\n self.entity_span_mlp = BertLinear(input_size=self.ent_output_size,\\n output_size=self.ent_output_size,\\n activation=self.activation,\\n dropout=cfg.dropout)\\n self.entity_span_decoder = VanillaSoftmaxDecoder(hidden_size=self.ent_output_size,\\n label_size=6)\\n\\n self.global_position_embedding = nn.Embedding(150, 200)\\n self.global_position_embedding.weight.data.normal_(mean=0.0, std=0.02)\\n self.masked_token_mlp = BertLinear(input_size=self.bert_encoder.get_output_dims() + 200,\\n output_size=self.bert_encoder.get_output_dims(),\\n activation=self.activation,\\n dropout=cfg.dropout)\\n self.masked_token_decoder = nn.Linear(self.bert_encoder.get_output_dims(),\\n 28996,\\n bias=False)\\n self.masked_token_decoder.weight.data.normal_(mean=0.0, std=0.02)\\n self.masked_token_decoder_bias = nn.Parameter(torch.zeros(28996))\\n\\n self.position_embedding = nn.Embedding(7, self.position_embedding_dims)\\n self.position_embedding.weight.data.normal_(mean=0.0, std=0.02)\\n self.attention_encoder = PosAwareAttEncoder(self.ent_output_size,\\n self.bert_encoder.get_output_dims(),\\n 2 * self.position_embedding_dims,\\n self.att_size,\\n activation=self.activation,\\n dropout=cfg.dropout)\\n\\n self.mlp_head1 = BertLinear(self.ent_output_size,\\n self.bert_encoder.get_output_dims(),\\n activation=self.activation,\\n dropout=cfg.dropout)\\n self.mlp_head2 = BertLinear(self.bert_encoder.get_output_dims(),\\n self.bert_encoder.get_output_dims(),\\n activation=self.activation,\\n dropout=cfg.dropout)\\n\\n self.masked_token_loss = nn.CrossEntropyLoss()\",\n \"def __init__(self, input_dim=6, hidden_dim=128, T=10, in_features=2*128 + 16, num_layers=1, out_features=1):\\n super(InputAttentionEncoder, self).__init__()\\n self.input_dim = input_dim\\n self.hidden_dim = hidden_dim\\n self.T = T\\n\\n in_features = hidden_dim * 2 + T - 1\\n\\n self.lstm_unit = nn.LSTM(input_size=input_dim, hidden_size=hidden_dim, num_layers=num_layers)\\n self.attn_linear = nn.Linear(in_features=in_features, out_features=out_features)\\n self.bn1 = nn.BatchNorm1d(out_features)\\n self.fc2 = nn.Linear(out_features, out_features)\",\n \"def __init__(self, wav2vec2_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([wav2vec2_layer.attention.q_proj.weight, wav2vec2_layer.attention.k_proj.weight, wav2vec2_layer.attention.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([wav2vec2_layer.attention.q_proj.bias, wav2vec2_layer.attention.k_proj.bias, wav2vec2_layer.attention.v_proj.bias]))\\n self.out_proj_weight = wav2vec2_layer.attention.out_proj.weight\\n self.out_proj_bias = wav2vec2_layer.attention.out_proj.bias\\n self.linear1_weight = wav2vec2_layer.feed_forward.intermediate_dense.weight\\n self.linear1_bias = wav2vec2_layer.feed_forward.intermediate_dense.bias\\n self.linear2_weight = wav2vec2_layer.feed_forward.output_dense.weight\\n self.linear2_bias = wav2vec2_layer.feed_forward.output_dense.bias\\n self.norm1_eps = wav2vec2_layer.layer_norm.eps\\n self.norm1_weight = wav2vec2_layer.layer_norm.weight\\n self.norm1_bias = wav2vec2_layer.layer_norm.bias\\n self.norm2_eps = wav2vec2_layer.final_layer_norm.eps\\n self.norm2_weight = wav2vec2_layer.final_layer_norm.weight\\n self.norm2_bias = wav2vec2_layer.final_layer_norm.bias\\n self.num_heads = wav2vec2_layer.attention.num_heads\\n self.embed_dim = wav2vec2_layer.attention.embed_dim\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def __init__(self,\\r\\n in_channels_1, in_channels_2, out_channels,\\r\\n kernel_size_1, kernel_size_2,\\r\\n stride_1, up_stride_2,\\r\\n padding_1, up_padding_2, output_padding=0,\\r\\n activation_in='relu', activation_out='lrelu',\\r\\n norm_in='bn', norm_out='none'):\\r\\n\\r\\n super(RefinementBlock, self).__init__()\\r\\n\\r\\n self.conv_1 = conv_block(\\r\\n in_channels=in_channels_1,\\r\\n out_channels=out_channels,\\r\\n kernel_size=kernel_size_1,\\r\\n stride=stride_1,\\r\\n padding=padding_1,\\r\\n norm='none',\\r\\n activation=activation_in\\r\\n )\\r\\n\\r\\n self.upconv_2 = upconv_block(\\r\\n in_channels=in_channels_2,\\r\\n out_channels=out_channels,\\r\\n kernel_size=kernel_size_2,\\r\\n stride=up_stride_2,\\r\\n padding=up_padding_2,\\r\\n output_padding=output_padding,\\r\\n norm=norm_in,\\r\\n activation='none'\\r\\n )\\r\\n\\r\\n self.conv_3 = conv_block(\\r\\n in_channels=out_channels,\\r\\n out_channels=out_channels,\\r\\n kernel_size=3,\\r\\n stride=1,\\r\\n padding=1,\\r\\n norm='none',\\r\\n activation='none'\\r\\n )\\r\\n\\r\\n self.conv_4 = conv_block(\\r\\n in_channels=out_channels,\\r\\n out_channels=out_channels,\\r\\n kernel_size=3,\\r\\n stride=1,\\r\\n padding=1,\\r\\n norm='none',\\r\\n activation='none'\\r\\n )\\r\\n\\r\\n self.out_act = _activation(act_type=activation_out)\\r\\n self.out_norm = _norm(norm_type=norm_out, channels=out_channels)\",\n \"def __init__(self, albert_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([albert_layer.attention.query.weight, albert_layer.attention.key.weight, albert_layer.attention.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([albert_layer.attention.query.bias, albert_layer.attention.key.bias, albert_layer.attention.value.bias]))\\n self.out_proj_weight = albert_layer.attention.dense.weight\\n self.out_proj_bias = albert_layer.attention.dense.bias\\n self.linear1_weight = albert_layer.ffn.weight\\n self.linear1_bias = albert_layer.ffn.bias\\n self.linear2_weight = albert_layer.ffn_output.weight\\n self.linear2_bias = albert_layer.ffn_output.bias\\n self.norm1_eps = albert_layer.attention.LayerNorm.eps\\n self.norm1_weight = albert_layer.attention.LayerNorm.weight\\n self.norm1_bias = albert_layer.attention.LayerNorm.bias\\n self.norm2_eps = albert_layer.full_layer_layer_norm.eps\\n self.norm2_weight = albert_layer.full_layer_layer_norm.weight\\n self.norm2_bias = albert_layer.full_layer_layer_norm.bias\\n self.num_heads = albert_layer.attention.num_attention_heads\\n self.embed_dim = albert_layer.attention.all_head_size\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def forward(self, x):\\n\\n if x.dim() == 3:\\n x = x.unsqueeze(1)\\n x = x.transpose(1, 3)\\n x = self.norm0(x)\\n x = x.transpose(1, 3)\\n\\n x = self.conv_block1(x, pool_size=(2, 2), pool_type=\\\"avg\\\")\\n x = F.dropout(x, p=0.2, training=self.training)\\n x = self.conv_block2(x, pool_size=(2, 2), pool_type=\\\"avg\\\")\\n x = F.dropout(x, p=0.2, training=self.training)\\n x = self.conv_block3(x, pool_size=(2, 2), pool_type=\\\"avg\\\")\\n x = F.dropout(x, p=0.2, training=self.training)\\n x3_out = self.conv_block4(x, pool_size=(2, 2), pool_type=\\\"avg\\\")\\n x = F.dropout(x3_out, p=0.2, training=self.training)\\n x2_out = self.conv_block5(x, pool_size=(2, 2), pool_type=\\\"avg\\\")\\n x = F.dropout(x2_out, p=0.2, training=self.training)\\n x1_out = self.conv_block6(x, pool_size=(1, 1), pool_type=\\\"avg\\\")\\n x = F.dropout(x1_out, p=0.2, training=self.training)\\n x = torch.mean(x, dim=3)\\n\\n (x1, _) = torch.max(x, dim=2)\\n x2 = torch.mean(x, dim=2)\\n x = x1 + x2\\n\\n # [B x 1 x emb_dim]\\n if not self.return_reps:\\n return x.unsqueeze(1)\\n\\n return x.unsqueeze(1), (x1_out, x2_out, x3_out)\",\n \"def __init__(self, whisper_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([whisper_layer.self_attn.q_proj.weight, whisper_layer.self_attn.k_proj.weight, whisper_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([whisper_layer.self_attn.q_proj.bias, torch.zeros_like(whisper_layer.self_attn.q_proj.bias), whisper_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = whisper_layer.self_attn.out_proj.weight\\n self.out_proj_bias = whisper_layer.self_attn.out_proj.bias\\n self.linear1_weight = whisper_layer.fc1.weight\\n self.linear1_bias = whisper_layer.fc1.bias\\n self.linear2_weight = whisper_layer.fc2.weight\\n self.linear2_bias = whisper_layer.fc2.bias\\n self.norm1_eps = whisper_layer.self_attn_layer_norm.eps\\n self.norm1_weight = whisper_layer.self_attn_layer_norm.weight\\n self.norm1_bias = whisper_layer.self_attn_layer_norm.bias\\n self.norm2_eps = whisper_layer.final_layer_norm.eps\\n self.norm2_weight = whisper_layer.final_layer_norm.weight\\n self.norm2_bias = whisper_layer.final_layer_norm.bias\\n self.num_heads = whisper_layer.self_attn.num_heads\\n self.embed_dim = whisper_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def forward_test(self, imgs, token_ids=None, segment_ids=None, input_mask=None, **kwargs):\\n imgs = imgs.reshape((-1,) + imgs.shape[2:])\\n if self.from_scratch:\\n imgs = imgs / 255.0\\n visual_token = self.extract_visual_feat(imgs) # b, d, T, h, w\\n B, D, T, H, W = visual_token.shape\\n B_text = token_ids.shape[0]\\n if B_text != B:\\n visual_token = visual_token.view(B_text, -1, D, T, H, W)\\n visual_token = visual_token.mean(dim=1)\\n B = B_text\\n token_ids = token_ids.reshape((-1, ) + token_ids.shape[2:])\\n segment_ids = segment_ids.reshape((-1, ) + segment_ids.shape[2:])\\n input_mask = input_mask.reshape((-1, ) + input_mask.shape[2:])\\n # only use the multimodal transformer for \\n if self.separate_test:\\n # text_only_out = self.multimodal_backbone.forward_text(token_ids, input_mask)\\n # visual_emb, text_emb = self.ssl_head(visual_token, text_only_out['last_hidden_state'], input_mask, token_ids)\\n text_only_out = self.text_backbone(token_ids, input_mask)\\n text_out_last_hidden_state = text_only_out['last_hidden_state']\\n visual_emb, text_emb = self.ssl_head(visual_token, text_out_last_hidden_state, input_mask, token_ids)\\n\\n return visual_emb, text_emb\\n \\n visual_token = visual_token.view(B, D, T, -1).permute(0, 2, 3, 1)\\n output = self.multimodal_backbone(visual_token, token_ids, segment_ids, input_mask)\\n all_cls_emb = output['last_hidden_state'][:, 0]\\n itm_output = self.itm_head(all_cls_emb)\\n itm_output = torch.softmax(itm_output, dim=-1)\\n return itm_output\",\n \"def identity_block(input_tensor, kernel_size, filters, stage, block_id, data_format='channels_last', train_bn=True):\\n\\n filter1, filter2, filter3 = filters\\n conv_name_base = 'res' + str(stage) + block_id + '_branch'\\n bn_name_base = 'bn' + str(stage) + block_id + '_branch'\\n\\n shortcut = input_tensor\\n\\n x = tl.layers.Conv2d(input_tensor, filter1, (1, 1), padding='VALID', data_format=data_format,\\n W_init=W_init, b_init=b_init, name=conv_name_base + '2a')\\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2a')\\n\\n x = tl.layers.Conv2d(x, filter2, (kernel_size, kernel_size), padding='SAME', data_format=data_format,\\n W_init=W_init, b_init=b_init, name=conv_name_base + '2b')\\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2b')\\n\\n x = tl.layers.Conv2d(x, filter3, (1, 1), padding='VALID', data_format=data_format,\\n W_init=W_init, b_init=b_init, name=conv_name_base + '2c')\\n x = tl.layers.BatchNormLayer(x, act=tf.nn.relu, is_train=train_bn, name=bn_name_base + '2c')\\n\\n x = tl.layers.ElementwiseLayer([x, shortcut], combine_fn=tf.add, act=tf.nn.relu, name=\\\"add\\\")\\n return x\",\n \"def inception_block_1a(X):\\n\\tX_3x3 = Conv2D(96, (1, 1), data_format='channels_first', name ='inception_3a_3x3_conv1')(X)\\n\\tX_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name = 'inception_3a_3x3_bn1')(X_3x3)\\n\\tX_3x3 = Activation('relu')(X_3x3)\\n\\tX_3x3 = ZeroPadding2D(padding=(1, 1), data_format='channels_first')(X_3x3)\\n\\tX_3x3 = Conv2D(128, (3, 3), data_format='channels_first', name='inception_3a_3x3_conv2')(X_3x3)\\n\\tX_3x3 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_3x3_bn2')(X_3x3)\\n\\tX_3x3 = Activation('relu')(X_3x3)\\n\\tX_5x5 = Conv2D(16, (1, 1), data_format='channels_first', name='inception_3a_5x5_conv1')(X)\\n\\tX_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn1')(X_5x5)\\n\\tX_5x5 = Activation('relu')(X_5x5)\\n\\tX_5x5 = ZeroPadding2D(padding=(2, 2), data_format='channels_first')(X_5x5)\\n\\tX_5x5 = Conv2D(32, (5, 5), data_format='channels_first', name='inception_3a_5x5_conv2')(X_5x5)\\n\\tX_5x5 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_5x5_bn2')(X_5x5)\\n\\tX_5x5 = Activation('relu')(X_5x5)\\n\\tX_pool = MaxPooling2D(pool_size=3, strides=2, data_format='channels_first')(X)\\n\\tX_pool = Conv2D(32, (1, 1), data_format='channels_first', name='inception_3a_pool_conv')(X_pool)\\n\\tX_pool = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_pool_bn')(X_pool)\\n\\tX_pool = Activation('relu')(X_pool)\\n\\tX_pool = ZeroPadding2D(padding=((3, 4), (3, 4)), data_format='channels_first')(X_pool)\\n\\tX_1x1 = Conv2D(64, (1, 1), data_format='channels_first', name='inception_3a_1x1_conv')(X)\\n\\tX_1x1 = BatchNormalization(axis=1, epsilon=0.00001, name='inception_3a_1x1_bn')(X_1x1)\\n\\tX_1x1 = Activation('relu')(X_1x1)\\n\\t# CONCAT\\n\\tinception = concatenate([X_3x3, X_5x5, X_pool, X_1x1], axis=1)\\n\\treturn inception\",\n \"def __init__(self, num_heads: int, size: int, dropout: float = 0.1):\\n super(MultiHeadedAttention, self).__init__()\\n\\n assert size % num_heads == 0\\n\\n self.head_size = head_size = size // num_heads\\n self.model_size = size\\n self.num_heads = num_heads\\n\\n self.k_layer = nn.Linear(size, num_heads * head_size)\\n self.v_layer = nn.Linear(size, num_heads * head_size)\\n self.q_layer = nn.Linear(size, num_heads * head_size)\\n\\n self.output_layer = nn.Linear(size, size)\\n self.softmax = nn.Softmax(dim=-1)\\n self.dropout = nn.Dropout(dropout)\",\n \"def __init__(self, idim, odim_tgt, odim_src, args, ignore_id=-1):\\n torch.nn.Module.__init__(self)\\n if args.transformer_attn_dropout_rate is None:\\n args.transformer_attn_dropout_rate = args.dropout_rate\\n\\n # special tokens and model dimensions\\n self.pad = 0\\n self.sos_tgt = odim_tgt - 1\\n self.eos_tgt = odim_tgt - 1\\n self.sos_src = odim_src - 1\\n self.eos_src = odim_src - 1\\n self.odim_tgt = odim_tgt\\n self.odim_src = odim_src\\n self.idim = idim\\n self.adim = args.adim\\n self.ignore_id = ignore_id\\n\\n # submodule\\n self.mtlalpha = getattr(args, \\\"mtlalpha\\\", 0.0)\\n self.asr_weight = getattr(args, \\\"asr_weight\\\", 0.0)\\n self.mt_weight = getattr(args, \\\"mt_weight\\\", 0.0)\\n self.num_decoders = getattr(args, \\\"num_decoders\\\", 2)\\n self.do_st = getattr(args, \\\"do_st\\\", True)\\n self.do_mt = getattr(args, \\\"do_mt\\\", self.mt_weight > 0.0)\\n self.do_asr = self.asr_weight > 0 and self.mtlalpha < 1\\n\\n # cross-attention parameters\\n self.cross_weight = getattr(args, \\\"cross_weight\\\", 0.0)\\n self.cross_self = getattr(args, \\\"cross_self\\\", False)\\n self.cross_src = getattr(args, \\\"cross_src\\\", False)\\n self.cross_operator = getattr(args, \\\"cross_operator\\\", None)\\n self.cross_to_asr = getattr(args, \\\"cross_to_asr\\\", False)\\n self.cross_to_st = getattr(args, \\\"cross_to_st\\\", False)\\n self.wait_k_asr = getattr(args, \\\"wait_k_asr\\\", 0)\\n self.wait_k_st = getattr(args, \\\"wait_k_st\\\", 0)\\n self.cross_src_from = getattr(args, \\\"cross_src_from\\\", \\\"embedding\\\")\\n self.cross_self_from = getattr(args, \\\"cross_self_from\\\", \\\"embedding\\\")\\n self.cross_shared = getattr(args, \\\"cross_shared\\\", False)\\n self.cross_weight_learnable = getattr(args, \\\"cross_weight_learnable\\\", False)\\n\\n # one-to-many models parameters\\n self.use_joint_dict = getattr(args, \\\"use_joint_dict\\\", True)\\n self.one_to_many = getattr(args, \\\"one_to_many\\\", False)\\n self.use_lid = getattr(args, \\\"use_lid\\\", False)\\n if self.use_joint_dict:\\n self.langs_dict = getattr(args, \\\"langs_dict_tgt\\\", None)\\n self.lang_tok = getattr(args, \\\"lang_tok\\\", None)\\n self.lang_tok_mt = getattr(args, \\\"lang_tok_mt\\\", None)\\n\\n self.subsample = get_subsample(args, \\n mode='mt' if self.do_mt else 'st', \\n arch='transformer')\\n self.reporter = MTReporter() if self.do_mt else Reporter() \\n self.normalize_before = getattr(args, \\\"normalize_before\\\", True)\\n\\n # Backward compatability\\n if self.cross_operator in [\\\"sum\\\", \\\"concat\\\"]:\\n if self.cross_self and self.cross_src:\\n self.cross_operator = \\\"self_src\\\" + self.cross_operator\\n elif self.cross_self:\\n self.cross_operator = \\\"self_\\\" + self.cross_operator\\n elif self.cross_src:\\n self.cross_operator = \\\"src_\\\" + self.cross_operator\\n if self.cross_operator:\\n assert self.cross_operator in ['self_sum', 'self_concat', 'src_sum', \\n 'src_concat', 'self_src_sum', 'self_src_concat']\\n\\n # Check parameters\\n if self.one_to_many:\\n self.use_lid = True\\n if not self.do_st:\\n assert (not self.cross_to_asr) and (not self.cross_to_st)\\n if self.cross_operator and 'sum' in self.cross_operator and self.cross_weight <= 0:\\n assert (not self.cross_to_asr) and (not self.cross_to_st)\\n if self.cross_to_asr or self.cross_to_st:\\n assert self.do_st and self.do_asr\\n assert self.cross_self or self.cross_src\\n assert bool(self.cross_operator) == (self.do_asr and (self.cross_to_asr or self.cross_to_st))\\n if self.cross_src_from != \\\"embedding\\\" or self.cross_self_from != \\\"embedding\\\":\\n assert self.normalize_before\\n if self.wait_k_asr > 0:\\n assert self.wait_k_st == 0\\n elif self.wait_k_st > 0:\\n assert self.wait_k_asr == 0\\n else:\\n assert self.wait_k_asr == 0\\n assert self.wait_k_st == 0\\n\\n logging.info(\\\"*** Cross attention parameters ***\\\")\\n if self.cross_to_asr:\\n logging.info(\\\"| Cross to ASR\\\")\\n if self.cross_to_st:\\n logging.info(\\\"| Cross to ST\\\")\\n if self.cross_self:\\n logging.info(\\\"| Cross at Self\\\")\\n if self.cross_src:\\n logging.info(\\\"| Cross at Source\\\")\\n if self.cross_to_asr or self.cross_to_st:\\n logging.info(f'| Cross operator: {self.cross_operator}')\\n logging.info(f'| Cross sum weight: {self.cross_weight}')\\n if self.cross_src:\\n logging.info(f'| Cross source from: {self.cross_src_from}')\\n if self.cross_self:\\n logging.info(f'| Cross self from: {self.cross_self_from}')\\n logging.info(f\\\"Use joint dictionary: {self.use_joint_dict}\\\")\\n \\n if (self.cross_src_from != \\\"embedding\\\" and self.cross_src) \\\\\\n and (not self.normalize_before):\\n logging.warning(f'WARNING: Resort to using \\\\\\n self.cross_src_from == embedding for cross at source attention.')\\n if (self.cross_self_from != \\\"embedding\\\" and self.cross_self) \\\\\\n and (not self.normalize_before):\\n logging.warning(f'WARNING: Resort to using \\\\\\n self.cross_self_from == embedding for cross at self attention.')\\n\\n # Adapters\\n self.use_adapters = getattr(args, \\\"use_adapters\\\", False)\\n self.use_adapters_in_enc = getattr(args, \\\"use_adapters_in_enc\\\", False)\\n adapter_names = getattr(args, \\\"adapters\\\", None)\\n adapter_reduction_factor = getattr(args, \\\"adapter_reduction_factor\\\", None)\\n adapter_reduction_factor_enc = getattr(args, \\\"adapter_reduction_factor_enc\\\", adapter_reduction_factor)\\n use_adapters_for_asr = getattr(args, \\\"use_adapters_for_asr\\\", True)\\n adapter_before_src_attn = getattr(args, \\\"adapter_before_src_attn\\\", False)\\n adapter_after_mha = getattr(args, \\\"adapter_after_mha\\\", False)\\n use_shared_adapters = getattr(args, \\\"use_shared_adapters\\\", False)\\n use_shared_adapters_enc = getattr(args, \\\"use_shared_adapters_enc\\\", False)\\n # if self.use_adapters and not use_adapters_for_asr:\\n # assert not self.do_asr or \\\\\\n # (self.do_asr and self.num_decoders != 1) or \\\\\\n # (self.do_asr and not self.do_st) # for backward compatibility\\n\\n if adapter_names:\\n if self.do_asr and not self.do_st:\\n adapter_names = [str(args.char_list_src.index(f'<2{l}>')) for l in adapter_names]\\n else:\\n adapter_names = [str(args.char_list_tgt.index(f'<2{l}>')) for l in adapter_names]\\n logging.info(f'| adapters = {adapter_names}')\\n\\n if self.do_st or self.do_asr:\\n logging.info(f'Speech encoder')\\n self.encoder = Encoder(\\n idim=idim,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.eunits,\\n num_blocks=args.elayers,\\n input_layer=getattr(args, \\\"transformer_input_layer\\\", \\\"conv2d\\\"),\\n dropout_rate=args.dropout_rate,\\n positional_dropout_rate=args.dropout_rate,\\n attention_dropout_rate=args.transformer_attn_dropout_rate,\\n adapter_names=adapter_names if self.use_adapters_in_enc else None,\\n reduction_factor=adapter_reduction_factor_enc,\\n adapter_after_mha=adapter_after_mha,\\n shared_adapters=use_shared_adapters_enc,\\n )\\n if self.do_st:\\n logging.info('ST decoder')\\n self.decoder = Decoder(\\n odim=odim_tgt,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n dropout_rate=args.dropout_rate,\\n positional_dropout_rate=args.dropout_rate,\\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n normalize_before=self.normalize_before,\\n cross_operator=self.cross_operator if self.cross_to_st else None,\\n cross_shared=self.cross_shared,\\n cross_weight_learnable=self.cross_weight_learnable,\\n cross_weight=self.cross_weight,\\n use_output_layer=True if (self.use_joint_dict or \\\\\\n (self.do_st and not self.do_asr)) else False,\\n adapter_names=adapter_names,\\n reduction_factor=adapter_reduction_factor,\\n adapter_before_src_attn=adapter_before_src_attn,\\n adapter_after_mha=adapter_after_mha,\\n shared_adapters=use_shared_adapters,\\n )\\n if self.do_asr:\\n logging.info('ASR decoder')\\n self.decoder_asr = Decoder(\\n odim=odim_src,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n dropout_rate=args.dropout_rate,\\n positional_dropout_rate=args.dropout_rate,\\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n normalize_before=self.normalize_before,\\n cross_operator=self.cross_operator if self.cross_to_asr else None,\\n cross_shared=self.cross_shared,\\n cross_weight_learnable=self.cross_weight_learnable,\\n cross_weight=self.cross_weight,\\n use_output_layer=True if (self.use_joint_dict or \\\\\\n (self.do_asr and not self.do_st)) else False,\\n adapter_names=adapter_names,\\n reduction_factor=adapter_reduction_factor,\\n adapter_before_src_attn=adapter_before_src_attn,\\n adapter_after_mha=adapter_after_mha,\\n shared_adapters=use_shared_adapters,\\n )\\n if self.num_decoders == 1 and self.do_st:\\n logging.info('*** Use shared decoders *** ')\\n self.decoder_asr = self.decoder\\n\\n if not self.use_joint_dict and (self.do_st and self.do_asr):\\n self.output_layer = torch.nn.Linear(args.adim, odim_tgt)\\n self.output_layer_asr = torch.nn.Linear(args.adim, odim_src)\\n\\n # submodule for MT task\\n if self.do_mt:\\n logging.info('MT encoder')\\n self.encoder_mt = Encoder(\\n idim=odim_src,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n input_layer='embed',\\n dropout_rate=args.dropout_rate,\\n positional_dropout_rate=args.dropout_rate,\\n attention_dropout_rate=args.transformer_attn_dropout_rate,\\n padding_idx=0\\n )\\n if not self.do_st:\\n logging.info('MT decoder')\\n self.decoder_mt = Decoder(\\n odim=odim_tgt,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n dropout_rate=args.dropout_rate,\\n positional_dropout_rate=args.dropout_rate,\\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n normalize_before=self.normalize_before,\\n use_output_layer=True,\\n )\\n self.reset_parameters(args) # place after the submodule initialization\\n if self.mtlalpha > 0.0:\\n self.ctc = CTC(odim_src, args.adim, args.dropout_rate, \\n ctc_type=args.ctc_type, reduce=True,\\n zero_infinity=True)\\n else:\\n self.ctc = None\\n\\n if self.asr_weight > 0 and (args.report_cer or args.report_wer):\\n from espnet.nets.e2e_asr_common import ErrorCalculator\\n self.error_calculator = ErrorCalculator(args.char_list_src,\\n args.sym_space, args.sym_blank,\\n args.report_cer, args.report_wer)\\n elif self.do_mt and getattr(args, \\\"report_bleu\\\", False):\\n from espnet.nets.e2e_mt_common import ErrorCalculator\\n self.error_calculator = ErrorCalculator(args.char_list_tgt,\\n args.sym_space,\\n args.report_bleu)\\n else:\\n self.error_calculator = None\\n self.rnnlm = None\\n\\n # criterion\\n if self.do_st:\\n self.criterion_st = LabelSmoothingLoss(self.odim_tgt, self.ignore_id, args.lsm_weight,\\n args.transformer_length_normalized_loss)\\n if self.do_asr:\\n self.criterion_asr = LabelSmoothingLoss(self.odim_src, self.ignore_id, args.lsm_weight,\\n args.transformer_length_normalized_loss)\\n if self.do_mt:\\n self.criterion_mt = LabelSmoothingLoss(self.odim_tgt, self.ignore_id, args.lsm_weight,\\n args.transformer_length_normalized_loss)\\n self.normalize_length = args.transformer_length_normalized_loss # for PPL\\n\\n # Language embedding layer\\n if self.lang_tok == \\\"encoder-pre-sum\\\":\\n self.language_embeddings = build_embedding(self.langs_dict, self.idim, \\n padding_idx=self.pad)\\n logging.info(f'language_embeddings: {self.language_embeddings}')\\n\\n # Backward compatability\\n if self.cross_operator:\\n if \\\"sum\\\" in self.cross_operator:\\n self.cross_operator = \\\"sum\\\"\\n if \\\"concat\\\" in self.cross_operator: \\n self.cross_operator = \\\"concat\\\"\",\n \"def forward(self, data_batch):\\n\\n x = data_batch[0]\\n im_info = data_batch[1]\\n gt_boxes = data_batch[2]\\n num_boxes = data_batch[3]\\n rel_mat = data_batch[4]\\n\\n if self.training:\\n self.iter_counter += 1\\n\\n input_imgs = x.clone()\\n\\n sources = list()\\n loc = list()\\n conf = list()\\n\\n self.batch_size = x.size(0)\\n\\n # apply vgg up to conv4_3 relu\\n if isinstance(self.base, nn.ModuleList):\\n for k,v in enumerate(self.base):\\n x = v(x)\\n else:\\n x = self.base(x)\\n\\n s = self.L2Norm(x)\\n sources.append(s)\\n base_feat = s\\n\\n # apply vgg up to fc7\\n if isinstance(self.conv5, nn.ModuleList):\\n for k,v in enumerate(self.conv5):\\n x = v(x)\\n else:\\n x = self.conv5(x)\\n sources.append(x)\\n\\n # apply extra layers and cache source layer outputs\\n for k, v in enumerate(self.extras):\\n x = F.relu(v(x), inplace=True)\\n if k % 2 == 1:\\n sources.append(x)\\n\\n # apply multibox head to source layers\\n for (x, l, c) in zip(sources, self.loc, self.conf):\\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\\n\\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\\n\\n loc = loc.view(loc.size(0), -1, 4)\\n conf = conf.view(conf.size(0), -1, self.num_classes)\\n\\n SSD_loss_cls = 0\\n SSD_loss_bbox = 0\\n if self.training:\\n predictions = (\\n loc,\\n conf,\\n self.priors.type_as(loc)\\n )\\n # targets = torch.cat([gt_boxes[:,:,:4] / self.size, gt_boxes[:,:,4:5]],dim=2)\\n targets = gt_boxes\\n SSD_loss_bbox, SSD_loss_cls = self.criterion(predictions, targets, num_boxes)\\n\\n conf = self.softmax(conf)\\n\\n # online data\\n if self.training:\\n if self.iter_counter > cfg.TRAIN.VMRN.ONLINEDATA_BEGIN_ITER:\\n obj_rois, obj_num = self._obj_det(conf, loc, self.batch_size, im_info)\\n obj_rois = obj_rois.type_as(gt_boxes)\\n obj_num = obj_num.type_as(num_boxes)\\n else:\\n obj_rois = torch.FloatTensor([]).type_as(gt_boxes)\\n obj_num = torch.LongTensor([]).type_as(num_boxes)\\n obj_labels = None\\n else:\\n # when testing, this is object detection results\\n # TODO: SUPPORT MULTI-IMAGE BATCH\\n obj_rois, obj_num = self._obj_det(conf, loc, self.batch_size, im_info)\\n if obj_rois.numel() > 0:\\n obj_labels = obj_rois[:, 5]\\n obj_rois = obj_rois[:, :5]\\n obj_rois = obj_rois.type_as(gt_boxes)\\n obj_num = obj_num.type_as(num_boxes)\\n else:\\n # there is no object detected\\n obj_labels = torch.Tensor([]).type_as(gt_boxes).long()\\n obj_rois = obj_rois.type_as(gt_boxes)\\n obj_num = obj_num.type_as(num_boxes)\\n\\n if self.training:\\n # offline data\\n for i in range(self.batch_size):\\n obj_rois = torch.cat([obj_rois,\\n torch.cat([(i * torch.ones(num_boxes[i].item(), 1)).type_as(gt_boxes),\\n (gt_boxes[i][:num_boxes[i]][:, 0:4])], 1)\\n ])\\n obj_num = torch.cat([obj_num, torch.Tensor([num_boxes[i]]).type_as(obj_num)])\\n\\n\\n obj_rois = Variable(obj_rois)\\n\\n VMRN_rel_loss_cls = 0\\n rel_cls_prob = torch.Tensor([]).type_as(obj_rois)\\n if (obj_num > 1).sum().item() > 0:\\n\\n obj_pair_feat = self.VMRN_obj_pair_feat_extractor(input_imgs, obj_rois, self.batch_size, obj_num)\\n # obj_pair_feat = obj_pair_feat.detach()\\n rel_cls_score = self.VMRN_rel_cls_score(obj_pair_feat)\\n\\n rel_cls_prob = F.softmax(rel_cls_score)\\n\\n self.rel_batch_size = obj_pair_feat.size(0)\\n\\n if self.training:\\n obj_pair_rel_label = self._generate_rel_labels(obj_rois, gt_boxes, obj_num, rel_mat)\\n obj_pair_rel_label = obj_pair_rel_label.type_as(gt_boxes).long()\\n\\n rel_not_keep = (obj_pair_rel_label == 0)\\n # no relationship is kept\\n if (rel_not_keep == 0).sum().item() > 0:\\n rel_keep = torch.nonzero(rel_not_keep == 0).view(-1)\\n\\n rel_cls_score = rel_cls_score[rel_keep]\\n\\n obj_pair_rel_label = obj_pair_rel_label[rel_keep]\\n obj_pair_rel_label -= 1\\n VMRN_rel_loss_cls = F.cross_entropy(rel_cls_score, obj_pair_rel_label)\\n else:\\n if (not cfg.TEST.VMRN.ISEX) and cfg.TRAIN.VMRN.ISEX:\\n rel_cls_prob = rel_cls_prob[::2, :]\\n\\n rel_result = None\\n if not self.training:\\n if obj_rois.numel() > 0:\\n pred_boxes = obj_rois.data[:,1:5]\\n pred_boxes[:, 0::2] /= im_info[0][3].item()\\n pred_boxes[:, 1::2] /= im_info[0][2].item()\\n rel_result = (pred_boxes, obj_labels, rel_cls_prob.data)\\n else:\\n rel_result = (obj_rois.data, obj_labels, rel_cls_prob.data)\\n\\n return loc, conf, rel_result, SSD_loss_bbox, SSD_loss_cls, VMRN_rel_loss_cls\",\n \"def __init__(self, in_channels, num_classes):\\n super(InceptionAux, self).__init__()\\n self.relu = nn.LeakyReLU()\\n self.dropout1 = nn.Dropout(p=0.5)\\n self.pool = nn.AvgPool2d(kernel_size=5, stride=3)\\n self.conv = conv_block(in_channels, 128, kernel_size=1)\\n self.fc1 = nn.Linear(12800, 1024)\\n self.fc2 = nn.Linear(1024, 420)\\n self.dropout2 = nn.Dropout(p=0.4)\\n self.fc3 = nn.Linear(420,128)\\n self.dropout3 = nn.Dropout(p=0.4)\\n self.fc4 = nn.Linear(128,num_classes)\",\n \"def __init__(self, num_heads: int, size: int, size_v: int, dropout: float = 0.1):\\n super(ContMultiHeadedAttention, self).__init__()\\n\\n assert size % num_heads == 0\\n\\n self.head_size = head_size = size // num_heads\\n self.model_size = size\\n self.num_heads = num_heads\\n\\n self.k_layer = nn.Linear(size, num_heads * head_size)\\n self.v_layer = nn.Linear(size_v, num_heads * head_size)\\n self.q_layer = nn.Linear(size, num_heads * head_size)\\n\\n self.output_layer = nn.Linear(size, size_v)\\n self.softmax = nn.Softmax(dim=-1)\\n self.dropout = nn.Dropout(dropout)\",\n \"def forward(self, x):\\n sources = list()\\n loc = list()\\n conf = list()\\n\\n # apply vgg up to conv4_3 relu\\n for k in range(23):\\n x = self.vgg[k](x)\\n\\n s = self.L2Norm(x)\\n sources.append(s)\\n\\n # apply vgg up to fc7\\n for k in range(23, len(self.vgg)):\\n x = self.vgg[k](x)\\n sources.append(x)\\n\\n # apply extra layers and cache source layer outputs\\n for k, v in enumerate(self.extras):\\n x = F.relu(v(x), inplace=True)\\n if k % 2 == 1:\\n sources.append(x)\\n\\n # apply multibox head to source layers\\n for (x, l, c) in zip(sources, self.loc, self.conf):\\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\\n\\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\\n if self.phase == \\\"test\\\":\\n output = self.detect(\\n loc.view(loc.size(0), -1, 4), # loc preds\\n self.softmax(conf.view(conf.size(0), -1,\\n self.num_classes)), # conf preds\\n self.priors.type(type(x.data)) # default boxes\\n )\\n else:\\n output = (\\n loc.view(loc.size(0), -1, 4),\\n conf.view(conf.size(0), -1, self.num_classes),\\n self.priors\\n )\\n return output\",\n \"def forward(self, enc_states, enc_len, dec_states):\\n if self.precomputed_enc_h is None:\\n\\n self.precomputed_enc_h = self.mlp_enc(enc_states)\\n self.mask = length_to_mask(\\n enc_len, max_len=enc_states.size(1), device=enc_states.device\\n )\\n\\n # multiply mask by 1/Ln for each row\\n self.prev_attn = self.mask * (1 / enc_len.float()).unsqueeze(1)\\n\\n # compute location-aware features\\n # [B, 1, L] -> [B, C, L]\\n attn_conv = self.conv_loc(self.prev_attn.unsqueeze(1))\\n # [B, C, L] -> [B, L, C] -> [B, L, F]\\n attn_conv = self.mlp_loc(attn_conv.transpose(1, 2))\\n\\n dec_h = self.mlp_dec(dec_states.unsqueeze(1))\\n attn = self.mlp_attn(\\n torch.tanh(self.precomputed_enc_h + dec_h + attn_conv)\\n ).squeeze(-1)\\n\\n # mask the padded frames\\n attn = attn.masked_fill(self.mask == 0, -np.inf)\\n attn = self.softmax(attn * self.scaling)\\n\\n # set prev_attn to current attn for the next timestep\\n self.prev_attn = attn.detach()\\n\\n # compute context vectors\\n # [B, 1, L] X [B, L, F]\\n context = torch.bmm(attn.unsqueeze(1), enc_states).squeeze(1)\\n context = self.mlp_out(context)\\n\\n return context, attn\",\n \"def __init__(self,\\n num_heads,\\n seq_len,\\n block=16,\\n different_layout_per_head=False,\\n num_local_blocks=4,\\n num_global_blocks=1,\\n attention='bidirectional',\\n horizental_global_attention=False,\\n num_differnt_global_patterns=1):\\n super().__init__(num_heads, seq_len, block, different_layout_per_head)\\n\\n if (seq_len % (block * num_local_blocks) != 0):\\n raise ValueError(\\n f'Sequence Length, {seq_len}, must be dividable by local window, {block*num_local_blocks}!'\\n )\\n self.num_local_blocks = num_local_blocks\\n\\n if (num_local_blocks % num_global_blocks != 0):\\n raise ValueError(\\n f'Number of blocks in a local window, {num_local_blocks}, must be dividable by number of global blocks, {num_global_blocks}!'\\n )\\n self.num_global_blocks = num_global_blocks\\n\\n if (attention != 'unidirectional' and attention != 'bidirectional'):\\n raise NotImplementedError(\\n 'only \\\\\\\"uni/bi-directional\\\\\\\" attentions are supported for now!')\\n self.attention = attention\\n\\n if (attention != 'bidirectional' and horizental_global_attention):\\n raise ValueError(\\n 'only \\\\\\\"bi-directional\\\\\\\" attentions can support horizental global attention!'\\n )\\n self.horizental_global_attention = horizental_global_attention\\n\\n if (num_differnt_global_patterns > 1 and not different_layout_per_head):\\n raise ValueError(\\n f'Number of different layouts cannot be more than one when you have set a single layout for all heads! Set different_layout_per_head to True.'\\n )\\n if (num_differnt_global_patterns > (num_local_blocks // num_global_blocks)):\\n raise ValueError(\\n f'Number of layout versions (num_differnt_global_patterns), {num_differnt_global_patterns}, cannot be larger than number of local window blocks divided by number of global blocks, {num_local_blocks} / {num_global_blocks} = {num_local_blocks//num_global_blocks}!'\\n )\\n self.num_differnt_global_patterns = num_differnt_global_patterns\\n self.make_layout()\",\n \"def mt_ic_step(self, dataset='coco', input_stream='img', lambda_coeff=1):\\n assert lambda_coeff >= 0\\n if lambda_coeff == 0:\\n return\\n params = self.params\\n name = 'model' if params.encoder_only else 'encoder'\\n model = getattr(self, name)\\n model.train()\\n\\n # _lang1 = _lang2 = 'en'\\n # lang2_id = params.lang2id[_lang2]\\n\\n # generate batch\\n\\n (x_src,len_src),(x2, len2),(x1, x1_mask, img_loc, img_id) = self.get_batch('txt2img', dataset, input_stream)\\n # convert fp16\\n #assign lang ids\\n\\n\\n lang0_id = params.lang2id[params.ft_lgs[0]]\\n lang_src = x_src.clone().fill_(lang0_id)\\n lang1_id = params.lang2id[params.ft_lgs[1]]\\n langs = x2.clone().fill_(lang1_id)\\n\\n\\n # target words to predict\\n alen = torch.arange(len2.max(), dtype=torch.long, device=len2.device)\\n pred_mask = alen[:, None] < len2[None] - 1 # do not predict anything given the last target word\\n y = x2[1:].masked_select(pred_mask[:-1]) # decide whether or not predict\\n assert len(y) == (len2 - 1).sum().item()\\n\\n #\\n len1 = x1_mask.sum(dim=1)\\n x1 = x1.transpose(0, 1)\\n img_loc = img_loc.transpose(0, 1)\\n\\n # cuda\\n\\n x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src = to_cuda(x1, len1, img_loc, x2, len2, y, x1_mask,langs,lang_src,x_src,len_src)\\n\\n if params.mt_only_text:\\n encoder_outputs = model('crossfwd', stream_='text', x=x_src, lengths=len_src, langs=lang_src, causal=False,refine_image=params.refine_image)\\n len_all = len_src\\n else:\\n encoder_outputs = model('jointfwd', x=x_src, lengths=len_src, x_img=x1, lengths_img=len1, causal=False,\\n langs=None,\\n image_loc=img_loc, refine_image=params.refine_image)\\n len_all = len_src + len1\\n\\n # encode source sentence\\n # enc1 = model('jo', stream_='img', x=x1, lengths=len1, langs=None, causal=False, cross_modal=True,\\n # image_loc=img_loc, refine_image=params.refine_image, refine_encoder=params.refine_encoder,\\n # image_dist=None)\\n enc1 = encoder_outputs.transpose(0, 1)\\n\\n\\n # decode target sentence\\n dec2 = model('crossfwd', stream_='text', x=x2, lengths=len2, langs=langs, causal=True, src_enc=enc1,\\n src_len=len_all)\\n\\n # loss\\n # the last word not apply prediction logic\\n _, loss = model('predict', tensor=dec2, pred_mask=pred_mask, y=y, get_scores=False)\\n self.stats[('IC-%s-%s' % (dataset, input_stream))].append(loss.item())\\n loss = lambda_coeff * loss\\n\\n # optimize\\n self.optimize(loss)\\n\\n # number of processed sentences / words\\n self.n_sentences += params.batch_size\\n self.stats['processed_s'] += len2.size(0)\\n self.stats['processed_w'] += (len2 - 1).sum().item()\",\n \"def __init__(self, idim, odim, args, ignore_id=-1):\\n torch.nn.Module.__init__(self)\\n if args.transformer_attn_dropout_rate is None:\\n args.transformer_attn_dropout_rate = args.dropout_rate\\n\\n self.mtlalpha = args.mtlalpha\\n if not hasattr(args, \\\"beta\\\"):\\n self.beta = 0.0\\n else:\\n self.beta = args.beta\\n self.num_steps = args.num_steps\\n self.tau = args.error_temperature\\n self.idx_blank = 0\\n self.masktoken = odim - 1\\n self.sos = odim - 1\\n self.eos = odim - 1\\n self.odim = odim\\n self.ignore_id = ignore_id\\n self.lsm_weight = args.lsm_weight\\n self.reporter = None\\n\\n self.subsample = [1]\\n self.encoder = Encoder(\\n idim=idim,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.eunits,\\n num_blocks=args.elayers,\\n input_layer=args.transformer_input_layer,\\n dropout_rate=args.dropout_rate,\\n death_rate=args.edeath_rate,\\n positional_dropout_rate=args.dropout_rate,\\n attention_dropout_rate=args.transformer_attn_dropout_rate\\n )\\n self.enc_odim = args.adim\\n\\n # Define alignment embedding module.\\n self.embed = torch.nn.Embedding(odim, self.enc_odim)\\n self.decoder = Decoder(\\n odim=odim,\\n attention_dim=args.adim,\\n attention_heads=args.aheads,\\n linear_units=args.dunits,\\n num_blocks=args.dlayers,\\n dropout_rate=args.dropout_rate,\\n death_rate=args.ddeath_rate,\\n positional_dropout_rate=args.dropout_rate,\\n self_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n src_attention_dropout_rate=args.transformer_attn_dropout_rate,\\n input_layer=\\\"embed_linear\\\",\\n pos_enc_class=ScaledPositionalEncoding\\n )\\n\\n # self.verbose = args.verbose\\n self.reset_parameters(args)\\n self.adim = args.adim\\n\\n # CTC directly works on top of encoder outputs.\\n self.ctc = CTC(odim, self.enc_odim, args.dropout_rate, ctc_type=args.ctc_type, reduce=True)\\n self.rnnlm = None\\n\\n # Decoder masking params\\n self.decoder_num_mask = args.decoder_num_mask\\n self.decoder_mask_width = args.decoder_mask_width\",\n \"def __init__(self, block, layers, num_classes=1000, width=64):\\n\\n torch.nn.Module.__init__(self) # Skip the parent constructor. This replaces it.\\n self._norm_layer = torch.nn.BatchNorm2d\\n self.inplanes = width\\n self.dilation = 1\\n self.groups = 1\\n self.base_width = 64\\n\\n # The initial convolutional layer.\\n self.conv1 = torch.nn.Conv2d(3, self.inplanes, kernel_size=7, stride=2, padding=3, bias=False)\\n self.bn1 = self._norm_layer(self.inplanes)\\n self.relu = torch.nn.ReLU(inplace=True)\\n self.maxpool = torch.nn.MaxPool2d(kernel_size=3, stride=2, padding=1)\\n\\n # The subsequent blocks.\\n self.layer1 = self._make_layer(block, width, layers[0])\\n self.layer2 = self._make_layer(block, width*2, layers[1], stride=2, dilate=False)\\n self.layer3 = self._make_layer(block, width*4, layers[2], stride=2, dilate=False)\\n self.layer4 = self._make_layer(block, width*8, layers[3], stride=2, dilate=False)\\n\\n # The last layers.\\n self.avgpool = torch.nn.AdaptiveAvgPool2d((1, 1))\\n self.fc = torch.nn.Linear(width*8*block.expansion, num_classes)\\n\\n # Default init.\\n for m in self.modules():\\n if isinstance(m, torch.nn.Conv2d):\\n torch.nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')\\n elif isinstance(m, torch.nn.BatchNorm2d):\\n torch.nn.init.constant_(m.weight, 1)\\n torch.nn.init.constant_(m.bias, 0)\",\n \"def __init__(self, id, node_type=NodeType.HIDDEN, activation=F.relu, layer_type=nn.Conv2d,\\n conv_window_size=3, conv_stride=1, max_pool_size=2):\\n\\n super(ModuleNEATNode, self).__init__(id, node_type)\\n\\n batch_norm_chance = 0.65 # chance that a new node will start with batch norm\\n use_batch_norm = random.random() < batch_norm_chance\\n\\n dropout_chance = 0.2 # chance that a new node will start with drop out\\n use_dropout = random.random() < dropout_chance\\n\\n max_pool_chance = 0.3 # chance that a new node will start with drop out\\n use_max_pool = random.random() < max_pool_chance\\n\\n self.activation = Mutagen(F.relu, F.leaky_relu, torch.sigmoid, F.relu6,\\n discreet_value=activation, name=\\\"activation function\\\",\\n mutation_chance=0.15) # TODO try add in Selu, Elu\\n\\n conv_out_features = 25 + random.randint(0, 25)\\n linear_out_features = 100 + random.randint(0, 100)\\n\\n linear_submutagens = \\\\\\n {\\n \\\"regularisation\\\": Mutagen(None, nn.BatchNorm1d,\\n discreet_value=nn.BatchNorm1d if use_batch_norm else None,\\n mutation_chance=0.15),\\n\\n \\\"dropout\\\": Mutagen(None, nn.Dropout, discreet_value=nn.Dropout if use_dropout else None, sub_mutagens=\\n {\\n nn.Dropout: {\\n \\\"dropout_factor\\\": Mutagen(value_type=ValueType.CONTINUOUS, current_value=0.15, start_range=0,\\n end_range=0.75)}\\n }, mutation_chance=0.08),\\n\\n \\\"out_features\\\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=linear_out_features,\\n start_range=10,\\n end_range=1024, name=\\\"num out features\\\", mutation_chance=0.22,\\n distance_weighting=Props.LAYER_SIZE_COEFFICIENT if Config.allow_attribute_distance else 0)\\n }\\n\\n conv_submutagens = {\\n \\\"conv_window_size\\\": Mutagen(3, 5, 7, discreet_value=conv_window_size, mutation_chance=0.13),\\n\\n \\\"conv_stride\\\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=conv_stride, start_range=1,\\n end_range=5),\\n\\n \\\"reduction\\\": Mutagen(None, nn.MaxPool2d, discreet_value=nn.MaxPool2d if use_max_pool else None,\\n sub_mutagens=\\n {\\n nn.MaxPool2d: {\\\"pool_size\\\": Mutagen(\\n value_type=ValueType.WHOLE_NUMBERS, current_value=max_pool_size, start_range=2,\\n end_range=5)}\\n }, mutation_chance=0.15),\\n\\n \\\"regularisation\\\": Mutagen(None, nn.BatchNorm2d, discreet_value=nn.BatchNorm2d if use_batch_norm else None,\\n mutation_chance=0.15),\\n\\n \\\"dropout\\\": Mutagen(None, nn.Dropout2d, discreet_value=nn.Dropout2d if use_dropout else None, sub_mutagens=\\n {\\n nn.Dropout2d: {\\n \\\"dropout_factor\\\": Mutagen(value_type=ValueType.CONTINUOUS, current_value=0.1,\\n start_range=0, end_range=0.75)}\\n }, mutation_chance=0.08),\\n\\n \\\"out_features\\\": Mutagen(value_type=ValueType.WHOLE_NUMBERS, current_value=conv_out_features, start_range=1,\\n end_range=100, name=\\\"num out features\\\", mutation_chance=0.22,\\n distance_weighting=Props.LAYER_SIZE_COEFFICIENT if Config.allow_attribute_distance else 0)\\n }\\n\\n if use_linears and not use_convs:\\n self.layer_type = Mutagen(nn.Linear, discreet_value=nn.Linear,\\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\\n sub_mutagens={nn.Linear: linear_submutagens}\\n )\\n if use_convs and not use_linears:\\n self.layer_type = Mutagen(nn.Conv2d, discreet_value=nn.Conv2d,\\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\\n sub_mutagens={nn.Conv2d: conv_submutagens})\\n if use_convs and use_linears:\\n self.layer_type = Mutagen(nn.Conv2d, nn.Linear, discreet_value=layer_type,\\n distance_weighting=Props.LAYER_TYPE_COEFFICIENT if Config.allow_attribute_distance else 0,\\n sub_mutagens={\\n nn.Conv2d: conv_submutagens,\\n nn.Linear: linear_submutagens\\n }, name=\\\"deep layer type\\\", mutation_chance=0.08)\",\n \"def __init__(self, fsmt_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([fsmt_layer.self_attn.q_proj.weight, fsmt_layer.self_attn.k_proj.weight, fsmt_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([fsmt_layer.self_attn.q_proj.bias, fsmt_layer.self_attn.k_proj.bias, fsmt_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = fsmt_layer.self_attn.out_proj.weight\\n self.out_proj_bias = fsmt_layer.self_attn.out_proj.bias\\n self.linear1_weight = fsmt_layer.fc1.weight\\n self.linear1_bias = fsmt_layer.fc1.bias\\n self.linear2_weight = fsmt_layer.fc2.weight\\n self.linear2_bias = fsmt_layer.fc2.bias\\n self.norm1_eps = fsmt_layer.self_attn_layer_norm.eps\\n self.norm1_weight = fsmt_layer.self_attn_layer_norm.weight\\n self.norm1_bias = fsmt_layer.self_attn_layer_norm.bias\\n self.norm2_eps = fsmt_layer.final_layer_norm.eps\\n self.norm2_weight = fsmt_layer.final_layer_norm.weight\\n self.norm2_bias = fsmt_layer.final_layer_norm.bias\\n self.num_heads = fsmt_layer.self_attn.num_heads\\n self.embed_dim = fsmt_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def preprocess_module(mod):\\n\\n def alter_conv(attrs, inputs, tinfos, out_type):\\n new_attrs = dict(attrs)\\n data_info = tinfos[0]\\n weight_info = tinfos[1]\\n (desired_data_layout, desired_kernel_layout) = (\\\"NCHW\\\", \\\"OIHW\\\")\\n new_attrs[\\\"data_layout\\\"] = desired_data_layout\\n new_attrs[\\\"kernel_layout\\\"] = desired_kernel_layout\\n\\n if is_depthwise_conv2d(\\n data_info.shape,\\n attrs[\\\"data_layout\\\"],\\n weight_info.shape,\\n attrs[\\\"kernel_layout\\\"],\\n attrs[\\\"groups\\\"],\\n ):\\n dkl = desired_kernel_layout\\n new_attrs[\\\"kernel_layout\\\"] = dkl[1] + dkl[0] + dkl[2] + dkl[3]\\n return relay.nn.conv2d(*inputs, **new_attrs)\\n\\n with OpAttrContext(\\\"nn.conv2d\\\", \\\"FTVMAlterOpLayout\\\", alter_conv):\\n seq = tvm.transform.Sequential(\\n [\\n transform.ConvertLayout({\\\"nn.conv2d\\\": [\\\"NCHW\\\", \\\"OIHW\\\"]}),\\n transform.ConvertLayout({\\\"nn.conv2d_transpose\\\": [\\\"NCHW\\\", \\\"OIHW\\\"]}),\\n transform.AlterOpLayout(),\\n transform.FoldConstant(),\\n ]\\n )\\n with tvm.transform.PassContext(opt_level=3):\\n preprocessed_mod = seq(mod)\\n return preprocessed_mod\",\n \"def __init__(self):\\n super(Backbone, self).__init__()\\n\\n # input size: (128, 282, 282)\\n # Block 1:\\n # relu + 4 conv + bn\\n self.conv1 = torch.nn.Conv2d(128, 64, kernel_size=3, stride=1, padding=0)\\n self.conv2 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\\n self.conv3 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\\n self.conv4 = torch.nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=0)\\n\\n self.bn1 = torch.nn.BatchNorm2d(64)\\n self.bn2 = torch.nn.BatchNorm2d(64)\\n self.bn3 = torch.nn.BatchNorm2d(64)\\n self.bn4 = torch.nn.BatchNorm2d(64)\\n\\n # Block 2:\\n # relu + 6 conv + stride 2 + bn\\n self.conv5 = torch.nn.Conv2d(64, 128, kernel_size=3, stride=2, padding=0)\\n self.conv6 = torch.nn.Conv2d(128, 128, kernel_size=3, stride=2, padding=0)\\n self.conv7 = torch.nn.Conv2d(128, 64, kernel_size=3, stride=1, padding=0)\\n self.conv8 = torch.nn.Conv2d(64, 32, kernel_size=3, stride=1, padding=0)\\n self.conv9 = torch.nn.Conv2d(32, 16, kernel_size=3, stride=1, padding=0)\\n self.conv10 = torch.nn.Conv2d(16, 8, kernel_size=3, stride=1, padding=0)\\n\\n self.bn5 = torch.nn.BatchNorm2d(128)\\n self.bn6 = torch.nn.BatchNorm2d(128)\\n self.bn7 = torch.nn.BatchNorm2d(64)\\n self.bn8 = torch.nn.BatchNorm2d(32)\\n self.bn9 = torch.nn.BatchNorm2d(16)\\n self.bn10 = torch.nn.BatchNorm2d(8)\\n\\n # Block 3:\\n # 2 fully connected with drop out.\\n\\n self.fc1 = torch.nn.Linear( 8 * 59 * 59, 32)\\n self.fc1_bn = torch.nn.BatchNorm1d(32)\\n self.fc_out = torch.nn.Linear(32, 3)\",\n \"def __init__(self, vit_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([vit_layer.attention.attention.query.weight, vit_layer.attention.attention.key.weight, vit_layer.attention.attention.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([vit_layer.attention.attention.query.bias, vit_layer.attention.attention.key.bias, vit_layer.attention.attention.value.bias]))\\n self.out_proj_weight = vit_layer.attention.output.dense.weight\\n self.out_proj_bias = vit_layer.attention.output.dense.bias\\n self.linear1_weight = vit_layer.intermediate.dense.weight\\n self.linear1_bias = vit_layer.intermediate.dense.bias\\n self.linear2_weight = vit_layer.output.dense.weight\\n self.linear2_bias = vit_layer.output.dense.bias\\n self.norm1_eps = vit_layer.layernorm_before.eps\\n self.norm1_weight = vit_layer.layernorm_before.weight\\n self.norm1_bias = vit_layer.layernorm_before.bias\\n self.norm2_eps = vit_layer.layernorm_after.eps\\n self.norm2_weight = vit_layer.layernorm_after.weight\\n self.norm2_bias = vit_layer.layernorm_after.bias\\n self.num_heads = vit_layer.attention.attention.num_attention_heads\\n self.embed_dim = int(vit_layer.attention.attention.attention_head_size * self.num_heads)\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def _Attention(self, name, is_causal=True):\\n p = self.params\\n tr_atten_p = TransformerAttentionLayer.Params().Set(\\n name='transformer_atten',\\n input_dim=p.model_dim,\\n hidden_dim=p.attention_hidden_dim or p.model_dim,\\n is_masked=is_causal,\\n num_heads=p.num_heads,\\n residual_dropout_prob=p.residual_dropout_prob,\\n atten_dropout_prob=p.atten_dropout_prob,\\n fprop_dtype=p.fprop_dtype,\\n add_unnormalized_input=p.selfatten_add_unnormalized_input,\\n )\\n tr_atten_p.atten_tpl.use_bias = p.use_bias\\n tr_atten_p.atten_tpl.enable_value_proj = p.selfatten_enable_value_proj\\n tr_atten_p.atten_tpl.enable_query_scale = p.enable_query_scale\\n tr_atten_p.atten_tpl.enable_per_dim_scale = p.enable_per_dim_scale\\n tr_atten_p.atten_tpl.device_mesh = p.device_mesh\\n tr_atten_p.atten_tpl.weight_split_dims_mapping = (\\n p.weight_split_dims_mapping.dnh)\\n tr_atten_p.atten_tpl.activation_split_dims_mapping.blnh = (\\n p.activation_split_dims_mapping.blnh)\\n tr_atten_p.atten_tpl.activation_split_dims_mapping.bld = (\\n p.activation_split_dims_mapping.bld)\\n if p.deterministic_dropout:\\n tr_atten_p.dropout_tpl = layers.DeterministicDropoutLayer.Params()\\n tr_atten_p.atten_p.dropout_tpl = layers.DeterministicDropoutLayer.Params()\\n\\n return self._Graph(\\n name,\\n ['i'], # input NestedMap with {vec, paddings}\\n ['o'], # output NestedMap with {vec, paddings}\\n ('i.vec->split_i',\\n self.MeshSplit('input_split', p.activation_split_dims_mapping.bld)),\\n ('split_i,split_i,i.paddings->o.vec,unused_prob', tr_atten_p),\\n ('i.paddings->o.paddings', self._Id('id')))\",\n \"def __init__(self, d_model=256, n_levels=4, n_heads=8, n_points=4):\\n super().__init__()\\n if d_model % n_heads != 0:\\n raise ValueError(\\\"d_model must be divisible by n_heads, but got {} and {}\\\".format(d_model, n_heads))\\n _d_per_head = d_model // n_heads\\n # you'd better set _d_per_head to a power of 2 which is more efficient in our CUDA implementation\\n if not _is_power_of_2(_d_per_head):\\n warnings.warn(\\n \\\"You'd better set d_model in MSDeformAttn to make the dimension of each attention head a power of 2 \\\"\\n \\\"which is more efficient in our CUDA implementation.\\\"\\n )\\n self.im2col_step = 64\\n ###\\n self.d_model = d_model\\n self.n_levels = n_levels\\n self.n_heads = n_heads\\n self.n_points = n_points\\n self.sampling_offsets = nn.Dense(d_model, n_heads * n_levels * n_points * 2)\\n self.attention_weights = nn.Dense(d_model, n_heads * n_levels * n_points)\\n self.value_proj = nn.Dense(d_model, d_model)\\n self.output_proj = nn.Dense(d_model, d_model)\",\n \"def identity_block(inputs, kernel_size, filters, stage, block,\\n training):\\n nb_filter1, nb_filter2, nb_filter3 = filters\\n\\n scope_name = 'conv'+str(stage)+block+'_branch'\\n scope_name = 'bn'+str(stage)+block+'_branch'\\n\\n x = slim.conv2d(inputs,nb_filter1,[1,1],stride=[1,1],padding='VALID',activation_fn=None,\\n trainable=training,scope=scope_name+'2a')\\n x = slim.batch_norm(x,scope=scope_name+'2a',is_training=training)\\n x = tf.nn.relu(x)\\n\\n\\n x = slim.conv2d(x,nb_filter2,[kernel_size,kernel_size],stride=[1,1],padding='SAME',\\n activation_fn=None,trainable=training,scope=scope_name+'2b')\\n x = slim.batch_norm(x,scope=scope_name+'2b',is_training=training)\\n x = tf.nn.relu(x)\\n\\n\\n x = slim.conv2d(x,nb_filter3,[1,1],stride=[1,1],padding='VALID',\\n activation_fn=None,trainable=training,scope=scope_name+'2c')\\n x = slim.batch_norm(x,scope=scope_name+'2c',is_training=training)\\n\\n \\n x = tf.add(x,inputs)\\n x = tf.nn.relu(x,name='res'+str(stage)+block+\\\"_out\\\")\\n return x\",\n \"def forward(self, x):\\n sources = list()\\n loc = list()\\n conf = list()\\n\\n # apply vgg up to conv4_3 relu\\n #print('Reached start of vgg')\\n for k in self.vgg._modules.keys():\\n if int(k) < 23:\\n #print('Reached ' + k + ' ', x.size())\\n x = self.vgg._modules[k].cuda()(x)\\n #print('Reached L2Norm')\\n s = self.L2Norm(x)\\n sources.append(s)\\n\\n #print('Reached after L2Norm')\\n # apply vgg up to fc7\\n for k in self.vgg._modules.keys():\\n if int(k) >= 23:\\n #print('Reached ' + k + ' ', x.size())\\n x = self.vgg._modules[k].cuda()(x)\\n sources.append(x)\\n #print('Reached end of VGG')\\n\\n # apply extra layers and cache source layer outputs\\n for k, v in enumerate(self.extras):\\n x = F.relu(v(x), inplace=True)\\n if k % 2 == 1:\\n sources.append(x)\\n\\n # apply multibox head to source layers\\n for (x, l, c) in zip(sources, self.loc, self.conf):\\n loc.append(l(x).permute(0, 2, 3, 1).contiguous())\\n conf.append(c(x).permute(0, 2, 3, 1).contiguous())\\n\\n loc = torch.cat([o.view(o.size(0), -1) for o in loc], 1)\\n conf = torch.cat([o.view(o.size(0), -1) for o in conf], 1)\\n\\n if self.phase == \\\"test\\\":\\n output = self.detect(\\n loc.view(loc.size(0), -1, 4), # loc preds\\n self.softmax(conf.view(-1, self.num_classes)), # conf preds\\n self.priors # default boxes\\n )\\n else:\\n output = (\\n loc.view(loc.size(0), -1, 4),\\n conf.view(conf.size(0), -1, self.num_classes),\\n self.priors\\n )\\n return output\",\n \"def __init__(\\n self,\\n encoder,\\n num_classes=20,\\n shape_size=16,\\n mask_conv=True,\\n min_confidence=0.2,\\n min_remain=0.5,\\n mask_threshold=0.4,\\n ):\\n super(PaPs, self).__init__()\\n self.encoder = encoder\\n self.shape_size = shape_size\\n self.num_classes = num_classes\\n self.min_scale = 1 / shape_size\\n self.register_buffer(\\\"min_confidence\\\", torch.tensor([min_confidence]))\\n self.min_remain = min_remain\\n self.mask_threshold = mask_threshold\\n self.center_extractor = CenterExtractor()\\n\\n enc_dim = encoder.enc_dim\\n stack_dim = encoder.stack_dim\\n self.heatmap_conv = nn.Sequential(\\n ConvLayer(nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\\n padding_mode=\\\"reflect\\\"),\\n nn.Sigmoid(),\\n )\\n\\n self.saliency_conv = ConvLayer(\\n nkernels=[enc_dim, 32, 1], last_relu=False, k=3, p=1,\\n padding_mode=\\\"reflect\\\"\\n )\\n\\n self.shape_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, shape_size ** 2),\\n )\\n\\n self.size_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, stack_dim // 4),\\n nn.BatchNorm1d(stack_dim // 4),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 4, 2),\\n nn.Softplus(),\\n )\\n\\n self.class_mlp = nn.Sequential(\\n nn.Linear(stack_dim, stack_dim // 2),\\n nn.BatchNorm1d(stack_dim // 2),\\n nn.ReLU(),\\n nn.Linear(stack_dim // 2, stack_dim // 4),\\n nn.Linear(stack_dim // 4, num_classes),\\n )\\n\\n if mask_conv:\\n self.mask_cnn = nn.Sequential(\\n nn.Conv2d(in_channels=1, out_channels=16, kernel_size=3, padding=1),\\n nn.GroupNorm(num_channels=16, num_groups=1),\\n nn.ReLU(),\\n nn.Conv2d(in_channels=16, out_channels=16, kernel_size=3, padding=1),\\n nn.ReLU(),\\n nn.Conv2d(in_channels=16, out_channels=1, kernel_size=3, padding=1),\\n )\\n else:\\n self.mask_cnn = None\",\n \"def __init__(self, args):\\n \\n super(MicroNet, self).__init__()\\n self.conv1 = nn.Conv2d(3, 1, kernel_size=1)\\n self.conv2 = nn.Conv2d(1, 29, kernel_size=5)\\n self.maxpool2 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\\n self.conv3 = nn.Conv2d(29, 59, kernel_size=3)\\n self.maxpool3 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\\n self.conv4 = nn.Conv2d(59, 74, kernel_size=3)\\n self.maxpool4 = nn.MaxPool2d(3, stride=2 , ceil_mode=True)\\n self.conv2_drop = nn.Dropout2d()\\n self.conv3_drop = nn.Dropout2d()\\n self.fc1 = nn.Linear(1184, 300)\\n self.fc2 = nn.Linear(300, args.num_classes)\\n self.conv0_bn = nn.BatchNorm2d(3)\\n self.conv1_bn = nn.BatchNorm2d(1)\\n self.conv2_bn = nn.BatchNorm2d(29)\\n self.conv3_bn = nn.BatchNorm2d(59)\\n self.conv4_bn = nn.BatchNorm2d(74)\\n self.dense1_bn = nn.BatchNorm1d(300)\",\n \"def td_identity_block(input_tensor, kernel_size, filters, stage, block, use_conv_bias=True,\\n weight_regularizer=None, bias_regularizer=None, bn_training=False, separate_scale=False):\\n filters1, filters2, filters3 = filters\\n bn_axis = 3\\n conv_name_base = 'res' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n scale_name_base = 'scale' + str(stage) + block + '_branch'\\n eps = 1e-5\\n\\n x = TimeDistributed(Conv2D(filters1, (1, 1), use_bias=use_conv_bias,\\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer,\\n ),\\n name=conv_name_base + '2a')(input_tensor)\\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\\n name=bn_name_base + '2a',)(x, training=bn_training)\\n if separate_scale:\\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2a')(x, training=bn_training)\\n x = Activation('relu')(x)\\n\\n x = TimeDistributed(Conv2D(filters2, kernel_size, padding='same', use_bias=use_conv_bias,\\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer\\n ),\\n name=conv_name_base + '2b')(x)\\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\\n name=bn_name_base + '2b')(x, training=bn_training)\\n if separate_scale:\\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2b')(x, training=bn_training)\\n x = Activation('relu')(x)\\n\\n x = TimeDistributed(Conv2D(filters3, (1, 1), use_bias=use_conv_bias,\\n kernel_regularizer=weight_regularizer, bias_regularizer=bias_regularizer\\n ),\\n name=conv_name_base + '2c')(x)\\n x = TimeDistributed(BatchNormalization(epsilon=eps, axis=bn_axis, trainable=bn_training),\\n name=bn_name_base + '2c')(x, training=bn_training)\\n if separate_scale:\\n x = TimeDistributed(Scale(axis=bn_axis, trainable=bn_training), name=scale_name_base + '2c')(x, training=bn_training)\\n\\n x = layers.add([x, input_tensor])\\n x = TimeDistributed(Activation('relu'))(x)\\n return x\",\n \"def TextEncBlock(L,d,scope=\\\"TextEncBlock\\\"):\\n e = L.shape.as_list()[2]\\n\\n with tf.variable_scope(scope):\\n conv_params = {\\\"filters\\\":2*d,\\\"kernel_size\\\":1,\\\"padding\\\":\\\"same\\\"} \\n with tf.variable_scope(\\\"C_block1\\\"):\\n L1 = tf.nn.relu(conv1d(inputs=L,**conv_params)) # relu(conv)\\n L2 = conv1d(inputs=L1,**conv_params) # conv\\n # hc_dilation_blocks \\n L3_HC1_1 = hc_dilation_block(L2,local_scope=\\\"HC_dilation_block1\\\")\\n L3_HC1_2 = hc_dilation_block(L3_HC1_1,local_scope=\\\"HC_dilation_block2\\\")\\n # highway_conv_block2\\n with tf.variable_scope(\\\"HC_block2\\\"):\\n L4_HC2_1 = highway_activation_conv(L3_HC1_2,kernel_size=3,scope=\\\"HC1\\\")\\n L4_HC2_2 = highway_activation_conv(L4_HC2_1,kernel_size=3,scope=\\\"HC2\\\")\\n # highway_conv_block3\\n with tf.variable_scope(\\\"HC_block3\\\"):\\n L5_HC3_1 = highway_activation_conv(L4_HC2_2,kernel_size=1,scope=\\\"HC1\\\")\\n KV = highway_activation_conv(L5_HC3_1,kernel_size=1,scope=\\\"HC2\\\")\\n\\n return KV[:,:,:d], KV[:,:,d:] # key, value\",\n \"def __init__(self, bart_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([bart_layer.self_attn.q_proj.weight, bart_layer.self_attn.k_proj.weight, bart_layer.self_attn.v_proj.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([bart_layer.self_attn.q_proj.bias, bart_layer.self_attn.k_proj.bias, bart_layer.self_attn.v_proj.bias]))\\n self.out_proj_weight = bart_layer.self_attn.out_proj.weight\\n self.out_proj_bias = bart_layer.self_attn.out_proj.bias\\n self.linear1_weight = bart_layer.fc1.weight\\n self.linear1_bias = bart_layer.fc1.bias\\n self.linear2_weight = bart_layer.fc2.weight\\n self.linear2_bias = bart_layer.fc2.bias\\n self.norm1_eps = bart_layer.self_attn_layer_norm.eps\\n self.norm1_weight = bart_layer.self_attn_layer_norm.weight\\n self.norm1_bias = bart_layer.self_attn_layer_norm.bias\\n self.norm2_eps = bart_layer.final_layer_norm.eps\\n self.norm2_weight = bart_layer.final_layer_norm.weight\\n self.norm2_bias = bart_layer.final_layer_norm.bias\\n self.num_heads = bart_layer.self_attn.num_heads\\n self.embed_dim = bart_layer.self_attn.embed_dim\\n self.is_last_layer = False\\n self.validate_bettertransformer()\",\n \"def __init__(self, vilt_layer, config):\\n super().__init__(config)\\n self.in_proj_weight = nn.Parameter(torch.cat([vilt_layer.attention.attention.query.weight, vilt_layer.attention.attention.key.weight, vilt_layer.attention.attention.value.weight]))\\n self.in_proj_bias = nn.Parameter(torch.cat([vilt_layer.attention.attention.query.bias, vilt_layer.attention.attention.key.bias, vilt_layer.attention.attention.value.bias]))\\n self.out_proj_weight = vilt_layer.attention.output.dense.weight\\n self.out_proj_bias = vilt_layer.attention.output.dense.bias\\n self.linear1_weight = vilt_layer.intermediate.dense.weight\\n self.linear1_bias = vilt_layer.intermediate.dense.bias\\n self.linear2_weight = vilt_layer.output.dense.weight\\n self.linear2_bias = vilt_layer.output.dense.bias\\n self.norm1_eps = vilt_layer.layernorm_before.eps\\n self.norm1_weight = vilt_layer.layernorm_before.weight\\n self.norm1_bias = vilt_layer.layernorm_before.bias\\n self.norm2_eps = vilt_layer.layernorm_after.eps\\n self.norm2_weight = vilt_layer.layernorm_after.weight\\n self.norm2_bias = vilt_layer.layernorm_after.bias\\n self.num_heads = vilt_layer.attention.attention.num_attention_heads\\n self.embed_dim = int(vilt_layer.attention.attention.attention_head_size * self.num_heads)\\n self.is_last_layer = False\\n self.norm_first = True\\n self.validate_bettertransformer()\",\n \"def __init__(self, dm, h, hidden, drop_rate=0.1):\\n super().__init__()\\n\\n self.mha1 = MultiHeadAttention(dm, h)\\n self.mha2 = MultiHeadAttention(dm, h)\\n\\n self.ffn = point_wise_feed_forward_network(dm, hidden)\\n\\n self.layernorm1 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n self.layernorm2 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n self.layernorm3 = tf.keras.layers.LayerNormalization(epsilon=1e-6)\\n\\n self.dropout1 = tf.keras.layers.Dropout(drop_rate)\\n self.dropout2 = tf.keras.layers.Dropout(drop_rate)\\n self.dropout3 = tf.keras.layers.Dropout(drop_rate)\",\n \"def __init__(self):\\n super(Generator, self).__init__()\\n\\n # output feature map will have the size of 8x8\\n self.first_deconv = TransitionUp(in_channels=3, out_channels=256, stride=2, padding=1, kernel_size=4)\\n self.first_batch_norm = nn.BatchNorm2d(256)\\n\\n # output feature map will have the size of 16x16\\n self.second_deconv = TransitionUp(in_channels=256, out_channels=128, stride=2, padding=1, kernel_size=4)\\n self.second_batch_norm = nn.BatchNorm2d(128)\\n\\n # output feature map will have the size of 32x32\\n self.third_deconv = TransitionUp(in_channels=128, out_channels=32, stride=2, padding=1, kernel_size=4)\\n self.third_batch_norm = nn.BatchNorm2d(32)\\n\\n # output feature map will have the size of 64x64\\n self.fourth_deconv = TransitionUp(in_channels=32, out_channels=16, stride=2, padding=1, kernel_size=4)\\n self.fourth_batch_norm = nn.BatchNorm2d(16)\\n\\n # output feature map will have the size of 128x128\\n self.fifth_deconv = TransitionUp(in_channels=16, out_channels=16, stride=1, padding=1, kernel_size=4)\\n self.fifth_batch_norm = nn.BatchNorm2d(16)\\n\\n # output feature map will have the size of 128x128\\n self.sixth_deconv = TransitionUp(in_channels=16, out_channels=16, stride=1, padding=1, kernel_size=4)\\n self.sixth_batch_norm = nn.BatchNorm2d(16)\\n\\n # output feature map will have the size of 128x128\\n self.seventh_deconv = TransitionUp(in_channels=16, out_channels=3, stride=1, padding=1, kernel_size=4)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.6663586","0.6597379","0.64268494","0.6360674","0.6338788","0.6257444","0.6235589","0.6209039","0.60610014","0.59759337","0.59740186","0.59430027","0.590821","0.5900392","0.58956826","0.5895034","0.58770245","0.58639365","0.5794944","0.5788988","0.5763849","0.57396513","0.57335454","0.5717228","0.5687134","0.5684168","0.5671052","0.5664152","0.5664052","0.5639643","0.56332403","0.5623402","0.5618749","0.5616841","0.5599451","0.55970484","0.5594293","0.55888164","0.558881","0.55871254","0.55855006","0.5576405","0.5553441","0.55489033","0.5547713","0.55449694","0.5538067","0.5537847","0.5528931","0.5525154","0.552089","0.55190223","0.5516318","0.5513863","0.5505853","0.55054265","0.5504818","0.5504414","0.5495378","0.5476525","0.54702646","0.54635566","0.54567075","0.5454969","0.5449905","0.54470944","0.5444188","0.5433262","0.5433254","0.54294515","0.5429044","0.54193884","0.54168814","0.5416622","0.5415773","0.5414807","0.5414179","0.541274","0.53953415","0.53948796","0.5393287","0.53889835","0.53887725","0.53870195","0.5385905","0.5382028","0.5375711","0.53697115","0.5367631","0.53643805","0.53629297","0.53619504","0.536036","0.53591424","0.5351359","0.53491396","0.53458726","0.5344204","0.53344744","0.53327936"],"string":"[\n \"0.6663586\",\n \"0.6597379\",\n \"0.64268494\",\n \"0.6360674\",\n \"0.6338788\",\n \"0.6257444\",\n \"0.6235589\",\n \"0.6209039\",\n \"0.60610014\",\n \"0.59759337\",\n \"0.59740186\",\n \"0.59430027\",\n \"0.590821\",\n \"0.5900392\",\n \"0.58956826\",\n \"0.5895034\",\n \"0.58770245\",\n \"0.58639365\",\n \"0.5794944\",\n \"0.5788988\",\n \"0.5763849\",\n \"0.57396513\",\n \"0.57335454\",\n \"0.5717228\",\n \"0.5687134\",\n \"0.5684168\",\n \"0.5671052\",\n \"0.5664152\",\n \"0.5664052\",\n \"0.5639643\",\n \"0.56332403\",\n \"0.5623402\",\n \"0.5618749\",\n \"0.5616841\",\n \"0.5599451\",\n \"0.55970484\",\n \"0.5594293\",\n \"0.55888164\",\n \"0.558881\",\n \"0.55871254\",\n \"0.55855006\",\n \"0.5576405\",\n \"0.5553441\",\n \"0.55489033\",\n \"0.5547713\",\n \"0.55449694\",\n \"0.5538067\",\n \"0.5537847\",\n \"0.5528931\",\n \"0.5525154\",\n \"0.552089\",\n \"0.55190223\",\n \"0.5516318\",\n \"0.5513863\",\n \"0.5505853\",\n \"0.55054265\",\n \"0.5504818\",\n \"0.5504414\",\n \"0.5495378\",\n \"0.5476525\",\n \"0.54702646\",\n \"0.54635566\",\n \"0.54567075\",\n \"0.5454969\",\n \"0.5449905\",\n \"0.54470944\",\n \"0.5444188\",\n \"0.5433262\",\n \"0.5433254\",\n \"0.54294515\",\n \"0.5429044\",\n \"0.54193884\",\n \"0.54168814\",\n \"0.5416622\",\n \"0.5415773\",\n \"0.5414807\",\n \"0.5414179\",\n \"0.541274\",\n \"0.53953415\",\n \"0.53948796\",\n \"0.5393287\",\n \"0.53889835\",\n \"0.53887725\",\n \"0.53870195\",\n \"0.5385905\",\n \"0.5382028\",\n \"0.5375711\",\n \"0.53697115\",\n \"0.5367631\",\n \"0.53643805\",\n \"0.53629297\",\n \"0.53619504\",\n \"0.536036\",\n \"0.53591424\",\n \"0.5351359\",\n \"0.53491396\",\n \"0.53458726\",\n \"0.5344204\",\n \"0.53344744\",\n \"0.53327936\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6992421"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90880,"cells":{"query":{"kind":"string","value":"Pass Embedded Patches through the layers"},"document":{"kind":"string","value":"def call(self, inputs):\n # Feed attention\n norm_attention = self.norm_attention(inputs)\n\n attention = self.attention(query=norm_attention, value=norm_attention)\n\n # Skip Connection\n attention += inputs\n\n # Feed MLP\n outputs = self.mlp(self.norm_mlp(attention))\n\n # Skip Connection\n outputs += attention\n\n return outputs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def run(layers):","def build_layers(self):\n raise NotImplementedError","def layers(self): # -> LayerView:\n ...","def UpdateLayers(self):\n pass","def layers(self, layers):\n self._layers = layers\n prev = None\n for layer in layers:\n if not layer.inputs and prev is not None:\n layer.inputs = [prev]\n prev = layer","def layers(self, layers):\n\n self._layers = layers","def __init__(self, layers):\n\n\t\tself.layers = layers","def handle_layers(context, model, toplayer, layerids, materials, update, import_hidden=False):\n #setup main container to hold all layer collections\n layer_col_id=\"Layers\"\n if not layer_col_id in context.blend_data.collections:\n layer_col = context.blend_data.collections.new(name=layer_col_id)\n try:\n toplayer.children.link(layer_col)\n except Exception:\n pass\n else:\n #If \"Layers\" collection is in place, we assume the plugin had imported 3dm before\n layer_col = context.blend_data.collections[layer_col_id]\n\n # build lookup table for LayerTable index\n # from GUID, create collection for each\n # layer\n for lid, l in enumerate(model.Layers):\n if not l.Visible and not import_hidden:\n continue\n lcol = utils.get_iddata(context.blend_data.collections, l.Id, l.Name, None)\n layerids[str(l.Id)] = (lid, lcol)\n utils.tag_data(layerids[str(l.Id)][1], l.Id, l.Name)\n '''\n matname = l.Name + \"+\" + str(l.Id)\n if matname not in materials:\n laymat = utils.get_iddata(context.blend_data.materials, l.Id, l.Name, None)\n if update:\n\t laymat.use_nodes = True\n\t r, g, b, _ = l.Color\n\t principled = PrincipledBSDFWrapper(laymat, is_readonly=False)\n\t principled.base_color = (r/255.0, g/255.0, b/255.0)\n materials[matname] = laymat\n '''\n # second pass so we can link layers to each other\n for l in model.Layers:\n # link up layers to their parent layers\n if str(l.ParentLayerId) in layerids:\n parentlayer = layerids[str(l.ParentLayerId)][1]\n try:\n parentlayer.children.link(layerids[str(l.Id)][1])\n except Exception:\n pass\n # or to the top collection if no parent layer was found\n else:\n try:\n layer_col.children.link(layerids[str(l.Id)][1])\n except Exception:\n pass","def consume_layer(self, reports):\n layer_list = []\n layer_count = 1\n for report in reports:\n layer = create_image_layer(report)\n layer.layer_index = layer_count\n layer_list.append(layer)\n layer_count += 1\n return layer_list","def __call__(cls, *args, **kwargs):\n layer = super(LayerAspect, cls).__call__(*args, **kwargs)\n\n if Job.Current:\n Job.Current.addLayer(layer)\n \n layer.afterInit()\n return layer","def layers(self, layers):\n self._layers = layers\n self.thetas = []\n prev = None\n for layer in layers:\n if not layer.inputs and prev is not None:\n layer.inputs = [prev]\n prev = layer\n self.thetas.extend(layer.thetas())","def layers(self, x):\n raise NotImplementedError","def setup_layer_structure(self):\n self.page_rank_convolution_1 = self.layer(self.feature_number, self.args.layers[0], self.args.iterations, self.args.alpha)\n self.page_rank_convolution_2 = self.layer(self.args.layers[0], self.args.layers[1], self.args.iterations, self.args.alpha)\n self.page_rank_convolution_3 = self.layer(self.args.layers[1], self.class_number, self.args.iterations, self.args.alpha)","def build_models(config):\n inputs = Input(shape=(config['patch_height'], config['patch_width'], config['depth'], config['channel']),\n name='patchimg')\n\n kernelinitfun = keras.initializers.RandomNormal(mean=0.0, stddev=0.1, seed=None)\n activationfun = 'relu'\n # kernelinitfun = 'glorot_normal'\n\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_1')(inputs)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_2')(x)\n # x = Dropout(0.3)(x)\n x = BatchNormalization(name='bn1')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp1', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_1')(x)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_2')(x)\n # x = Dropout(0.2)(x)\n x = BatchNormalization(name='bn2')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp2', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_1')(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_2')(x)\n # x = Dropout(0.5)(x)\n x = BatchNormalization(name='bn3')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp3', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_1')(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_2')(x)\n # x = Dropout(0.5)(x)\n x = BatchNormalization(name='bn4')(x)\n x = Activation(activationfun)(x)\n\n x4 = Flatten(name='aux_fx')(x)\n\n source_classifier = Dropout(0.5)(x4)\n source_classifier = Dense(512, activation='softmax', name=\"mo1\")(source_classifier)\n source_classifier = Dropout(0.5)(source_classifier)\n source_classifier = Dense(128, activation='softmax', name=\"mo2\")(source_classifier)\n # source_classifier = Dropout(0.3)(source_classifier)\n source_classifier = Dense(1, name=\"mo\")(source_classifier)\n\n domain_classifier = Dense(32, activation='linear', name=\"do4\")(x4)\n domain_classifier = BatchNormalization(name=\"do5\")(domain_classifier)\n domain_classifier = Activation(\"elu\", name=\"do6\")(domain_classifier)\n domain_classifier = Dropout(0.5)(domain_classifier)\n\n domain_classifier = Dense(2, activation='softmax', name=\"do\")(domain_classifier)\n\n adamop = keras.optimizers.Adam(learning_rate=1e-5, beta_1=0.9, beta_2=0.999, amsgrad=False)\n comb_model = Model(inputs=inputs, outputs=[source_classifier, domain_classifier])\n comb_model.compile(optimizer=adamop,\n loss={'mo': 'mae', 'do': 'categorical_crossentropy'},\n loss_weights={'mo': 1, 'do': 2}, metrics=['accuracy'], )\n\n source_classification_model = Model(inputs=inputs, outputs=[source_classifier])\n source_classification_model.compile(optimizer=adamop,\n loss={'mo': 'mae'}, metrics=['accuracy'], )\n\n domain_classification_model = Model(inputs=inputs, outputs=[domain_classifier])\n domain_classification_model.compile(optimizer=adamop,\n loss={'do': 'categorical_crossentropy'}, metrics=['accuracy'])\n\n embeddings_model = Model(inputs=inputs, outputs=[x4])\n embeddings_model.compile(optimizer=adamop, loss='categorical_crossentropy', metrics=['accuracy'])\n\n return comb_model, source_classification_model, domain_classification_model, embeddings_model","def __init__(self, parent, layer):\n pass","def __init__(self, kernel_size, filters, stage, block):\n super().__init__(name='identity' + str(stage) + block)\n filters1, filters2, filters3 = filters\n if K.image_data_format() == 'channels_last':\n bn_axis = 3\n else:\n bn_axis = 1\n\n conv_name_base = 'res' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n self.conv1 = layers.Conv2D(\n filters1, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2a')\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\n self.act1 = layers.Activation('relu')\n\n self.conv2 = layers.Conv2D(\n filters2,\n kernel_size,\n padding='same',\n kernel_initializer='he_normal',\n name=conv_name_base + '2b')\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\n self.act2 = layers.Activation('relu')\n\n self.conv3 = layers.Conv2D(\n filters3, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2c')\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\n\n self.add = layers.Add()\n self.act = layers.Activation('relu')","def setup_layers(self):\n if self.args.model == \"exact\":\n self.layer = PPNPLayer\n else:\n self.layer = APPNPLayer\n self.setup_layer_structure()","def set_hooks(self):\n for layer_idx, layer in enumerate(self.vgg_features):\n if layer_idx in self.layers_to_watch:\n layer.register_forward_hook(self.layer_watch_hooks(layer_idx))","def __init__(self, kernel_size, filters, stage, block, strides=(2, 2)):\n super().__init__(name='conv_block' + str(stage) + block)\n filters1, filters2, filters3 = filters\n if K.image_data_format() == 'channels_last':\n bn_axis = 3\n else:\n bn_axis = 1\n conv_name_base = 'res' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n self.conv1 = layers.Conv2D(\n filters1, (1, 1),\n strides=strides,\n kernel_initializer='he_normal',\n name=conv_name_base + '2a')\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\n self.act1 = layers.Activation('relu')\n\n self.conv2 = layers.Conv2D(\n filters2,\n kernel_size,\n padding='same',\n kernel_initializer='he_normal',\n name=conv_name_base + '2b')\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\n self.act2 = layers.Activation('relu')\n\n self.conv3 = layers.Conv2D(\n filters3, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2c')\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\n\n self.shortcut_conv = layers.Conv2D(\n filters3, (1, 1),\n strides=strides,\n kernel_initializer='he_normal',\n name=conv_name_base + '1')\n self.shortcut_bn = layers.BatchNormalization(\n axis=bn_axis, name=bn_name_base + '1')\n\n self.add = layers.Add()\n self.act = layers.Activation('relu')","def _build(self):\n with tf.variable_scope (self.name + '_architecutre') as scope:\n images_square = unflatten_layer ( self.images )\n visualize_images(images_square)\n\n # Conv Layer 1\n conv1_out, params = conv_2d_layer ( input = images_square,\n neurons = CONV_1_N,\n filter_size = CONV_1_FILT,\n name = 'enc_conv_1',\n visualize = True )\n process_params(params, name = self.name)\n e1_params = params\n pool1_out = max_pool_2d_layer ( input = conv1_out, name = 'enc_pool_1')\n # lrn1_out = local_response_normalization_layer (pool1_out, name = 'lrn_1' )\n\n # Conv Layer 2\n conv2_out, params = conv_2d_layer ( input = pool1_out,\n neurons = CONV_2_N,\n filter_size = CONV_2_FILT,\n name = 'enc_conv_2' )\n process_params(params, name = self.name)\n e2_params = params\n pool2_out = max_pool_2d_layer ( input = conv2_out, name = 'enc_pool_2')\n # lrn2_out = local_response_normalization_layer (pool2_out, name = 'lrn_2' )\n\n flattened = flatten_layer(pool2_out)\n\n # Dropout Layer 1 \n flattened_dropout = dropout_layer ( input = flattened,\n prob = self.dropout_prob,\n name = 'enc_dropout_1') \n\n # Dot Product Layer 1\n fc1_out, params = dot_product_layer ( input = flattened_dropout,\n neurons = HIDDEN_1,\n name = 'enc_dot_1')\n process_params(params, name = self.name)\n e3_params = params \n\n # Dropout Layer 2 \n fc1_out_dropout = dropout_layer ( input = fc1_out,\n prob = self.dropout_prob,\n name = 'enc_dropout_2')\n # Dot Product Layer 2\n fc2_out, params = dot_product_layer ( input = fc1_out_dropout, \n neurons = HIDDEN_2,\n name = 'enc_dot_2')\n process_params(params, name = self.name)\n e4_params = params \n\n # Dropout Layer 3 \n fc2_out_dropout = dropout_layer ( input = fc2_out,\n prob = self.dropout_prob,\n name = 'enc_dropout_3')\n \n # Dot Product Layer 2\n self.codeword, params = dot_product_layer ( input = fc2_out_dropout, \n neurons = CODEWORD_LENGTH,\n activation = CODE_ACTIVATION,\n name = 'enc_dot_2')\n process_params(params, name = self.name)\n process_codeword_normalization_regularizer(self.codeword, \n coeff = AUTOENCODER_CODEWORD_COEFF,\n name = self.name)\n e5_params = params \n # tf.summary.histogram('codewords', self.codeword)\n # self.hash = threshold_layer ( input = self.codeword,\n # name = 'hash')\n # process_hash_regularizer(self.codeword, coeff = AUTOENCODER_HASH_COEFF,\n # name = self.name)\n\n # Decoder ... \n decoder_1_out, params = dot_product_layer ( input = self.codeword, \n neurons = HIDDEN_2,\n params = [tf.transpose(e5_params[0]), None],\n name = 'decoder_dot_1')\n d1_params = params\n process_params([params[1]], name = self.name)\n \n dec_1_out_dropout = dropout_layer ( input = decoder_1_out,\n prob = self.dropout_prob,\n name = 'dec_dropout_1')\n\n decoder_2_out, params = dot_product_layer ( input = dec_1_out_dropout, \n neurons = HIDDEN_1,\n params = [tf.transpose(e4_params[0]), None],\n name = 'decoder_dot_2')\n d2_params = params\n process_params([params[1]], name = self.name)\n \n # dropout 2\n dec_2_out_dropout = dropout_layer ( input = decoder_2_out,\n prob = self.dropout_prob,\n name = 'dec_dropout_2')\n\n decoder_3_out, params = dot_product_layer ( input = dec_2_out_dropout, \n neurons = 1250,\n params = [tf.transpose(e3_params[0]), None],\n name = 'decoder_dot_3')\n d3_params = params\n process_params([params[1]], name = self.name)\n\n # DeConv Layer 1\n # The output shapes need to be changed according to architecture.\n\n dec_3_square = unflatten_layer ( decoder_3_out, channels = CONV_2_N )\n upsample_1 = upsampling_layer (dec_3_square, size = (10,10), name = 'dec_upsampling_1')\n\n deconv1_out, params = deconv_2d_layer ( input = upsample_1,\n neurons = CONV_1_N,\n filter_size = CONV_2_FILT,\n output_shape = (12,12),\n # n_outs = MINI_BATCH_SIZE,\n stride = (1,1,1,1), \n params = [e2_params[0], None], \n name = 'dec_deconv_1' )\n\n process_params([params[1]], name = self.name)\n d4_params = params\n\n # DeConv Layer 2\n upsample_2 = upsampling_layer (deconv1_out, size = (24,24), name = 'dec_upsampling_2')\n decoded_images_square, params = deconv_2d_layer ( input = upsample_2,\n neurons = 1,\n filter_size = CONV_1_FILT,\n stride = (1,1,1,1),\n output_shape = (28,28),\n # n_outs = MINI_BATCH_SIZE, \n params = [e1_params[0], None], \n activation = 'tanh', \n name = 'dec_deconv_2' )\n \n process_params([params[1]], name = self.name)\n d5_params = params \n \n self.decoded = flatten_layer (decoded_images_square, in_shp = [-1, 28, 28, 1])\n visualize_images(decoded_images_square, name = 'decoded')\n # This is because transpose don't initialize.\n self.params = [ [e5_params[0], d1_params[1] ],\n [e4_params[0], d2_params[1] ],\n [e3_params[0], d3_params[1] ],\n [e2_params[0], d4_params[1] ],\n [e1_params[0], d5_params[1] ] ]\n\n with tf.variable_scope (self.name + '_objectives') as scope: \n with tf.variable_scope( self.name + '_decoder_error') as scope:\n reconstruction_error = rmse(self.images, self.decoded) \n tf.add_to_collection( self.name + '_objectives', reconstruction_error ) \n tf.summary.scalar('reconstruction_error', reconstruction_error)\n\n self._cook_optimizer( \n lr = AUTOENCODER_LR, \n optimizer = AUTOENCODER_OPTIMIZER,\n l1_coeff = AUTOENCODER_L1_COEFF,\n l2_coeff = AUTOENCODER_WEIGHT_DECAY_COEFF)","def setup_layers(self, params):\n\n #\n # Input layer\n #\n\n # Input and encoding layers\n layer_stack = [\n tf.placeholder(\n tf.int32,\n name=\"input\",\n shape=[None, params.in_size]\n )\n ]\n\n in_layer = layer_stack[0]\n\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\n character_depth = len(good_chars) + 2\n\n layer_stack.append(\n self.make_onehot_encode_layer(\n layer_stack[-1],\n character_depth\n )\n )\n\n # Add conv layers\n for idx, num_nodes in enumerate(params.conv_nodes):\n layer_stack.append(\n self.make_conv_layer(\n layer_stack[-1],\n num_nodes,\n 'conv_{}'.format(idx),\n params.conv_params\n )\n )\n\n # Flatten convolutional layers\n layer_stack.append(\n tf.contrib.layers.flatten(\n layer_stack[-1]\n )\n )\n\n # Add dense layers\n for idx, num_nodes in enumerate(params.dense_nodes):\n layer_stack.append(\n self.make_dense_layer(\n layer_stack[-1],\n num_nodes,\n 'dense_{}'.format(idx),\n params.dense_params\n )\n )\n\n # Output layer is broken into two pieces. The pre/post the application\n # of the activation function. That's because the loss functions\n # for cross-entropy rely on the pre-activation scores in preact\n\n preact_out_layer, out_layer = self.make_dense_output_layer(\n layer_stack[-1],\n params.out_size,\n params.output_params\n )\n\n target_layer = tf.placeholder(\n \"float\",\n name=\"target\",\n shape=[None, params.out_size]\n )\n\n if params.output_params.activation in ['sigmoid']:\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\n else:\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\n\n return in_layer, out_layer, target_layer, loss","def __call__(self, layers, inputs, name=None):\r\n def call_layer(l, x):\r\n if hasattr(l, \"call_by_computation_context\"):\r\n return l.call_by_computation_context(x, *self.args, **self.kwargs)\r\n else:\r\n return l(x, *self.args, **self.kwargs)\r\n\r\n if not isinstance(layers, (list, tuple)):\r\n # Single layer, just call\r\n return call_layer(layers, inputs)\r\n else:\r\n # Flatten the layers\r\n layers = tf.nest.flatten(layers)\r\n outputs = inputs\r\n\r\n # Add a name scope for wrapping\r\n if name is not None:\r\n scope_name = name\r\n else:\r\n scope_name = commonprefix([layer.name for layer in layers]).strip(\"-/\\\\?_ \\t\\n\") or \"ComputationContext\"\r\n with tf.name_scope(scope_name):\r\n for l in layers:\r\n outputs = call_layer(l, outputs)\r\n\r\n return outputs","def __init__(self, layer_nest):\n self._layer_nest = layer_nest\n super().__init__()","def generate_legacy_layers(self, images_map, content_retriever):\n pass","def generate_legacy_layers(self, images_map, content_retriever):\n pass","def preview_layerset(ls, size = 100, spacing = 100):\n D = Device()\n scale = size/100\n num_layers = len(ls._layers)\n matrix_size = int(np.ceil(np.sqrt(num_layers)))\n sorted_layers = sorted(ls._layers.values(),\n key = lambda x: (x.gds_layer, x.gds_datatype))\n for n, layer in enumerate(sorted_layers):\n R = rectangle(size = (100*scale, 100*scale), layer = layer)\n T = text(text = '%s\\n%s / %s'\\\n % (layer.name, layer.gds_layer, layer.gds_datatype),\n size = 20*scale,\n justify = 'center',\n layer = layer)\n\n T.move((50*scale, -20*scale))\n xloc = n % matrix_size\n yloc = int(n // matrix_size)\n D.add_ref(R)\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\n D.add_ref(T)\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\n return D","def add_layers(self, layers):\n\n existing_layers = self.layers\n assert len(existing_layers) > 0\n for layer in layers:\n assert layer.get_mlp() is None\n layer.set_mlp(self)\n layer.set_input_space(existing_layers[-1].get_output_space())\n existing_layers.append(layer)\n assert layer.layer_name not in self.layer_names\n self.layer_names.add(layer.layer_name)","def build(self):\n\n layers = GiraffeLayer.get_all_structural()\n \n for layer in layers:\n\n self.add_objects_from_layer(layer)\n\n return self","def init_nnpatches(self):\n # nnpatches = self._generate_nnpatches()\n\n # Register `nnpatches` to this `nnmodel`\n # Refer: NNFramework._init_model_params(...)\n\n # # Register this `nnmodel` to `nnpatch`\n # for nnpatch in nnpatches:\n # nnpatch.add_nnmodels(self)","def set_up_rlayer(self, rlname, rlname_other=None, include_layers=None,\n exclude_layers=None, mask_layers=None):\n scene = self.set_as_active()\n layer_numbers = constants.layer_numbers\n w_var.rlname = rlname\n\n if include_layers is None:\n include_layers = w_var.layer_numbers_all_used\n\n if exclude_layers is None:\n exclude_layers = []\n\n if mask_layers is None:\n mask_layers = []\n\n if w_var.cb_clear_rlayers:\n for layer in scene.render.layers[:-1]:\n scene.render.layers.remove(layer)\n\n scene.render.layers.active.name = rlname\n scene.render.layers.active.use = True\n\n new_rlayer = scene.render.layers.active\n\n # if not clearing render layers: creates new one\n else:\n new_rlayer = scene.render.layers.new(rlname)\n scene.render.layers.active = new_rlayer\n\n # there needs to be two render layers in the same scene for freestyle compositing\n if w_var.cb_composited:\n w_var.rlname_other = rlname_other\n other_rlayer = scene.render.layers.new(rlname_other)\n other_rlayer.layers[19] = True\n scene.render.layers[rlname_other].layers_zmask = (False,) * 20\n\n if w_var.cb_ao:\n scene.render.layers[rlname].use_pass_ambient_occlusion = True\n\n if w_var.cb_composited:\n scene.render.layers[rlname_other].use_pass_ambient_occlusion = True\n\n # because I can't deactivate a layer if it is the only active one\n new_rlayer.layers[19] = True\n \n scene.render.layers[rlname].layers_exclude = (False,) * 20\n scene.render.layers[rlname].layers_zmask = (False,) * 20\n\n for i in layer_numbers:\n if w_var.cb_composited:\n if i in w_var.layer_numbers_affected:\n scene.render.layers[rlname].layers[i] = True\n scene.render.layers[rlname_other].layers_zmask[i] = True\n\n else:\n scene.render.layers[rlname].layers[i] = False\n\n if i in w_var.layer_numbers_other:\n scene.render.layers[rlname_other].layers[i] = True\n\n else:\n scene.render.layers[rlname_other].layers[i] = False\n\n else:\n if i in include_layers:\n scene.render.layers[rlname].layers[i] = True\n\n else:\n scene.render.layers[rlname].layers[i] = False\n\n if i in mask_layers:\n scene.render.layers[rlname].layers_zmask[i] = True\n\n if i in exclude_layers:\n scene.render.layers[rlname].layers_exclude[i] = True","def draw_layers(self):\n\t\tfor z in xrange(0,16):\n\t\t\t#create surface for this layer\n\t\t\tsrf = pygame.Surface((16,128))\n\t\t\tfor x in xrange(0,16):\n\t\t\t\tfor y in xrange(0,128):\n\t\t\t\t\tv = self.data[ self.xyz_to_offset( x,y,z) ]\n\t\t\t\t\tif v != 0:\n\t\t\t\t\t\tsrf.fill( BLOCKS.get(v, [0,0])[1], \t(x, 127 -y, 1, 1 ))\n\t\t\t#save layer to dict for this chunk\n\t\t\tself.layers[z] = srf","def process_layers(size: Sequence[int],\n layers: Union[Source, Layer, Sequence[Layer]],\n a: Union[None, np.ndarray] = None,\n status: Union[None, Queue] = None) -> np.ndarray:\n # If no destination array was sent, it means we are either at the\n # start of the layer processing or we are starting the processing\n # of a layer group. If we are starting processing, then this will\n # contain the final image data. If we are starting the processing\n # of a layer group, this allows the layers in the group to only\n # blend with the other layers in the group before they blend with\n # the layers outside the group.\n if a is None:\n a = np.zeros(size, dtype=float)\n\n # If we got a sequence of layers, we process them recursively and\n # return the result.\n if isinstance(layers, Sequence):\n for layer in layers:\n a = process_layers(size, layer, a, status)\n return a\n\n # If we got a source layer, process it.\n if isinstance(layers.source, Source):\n if status is not None:\n src_name = layers.source.__class__.__name__\n status.put((sw.MSG, f'Rendering {src_name}...'))\n kwargs = {\n 'source': layers.source,\n 'size': size,\n 'location': layers.location,\n 'filters': layers.filters,\n }\n b = render_source(**kwargs)\n if status is not None:\n src_name = layers.source.__class__.__name__\n status.put((sw.MSG, f'Rendered {src_name}.'))\n status.put((sw.PROG,))\n\n # Otherwise we got a container layer, process its source and run\n # any filters that are set on the layer.\n else:\n new_size = f.preprocess(size, layers.filters)\n b = process_layers(new_size, layers.source, None, status)\n b = f.process(b, layers.filters)\n b = f.postprocess(b, layers.filters)\n\n # There are two possibilities for how the layers should be\n # blended: masked or unmasked. Masked blends will have a\n # Source in the mask attribute, which needs to be sent\n # to the blending operation.\n if layers.mask is not None:\n if status is not None:\n src_name = layers.mask.__class__.__name__\n status.put((sw.MSG, f'Rendering {src_name}...'))\n kwargs = {\n 'source': layers.mask,\n 'size': size,\n 'filters': layers.mask_filters,\n }\n mask = render_source(**kwargs)\n if status is not None:\n src_name = layers.mask.__class__.__name__\n status.put((sw.MSG, f'Rendered {src_name}.'))\n status.put((sw.PROG,))\n a, b, mask = _normalize_color_space(a, b, mask)\n return layers.blend(a, b, layers.blend_amount, mask)\n\n # Unmasked layers have no mask, so Blend the image data and return.\n a, b = _normalize_color_space(a, b)\n return layers.blend(a, b, layers.blend_amount)","def make_layers(self):\n w, h = self.image.get_size()\n shrink = pg.transform.smoothscale(self.image, (w//2, h//2))\n self.mid_image = tools.tile_surface((w,h), shrink, True)\n shrink = pg.transform.smoothscale(self.image, (w//4, h//4))\n self.base = tools.tile_surface(prepare.SCREEN_SIZE, shrink, True)","def __initAvailableLayerTypes(self):\n from backend.caffe.path_loader import PathLoader\n caffe = PathLoader().importCaffe()\n layerNameMainParts = list(caffe.layer_type_list())\n\n res = {}\n paramsPerLayerType = {}\n\n # calculate common parameters of all layer types\n # by removing all which will be used for one specific layer type only\n # also keep in mind which ones have been removed to readd them to specific layers\n commonParams = self._availableParameterGroupDescriptors[\"LayerParameter\"].parameter() #use .parameter() on purpose\n layerSpecificParameters = set()\n for nameMainPart in layerNameMainParts:\n specificParamsName = [nameMainPart + \"Parameter\"]\n if moreLayerNameParameter.has_key(nameMainPart):\n specificParamsName.append( moreLayerNameParameter[nameMainPart])\n paramsPerLayerType[nameMainPart] = {}\n for key, value in commonParams.items():\n if value.isParameterGroup() and value.parameterName() in specificParamsName:\n paramsPerLayerType[nameMainPart][key] = value\n layerSpecificParameters.add(key)\n\n\n # special case: shared params for loss layers\n key = \"loss_param\"\n value = commonParams[key]\n del commonParams[key]\n for nameMainPart in layerNameMainParts:\n if LayerType.getCategoryByName(nameMainPart) == LayerType.CATEGORY_LOSS:\n paramsPerLayerType[nameMainPart][key] = value\n\n # TODO is there a special case for the TransformationParameter?\n\n # create each layer type after one another\n for nameMainPart in layerNameMainParts:\n\n # add common params to the specific ones\n layerTypeParam = paramsPerLayerType[nameMainPart].keys()\n paramsPerLayerType[nameMainPart].update(commonParams)\n\n irrelevant = layerSpecificParameters.difference(layerTypeParam)\n res[nameMainPart] = LayerType(nameMainPart, paramsPerLayerType[nameMainPart], layerTypeParam, irrelevant)\n\n self._commonParams = commonParams\n self._availableLayerTypes = res","def EmbeddingLayers_pooling(pretrained=False, progress=True, **kwargs):\n kwargs['groups'] = 32\n kwargs['width_per_group'] = 4\n # return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,\n # **kwargs) \n return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],\n pretrained, progress, **kwargs)","def __init__(self,layer_def,input,input_shape,rs,clone_from=None):\n batch_size = int(layer_def.find(\"batchsize\").text)\n image_size = int(layer_def.find(\"imagesize\").text)\n image_channels = int(layer_def.find(\"imagechannels\").text)\n self.layer_name = layer_def.attrib[\"name\"]\n self.init(input, input_shape, batch_size,image_size,image_channels)","def feed(self, *args):\n\t\tprint(\"args: \", args)\n\t\tassert len(args) != 0\n\t\tself.terminals = []\n\t\tfor fed_layer in args:#\n\t\t\ttry:\n\t\t\t\tis_str = isinstance(fed_layer, basestring)\n\t\t\texcept NameError:\n\t\t\t\tis_str = isinstance(fed_layer, str)\n\t\t\tif is_str:\n\t\t\t\ttry:\n\t\t\t\t\tfed_layer = self.layers[fed_layer]\n\t\t\t\t\tprint(fed_layer)\n\t\t\t\texcept KeyError:\n\t\t\t\t\traise KeyError('Unknown layer name fed: %s' % fed_layer)\n\t\t\tself.terminals.append(fed_layer)\n\t\treturn self","def append_layer(self, *args, **kwargs) :\n \n self.insert_layer(len(self._layers), *args, **kwargs)","def get_data_from_layers(vk4_container, layers, step, is_image=False):\n log.debug(\"Entering get_data_from_layers()\")\n holder = []\n for x in range(0, len(layers), step):\n lay = layers[x:x + step]\n log.debug(\"In get_data_from_layers()\\n\\tCurrent layer to append: %s\" % lay)\n if lay not in layer_dict.keys():\n log.error(\"In get_data_from_layers()\\n\\tLayer {}: {} from param \" \\\n \"layers: {} is not a valid layer. See documentation.\"\n .format(x, lay, layers))\n holder.append(vk4_container.get_single_color_values(layer_dict[lay][0], layer_dict[lay][1]))\n\n if is_image:\n log.debug(\"Exiting get_data_from_layers() where is_image is {}\".format(is_image))\n return create_array_from_rgb_layers(vk4_container, holder)\n else:\n log.debug(\"Exiting get_data_from_layers() where is_image is {}\".format(is_image))\n return create_composite_rgb_values(vk4_container, holder)\n # return create_separate_rgb_values(vk4_container, holder)","def initialize_layers(self, layers_config: dict, inputs=None):\n layers_config = layers_config.copy()\n input_lyrs = []\n initiated_layers = OrderedDict()\n wrp_layer = None # indicator for wrapper layers\n first_layer = True\n\n for lyr, lyr_args in layers_config.items():\n\n lyr_config, lyr_inputs, named_outs, call_args = self.deconstruct_lyr_args(lyr, lyr_args)\n\n lyr_name, args, lyr_config, activation = self.check_lyr_config(lyr, lyr_config)\n\n if K.BACKEND == 'pytorch':\n\n if first_layer:\n first_layer = False\n\n if callable(lyr_config):\n lyr_initiated = lyr_config\n else:\n lyr_initiated = TORCH_LAYERS[lyr_name](**lyr_config)\n setattr(self, lyr, lyr_initiated)\n initiated_layers[lyr] = {\"layer\": lyr_initiated, \"named_outs\": named_outs, 'call_args': call_args,\n 'inputs': lyr_inputs}\n\n else:\n # may be user has defined layers without input layer, in this case add Input layer as first layer\n if first_layer:\n if inputs is not None: # This method was called by providing it inputs.\n assert isinstance(inputs, tf.Tensor)\n # since inputs have been defined, all the layers that will be added will be next to first layer\n first_layer = False\n layer_outputs = inputs\n initiated_layers[layer_outputs.name] = {'layer': layer_outputs, 'tf_name': lyr_name}\n\n elif lyr_name != \"Input\":\n if 'input_shape' in lyr_config: # input_shape is given in the first layer so make input layer\n initialized_layer = LAYERS[\"Input\"](shape=lyr_config['input_shape'])\n else:\n # for simple dense layer based models, lookback will not be used\n def_shape = (self.num_ins,) if self.lookback == 1 else (self.lookback, self.num_ins)\n initialized_layer = LAYERS[\"Input\"](shape=def_shape)\n\n # first layer is built so next iterations will not be for first layer\n first_layer = False\n # put the first layer in memory to be used for model compilation\n # add th layer which the user had specified as first layer\n initiated_layers[initialized_layer.name] = {'layer': initialized_layer,\n 'tf_name': lyr_name}\n input_lyrs.append(initialized_layer)\n\n # The inputs to the layer have not been specified, so either it is an Input layer\n if lyr_inputs is None:\n # or it uses the previous outputs as inputs\n if lyr_name == \"Input\":\n # it is an Input layer, hence should not be called\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'tf_name': lyr_name}\n input_lyrs.append(initialized_layer)\n else:\n # it is executable and uses previous outputs as inputs\n if lyr_name in ACTIVATION_LAYERS:\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\n wrp_layer = LAYERS[lyr_name]\n # because wrapper layer name is property\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\n 'tf_name': lyr_name}\n continue\n elif \"LAMBDA\" in lyr_name.upper():\n # lyr_config is serialized lambda layer, which needs to be deserialized\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\n # layers_config['lambda']['config'] still contails lambda, so we need to replace the python\n # object (lambda) with the serialized version (lyr_config) so that it can be saved as json file.\n layers_config[lyr]['config'] = lyr_config\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n else:\n if wrp_layer is not None:\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n wrp_layer = None\n else:\n if lyr_name == \"TemporalFusionTransformer\":\n lyr_config['return_attention_components'] = True\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n\n else: # The inputs to this layer have been specified so they must exist in lyr_cache.\n # it is an executable\n if lyr_name in ACTIVATION_LAYERS:\n\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\n wrp_layer = LAYERS[lyr_name]\n # because wrapper layer name is property\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\n 'tf_name': lyr_name}\n continue\n elif \"LAMBDA\" in lyr_name.upper():\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\n layers_config[lyr]['config'] = lyr_config\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n else:\n if wrp_layer is not None:\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n wrp_layer = None\n else:\n layer_initialized = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': layer_initialized,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n\n if activation is not None: # put the string back to dictionary to be saved in config file\n lyr_config['activation'] = activation\n\n first_layer = False\n\n self.jsonize_lyr_config(lyr_config)\n\n # inputs = [] todo, indentify input layers\n # for k,v in lyr_cache.items():\n # since the model is not build yet and we have access to only output tensors of each list, this is probably\n # # the only way to know that how many `Input` layers were encountered during the run of this method. Each\n # tensor (except TimeDistributed) has .op.inputs attribute, which is empty if a tensor represents output of Input layer.\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\n # if k.upper() != \"TIMEDISTRIBUTED\" and hasattr(v, 'op'):\n # if hasattr(v.op, 'inputs'):\n # _ins = v.op.inputs\n # if len(_ins) == 0:\n # inputs.append(v)\n # else: # not sure if this is the proper way of checking if a layer receives an input or not!\n # if hasattr(v, '_keras_mask'):\n # inputs.append(v)\n\n setattr(self, 'initiated_layers', initiated_layers)\n setattr(self, 'input_lyrs', input_lyrs)\n\n\n # todo,\n # # for case when {Input -> Dense, Input_1}, this method wrongly makes Input_1 as output so in such case use\n # # {Input_1, Input -> Dense }, thus it makes Dense as output and first 2 as inputs, so throwing warning\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\n # if len(layer_outputs.op.inputs) < 1:\n # print(\"Warning: the output is of Input tensor class type\")\n # else:\n # if 'op' not in dir(layer_outputs): # layer_outputs does not have `op`, which means it has no incoming node\n # print(\"Warning: the output is of Input tensor class type\")\n\n # outs = None\n #if BACKEND == 'tensorflow':\n # outs = self.call(input_lyrs)\n # setattr(self, 'output_lyrs', outs)\n # if BACKEND == 'tensorflow':\n # ## Reinitial\n # super(Model, self).__init__(\n # inputs=input_lyrs,\n # outputs=outs)\n #MODEL.__init__(self, inputs=inputs, outputs=outs)\n\n return input_lyrs # , outs","def build(width, height, depth, classes, stages, filters, include_top, pooling,\n reg=1e-3, bnEps=2e-5, bnMom=0.0):\n inputShape = (height, width, depth)\n chanDim = -1\n\n if K.image_data_format() == \"channels_first\":\n inputShape = (depth, height, width)\n chanDim = 1\n\n inputs = Input(shape=inputShape)\n\n\n # block 1 (initial conv block)\n x = ZeroPadding2D(padding=((3, 3), (3, 3)), name='conv1_pad')(inputs)\n x = Conv2D(64, (7,7), use_bias=False, strides=(2,2),\n kernel_initializer=\"he_normal\", kernel_regularizer=l2(reg))(x)\n x = BatchNormalization(axis=chanDim, name=\"bn_conv1\")(x)\n x = Activation(\"relu\")(x)\n x = ZeroPadding2D(padding=((1,1), (1,1)), name=\"pool1_pad\")(x)\n x = MaxPooling2D(3, strides=2)(x)\n\n for i in range(0, len(stages)):\n stride = (1,1) if i == 0 else (2,2) # block 2 (projection block) w stride(1,1)\n\n print(\"Stage {}, Stride={}\".format(i, stride))\n x = SEResNet.residual_module(x, filters[i+1], stride,\n chanDim=chanDim, red=True, bnEps=bnEps, bnMom=bnMom)\n for j in range(0, stages[i] + 1): #stacking res block to each depth layer\n x = SEResNet.residual_module(x, filters[i+1], stride=(1,1),\n chanDim=chanDim, bnEps=bnEps,\n bnMom=bnMom)\n x = BatchNormalization(axis=chanDim, epsilon=bnEps,\n momentum=bnMom)(x)\n x = Activation(\"relu\")(x)\n\n if include_top:\n x = GlobalAveragePooling2D()(x)\n x = Dense(classes, use_bias=False, kernel_regularizer=l2(reg),\n activation='softmax')(x)\n else:\n if pooling == 'avg':\n print(\"Adding average pool\")\n x = GlobalAveragePooling2D()(x)\n elif pooling == 'max':\n x = GlobalMaxPooling2D()(x)\n\n model = Model(inputs=inputs, outputs=x, name=\"SEResNet\")\n return model","def convolutional_block(input_tensor, kernel_size, filters, stage, block, strides = (2, 2)):\n chanDim = 3\n eps = 1.1e-5\n \n if K.image_data_format() == \"channels_first\":\n inputShape = (depth, height, width)\n chanDim = 1\n\n # defining name basis\n conv_name_base = 'res152' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n # Retrieve Filters\n filters1, filters2, filters3 = filters\n \n ##### MAIN PATH #####\n # First component of main path \n X = Conv2D(filters = filters1, kernel_size = (1, 1), strides = strides, name = conv_name_base + '2a')(input_tensor)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2a')(X)\n X = Activation('relu', name = conv_name_base + '2a_relu')(X)\n \n # Second component of main path \n X = ZeroPadding2D(padding = (1, 1), name = conv_name_base + '2b_zeropadding')(X)\n X = Conv2D(filters = filters2, kernel_size = (kernel_size, kernel_size), name = conv_name_base + '2b')(X)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2b')(X)\n X = Activation('relu', name = conv_name_base + \"2b_relu\")(X)\n \n # Third component of main path \n X = Conv2D(filters = filters3, kernel_size = (1, 1), name = conv_name_base + '2c')(X)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2c')(X)\n \n ##### SHORTCUT PATH #### \n X_shortcut = Conv2D(filters = filters3, kernel_size = (1, 1), strides = strides, name = conv_name_base + '1')(input_tensor)\n X_shortcut = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '1')(X_shortcut)\n \n # Final step: Add shortcut value to main path, and pass it through a RELU activation\n X = Add()([X, X_shortcut])\n X = Activation('relu', name = \"res152\" + str(stage) + block + \"_relu\")(X) \n \n return X","def init_layers(self):\n\n # get caching layers activated\n caching_layers = G3WCachingLayer.objects.all()\n for caching_layer in caching_layers:\n self.add_layer(str(caching_layer), caching_layer)","def _init_layers(self):\n cls_branch = []\n for _ in range(self.num_reg_fcs):\n cls_branch.append(Linear(self.embed_dims, self.embed_dims))\n cls_branch.append(nn.LayerNorm(self.embed_dims))\n cls_branch.append(nn.ReLU(inplace=True))\n cls_branch.append(Linear(self.embed_dims, self.cls_out_channels))\n fc_cls = nn.Sequential(*cls_branch)\n\n reg_branch = []\n for _ in range(self.num_reg_fcs):\n reg_branch.append(Linear(self.embed_dims, self.embed_dims))\n reg_branch.append(nn.ReLU())\n reg_branch.append(Linear(self.embed_dims, self.code_size))\n reg_branch = nn.Sequential(*reg_branch)\n\n past_traj_reg_branch = []\n for _ in range(self.num_reg_fcs):\n past_traj_reg_branch.append(\n Linear(self.embed_dims, self.embed_dims))\n past_traj_reg_branch.append(nn.ReLU())\n past_traj_reg_branch.append(\n Linear(self.embed_dims, (self.past_steps + self.fut_steps)*2))\n past_traj_reg_branch = nn.Sequential(*past_traj_reg_branch)\n\n def _get_clones(module, N):\n return nn.ModuleList([copy.deepcopy(module) for i in range(N)])\n\n # last reg_branch is used to generate proposal from\n # encode feature map when as_two_stage is True.\n num_pred = (self.transformer.decoder.num_layers + 1) if \\\n self.as_two_stage else self.transformer.decoder.num_layers\n\n if self.with_box_refine:\n self.cls_branches = _get_clones(fc_cls, num_pred)\n self.reg_branches = _get_clones(reg_branch, num_pred)\n self.past_traj_reg_branches = _get_clones(\n past_traj_reg_branch, num_pred)\n else:\n self.cls_branches = nn.ModuleList(\n [fc_cls for _ in range(num_pred)])\n self.reg_branches = nn.ModuleList(\n [reg_branch for _ in range(num_pred)])\n self.past_traj_reg_branches = nn.ModuleList(\n [past_traj_reg_branch for _ in range(num_pred)])\n if not self.as_two_stage:\n self.bev_embedding = nn.Embedding(\n self.bev_h * self.bev_w, self.embed_dims)","def __call__(\n self, inputs, training, get_all_layers=None, skip_final_dense=False,\n skip_final_dense_with_pool=False, final_pooling=None):\n global ENDING_POINTS\n ENDING_POINTS = []\n\n with self._model_variable_scope():\n inputs = self._preprocess_data(inputs)\n\n inputs = conv3d_fixed_padding(\n inputs=inputs, filters=self.num_filters, \n kernel_size=self.kernel_size, time_kernel_size=self.time_kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = tf.identity(inputs, 'initial_conv')\n\n # We do not include batch normalization or activation functions in V2\n # for the initial conv1 because the first ResNet unit will perform these\n # for both the shortcut and non-shortcut paths as part of the first\n # block's projection. Cf. Appendix of [2].\n if self.resnet_version == 1:\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n\n _fp_size = self.first_pool_size\n _fp_stride = self.first_pool_stride\n if self.first_pool_size:\n inputs = tf.layers.max_pooling3d(\n inputs=inputs, \n pool_size=[_fp_size, _fp_size, _fp_size],\n strides=[_fp_stride, _fp_stride, _fp_stride], \n padding='SAME',\n data_format=self.data_format)\n inputs = tf.identity(inputs, 'initial_max_pool')\n\n ENDING_POINTS.append(inputs)\n\n for i, num_blocks in enumerate(self.block_sizes):\n num_filters = self.num_filters * (2**i)\n inputs = block_layer(\n inputs=inputs, filters=num_filters, bottleneck=self.bottleneck,\n block_fn=self.block_fn, blocks=num_blocks,\n strides=self.block_strides[i], training=training,\n name='block_layer{}'.format(i + 1), data_format=self.data_format)\n curr_res_name = 'res{}'.format(i + 1)\n\n # Only apply the BN and ReLU for model that does pre_activation in each\n # building/bottleneck block, eg resnet V2.\n if self.pre_activation:\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n ENDING_POINTS.append(inputs)\n\n if skip_final_dense:\n bs = inputs.get_shape().as_list()[0]\n # Along the temporal dimension\n inputs = tf.reduce_mean(inputs, axis=2)\n if final_pooling is None:\n inputs = tf.reshape(inputs, [bs, -1])\n return inputs\n if final_pooling == 1:\n axes = [2, 3] if self.data_format == 'channels_first' else [1, 2]\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\n return tf.reshape(inputs, [bs, 1 * 1 * self.final_size])\n \n # The current top layer has shape\n # `batch_size x pool_size x pool_size x final_size`.\n # ResNet does an Average Pooling layer over pool_size,\n # but that is the same as doing a reduce_mean. We do a reduce_mean\n # here because it performs better than AveragePooling2D.\n axes = [2, 3, 4] if self.data_format == 'channels_first' else [1, 2, 3]\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\n inputs = tf.identity(inputs, 'final_reduce_mean')\n\n if skip_final_dense_with_pool:\n return inputs\n\n inputs = self._get_final_dense(inputs)\n if not get_all_layers:\n return inputs\n else:\n return inputs, ENDING_POINTS","def make_feature_layers(self, config):\n raise NotImplementedError","def add_layer_data(prj_id, expl_id, exploration):\n layers = tbl_explsoilboring[\n (tbl_explsoilboring.lng_KeyProject == prj_id) &\n (tbl_explsoilboring.txt_KeyExplorationName == expl_id)]\n\n # add new column with layer numbers\n layers['layer'] = [i for i in range(1, len(layers.index) + 1)]\n\n # get SPT data\n spt_data = tbl_explboring[\n (tbl_explboring.lng_KeyProject == prj_id) &\n (tbl_explboring.txt_KeyExplorationName == expl_id)]\n\n if not spt_data.empty and not layers.empty:\n spt_depths = spt_data.dbl_KeyDepth\n layer_i = []\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\n for d in spt_depths:\n if d <= max_d:\n layr = layers[\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\n else:\n layr = None\n layer_i.append(layr)\n spt_data['layer'] = layer_i\n avg_spt = spt_data[\n ['layer', 'dbl_FieldBlowCount']].groupby('layer').agg('mean')\n else:\n avg_spt = pd.DataFrame()\n\n # get lab data\n lab_data = tbl_expllab[\n (tbl_expllab.lng_KeyProject == prj_id) &\n (tbl_expllab.txt_KeyExplorationName == expl_id)]\n\n if not lab_data.empty and not layers.empty:\n lab_depths = lab_data.dbl_KeyDepth\n layer_i = []\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\n for d in lab_depths:\n if d <= max_d:\n layr = layers[\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\n else:\n layr = None\n layer_i.append(layr)\n lab_data['layer'] = layer_i\n avg_lab = lab_data[\n ['layer', 'dbl_MoistureContent', 'dbl_TotalUnitWeight',\n 'dbl_LiquidLimit', 'dbl_PlasticityIndex', 'dbl_Cohesion']\n ].groupby('layer').agg('mean')\n else:\n avg_lab = pd.DataFrame()\n\n for i in layers.index:\n index = int(layers.layer[i])\n if index == 1:\n height = tbl_explsoilboring['dbl_KeyDepthToBottom'][i] * 0.00328084\n else:\n height = (tbl_explsoilboring['dbl_KeyDepthToBottom'][i]\n - tbl_explsoilboring['dbl_KeyDepthToBottom'][i-1]) \\\n * 0.00328084\n soil_type = tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i] \\\n if not pd.isna(\n tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i]) else None\n soil_type = primary_explsoil[soil_type] if soil_type else None\n uscs_type = tbl_explsoilboring['txt_USCSCode'][i] if not \\\n pd.isna(tbl_explsoilboring['txt_USCSCode'][i]) else None\n if not soil_type and uscs_type:\n soil_type = uscs_type\n description = tbl_explsoilboring['txt_LayerDescription'][i] \\\n if not pd.isna(tbl_explsoilboring['txt_LayerDescription'][i]) \\\n else None\n nval = tbl_explsoilboring['dbl_BlowCountInterpreted'][i] \\\n if not pd.isna(tbl_explsoilboring['dbl_BlowCountInterpreted'][i]) \\\n else None\n nval_from_avg_spt = avg_spt[avg_spt.index == index]\n if not nval and not nval_from_avg_spt.empty:\n nval = nval_from_avg_spt.values[0][0] if not pd.isna(\n nval_from_avg_spt.values[0][0]) else None\n angle = tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i] \\\n if not pd.isna(\n tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i]) else None\n data_from_avg_lab = avg_lab[avg_lab.index == index]\n cohesion = tbl_explsoilboring['dbl_CohesionInterpreted'][i] \\\n * 2.0885434273e-05 if not pd.isna(\n tbl_explsoilboring['dbl_CohesionInterpreted'][i]) else None\n if not cohesion and not data_from_avg_lab.empty:\n cohesion = data_from_avg_lab.values[0][4] * 2.0885434273e-05 \\\n if not pd.isna(data_from_avg_lab.values[0][4]) else None\n if cohesion:\n cohesion = cohesion if cohesion < 99 else None\n tuw = tbl_explsoilboring['dbl_UnitWeightInterpreted'][i] * 0.062428 \\\n if not pd.isna(tbl_explsoilboring['dbl_UnitWeightInterpreted'][i]) \\\n else None\n if not tuw and not data_from_avg_lab.empty:\n tuw = data_from_avg_lab.values[0][1] * 0.062428 if not pd.isna(\n data_from_avg_lab.values[0][1]) else None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][0]):\n water_content = int(data_from_avg_lab.values[0][0])\n else:\n water_content = None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][2]):\n liquid_limit = int(data_from_avg_lab.values[0][2])\n else:\n liquid_limit = None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][3]):\n plasticity = int(data_from_avg_lab.values[0][3])\n else:\n plasticity = None\n layer = Layers(\n boring=exploration,\n index=index,\n soil_type=soil_type,\n height=height,\n description=description,\n nval=nval,\n ssuu=cohesion,\n friction_angle=angle,\n tuw=tuw,\n water_content=water_content,\n liquid_limit=liquid_limit,\n plasticity=plasticity\n )\n db.session.add(layer)","def layers(self):\n return self['layers']","def expandWithoutMutex(self, previousLayer):\n previousLayerProposition = previousLayer.getPropositionLayer()\n \"*** YOUR CODE HERE ***\"","def __init__(self, layers) -> None:\n super(FeaturizerPhi, self).__init__()\n self.layers = layers","def filesystem_layers(self):\n pass","def renderLayerPostProcess(*args, keepImages: bool=True, sceneName: Union[AnyStr, bool]=\"\",\n q=True, query=True, **kwargs)->Union[None, Any]:\n pass","def update(self):\r\n\r\n # Update the vision frames in the system\r\n self._system.update()\r\n\r\n # Create blank PIL images to hold the video streams\r\n layered = PIL.Image.new('RGBA', (400, 400))\r\n stacked = PIL.Image.new('RGBA', (200, 800))\r\n control = PIL.Image.new('RGBA', (600, 800))\r\n\r\n focalpoint = self._system[self._appString[\"device\"].get()].focalpoint()\r\n # print(focalpoint)\r\n\r\n # Get each vision key and vision for the selected device\r\n visionList = [(visionKey, vision) for visionKey, vision in self._system[self._appString[\"device\"].get()]]\r\n\r\n # Loop through each vision in the vision list\r\n for i, (visionKey, vision) in enumerate(visionList):\r\n\r\n # Grab the frames from the vision when it is \"curr\"\r\n frameList = [frame for frameKey, frame in vision if frameKey==self._appString[\"frame\"].get()]\r\n\r\n # Loop through each frame in the frame list\r\n for frame in frameList:\r\n\r\n # Get the properties and turn the image into RGBA\r\n ratio, size = vision.properties()\r\n rgbFrame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGBA)\r\n\r\n # print(rgbFrame.shape)\r\n width, height, channels = rgbFrame.shape\r\n\r\n # Paste the images together in layered\r\n\r\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (int(400 * ratio), int(400 * ratio))))\r\n layered.paste(imgFrame, (int(200 * (1 - ratio)), int(200 * (1 - ratio))))\r\n\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 // width)), int(200 * (1 - ratio) - focalpoint[1] * (200 // height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1)), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200/width) / ratio), int(200 * (1 - ratio) - focalpoint[1] * (200/height) / ratio)))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1) / 200), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1) / 200)))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (400//width * (1- ratio))), int(200 * (1 - ratio) - focalpoint[1] * (400//height * (1 - ratio)))))\r\n\r\n # Paste the images together in stacked\r\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (200, 200)))\r\n stacked.paste(imgFrame, (0, 200 * i))\r\n\r\n # Add the stacked image to the canvas\r\n self._pilFrames[\"stacked\"] = PIL.ImageTk.PhotoImage(image=stacked)\r\n self._appCanvas[\"stacked\"].create_image(100, 0, image=self._pilFrames[\"stacked\"], anchor=tkinter.NW)\r\n\r\n # Add the layered image to the canvas\r\n self._pilFrames[\"layered\"] = PIL.ImageTk.PhotoImage(image=layered)\r\n self._appCanvas[\"layered\"].create_image(0, 0, image=self._pilFrames[\"layered\"], anchor=tkinter.NW)\r\n\r\n # Add the control image to the canvas\r\n imgFrame = cv2.cvtColor(self._system[self._appString[\"device\"].get()][self._appString[\"vision\"].get()][self._appString[\"frame\"].get()], cv2.COLOR_BGR2RGBA)\r\n control = PIL.Image.fromarray(cv2.resize(imgFrame, (600, 600)))\r\n self._pilFrames[\"control\"] = PIL.ImageTk.PhotoImage(image=control)\r\n self._appCanvas[\"control\"].create_image(100, 90, image=self._pilFrames[\"control\"], anchor=tkinter.NW)\r\n\r\n # Continue to update with a delay of 15\r\n self.after(15, self.update)","def _get_layers(self) :\n \n return self._layers","def layers(self):\r\n return self._flc.layers","def initLayers(self):\n self.addRow(self.currentLayerData)\n self.selected = self.jobRow.child(0)\n self.setSelectedFromItem(self.selected)","def num_layers(self): # -> int:\n ...","def get_patches(rimage, gimage, mimage, num_patches=48, patch_size=80, patch_stride=80):\n num_FSpatches = 16\n num_RApatches = 32\n rpatches = []\n gpatches = []\n mpatches = []\n #R_imgs = ((rimage+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'rainy.jpg', R_imgs[0,:,:,:])\n for i in range(int(math.sqrt(num_FSpatches))):\n for j in range(int(math.sqrt(num_FSpatches))):\n point_x = patch_stride*i\n point_y = patch_stride*j\n rpatch = rimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n #print(point_x)\n #print(point_y)\n #print(point_y+patch_size)\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'patch_%d_%d.jpg'%(i,j), P_imgs[0,:,:,:])\n #print(rpatch.shape)\n rpatches.append(rpatch)\n #print(np.array(rpatches).shape)\n gpatch = gimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n gpatches.append(gpatch)\n mpatch = mimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n mpatches.append(mpatch)\n\n for k in range(num_RApatches):\n point1 = random.randint(0,240) # 116 comes from the image source size (320) - the patch dimension (80)\n point2 = random.randint(0,240)\n #rpatch = tf.image.crop_to_bounding_box(rimage, point1, point2, patch_size, patch_size)\n rpatch = rimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'patch_%d.jpg'%i, P_imgs[0,:,:,:])\n #print(rpatch.shape)\n rpatches.append(rpatch)\n #print(np.array(rpatches).shape)\n gpatch = gimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n gpatches.append(gpatch)\n mpatch = mimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n mpatches.append(mpatch)\n\n rpatches = np.array(rpatches)\n rpatches = np.squeeze(rpatches)\n #print(rpatches.shape)\n gpatches = np.array(gpatches)\n gpatches = np.squeeze(gpatches)\n mpatches = np.array(mpatches)\n mpatches = np.squeeze(mpatches)\n #assert rpatches.get_shape().dims == [num_patches, patch_size, patch_size, 3]\n assert rpatches.shape == (num_patches, patch_size, patch_size, 3)\n return rpatches, gpatches, mpatches","def _init_layers(self):\n self._init_predictor()\n if self.use_edge_fusion:\n self._init_edge_module()","def lightened_v2(inputs, is_training=True,\n dropout_keep_prob=0.8,\n reuse=None,\n scope='LightenedV1'):\n end_points = {}\n \n with tf.variable_scope(scope, 'LightenedV1', [inputs], reuse=reuse):\n with slim.arg_scope([slim.batch_norm, slim.dropout],\n is_training=is_training):\n with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],\n stride=1, padding='SAME'):\n \n endpoints = {}\n \n net=conv(inputs, 48, 9, stride=1, padding='VALID', scope='conv1_9x9')\n end_points['conv1_9x9'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool1')\n end_points['pool1'] = net\n \n net=conv(net, 96, 5, stride=1, padding='VALID', scope='conv2_5x5')\n end_points['conv2_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool2')\n end_points['pool2'] = net\n\n net=conv(net, 128, 5, stride=1, padding='VALID', scope='conv3_5x5')\n end_points['conv3_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool3')\n end_points['pool3'] = net\n\n net=conv(net, 192, 4, stride=1, padding='VALID', scope='conv4_4x4')\n end_points['conv4_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool4')\n end_points['pool4'] = net\n\n with tf.variable_scope('Logits'):\n net = slim.flatten(net)\n net = slim.dropout(net, dropout_keep_prob, is_training=is_training,\n scope='Dropout') \n endpoints['PreLogitsFlatten'] = net\n \n return net, end_points","def layerListObject(self, parent):\n pass","def __init__(self, batch_env):\n\n dims = [84, 84]\n nature_transform = lambda o: tf.image.rgb_to_grayscale( # pylint: disable=g-long-lambda\n tf.image.resize_images(o, dims))\n\n super(WarpFrameWrapper, self).__init__(batch_env, transform_observation=(\n nature_transform, dims, tf.float32))","def update_layer(LayerId=None, Name=None, Shortname=None, Attributes=None, CloudWatchLogsConfiguration=None, CustomInstanceProfileArn=None, CustomJson=None, CustomSecurityGroupIds=None, Packages=None, VolumeConfigurations=None, EnableAutoHealing=None, AutoAssignElasticIps=None, AutoAssignPublicIps=None, CustomRecipes=None, InstallUpdatesOnBoot=None, UseEbsOptimizedInstances=None, LifecycleEventConfiguration=None):\n pass","def addExportLayerToCoreml(builder):\n outputNames = [output.name for output in builder.spec.description.output]\n\n for i, outputName in enumerate(outputNames):\n # formulas: https://github.com/ultralytics/yolov5/issues/471\n builder.add_activation(\n name=f\"sigmoid_{outputName}\",\n non_linearity=\"SIGMOID\",\n input_name=outputName,\n output_name=f\"{outputName}_sigmoid\",\n )\n\n ### Coordinates calculation ###\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2) -> nC = 640 / strides[i]\n builder.add_slice(\n name=f\"slice_coordinates_xy_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_sliced_coordinates_xy\",\n axis=\"width\",\n start_index=0,\n end_index=2,\n )\n # x,y * 2\n builder.add_elementwise(\n name=f\"multiply_xy_by_two_{outputName}\",\n input_names=[f\"{outputName}_sliced_coordinates_xy\"],\n output_name=f\"{outputName}_multiplied_xy_by_two\",\n mode=\"MULTIPLY\",\n alpha=2,\n )\n # x,y * 2 - 0.5\n builder.add_elementwise(\n name=f\"subtract_0_5_from_xy_{outputName}\",\n input_names=[f\"{outputName}_multiplied_xy_by_two\"],\n output_name=f\"{outputName}_subtracted_0_5_from_xy\",\n mode=\"ADD\",\n alpha=-0.5,\n )\n grid = make_grid(featureMapDimensions[i], featureMapDimensions[i]).numpy()\n # x,y * 2 - 0.5 + grid[i]\n builder.add_bias(\n name=f\"add_grid_from_xy_{outputName}\",\n input_name=f\"{outputName}_subtracted_0_5_from_xy\",\n output_name=f\"{outputName}_added_grid_xy\",\n b=grid,\n shape_bias=grid.shape,\n )\n # (x,y * 2 - 0.5 + grid[i]) * stride[i]\n builder.add_elementwise(\n name=f\"multiply_xy_by_stride_{outputName}\",\n input_names=[f\"{outputName}_added_grid_xy\"],\n output_name=f\"{outputName}_calculated_xy\",\n mode=\"MULTIPLY\",\n alpha=strides[i],\n )\n\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2)\n builder.add_slice(\n name=f\"slice_coordinates_wh_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_sliced_coordinates_wh\",\n axis=\"width\",\n start_index=2,\n end_index=4,\n )\n # w,h * 2\n builder.add_elementwise(\n name=f\"multiply_wh_by_two_{outputName}\",\n input_names=[f\"{outputName}_sliced_coordinates_wh\"],\n output_name=f\"{outputName}_multiplied_wh_by_two\",\n mode=\"MULTIPLY\",\n alpha=2,\n )\n # (w,h * 2) ** 2\n builder.add_unary(\n name=f\"power_wh_{outputName}\",\n input_name=f\"{outputName}_multiplied_wh_by_two\",\n output_name=f\"{outputName}_power_wh\",\n mode=\"power\",\n alpha=2,\n )\n # (w,h * 2) ** 2 * anchor_grid[i]\n anchor = (\n anchorGrid[i]\n .expand(-1, featureMapDimensions[i], featureMapDimensions[i], -1)\n .numpy()\n )\n builder.add_load_constant_nd(\n name=f\"anchors_{outputName}\",\n output_name=f\"{outputName}_anchors\",\n constant_value=anchor,\n shape=anchor.shape,\n )\n builder.add_elementwise(\n name=f\"multiply_wh_with_achors_{outputName}\",\n input_names=[f\"{outputName}_power_wh\", f\"{outputName}_anchors\"],\n output_name=f\"{outputName}_calculated_wh\",\n mode=\"MULTIPLY\",\n )\n\n builder.add_concat_nd(\n name=f\"concat_coordinates_{outputName}\",\n input_names=[f\"{outputName}_calculated_xy\", f\"{outputName}_calculated_wh\"],\n output_name=f\"{outputName}_raw_coordinates\",\n axis=-1,\n )\n builder.add_scale(\n name=f\"normalize_coordinates_{outputName}\",\n input_name=f\"{outputName}_raw_coordinates\",\n output_name=f\"{outputName}_raw_normalized_coordinates\",\n W=torch.tensor([1 / 640]).numpy(),\n b=0,\n has_bias=False,\n )\n\n ### Confidence calculation ###\n builder.add_slice(\n name=f\"slice_object_confidence_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_object_confidence\",\n axis=\"width\",\n start_index=4,\n end_index=5,\n )\n builder.add_slice(\n name=f\"slice_label_confidence_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_label_confidence\",\n axis=\"width\",\n start_index=5,\n end_index=0,\n )\n # confidence = object_confidence * label_confidence\n builder.add_multiply_broadcastable(\n name=f\"multiply_object_label_confidence_{outputName}\",\n input_names=[\n f\"{outputName}_label_confidence\",\n f\"{outputName}_object_confidence\",\n ],\n output_name=f\"{outputName}_raw_confidence\",\n )\n\n # input: (1, 3, nC, nC, 85), output: (3 * nc^2, 85)\n builder.add_flatten_to_2d(\n name=f\"flatten_confidence_{outputName}\",\n input_name=f\"{outputName}_raw_confidence\",\n output_name=f\"{outputName}_flatten_raw_confidence\",\n axis=-1,\n )\n builder.add_flatten_to_2d(\n name=f\"flatten_coordinates_{outputName}\",\n input_name=f\"{outputName}_raw_normalized_coordinates\",\n output_name=f\"{outputName}_flatten_raw_coordinates\",\n axis=-1,\n )\n\n builder.add_concat_nd(\n name=\"concat_confidence\",\n input_names=[\n f\"{outputName}_flatten_raw_confidence\" for outputName in outputNames\n ],\n output_name=\"raw_confidence\",\n axis=-2,\n )\n builder.add_concat_nd(\n name=\"concat_coordinates\",\n input_names=[\n f\"{outputName}_flatten_raw_coordinates\" for outputName in outputNames\n ],\n output_name=\"raw_coordinates\",\n axis=-2,\n )\n\n builder.set_output(\n output_names=[\"raw_confidence\", \"raw_coordinates\"],\n output_dims=[(25200, numberOfClassLabels), (25200, 4)],\n )","def stage(self, stage):\n self._stage = stage\n self._layer = Sdf.Layer.CreateAnonymous()\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)","def __init__(self,\n decoder_options,\n panoptic_deeplab_options,\n bn_layer=tf.keras.layers.BatchNormalization):\n super(PanopticDeepLab, self).__init__(name='PanopticDeepLab')\n\n low_level_feature_keys = [\n item.feature_key for item in panoptic_deeplab_options.low_level\n ]\n low_level_channels_project = [\n item.channels_project for item in panoptic_deeplab_options.low_level\n ]\n\n self._semantic_decoder = PanopticDeepLabSingleDecoder(\n high_level_feature_name=decoder_options.feature_key,\n low_level_feature_names=low_level_feature_keys,\n low_level_channels_project=low_level_channels_project,\n aspp_output_channels=decoder_options.aspp_channels,\n decoder_output_channels=decoder_options.decoder_channels,\n atrous_rates=decoder_options.atrous_rates,\n name='semantic_decoder',\n aspp_use_only_1x1_proj_conv=decoder_options.aspp_use_only_1x1_proj_conv,\n decoder_conv_type=decoder_options.decoder_conv_type,\n bn_layer=bn_layer)\n self._semantic_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.semantic_head.head_channels,\n panoptic_deeplab_options.semantic_head.output_channels,\n common.PRED_SEMANTIC_LOGITS_KEY,\n name='semantic_head',\n conv_type=panoptic_deeplab_options.semantic_head.head_conv_type,\n bn_layer=bn_layer)\n\n self._instance_decoder = None\n self._instance_center_head = None\n self._instance_regression_head = None\n\n if panoptic_deeplab_options.instance.enable:\n if panoptic_deeplab_options.instance.low_level_override:\n low_level_options = panoptic_deeplab_options.instance.low_level_override\n else:\n low_level_options = panoptic_deeplab_options.low_level\n\n # If instance_decoder is set, use those options; otherwise reuse the\n # architecture as defined for the semantic decoder.\n if panoptic_deeplab_options.instance.HasField(\n 'instance_decoder_override'):\n decoder_options = (panoptic_deeplab_options.instance\n .instance_decoder_override)\n\n low_level_feature_keys = [item.feature_key for item in low_level_options]\n low_level_channels_project = [\n item.channels_project for item in low_level_options\n ]\n\n self._instance_decoder = PanopticDeepLabSingleDecoder(\n high_level_feature_name=decoder_options.feature_key,\n low_level_feature_names=low_level_feature_keys,\n low_level_channels_project=low_level_channels_project,\n aspp_output_channels=decoder_options.aspp_channels,\n decoder_output_channels=decoder_options.decoder_channels,\n atrous_rates=decoder_options.atrous_rates,\n name='instance_decoder',\n aspp_use_only_1x1_proj_conv=(\n decoder_options.aspp_use_only_1x1_proj_conv),\n decoder_conv_type=decoder_options.decoder_conv_type,\n bn_layer=bn_layer)\n self._instance_center_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.instance.center_head.head_channels,\n panoptic_deeplab_options.instance.center_head.output_channels,\n common.PRED_CENTER_HEATMAP_KEY,\n name='instance_center_head',\n conv_type=(\n panoptic_deeplab_options.instance.center_head.head_conv_type),\n bn_layer=bn_layer)\n self._instance_regression_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.instance.regression_head.head_channels,\n panoptic_deeplab_options.instance.regression_head.output_channels,\n common.PRED_OFFSET_MAP_KEY,\n name='instance_regression_head',\n conv_type=(\n panoptic_deeplab_options.instance.regression_head.head_conv_type),\n bn_layer=bn_layer)","def __init__(self,\n high_level_feature_name,\n low_level_feature_names,\n low_level_channels_project,\n aspp_output_channels,\n decoder_output_channels,\n atrous_rates,\n name,\n aspp_use_only_1x1_proj_conv=False,\n decoder_conv_type='depthwise_separable_conv',\n bn_layer=tf.keras.layers.BatchNormalization):\n super(PanopticDeepLabSingleDecoder, self).__init__(name=name)\n self._channel_axis = 3\n\n self._aspp = aspp.ASPP(\n aspp_output_channels,\n atrous_rates,\n aspp_use_only_1x1_proj_conv=aspp_use_only_1x1_proj_conv,\n name='aspp',\n bn_layer=bn_layer)\n self._high_level_feature_name = high_level_feature_name\n\n if len(low_level_feature_names) != len(low_level_channels_project):\n raise ValueError('The Panoptic-DeepLab decoder requires the same number '\n 'of low-level features as the number of low-level '\n 'projection channels. But got %d and %d.'\n % (len(low_level_feature_names),\n len(low_level_channels_project)))\n\n self._low_level_feature_names = low_level_feature_names\n\n for i, channels_project in enumerate(low_level_channels_project):\n # Check if channel sizes increases and issue a warning.\n if i > 0 and low_level_channels_project[i - 1] < channels_project:\n logging.warning(\n 'The low level projection channels usually do not '\n 'increase for features with higher spatial resolution. '\n 'Please make sure, this behavior is intended.')\n current_low_level_conv_name, current_fusion_conv_name = (\n utils.get_low_level_conv_fusion_conv_current_names(i))\n utils.safe_setattr(\n self, current_low_level_conv_name, convolutions.Conv2DSame(\n channels_project,\n kernel_size=1,\n name=utils.get_layer_name(current_low_level_conv_name),\n use_bias=False,\n use_bn=True,\n bn_layer=bn_layer,\n activation='relu'))\n\n utils.safe_setattr(\n self, current_fusion_conv_name, convolutions.StackedConv2DSame(\n conv_type=decoder_conv_type,\n num_layers=1,\n output_channels=decoder_output_channels,\n kernel_size=5,\n name=utils.get_layer_name(current_fusion_conv_name),\n use_bias=False,\n use_bn=True,\n bn_layer=bn_layer,\n activation='relu'))","def develop_image(pipeline, camera=None, batch=None, image=None, patch_size=0, patches=2, root_dir='./data', pipeline_args=None):\n\n if camera is not None:\n supported_cameras = fsutil.listdir(os.path.join(root_dir, 'models', 'nip'), '.*')\n if camera not in supported_cameras:\n raise ValueError('Camera data not found ({})! Available cameras: {}'.format(camera, ', '.join(supported_cameras)))\n root_dirname = os.path.join(root_dir, 'models', 'nip', camera)\n data_dirname = os.path.join(root_dir, 'raw', 'training_data', camera)\n\n if patch_size != 0 and (patch_size < 4 or patch_size > 2048):\n raise ValueError('Patch size seems to be invalid!')\n\n # Lazy imports to minimize delay for invalid command line parameters\n import numpy as np\n import imageio as io\n import matplotlib.pyplot as plt\n import tensorflow as tf\n from models import pipelines\n\n # Construct the NIP model ---------------------------------------------------------------------\n \n if os.path.isdir(pipeline):\n # Restore a NIP model from a training log\n model = tfmodel.restore(pipeline, pipelines)\n else:\n # Construct the NIP model from class name (and optional arguments)\n if pipeline_args is None:\n model = getattr(pipelines, pipeline)()\n else:\n model = getattr(pipelines, pipeline)(**pipeline_args)\n\n loaded_model = False\n candidate_dirs = [os.path.join(root_dirname, model.model_code), os.path.join(root_dirname)]\n for candidate in candidate_dirs:\n if os.path.isdir(candidate):\n model.load_model(candidate)\n loaded_model = True\n break\n\n if not loaded_model:\n raise FileNotFoundError(f'Could not find the corresponding model: {candidate_dirs}')\n\n # Load image(s) -------------------------------------------------------------------------------\n \n if image is None and batch is not None:\n print('Loading a batch of {} images'.format(batch))\n data = dataset.Dataset(data_dirname, n_images=0, v_images=batch, val_rgb_patch_size=patch_size or 256, val_n_patches=patches)\n sample_x, sample_y = data.next_validation_batch(0, data.count_validation)\n\n with open('config/cameras.json') as f:\n cameras = json.load(f)\n cfa, srgb = cameras[camera]['cfa'], np.array(cameras[camera]['srgb'])\n\n elif image is not None:\n print('Loading a RAW image {}'.format(image))\n sample_x, cfa, srgb, _ = raw.unpack(image, expand=True)\n sample_y = raw.process(image, brightness=None, expand=True)\n\n if isinstance(model, pipelines.ClassicISP):\n print('Configuring ISP to CFA: {} & sRGB {}'.format(cfa, srgb.round(2).tolist()))\n model.set_cfa_pattern(cfa)\n model.set_srgb_conversion(srgb)\n\n sample_Y = model.process(sample_x).numpy()\n\n if patch_size > 0:\n xx = (sample_y.shape[2] - patch_size) // 2\n yy = (sample_y.shape[1] - patch_size) // 2\n sample_y = sample_y[:, yy:yy+patch_size, xx:xx+patch_size, :]\n sample_Y = sample_Y[:, yy:yy+patch_size, xx:xx+patch_size, :]\n\n psnrs = metrics.psnr(sample_y, sample_Y)\n ssims = metrics.ssim(sample_y, sample_Y)\n\n print('sample x: {}'.format(sample_x.shape))\n print('sample y: {}'.format(sample_y.shape))\n print('sample Y: {}'.format(sample_Y.shape))\n\n # Plot images ---------------------------------------------------------------------------------\n if len(sample_y) > 1:\n sample_y = plots.thumbnails(sample_y, batch, True)\n sample_Y = plots.thumbnails(sample_Y, batch, True)\n else:\n sample_y = sample_y.squeeze()\n sample_Y = sample_Y.squeeze()\n\n print('thumbnails: {}'.format(sample_y.shape))\n\n ncols = 1 if sample_y.shape[1] > sample_y.shape[0] else 2\n nrows = 2 if ncols == 1 else 1\n fig, axes = plt.subplots(nrows, ncols)\n\n plots.image(sample_Y, '{}, PSNR={:.1f} dB, SSIM={:.2f} : {{}}'.format(model.model_code, float(psnrs.mean()), float(ssims.mean())), axes=axes[0])\n plots.image(sample_y, 'Target RGB images () : {}', axes=axes[1])\n\n plt.show()\n plt.close()","def SetUpLayerManager(self):\n pass","def inception_network():\n input = K.Input(shape=(224, 224, 3))\n\n conv_1 = K.layers.Conv2D(\n filters=64,\n kernel_size=7,\n strides=2,\n padding='same',\n activation='relu'\n )(input)\n\n pool_1 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(conv_1)\n\n conv_2 = K.layers.Conv2D(\n filters=64,\n kernel_size=1,\n strides=1,\n padding='valid',\n activation='relu'\n )(pool_1)\n\n conv_3 = K.layers.Conv2D(\n filters=192,\n kernel_size=3,\n strides=1,\n padding='same',\n activation='relu'\n )(conv_2)\n\n pool_2 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(conv_3)\n\n inception_1 = inception_block(pool_2, (64, 96, 128, 16, 32, 32))\n inception_2 = inception_block(inception_1, (128, 128, 192, 32, 96, 64))\n\n pool_3 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(inception_2)\n\n inception_3 = inception_block(pool_3, (192, 96, 208, 16, 48, 64))\n inception_4 = inception_block(inception_3, (160, 112, 224, 24, 64, 64))\n inception_5 = inception_block(inception_4, (128, 128, 256, 24, 64, 64))\n inception_6 = inception_block(inception_5, (112, 144, 288, 32, 64, 64))\n inception_7 = inception_block(inception_6, (256, 160, 320, 32, 128, 128))\n\n pool_4 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same',\n )(inception_7)\n\n inception_8 = inception_block(pool_4, (256, 160, 320, 32, 128, 128))\n inception_9 = inception_block(inception_8, (384, 192, 384, 48, 128, 128))\n\n pool_5 = K.layers.AveragePooling2D(\n pool_size=7,\n strides=1,\n padding='valid'\n )(inception_9)\n\n dropout = K.layers.Dropout(rate=0.4)(pool_5)\n linear = K.layers.Dense(units=1000, activation='softmax')(dropout)\n\n return K.Model(inputs=input, outputs=linear)","def __init__(self):\n super(GatherLastLayer, self).__init__()","def _save_data(self, outputs, latents):\n list(map(lambda e, v: e.append(v), self.curr_patches, outputs))\n list(map(lambda e, v: e.append(v), self.curr_latents, latents))\n self.count_patches += len(self.curr_patches[0][-1])\n\n while self.count_patches >= self.patches_ref:\n aux_patches = list(map(lambda e: tf.concat(e, axis=0),\n self.curr_patches))\n aux_latents = list(map(lambda e: tf.concat(e, axis=0),\n self.curr_latents))\n\n data_name = str((self.out_folder / str(Folders.RAW_DATA))\n / self.curr_img.stem)\n for lvl, (aux_p, aux_l) in enumerate(zip(aux_patches, aux_latents)):\n patches, remaining = tf.split(\n aux_p, [self.patches_ref, len(aux_p) - self.patches_ref],\n axis=0)\n self.curr_patches[lvl] = [remaining]\n latent, remaining = tf.split(\n aux_l, [self.patches_ref, len(aux_l) - self.patches_ref],\n axis=0)\n self.curr_latents[lvl] = [remaining]\n\n patch_size = len(patches[0])\n with Image.open(self.curr_img) as img:\n # width, height -> height, width (to be compatible with\n # arrays)\n img_size = np.array(img.size)[::-1]\n p_size = np.ceil(img_size / patch_size).astype(int)\n\n img = tf.reshape(\n patches, [*p_size, *self.m_cfg['configs']['patch']])\n img = tf.transpose(img, [0, 2, 1, 3, 4])\n img = tf.reshape(img, [*patch_size * p_size, 3])\n img = tf.image.convert_image_dtype(img, tf.uint8)\n img = img[:img_size[0], :img_size[1], :]\n\n latent_name = data_name + '_latent' + str(lvl)\n serialized = tf.io.serialize_tensor(latent)\n tf.io.write_file(latent_name, serialized)\n\n img_name = data_name + '_' + str(lvl) + self.curr_img.suffix\n tf.io.write_file(img_name, tf.image.encode_png(\n img, compression=0))\n\n with open(Path(data_name).with_suffix('.txt'), 'w') as file:\n file.write(str(self.curr_img))\n\n self.count_patches = len(self.curr_patches[0][0])\n if len(self.img_paths):\n self.curr_img = Path(self.img_paths.pop(0))\n self.patches_ref = self.calc_n_patches(\n self.curr_img, self.m_cfg['configs']['patch'][0])","def multi_run_wrapper(args):\n\treturn img_preprocessing(*args)","def registerPatches(self,pl):\n self.set('patchmesh.patches',pl)","def _set_freeze_layers(self):\n for layer in self.encoder.layers[:self.freeze_layers]:\n layer.trainable = False","def __call__(self, inputs, training):\n\n with self._model_variable_scope():\n if self.data_format == 'channels_first':\n # Convert the inputs from channels_last (NHWC) to channels_first (NCHW).\n # This provides a large performance boost on GPU. See\n # https://www.tensorflow.org/performance/performance_guide#data_formats\n inputs = tf.transpose(a=inputs, perm=[0, 3, 1, 2])\n\n with tf.variable_scope('stem'):\n inputs = conv2d_fixed_padding(\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n inputs = tf.identity(inputs, 'stem_conv1')\n\n inputs = conv2d_fixed_padding(\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n inputs = tf.identity(inputs, 'stem_conv2')\n\n # stage 1\n with tf.variable_scope('stage1'):\n stage1_out = block_layer(\n inputs=inputs, inp_filters=self.num_filters,\n filters=self.num_filters, bottleneck=True,\n block_fn=_bottleneck_block_v1, blocks=self.block_sizes[0],\n strides=1, training=training,\n name='stage1', data_format=self.data_format)\n\n pre_stage_channels = self.num_filters * 4\n stagex_input = [stage1_out]\n num_inchannels = [pre_stage_channels]\n\n # stage 2 to last stage\n for i, num_modules in enumerate(self.module_sizes[1:], 2):\n with tf.variable_scope('stage{}'.format(i)):\n num_channels = [self.hrnet_size * 2 ** n for n in range(i)]\n transn = Transition(\n stagex_input[-1], num_outchannels=num_channels[-1], \n training=training, data_format=self.data_format)\n stagex_input.append(transn)\n num_inchannels.append(num_inchannels[-1])\n\n for j in range(num_modules):\n if i == len(self.module_sizes) and j == num_modules - 1:\n multi_scale_output = False\n else:\n multi_scale_output = True\n\n stagex_input = HighResolutionModule(\n stagex_input, num_branches=i, \n num_inchannels=num_inchannels, num_channels=num_channels,\n bottleneck=self.bottleneck, block_fn=self.block_fn,\n num_blocks=self.block_sizes[i - 1], training=training,\n name='stage{}_block{}'.format(i, j), data_format=self.data_format,\n multi_scale_output=multi_scale_output)\n num_inchannels = num_channels\n\n return stagex_input","def __call__(self, **kwargs):\n segname = 'block_{}_expand_relu'\n blocks = [13, 6, 3, 1]\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\n backbone_out = self._backbone.get_layer('block_16_project')\n\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\n p4 = self._fpn_block(p5, skips[1].output)\n p3 = self._fpn_block(p4, skips[2].output)\n p2 = self._fpn_block(p3, skips[3].output)\n\n s5 = self._conv_block(p5, 128)\n s4 = self._conv_block(p4, 128)\n s3 = self._conv_block(p3, 128)\n s2 = self._conv_block(p2, 128)\n\n s5 = tf.keras.layers.UpSampling2D(\n size=(8, 8),\n interpolation='nearest'\n )(s5)\n\n s4 = tf.keras.layers.UpSampling2D(\n size=(4, 4),\n interpolation='nearest'\n )(s4)\n\n s3 = tf.keras.layers.UpSampling2D(\n size=(2, 2),\n interpolation='nearest'\n )(s3)\n\n concat = [s5, s4, s3, s2]\n x = tf.keras.layers.Concatenate()(concat)\n x = tf.keras.layers.Conv2D(\n 64,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n x = tf.keras.layers.BatchNormalization()(x)\n x = tf.keras.layers.Activation('relu')(x)\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\n\n x = tf.keras.layers.Conv2D(\n 1,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n out = tf.keras.layers.Activation('sigmoid')(x)\n model = tf.keras.models.Model(\n inputs=self._backbone.input,\n outputs=out\n )\n\n return model","def _augment_pipeline_cfg(self):","def add_bluprint_layers(self, models_dict, graph, class_names, dimensions,\n show_activation, show_constant_input):\n top_model = models_dict['top-model']\n # Get the top-model color list containing\n # the base color and the layer shades\n top_model_color = top_model[1]\n # get the model\n top_model = top_model[0]\n\n # Get the layers of the model\n layers = top_model[\"config\"][\"layers\"]\n # Loop through the layers\n for layer in layers:\n # If the layer is not a model\n if layer[\"class_name\"] != \"Model\":\n # Get the layer name\n layer_name = layer[\"name\"]\n # If label only layer's class name\n if class_names:\n # Get the layer's information\n layer_info = self.viz_utils.query_layer(layer_name,\n models_dict)\n # Get the layer's class name\n layer_class = layer_info['class_name']\n # If the layer is a a constant input layer,\n # manually specify the class name\n if layer_name.find('constant_input') != -1:\n layer_class = 'Constant Input'\n # Depending on the class name\n # find the the layer shade\n # If the layer is a constant_input layer\n # the color is black\n model_color = top_model_color[1].get(layer_class, \"black\")\n else:\n # If don't use class names for layers\n # then use the layer name from the JSON\n layer_class = layer_name\n model_color = top_model_color[0]\n\n\n # Add the node to the graph\n graph = self.viz_utils.add_nodes(layer_name, graph,\n layer_class, model_color,\n dimensions, show_constant_input)\n\n # Add Blueprint Inbound Edges\n graph = self.connect_blueprint_inbounds(models_dict,\n layer, graph,\n class_names, dimensions,\n show_activation, show_constant_input)\n else:\n # Add Softmod\n graph = self.connect_softmod_in_blueprint(models_dict,\n layer, graph, class_names,\n dimensions, show_activation, show_constant_input)\n\n return graph","def intermediate_layer(layer, filters, kernel_size):\n layer = Conv2D(filters=filters, kernel_size=kernel_size,\n strides=(2, 2), padding=\"same\")(layer)\n layer = LeakyReLU(alpha=LEAKY_RELU_ALPHA)(layer)\n return layer","def load_registration_as_layers() -> None:\n viewer = getattr(widget, \"viewer\").value\n registration_directory = pathlib.Path(\n getattr(widget, \"registration_output_folder\").value\n )\n add_registered_image_layers(\n viewer, registration_directory=registration_directory\n )","def deep_dream_of_extreme_control(FLAGS,model,input_images=[],num_iterations=10,step_size=0.1):\n if len(input_images) == 0:\n # use predefined images\n img_dir='/esat/opal/kkelchte/docker_home/pilot_data/visualization_images'\n input_images=sorted([img_dir+'/'+f for f in os.listdir(img_dir)])\n\n print(\"[tools.py]: extracting deep dream maps of {0} in {1}\".format([os.path.basename(i) for i in input_images], os.path.dirname(input_images[0])))\n \n # experts=np.asarray([[k]*(FLAGS.action_quantity if FLAGS.discrete else 1) for v in sorted(model.factor_offsets.values()) for k in model.factor_offsets.keys() if model.factor_offsets[k]==v]).flatten()\n\n inputs = load_images(input_images, model.input_size[1:])\n \n # collect gradients for output endpoint of evaluation model\n grads={}\n with tf.device('/cpu:0'):\n output_tensor = model.endpoints['eval']['outputs']\n for i in range(output_tensor.shape[1].value):\n layer_loss = output_tensor[:,i]\n gradients = tf.gradients(layer_loss, model.inputs)[0]\n gradients /= (tf.sqrt(tf.reduce_mean(tf.square(gradients))) + 1e-5)\n grads[output_tensor.name+'_'+str(i)]=gradients\n\n\n # apply gradient ascent for all outputs and each input image\n # if number of outputs ==1 apply gradient descent for contrast\n if len(grads.keys())== 1:\n opposite_results={}\n else:\n opposite_results=None\n\n import copy\n results = {}\n for gk in grads.keys(): \n results[gk]=copy.deepcopy(inputs)\n if isinstance(opposite_results,dict): opposite_results[gk]=copy.deepcopy(inputs)\n\n for step in range(num_iterations):\n if step%10==0: print \"{0} step: {1}\".format(time.ctime(), step)\n for i,gk in enumerate(sorted(grads.keys())):\n results[gk] += step_size * model.sess.run(grads[gk], {model.inputs: results[gk]})\n if isinstance(opposite_results,dict):\n opposite_results[gk] -= step_size * model.sess.run(grads[gk], {model.inputs: opposite_results[gk]})\n\n # Normalize results within 0:1 range\n clean_results={}\n for gk in results.keys():\n clean_results[gk]=[]\n for i in range(results[gk].shape[0]):\n clean_results[gk].append(deprocess_image(results[gk][i], one_channel=True))\n # results[gk][i]=deprocess_image(results[gk][i], one_channel=True)\n if isinstance(opposite_results,dict):\n opposite_results[gk][i]=deprocess_image(opposite_results[gk][i])\n\n # combine adjust input images in one overview image\n # one column for each input image\n # one row with each extreme control for separate and difference images\n num_rows=1+len(results.keys())\n fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\n # fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\n # add original images in first row\n for i in range(axes.shape[1]):\n axes[0, i].set_title(os.path.basename(input_images[i]).split('.')[0])\n axes[0, i].imshow(matplotlibprove(inputs[i]), cmap='inferno')\n axes[0, i].axis('off')\n\n # add for each filter the modified input\n row_index=1\n for gk in sorted(results.keys()):\n for i in range(axes.shape[1]):\n # print gk\n # axes[row_index, i].set_title('Grad Asc: '+gk.split('/')[1]+'/'+gk[-1]) \n axes[row_index, i].set_title('Grad Asc: '+gk)\n # axes[row_index, i].set_title(experts[row_index-1])\n\n axes[row_index, i].imshow(np.concatenate((inputs[i],np.expand_dims(clean_results[gk][i],axis=2)), axis=2), cmap='inferno')\n # axes[row_index, i].imshow(matplotlibprove(results[gk][i]), cmap='inferno')\n axes[row_index, i].axis('off')\n row_index+=1\n # In cas of continouos controls: visualize the gradient descent and difference\n # if isinstance(opposite_results,dict):\n # for gk in opposite_results.keys():\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Grad Desc: '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Grad Desc: '+gk) \n # axes[row_index, i].imshow(matplotlibprove(opposite_results[gk][i]), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n \n # # add difference\n # for gk in opposite_results.keys():\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Diff: '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Diff: '+gk) \n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((opposite_results[gk][i]-results[gk][i])**2)), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n # else:\n # # add difference between 2 exteme actions\n # gk_left=sorted(results.keys())[0]\n # gk_right=sorted(results.keys())[-1]\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Diff : '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Diff : '+gk) \n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((results[gk_left][i]-results[gk_right][i])**2)), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n \n \n plt.savefig(FLAGS.summary_dir+FLAGS.log_tag+'/control_dream_maps.jpg',bbox_inches='tight')\n # plt.show()","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n TransformerDecoderBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n multi_channel_cross_attention=self.multi_channel_cross_attention,\n name=(\"layer_%d\" % i)))\n super(TransformerDecoder, self).build(unused_input_shapes)","def __init__(self, layer_list_info):\n self.layer_list_info = layer_list_info","def GetShapesOfLayer(self, *args):\n return _XCAFDoc.XCAFDoc_LayerTool_GetShapesOfLayer(self, *args)","def depth_rendering(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\n\n x_indices = np.arange(lf_size[1])\n y_indices = np.arange(lf_size[2])\n b_indices = np.arange(lf_size[0])\n\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\n\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\n stacked = np.stack(grid, 2)\n positions = stacked.reshape(-1, 2)\n\n # Compute the distance from each lf position from the reference view\n # Repeat the elements of this to match the size of the disparity map\n ref_pos = np.array(\n [lf_one_way // 2, lf_one_way // 2])\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\n\n\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\n\n # Compute the shifted pixels\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\n\n #indices for linear interpolation in a square around the central point\n x_low = np.around(x_shifted).astype(int)\n #x_high = x_low + 1\n\n y_low = np.around(y_shifted).astype(int)\n #y_high = y_low + 1\n\n #Place co-ordinates outside the image back into the image\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\n #x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\n #y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\n\n #Gather the interpolation points\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\n #interp_pts_2 = np.stack((x_low_clip, y_high_clip))\n #interp_pts_3 = np.stack((x_high_clip, y_low_clip))\n #interp_pts_4 = np.stack((x_high_clip, y_high_clip))\n\n #Index into the images\n desired_shape = lf_size\n res_1 = torch_big_sample(ref_view, interp_pts_1, desired_shape)\n return res_1\n res_2 = torch_big_sample(ref_view, interp_pts_2, desired_shape)\n res_3 = torch_big_sample(ref_view, interp_pts_3, desired_shape)\n res_4 = torch_big_sample(ref_view, interp_pts_4, desired_shape)\n\n #Compute interpolation weights\n x_low_f = x_low.astype(np.float32)\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\n d_x_high = 1.0 - d_x_low\n y_low_f = y_low.astype(np.float32)\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\n d_y_high = 1.0 - d_y_low\n\n w1 = torch.from_numpy(d_x_low * d_y_low)\n w2 = torch.from_numpy(d_x_low * d_y_high)\n w3 = torch.from_numpy(d_x_high * d_y_low)\n w4 = torch.from_numpy(d_x_high * d_y_high)\n\n #THEY AGREE AT THIS POINT\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape), res_1)\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape), res_2)\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape), res_3)\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape), res_4)\n\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\n torch.add(novel_view, weighted_4, out=novel_view)\n return novel_view","def process_layer(layer_data):\n layer_name = layer_data['name']\n\n # Instantiate layer.\n from tensorflow.python.keras._impl.keras.layers import deserialize as deserialize_layer # pylint: disable=g-import-not-at-top\n\n layer = deserialize_layer(layer_data, custom_objects=custom_objects)\n created_layers[layer_name] = layer\n\n # Gather layer inputs.\n inbound_nodes_data = layer_data['inbound_nodes']\n for node_data in inbound_nodes_data:\n # We don't process nodes (i.e. make layer calls)\n # on the fly because the inbound node may not yet exist,\n # in case of layer shared at different topological depths\n # (e.g. a model such as A(B(A(B(x)))))\n add_unprocessed_node(layer, node_data)","def __init__(self, stage, layer=None, proxyPrimPrefix=\"\", renderPrimPrefix=\"\"):\n\n self._stage = stage\n\n # Layer\n if layer is None:\n self._layer = Sdf.Layer.CreateAnonymous()\n elif isinstance(layer, Sdf.Layer):\n self._layer = layer\n elif isinstance(layer, str):\n self._layer = Sdf.Layer.Open(layer)\n\n self._layer.SetMuted(True)\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\n self._proxyPrimPrefix = proxyPrimPrefix\n self._renderPrimPrefix = renderPrimPrefix","def __reslayer(self, inputs, in_filters, out_filters, stride=1):\n with tf.variable_scope('sub1'):\n kernel = tf.get_variable('weights', [3, 3, in_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n conv = tf.nn.conv2d(inputs, kernel, [1, stride, stride, 1],\n padding='SAME',\n name='conv')\n batch_norm = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\n conv = tf.nn.elu(batch_norm, 'elu')\n\n with tf.variable_scope('sub2'):\n kernel = tf.get_variable('weights',\n [3, 3, out_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n conv = tf.nn.conv2d(conv, kernel, [1, 1, 1, 1], padding='SAME',\n name='conv1')\n bias = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\n\n with tf.variable_scope('subadd'):\n if in_filters != out_filters:\n kernel = tf.get_variable('weights', [1, 1, in_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n inputs = tf.nn.conv2d(\n inputs, kernel, [1, stride, stride, 1], padding='SAME')\n bias += inputs\n conv = tf.nn.elu(bias, 'elu')\n\n num = np.power(2, np.floor(np.log2(out_filters) / 2))\n\n grid = self.__put_activations_on_grid(conv, (int(num),\n int(out_filters /\n num)))\n tf.summary.image('sub2/activations', grid, max_outputs=1)\n\n return conv","def depth_rendering_gpu(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\n\n x_indices = np.arange(lf_size[1])\n y_indices = np.arange(lf_size[2])\n b_indices = np.arange(lf_size[0])\n\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\n\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\n stacked = np.stack(grid, 2)\n positions = stacked.reshape(-1, 2)\n\n # Compute the distance from each lf position from the reference view\n # Repeat the elements of this to match the size of the disparity map\n ref_pos = np.array(\n [lf_one_way // 2, lf_one_way // 2])\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\n\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\n\n # Compute the shifted pixels\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\n\n #indices for linear interpolation in a square around the central point\n x_low = np.floor(x_shifted).astype(int)\n x_high = x_low + 1\n\n y_low = np.floor(y_shifted).astype(int)\n y_high = y_low + 1\n\n #Place co-ordinates outside the image back into the image\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\n x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\n y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\n\n #Gather the interpolation points\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\n interp_pts_2 = np.stack((x_low_clip, y_high_clip))\n interp_pts_3 = np.stack((x_high_clip, y_low_clip))\n interp_pts_4 = np.stack((x_high_clip, y_high_clip))\n\n #Index into the images\n desired_shape = lf_size\n ref_view = torch.tensor(ref_view, dtype=torch.float32).cuda()\n res_1 = torch_tensor_sample(ref_view, interp_pts_1, desired_shape)\n res_2 = torch_tensor_sample(ref_view, interp_pts_2, desired_shape)\n res_3 = torch_tensor_sample(ref_view, interp_pts_3, desired_shape)\n res_4 = torch_tensor_sample(ref_view, interp_pts_4, desired_shape)\n\n #Compute interpolation weights\n x_low_f = x_low.astype(np.float32)\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\n d_x_high = 1.0 - d_x_low\n y_low_f = y_low.astype(np.float32)\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\n d_y_high = 1.0 - d_y_low\n\n w1 = torch.from_numpy(d_x_low * d_y_low)\n w2 = torch.from_numpy(d_x_low * d_y_high)\n w3 = torch.from_numpy(d_x_high * d_y_low)\n w4 = torch.from_numpy(d_x_high * d_y_high)\n\n #THEY AGREE AT THIS POINT\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape).cuda(), res_1)\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape).cuda(), res_2)\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape).cuda(), res_3)\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape).cuda(), res_4)\n\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\n torch.add(novel_view, weighted_4, out=novel_view)\n return novel_view","def __call__(self, inputs, *args, **kwargs):\n out = inputs\n for i, layer in enumerate(self.layers):\n if i == 0:\n out = layer(out, *args, **kwargs)\n else:\n out = layer(out)\n return out","def add_convolution_layers(self):\n with tf.variable_scope(\"conv_layer_1\"):\n # our input to this layer is from word embeddings. which would be of\n # shape (batch_size, number of words in the sentence, dimension of\n # embedding). For us dimension of embedding is the number of input\n # channels, self.out_channels_1 is the output channels.\n\n # First we have to define a filter. The shape of the filter would be [\n # self.filter_size_1, dimension of embedding, self.out_channels_1]\n filter1 = tf.get_variable(\"filter1\", dtype=tf.float32,\n shape=[self.filter_size_1,\n self.word_embeddings.shape[2],\n self.out_channels_1])\n conv_layer_1 = tf.nn.conv1d(self.word_embeddings, filters=filter1,\n stride=self.strides_1, padding=\"SAME\")\n conv_layer_1 = tf.nn.tanh(conv_layer_1)\n # now add pooling layer\n print(\"adding pooling to layer1\")\n\n if self.max_pooling:\n print(\"Adding max pooling layer\")\n pooling_layer = tf.layers.max_pooling1d\n else:\n print(\"Adding average pooling layer\")\n pooling_layer = tf.layers.average_pooling1d\n conv_layer_1 = pooling_layer(conv_layer_1,\n pool_size=self.pool_size_1,\n strides=self.strides_1,\n padding=\"SAME\")\n\n with tf.variable_scope(\"conv_layer_2\"):\n print(\"adding convolution layer 2\")\n # Input to this layer is conv_layer_1 from above.\n # get another filter now. Shape of this filter would be [\n # self.filter_size_2, self.out_channels_1, self.out_channels_2]\n filter2 = tf.get_variable(\"filter2\", dtype=tf.float32,\n shape=[self.filter_size_2,\n self.out_channels_1,\n self.out_channels_2])\n conv_layer_2 = tf.nn.conv1d(conv_layer_1, filters=filter2,\n stride=self.strides_2, padding=\"SAME\")\n conv_layer_2 = tf.nn.tanh(conv_layer_2)\n print(\"adding pooling to layer2\")\n conv_layer_2 = pooling_layer(conv_layer_2,\n pool_size=self.pool_size_2,\n strides=self.strides_2,\n padding=\"SAME\")\n\n self.cnn_layer = tf.nn.dropout(conv_layer_2, self.dropout_holder)\n self.word_reps = self.cnn_layer\n if self.use_additional and self.hybrid:\n self.cnn_layer = tf.concat([self.word_reps, self.word_embeddings_2], \n axis=-1)\n self.word_reps = self.cnn_layer","def _init_layers(self) -> None:\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\n self.embed_dims = self.self_attn.embed_dims\n self.ffn = FFN(**self.ffn_cfg)\n norms_list = [\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\n for _ in range(3)\n ]\n self.norms = ModuleList(norms_list)","def _add_layer(self, layer_dict, layer_name, input_layers, merge_mode=None, share_params_with=None):\n util.colorprint(layer_name, 'teal')\n \n layer_dict = dict(layer_dict)\n util.colorprint(layer_dict, 'red')\n \n if share_params_with is not None:\n print \"Warning: ignoring share_params_with\"\n \n layer_options = layer_dict[\"options\"]\n layer=None\n if layer_dict[\"type\"]==\"conv2d\":\n #TODO: remove below\n nb_filter, nb_row, nb_col = 3,3,3\n layer = keras.layers.convolutional.Convolution2D(nb_filter, nb_row, nb_col, **layer_options)\n elif layer_dict[\"type\"]==\"dense\":\n dim = layer_dict[\"output_dim\"]\n # del layer_options[\"output_dim\"]\n layer = keras.layers.core.Dense(dim, **layer_options) \n else:\n print \"Ursol Major\"\n RaiseError()\n # TODO: one of the layers is a string\n if isinstance(input_layers, list):\n #this means that there is input from a loop to this layer\n self.model.add_node(layer, name=layer_name, inputs=input_layers, merge_mode=merge_mode)\n else:\n self.model.add_node(layer, name=layer_name, input=input_layers)\n\n return layer_name","def _embed_composite_feature(self, features, embedding_layers):\n embedding_list = []\n for i in range(len(embedding_layers)):\n embedding_list.append(embedding_layers[i](features[:, :, i]))\n embedding = tf.add_n(embedding_list)\n return embedding","def __init__(self, config_data, in_dims, layer_num):\n self.layer_num = layer_num\n self.in_dims = in_dims\n self.out_dims = in_dims[\"width\"] * in_dims[\"height\"] * in_dims[\"channels\"]\n self.layer_type = config_data[\"type\"]\n self.name = config_data[\"name\"]\n self.params = []","def test_2d_local_convolution(batch_exp, input_height, input_width, in_channels, filters, drawer, is_first):\n batch_size = 2 ** batch_exp\n if is_first:\n data_format = \"channels_first\"\n else:\n data_format = \"channels_last\"\n kernel_height = drawer.draw(integers(min_value=1, max_value=input_height-1))\n kernel_width = drawer.draw(integers(min_value=1, max_value=input_width-1))\n\n if data_format == \"channels_last\":\n input_shape = (input_height, input_width, in_channels)\n else:\n input_shape = (in_channels, input_height, input_width)\n\n # Create keras model\n keras_model = keras.Sequential()\n keras_local = keras.layers.LocallyConnected2D(filters, (kernel_height, kernel_width),\n use_bias=True,\n bias_initializer=\"random_uniform\",\n input_shape=input_shape,\n data_format=data_format)\n keras_model.add(keras_local)\n\n # Setup weights\n weights = drawer.draw(arrays(shape=keras_local.kernel.shape.as_list(), dtype=np.float32, elements=floats(min_value=-1e6, max_value=1e6)))\n keras_model.set_weights([weights])\n\n # Translate keras to PyTorch model\n torch_model = translate.translate_layer(keras_local)[0].to(device)\n\n keras_input = np.random.uniform(-10, 10, (batch_size,) + input_shape).astype(np.float32)\n keras_output = keras_model.predict(keras_input)\n\n torch_input = torch.Tensor(keras_input).to(device)\n torch_output = torch_model(torch_input).cpu().data.numpy()\n nan_fill = np.random.uniform(-5, 5)\n assert (np.isnan(keras_output) == np.isnan(torch_output)).all()\n keras_output[np.isnan(keras_output)] = nan_fill\n torch_output[np.isnan(torch_output)] = nan_fill\n assert np.isclose(keras_output, torch_output, atol=1e-3, rtol=1e-3).all()","def multiply(self, layer):\n pass","def forward(self,bottom,top):\n # imgmaps = self.batch_loader.mixup_gen()\n # print(\"DataLayer forward!!\")\n trainX, trainY = self.batch_loader.batch_imgs()\n # print(\"trainX:\",trainX.shape)\n # print(\"trainY:\",trainY.shape)\n # print(\"trainY:\", trainY)\n # print(\"top[0].data.shape:\",top[0].data.shape)\n # print(\"top[1].data.shape:\", top[1].data.shape)\n top[0].data[:, ...] = trainX\n top[1].data[:, ...] = trainY\n # print(\"DataLayer forward!!\")","def apply_net_tiled(net,img):\n\n # borders = [8,20,20] ## border width within each patch that is thrown away after prediction\n # patchshape_padded = [32,240,240] ## the size of the patch that we feed into the net. must be divisible by 8 or net fails.\n # patchshape = [16,200,200] ## must be divisible by 8 to avoid artifacts.\n # stride = [16,200,200] ## same as patchshape in this case\n def f(n,m): return (ceil(n/m)*m)-n ## f(n,m) gives padding needed for n to be divisible by m\n def g(n,m): return (floor(n/m)*m)-n ## f(n,m) gives un-padding needed for n to be divisible by m\n\n a,b,c = img.shape[1:]\n q,r,s = f(a,8),f(b,8),f(c,8) ## calculate extra border needed for stride % 8 = 0\n\n ZPAD,YPAD,XPAD = 8,24,24\n\n img_padded = np.pad(img,[(0,0),(ZPAD,ZPAD+q),(YPAD,YPAD+r),(XPAD,XPAD+s)],mode='constant') ## pad for patch borders\n output = np.zeros(img.shape)\n\n zs = np.r_[:a:16]\n ys = np.r_[:b:200]\n xs = np.r_[:c:200]\n\n for x,y,z in itertools.product(xs,ys,zs):\n qe,re,se = min(z+16,a+q),min(y+200,b+r),min(x+200,c+s)\n ae,be,ce = min(z+16,a),min(y+200,b),min(x+200,c)\n patch = img_padded[:,z:qe+2*ZPAD,y:re+2*YPAD,x:se+2*XPAD]\n patch = torch.from_numpy(patch).cuda().float()\n patch = net(patch[None])[0,:,ZPAD:-ZPAD,YPAD:-YPAD,XPAD:-XPAD].detach().cpu().numpy()\n output[:,z:ae,y:be,x:ce] = patch[:,:ae-z,:be-y,:ce-x]\n\n return output","def layeredTexturePort(*args, annotation: Union[AnyStr, bool]=\"\", backgroundColor:\n Union[List[float, float, float], bool]=None, defineTemplate: AnyStr=\"\",\n docTag: Union[AnyStr, bool]=\"\", dragCallback: Script=None, dropCallback:\n Script=None, enable: bool=True, enableBackground: bool=True,\n enableKeyboardFocus: bool=True, exists: bool=True, fullPathName:\n bool=True, height: Union[int, bool]=0, highlightColor: Union[List[float,\n float, float], bool]=None, isObscured: bool=True, manage: bool=True,\n noBackground: bool=True, node: name=None, numberOfPopupMenus: bool=True,\n parent: Union[AnyStr, bool]=\"\", popupMenuArray: bool=True,\n preventOverride: bool=True, selectedAlphaControl: AnyStr=\"\",\n selectedBlendModeControl: AnyStr=\"\", selectedColorControl: AnyStr=\"\",\n selectedIsVisibleControl: AnyStr=\"\", statusBarMessage: AnyStr=\"\",\n useTemplate: AnyStr=\"\", visible: bool=True, visibleChangeCommand:\n Union[Script, bool]=None, width: Union[int, bool]=0, q=True, query=True,\n e=True, edit=True, **kwargs)->Union[AnyStr, Any]:\n pass"],"string":"[\n \"def run(layers):\",\n \"def build_layers(self):\\n raise NotImplementedError\",\n \"def layers(self): # -> LayerView:\\n ...\",\n \"def UpdateLayers(self):\\n pass\",\n \"def layers(self, layers):\\n self._layers = layers\\n prev = None\\n for layer in layers:\\n if not layer.inputs and prev is not None:\\n layer.inputs = [prev]\\n prev = layer\",\n \"def layers(self, layers):\\n\\n self._layers = layers\",\n \"def __init__(self, layers):\\n\\n\\t\\tself.layers = layers\",\n \"def handle_layers(context, model, toplayer, layerids, materials, update, import_hidden=False):\\n #setup main container to hold all layer collections\\n layer_col_id=\\\"Layers\\\"\\n if not layer_col_id in context.blend_data.collections:\\n layer_col = context.blend_data.collections.new(name=layer_col_id)\\n try:\\n toplayer.children.link(layer_col)\\n except Exception:\\n pass\\n else:\\n #If \\\"Layers\\\" collection is in place, we assume the plugin had imported 3dm before\\n layer_col = context.blend_data.collections[layer_col_id]\\n\\n # build lookup table for LayerTable index\\n # from GUID, create collection for each\\n # layer\\n for lid, l in enumerate(model.Layers):\\n if not l.Visible and not import_hidden:\\n continue\\n lcol = utils.get_iddata(context.blend_data.collections, l.Id, l.Name, None)\\n layerids[str(l.Id)] = (lid, lcol)\\n utils.tag_data(layerids[str(l.Id)][1], l.Id, l.Name)\\n '''\\n matname = l.Name + \\\"+\\\" + str(l.Id)\\n if matname not in materials:\\n laymat = utils.get_iddata(context.blend_data.materials, l.Id, l.Name, None)\\n if update:\\n\\t laymat.use_nodes = True\\n\\t r, g, b, _ = l.Color\\n\\t principled = PrincipledBSDFWrapper(laymat, is_readonly=False)\\n\\t principled.base_color = (r/255.0, g/255.0, b/255.0)\\n materials[matname] = laymat\\n '''\\n # second pass so we can link layers to each other\\n for l in model.Layers:\\n # link up layers to their parent layers\\n if str(l.ParentLayerId) in layerids:\\n parentlayer = layerids[str(l.ParentLayerId)][1]\\n try:\\n parentlayer.children.link(layerids[str(l.Id)][1])\\n except Exception:\\n pass\\n # or to the top collection if no parent layer was found\\n else:\\n try:\\n layer_col.children.link(layerids[str(l.Id)][1])\\n except Exception:\\n pass\",\n \"def consume_layer(self, reports):\\n layer_list = []\\n layer_count = 1\\n for report in reports:\\n layer = create_image_layer(report)\\n layer.layer_index = layer_count\\n layer_list.append(layer)\\n layer_count += 1\\n return layer_list\",\n \"def __call__(cls, *args, **kwargs):\\n layer = super(LayerAspect, cls).__call__(*args, **kwargs)\\n\\n if Job.Current:\\n Job.Current.addLayer(layer)\\n \\n layer.afterInit()\\n return layer\",\n \"def layers(self, layers):\\n self._layers = layers\\n self.thetas = []\\n prev = None\\n for layer in layers:\\n if not layer.inputs and prev is not None:\\n layer.inputs = [prev]\\n prev = layer\\n self.thetas.extend(layer.thetas())\",\n \"def layers(self, x):\\n raise NotImplementedError\",\n \"def setup_layer_structure(self):\\n self.page_rank_convolution_1 = self.layer(self.feature_number, self.args.layers[0], self.args.iterations, self.args.alpha)\\n self.page_rank_convolution_2 = self.layer(self.args.layers[0], self.args.layers[1], self.args.iterations, self.args.alpha)\\n self.page_rank_convolution_3 = self.layer(self.args.layers[1], self.class_number, self.args.iterations, self.args.alpha)\",\n \"def build_models(config):\\n inputs = Input(shape=(config['patch_height'], config['patch_width'], config['depth'], config['channel']),\\n name='patchimg')\\n\\n kernelinitfun = keras.initializers.RandomNormal(mean=0.0, stddev=0.1, seed=None)\\n activationfun = 'relu'\\n # kernelinitfun = 'glorot_normal'\\n\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_1')(inputs)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_2')(x)\\n # x = Dropout(0.3)(x)\\n x = BatchNormalization(name='bn1')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp1', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_1')(x)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_2')(x)\\n # x = Dropout(0.2)(x)\\n x = BatchNormalization(name='bn2')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp2', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_1')(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_2')(x)\\n # x = Dropout(0.5)(x)\\n x = BatchNormalization(name='bn3')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp3', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_1')(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_2')(x)\\n # x = Dropout(0.5)(x)\\n x = BatchNormalization(name='bn4')(x)\\n x = Activation(activationfun)(x)\\n\\n x4 = Flatten(name='aux_fx')(x)\\n\\n source_classifier = Dropout(0.5)(x4)\\n source_classifier = Dense(512, activation='softmax', name=\\\"mo1\\\")(source_classifier)\\n source_classifier = Dropout(0.5)(source_classifier)\\n source_classifier = Dense(128, activation='softmax', name=\\\"mo2\\\")(source_classifier)\\n # source_classifier = Dropout(0.3)(source_classifier)\\n source_classifier = Dense(1, name=\\\"mo\\\")(source_classifier)\\n\\n domain_classifier = Dense(32, activation='linear', name=\\\"do4\\\")(x4)\\n domain_classifier = BatchNormalization(name=\\\"do5\\\")(domain_classifier)\\n domain_classifier = Activation(\\\"elu\\\", name=\\\"do6\\\")(domain_classifier)\\n domain_classifier = Dropout(0.5)(domain_classifier)\\n\\n domain_classifier = Dense(2, activation='softmax', name=\\\"do\\\")(domain_classifier)\\n\\n adamop = keras.optimizers.Adam(learning_rate=1e-5, beta_1=0.9, beta_2=0.999, amsgrad=False)\\n comb_model = Model(inputs=inputs, outputs=[source_classifier, domain_classifier])\\n comb_model.compile(optimizer=adamop,\\n loss={'mo': 'mae', 'do': 'categorical_crossentropy'},\\n loss_weights={'mo': 1, 'do': 2}, metrics=['accuracy'], )\\n\\n source_classification_model = Model(inputs=inputs, outputs=[source_classifier])\\n source_classification_model.compile(optimizer=adamop,\\n loss={'mo': 'mae'}, metrics=['accuracy'], )\\n\\n domain_classification_model = Model(inputs=inputs, outputs=[domain_classifier])\\n domain_classification_model.compile(optimizer=adamop,\\n loss={'do': 'categorical_crossentropy'}, metrics=['accuracy'])\\n\\n embeddings_model = Model(inputs=inputs, outputs=[x4])\\n embeddings_model.compile(optimizer=adamop, loss='categorical_crossentropy', metrics=['accuracy'])\\n\\n return comb_model, source_classification_model, domain_classification_model, embeddings_model\",\n \"def __init__(self, parent, layer):\\n pass\",\n \"def __init__(self, kernel_size, filters, stage, block):\\n super().__init__(name='identity' + str(stage) + block)\\n filters1, filters2, filters3 = filters\\n if K.image_data_format() == 'channels_last':\\n bn_axis = 3\\n else:\\n bn_axis = 1\\n\\n conv_name_base = 'res' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n self.conv1 = layers.Conv2D(\\n filters1, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2a')\\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\\n self.act1 = layers.Activation('relu')\\n\\n self.conv2 = layers.Conv2D(\\n filters2,\\n kernel_size,\\n padding='same',\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2b')\\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\\n self.act2 = layers.Activation('relu')\\n\\n self.conv3 = layers.Conv2D(\\n filters3, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2c')\\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\\n\\n self.add = layers.Add()\\n self.act = layers.Activation('relu')\",\n \"def setup_layers(self):\\n if self.args.model == \\\"exact\\\":\\n self.layer = PPNPLayer\\n else:\\n self.layer = APPNPLayer\\n self.setup_layer_structure()\",\n \"def set_hooks(self):\\n for layer_idx, layer in enumerate(self.vgg_features):\\n if layer_idx in self.layers_to_watch:\\n layer.register_forward_hook(self.layer_watch_hooks(layer_idx))\",\n \"def __init__(self, kernel_size, filters, stage, block, strides=(2, 2)):\\n super().__init__(name='conv_block' + str(stage) + block)\\n filters1, filters2, filters3 = filters\\n if K.image_data_format() == 'channels_last':\\n bn_axis = 3\\n else:\\n bn_axis = 1\\n conv_name_base = 'res' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n self.conv1 = layers.Conv2D(\\n filters1, (1, 1),\\n strides=strides,\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2a')\\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\\n self.act1 = layers.Activation('relu')\\n\\n self.conv2 = layers.Conv2D(\\n filters2,\\n kernel_size,\\n padding='same',\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2b')\\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\\n self.act2 = layers.Activation('relu')\\n\\n self.conv3 = layers.Conv2D(\\n filters3, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2c')\\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\\n\\n self.shortcut_conv = layers.Conv2D(\\n filters3, (1, 1),\\n strides=strides,\\n kernel_initializer='he_normal',\\n name=conv_name_base + '1')\\n self.shortcut_bn = layers.BatchNormalization(\\n axis=bn_axis, name=bn_name_base + '1')\\n\\n self.add = layers.Add()\\n self.act = layers.Activation('relu')\",\n \"def _build(self):\\n with tf.variable_scope (self.name + '_architecutre') as scope:\\n images_square = unflatten_layer ( self.images )\\n visualize_images(images_square)\\n\\n # Conv Layer 1\\n conv1_out, params = conv_2d_layer ( input = images_square,\\n neurons = CONV_1_N,\\n filter_size = CONV_1_FILT,\\n name = 'enc_conv_1',\\n visualize = True )\\n process_params(params, name = self.name)\\n e1_params = params\\n pool1_out = max_pool_2d_layer ( input = conv1_out, name = 'enc_pool_1')\\n # lrn1_out = local_response_normalization_layer (pool1_out, name = 'lrn_1' )\\n\\n # Conv Layer 2\\n conv2_out, params = conv_2d_layer ( input = pool1_out,\\n neurons = CONV_2_N,\\n filter_size = CONV_2_FILT,\\n name = 'enc_conv_2' )\\n process_params(params, name = self.name)\\n e2_params = params\\n pool2_out = max_pool_2d_layer ( input = conv2_out, name = 'enc_pool_2')\\n # lrn2_out = local_response_normalization_layer (pool2_out, name = 'lrn_2' )\\n\\n flattened = flatten_layer(pool2_out)\\n\\n # Dropout Layer 1 \\n flattened_dropout = dropout_layer ( input = flattened,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_1') \\n\\n # Dot Product Layer 1\\n fc1_out, params = dot_product_layer ( input = flattened_dropout,\\n neurons = HIDDEN_1,\\n name = 'enc_dot_1')\\n process_params(params, name = self.name)\\n e3_params = params \\n\\n # Dropout Layer 2 \\n fc1_out_dropout = dropout_layer ( input = fc1_out,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_2')\\n # Dot Product Layer 2\\n fc2_out, params = dot_product_layer ( input = fc1_out_dropout, \\n neurons = HIDDEN_2,\\n name = 'enc_dot_2')\\n process_params(params, name = self.name)\\n e4_params = params \\n\\n # Dropout Layer 3 \\n fc2_out_dropout = dropout_layer ( input = fc2_out,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_3')\\n \\n # Dot Product Layer 2\\n self.codeword, params = dot_product_layer ( input = fc2_out_dropout, \\n neurons = CODEWORD_LENGTH,\\n activation = CODE_ACTIVATION,\\n name = 'enc_dot_2')\\n process_params(params, name = self.name)\\n process_codeword_normalization_regularizer(self.codeword, \\n coeff = AUTOENCODER_CODEWORD_COEFF,\\n name = self.name)\\n e5_params = params \\n # tf.summary.histogram('codewords', self.codeword)\\n # self.hash = threshold_layer ( input = self.codeword,\\n # name = 'hash')\\n # process_hash_regularizer(self.codeword, coeff = AUTOENCODER_HASH_COEFF,\\n # name = self.name)\\n\\n # Decoder ... \\n decoder_1_out, params = dot_product_layer ( input = self.codeword, \\n neurons = HIDDEN_2,\\n params = [tf.transpose(e5_params[0]), None],\\n name = 'decoder_dot_1')\\n d1_params = params\\n process_params([params[1]], name = self.name)\\n \\n dec_1_out_dropout = dropout_layer ( input = decoder_1_out,\\n prob = self.dropout_prob,\\n name = 'dec_dropout_1')\\n\\n decoder_2_out, params = dot_product_layer ( input = dec_1_out_dropout, \\n neurons = HIDDEN_1,\\n params = [tf.transpose(e4_params[0]), None],\\n name = 'decoder_dot_2')\\n d2_params = params\\n process_params([params[1]], name = self.name)\\n \\n # dropout 2\\n dec_2_out_dropout = dropout_layer ( input = decoder_2_out,\\n prob = self.dropout_prob,\\n name = 'dec_dropout_2')\\n\\n decoder_3_out, params = dot_product_layer ( input = dec_2_out_dropout, \\n neurons = 1250,\\n params = [tf.transpose(e3_params[0]), None],\\n name = 'decoder_dot_3')\\n d3_params = params\\n process_params([params[1]], name = self.name)\\n\\n # DeConv Layer 1\\n # The output shapes need to be changed according to architecture.\\n\\n dec_3_square = unflatten_layer ( decoder_3_out, channels = CONV_2_N )\\n upsample_1 = upsampling_layer (dec_3_square, size = (10,10), name = 'dec_upsampling_1')\\n\\n deconv1_out, params = deconv_2d_layer ( input = upsample_1,\\n neurons = CONV_1_N,\\n filter_size = CONV_2_FILT,\\n output_shape = (12,12),\\n # n_outs = MINI_BATCH_SIZE,\\n stride = (1,1,1,1), \\n params = [e2_params[0], None], \\n name = 'dec_deconv_1' )\\n\\n process_params([params[1]], name = self.name)\\n d4_params = params\\n\\n # DeConv Layer 2\\n upsample_2 = upsampling_layer (deconv1_out, size = (24,24), name = 'dec_upsampling_2')\\n decoded_images_square, params = deconv_2d_layer ( input = upsample_2,\\n neurons = 1,\\n filter_size = CONV_1_FILT,\\n stride = (1,1,1,1),\\n output_shape = (28,28),\\n # n_outs = MINI_BATCH_SIZE, \\n params = [e1_params[0], None], \\n activation = 'tanh', \\n name = 'dec_deconv_2' )\\n \\n process_params([params[1]], name = self.name)\\n d5_params = params \\n \\n self.decoded = flatten_layer (decoded_images_square, in_shp = [-1, 28, 28, 1])\\n visualize_images(decoded_images_square, name = 'decoded')\\n # This is because transpose don't initialize.\\n self.params = [ [e5_params[0], d1_params[1] ],\\n [e4_params[0], d2_params[1] ],\\n [e3_params[0], d3_params[1] ],\\n [e2_params[0], d4_params[1] ],\\n [e1_params[0], d5_params[1] ] ]\\n\\n with tf.variable_scope (self.name + '_objectives') as scope: \\n with tf.variable_scope( self.name + '_decoder_error') as scope:\\n reconstruction_error = rmse(self.images, self.decoded) \\n tf.add_to_collection( self.name + '_objectives', reconstruction_error ) \\n tf.summary.scalar('reconstruction_error', reconstruction_error)\\n\\n self._cook_optimizer( \\n lr = AUTOENCODER_LR, \\n optimizer = AUTOENCODER_OPTIMIZER,\\n l1_coeff = AUTOENCODER_L1_COEFF,\\n l2_coeff = AUTOENCODER_WEIGHT_DECAY_COEFF)\",\n \"def setup_layers(self, params):\\n\\n #\\n # Input layer\\n #\\n\\n # Input and encoding layers\\n layer_stack = [\\n tf.placeholder(\\n tf.int32,\\n name=\\\"input\\\",\\n shape=[None, params.in_size]\\n )\\n ]\\n\\n in_layer = layer_stack[0]\\n\\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\\n character_depth = len(good_chars) + 2\\n\\n layer_stack.append(\\n self.make_onehot_encode_layer(\\n layer_stack[-1],\\n character_depth\\n )\\n )\\n\\n # Add conv layers\\n for idx, num_nodes in enumerate(params.conv_nodes):\\n layer_stack.append(\\n self.make_conv_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'conv_{}'.format(idx),\\n params.conv_params\\n )\\n )\\n\\n # Flatten convolutional layers\\n layer_stack.append(\\n tf.contrib.layers.flatten(\\n layer_stack[-1]\\n )\\n )\\n\\n # Add dense layers\\n for idx, num_nodes in enumerate(params.dense_nodes):\\n layer_stack.append(\\n self.make_dense_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'dense_{}'.format(idx),\\n params.dense_params\\n )\\n )\\n\\n # Output layer is broken into two pieces. The pre/post the application\\n # of the activation function. That's because the loss functions\\n # for cross-entropy rely on the pre-activation scores in preact\\n\\n preact_out_layer, out_layer = self.make_dense_output_layer(\\n layer_stack[-1],\\n params.out_size,\\n params.output_params\\n )\\n\\n target_layer = tf.placeholder(\\n \\\"float\\\",\\n name=\\\"target\\\",\\n shape=[None, params.out_size]\\n )\\n\\n if params.output_params.activation in ['sigmoid']:\\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\\n else:\\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\\n\\n return in_layer, out_layer, target_layer, loss\",\n \"def __call__(self, layers, inputs, name=None):\\r\\n def call_layer(l, x):\\r\\n if hasattr(l, \\\"call_by_computation_context\\\"):\\r\\n return l.call_by_computation_context(x, *self.args, **self.kwargs)\\r\\n else:\\r\\n return l(x, *self.args, **self.kwargs)\\r\\n\\r\\n if not isinstance(layers, (list, tuple)):\\r\\n # Single layer, just call\\r\\n return call_layer(layers, inputs)\\r\\n else:\\r\\n # Flatten the layers\\r\\n layers = tf.nest.flatten(layers)\\r\\n outputs = inputs\\r\\n\\r\\n # Add a name scope for wrapping\\r\\n if name is not None:\\r\\n scope_name = name\\r\\n else:\\r\\n scope_name = commonprefix([layer.name for layer in layers]).strip(\\\"-/\\\\\\\\?_ \\\\t\\\\n\\\") or \\\"ComputationContext\\\"\\r\\n with tf.name_scope(scope_name):\\r\\n for l in layers:\\r\\n outputs = call_layer(l, outputs)\\r\\n\\r\\n return outputs\",\n \"def __init__(self, layer_nest):\\n self._layer_nest = layer_nest\\n super().__init__()\",\n \"def generate_legacy_layers(self, images_map, content_retriever):\\n pass\",\n \"def generate_legacy_layers(self, images_map, content_retriever):\\n pass\",\n \"def preview_layerset(ls, size = 100, spacing = 100):\\n D = Device()\\n scale = size/100\\n num_layers = len(ls._layers)\\n matrix_size = int(np.ceil(np.sqrt(num_layers)))\\n sorted_layers = sorted(ls._layers.values(),\\n key = lambda x: (x.gds_layer, x.gds_datatype))\\n for n, layer in enumerate(sorted_layers):\\n R = rectangle(size = (100*scale, 100*scale), layer = layer)\\n T = text(text = '%s\\\\n%s / %s'\\\\\\n % (layer.name, layer.gds_layer, layer.gds_datatype),\\n size = 20*scale,\\n justify = 'center',\\n layer = layer)\\n\\n T.move((50*scale, -20*scale))\\n xloc = n % matrix_size\\n yloc = int(n // matrix_size)\\n D.add_ref(R)\\\\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\\n D.add_ref(T)\\\\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\\n return D\",\n \"def add_layers(self, layers):\\n\\n existing_layers = self.layers\\n assert len(existing_layers) > 0\\n for layer in layers:\\n assert layer.get_mlp() is None\\n layer.set_mlp(self)\\n layer.set_input_space(existing_layers[-1].get_output_space())\\n existing_layers.append(layer)\\n assert layer.layer_name not in self.layer_names\\n self.layer_names.add(layer.layer_name)\",\n \"def build(self):\\n\\n layers = GiraffeLayer.get_all_structural()\\n \\n for layer in layers:\\n\\n self.add_objects_from_layer(layer)\\n\\n return self\",\n \"def init_nnpatches(self):\\n # nnpatches = self._generate_nnpatches()\\n\\n # Register `nnpatches` to this `nnmodel`\\n # Refer: NNFramework._init_model_params(...)\\n\\n # # Register this `nnmodel` to `nnpatch`\\n # for nnpatch in nnpatches:\\n # nnpatch.add_nnmodels(self)\",\n \"def set_up_rlayer(self, rlname, rlname_other=None, include_layers=None,\\n exclude_layers=None, mask_layers=None):\\n scene = self.set_as_active()\\n layer_numbers = constants.layer_numbers\\n w_var.rlname = rlname\\n\\n if include_layers is None:\\n include_layers = w_var.layer_numbers_all_used\\n\\n if exclude_layers is None:\\n exclude_layers = []\\n\\n if mask_layers is None:\\n mask_layers = []\\n\\n if w_var.cb_clear_rlayers:\\n for layer in scene.render.layers[:-1]:\\n scene.render.layers.remove(layer)\\n\\n scene.render.layers.active.name = rlname\\n scene.render.layers.active.use = True\\n\\n new_rlayer = scene.render.layers.active\\n\\n # if not clearing render layers: creates new one\\n else:\\n new_rlayer = scene.render.layers.new(rlname)\\n scene.render.layers.active = new_rlayer\\n\\n # there needs to be two render layers in the same scene for freestyle compositing\\n if w_var.cb_composited:\\n w_var.rlname_other = rlname_other\\n other_rlayer = scene.render.layers.new(rlname_other)\\n other_rlayer.layers[19] = True\\n scene.render.layers[rlname_other].layers_zmask = (False,) * 20\\n\\n if w_var.cb_ao:\\n scene.render.layers[rlname].use_pass_ambient_occlusion = True\\n\\n if w_var.cb_composited:\\n scene.render.layers[rlname_other].use_pass_ambient_occlusion = True\\n\\n # because I can't deactivate a layer if it is the only active one\\n new_rlayer.layers[19] = True\\n \\n scene.render.layers[rlname].layers_exclude = (False,) * 20\\n scene.render.layers[rlname].layers_zmask = (False,) * 20\\n\\n for i in layer_numbers:\\n if w_var.cb_composited:\\n if i in w_var.layer_numbers_affected:\\n scene.render.layers[rlname].layers[i] = True\\n scene.render.layers[rlname_other].layers_zmask[i] = True\\n\\n else:\\n scene.render.layers[rlname].layers[i] = False\\n\\n if i in w_var.layer_numbers_other:\\n scene.render.layers[rlname_other].layers[i] = True\\n\\n else:\\n scene.render.layers[rlname_other].layers[i] = False\\n\\n else:\\n if i in include_layers:\\n scene.render.layers[rlname].layers[i] = True\\n\\n else:\\n scene.render.layers[rlname].layers[i] = False\\n\\n if i in mask_layers:\\n scene.render.layers[rlname].layers_zmask[i] = True\\n\\n if i in exclude_layers:\\n scene.render.layers[rlname].layers_exclude[i] = True\",\n \"def draw_layers(self):\\n\\t\\tfor z in xrange(0,16):\\n\\t\\t\\t#create surface for this layer\\n\\t\\t\\tsrf = pygame.Surface((16,128))\\n\\t\\t\\tfor x in xrange(0,16):\\n\\t\\t\\t\\tfor y in xrange(0,128):\\n\\t\\t\\t\\t\\tv = self.data[ self.xyz_to_offset( x,y,z) ]\\n\\t\\t\\t\\t\\tif v != 0:\\n\\t\\t\\t\\t\\t\\tsrf.fill( BLOCKS.get(v, [0,0])[1], \\t(x, 127 -y, 1, 1 ))\\n\\t\\t\\t#save layer to dict for this chunk\\n\\t\\t\\tself.layers[z] = srf\",\n \"def process_layers(size: Sequence[int],\\n layers: Union[Source, Layer, Sequence[Layer]],\\n a: Union[None, np.ndarray] = None,\\n status: Union[None, Queue] = None) -> np.ndarray:\\n # If no destination array was sent, it means we are either at the\\n # start of the layer processing or we are starting the processing\\n # of a layer group. If we are starting processing, then this will\\n # contain the final image data. If we are starting the processing\\n # of a layer group, this allows the layers in the group to only\\n # blend with the other layers in the group before they blend with\\n # the layers outside the group.\\n if a is None:\\n a = np.zeros(size, dtype=float)\\n\\n # If we got a sequence of layers, we process them recursively and\\n # return the result.\\n if isinstance(layers, Sequence):\\n for layer in layers:\\n a = process_layers(size, layer, a, status)\\n return a\\n\\n # If we got a source layer, process it.\\n if isinstance(layers.source, Source):\\n if status is not None:\\n src_name = layers.source.__class__.__name__\\n status.put((sw.MSG, f'Rendering {src_name}...'))\\n kwargs = {\\n 'source': layers.source,\\n 'size': size,\\n 'location': layers.location,\\n 'filters': layers.filters,\\n }\\n b = render_source(**kwargs)\\n if status is not None:\\n src_name = layers.source.__class__.__name__\\n status.put((sw.MSG, f'Rendered {src_name}.'))\\n status.put((sw.PROG,))\\n\\n # Otherwise we got a container layer, process its source and run\\n # any filters that are set on the layer.\\n else:\\n new_size = f.preprocess(size, layers.filters)\\n b = process_layers(new_size, layers.source, None, status)\\n b = f.process(b, layers.filters)\\n b = f.postprocess(b, layers.filters)\\n\\n # There are two possibilities for how the layers should be\\n # blended: masked or unmasked. Masked blends will have a\\n # Source in the mask attribute, which needs to be sent\\n # to the blending operation.\\n if layers.mask is not None:\\n if status is not None:\\n src_name = layers.mask.__class__.__name__\\n status.put((sw.MSG, f'Rendering {src_name}...'))\\n kwargs = {\\n 'source': layers.mask,\\n 'size': size,\\n 'filters': layers.mask_filters,\\n }\\n mask = render_source(**kwargs)\\n if status is not None:\\n src_name = layers.mask.__class__.__name__\\n status.put((sw.MSG, f'Rendered {src_name}.'))\\n status.put((sw.PROG,))\\n a, b, mask = _normalize_color_space(a, b, mask)\\n return layers.blend(a, b, layers.blend_amount, mask)\\n\\n # Unmasked layers have no mask, so Blend the image data and return.\\n a, b = _normalize_color_space(a, b)\\n return layers.blend(a, b, layers.blend_amount)\",\n \"def make_layers(self):\\n w, h = self.image.get_size()\\n shrink = pg.transform.smoothscale(self.image, (w//2, h//2))\\n self.mid_image = tools.tile_surface((w,h), shrink, True)\\n shrink = pg.transform.smoothscale(self.image, (w//4, h//4))\\n self.base = tools.tile_surface(prepare.SCREEN_SIZE, shrink, True)\",\n \"def __initAvailableLayerTypes(self):\\n from backend.caffe.path_loader import PathLoader\\n caffe = PathLoader().importCaffe()\\n layerNameMainParts = list(caffe.layer_type_list())\\n\\n res = {}\\n paramsPerLayerType = {}\\n\\n # calculate common parameters of all layer types\\n # by removing all which will be used for one specific layer type only\\n # also keep in mind which ones have been removed to readd them to specific layers\\n commonParams = self._availableParameterGroupDescriptors[\\\"LayerParameter\\\"].parameter() #use .parameter() on purpose\\n layerSpecificParameters = set()\\n for nameMainPart in layerNameMainParts:\\n specificParamsName = [nameMainPart + \\\"Parameter\\\"]\\n if moreLayerNameParameter.has_key(nameMainPart):\\n specificParamsName.append( moreLayerNameParameter[nameMainPart])\\n paramsPerLayerType[nameMainPart] = {}\\n for key, value in commonParams.items():\\n if value.isParameterGroup() and value.parameterName() in specificParamsName:\\n paramsPerLayerType[nameMainPart][key] = value\\n layerSpecificParameters.add(key)\\n\\n\\n # special case: shared params for loss layers\\n key = \\\"loss_param\\\"\\n value = commonParams[key]\\n del commonParams[key]\\n for nameMainPart in layerNameMainParts:\\n if LayerType.getCategoryByName(nameMainPart) == LayerType.CATEGORY_LOSS:\\n paramsPerLayerType[nameMainPart][key] = value\\n\\n # TODO is there a special case for the TransformationParameter?\\n\\n # create each layer type after one another\\n for nameMainPart in layerNameMainParts:\\n\\n # add common params to the specific ones\\n layerTypeParam = paramsPerLayerType[nameMainPart].keys()\\n paramsPerLayerType[nameMainPart].update(commonParams)\\n\\n irrelevant = layerSpecificParameters.difference(layerTypeParam)\\n res[nameMainPart] = LayerType(nameMainPart, paramsPerLayerType[nameMainPart], layerTypeParam, irrelevant)\\n\\n self._commonParams = commonParams\\n self._availableLayerTypes = res\",\n \"def EmbeddingLayers_pooling(pretrained=False, progress=True, **kwargs):\\n kwargs['groups'] = 32\\n kwargs['width_per_group'] = 4\\n # return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,\\n # **kwargs) \\n return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],\\n pretrained, progress, **kwargs)\",\n \"def __init__(self,layer_def,input,input_shape,rs,clone_from=None):\\n batch_size = int(layer_def.find(\\\"batchsize\\\").text)\\n image_size = int(layer_def.find(\\\"imagesize\\\").text)\\n image_channels = int(layer_def.find(\\\"imagechannels\\\").text)\\n self.layer_name = layer_def.attrib[\\\"name\\\"]\\n self.init(input, input_shape, batch_size,image_size,image_channels)\",\n \"def feed(self, *args):\\n\\t\\tprint(\\\"args: \\\", args)\\n\\t\\tassert len(args) != 0\\n\\t\\tself.terminals = []\\n\\t\\tfor fed_layer in args:#\\n\\t\\t\\ttry:\\n\\t\\t\\t\\tis_str = isinstance(fed_layer, basestring)\\n\\t\\t\\texcept NameError:\\n\\t\\t\\t\\tis_str = isinstance(fed_layer, str)\\n\\t\\t\\tif is_str:\\n\\t\\t\\t\\ttry:\\n\\t\\t\\t\\t\\tfed_layer = self.layers[fed_layer]\\n\\t\\t\\t\\t\\tprint(fed_layer)\\n\\t\\t\\t\\texcept KeyError:\\n\\t\\t\\t\\t\\traise KeyError('Unknown layer name fed: %s' % fed_layer)\\n\\t\\t\\tself.terminals.append(fed_layer)\\n\\t\\treturn self\",\n \"def append_layer(self, *args, **kwargs) :\\n \\n self.insert_layer(len(self._layers), *args, **kwargs)\",\n \"def get_data_from_layers(vk4_container, layers, step, is_image=False):\\n log.debug(\\\"Entering get_data_from_layers()\\\")\\n holder = []\\n for x in range(0, len(layers), step):\\n lay = layers[x:x + step]\\n log.debug(\\\"In get_data_from_layers()\\\\n\\\\tCurrent layer to append: %s\\\" % lay)\\n if lay not in layer_dict.keys():\\n log.error(\\\"In get_data_from_layers()\\\\n\\\\tLayer {}: {} from param \\\" \\\\\\n \\\"layers: {} is not a valid layer. See documentation.\\\"\\n .format(x, lay, layers))\\n holder.append(vk4_container.get_single_color_values(layer_dict[lay][0], layer_dict[lay][1]))\\n\\n if is_image:\\n log.debug(\\\"Exiting get_data_from_layers() where is_image is {}\\\".format(is_image))\\n return create_array_from_rgb_layers(vk4_container, holder)\\n else:\\n log.debug(\\\"Exiting get_data_from_layers() where is_image is {}\\\".format(is_image))\\n return create_composite_rgb_values(vk4_container, holder)\\n # return create_separate_rgb_values(vk4_container, holder)\",\n \"def initialize_layers(self, layers_config: dict, inputs=None):\\n layers_config = layers_config.copy()\\n input_lyrs = []\\n initiated_layers = OrderedDict()\\n wrp_layer = None # indicator for wrapper layers\\n first_layer = True\\n\\n for lyr, lyr_args in layers_config.items():\\n\\n lyr_config, lyr_inputs, named_outs, call_args = self.deconstruct_lyr_args(lyr, lyr_args)\\n\\n lyr_name, args, lyr_config, activation = self.check_lyr_config(lyr, lyr_config)\\n\\n if K.BACKEND == 'pytorch':\\n\\n if first_layer:\\n first_layer = False\\n\\n if callable(lyr_config):\\n lyr_initiated = lyr_config\\n else:\\n lyr_initiated = TORCH_LAYERS[lyr_name](**lyr_config)\\n setattr(self, lyr, lyr_initiated)\\n initiated_layers[lyr] = {\\\"layer\\\": lyr_initiated, \\\"named_outs\\\": named_outs, 'call_args': call_args,\\n 'inputs': lyr_inputs}\\n\\n else:\\n # may be user has defined layers without input layer, in this case add Input layer as first layer\\n if first_layer:\\n if inputs is not None: # This method was called by providing it inputs.\\n assert isinstance(inputs, tf.Tensor)\\n # since inputs have been defined, all the layers that will be added will be next to first layer\\n first_layer = False\\n layer_outputs = inputs\\n initiated_layers[layer_outputs.name] = {'layer': layer_outputs, 'tf_name': lyr_name}\\n\\n elif lyr_name != \\\"Input\\\":\\n if 'input_shape' in lyr_config: # input_shape is given in the first layer so make input layer\\n initialized_layer = LAYERS[\\\"Input\\\"](shape=lyr_config['input_shape'])\\n else:\\n # for simple dense layer based models, lookback will not be used\\n def_shape = (self.num_ins,) if self.lookback == 1 else (self.lookback, self.num_ins)\\n initialized_layer = LAYERS[\\\"Input\\\"](shape=def_shape)\\n\\n # first layer is built so next iterations will not be for first layer\\n first_layer = False\\n # put the first layer in memory to be used for model compilation\\n # add th layer which the user had specified as first layer\\n initiated_layers[initialized_layer.name] = {'layer': initialized_layer,\\n 'tf_name': lyr_name}\\n input_lyrs.append(initialized_layer)\\n\\n # The inputs to the layer have not been specified, so either it is an Input layer\\n if lyr_inputs is None:\\n # or it uses the previous outputs as inputs\\n if lyr_name == \\\"Input\\\":\\n # it is an Input layer, hence should not be called\\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'tf_name': lyr_name}\\n input_lyrs.append(initialized_layer)\\n else:\\n # it is executable and uses previous outputs as inputs\\n if lyr_name in ACTIVATION_LAYERS:\\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\\n wrp_layer = LAYERS[lyr_name]\\n # because wrapper layer name is property\\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\\n 'tf_name': lyr_name}\\n continue\\n elif \\\"LAMBDA\\\" in lyr_name.upper():\\n # lyr_config is serialized lambda layer, which needs to be deserialized\\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\\n # layers_config['lambda']['config'] still contails lambda, so we need to replace the python\\n # object (lambda) with the serialized version (lyr_config) so that it can be saved as json file.\\n layers_config[lyr]['config'] = lyr_config\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n else:\\n if wrp_layer is not None:\\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n wrp_layer = None\\n else:\\n if lyr_name == \\\"TemporalFusionTransformer\\\":\\n lyr_config['return_attention_components'] = True\\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n\\n else: # The inputs to this layer have been specified so they must exist in lyr_cache.\\n # it is an executable\\n if lyr_name in ACTIVATION_LAYERS:\\n\\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\\n wrp_layer = LAYERS[lyr_name]\\n # because wrapper layer name is property\\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\\n 'tf_name': lyr_name}\\n continue\\n elif \\\"LAMBDA\\\" in lyr_name.upper():\\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\\n layers_config[lyr]['config'] = lyr_config\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n else:\\n if wrp_layer is not None:\\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n wrp_layer = None\\n else:\\n layer_initialized = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': layer_initialized,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n\\n if activation is not None: # put the string back to dictionary to be saved in config file\\n lyr_config['activation'] = activation\\n\\n first_layer = False\\n\\n self.jsonize_lyr_config(lyr_config)\\n\\n # inputs = [] todo, indentify input layers\\n # for k,v in lyr_cache.items():\\n # since the model is not build yet and we have access to only output tensors of each list, this is probably\\n # # the only way to know that how many `Input` layers were encountered during the run of this method. Each\\n # tensor (except TimeDistributed) has .op.inputs attribute, which is empty if a tensor represents output of Input layer.\\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\\n # if k.upper() != \\\"TIMEDISTRIBUTED\\\" and hasattr(v, 'op'):\\n # if hasattr(v.op, 'inputs'):\\n # _ins = v.op.inputs\\n # if len(_ins) == 0:\\n # inputs.append(v)\\n # else: # not sure if this is the proper way of checking if a layer receives an input or not!\\n # if hasattr(v, '_keras_mask'):\\n # inputs.append(v)\\n\\n setattr(self, 'initiated_layers', initiated_layers)\\n setattr(self, 'input_lyrs', input_lyrs)\\n\\n\\n # todo,\\n # # for case when {Input -> Dense, Input_1}, this method wrongly makes Input_1 as output so in such case use\\n # # {Input_1, Input -> Dense }, thus it makes Dense as output and first 2 as inputs, so throwing warning\\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\\n # if len(layer_outputs.op.inputs) < 1:\\n # print(\\\"Warning: the output is of Input tensor class type\\\")\\n # else:\\n # if 'op' not in dir(layer_outputs): # layer_outputs does not have `op`, which means it has no incoming node\\n # print(\\\"Warning: the output is of Input tensor class type\\\")\\n\\n # outs = None\\n #if BACKEND == 'tensorflow':\\n # outs = self.call(input_lyrs)\\n # setattr(self, 'output_lyrs', outs)\\n # if BACKEND == 'tensorflow':\\n # ## Reinitial\\n # super(Model, self).__init__(\\n # inputs=input_lyrs,\\n # outputs=outs)\\n #MODEL.__init__(self, inputs=inputs, outputs=outs)\\n\\n return input_lyrs # , outs\",\n \"def build(width, height, depth, classes, stages, filters, include_top, pooling,\\n reg=1e-3, bnEps=2e-5, bnMom=0.0):\\n inputShape = (height, width, depth)\\n chanDim = -1\\n\\n if K.image_data_format() == \\\"channels_first\\\":\\n inputShape = (depth, height, width)\\n chanDim = 1\\n\\n inputs = Input(shape=inputShape)\\n\\n\\n # block 1 (initial conv block)\\n x = ZeroPadding2D(padding=((3, 3), (3, 3)), name='conv1_pad')(inputs)\\n x = Conv2D(64, (7,7), use_bias=False, strides=(2,2),\\n kernel_initializer=\\\"he_normal\\\", kernel_regularizer=l2(reg))(x)\\n x = BatchNormalization(axis=chanDim, name=\\\"bn_conv1\\\")(x)\\n x = Activation(\\\"relu\\\")(x)\\n x = ZeroPadding2D(padding=((1,1), (1,1)), name=\\\"pool1_pad\\\")(x)\\n x = MaxPooling2D(3, strides=2)(x)\\n\\n for i in range(0, len(stages)):\\n stride = (1,1) if i == 0 else (2,2) # block 2 (projection block) w stride(1,1)\\n\\n print(\\\"Stage {}, Stride={}\\\".format(i, stride))\\n x = SEResNet.residual_module(x, filters[i+1], stride,\\n chanDim=chanDim, red=True, bnEps=bnEps, bnMom=bnMom)\\n for j in range(0, stages[i] + 1): #stacking res block to each depth layer\\n x = SEResNet.residual_module(x, filters[i+1], stride=(1,1),\\n chanDim=chanDim, bnEps=bnEps,\\n bnMom=bnMom)\\n x = BatchNormalization(axis=chanDim, epsilon=bnEps,\\n momentum=bnMom)(x)\\n x = Activation(\\\"relu\\\")(x)\\n\\n if include_top:\\n x = GlobalAveragePooling2D()(x)\\n x = Dense(classes, use_bias=False, kernel_regularizer=l2(reg),\\n activation='softmax')(x)\\n else:\\n if pooling == 'avg':\\n print(\\\"Adding average pool\\\")\\n x = GlobalAveragePooling2D()(x)\\n elif pooling == 'max':\\n x = GlobalMaxPooling2D()(x)\\n\\n model = Model(inputs=inputs, outputs=x, name=\\\"SEResNet\\\")\\n return model\",\n \"def convolutional_block(input_tensor, kernel_size, filters, stage, block, strides = (2, 2)):\\n chanDim = 3\\n eps = 1.1e-5\\n \\n if K.image_data_format() == \\\"channels_first\\\":\\n inputShape = (depth, height, width)\\n chanDim = 1\\n\\n # defining name basis\\n conv_name_base = 'res152' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n # Retrieve Filters\\n filters1, filters2, filters3 = filters\\n \\n ##### MAIN PATH #####\\n # First component of main path \\n X = Conv2D(filters = filters1, kernel_size = (1, 1), strides = strides, name = conv_name_base + '2a')(input_tensor)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2a')(X)\\n X = Activation('relu', name = conv_name_base + '2a_relu')(X)\\n \\n # Second component of main path \\n X = ZeroPadding2D(padding = (1, 1), name = conv_name_base + '2b_zeropadding')(X)\\n X = Conv2D(filters = filters2, kernel_size = (kernel_size, kernel_size), name = conv_name_base + '2b')(X)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2b')(X)\\n X = Activation('relu', name = conv_name_base + \\\"2b_relu\\\")(X)\\n \\n # Third component of main path \\n X = Conv2D(filters = filters3, kernel_size = (1, 1), name = conv_name_base + '2c')(X)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2c')(X)\\n \\n ##### SHORTCUT PATH #### \\n X_shortcut = Conv2D(filters = filters3, kernel_size = (1, 1), strides = strides, name = conv_name_base + '1')(input_tensor)\\n X_shortcut = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '1')(X_shortcut)\\n \\n # Final step: Add shortcut value to main path, and pass it through a RELU activation\\n X = Add()([X, X_shortcut])\\n X = Activation('relu', name = \\\"res152\\\" + str(stage) + block + \\\"_relu\\\")(X) \\n \\n return X\",\n \"def init_layers(self):\\n\\n # get caching layers activated\\n caching_layers = G3WCachingLayer.objects.all()\\n for caching_layer in caching_layers:\\n self.add_layer(str(caching_layer), caching_layer)\",\n \"def _init_layers(self):\\n cls_branch = []\\n for _ in range(self.num_reg_fcs):\\n cls_branch.append(Linear(self.embed_dims, self.embed_dims))\\n cls_branch.append(nn.LayerNorm(self.embed_dims))\\n cls_branch.append(nn.ReLU(inplace=True))\\n cls_branch.append(Linear(self.embed_dims, self.cls_out_channels))\\n fc_cls = nn.Sequential(*cls_branch)\\n\\n reg_branch = []\\n for _ in range(self.num_reg_fcs):\\n reg_branch.append(Linear(self.embed_dims, self.embed_dims))\\n reg_branch.append(nn.ReLU())\\n reg_branch.append(Linear(self.embed_dims, self.code_size))\\n reg_branch = nn.Sequential(*reg_branch)\\n\\n past_traj_reg_branch = []\\n for _ in range(self.num_reg_fcs):\\n past_traj_reg_branch.append(\\n Linear(self.embed_dims, self.embed_dims))\\n past_traj_reg_branch.append(nn.ReLU())\\n past_traj_reg_branch.append(\\n Linear(self.embed_dims, (self.past_steps + self.fut_steps)*2))\\n past_traj_reg_branch = nn.Sequential(*past_traj_reg_branch)\\n\\n def _get_clones(module, N):\\n return nn.ModuleList([copy.deepcopy(module) for i in range(N)])\\n\\n # last reg_branch is used to generate proposal from\\n # encode feature map when as_two_stage is True.\\n num_pred = (self.transformer.decoder.num_layers + 1) if \\\\\\n self.as_two_stage else self.transformer.decoder.num_layers\\n\\n if self.with_box_refine:\\n self.cls_branches = _get_clones(fc_cls, num_pred)\\n self.reg_branches = _get_clones(reg_branch, num_pred)\\n self.past_traj_reg_branches = _get_clones(\\n past_traj_reg_branch, num_pred)\\n else:\\n self.cls_branches = nn.ModuleList(\\n [fc_cls for _ in range(num_pred)])\\n self.reg_branches = nn.ModuleList(\\n [reg_branch for _ in range(num_pred)])\\n self.past_traj_reg_branches = nn.ModuleList(\\n [past_traj_reg_branch for _ in range(num_pred)])\\n if not self.as_two_stage:\\n self.bev_embedding = nn.Embedding(\\n self.bev_h * self.bev_w, self.embed_dims)\",\n \"def __call__(\\n self, inputs, training, get_all_layers=None, skip_final_dense=False,\\n skip_final_dense_with_pool=False, final_pooling=None):\\n global ENDING_POINTS\\n ENDING_POINTS = []\\n\\n with self._model_variable_scope():\\n inputs = self._preprocess_data(inputs)\\n\\n inputs = conv3d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, \\n kernel_size=self.kernel_size, time_kernel_size=self.time_kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = tf.identity(inputs, 'initial_conv')\\n\\n # We do not include batch normalization or activation functions in V2\\n # for the initial conv1 because the first ResNet unit will perform these\\n # for both the shortcut and non-shortcut paths as part of the first\\n # block's projection. Cf. Appendix of [2].\\n if self.resnet_version == 1:\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n\\n _fp_size = self.first_pool_size\\n _fp_stride = self.first_pool_stride\\n if self.first_pool_size:\\n inputs = tf.layers.max_pooling3d(\\n inputs=inputs, \\n pool_size=[_fp_size, _fp_size, _fp_size],\\n strides=[_fp_stride, _fp_stride, _fp_stride], \\n padding='SAME',\\n data_format=self.data_format)\\n inputs = tf.identity(inputs, 'initial_max_pool')\\n\\n ENDING_POINTS.append(inputs)\\n\\n for i, num_blocks in enumerate(self.block_sizes):\\n num_filters = self.num_filters * (2**i)\\n inputs = block_layer(\\n inputs=inputs, filters=num_filters, bottleneck=self.bottleneck,\\n block_fn=self.block_fn, blocks=num_blocks,\\n strides=self.block_strides[i], training=training,\\n name='block_layer{}'.format(i + 1), data_format=self.data_format)\\n curr_res_name = 'res{}'.format(i + 1)\\n\\n # Only apply the BN and ReLU for model that does pre_activation in each\\n # building/bottleneck block, eg resnet V2.\\n if self.pre_activation:\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n ENDING_POINTS.append(inputs)\\n\\n if skip_final_dense:\\n bs = inputs.get_shape().as_list()[0]\\n # Along the temporal dimension\\n inputs = tf.reduce_mean(inputs, axis=2)\\n if final_pooling is None:\\n inputs = tf.reshape(inputs, [bs, -1])\\n return inputs\\n if final_pooling == 1:\\n axes = [2, 3] if self.data_format == 'channels_first' else [1, 2]\\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\\n return tf.reshape(inputs, [bs, 1 * 1 * self.final_size])\\n \\n # The current top layer has shape\\n # `batch_size x pool_size x pool_size x final_size`.\\n # ResNet does an Average Pooling layer over pool_size,\\n # but that is the same as doing a reduce_mean. We do a reduce_mean\\n # here because it performs better than AveragePooling2D.\\n axes = [2, 3, 4] if self.data_format == 'channels_first' else [1, 2, 3]\\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\\n inputs = tf.identity(inputs, 'final_reduce_mean')\\n\\n if skip_final_dense_with_pool:\\n return inputs\\n\\n inputs = self._get_final_dense(inputs)\\n if not get_all_layers:\\n return inputs\\n else:\\n return inputs, ENDING_POINTS\",\n \"def make_feature_layers(self, config):\\n raise NotImplementedError\",\n \"def add_layer_data(prj_id, expl_id, exploration):\\n layers = tbl_explsoilboring[\\n (tbl_explsoilboring.lng_KeyProject == prj_id) &\\n (tbl_explsoilboring.txt_KeyExplorationName == expl_id)]\\n\\n # add new column with layer numbers\\n layers['layer'] = [i for i in range(1, len(layers.index) + 1)]\\n\\n # get SPT data\\n spt_data = tbl_explboring[\\n (tbl_explboring.lng_KeyProject == prj_id) &\\n (tbl_explboring.txt_KeyExplorationName == expl_id)]\\n\\n if not spt_data.empty and not layers.empty:\\n spt_depths = spt_data.dbl_KeyDepth\\n layer_i = []\\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\\n for d in spt_depths:\\n if d <= max_d:\\n layr = layers[\\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\\n else:\\n layr = None\\n layer_i.append(layr)\\n spt_data['layer'] = layer_i\\n avg_spt = spt_data[\\n ['layer', 'dbl_FieldBlowCount']].groupby('layer').agg('mean')\\n else:\\n avg_spt = pd.DataFrame()\\n\\n # get lab data\\n lab_data = tbl_expllab[\\n (tbl_expllab.lng_KeyProject == prj_id) &\\n (tbl_expllab.txt_KeyExplorationName == expl_id)]\\n\\n if not lab_data.empty and not layers.empty:\\n lab_depths = lab_data.dbl_KeyDepth\\n layer_i = []\\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\\n for d in lab_depths:\\n if d <= max_d:\\n layr = layers[\\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\\n else:\\n layr = None\\n layer_i.append(layr)\\n lab_data['layer'] = layer_i\\n avg_lab = lab_data[\\n ['layer', 'dbl_MoistureContent', 'dbl_TotalUnitWeight',\\n 'dbl_LiquidLimit', 'dbl_PlasticityIndex', 'dbl_Cohesion']\\n ].groupby('layer').agg('mean')\\n else:\\n avg_lab = pd.DataFrame()\\n\\n for i in layers.index:\\n index = int(layers.layer[i])\\n if index == 1:\\n height = tbl_explsoilboring['dbl_KeyDepthToBottom'][i] * 0.00328084\\n else:\\n height = (tbl_explsoilboring['dbl_KeyDepthToBottom'][i]\\n - tbl_explsoilboring['dbl_KeyDepthToBottom'][i-1]) \\\\\\n * 0.00328084\\n soil_type = tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i] \\\\\\n if not pd.isna(\\n tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i]) else None\\n soil_type = primary_explsoil[soil_type] if soil_type else None\\n uscs_type = tbl_explsoilboring['txt_USCSCode'][i] if not \\\\\\n pd.isna(tbl_explsoilboring['txt_USCSCode'][i]) else None\\n if not soil_type and uscs_type:\\n soil_type = uscs_type\\n description = tbl_explsoilboring['txt_LayerDescription'][i] \\\\\\n if not pd.isna(tbl_explsoilboring['txt_LayerDescription'][i]) \\\\\\n else None\\n nval = tbl_explsoilboring['dbl_BlowCountInterpreted'][i] \\\\\\n if not pd.isna(tbl_explsoilboring['dbl_BlowCountInterpreted'][i]) \\\\\\n else None\\n nval_from_avg_spt = avg_spt[avg_spt.index == index]\\n if not nval and not nval_from_avg_spt.empty:\\n nval = nval_from_avg_spt.values[0][0] if not pd.isna(\\n nval_from_avg_spt.values[0][0]) else None\\n angle = tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i] \\\\\\n if not pd.isna(\\n tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i]) else None\\n data_from_avg_lab = avg_lab[avg_lab.index == index]\\n cohesion = tbl_explsoilboring['dbl_CohesionInterpreted'][i] \\\\\\n * 2.0885434273e-05 if not pd.isna(\\n tbl_explsoilboring['dbl_CohesionInterpreted'][i]) else None\\n if not cohesion and not data_from_avg_lab.empty:\\n cohesion = data_from_avg_lab.values[0][4] * 2.0885434273e-05 \\\\\\n if not pd.isna(data_from_avg_lab.values[0][4]) else None\\n if cohesion:\\n cohesion = cohesion if cohesion < 99 else None\\n tuw = tbl_explsoilboring['dbl_UnitWeightInterpreted'][i] * 0.062428 \\\\\\n if not pd.isna(tbl_explsoilboring['dbl_UnitWeightInterpreted'][i]) \\\\\\n else None\\n if not tuw and not data_from_avg_lab.empty:\\n tuw = data_from_avg_lab.values[0][1] * 0.062428 if not pd.isna(\\n data_from_avg_lab.values[0][1]) else None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][0]):\\n water_content = int(data_from_avg_lab.values[0][0])\\n else:\\n water_content = None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][2]):\\n liquid_limit = int(data_from_avg_lab.values[0][2])\\n else:\\n liquid_limit = None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][3]):\\n plasticity = int(data_from_avg_lab.values[0][3])\\n else:\\n plasticity = None\\n layer = Layers(\\n boring=exploration,\\n index=index,\\n soil_type=soil_type,\\n height=height,\\n description=description,\\n nval=nval,\\n ssuu=cohesion,\\n friction_angle=angle,\\n tuw=tuw,\\n water_content=water_content,\\n liquid_limit=liquid_limit,\\n plasticity=plasticity\\n )\\n db.session.add(layer)\",\n \"def layers(self):\\n return self['layers']\",\n \"def expandWithoutMutex(self, previousLayer):\\n previousLayerProposition = previousLayer.getPropositionLayer()\\n \\\"*** YOUR CODE HERE ***\\\"\",\n \"def __init__(self, layers) -> None:\\n super(FeaturizerPhi, self).__init__()\\n self.layers = layers\",\n \"def filesystem_layers(self):\\n pass\",\n \"def renderLayerPostProcess(*args, keepImages: bool=True, sceneName: Union[AnyStr, bool]=\\\"\\\",\\n q=True, query=True, **kwargs)->Union[None, Any]:\\n pass\",\n \"def update(self):\\r\\n\\r\\n # Update the vision frames in the system\\r\\n self._system.update()\\r\\n\\r\\n # Create blank PIL images to hold the video streams\\r\\n layered = PIL.Image.new('RGBA', (400, 400))\\r\\n stacked = PIL.Image.new('RGBA', (200, 800))\\r\\n control = PIL.Image.new('RGBA', (600, 800))\\r\\n\\r\\n focalpoint = self._system[self._appString[\\\"device\\\"].get()].focalpoint()\\r\\n # print(focalpoint)\\r\\n\\r\\n # Get each vision key and vision for the selected device\\r\\n visionList = [(visionKey, vision) for visionKey, vision in self._system[self._appString[\\\"device\\\"].get()]]\\r\\n\\r\\n # Loop through each vision in the vision list\\r\\n for i, (visionKey, vision) in enumerate(visionList):\\r\\n\\r\\n # Grab the frames from the vision when it is \\\"curr\\\"\\r\\n frameList = [frame for frameKey, frame in vision if frameKey==self._appString[\\\"frame\\\"].get()]\\r\\n\\r\\n # Loop through each frame in the frame list\\r\\n for frame in frameList:\\r\\n\\r\\n # Get the properties and turn the image into RGBA\\r\\n ratio, size = vision.properties()\\r\\n rgbFrame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGBA)\\r\\n\\r\\n # print(rgbFrame.shape)\\r\\n width, height, channels = rgbFrame.shape\\r\\n\\r\\n # Paste the images together in layered\\r\\n\\r\\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (int(400 * ratio), int(400 * ratio))))\\r\\n layered.paste(imgFrame, (int(200 * (1 - ratio)), int(200 * (1 - ratio))))\\r\\n\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 // width)), int(200 * (1 - ratio) - focalpoint[1] * (200 // height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1)), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200/width) / ratio), int(200 * (1 - ratio) - focalpoint[1] * (200/height) / ratio)))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1) / 200), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1) / 200)))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (400//width * (1- ratio))), int(200 * (1 - ratio) - focalpoint[1] * (400//height * (1 - ratio)))))\\r\\n\\r\\n # Paste the images together in stacked\\r\\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (200, 200)))\\r\\n stacked.paste(imgFrame, (0, 200 * i))\\r\\n\\r\\n # Add the stacked image to the canvas\\r\\n self._pilFrames[\\\"stacked\\\"] = PIL.ImageTk.PhotoImage(image=stacked)\\r\\n self._appCanvas[\\\"stacked\\\"].create_image(100, 0, image=self._pilFrames[\\\"stacked\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Add the layered image to the canvas\\r\\n self._pilFrames[\\\"layered\\\"] = PIL.ImageTk.PhotoImage(image=layered)\\r\\n self._appCanvas[\\\"layered\\\"].create_image(0, 0, image=self._pilFrames[\\\"layered\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Add the control image to the canvas\\r\\n imgFrame = cv2.cvtColor(self._system[self._appString[\\\"device\\\"].get()][self._appString[\\\"vision\\\"].get()][self._appString[\\\"frame\\\"].get()], cv2.COLOR_BGR2RGBA)\\r\\n control = PIL.Image.fromarray(cv2.resize(imgFrame, (600, 600)))\\r\\n self._pilFrames[\\\"control\\\"] = PIL.ImageTk.PhotoImage(image=control)\\r\\n self._appCanvas[\\\"control\\\"].create_image(100, 90, image=self._pilFrames[\\\"control\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Continue to update with a delay of 15\\r\\n self.after(15, self.update)\",\n \"def _get_layers(self) :\\n \\n return self._layers\",\n \"def layers(self):\\r\\n return self._flc.layers\",\n \"def initLayers(self):\\n self.addRow(self.currentLayerData)\\n self.selected = self.jobRow.child(0)\\n self.setSelectedFromItem(self.selected)\",\n \"def num_layers(self): # -> int:\\n ...\",\n \"def get_patches(rimage, gimage, mimage, num_patches=48, patch_size=80, patch_stride=80):\\n num_FSpatches = 16\\n num_RApatches = 32\\n rpatches = []\\n gpatches = []\\n mpatches = []\\n #R_imgs = ((rimage+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'rainy.jpg', R_imgs[0,:,:,:])\\n for i in range(int(math.sqrt(num_FSpatches))):\\n for j in range(int(math.sqrt(num_FSpatches))):\\n point_x = patch_stride*i\\n point_y = patch_stride*j\\n rpatch = rimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n #print(point_x)\\n #print(point_y)\\n #print(point_y+patch_size)\\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'patch_%d_%d.jpg'%(i,j), P_imgs[0,:,:,:])\\n #print(rpatch.shape)\\n rpatches.append(rpatch)\\n #print(np.array(rpatches).shape)\\n gpatch = gimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n gpatches.append(gpatch)\\n mpatch = mimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n mpatches.append(mpatch)\\n\\n for k in range(num_RApatches):\\n point1 = random.randint(0,240) # 116 comes from the image source size (320) - the patch dimension (80)\\n point2 = random.randint(0,240)\\n #rpatch = tf.image.crop_to_bounding_box(rimage, point1, point2, patch_size, patch_size)\\n rpatch = rimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'patch_%d.jpg'%i, P_imgs[0,:,:,:])\\n #print(rpatch.shape)\\n rpatches.append(rpatch)\\n #print(np.array(rpatches).shape)\\n gpatch = gimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n gpatches.append(gpatch)\\n mpatch = mimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n mpatches.append(mpatch)\\n\\n rpatches = np.array(rpatches)\\n rpatches = np.squeeze(rpatches)\\n #print(rpatches.shape)\\n gpatches = np.array(gpatches)\\n gpatches = np.squeeze(gpatches)\\n mpatches = np.array(mpatches)\\n mpatches = np.squeeze(mpatches)\\n #assert rpatches.get_shape().dims == [num_patches, patch_size, patch_size, 3]\\n assert rpatches.shape == (num_patches, patch_size, patch_size, 3)\\n return rpatches, gpatches, mpatches\",\n \"def _init_layers(self):\\n self._init_predictor()\\n if self.use_edge_fusion:\\n self._init_edge_module()\",\n \"def lightened_v2(inputs, is_training=True,\\n dropout_keep_prob=0.8,\\n reuse=None,\\n scope='LightenedV1'):\\n end_points = {}\\n \\n with tf.variable_scope(scope, 'LightenedV1', [inputs], reuse=reuse):\\n with slim.arg_scope([slim.batch_norm, slim.dropout],\\n is_training=is_training):\\n with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],\\n stride=1, padding='SAME'):\\n \\n endpoints = {}\\n \\n net=conv(inputs, 48, 9, stride=1, padding='VALID', scope='conv1_9x9')\\n end_points['conv1_9x9'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool1')\\n end_points['pool1'] = net\\n \\n net=conv(net, 96, 5, stride=1, padding='VALID', scope='conv2_5x5')\\n end_points['conv2_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool2')\\n end_points['pool2'] = net\\n\\n net=conv(net, 128, 5, stride=1, padding='VALID', scope='conv3_5x5')\\n end_points['conv3_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool3')\\n end_points['pool3'] = net\\n\\n net=conv(net, 192, 4, stride=1, padding='VALID', scope='conv4_4x4')\\n end_points['conv4_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool4')\\n end_points['pool4'] = net\\n\\n with tf.variable_scope('Logits'):\\n net = slim.flatten(net)\\n net = slim.dropout(net, dropout_keep_prob, is_training=is_training,\\n scope='Dropout') \\n endpoints['PreLogitsFlatten'] = net\\n \\n return net, end_points\",\n \"def layerListObject(self, parent):\\n pass\",\n \"def __init__(self, batch_env):\\n\\n dims = [84, 84]\\n nature_transform = lambda o: tf.image.rgb_to_grayscale( # pylint: disable=g-long-lambda\\n tf.image.resize_images(o, dims))\\n\\n super(WarpFrameWrapper, self).__init__(batch_env, transform_observation=(\\n nature_transform, dims, tf.float32))\",\n \"def update_layer(LayerId=None, Name=None, Shortname=None, Attributes=None, CloudWatchLogsConfiguration=None, CustomInstanceProfileArn=None, CustomJson=None, CustomSecurityGroupIds=None, Packages=None, VolumeConfigurations=None, EnableAutoHealing=None, AutoAssignElasticIps=None, AutoAssignPublicIps=None, CustomRecipes=None, InstallUpdatesOnBoot=None, UseEbsOptimizedInstances=None, LifecycleEventConfiguration=None):\\n pass\",\n \"def addExportLayerToCoreml(builder):\\n outputNames = [output.name for output in builder.spec.description.output]\\n\\n for i, outputName in enumerate(outputNames):\\n # formulas: https://github.com/ultralytics/yolov5/issues/471\\n builder.add_activation(\\n name=f\\\"sigmoid_{outputName}\\\",\\n non_linearity=\\\"SIGMOID\\\",\\n input_name=outputName,\\n output_name=f\\\"{outputName}_sigmoid\\\",\\n )\\n\\n ### Coordinates calculation ###\\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2) -> nC = 640 / strides[i]\\n builder.add_slice(\\n name=f\\\"slice_coordinates_xy_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_sliced_coordinates_xy\\\",\\n axis=\\\"width\\\",\\n start_index=0,\\n end_index=2,\\n )\\n # x,y * 2\\n builder.add_elementwise(\\n name=f\\\"multiply_xy_by_two_{outputName}\\\",\\n input_names=[f\\\"{outputName}_sliced_coordinates_xy\\\"],\\n output_name=f\\\"{outputName}_multiplied_xy_by_two\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=2,\\n )\\n # x,y * 2 - 0.5\\n builder.add_elementwise(\\n name=f\\\"subtract_0_5_from_xy_{outputName}\\\",\\n input_names=[f\\\"{outputName}_multiplied_xy_by_two\\\"],\\n output_name=f\\\"{outputName}_subtracted_0_5_from_xy\\\",\\n mode=\\\"ADD\\\",\\n alpha=-0.5,\\n )\\n grid = make_grid(featureMapDimensions[i], featureMapDimensions[i]).numpy()\\n # x,y * 2 - 0.5 + grid[i]\\n builder.add_bias(\\n name=f\\\"add_grid_from_xy_{outputName}\\\",\\n input_name=f\\\"{outputName}_subtracted_0_5_from_xy\\\",\\n output_name=f\\\"{outputName}_added_grid_xy\\\",\\n b=grid,\\n shape_bias=grid.shape,\\n )\\n # (x,y * 2 - 0.5 + grid[i]) * stride[i]\\n builder.add_elementwise(\\n name=f\\\"multiply_xy_by_stride_{outputName}\\\",\\n input_names=[f\\\"{outputName}_added_grid_xy\\\"],\\n output_name=f\\\"{outputName}_calculated_xy\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=strides[i],\\n )\\n\\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2)\\n builder.add_slice(\\n name=f\\\"slice_coordinates_wh_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_sliced_coordinates_wh\\\",\\n axis=\\\"width\\\",\\n start_index=2,\\n end_index=4,\\n )\\n # w,h * 2\\n builder.add_elementwise(\\n name=f\\\"multiply_wh_by_two_{outputName}\\\",\\n input_names=[f\\\"{outputName}_sliced_coordinates_wh\\\"],\\n output_name=f\\\"{outputName}_multiplied_wh_by_two\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=2,\\n )\\n # (w,h * 2) ** 2\\n builder.add_unary(\\n name=f\\\"power_wh_{outputName}\\\",\\n input_name=f\\\"{outputName}_multiplied_wh_by_two\\\",\\n output_name=f\\\"{outputName}_power_wh\\\",\\n mode=\\\"power\\\",\\n alpha=2,\\n )\\n # (w,h * 2) ** 2 * anchor_grid[i]\\n anchor = (\\n anchorGrid[i]\\n .expand(-1, featureMapDimensions[i], featureMapDimensions[i], -1)\\n .numpy()\\n )\\n builder.add_load_constant_nd(\\n name=f\\\"anchors_{outputName}\\\",\\n output_name=f\\\"{outputName}_anchors\\\",\\n constant_value=anchor,\\n shape=anchor.shape,\\n )\\n builder.add_elementwise(\\n name=f\\\"multiply_wh_with_achors_{outputName}\\\",\\n input_names=[f\\\"{outputName}_power_wh\\\", f\\\"{outputName}_anchors\\\"],\\n output_name=f\\\"{outputName}_calculated_wh\\\",\\n mode=\\\"MULTIPLY\\\",\\n )\\n\\n builder.add_concat_nd(\\n name=f\\\"concat_coordinates_{outputName}\\\",\\n input_names=[f\\\"{outputName}_calculated_xy\\\", f\\\"{outputName}_calculated_wh\\\"],\\n output_name=f\\\"{outputName}_raw_coordinates\\\",\\n axis=-1,\\n )\\n builder.add_scale(\\n name=f\\\"normalize_coordinates_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_coordinates\\\",\\n output_name=f\\\"{outputName}_raw_normalized_coordinates\\\",\\n W=torch.tensor([1 / 640]).numpy(),\\n b=0,\\n has_bias=False,\\n )\\n\\n ### Confidence calculation ###\\n builder.add_slice(\\n name=f\\\"slice_object_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_object_confidence\\\",\\n axis=\\\"width\\\",\\n start_index=4,\\n end_index=5,\\n )\\n builder.add_slice(\\n name=f\\\"slice_label_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_label_confidence\\\",\\n axis=\\\"width\\\",\\n start_index=5,\\n end_index=0,\\n )\\n # confidence = object_confidence * label_confidence\\n builder.add_multiply_broadcastable(\\n name=f\\\"multiply_object_label_confidence_{outputName}\\\",\\n input_names=[\\n f\\\"{outputName}_label_confidence\\\",\\n f\\\"{outputName}_object_confidence\\\",\\n ],\\n output_name=f\\\"{outputName}_raw_confidence\\\",\\n )\\n\\n # input: (1, 3, nC, nC, 85), output: (3 * nc^2, 85)\\n builder.add_flatten_to_2d(\\n name=f\\\"flatten_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_confidence\\\",\\n output_name=f\\\"{outputName}_flatten_raw_confidence\\\",\\n axis=-1,\\n )\\n builder.add_flatten_to_2d(\\n name=f\\\"flatten_coordinates_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_normalized_coordinates\\\",\\n output_name=f\\\"{outputName}_flatten_raw_coordinates\\\",\\n axis=-1,\\n )\\n\\n builder.add_concat_nd(\\n name=\\\"concat_confidence\\\",\\n input_names=[\\n f\\\"{outputName}_flatten_raw_confidence\\\" for outputName in outputNames\\n ],\\n output_name=\\\"raw_confidence\\\",\\n axis=-2,\\n )\\n builder.add_concat_nd(\\n name=\\\"concat_coordinates\\\",\\n input_names=[\\n f\\\"{outputName}_flatten_raw_coordinates\\\" for outputName in outputNames\\n ],\\n output_name=\\\"raw_coordinates\\\",\\n axis=-2,\\n )\\n\\n builder.set_output(\\n output_names=[\\\"raw_confidence\\\", \\\"raw_coordinates\\\"],\\n output_dims=[(25200, numberOfClassLabels), (25200, 4)],\\n )\",\n \"def stage(self, stage):\\n self._stage = stage\\n self._layer = Sdf.Layer.CreateAnonymous()\\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\",\n \"def __init__(self,\\n decoder_options,\\n panoptic_deeplab_options,\\n bn_layer=tf.keras.layers.BatchNormalization):\\n super(PanopticDeepLab, self).__init__(name='PanopticDeepLab')\\n\\n low_level_feature_keys = [\\n item.feature_key for item in panoptic_deeplab_options.low_level\\n ]\\n low_level_channels_project = [\\n item.channels_project for item in panoptic_deeplab_options.low_level\\n ]\\n\\n self._semantic_decoder = PanopticDeepLabSingleDecoder(\\n high_level_feature_name=decoder_options.feature_key,\\n low_level_feature_names=low_level_feature_keys,\\n low_level_channels_project=low_level_channels_project,\\n aspp_output_channels=decoder_options.aspp_channels,\\n decoder_output_channels=decoder_options.decoder_channels,\\n atrous_rates=decoder_options.atrous_rates,\\n name='semantic_decoder',\\n aspp_use_only_1x1_proj_conv=decoder_options.aspp_use_only_1x1_proj_conv,\\n decoder_conv_type=decoder_options.decoder_conv_type,\\n bn_layer=bn_layer)\\n self._semantic_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.semantic_head.head_channels,\\n panoptic_deeplab_options.semantic_head.output_channels,\\n common.PRED_SEMANTIC_LOGITS_KEY,\\n name='semantic_head',\\n conv_type=panoptic_deeplab_options.semantic_head.head_conv_type,\\n bn_layer=bn_layer)\\n\\n self._instance_decoder = None\\n self._instance_center_head = None\\n self._instance_regression_head = None\\n\\n if panoptic_deeplab_options.instance.enable:\\n if panoptic_deeplab_options.instance.low_level_override:\\n low_level_options = panoptic_deeplab_options.instance.low_level_override\\n else:\\n low_level_options = panoptic_deeplab_options.low_level\\n\\n # If instance_decoder is set, use those options; otherwise reuse the\\n # architecture as defined for the semantic decoder.\\n if panoptic_deeplab_options.instance.HasField(\\n 'instance_decoder_override'):\\n decoder_options = (panoptic_deeplab_options.instance\\n .instance_decoder_override)\\n\\n low_level_feature_keys = [item.feature_key for item in low_level_options]\\n low_level_channels_project = [\\n item.channels_project for item in low_level_options\\n ]\\n\\n self._instance_decoder = PanopticDeepLabSingleDecoder(\\n high_level_feature_name=decoder_options.feature_key,\\n low_level_feature_names=low_level_feature_keys,\\n low_level_channels_project=low_level_channels_project,\\n aspp_output_channels=decoder_options.aspp_channels,\\n decoder_output_channels=decoder_options.decoder_channels,\\n atrous_rates=decoder_options.atrous_rates,\\n name='instance_decoder',\\n aspp_use_only_1x1_proj_conv=(\\n decoder_options.aspp_use_only_1x1_proj_conv),\\n decoder_conv_type=decoder_options.decoder_conv_type,\\n bn_layer=bn_layer)\\n self._instance_center_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.instance.center_head.head_channels,\\n panoptic_deeplab_options.instance.center_head.output_channels,\\n common.PRED_CENTER_HEATMAP_KEY,\\n name='instance_center_head',\\n conv_type=(\\n panoptic_deeplab_options.instance.center_head.head_conv_type),\\n bn_layer=bn_layer)\\n self._instance_regression_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.instance.regression_head.head_channels,\\n panoptic_deeplab_options.instance.regression_head.output_channels,\\n common.PRED_OFFSET_MAP_KEY,\\n name='instance_regression_head',\\n conv_type=(\\n panoptic_deeplab_options.instance.regression_head.head_conv_type),\\n bn_layer=bn_layer)\",\n \"def __init__(self,\\n high_level_feature_name,\\n low_level_feature_names,\\n low_level_channels_project,\\n aspp_output_channels,\\n decoder_output_channels,\\n atrous_rates,\\n name,\\n aspp_use_only_1x1_proj_conv=False,\\n decoder_conv_type='depthwise_separable_conv',\\n bn_layer=tf.keras.layers.BatchNormalization):\\n super(PanopticDeepLabSingleDecoder, self).__init__(name=name)\\n self._channel_axis = 3\\n\\n self._aspp = aspp.ASPP(\\n aspp_output_channels,\\n atrous_rates,\\n aspp_use_only_1x1_proj_conv=aspp_use_only_1x1_proj_conv,\\n name='aspp',\\n bn_layer=bn_layer)\\n self._high_level_feature_name = high_level_feature_name\\n\\n if len(low_level_feature_names) != len(low_level_channels_project):\\n raise ValueError('The Panoptic-DeepLab decoder requires the same number '\\n 'of low-level features as the number of low-level '\\n 'projection channels. But got %d and %d.'\\n % (len(low_level_feature_names),\\n len(low_level_channels_project)))\\n\\n self._low_level_feature_names = low_level_feature_names\\n\\n for i, channels_project in enumerate(low_level_channels_project):\\n # Check if channel sizes increases and issue a warning.\\n if i > 0 and low_level_channels_project[i - 1] < channels_project:\\n logging.warning(\\n 'The low level projection channels usually do not '\\n 'increase for features with higher spatial resolution. '\\n 'Please make sure, this behavior is intended.')\\n current_low_level_conv_name, current_fusion_conv_name = (\\n utils.get_low_level_conv_fusion_conv_current_names(i))\\n utils.safe_setattr(\\n self, current_low_level_conv_name, convolutions.Conv2DSame(\\n channels_project,\\n kernel_size=1,\\n name=utils.get_layer_name(current_low_level_conv_name),\\n use_bias=False,\\n use_bn=True,\\n bn_layer=bn_layer,\\n activation='relu'))\\n\\n utils.safe_setattr(\\n self, current_fusion_conv_name, convolutions.StackedConv2DSame(\\n conv_type=decoder_conv_type,\\n num_layers=1,\\n output_channels=decoder_output_channels,\\n kernel_size=5,\\n name=utils.get_layer_name(current_fusion_conv_name),\\n use_bias=False,\\n use_bn=True,\\n bn_layer=bn_layer,\\n activation='relu'))\",\n \"def develop_image(pipeline, camera=None, batch=None, image=None, patch_size=0, patches=2, root_dir='./data', pipeline_args=None):\\n\\n if camera is not None:\\n supported_cameras = fsutil.listdir(os.path.join(root_dir, 'models', 'nip'), '.*')\\n if camera not in supported_cameras:\\n raise ValueError('Camera data not found ({})! Available cameras: {}'.format(camera, ', '.join(supported_cameras)))\\n root_dirname = os.path.join(root_dir, 'models', 'nip', camera)\\n data_dirname = os.path.join(root_dir, 'raw', 'training_data', camera)\\n\\n if patch_size != 0 and (patch_size < 4 or patch_size > 2048):\\n raise ValueError('Patch size seems to be invalid!')\\n\\n # Lazy imports to minimize delay for invalid command line parameters\\n import numpy as np\\n import imageio as io\\n import matplotlib.pyplot as plt\\n import tensorflow as tf\\n from models import pipelines\\n\\n # Construct the NIP model ---------------------------------------------------------------------\\n \\n if os.path.isdir(pipeline):\\n # Restore a NIP model from a training log\\n model = tfmodel.restore(pipeline, pipelines)\\n else:\\n # Construct the NIP model from class name (and optional arguments)\\n if pipeline_args is None:\\n model = getattr(pipelines, pipeline)()\\n else:\\n model = getattr(pipelines, pipeline)(**pipeline_args)\\n\\n loaded_model = False\\n candidate_dirs = [os.path.join(root_dirname, model.model_code), os.path.join(root_dirname)]\\n for candidate in candidate_dirs:\\n if os.path.isdir(candidate):\\n model.load_model(candidate)\\n loaded_model = True\\n break\\n\\n if not loaded_model:\\n raise FileNotFoundError(f'Could not find the corresponding model: {candidate_dirs}')\\n\\n # Load image(s) -------------------------------------------------------------------------------\\n \\n if image is None and batch is not None:\\n print('Loading a batch of {} images'.format(batch))\\n data = dataset.Dataset(data_dirname, n_images=0, v_images=batch, val_rgb_patch_size=patch_size or 256, val_n_patches=patches)\\n sample_x, sample_y = data.next_validation_batch(0, data.count_validation)\\n\\n with open('config/cameras.json') as f:\\n cameras = json.load(f)\\n cfa, srgb = cameras[camera]['cfa'], np.array(cameras[camera]['srgb'])\\n\\n elif image is not None:\\n print('Loading a RAW image {}'.format(image))\\n sample_x, cfa, srgb, _ = raw.unpack(image, expand=True)\\n sample_y = raw.process(image, brightness=None, expand=True)\\n\\n if isinstance(model, pipelines.ClassicISP):\\n print('Configuring ISP to CFA: {} & sRGB {}'.format(cfa, srgb.round(2).tolist()))\\n model.set_cfa_pattern(cfa)\\n model.set_srgb_conversion(srgb)\\n\\n sample_Y = model.process(sample_x).numpy()\\n\\n if patch_size > 0:\\n xx = (sample_y.shape[2] - patch_size) // 2\\n yy = (sample_y.shape[1] - patch_size) // 2\\n sample_y = sample_y[:, yy:yy+patch_size, xx:xx+patch_size, :]\\n sample_Y = sample_Y[:, yy:yy+patch_size, xx:xx+patch_size, :]\\n\\n psnrs = metrics.psnr(sample_y, sample_Y)\\n ssims = metrics.ssim(sample_y, sample_Y)\\n\\n print('sample x: {}'.format(sample_x.shape))\\n print('sample y: {}'.format(sample_y.shape))\\n print('sample Y: {}'.format(sample_Y.shape))\\n\\n # Plot images ---------------------------------------------------------------------------------\\n if len(sample_y) > 1:\\n sample_y = plots.thumbnails(sample_y, batch, True)\\n sample_Y = plots.thumbnails(sample_Y, batch, True)\\n else:\\n sample_y = sample_y.squeeze()\\n sample_Y = sample_Y.squeeze()\\n\\n print('thumbnails: {}'.format(sample_y.shape))\\n\\n ncols = 1 if sample_y.shape[1] > sample_y.shape[0] else 2\\n nrows = 2 if ncols == 1 else 1\\n fig, axes = plt.subplots(nrows, ncols)\\n\\n plots.image(sample_Y, '{}, PSNR={:.1f} dB, SSIM={:.2f} : {{}}'.format(model.model_code, float(psnrs.mean()), float(ssims.mean())), axes=axes[0])\\n plots.image(sample_y, 'Target RGB images () : {}', axes=axes[1])\\n\\n plt.show()\\n plt.close()\",\n \"def SetUpLayerManager(self):\\n pass\",\n \"def inception_network():\\n input = K.Input(shape=(224, 224, 3))\\n\\n conv_1 = K.layers.Conv2D(\\n filters=64,\\n kernel_size=7,\\n strides=2,\\n padding='same',\\n activation='relu'\\n )(input)\\n\\n pool_1 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(conv_1)\\n\\n conv_2 = K.layers.Conv2D(\\n filters=64,\\n kernel_size=1,\\n strides=1,\\n padding='valid',\\n activation='relu'\\n )(pool_1)\\n\\n conv_3 = K.layers.Conv2D(\\n filters=192,\\n kernel_size=3,\\n strides=1,\\n padding='same',\\n activation='relu'\\n )(conv_2)\\n\\n pool_2 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(conv_3)\\n\\n inception_1 = inception_block(pool_2, (64, 96, 128, 16, 32, 32))\\n inception_2 = inception_block(inception_1, (128, 128, 192, 32, 96, 64))\\n\\n pool_3 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(inception_2)\\n\\n inception_3 = inception_block(pool_3, (192, 96, 208, 16, 48, 64))\\n inception_4 = inception_block(inception_3, (160, 112, 224, 24, 64, 64))\\n inception_5 = inception_block(inception_4, (128, 128, 256, 24, 64, 64))\\n inception_6 = inception_block(inception_5, (112, 144, 288, 32, 64, 64))\\n inception_7 = inception_block(inception_6, (256, 160, 320, 32, 128, 128))\\n\\n pool_4 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same',\\n )(inception_7)\\n\\n inception_8 = inception_block(pool_4, (256, 160, 320, 32, 128, 128))\\n inception_9 = inception_block(inception_8, (384, 192, 384, 48, 128, 128))\\n\\n pool_5 = K.layers.AveragePooling2D(\\n pool_size=7,\\n strides=1,\\n padding='valid'\\n )(inception_9)\\n\\n dropout = K.layers.Dropout(rate=0.4)(pool_5)\\n linear = K.layers.Dense(units=1000, activation='softmax')(dropout)\\n\\n return K.Model(inputs=input, outputs=linear)\",\n \"def __init__(self):\\n super(GatherLastLayer, self).__init__()\",\n \"def _save_data(self, outputs, latents):\\n list(map(lambda e, v: e.append(v), self.curr_patches, outputs))\\n list(map(lambda e, v: e.append(v), self.curr_latents, latents))\\n self.count_patches += len(self.curr_patches[0][-1])\\n\\n while self.count_patches >= self.patches_ref:\\n aux_patches = list(map(lambda e: tf.concat(e, axis=0),\\n self.curr_patches))\\n aux_latents = list(map(lambda e: tf.concat(e, axis=0),\\n self.curr_latents))\\n\\n data_name = str((self.out_folder / str(Folders.RAW_DATA))\\n / self.curr_img.stem)\\n for lvl, (aux_p, aux_l) in enumerate(zip(aux_patches, aux_latents)):\\n patches, remaining = tf.split(\\n aux_p, [self.patches_ref, len(aux_p) - self.patches_ref],\\n axis=0)\\n self.curr_patches[lvl] = [remaining]\\n latent, remaining = tf.split(\\n aux_l, [self.patches_ref, len(aux_l) - self.patches_ref],\\n axis=0)\\n self.curr_latents[lvl] = [remaining]\\n\\n patch_size = len(patches[0])\\n with Image.open(self.curr_img) as img:\\n # width, height -> height, width (to be compatible with\\n # arrays)\\n img_size = np.array(img.size)[::-1]\\n p_size = np.ceil(img_size / patch_size).astype(int)\\n\\n img = tf.reshape(\\n patches, [*p_size, *self.m_cfg['configs']['patch']])\\n img = tf.transpose(img, [0, 2, 1, 3, 4])\\n img = tf.reshape(img, [*patch_size * p_size, 3])\\n img = tf.image.convert_image_dtype(img, tf.uint8)\\n img = img[:img_size[0], :img_size[1], :]\\n\\n latent_name = data_name + '_latent' + str(lvl)\\n serialized = tf.io.serialize_tensor(latent)\\n tf.io.write_file(latent_name, serialized)\\n\\n img_name = data_name + '_' + str(lvl) + self.curr_img.suffix\\n tf.io.write_file(img_name, tf.image.encode_png(\\n img, compression=0))\\n\\n with open(Path(data_name).with_suffix('.txt'), 'w') as file:\\n file.write(str(self.curr_img))\\n\\n self.count_patches = len(self.curr_patches[0][0])\\n if len(self.img_paths):\\n self.curr_img = Path(self.img_paths.pop(0))\\n self.patches_ref = self.calc_n_patches(\\n self.curr_img, self.m_cfg['configs']['patch'][0])\",\n \"def multi_run_wrapper(args):\\n\\treturn img_preprocessing(*args)\",\n \"def registerPatches(self,pl):\\n self.set('patchmesh.patches',pl)\",\n \"def _set_freeze_layers(self):\\n for layer in self.encoder.layers[:self.freeze_layers]:\\n layer.trainable = False\",\n \"def __call__(self, inputs, training):\\n\\n with self._model_variable_scope():\\n if self.data_format == 'channels_first':\\n # Convert the inputs from channels_last (NHWC) to channels_first (NCHW).\\n # This provides a large performance boost on GPU. See\\n # https://www.tensorflow.org/performance/performance_guide#data_formats\\n inputs = tf.transpose(a=inputs, perm=[0, 3, 1, 2])\\n\\n with tf.variable_scope('stem'):\\n inputs = conv2d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n inputs = tf.identity(inputs, 'stem_conv1')\\n\\n inputs = conv2d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n inputs = tf.identity(inputs, 'stem_conv2')\\n\\n # stage 1\\n with tf.variable_scope('stage1'):\\n stage1_out = block_layer(\\n inputs=inputs, inp_filters=self.num_filters,\\n filters=self.num_filters, bottleneck=True,\\n block_fn=_bottleneck_block_v1, blocks=self.block_sizes[0],\\n strides=1, training=training,\\n name='stage1', data_format=self.data_format)\\n\\n pre_stage_channels = self.num_filters * 4\\n stagex_input = [stage1_out]\\n num_inchannels = [pre_stage_channels]\\n\\n # stage 2 to last stage\\n for i, num_modules in enumerate(self.module_sizes[1:], 2):\\n with tf.variable_scope('stage{}'.format(i)):\\n num_channels = [self.hrnet_size * 2 ** n for n in range(i)]\\n transn = Transition(\\n stagex_input[-1], num_outchannels=num_channels[-1], \\n training=training, data_format=self.data_format)\\n stagex_input.append(transn)\\n num_inchannels.append(num_inchannels[-1])\\n\\n for j in range(num_modules):\\n if i == len(self.module_sizes) and j == num_modules - 1:\\n multi_scale_output = False\\n else:\\n multi_scale_output = True\\n\\n stagex_input = HighResolutionModule(\\n stagex_input, num_branches=i, \\n num_inchannels=num_inchannels, num_channels=num_channels,\\n bottleneck=self.bottleneck, block_fn=self.block_fn,\\n num_blocks=self.block_sizes[i - 1], training=training,\\n name='stage{}_block{}'.format(i, j), data_format=self.data_format,\\n multi_scale_output=multi_scale_output)\\n num_inchannels = num_channels\\n\\n return stagex_input\",\n \"def __call__(self, **kwargs):\\n segname = 'block_{}_expand_relu'\\n blocks = [13, 6, 3, 1]\\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\\n backbone_out = self._backbone.get_layer('block_16_project')\\n\\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\\n p4 = self._fpn_block(p5, skips[1].output)\\n p3 = self._fpn_block(p4, skips[2].output)\\n p2 = self._fpn_block(p3, skips[3].output)\\n\\n s5 = self._conv_block(p5, 128)\\n s4 = self._conv_block(p4, 128)\\n s3 = self._conv_block(p3, 128)\\n s2 = self._conv_block(p2, 128)\\n\\n s5 = tf.keras.layers.UpSampling2D(\\n size=(8, 8),\\n interpolation='nearest'\\n )(s5)\\n\\n s4 = tf.keras.layers.UpSampling2D(\\n size=(4, 4),\\n interpolation='nearest'\\n )(s4)\\n\\n s3 = tf.keras.layers.UpSampling2D(\\n size=(2, 2),\\n interpolation='nearest'\\n )(s3)\\n\\n concat = [s5, s4, s3, s2]\\n x = tf.keras.layers.Concatenate()(concat)\\n x = tf.keras.layers.Conv2D(\\n 64,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n x = tf.keras.layers.BatchNormalization()(x)\\n x = tf.keras.layers.Activation('relu')(x)\\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\\n\\n x = tf.keras.layers.Conv2D(\\n 1,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n out = tf.keras.layers.Activation('sigmoid')(x)\\n model = tf.keras.models.Model(\\n inputs=self._backbone.input,\\n outputs=out\\n )\\n\\n return model\",\n \"def _augment_pipeline_cfg(self):\",\n \"def add_bluprint_layers(self, models_dict, graph, class_names, dimensions,\\n show_activation, show_constant_input):\\n top_model = models_dict['top-model']\\n # Get the top-model color list containing\\n # the base color and the layer shades\\n top_model_color = top_model[1]\\n # get the model\\n top_model = top_model[0]\\n\\n # Get the layers of the model\\n layers = top_model[\\\"config\\\"][\\\"layers\\\"]\\n # Loop through the layers\\n for layer in layers:\\n # If the layer is not a model\\n if layer[\\\"class_name\\\"] != \\\"Model\\\":\\n # Get the layer name\\n layer_name = layer[\\\"name\\\"]\\n # If label only layer's class name\\n if class_names:\\n # Get the layer's information\\n layer_info = self.viz_utils.query_layer(layer_name,\\n models_dict)\\n # Get the layer's class name\\n layer_class = layer_info['class_name']\\n # If the layer is a a constant input layer,\\n # manually specify the class name\\n if layer_name.find('constant_input') != -1:\\n layer_class = 'Constant Input'\\n # Depending on the class name\\n # find the the layer shade\\n # If the layer is a constant_input layer\\n # the color is black\\n model_color = top_model_color[1].get(layer_class, \\\"black\\\")\\n else:\\n # If don't use class names for layers\\n # then use the layer name from the JSON\\n layer_class = layer_name\\n model_color = top_model_color[0]\\n\\n\\n # Add the node to the graph\\n graph = self.viz_utils.add_nodes(layer_name, graph,\\n layer_class, model_color,\\n dimensions, show_constant_input)\\n\\n # Add Blueprint Inbound Edges\\n graph = self.connect_blueprint_inbounds(models_dict,\\n layer, graph,\\n class_names, dimensions,\\n show_activation, show_constant_input)\\n else:\\n # Add Softmod\\n graph = self.connect_softmod_in_blueprint(models_dict,\\n layer, graph, class_names,\\n dimensions, show_activation, show_constant_input)\\n\\n return graph\",\n \"def intermediate_layer(layer, filters, kernel_size):\\n layer = Conv2D(filters=filters, kernel_size=kernel_size,\\n strides=(2, 2), padding=\\\"same\\\")(layer)\\n layer = LeakyReLU(alpha=LEAKY_RELU_ALPHA)(layer)\\n return layer\",\n \"def load_registration_as_layers() -> None:\\n viewer = getattr(widget, \\\"viewer\\\").value\\n registration_directory = pathlib.Path(\\n getattr(widget, \\\"registration_output_folder\\\").value\\n )\\n add_registered_image_layers(\\n viewer, registration_directory=registration_directory\\n )\",\n \"def deep_dream_of_extreme_control(FLAGS,model,input_images=[],num_iterations=10,step_size=0.1):\\n if len(input_images) == 0:\\n # use predefined images\\n img_dir='/esat/opal/kkelchte/docker_home/pilot_data/visualization_images'\\n input_images=sorted([img_dir+'/'+f for f in os.listdir(img_dir)])\\n\\n print(\\\"[tools.py]: extracting deep dream maps of {0} in {1}\\\".format([os.path.basename(i) for i in input_images], os.path.dirname(input_images[0])))\\n \\n # experts=np.asarray([[k]*(FLAGS.action_quantity if FLAGS.discrete else 1) for v in sorted(model.factor_offsets.values()) for k in model.factor_offsets.keys() if model.factor_offsets[k]==v]).flatten()\\n\\n inputs = load_images(input_images, model.input_size[1:])\\n \\n # collect gradients for output endpoint of evaluation model\\n grads={}\\n with tf.device('/cpu:0'):\\n output_tensor = model.endpoints['eval']['outputs']\\n for i in range(output_tensor.shape[1].value):\\n layer_loss = output_tensor[:,i]\\n gradients = tf.gradients(layer_loss, model.inputs)[0]\\n gradients /= (tf.sqrt(tf.reduce_mean(tf.square(gradients))) + 1e-5)\\n grads[output_tensor.name+'_'+str(i)]=gradients\\n\\n\\n # apply gradient ascent for all outputs and each input image\\n # if number of outputs ==1 apply gradient descent for contrast\\n if len(grads.keys())== 1:\\n opposite_results={}\\n else:\\n opposite_results=None\\n\\n import copy\\n results = {}\\n for gk in grads.keys(): \\n results[gk]=copy.deepcopy(inputs)\\n if isinstance(opposite_results,dict): opposite_results[gk]=copy.deepcopy(inputs)\\n\\n for step in range(num_iterations):\\n if step%10==0: print \\\"{0} step: {1}\\\".format(time.ctime(), step)\\n for i,gk in enumerate(sorted(grads.keys())):\\n results[gk] += step_size * model.sess.run(grads[gk], {model.inputs: results[gk]})\\n if isinstance(opposite_results,dict):\\n opposite_results[gk] -= step_size * model.sess.run(grads[gk], {model.inputs: opposite_results[gk]})\\n\\n # Normalize results within 0:1 range\\n clean_results={}\\n for gk in results.keys():\\n clean_results[gk]=[]\\n for i in range(results[gk].shape[0]):\\n clean_results[gk].append(deprocess_image(results[gk][i], one_channel=True))\\n # results[gk][i]=deprocess_image(results[gk][i], one_channel=True)\\n if isinstance(opposite_results,dict):\\n opposite_results[gk][i]=deprocess_image(opposite_results[gk][i])\\n\\n # combine adjust input images in one overview image\\n # one column for each input image\\n # one row with each extreme control for separate and difference images\\n num_rows=1+len(results.keys())\\n fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\\n # fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\\n # add original images in first row\\n for i in range(axes.shape[1]):\\n axes[0, i].set_title(os.path.basename(input_images[i]).split('.')[0])\\n axes[0, i].imshow(matplotlibprove(inputs[i]), cmap='inferno')\\n axes[0, i].axis('off')\\n\\n # add for each filter the modified input\\n row_index=1\\n for gk in sorted(results.keys()):\\n for i in range(axes.shape[1]):\\n # print gk\\n # axes[row_index, i].set_title('Grad Asc: '+gk.split('/')[1]+'/'+gk[-1]) \\n axes[row_index, i].set_title('Grad Asc: '+gk)\\n # axes[row_index, i].set_title(experts[row_index-1])\\n\\n axes[row_index, i].imshow(np.concatenate((inputs[i],np.expand_dims(clean_results[gk][i],axis=2)), axis=2), cmap='inferno')\\n # axes[row_index, i].imshow(matplotlibprove(results[gk][i]), cmap='inferno')\\n axes[row_index, i].axis('off')\\n row_index+=1\\n # In cas of continouos controls: visualize the gradient descent and difference\\n # if isinstance(opposite_results,dict):\\n # for gk in opposite_results.keys():\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Grad Desc: '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Grad Desc: '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(opposite_results[gk][i]), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n \\n # # add difference\\n # for gk in opposite_results.keys():\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Diff: '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Diff: '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((opposite_results[gk][i]-results[gk][i])**2)), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n # else:\\n # # add difference between 2 exteme actions\\n # gk_left=sorted(results.keys())[0]\\n # gk_right=sorted(results.keys())[-1]\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Diff : '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Diff : '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((results[gk_left][i]-results[gk_right][i])**2)), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n \\n \\n plt.savefig(FLAGS.summary_dir+FLAGS.log_tag+'/control_dream_maps.jpg',bbox_inches='tight')\\n # plt.show()\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n TransformerDecoderBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n multi_channel_cross_attention=self.multi_channel_cross_attention,\\n name=(\\\"layer_%d\\\" % i)))\\n super(TransformerDecoder, self).build(unused_input_shapes)\",\n \"def __init__(self, layer_list_info):\\n self.layer_list_info = layer_list_info\",\n \"def GetShapesOfLayer(self, *args):\\n return _XCAFDoc.XCAFDoc_LayerTool_GetShapesOfLayer(self, *args)\",\n \"def depth_rendering(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\\n\\n x_indices = np.arange(lf_size[1])\\n y_indices = np.arange(lf_size[2])\\n b_indices = np.arange(lf_size[0])\\n\\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\\n\\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\\n stacked = np.stack(grid, 2)\\n positions = stacked.reshape(-1, 2)\\n\\n # Compute the distance from each lf position from the reference view\\n # Repeat the elements of this to match the size of the disparity map\\n ref_pos = np.array(\\n [lf_one_way // 2, lf_one_way // 2])\\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\\n\\n\\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\\n\\n # Compute the shifted pixels\\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\\n\\n #indices for linear interpolation in a square around the central point\\n x_low = np.around(x_shifted).astype(int)\\n #x_high = x_low + 1\\n\\n y_low = np.around(y_shifted).astype(int)\\n #y_high = y_low + 1\\n\\n #Place co-ordinates outside the image back into the image\\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\\n #x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\\n #y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\\n\\n #Gather the interpolation points\\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\\n #interp_pts_2 = np.stack((x_low_clip, y_high_clip))\\n #interp_pts_3 = np.stack((x_high_clip, y_low_clip))\\n #interp_pts_4 = np.stack((x_high_clip, y_high_clip))\\n\\n #Index into the images\\n desired_shape = lf_size\\n res_1 = torch_big_sample(ref_view, interp_pts_1, desired_shape)\\n return res_1\\n res_2 = torch_big_sample(ref_view, interp_pts_2, desired_shape)\\n res_3 = torch_big_sample(ref_view, interp_pts_3, desired_shape)\\n res_4 = torch_big_sample(ref_view, interp_pts_4, desired_shape)\\n\\n #Compute interpolation weights\\n x_low_f = x_low.astype(np.float32)\\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\\n d_x_high = 1.0 - d_x_low\\n y_low_f = y_low.astype(np.float32)\\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\\n d_y_high = 1.0 - d_y_low\\n\\n w1 = torch.from_numpy(d_x_low * d_y_low)\\n w2 = torch.from_numpy(d_x_low * d_y_high)\\n w3 = torch.from_numpy(d_x_high * d_y_low)\\n w4 = torch.from_numpy(d_x_high * d_y_high)\\n\\n #THEY AGREE AT THIS POINT\\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape), res_1)\\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape), res_2)\\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape), res_3)\\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape), res_4)\\n\\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\\n torch.add(novel_view, weighted_4, out=novel_view)\\n return novel_view\",\n \"def process_layer(layer_data):\\n layer_name = layer_data['name']\\n\\n # Instantiate layer.\\n from tensorflow.python.keras._impl.keras.layers import deserialize as deserialize_layer # pylint: disable=g-import-not-at-top\\n\\n layer = deserialize_layer(layer_data, custom_objects=custom_objects)\\n created_layers[layer_name] = layer\\n\\n # Gather layer inputs.\\n inbound_nodes_data = layer_data['inbound_nodes']\\n for node_data in inbound_nodes_data:\\n # We don't process nodes (i.e. make layer calls)\\n # on the fly because the inbound node may not yet exist,\\n # in case of layer shared at different topological depths\\n # (e.g. a model such as A(B(A(B(x)))))\\n add_unprocessed_node(layer, node_data)\",\n \"def __init__(self, stage, layer=None, proxyPrimPrefix=\\\"\\\", renderPrimPrefix=\\\"\\\"):\\n\\n self._stage = stage\\n\\n # Layer\\n if layer is None:\\n self._layer = Sdf.Layer.CreateAnonymous()\\n elif isinstance(layer, Sdf.Layer):\\n self._layer = layer\\n elif isinstance(layer, str):\\n self._layer = Sdf.Layer.Open(layer)\\n\\n self._layer.SetMuted(True)\\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\\n self._proxyPrimPrefix = proxyPrimPrefix\\n self._renderPrimPrefix = renderPrimPrefix\",\n \"def __reslayer(self, inputs, in_filters, out_filters, stride=1):\\n with tf.variable_scope('sub1'):\\n kernel = tf.get_variable('weights', [3, 3, in_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n conv = tf.nn.conv2d(inputs, kernel, [1, stride, stride, 1],\\n padding='SAME',\\n name='conv')\\n batch_norm = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\\n conv = tf.nn.elu(batch_norm, 'elu')\\n\\n with tf.variable_scope('sub2'):\\n kernel = tf.get_variable('weights',\\n [3, 3, out_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n conv = tf.nn.conv2d(conv, kernel, [1, 1, 1, 1], padding='SAME',\\n name='conv1')\\n bias = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\\n\\n with tf.variable_scope('subadd'):\\n if in_filters != out_filters:\\n kernel = tf.get_variable('weights', [1, 1, in_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n inputs = tf.nn.conv2d(\\n inputs, kernel, [1, stride, stride, 1], padding='SAME')\\n bias += inputs\\n conv = tf.nn.elu(bias, 'elu')\\n\\n num = np.power(2, np.floor(np.log2(out_filters) / 2))\\n\\n grid = self.__put_activations_on_grid(conv, (int(num),\\n int(out_filters /\\n num)))\\n tf.summary.image('sub2/activations', grid, max_outputs=1)\\n\\n return conv\",\n \"def depth_rendering_gpu(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\\n\\n x_indices = np.arange(lf_size[1])\\n y_indices = np.arange(lf_size[2])\\n b_indices = np.arange(lf_size[0])\\n\\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\\n\\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\\n stacked = np.stack(grid, 2)\\n positions = stacked.reshape(-1, 2)\\n\\n # Compute the distance from each lf position from the reference view\\n # Repeat the elements of this to match the size of the disparity map\\n ref_pos = np.array(\\n [lf_one_way // 2, lf_one_way // 2])\\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\\n\\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\\n\\n # Compute the shifted pixels\\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\\n\\n #indices for linear interpolation in a square around the central point\\n x_low = np.floor(x_shifted).astype(int)\\n x_high = x_low + 1\\n\\n y_low = np.floor(y_shifted).astype(int)\\n y_high = y_low + 1\\n\\n #Place co-ordinates outside the image back into the image\\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\\n x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\\n y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\\n\\n #Gather the interpolation points\\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\\n interp_pts_2 = np.stack((x_low_clip, y_high_clip))\\n interp_pts_3 = np.stack((x_high_clip, y_low_clip))\\n interp_pts_4 = np.stack((x_high_clip, y_high_clip))\\n\\n #Index into the images\\n desired_shape = lf_size\\n ref_view = torch.tensor(ref_view, dtype=torch.float32).cuda()\\n res_1 = torch_tensor_sample(ref_view, interp_pts_1, desired_shape)\\n res_2 = torch_tensor_sample(ref_view, interp_pts_2, desired_shape)\\n res_3 = torch_tensor_sample(ref_view, interp_pts_3, desired_shape)\\n res_4 = torch_tensor_sample(ref_view, interp_pts_4, desired_shape)\\n\\n #Compute interpolation weights\\n x_low_f = x_low.astype(np.float32)\\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\\n d_x_high = 1.0 - d_x_low\\n y_low_f = y_low.astype(np.float32)\\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\\n d_y_high = 1.0 - d_y_low\\n\\n w1 = torch.from_numpy(d_x_low * d_y_low)\\n w2 = torch.from_numpy(d_x_low * d_y_high)\\n w3 = torch.from_numpy(d_x_high * d_y_low)\\n w4 = torch.from_numpy(d_x_high * d_y_high)\\n\\n #THEY AGREE AT THIS POINT\\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape).cuda(), res_1)\\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape).cuda(), res_2)\\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape).cuda(), res_3)\\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape).cuda(), res_4)\\n\\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\\n torch.add(novel_view, weighted_4, out=novel_view)\\n return novel_view\",\n \"def __call__(self, inputs, *args, **kwargs):\\n out = inputs\\n for i, layer in enumerate(self.layers):\\n if i == 0:\\n out = layer(out, *args, **kwargs)\\n else:\\n out = layer(out)\\n return out\",\n \"def add_convolution_layers(self):\\n with tf.variable_scope(\\\"conv_layer_1\\\"):\\n # our input to this layer is from word embeddings. which would be of\\n # shape (batch_size, number of words in the sentence, dimension of\\n # embedding). For us dimension of embedding is the number of input\\n # channels, self.out_channels_1 is the output channels.\\n\\n # First we have to define a filter. The shape of the filter would be [\\n # self.filter_size_1, dimension of embedding, self.out_channels_1]\\n filter1 = tf.get_variable(\\\"filter1\\\", dtype=tf.float32,\\n shape=[self.filter_size_1,\\n self.word_embeddings.shape[2],\\n self.out_channels_1])\\n conv_layer_1 = tf.nn.conv1d(self.word_embeddings, filters=filter1,\\n stride=self.strides_1, padding=\\\"SAME\\\")\\n conv_layer_1 = tf.nn.tanh(conv_layer_1)\\n # now add pooling layer\\n print(\\\"adding pooling to layer1\\\")\\n\\n if self.max_pooling:\\n print(\\\"Adding max pooling layer\\\")\\n pooling_layer = tf.layers.max_pooling1d\\n else:\\n print(\\\"Adding average pooling layer\\\")\\n pooling_layer = tf.layers.average_pooling1d\\n conv_layer_1 = pooling_layer(conv_layer_1,\\n pool_size=self.pool_size_1,\\n strides=self.strides_1,\\n padding=\\\"SAME\\\")\\n\\n with tf.variable_scope(\\\"conv_layer_2\\\"):\\n print(\\\"adding convolution layer 2\\\")\\n # Input to this layer is conv_layer_1 from above.\\n # get another filter now. Shape of this filter would be [\\n # self.filter_size_2, self.out_channels_1, self.out_channels_2]\\n filter2 = tf.get_variable(\\\"filter2\\\", dtype=tf.float32,\\n shape=[self.filter_size_2,\\n self.out_channels_1,\\n self.out_channels_2])\\n conv_layer_2 = tf.nn.conv1d(conv_layer_1, filters=filter2,\\n stride=self.strides_2, padding=\\\"SAME\\\")\\n conv_layer_2 = tf.nn.tanh(conv_layer_2)\\n print(\\\"adding pooling to layer2\\\")\\n conv_layer_2 = pooling_layer(conv_layer_2,\\n pool_size=self.pool_size_2,\\n strides=self.strides_2,\\n padding=\\\"SAME\\\")\\n\\n self.cnn_layer = tf.nn.dropout(conv_layer_2, self.dropout_holder)\\n self.word_reps = self.cnn_layer\\n if self.use_additional and self.hybrid:\\n self.cnn_layer = tf.concat([self.word_reps, self.word_embeddings_2], \\n axis=-1)\\n self.word_reps = self.cnn_layer\",\n \"def _init_layers(self) -> None:\\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\\n self.embed_dims = self.self_attn.embed_dims\\n self.ffn = FFN(**self.ffn_cfg)\\n norms_list = [\\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\\n for _ in range(3)\\n ]\\n self.norms = ModuleList(norms_list)\",\n \"def _add_layer(self, layer_dict, layer_name, input_layers, merge_mode=None, share_params_with=None):\\n util.colorprint(layer_name, 'teal')\\n \\n layer_dict = dict(layer_dict)\\n util.colorprint(layer_dict, 'red')\\n \\n if share_params_with is not None:\\n print \\\"Warning: ignoring share_params_with\\\"\\n \\n layer_options = layer_dict[\\\"options\\\"]\\n layer=None\\n if layer_dict[\\\"type\\\"]==\\\"conv2d\\\":\\n #TODO: remove below\\n nb_filter, nb_row, nb_col = 3,3,3\\n layer = keras.layers.convolutional.Convolution2D(nb_filter, nb_row, nb_col, **layer_options)\\n elif layer_dict[\\\"type\\\"]==\\\"dense\\\":\\n dim = layer_dict[\\\"output_dim\\\"]\\n # del layer_options[\\\"output_dim\\\"]\\n layer = keras.layers.core.Dense(dim, **layer_options) \\n else:\\n print \\\"Ursol Major\\\"\\n RaiseError()\\n # TODO: one of the layers is a string\\n if isinstance(input_layers, list):\\n #this means that there is input from a loop to this layer\\n self.model.add_node(layer, name=layer_name, inputs=input_layers, merge_mode=merge_mode)\\n else:\\n self.model.add_node(layer, name=layer_name, input=input_layers)\\n\\n return layer_name\",\n \"def _embed_composite_feature(self, features, embedding_layers):\\n embedding_list = []\\n for i in range(len(embedding_layers)):\\n embedding_list.append(embedding_layers[i](features[:, :, i]))\\n embedding = tf.add_n(embedding_list)\\n return embedding\",\n \"def __init__(self, config_data, in_dims, layer_num):\\n self.layer_num = layer_num\\n self.in_dims = in_dims\\n self.out_dims = in_dims[\\\"width\\\"] * in_dims[\\\"height\\\"] * in_dims[\\\"channels\\\"]\\n self.layer_type = config_data[\\\"type\\\"]\\n self.name = config_data[\\\"name\\\"]\\n self.params = []\",\n \"def test_2d_local_convolution(batch_exp, input_height, input_width, in_channels, filters, drawer, is_first):\\n batch_size = 2 ** batch_exp\\n if is_first:\\n data_format = \\\"channels_first\\\"\\n else:\\n data_format = \\\"channels_last\\\"\\n kernel_height = drawer.draw(integers(min_value=1, max_value=input_height-1))\\n kernel_width = drawer.draw(integers(min_value=1, max_value=input_width-1))\\n\\n if data_format == \\\"channels_last\\\":\\n input_shape = (input_height, input_width, in_channels)\\n else:\\n input_shape = (in_channels, input_height, input_width)\\n\\n # Create keras model\\n keras_model = keras.Sequential()\\n keras_local = keras.layers.LocallyConnected2D(filters, (kernel_height, kernel_width),\\n use_bias=True,\\n bias_initializer=\\\"random_uniform\\\",\\n input_shape=input_shape,\\n data_format=data_format)\\n keras_model.add(keras_local)\\n\\n # Setup weights\\n weights = drawer.draw(arrays(shape=keras_local.kernel.shape.as_list(), dtype=np.float32, elements=floats(min_value=-1e6, max_value=1e6)))\\n keras_model.set_weights([weights])\\n\\n # Translate keras to PyTorch model\\n torch_model = translate.translate_layer(keras_local)[0].to(device)\\n\\n keras_input = np.random.uniform(-10, 10, (batch_size,) + input_shape).astype(np.float32)\\n keras_output = keras_model.predict(keras_input)\\n\\n torch_input = torch.Tensor(keras_input).to(device)\\n torch_output = torch_model(torch_input).cpu().data.numpy()\\n nan_fill = np.random.uniform(-5, 5)\\n assert (np.isnan(keras_output) == np.isnan(torch_output)).all()\\n keras_output[np.isnan(keras_output)] = nan_fill\\n torch_output[np.isnan(torch_output)] = nan_fill\\n assert np.isclose(keras_output, torch_output, atol=1e-3, rtol=1e-3).all()\",\n \"def multiply(self, layer):\\n pass\",\n \"def forward(self,bottom,top):\\n # imgmaps = self.batch_loader.mixup_gen()\\n # print(\\\"DataLayer forward!!\\\")\\n trainX, trainY = self.batch_loader.batch_imgs()\\n # print(\\\"trainX:\\\",trainX.shape)\\n # print(\\\"trainY:\\\",trainY.shape)\\n # print(\\\"trainY:\\\", trainY)\\n # print(\\\"top[0].data.shape:\\\",top[0].data.shape)\\n # print(\\\"top[1].data.shape:\\\", top[1].data.shape)\\n top[0].data[:, ...] = trainX\\n top[1].data[:, ...] = trainY\\n # print(\\\"DataLayer forward!!\\\")\",\n \"def apply_net_tiled(net,img):\\n\\n # borders = [8,20,20] ## border width within each patch that is thrown away after prediction\\n # patchshape_padded = [32,240,240] ## the size of the patch that we feed into the net. must be divisible by 8 or net fails.\\n # patchshape = [16,200,200] ## must be divisible by 8 to avoid artifacts.\\n # stride = [16,200,200] ## same as patchshape in this case\\n def f(n,m): return (ceil(n/m)*m)-n ## f(n,m) gives padding needed for n to be divisible by m\\n def g(n,m): return (floor(n/m)*m)-n ## f(n,m) gives un-padding needed for n to be divisible by m\\n\\n a,b,c = img.shape[1:]\\n q,r,s = f(a,8),f(b,8),f(c,8) ## calculate extra border needed for stride % 8 = 0\\n\\n ZPAD,YPAD,XPAD = 8,24,24\\n\\n img_padded = np.pad(img,[(0,0),(ZPAD,ZPAD+q),(YPAD,YPAD+r),(XPAD,XPAD+s)],mode='constant') ## pad for patch borders\\n output = np.zeros(img.shape)\\n\\n zs = np.r_[:a:16]\\n ys = np.r_[:b:200]\\n xs = np.r_[:c:200]\\n\\n for x,y,z in itertools.product(xs,ys,zs):\\n qe,re,se = min(z+16,a+q),min(y+200,b+r),min(x+200,c+s)\\n ae,be,ce = min(z+16,a),min(y+200,b),min(x+200,c)\\n patch = img_padded[:,z:qe+2*ZPAD,y:re+2*YPAD,x:se+2*XPAD]\\n patch = torch.from_numpy(patch).cuda().float()\\n patch = net(patch[None])[0,:,ZPAD:-ZPAD,YPAD:-YPAD,XPAD:-XPAD].detach().cpu().numpy()\\n output[:,z:ae,y:be,x:ce] = patch[:,:ae-z,:be-y,:ce-x]\\n\\n return output\",\n \"def layeredTexturePort(*args, annotation: Union[AnyStr, bool]=\\\"\\\", backgroundColor:\\n Union[List[float, float, float], bool]=None, defineTemplate: AnyStr=\\\"\\\",\\n docTag: Union[AnyStr, bool]=\\\"\\\", dragCallback: Script=None, dropCallback:\\n Script=None, enable: bool=True, enableBackground: bool=True,\\n enableKeyboardFocus: bool=True, exists: bool=True, fullPathName:\\n bool=True, height: Union[int, bool]=0, highlightColor: Union[List[float,\\n float, float], bool]=None, isObscured: bool=True, manage: bool=True,\\n noBackground: bool=True, node: name=None, numberOfPopupMenus: bool=True,\\n parent: Union[AnyStr, bool]=\\\"\\\", popupMenuArray: bool=True,\\n preventOverride: bool=True, selectedAlphaControl: AnyStr=\\\"\\\",\\n selectedBlendModeControl: AnyStr=\\\"\\\", selectedColorControl: AnyStr=\\\"\\\",\\n selectedIsVisibleControl: AnyStr=\\\"\\\", statusBarMessage: AnyStr=\\\"\\\",\\n useTemplate: AnyStr=\\\"\\\", visible: bool=True, visibleChangeCommand:\\n Union[Script, bool]=None, width: Union[int, bool]=0, q=True, query=True,\\n e=True, edit=True, **kwargs)->Union[AnyStr, Any]:\\n pass\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.64498407","0.6189903","0.6165279","0.5999772","0.5932045","0.5875921","0.5680412","0.5599581","0.5518509","0.5489415","0.5385347","0.5311613","0.53089815","0.5294126","0.5293941","0.52928007","0.5287579","0.5253471","0.52430594","0.5234675","0.5187667","0.5184719","0.51756847","0.5174631","0.5174631","0.51700413","0.5121166","0.5120242","0.5111518","0.51053774","0.5101962","0.5079799","0.5071217","0.5069473","0.5055256","0.5047029","0.50431156","0.50379586","0.5035134","0.5030663","0.50138134","0.5010369","0.5002769","0.4993102","0.49889123","0.49834126","0.49804276","0.49778217","0.49676228","0.49628872","0.49607155","0.49590054","0.49585548","0.49464065","0.4936438","0.49362716","0.49264815","0.49202877","0.49060592","0.48991746","0.48930696","0.4889569","0.48893136","0.48881912","0.4881248","0.48786336","0.48731115","0.4872928","0.4868599","0.48685786","0.48681092","0.48629642","0.48580903","0.48571175","0.48501977","0.48449942","0.4842994","0.4840277","0.48296258","0.48248196","0.48238403","0.481574","0.47991833","0.47967663","0.47955006","0.4794905","0.4787943","0.47820505","0.47782245","0.47773522","0.47743624","0.47693402","0.47674328","0.47658128","0.47632277","0.47626638","0.4762577","0.4760584","0.47557127","0.47537267","0.4751996"],"string":"[\n \"0.64498407\",\n \"0.6189903\",\n \"0.6165279\",\n \"0.5999772\",\n \"0.5932045\",\n \"0.5875921\",\n \"0.5680412\",\n \"0.5599581\",\n \"0.5518509\",\n \"0.5489415\",\n \"0.5385347\",\n \"0.5311613\",\n \"0.53089815\",\n \"0.5294126\",\n \"0.5293941\",\n \"0.52928007\",\n \"0.5287579\",\n \"0.5253471\",\n \"0.52430594\",\n \"0.5234675\",\n \"0.5187667\",\n \"0.5184719\",\n \"0.51756847\",\n \"0.5174631\",\n \"0.5174631\",\n \"0.51700413\",\n \"0.5121166\",\n \"0.5120242\",\n \"0.5111518\",\n \"0.51053774\",\n \"0.5101962\",\n \"0.5079799\",\n \"0.5071217\",\n \"0.5069473\",\n \"0.5055256\",\n \"0.5047029\",\n \"0.50431156\",\n \"0.50379586\",\n \"0.5035134\",\n \"0.5030663\",\n \"0.50138134\",\n \"0.5010369\",\n \"0.5002769\",\n \"0.4993102\",\n \"0.49889123\",\n \"0.49834126\",\n \"0.49804276\",\n \"0.49778217\",\n \"0.49676228\",\n \"0.49628872\",\n \"0.49607155\",\n \"0.49590054\",\n \"0.49585548\",\n \"0.49464065\",\n \"0.4936438\",\n \"0.49362716\",\n \"0.49264815\",\n \"0.49202877\",\n \"0.49060592\",\n \"0.48991746\",\n \"0.48930696\",\n \"0.4889569\",\n \"0.48893136\",\n \"0.48881912\",\n \"0.4881248\",\n \"0.48786336\",\n \"0.48731115\",\n \"0.4872928\",\n \"0.4868599\",\n \"0.48685786\",\n \"0.48681092\",\n \"0.48629642\",\n \"0.48580903\",\n \"0.48571175\",\n \"0.48501977\",\n \"0.48449942\",\n \"0.4842994\",\n \"0.4840277\",\n \"0.48296258\",\n \"0.48248196\",\n \"0.48238403\",\n \"0.481574\",\n \"0.47991833\",\n \"0.47967663\",\n \"0.47955006\",\n \"0.4794905\",\n \"0.4787943\",\n \"0.47820505\",\n \"0.47782245\",\n \"0.47773522\",\n \"0.47743624\",\n \"0.47693402\",\n \"0.47674328\",\n \"0.47658128\",\n \"0.47632277\",\n \"0.47626638\",\n \"0.4762577\",\n \"0.4760584\",\n \"0.47557127\",\n \"0.47537267\",\n \"0.4751996\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90881,"cells":{"query":{"kind":"string","value":"Transformer Encoder which comprises several transformer layers"},"document":{"kind":"string","value":"def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\n super(TransformerEncoder, self).__init__()\n\n # Create num_layers of TransformerBlock\n self.encoder = Sequential(\n [\n TransformerBlock(num_heads=num_heads,\n D=D,\n hidden_layers=[mlp_dim, D],\n dropout=dropout,\n norm_eps=norm_eps)\n for _ in range(num_layers)\n ]\n )"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def encoder(enc_input, attn_bias, n_layer, n_head,\n d_key, d_value, d_model, d_inner_hid, pos_enc,\n preporstprocess_dropout, attention_dropout,\n relu_dropout, preprocess_cmd='n',\n postprocess_cmd='da'):\n for i in range(n_layer):\n enc_output = encoder_layer(enc_input, attn_bias, n_head,\n d_key, d_value, d_model,d_inner_hid, pos_enc,\n prepostprocess_dropout, attention_dropout,relu_dropout,\n preprocess_cmd, postprocess_cmd\n )\n enc_input = enc_output\n enc_output = pre_process_layer(enc_output,\n preprocess_cmd, preporstprocess_dropout)\n return enc_output","def encoder(self, inputs):\n pass","def build_encoder(opt, embeddings, structure_embeddings):\n return TransformerEncoder(opt.enc_layers, opt.enc_rnn_size, opt.heads, opt.transformer_ff, opt.dropout, embeddings, structure_embeddings)","def _define_encoder(self):\n self.encoder = nn.Sequential(nn.Conv2d(in_channels=3, out_channels=32, kernel_size=4, stride=2, padding=1), # B, 32, 32, 32\n nn.SELU(),\n nn.Conv2d(32, 32, 4, 2, 1), # B, 32, 16, 16\n nn.SELU(),\n nn.Conv2d(32, 64, 4, 2, 1), # B, 64, 8, 8\n nn.SELU(),\n nn.Conv2d(64, 64, 4, 2, 1), # B, 64, 4, 4\n nn.SELU(),\n nn.Conv2d(64, 256, 4, 1), # B, 256, 1, 1\n nn.SELU(),\n View((-1, 256 * 1 * 1)), # B, 256\n nn.Linear(256, self.encoding_shape * 2), # B, z_dim*2\n )","def _define_encoder(self):\n self.encoder = nn.Sequential(View((-1, 64 * 64 * 3)),\n nn.Linear(64 * 64 * 3, 5120, bias=False), nn.SELU(),\n nn.BatchNorm1d(5120),\n nn.Linear(5120, 2560, bias=False), nn.SELU(),\n nn.BatchNorm1d(2560),\n nn.Linear(2560, 512, bias=False), nn.SELU(),\n nn.BatchNorm1d(512),\n nn.Linear(512, 128, bias=False), nn.SELU(),\n nn.BatchNorm1d(128),\n nn.Linear(128, self.encoding_shape, bias=False), nn.SELU(),\n )","def fit_transform(self):\n if self.enc_types == \"label\":\n return self._label_encoding()\n elif self.enc_types == \"ohe\":\n return self._one_hot_encoder()\n elif self.enc_types == \"binary\":\n return self._binarization()\n else:\n raise Exception(\"Encoding type not understood\")","def encoder_layer(enc_input, attn_bias, n_head, d_key,\n d_value, d_model, d_inner_hid, pos_enc, prepostprocess_dropout,\n attention_dropout, relu_dropout, preprocess_cmd='n',\n postprocess_cmd='da'):\n attn_output = multi_head_attention(\n pre_process_layer(enc_input, preprocess_cmd, prepostprocess_dropout),\n None, None, attn_bias, d_key, d_value, d_model, pos_enc,\n n_head, attention_dropout\n )\n attn_output = post_process_layer(enc_input, attn_output,\n postprocess_cmd, prepostprocess_dropout)\n ffd_output = positionwise_feed_forward(\n pre_process_layer(attn_output, preprocess_cmd, prepostprocess_dropout),\n d_inner_hid, d_model, relu_dropout\n )\n return post_process_layer(attn_output, ffd_output,\n postprocess_cmd, prepostprocess_dropout)","def TransformerEncoderLayer(self, name, is_causal=True):\n # Hack to be compatible with ckpt generated by self._rep\n return self._Seq(\n name,\n self._Seq('block', self._Attention('self_atten', is_causal=is_causal),\n self.Feedforward('ff')))","def TransformerEncoderStack(self, name, num_layers=1):\n blocks = [\n self.TransformerEncoderLayer(name='iter_{:0>3d}'.format(d))\n for d in range(num_layers)\n ]\n return self.Stack(name, blocks)","def encode(self):\n with tf.name_scope(\"encode\"):\n self.encoder_inputs = tf.layers.dense(\n inputs=self.encoder_inputs,\n units=self.options['hidden_size'], activation=None, use_bias=True,\n kernel_initializer=tf.keras.initializers.he_normal(seed=None),\n bias_initializer=tf.zeros_initializer(),\n kernel_regularizer=None, bias_regularizer=None, activity_regularizer=None,\n kernel_constraint=None, bias_constraint=None, trainable=True,\n name=None, reuse=None)\n self.encoder_inputs = tf.layers.batch_normalization(self.encoder_inputs,\n axis=-1,\n momentum=0.99,\n epsilon=0.001,\n center=True,\n scale=True,\n beta_initializer=tf.zeros_initializer(),\n gamma_initializer=tf.ones_initializer(),\n moving_mean_initializer=tf.zeros_initializer(),\n moving_variance_initializer=tf.ones_initializer(),\n training=self.is_training,\n trainable=True,\n renorm=False,\n renorm_momentum=0.99)\n # Prepare inputs to the layer stack by adding positional encodings and\n # applying dropout.\n # embedded_inputs = self.embedding_softmax_layer(inputs)\n #\n inputs_padding = transformer_model_utils.get_padding(tf.cast(\n tf.reduce_max(100*self.encoder_inputs, [-1]),\n dtype=tf.int32))\n\n with tf.name_scope(\"add_pos_encoding\"):\n length = tf.shape(self.encoder_inputs)[1]\n pos_encoding = transformer_model_utils.get_position_encoding(\n length, self.options[\"hidden_size\"])\n encoder_inputs = self.encoder_inputs + pos_encoding\n\n if self.is_training:\n encoder_inputs = tf.nn.dropout(\n encoder_inputs, 1 - self.options[\"layer_postprocess_dropout\"])\n\n return self.encoder_stack(encoder_inputs, self.attention_bias, inputs_padding)","def build_encoder(tparams, options):\n\t# word embedding (source)\n\tembedding = tensor.tensor3('embedding', dtype='float32')\n\tx_mask = tensor.matrix('x_mask', dtype='float32')\n\n\t# encoder\n\tproj = get_layer(options['encoder'])[1](tparams, embedding, options,\n\t\t\t\t\t\t\t\t\t\t\tprefix='encoder',\n\t\t\t\t\t\t\t\t\t\t\tmask=x_mask)\n\tctx = proj[0][-1]\n\n\treturn embedding, x_mask, ctx","def _create_encoder(self):\n\n def _init_weights(layer):\n \"\"\"Initializes the weights of a layer based on type.\"\"\"\n if isinstance(layer, (nn.Conv2d, nn.Linear)):\n torch.nn.init.xavier_uniform_(layer.weight)\n try:\n # Some layers may not have biases, so catch the exception and pass.\n layer.bias.data.fill_(0.0)\n except AttributeError:\n pass\n\n kernel_size = 5\n pad = 2\n input_channels = 1\n first_conv_channels = 6\n second_conv_channels = 16\n max_pool_kernel = 2\n linear_size = 120\n n_pixels = 7\n\n encoder = nn.Sequential(\n nn.Conv2d(\n input_channels, first_conv_channels, kernel_size, padding=pad),\n nn.BatchNorm2d(first_conv_channels),\n nn.ReLU(),\n nn.MaxPool2d(max_pool_kernel),\n nn.Conv2d(\n first_conv_channels, second_conv_channels, kernel_size,\n padding=pad),\n nn.BatchNorm2d(second_conv_channels),\n nn.ReLU(),\n nn.MaxPool2d(max_pool_kernel),\n utils.Flatten(),\n nn.Linear(n_pixels * n_pixels * self.n_digits * second_conv_channels,\n linear_size),\n nn.BatchNorm1d(linear_size),\n nn.ReLU(),\n nn.Linear(linear_size, self.embedding_dim),\n nn.Linear(self.embedding_dim, self.n_classes, bias=False),\n )\n\n encoder.apply(_init_weights)\n\n # This is the empirical approximation for initialization the vMF\n # distributions for each class in the final layer.\n if self.use_vmf:\n utils.vmf_class_weight_init(encoder[-1].weight, self.kappa_confidence,\n self.embedding_dim)\n\n return encoder","def encode(data, encoder):\n # Get the list of hidden depths\n\thd = encoder.hidden_depths\n # Find the middle hidden layer\n\tmiddle_layer_index = (len(hd)-1)/2\n # Initialize empty container for the encoded data\n\tdata_encoded = np.zeros((data.shape[0],hd[middle_layer_index]))\n\tfor i, d_ in enumerate(data):\n # feed forward, get all the activations, and just keep\n # the middle layer, which is the encoding\n\t\tx, z_container, x_container = encoder.ff(d_,True,True)\n\t\tx_encoded = x_container[1+middle_layer_index]\n\t\tdata_encoded[i] = x_encoded\n\t#\n\treturn data_encoded","def make_encoder(self, input_size: int, latent_size: int) -> nn.Module:\n pass","def __init__(self,\n input_dim,\n dec_seq_len,\n out_seq_len,\n d_model=512,\n nhead=8,\n num_encoder_layers=6,\n num_decoder_layers=6,\n dim_feedforward=2048,\n dropout=0.1,\n activation='relu',\n custom_encoder=None,\n custom_decoder=None):\n super(TransformerTS, self).__init__()\n self.transform = nn.Transformer(\n d_model=d_model,\n nhead=nhead,\n num_encoder_layers=num_encoder_layers,\n num_decoder_layers=num_decoder_layers,\n dim_feedforward=dim_feedforward,\n dropout=dropout,\n activation=activation,\n custom_encoder=custom_encoder,\n custom_decoder=custom_decoder\n )\n self.pos = PositionalEncoding(d_model)\n self.enc_input_fc = nn.Linear(input_dim, d_model)\n self.dec_input_fc = nn.Linear(input_dim, d_model)\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\n self.dec_seq_len = dec_seq_len","def TransformerEncoderLayer(self, name, stride=1, first_n=None,\n ff_hidden_dim=None, num_heads=None,\n use_moe=False):\n p = self.params\n if ff_hidden_dim is None:\n ff_hidden_dim = p.ff_hidden_dim\n if num_heads is None:\n num_heads = p.num_heads\n if use_moe:\n ffw_p = self.MoE('moe', ff_hidden_dim=ff_hidden_dim)\n else:\n ffw_p = self.Feedforward('ff', ff_hidden_dim=ff_hidden_dim)\n return self._Seq(name, self._Seq(\n 'block',\n self._StridedAttention('self_atten', stride=stride,\n first_n=first_n, num_heads=num_heads),\n ffw_p))","def beam_init(self, encoder_outputs, K):\n encoder_outputs = TransformerEncoderOutput(\n repeat_batch(encoder_outputs.output, K),\n repeat_batch(encoder_outputs.src_mask, K)\n )\n return encoder_outputs","def transformer_prepare_encoder(inputs, hparams, features=None):\n ishape_static = inputs.shape.as_list()\n encoder_input = inputs\n\n # Usual case - not a packed dataset.\n encoder_padding = common_attention.embedding_to_padding(encoder_input)\n ignore_padding = common_attention.attention_bias_ignore_padding(\n encoder_padding)\n encoder_self_attention_bias = ignore_padding\n encoder_decoder_attention_bias = ignore_padding\n inputs_position = None\n if hparams.proximity_bias:\n encoder_self_attention_bias += common_attention.attention_bias_proximal(\n common_layers.shape_list(inputs)[1])\n\n if inputs_position is not None:\n encoder_input = common_attention.add_timing_signal_1d_given_position(\n encoder_input, inputs_position)\n else:\n encoder_input = common_attention.add_timing_signal_1d(\n encoder_input)\n return (encoder_input, encoder_self_attention_bias,\n encoder_decoder_attention_bias)","def encode(self, img):\n with tf.variable_scope('encoder'):\n #conv1 = self.conv_layer(\n # img, [5, 5], [3, 32], stride=2, initializer_type=1, name='conv1')\n #conv2 = self.conv_layer(\n # conv1, [5, 5], [32, 32], stride=2, initializer_type=1, name='conv2')\n conv3 = self.conv_layer(\n img, [5, 5], [self.in_channels, 64], stride=2, initializer_type=1, name='conv3')\n #conv4 = self.conv_bn_layer(\n conv4 = self.conv_layer(\n conv3, [5, 5], [64, 128], stride=2, #is_training=self.is_training,\n initializer_type=1, name='conv4')\n #conv5 = self.conv_bn_layer(\n conv5 = self.conv_layer(\n conv4, [5, 5], [128, 256], stride=2, #is_training=self.is_training,\n initializer_type=1, name='conv5')\n shape = conv5.get_shape().as_list()\n feature_map_size = shape[1]*shape[2]*shape[3]\n conv5_flat = tf.reshape(\n conv5, [-1, feature_map_size], 'conv5_flat')\n #fc6 = self.fc_bn_layer(conv5_flat, 1024, is_training=self.is_training,\n fc6 = self.fc_layer(conv5_flat, 1024,\n initializer_type=1, name='fc6')\n #fc7 = self.fc_layer(fc6, 1024, initializer_type=1, name='fc7')\n return fc6, shape","def encode(self, inputs, target_space, hparams, features=None, losses=None):\n del losses\n\n inputs = common_layers.flatten4d3d(inputs)\n\n encoder_input, self_attention_bias, encoder_decoder_attention_bias = (\n transformer.transformer_prepare_encoder(\n inputs, target_space, hparams, features=features))\n\n encoder_input = tf.nn.dropout(encoder_input,\n 1.0 - hparams.layer_prepostprocess_dropout)\n\n (encoder_output, encoder_extra_output) = (\n my_universal_transformer_util.universal_transformer_encoder(\n encoder_input,\n self_attention_bias,\n hparams,\n nonpadding=transformer.features_to_nonpadding(features, \"inputs\"),\n save_weights_to=self.attention_weights))\n\n return encoder_output, encoder_decoder_attention_bias, encoder_extra_output","def forward(self, src, mask):\n bs = src.shape[0]\n src = src.permute(2, 0, 1)\n m = src \n enc_embed = self.enc_embed.weight.unsqueeze(1).repeat(1, bs, 1)\n for layer in self.encoder_layers:\n m = layer(m,\n pos=enc_embed,\n src_mask = mask\n )\n return m.permute(1, 2, 0), enc_embed.permute(1, 2, 0)","def encode_all(self, inputs, encoder):\n input_shape = tf.shape(inputs)\n num_timesteps, batch_size = input_shape[0], input_shape[1]\n reshaped_inputs = tf.reshape(inputs, [-1, inputs.shape[-1]])\n inputs_encoded = encoder(reshaped_inputs)\n inputs_encoded = tf.reshape(inputs_encoded, [num_timesteps, batch_size, encoder.output_size])\n return inputs_encoded","def encoder_model(self, input_shape, layers, filters, latent_dims):\n\n l = []\n l.append(Input(shape=input_shape, name='encoder_input'))\n\n for i in range(0,layers):\n l.append(Conv2D(filters[i], (5,5), padding='same',\n data_format='channels_last', name='Conv_'+str(i),\n activation='relu')(l[i*2]))\n l.append(MaxPooling2D(pool_size=(2, 2),\n data_format='channels_last',\n name='Pool_'+str(i))(l[i*2+1]))\n\n conv_out = Flatten(name='Flatten_dims')(l[-1])\n\n self.z_mean = Dense(latent_dims, name='z_mean')(conv_out)\n self.z_sigma = Dense(latent_dims, name='z_sigma')(conv_out)\n self.z = Lambda(VAE_sampling, output_shape=(latent_dims,), name='z')([self.z_mean,self.z_sigma])\n encoder = Model(l[0], [self.z_mean,self.z_sigma,self.z], name='encoder')\n\n encoder.summary()\n\n return encoder","def __init__(self, args):\n super(Transformer, self).__init__()\n\n self.name = 'Transformer'\n if len(args.max_relative_pos) != args.nlayers:\n assert len(args.max_relative_pos) == 1\n args.max_relative_pos = args.max_relative_pos * args.nlayers\n\n self.embedder = Embedder(args)\n self.encoder = Encoder(args, self.embedder.enc_input_size)\n self.decoder_loc = nn.Linear(self.embedder.enc_input_size, 1)\n self.decoder_fix = nn.Linear(self.embedder.enc_input_size, 1)\n self.nobug_embedding = nn.Parameter(torch.zeros(1))\n self.criterion = nn.CrossEntropyLoss(reduction='none')","def build_encoder_bi(tparams, options):\n\t# word embedding (source)\n\tembedding = tensor.tensor3('embedding', dtype='float32')\n\tembeddingr = embedding[::-1]\n\tx_mask = tensor.matrix('x_mask', dtype='float32')\n\txr_mask = x_mask[::-1]\n\n\t# encoder\n\tproj = get_layer(options['encoder'])[1](tparams, embedding, options,\n\t\t\t\t\t\t\t\t\t\t\tprefix='encoder',\n\t\t\t\t\t\t\t\t\t\t\tmask=x_mask)\n\tprojr = get_layer(options['encoder'])[1](tparams, embeddingr, options,\n\t\t\t\t\t\t\t\t\t\t\t prefix='encoder_r',\n\t\t\t\t\t\t\t\t\t\t\t mask=xr_mask)\n\n\tctx = tensor.concatenate([proj[0][-1], projr[0][-1]], axis=1)\n\n\treturn embedding, x_mask, ctx","def build_encoder(img_shape):\n input_img = Input(shape=(img_shape)) \n x = Conv2D(16, (3, 3), activation='tanh', padding='same')(input_img)\n x = AveragePooling2D((2, 2), padding='same')(x)\n x = Conv2D(8, (3, 3), padding='same')(x)\n x = LeakyReLU()(x)\n x = AveragePooling2D((2, 2), padding='same')(x)\n x = Conv2D(8, (3, 3), padding='same')(x)\n x = LeakyReLU()(x)\n x = Dropout(0.5)(x)\n x = AveragePooling2D((2, 2), padding='same')(x)\n x = Conv2D(1, (3, 3), padding='same')(x)\n x = LeakyReLU()(x)\n x = AveragePooling2D((3, 3), padding='same')(x)\n x = Flatten()(x)\n encoded = Dense(540, activation='tanh')(x)\n Encoder=Model(input_img,encoded,name='encoder')\n return input_img,encoded,Encoder","def encoder(self, x):\n # Input (batch_size, Channels=1, Width=28, Height=28)\n x = self.enc_conv(x) # (batch_size, lin_before_latent_dim)\n mean = self.enc_mean(x) # (batch_size, latent_dim)\n logvar = self.enc_logvar(x) # (batch_size, latent_dim)\n return mean, logvar","def forward(self, x):\n # Encoder1 --block1\n encode_block1 = self.conv_encoder1(x)\n if self.residus[0] == 1:\n encode_block1 += self.residual_shortcut1(x)\n encode_pool1 = self.max_pool_encoder1(encode_block1)\n\n # Encoder2 --block2\n encode_block2 = self.conv_encoder2(encode_pool1)\n if self.residus[1] == 1:\n encode_block2 += self.residual_shortcut2(encode_pool1)\n encode_pool2 = self.max_pool_encoder2(encode_block2)\n \n # Encoder3 --block3\n encode_block3 = self.conv_encoder3(encode_pool2)\n if self.residus[2] == 1:\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\n encode_pool3 = self.max_pool_encoder3(encode_block3)\n \n # Encoder4 --block4\n encode_block4 = self.conv_encoder4(encode_pool3) \n if self.residus[3] == 1:\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\n encode_pool4 = self.max_pool_encoder4(encode_block4)\n\n # Transitional block \n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\n \n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\n \n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\n\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\n \n convTranspose_transitional = self.convTranspose_transitional(middle_block) \n # Decoder4 --block5\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \n if self.residus[4] == 1:\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\n\n #--block6\n cat_layer3 = self.conv_decoder4(decode_block4) \n if self.residus[5] == 1:\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\n \n \n # Decoder3 --block7\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\n if self.residus[6] == 1:\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\n \n #--block8\n cat_layer2 = self.conv_decoder3(decode_block3) \n if self.residus[7] == 1:\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\n \n # Decoder2 --block9\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \n if self.residus[8] == 1:\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\n \n #--block10\n cat_layer1 = self.conv_decoder2(decode_block2)\n if self.residus[9] == 1:\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\n \n # Decoder1 --block11\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \n if self.residus[10] == 1:\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\n \n #--block12\n final_layer = self.final_layer(decode_block1)\n if self.residus[11] == 1:\n final_layer += self.residual_shortcut_final_layer(decode_block1)\n \n \n return final_layer","def encode_input(self, x_tensor, inp_lens_tensor):\r\n input_emb = self.input_emb.forward(x_tensor)\r\n enc_output_each_word, enc_context_mask, enc_final_states = self.encoder(input_emb, inp_lens_tensor)\r\n enc_final_states_reshaped = (enc_final_states[0].unsqueeze(0), enc_final_states[1].unsqueeze(0))\r\n # print('lest go', enc_final_states_reshaped[1].shape)\r\n return enc_output_each_word, enc_context_mask, enc_final_states_reshaped","def build_label_transform():\n\n return NALabelEncoder()","def transform(self, X):\n for i,f in enumerate(self.features):\n X[f] = self._label_encoders_[i].transform(X[f])\n return X","def _define_encoder(self):\n raise NotImplementedError","def __init__(self, momentum: float = .5):\n super(VanillaEncoder, self).__init__()\n self.conv1 = PointNetConv2Layer(64, momentum)\n self.conv2 = PointNetConv2Layer(64, momentum)\n self.conv3 = PointNetConv2Layer(64, momentum)\n self.conv4 = PointNetConv2Layer(128, momentum)\n self.conv5 = PointNetConv2Layer(1024, momentum)","def encoder_inference(self, features, states):\n with tf.name_scope(f\"{self.name}_encoder\"):\n outputs = tf.expand_dims(features, axis=0)\n outputs, new_states = self.encoder.recognize(outputs, states)\n return tf.squeeze(outputs, axis=0), new_states","def __call__(self, encoder_inputs, attention_bias, inputs_padding):\n for n, layer in enumerate(self.layers):\n # Run inputs through the sublayers.\n self_attention_layer = layer[0]\n feed_forward_network = layer[1]\n\n with tf.variable_scope(\"encoder_layer_%d\" % n):\n with tf.variable_scope(\"self_attention\"):\n encoder_inputs = self_attention_layer(encoder_inputs, attention_bias)\n with tf.variable_scope(\"ffn\"):\n encoder_inputs = feed_forward_network(encoder_inputs, inputs_padding)\n\n with tf.variable_scope(\"encoder_output\"):\n output = self.output_normalization(encoder_inputs)\n\n return output # self.output_normalization(encoder_inputs)","def pretrain_forward(self, inp):\n return self.encoder(inp)","def encoder(self, features=[8], name=\"encoder\") -> KM.Model:\n input_tensor = KL.Input(\n shape=(32, 32, 3)\n ) # shape of images for cifar10 dataset\n encoded = KL.Conv2D(\n features[0],\n 3,\n strides=(2, 2),\n padding=\"same\",\n use_bias=False,\n name=name + f\"_conv_{1}\",\n )(input_tensor)\n encoded = KL.Activation(\"relu\")(KL.BatchNormalization()(encoded))\n encoded_list = [encoded]\n\n # Prepare the skip tensor from input\n skip_input_tensor = KL.Activation(\"relu\")(\n KL.BatchNormalization()(\n KL.Conv2D(features[0], 1, strides=1, use_bias=False)(input_tensor)\n )\n )\n skip_input_tensor = KL.SpatialDropout2D(rate=0.2)(skip_input_tensor)\n skip_input_tensor = KL.AveragePooling2D(pool_size=(2, 2), strides=2)(\n skip_input_tensor\n )\n skip_tensors = tf.concat(\n [\n skip_input_tensor, # Routing info from input tensor to next levels\n encoded, # Routes info from second level to next levels\n ],\n axis=-1,\n )\n for i, feature_num in enumerate(features[1:], start=2):\n encoded, skip_tensors = conv_block(\n encoded,\n skip_tensors,\n features_in=features[i - 2],\n features_out=feature_num,\n name=name + f\"_conv_{i}\",\n )\n encoded_list.append(encoded)\n return KM.Model(inputs=input_tensor, outputs=encoded_list, name=name)","def __call__(self, *args, **kwargs):\n return self.encoder_model(args[0])","def __call__(self, *args, **kwargs):\n return self.encoder_model(args[0])","def encode(self, x):\n self.eval()\n x = torch.as_tensor(x).unsqueeze(0)\n if self.do_mt:\n enc_output, _ = self.encoder_mt(x, None)\n else:\n enc_output, _ = self.encoder(x, None)\n return enc_output.squeeze(0)","def forward(self, x):\n if self.training:\n x = self.input_pert(x)\n x = self.encoder(x)\n x = self.decoder(x)\n return x","def encoder(x, h_dim, reuse=False, reorder_layers=True, training=False):\n if(reorder_layers):\n block = conv_block_reordered\n else:\n block = conv_block\n with tf.variable_scope('encoder', reuse=reuse):\n net = block(x, h_dim, name='conv_1', training=training, block_idx=0)\n net = block(net, h_dim, name='conv_2', training=training, block_idx=1)\n net = block(net, h_dim, name='conv_3', training=training, block_idx=2)\n net = block(net, h_dim, name='conv_4', training=training, block_idx=3)\n embeddings = tf.contrib.layers.flatten(net)\n return embeddings","def complex_encoder(self, x, reuse=True): \n\t\tenc_layers = self.enc_layers\n\t\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\n\t\twith tf.variable_scope('model', reuse=reuse):\n\t\t\twith tf.variable_scope('encoder', reuse=reuse):\n\t\t\t\tconv1 = tf.layers.conv2d(x, 32, 4, strides=(2,2), padding=\"SAME\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv1')\n\t\t\t\tconv2 = tf.layers.conv2d(conv1, 32, 4, strides=(2,2), padding=\"SAME\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv2')\n\t\t\t\tconv3 = tf.layers.conv2d(conv2, 64, 4, strides=(2,2), padding=\"SAME\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv3')\n\t\t\t\tconv4 = tf.layers.conv2d(conv3, 64, 4, strides=(2,2), padding=\"SAME\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv4')\n\t\t\t\tconv5 = tf.layers.conv2d(conv4, 256, 4, padding=\"VALID\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv5')\n\t\t\t\tflattened = tf.reshape(conv5, (-1, 256*1*1))\n\t\t\t\tz_mean = tf.layers.dense(flattened, enc_layers[-1], activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-final')\n\t\treturn z_mean","def call(self, inputs, *args, **kwargs):\n outputs = self.encoder(inputs, *args, **kwargs)\n return outputs","def forward(self, x):\n # Encode\n encode_block1 = self.conv_encoder1(x)\n encode_pool1 = self.max_pool_encoder1(encode_block1)\n encode_block2 = self.conv_encoder2(encode_pool1)\n encode_pool2 = self.max_pool_encoder2(encode_block2)\n encode_block3 = self.conv_encoder3(encode_pool2)\n encode_pool3 = self.max_pool_encoder3(encode_block3)\n encode_block4 = self.conv_encoder4(encode_pool3)\n encode_pool4 = self.max_pool_encoder4(encode_block4)\n\n # Transitional block\n middle_block = self.transitional_block(encode_pool4)\n\n # Decode\n decode_block4 = torch.cat((middle_block, encode_block4), 1)\n cat_layer3 = self.conv_decoder4(decode_block4)\n decode_block3 = torch.cat((cat_layer3, encode_block3), 1)\n cat_layer2 = self.conv_decoder3(decode_block3)\n decode_block2 = torch.cat((cat_layer2, encode_block2), 1)\n cat_layer1 = self.conv_decoder2(decode_block2)\n decode_block1 = torch.cat((cat_layer1, encode_block1), 1)\n final_layer = self.final_layer(decode_block1)\n return final_layer","def encode(self,x,x_len):\n\n ## Check to see if batch_size parameter is fixed or base on input batch\n cur_batch_size = x.size()[1]\n encode_init_state = self.encoder.initialize_hidden_state(cur_batch_size)\n encoder_state, encoder_outputs = self.encoder.forward(x, encode_init_state, x_len)\n\n return encoder_outputs, encoder_state","def forward(self, x):\n # Get results of encoder network\n h1 = self.encode_nn(x)\n\n # latent space\n mu = self.encode_mu(h1)\n log_var = self.encode_log_var(h1)\n\n # Reparameterize\n z = self.reparameterize(mu, log_var)\n return z, mu, log_var","def forward(self, x):\n # Get results of encoder network\n h1 = self.encode_nn(x)\n\n # latent space\n mu = self.encode_mu(h1)\n log_var = self.encode_log_var(h1)\n\n # Reparameterize\n z = self.reparameterize(mu, log_var)\n return z, mu, log_var","def base_encoder(cls, data, init_encoder, downsize_encoder, input_encoder):\n #todo: maybe do positional encoding before passing to init_encoder\n data = downsize_encoder(init_encoder(data))\n return input_encoder(data)","def transform(self, X):\n \n # Check is fit had been called\n check_is_fitted(self, ['encoder_dict_'])\n \n # Check that the input is of the same shape as the one passed\n # during fit.\n if X.shape[1] != self.input_shape_[1]:\n raise ValueError('Number of columns in dataset is different from training set used to fit the encoder')\n \n X = X.copy()\n for feature in self.variables:\n for category in self.encoder_dict_[feature]:\n X[str(feature) + '_' + str(category)] = np.where(X[feature] == category, 1, 0)\n \n # drop the original non-encoded variables.\n X.drop(labels=self.variables, axis=1, inplace=True)\n \n return X","def encode(self, x):\n mu = []\n logvar = []\n for i in range(self.n_views):\n mu_, logvar_ = self.encoders[i](x[i])\n mu.append(mu_)\n logvar.append(logvar_)\n mu = torch.stack(mu)\n logvar = torch.stack(logvar)\n mu_out, logvar_out = self.join_z(mu, logvar)\n qz_x = hydra.utils.instantiate( \n self.cfg.encoder.default.enc_dist, loc=mu_out, scale=logvar_out.exp().pow(0.5)\n )\n return [qz_x]","def forward(self, *args):\r\n enc_src, src_mask, src_inp = self.unified_encoder(*args)\r\n batch_size, _, hid_dim = src_inp.shape\r\n device = src_inp.device\r\n # enc_src = [batch size, src len, hid dim]\r\n # src_mask = [batch size, src len]\r\n\r\n trg_inp = torch.cat((torch.zeros((batch_size, 1, hid_dim), device=device), src_inp[:, :-1, :]), 1)\r\n decoder_output, attention = self.decoder(trg_inp, enc_src, src_mask)\r\n\r\n # decoder_output = [batch size, trg len, hid dim]\r\n # attention = [batch size, n heads, trg len, src len]\r\n\r\n outputs = self.output_layer(decoder_output)\r\n\r\n return outputs, attention","def _apply_encoder(self, frame, prop, encoder, encoder_type=\"category\"):\n pass","def forward(self, x, mask):\n x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, mask))\n #print('encoder')\n #print(x.shape)\n return self.sublayer[1](x, self.feed_forward)","def autoencoder3(ds, compression_factor=16, input_noise=0.2, dropout_p=0.1, activ='tanh', final_activ='tanh'):\n # compression_factor=20\n print('DS shape: {}'.format(ds.shape))\n in_dims = np.prod(ds.shape[1:])\n encoding_dim = int(in_dims // compression_factor)\n in_shape = ds[0].shape\n print('Input Dims: {}, input shape: {}, encoding dims: {}'.format(in_dims, in_shape, encoding_dim))\n\n # this is our input placeholder\n input_img = Input(shape=(in_dims,))\n encoded = GaussianNoise(input_noise)(input_img)\n\n encoded = Dense(encoding_dim * 4, activation=activ, activity_regularizer=regularizers.activity_l1(10e-5))(encoded)\n # encoded = Dense(encoding_dim*4, activation='sigmoid')(input_img)\n encoded = BatchNormalization()(encoded)\n encoded = Dropout(dropout_p)(encoded) # batch norm before dropout\n # encoded = Dense(encoding_dim*3, activation=activ)(encoded)\n # encoded = Dropout(dropout_p)(encoded)\n encoded = Dense(encoding_dim * 2, activation=activ)(encoded)\n encoded = Dropout(dropout_p)(encoded)\n\n encoded = Dense(encoding_dim, activation=activ)(encoded)\n # Middle Noise\n encoded = GaussianNoise(0.02)(encoded)\n\n # DECODED LAYER\n # \"decoded\" is the lossy reconstruction of the input\n decoded = Dense(encoding_dim * 2, activation=activ)(encoded)\n # decoded = Dropout(dropout_p)(decoded)\n decoded = Dense(encoding_dim * 4, activation=activ)(decoded)\n decoded = Dropout(dropout_p)(decoded)\n decoded = Dense(in_dims, activation=final_activ)(decoded)\n\n # MODEL\n autoencoder = Model(input=input_img, output=decoded)\n\n # SEPERATE ENCODER MODEL\n encoder = Model(input=input_img, output=encoded)\n\n # create a placeholder for an encoded (32-dimensional) input\n encoded_input = Input(shape=(encoding_dim,))\n\n # retrieve the last layer of the autoencoder model\n decoder_layer0 = autoencoder.layers[-4]\n decoder_layer1 = autoencoder.layers[-3]\n decoder_layer2 = autoencoder.layers[-2]\n decoder_layer3 = autoencoder.layers[-1]\n # todo: make this into a dedicated unrolling function\n\n # create the decoder model - unrolling the model as we go\n decoder = Model(input=encoded_input, output=decoder_layer3(decoder_layer2(\n decoder_layer1(decoder_layer0(encoded_input)))))\n\n # model.add(GaussianNoise(0.1), input_shape=(n_input_len,))\n autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')\n\n autoencoder.model_name = 'Autoencoder 1'\n return autoencoder, encoder, decoder","def encoder(input_layer, num_filters, size=(4, 4), instance_norm=True):\n # Initialize the weights\n init = tf.random_normal_initializer(0.0, 0.02)\n\n # Convolutional Layer\n encoder_layer = Conv2D(filters=num_filters, kernel_size=size, strides=(2, 2), padding='same',\n kernel_initializer=init, use_bias=False)(input_layer)\n\n # Instance normalization\n if instance_norm:\n encoder_layer = tfa.layers.InstanceNormalization()(encoder_layer)\n\n # Leaky ReLU\n encoder_layer = LeakyReLU(alpha=0.2)(encoder_layer)\n return encoder_layer","def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\n latent_dimension):\n super(VAEEncoder, self).__init__()\n self.latent_dimension = latent_dimension\n\n # Reduce dimension up to second last layer of Encoder\n self.encode_nn = nn.Sequential(\n nn.Linear(in_dimension, layer_1d),\n nn.ReLU(),\n nn.Linear(layer_1d, layer_2d),\n nn.ReLU(),\n nn.Linear(layer_2d, layer_3d),\n nn.ReLU()\n )\n\n # Latent space mean\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\n\n # Latent space variance\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)","def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\n latent_dimension):\n super(VAEEncoder, self).__init__()\n self.latent_dimension = latent_dimension\n\n # Reduce dimension up to second last layer of Encoder\n self.encode_nn = nn.Sequential(\n nn.Linear(in_dimension, layer_1d),\n nn.ReLU(),\n nn.Linear(layer_1d, layer_2d),\n nn.ReLU(),\n nn.Linear(layer_2d, layer_3d),\n nn.ReLU()\n )\n\n # Latent space mean\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\n\n # Latent space variance\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)","def transformer_layers(self):\n return self._transformer_layers","def common_layers_with_encoder(self):\n return [\n self.self_attention, self.self_attention_layer_norm,\n self.intermediate_dense, self.output_dense, self.output_layer_norm\n ]","def autoencoder(self, data):\n with tf.variable_scope(\"autoencoder\"):\n latent = self.encoder(data)\n _, output = self.decoder(latent)\n\n return output, latent","def encoder(self, x, reuse=True):\n\t\te = x\n\t\tenc_layers = self.enc_layers\n\t\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\n\t\twith tf.variable_scope('model', reuse=reuse):\n\t\t\twith tf.variable_scope('encoder', reuse=reuse):\n\t\t\t\tfor layer_idx, layer_dim in enumerate(enc_layers[:-1]):\n\t\t\t\t\te = tf.layers.dense(e, layer_dim, activation=tf.nn.leaky_relu, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(layer_idx))\n\t\t\t\tif self.channel_model == 'bsc':\n\t\t\t\t\tz_mean = tf.layers.dense(e, self.z_dim, activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(len(enc_layers)-1))\n\t\t\t\telse:\n\t\t\t\t\t# N x D x 2 for erasure channel\n\t\t\t\t\tz_mean = tf.layers.dense(e, self.z_dim * 2, activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(len(enc_layers)-1))\n\t\t\t\t\tz_mean = tf.reshape(z_mean, (-1, self.z_dim, 2))\n\t\treturn z_mean","def build_encoder(shift):\n ### TODO.","def autoencoder(input_dims, filters, latent_dims):\n e_inputs = keras.Input(input_dims)\n d_inputs = keras.Input(latent_dims)\n\n encoder = e_inputs\n for f in filters:\n encoder = keras.layers.Conv2D(\n f, (3, 3), activation='relu', padding='same')(encoder)\n encoder = keras.layers.MaxPooling2D((2, 2), padding='same')(encoder)\n\n decoder = d_inputs\n for i in reversed(range(1, len(filters))):\n decoder = keras.layers.Conv2D(\n filters[i], (3, 3), activation='relu', padding='same')(decoder)\n decoder = keras.layers.UpSampling2D((2, 2))(decoder)\n\n decoder = keras.layers.Conv2D(\n filters[0], (3, 3), activation='relu', padding='valid')(decoder)\n decoder = keras.layers.UpSampling2D((2, 2))(decoder)\n decoder = keras.layers.Conv2D(input_dims[-1], (3, 3),\n activation='sigmoid',\n padding='same')(decoder)\n\n encoder = keras.Model(e_inputs, encoder)\n decoder = keras.Model(d_inputs, decoder)\n\n auto = keras.Model(e_inputs, decoder(encoder(e_inputs)))\n auto.compile(optimizer=\"adam\", loss=\"binary_crossentropy\")\n\n return encoder, decoder, auto","def encoder(x, reuse=False):\n if reuse:\n tf.get_variable_scope().reuse_variables()\n with tf.name_scope('Encoder'):\n e_dense_1 = tf.nn.relu(dense(x, input_dim, n_l1, 'e_dense_1'))\n e_dense_2 = tf.nn.relu(dense(e_dense_1, n_l1, n_l2, 'e_dense_2'))\n latent_variable = dense(e_dense_2, n_l2, z_dim, 'e_latent_variable')\n return latent_variable","def convert(encoder, bert_model):\n num_layers = encoder._config[\"num_layers\"]\n num_attention_heads = encoder._config[\"num_attention_heads\"]\n hidden_size = encoder._config[\"hidden_size\"]\n head_size = hidden_size // num_attention_heads\n assert head_size * num_attention_heads == hidden_size\n encoder._embedding_layer.set_weights(\n [bert_model[\"embeddings.word_embeddings.weight\"]])\n encoder._embedding_norm_layer.set_weights([\n bert_model[\"embeddings.LayerNorm.weight\"],\n bert_model[\"embeddings.LayerNorm.bias\"]\n ])\n encoder._type_embedding_layer.set_weights(\n [bert_model[\"embeddings.token_type_embeddings.weight\"]])\n encoder._position_embedding_layer.set_weights(\n [bert_model[\"embeddings.position_embeddings.weight\"]])\n for layer_num in range(num_layers):\n encoder._transformer_layers[\n layer_num]._attention_layer._key_dense.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.attention.self.key.weight\"].T\n .reshape((hidden_size, num_attention_heads, head_size)),\n bert_model[f\"encoder.layer.{layer_num}.attention.self.key.bias\"]\n .reshape((num_attention_heads, head_size))\n ])\n encoder._transformer_layers[\n layer_num]._attention_layer._query_dense.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.attention.self.query.weight\"]\n .T.reshape((hidden_size, num_attention_heads, head_size)),\n bert_model[f\"encoder.layer.{layer_num}.attention.self.query.bias\"]\n .reshape((num_attention_heads, head_size))\n ])\n encoder._transformer_layers[\n layer_num]._attention_layer._value_dense.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.attention.self.value.weight\"]\n .T.reshape((hidden_size, num_attention_heads, head_size)),\n bert_model[f\"encoder.layer.{layer_num}.attention.self.value.bias\"]\n .reshape((num_attention_heads, head_size))\n ])\n encoder._transformer_layers[\n layer_num]._attention_layer._output_dense.set_weights([\n bert_model[\n f\"encoder.layer.{layer_num}.attention.output.dense.weight\"].T\n .reshape((num_attention_heads, head_size, hidden_size)),\n bert_model[f\"encoder.layer.{layer_num}.attention.output.dense.bias\"]\n ])\n encoder._transformer_layers[layer_num]._attention_layer_norm.set_weights([\n bert_model[\n f\"encoder.layer.{layer_num}.attention.output.LayerNorm.weight\"],\n bert_model[f\"encoder.layer.{layer_num}.attention.output.LayerNorm.bias\"]\n ])\n encoder._transformer_layers[layer_num]._intermediate_dense.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.intermediate.dense.weight\"].T,\n bert_model[f\"encoder.layer.{layer_num}.intermediate.dense.bias\"]\n ])\n encoder._transformer_layers[layer_num]._output_dense.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.output.dense.weight\"].T,\n bert_model[f\"encoder.layer.{layer_num}.output.dense.bias\"]\n ])\n encoder._transformer_layers[layer_num]._output_layer_norm.set_weights([\n bert_model[f\"encoder.layer.{layer_num}.output.LayerNorm.weight\"],\n bert_model[f\"encoder.layer.{layer_num}.output.LayerNorm.bias\"]\n ])","def init_encoder(self):\n\n vgg = models.vgg16(pretrained=True)\n\n blocks = [self.layer_1,\n self.layer_2,\n self.layer_3,\n self.layer_4,\n self.layer_5]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data\n\n blocks = [self.layer_11,\n self.layer_12,\n self.layer_13,\n self.layer_14,\n self.layer_15]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data","def transform():","def init_encoder(self):\n\n vgg = models.vgg16(pretrained=True)\n\n blocks = [self.layer_1,\n self.layer_2,\n self.layer_3,\n self.layer_4,\n self.layer_5]\n\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\n features = list(vgg.features.children())\n\n vgg_layers = []\n for _layer in features:\n if isinstance(_layer, nn.Conv2d):\n vgg_layers.append(_layer)\n\n merged_layers = []\n for idx, conv_block in enumerate(blocks):\n if idx < 2:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit]\n else:\n units = [conv_block.conv1.cbr_unit,\n conv_block.conv2.cbr_unit,\n conv_block.conv3.cbr_unit]\n for _unit in units:\n for _layer in _unit:\n if isinstance(_layer, nn.Conv2d):\n merged_layers.append(_layer)\n\n assert len(vgg_layers) == len(merged_layers)\n\n for l1, l2 in zip(vgg_layers, merged_layers):\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\n assert l1.weight.size() == l2.weight.size()\n assert l1.bias.size() == l2.bias.size()\n l2.weight.data = l1.weight.data\n l2.bias.data = l1.bias.data","def forward(self, x):\n # Get results of encoder network\n q = self.encode_nn(x)\n\n return q","def conv_encoder(input_tensor, num_latent, is_training=True):\n del is_training\n\n e1 = tf.layers.conv2d(\n inputs=input_tensor,\n filters=32,\n kernel_size=4,\n strides=2,\n activation=tf.nn.relu,\n padding=\"same\",\n name=\"e1\",\n )\n e2 = tf.layers.conv2d(\n inputs=e1,\n filters=32,\n kernel_size=4,\n strides=2,\n activation=tf.nn.relu,\n padding=\"same\",\n name=\"e2\",\n )\n e3 = tf.layers.conv2d(\n inputs=e2,\n filters=64,\n kernel_size=2,\n strides=2,\n activation=tf.nn.relu,\n padding=\"same\",\n name=\"e3\",\n )\n e4 = tf.layers.conv2d(\n inputs=e3,\n filters=64,\n kernel_size=2,\n strides=2,\n activation=tf.nn.relu,\n padding=\"same\",\n name=\"e4\",\n )\n flat_e4 = tf.layers.flatten(e4)\n e5 = tf.layers.dense(flat_e4, 256, activation=tf.nn.relu, name=\"e5\")\n means = tf.layers.dense(e5, num_latent, activation=None, name=\"means\")\n log_var = tf.layers.dense(e5, num_latent, activation=None, name=\"log_var\")\n return means, log_var","def forward(self, x):\n # encode\n encode_block1 = self.conv_encode1(x)\n encode_pool1 = self.conv_maxpool1(encode_block1)\n encode_block2 = self.conv_encode2(encode_pool1)\n encode_pool2 = self.conv_maxpool2(encode_block2)\n encode_block3 = self.conv_encode3(encode_pool2)\n encode_pool3 = self.conv_maxpool3(encode_block3)\n # Bottleneck\n bottleneck1 = self.bottleneck(encode_pool3)\n # Decode\n decode_block3 = crop_and_concat(\n bottleneck1, encode_block3, crop=True)\n cat_layer2 = self.conv_decode3(decode_block3)\n decode_block2 = crop_and_concat(\n cat_layer2, encode_block2, crop=True)\n cat_layer1 = self.conv_decode2(decode_block2)\n decode_block1 = crop_and_concat(\n cat_layer1, encode_block1, crop=True)\n final_layer = self.final_layer(decode_block1)\n return final_layer","def transform(self):\n\n # Gets a serialized dict representation of the model, containing all its\n # layers, their connections and configuration. This is the main structure\n # which is used to understand model structure, and also manipulate it.\n #\n # config = {\n # 'input_layers': [ ... ],\n # 'layers': [{\n # 'inbound_nodes': [INPUT CONFIG OF LAYER],\n # 'name': 'LAYER_NAME',\n # 'config': { LAYER_CONFIG }\n # }, {\n # ...\n # }],\n # 'output_layers': [ ... ],\n # 'name': 'MODEL_NAME',\n #\n\n # Ensures old Keras serialization format\n self.model.use_legacy_config = True\n self._config = self.model.get_config()\n\n # Stores map of Transform -> List of layer names matched by transform.\n # Same transform should not match+replace the same layer more than once\n # to prevent infinite loops.\n self._transform_matched_layers_map = {}\n self._layer_weights_map = {}\n self._layer_names_and_weights_map = {}\n\n for layer in self.model.layers:\n self._layer_weights_map[layer.name] = self._get_keras_layer_weights(layer)\n self._layer_names_and_weights_map[\n layer.name] = self._get_keras_layer_names_and_weights(layer)\n\n # Maintains a current mutable copy of the metadata through transformation.\n self._layer_metadata_map = copy.deepcopy(self.layer_metadata)\n\n # We run an infinite loop and keep applying transformations as long as\n # patterns are found. This allows recursive pattern matching where a\n # modification by one transform may lead to another match.\n #\n # TODO(pulkitb): This leads to infinite loops with poor patterns which may\n # match their replacement. Add counters with limits to fix it.\n while True:\n match_found = False\n for transform in self.transforms:\n # A transform may find multiple instances of a pattern in the model.\n # Keep finding and replacing till done.\n while True:\n match_layer_node = self._find_pattern(\n transform.pattern(), self._get_matched_layers(transform))\n\n # Pattern did not match any layer. Move to next transform.\n if not match_layer_node:\n break\n\n self._store_successful_match(transform, match_layer_node)\n\n # Copying the match_layer_node ensures the replacement code can\n # freely modify the match.\n replacement_layer_node = transform.replacement(\n copy.deepcopy(match_layer_node))\n\n # If equal, the matched layers are being replaced with exactly the\n # same set of layers that were matched with the same config.\n # For Transforms, that may inadvertently do this we can end up in\n # an infinite loop. Move on if no meaningful change has been made.\n if match_layer_node == replacement_layer_node:\n continue\n\n match_found = True\n self._replace(match_layer_node, replacement_layer_node)\n\n # None of the transforms found a pattern. We can stop now.\n if not match_found:\n break\n\n custom_objects = {}\n for transform in self.transforms:\n custom_objects.update(transform.custom_objects())\n\n # Reconstruct model from the config, using the cloned layers.\n if self._is_functional_model(self.model):\n transformed_model = keras.Model.from_config(self._config, custom_objects)\n else:\n transformed_model = keras.Sequential.from_config(self._config,\n custom_objects)\n\n for layer in transformed_model.layers:\n weights = self._layer_weights_map.get(layer.name)\n if weights:\n self._set_layer_weights(layer, weights)\n else:\n names_and_weights = self._layer_names_and_weights_map.get(layer.name)\n if names_and_weights:\n self._set_layer_names_and_weights(layer, names_and_weights)\n\n # Ensures old Keras serialization format\n transformed_model.use_legacy_config = True\n return transformed_model, copy.deepcopy(self._layer_metadata_map)","def SemiAutoencoder(ds, compression_factor=16, input_noise=0.2, dropout_p=0.1, activ='tanh', final_activ='tanh'):\n # compression_factor=20\n print('DS shape: {}'.format(ds.shape))\n in_dims = np.prod(ds.shape[1:])\n encoding_dim = int(in_dims // compression_factor)\n in_shape = ds[0].shape\n print('Input Dims: {}, input shape: {}, encoding dims: {}'.format(in_dims, in_shape, encoding_dim))\n\n # this is our input placeholder\n input_img = Input(shape=(in_dims,))\n encoded = GaussianNoise(input_noise)(input_img)\n\n encoded = Dense(encoding_dim * 4, activation=activ, activity_regularizer=regularizers.activity_l1(10e-5))(encoded)\n # encoded = Dense(encoding_dim*4, activation='sigmoid')(input_img)\n encoded = BatchNormalization()(encoded)\n encoded = Dropout(dropout_p)(encoded) # batch norm before dropout\n # encoded = Dense(encoding_dim*3, activation=activ)(encoded)\n # encoded = Dropout(dropout_p)(encoded)\n encoded = Dense(encoding_dim * 2, activation=activ)(encoded)\n encoded = Dropout(dropout_p)(encoded)\n\n encoded = Dense(encoding_dim, activation=activ)(encoded)\n # Middle Noise\n encoded = GaussianNoise(0.02)(encoded)\n\n # DECODED LAYER\n # \"decoded\" is the lossy reconstruction of the input\n decoded = Dense(encoding_dim * 2, activation=activ)(encoded)\n # decoded = Dropout(dropout_p)(decoded)\n decoded = Dense(encoding_dim * 4, activation=activ)(decoded)\n decoded = Dropout(dropout_p)(decoded)\n decoded = Dense(in_dims, activation=final_activ)(decoded)\n\n # MODEL\n autoencoder = Model(input=input_img, output=decoded)\n\n # SEPERATE ENCODER MODEL\n encoder = Model(input=input_img, output=encoded)\n\n # create a placeholder for an encoded (32-dimensional) input\n encoded_input = Input(shape=(encoding_dim,))\n\n # retrieve the last layer of the autoencoder model\n decoder_layer0 = autoencoder.layers[-4]\n decoder_layer1 = autoencoder.layers[-3]\n decoder_layer2 = autoencoder.layers[-2]\n decoder_layer3 = autoencoder.layers[-1]\n # todo: make this into a dedicated unrolling function\n\n # create the decoder model - unrolling the model as we go\n decoder = Model(input=encoded_input, output=decoder_layer3(decoder_layer2(\n decoder_layer1(decoder_layer0(encoded_input)))))\n\n # model.add(GaussianNoise(0.1), input_shape=(n_input_len,))\n autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')\n\n autoencoder.model_name = 'Autoencoder 1'\n return autoencoder, encoder, decoder","def make_encoder(self, input_size, latent_size):\n return nn.Sequential(\n nn.Linear(input_size, 500),\n nn.ReLU(),\n nn.Linear(500, 200),\n nn.ReLU(),\n nn.Linear(200, latent_size * 2)\n # times 2 because this is the concatenated vector of latent mean and variance\n )","def _encoder(self, z_dimension, mcd):\n inputs = Input(shape=(2 * self.n_dims,), name=\"encoder_input\")\n x = inputs\n x = Dense(\n self.hidden_size,\n activation=self.activation_func,\n kernel_initializer=self.weight_init,\n )(x)\n x = Dropout(self.dropout_probability)(x, training=mcd)\n x = Dense(\n self.hidden_size // 2,\n activation=self.activation_func,\n kernel_initializer=self.weight_init,\n )(x)\n x = Dropout(self.dropout_probability)(x, training=mcd)\n\n x = Dense(\n self.hidden_size // 3,\n activation=self.activation_func,\n kernel_initializer=self.weight_init,\n )(x)\n x = Dropout(self.dropout_probability)(x, training=mcd)\n\n x = Dense(\n self.hidden_size // 4,\n activation=self.activation_func,\n kernel_initializer=self.weight_init,\n )(x)\n x = Dropout(self.dropout_probability)(x, training=mcd)\n\n # Latent mean\n z_mean = Dense(z_dimension, name=\"z_mean\")(x)\n \n # latent variance\n z_var = Dense(z_dimension, name=\"z_var\")(x)\n self.encoder = Model(inputs, [z_mean, z_var], name=\"encoder\")\n\n return self.encoder, inputs, z_mean, z_var","def convolutional_32_encoder(self, x, reuse=True): \n\t\tenc_layers = self.enc_layers\n\t\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\n\t\twith tf.variable_scope('model', reuse=reuse):\n\t\t\twith tf.variable_scope('encoder', reuse=reuse):\n\t\t\t\tconv1 = tf.layers.conv2d(x, 128, 2, strides=(2,2), padding=\"VALID\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv1')\n\t\t\t\tconv1 = tf.layers.batch_normalization(conv1)\n\n\t\t\t\tconv2 = tf.layers.conv2d(conv1, 256, 2, strides=(2,2), padding=\"VALID\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv2')\n\t\t\t\tconv2 = tf.layers.batch_normalization(conv2)\n\n\t\t\t\tconv3 = tf.layers.conv2d(conv2, 512, 2, strides=(2,2), padding=\"VALID\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv3')\n\t\t\t\tconv3 = tf.layers.batch_normalization(conv3)\n\n\t\t\t\tflattened = tf.contrib.layers.flatten(conv3)\n\t\t\t\tz_mean = tf.layers.dense(flattened, enc_layers[-1], activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-final')\n\t\treturn z_mean","def _encode(self):\n with tf.variable_scope('encoding'):\n self.sep_p_encodes, _ = bilstm_layer(self.p_emb, self.p_length, self.hidden_size)\n tf.get_variable_scope().reuse_variables()\n self.sep_q_encodes, _ = bilstm_layer(self.q_emb, self.q_length, self.hidden_size)\n if self.use_dropout:\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, 1-self.dropout)\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, 1-self.dropout)","def encode_input_for_decoder(x_tensor, inp_lens_tensor, model_input_emb: EmbeddingLayer, model_enc: RNNEncoder):\n input_emb = model_input_emb.forward(x_tensor)\n (enc_output_each_word, enc_context_mask, enc_final_states) = model_enc.forward(input_emb, inp_lens_tensor)\n enc_final_states_reshaped = (enc_final_states[0].unsqueeze(0), enc_final_states[1].unsqueeze(0))\n return (enc_output_each_word, enc_context_mask, enc_final_states_reshaped)","def forward(self, inputs_encoder, inputs_decoder):\n states_encoder = self.encoder(inputs_encoder)\n outputs_decoder, states_decoder = self.decoder(inputs_decoder, states_encoder)\n return outputs_decoder, states_decoder","def forward(self, *args):\r\n enc_src, _, _ = self.unified_encoder(*args)\r\n enc_src = enc_src.view(enc_src.shape[0], -1)\r\n y_pred = self.mlp(enc_src)\r\n return y_pred","def vae_encoder(self):\n return stax.serial(\n stax.Dense(self.hidden_dims[0], W_init=stax.randn()),\n stax.Relu,\n stax.Dense(self.hidden_dims[1], W_init=stax.randn()),\n stax.Relu,\n stax.FanOut(2),\n stax.parallel(\n stax.Dense(self.z_dim, W_init=stax.randn()), # mean\n stax.serial(stax.Dense(self.z_dim, W_init=stax.randn()), stax.Exp), # std -- i.e. diagonal covariance\n ),\n )","def setup_layers(self, params):\n\n #\n # Input layer\n #\n\n # Input and encoding layers\n layer_stack = [\n tf.placeholder(\n tf.int32,\n name=\"input\",\n shape=[None, params.in_size]\n )\n ]\n\n in_layer = layer_stack[0]\n\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\n character_depth = len(good_chars) + 2\n\n layer_stack.append(\n self.make_onehot_encode_layer(\n layer_stack[-1],\n character_depth\n )\n )\n\n # Add conv layers\n for idx, num_nodes in enumerate(params.conv_nodes):\n layer_stack.append(\n self.make_conv_layer(\n layer_stack[-1],\n num_nodes,\n 'conv_{}'.format(idx),\n params.conv_params\n )\n )\n\n # Flatten convolutional layers\n layer_stack.append(\n tf.contrib.layers.flatten(\n layer_stack[-1]\n )\n )\n\n # Add dense layers\n for idx, num_nodes in enumerate(params.dense_nodes):\n layer_stack.append(\n self.make_dense_layer(\n layer_stack[-1],\n num_nodes,\n 'dense_{}'.format(idx),\n params.dense_params\n )\n )\n\n # Output layer is broken into two pieces. The pre/post the application\n # of the activation function. That's because the loss functions\n # for cross-entropy rely on the pre-activation scores in preact\n\n preact_out_layer, out_layer = self.make_dense_output_layer(\n layer_stack[-1],\n params.out_size,\n params.output_params\n )\n\n target_layer = tf.placeholder(\n \"float\",\n name=\"target\",\n shape=[None, params.out_size]\n )\n\n if params.output_params.activation in ['sigmoid']:\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\n else:\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\n\n return in_layer, out_layer, target_layer, loss","def _fit_encoder(self, frame, prop, encoder_type=\"category\"):\n pass","def __init__(self):\n super(Encoder, self).__init__()\n\n self.lconv1 = nn.Sequential(\n nn.Conv1d(4, 64, kernel_size=9, padding=4),\n nn.BatchNorm1d(64),\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\n nn.BatchNorm1d(64),\n )\n\n self.conv1 = nn.Sequential(\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\n nn.BatchNorm1d(64),\n nn.ReLU(inplace=True),\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\n nn.BatchNorm1d(64),\n nn.ReLU(inplace=True),\n )\n\n self.lconv2 = nn.Sequential(\n nn.MaxPool1d(kernel_size=4, stride=4),\n nn.Conv1d(64, 96, kernel_size=9, padding=4),\n nn.BatchNorm1d(96),\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\n nn.BatchNorm1d(96),\n )\n\n self.conv2 = nn.Sequential(\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\n nn.BatchNorm1d(96),\n nn.ReLU(inplace=True),\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\n nn.BatchNorm1d(96),\n nn.ReLU(inplace=True),\n )\n\n self.lconv3 = nn.Sequential(\n nn.MaxPool1d(kernel_size=4, stride=4),\n nn.Conv1d(96, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n )\n\n self.conv3 = nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n )\n\n self.lconv4 = nn.Sequential(\n nn.MaxPool1d(kernel_size=5, stride=5),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n )\n\n self.conv4 = nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n )\n\n self.lconv5 = nn.Sequential(\n nn.MaxPool1d(kernel_size=5, stride=5),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n )\n\n self.conv5 = nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n )\n\n self.lconv6 = nn.Sequential(\n nn.MaxPool1d(kernel_size=5, stride=5),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n )\n\n self.conv6 = nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n )\n\n self.lconv7 = nn.Sequential(\n nn.MaxPool1d(kernel_size=2, stride=2),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n )\n\n self.conv7 = nn.Sequential(\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\n nn.BatchNorm1d(128),\n nn.ReLU(inplace=True),\n )","def build_autoencoder(input_dim):\r\n input_layer = Input(shape=(input_dim, 1))\r\n enc = Conv1D(filters=16, kernel_size=2, padding='same', activation='relu')(input_layer)\r\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\r\n enc = Conv1D(filters=32, kernel_size=2, padding='same', activation='relu')(enc)\r\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\r\n enc = Conv1D(filters=64, kernel_size=2, padding='same', activation='relu')(enc)\r\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\r\n enc = Flatten()(enc)\r\n enc = Dense(64)(enc)\r\n\r\n dec = Dense(200704)(enc)\r\n dec = Reshape((3136, 64))(dec)\r\n dec = Conv1D(filters=64, kernel_size=2, padding='same', activation='relu')(dec)\r\n dec = UpSampling1D(2)(dec)\r\n dec = Conv1D(filters=32, kernel_size=2, padding='same', activation='relu')(dec)\r\n dec = UpSampling1D(2)(dec)\r\n dec = Conv1D(filters=16, kernel_size=2, padding='same', activation='relu')(dec)\r\n dec = UpSampling1D(2)(dec)\r\n dec = Conv1D(filters=1, kernel_size=2, padding='same', activation='relu')(dec)\r\n\r\n autoencoder = Model(input_layer, dec)\r\n autoencoder.compile(optimizer='adam', loss='mean_squared_error', metrics=['accuracy'])\r\n autoencoder.summary()\r\n encoder = Model(input_layer, enc)\r\n return autoencoder, encoder","def base_encoder(cls, data, data_mask, init_encoder, downsize_encoder, input_encoder):\n data = downsize_encoder(init_encoder(data))\n return input_encoder(data, data_mask)","def __init__(self, AEs):\n \"\"\" the default view of the stacked autoencoders\"\"\"\n sa = AEs\n \"\"\" the encoder view of the stacked autoencoders \"\"\"\n ec = Cat([a.ec for a in sa])\n \"\"\" the decoder view of the stacked autoencoders \"\"\"\n dc = Cat([a.dc for a in reversed(sa)])\n\n self.sa = sa # default view\n self.ec = ec # encoder view\n self.dc = dc # decoder view\n\n nts = []\n nts.extend(ec)\n nts.extend(dc)\n super(SAE, self).__init__(nts)","def encode(self, x):\n h1 = F.relu(self.conv1(x))\n h2 = F.relu(self.conv2(h1))\n h3 = F.relu(self.conv3(h2))\n h4 = F.relu(self.conv4(h3))\n h4 = h4.reshape(-1)\n mu = self.fc1(h4)\n logvar = self.fc2(h4)\n\n # we take log of var as sigma cannot be negative \n # while neural network can output negative values\n sigma = torch.exp(logvar * 0.5)\n\n normal = torch.randn_like(sigma)\n\n return mu + sigma * normal, mu, logvar","def encoder(self, value):\n self._tensor.encoder = value","def transform(self, inputs: list, stage: str) -> datapack.DataPack:","def forward(self, x, neg=None):\n if neg is None:\n return self._target_encoder(tf.reshape(x, [-1]))\n else:\n return self._target_encoder(tf.reshape(x, [-1])), \\\n self._context_encoder(tf.reshape(x, [-1])), \\\n self._context_encoder(tf.reshape(neg, [-1]))","def transform(self):","def fc_encoder(input_tensor, num_latent, is_training=True):\n del is_training\n\n flattened = tf.layers.flatten(input_tensor)\n e1 = tf.layers.dense(flattened, 1200, activation=tf.nn.relu, name=\"e1\")\n e2 = tf.layers.dense(e1, 1200, activation=tf.nn.relu, name=\"e2\")\n means = tf.layers.dense(e2, num_latent, activation=None)\n log_var = tf.layers.dense(e2, num_latent, activation=None)\n return means, log_var","def encode(self, x, encode_params, name=None):\n values = [x] + list(encode_params)\n with tf.compat.v1.name_scope(name, 'gather_encoder_encode', values):\n x = tf.convert_to_tensor(x)\n encode_params = tf.nest.map_structure(tf.convert_to_tensor, encode_params)\n return self._encode_fn(x, encode_params)","def encoder_decoder_archi_gan(inputs, is_train):\n\n encoder_layers = []\n\n encoded = inputs\n\n encoder_layers.append(encoded)\n\n for i in range(config.encoder_layers):\n encoded = encoder_conv_block_gan(encoded, i, is_train)\n encoder_layers.append(encoded)\n \n encoder_layers.reverse()\n\n\n\n decoded = encoder_layers[0]\n\n for i in range(config.encoder_layers):\n decoded = decoder_conv_block_gan(decoded, encoder_layers[i+1], i, is_train)\n\n return decoded","def encoder(inputs):\n out = tf.layers.flatten(inputs)\n out = tf.layers.dense(out, 400, activation=tf.nn.relu)\n mean = tf.layers.dense(out, LATENT_SIZE)\n logstd = tf.layers.dense(out, LATENT_SIZE)\n return tf.distributions.Normal(loc=mean, scale=tf.exp(logstd))","def _build_encoder(self, hparams, is_training):\n with tf.variable_scope(\"trajectory_encoder\"):\n with tf.name_scope(\"source_placeholder\"):\n input_phs = list_ops.list_placeholder(self.num_gpu, (None, self.input_length, self.input_dims), tf.float32)\n for ph in input_phs:\n tf.add_to_collection('placeholder', ph)\n \n if hparams.encoder_type == \"rnn\":\n net = input_phs\n with tf.variable_scope(\"projection\"):\n net = self._build_input_projection(hparams, net, is_training)\n\n with tf.name_scope(\"batch_time_transpose\"):\n net = list_ops.list_transpose(net, perm=[1, 0, 2])\n\n with tf.variable_scope(\"rnn\"):\n net, state = self._build_rnn_encoder(hparams, net, is_training)\n\n if hparams.relu_reconfiguration:\n with tf.variable_scope(\"reconfiguration\"):\n net = list_ops.list_dense_with_bn(net,\n hparams.cnn_input_projector_filters[-1],\n is_training,\n self.bn_decay,\n seed=self.random_seed)\n\n elif hparams.encoder_type == \"cnn\":\n net = self._build_cnn_encoder(hparams, input_phs, is_training)\n state = None\n \n else:\n raise ValueError(\"Unknown encoder type {:s}.\".format(hparams.encoder_type))\n\n return net, state","def transform(self, X):\n ...","def transform(self, X):\n ..."],"string":"[\n \"def encoder(enc_input, attn_bias, n_layer, n_head,\\n d_key, d_value, d_model, d_inner_hid, pos_enc,\\n preporstprocess_dropout, attention_dropout,\\n relu_dropout, preprocess_cmd='n',\\n postprocess_cmd='da'):\\n for i in range(n_layer):\\n enc_output = encoder_layer(enc_input, attn_bias, n_head,\\n d_key, d_value, d_model,d_inner_hid, pos_enc,\\n prepostprocess_dropout, attention_dropout,relu_dropout,\\n preprocess_cmd, postprocess_cmd\\n )\\n enc_input = enc_output\\n enc_output = pre_process_layer(enc_output,\\n preprocess_cmd, preporstprocess_dropout)\\n return enc_output\",\n \"def encoder(self, inputs):\\n pass\",\n \"def build_encoder(opt, embeddings, structure_embeddings):\\n return TransformerEncoder(opt.enc_layers, opt.enc_rnn_size, opt.heads, opt.transformer_ff, opt.dropout, embeddings, structure_embeddings)\",\n \"def _define_encoder(self):\\n self.encoder = nn.Sequential(nn.Conv2d(in_channels=3, out_channels=32, kernel_size=4, stride=2, padding=1), # B, 32, 32, 32\\n nn.SELU(),\\n nn.Conv2d(32, 32, 4, 2, 1), # B, 32, 16, 16\\n nn.SELU(),\\n nn.Conv2d(32, 64, 4, 2, 1), # B, 64, 8, 8\\n nn.SELU(),\\n nn.Conv2d(64, 64, 4, 2, 1), # B, 64, 4, 4\\n nn.SELU(),\\n nn.Conv2d(64, 256, 4, 1), # B, 256, 1, 1\\n nn.SELU(),\\n View((-1, 256 * 1 * 1)), # B, 256\\n nn.Linear(256, self.encoding_shape * 2), # B, z_dim*2\\n )\",\n \"def _define_encoder(self):\\n self.encoder = nn.Sequential(View((-1, 64 * 64 * 3)),\\n nn.Linear(64 * 64 * 3, 5120, bias=False), nn.SELU(),\\n nn.BatchNorm1d(5120),\\n nn.Linear(5120, 2560, bias=False), nn.SELU(),\\n nn.BatchNorm1d(2560),\\n nn.Linear(2560, 512, bias=False), nn.SELU(),\\n nn.BatchNorm1d(512),\\n nn.Linear(512, 128, bias=False), nn.SELU(),\\n nn.BatchNorm1d(128),\\n nn.Linear(128, self.encoding_shape, bias=False), nn.SELU(),\\n )\",\n \"def fit_transform(self):\\n if self.enc_types == \\\"label\\\":\\n return self._label_encoding()\\n elif self.enc_types == \\\"ohe\\\":\\n return self._one_hot_encoder()\\n elif self.enc_types == \\\"binary\\\":\\n return self._binarization()\\n else:\\n raise Exception(\\\"Encoding type not understood\\\")\",\n \"def encoder_layer(enc_input, attn_bias, n_head, d_key,\\n d_value, d_model, d_inner_hid, pos_enc, prepostprocess_dropout,\\n attention_dropout, relu_dropout, preprocess_cmd='n',\\n postprocess_cmd='da'):\\n attn_output = multi_head_attention(\\n pre_process_layer(enc_input, preprocess_cmd, prepostprocess_dropout),\\n None, None, attn_bias, d_key, d_value, d_model, pos_enc,\\n n_head, attention_dropout\\n )\\n attn_output = post_process_layer(enc_input, attn_output,\\n postprocess_cmd, prepostprocess_dropout)\\n ffd_output = positionwise_feed_forward(\\n pre_process_layer(attn_output, preprocess_cmd, prepostprocess_dropout),\\n d_inner_hid, d_model, relu_dropout\\n )\\n return post_process_layer(attn_output, ffd_output,\\n postprocess_cmd, prepostprocess_dropout)\",\n \"def TransformerEncoderLayer(self, name, is_causal=True):\\n # Hack to be compatible with ckpt generated by self._rep\\n return self._Seq(\\n name,\\n self._Seq('block', self._Attention('self_atten', is_causal=is_causal),\\n self.Feedforward('ff')))\",\n \"def TransformerEncoderStack(self, name, num_layers=1):\\n blocks = [\\n self.TransformerEncoderLayer(name='iter_{:0>3d}'.format(d))\\n for d in range(num_layers)\\n ]\\n return self.Stack(name, blocks)\",\n \"def encode(self):\\n with tf.name_scope(\\\"encode\\\"):\\n self.encoder_inputs = tf.layers.dense(\\n inputs=self.encoder_inputs,\\n units=self.options['hidden_size'], activation=None, use_bias=True,\\n kernel_initializer=tf.keras.initializers.he_normal(seed=None),\\n bias_initializer=tf.zeros_initializer(),\\n kernel_regularizer=None, bias_regularizer=None, activity_regularizer=None,\\n kernel_constraint=None, bias_constraint=None, trainable=True,\\n name=None, reuse=None)\\n self.encoder_inputs = tf.layers.batch_normalization(self.encoder_inputs,\\n axis=-1,\\n momentum=0.99,\\n epsilon=0.001,\\n center=True,\\n scale=True,\\n beta_initializer=tf.zeros_initializer(),\\n gamma_initializer=tf.ones_initializer(),\\n moving_mean_initializer=tf.zeros_initializer(),\\n moving_variance_initializer=tf.ones_initializer(),\\n training=self.is_training,\\n trainable=True,\\n renorm=False,\\n renorm_momentum=0.99)\\n # Prepare inputs to the layer stack by adding positional encodings and\\n # applying dropout.\\n # embedded_inputs = self.embedding_softmax_layer(inputs)\\n #\\n inputs_padding = transformer_model_utils.get_padding(tf.cast(\\n tf.reduce_max(100*self.encoder_inputs, [-1]),\\n dtype=tf.int32))\\n\\n with tf.name_scope(\\\"add_pos_encoding\\\"):\\n length = tf.shape(self.encoder_inputs)[1]\\n pos_encoding = transformer_model_utils.get_position_encoding(\\n length, self.options[\\\"hidden_size\\\"])\\n encoder_inputs = self.encoder_inputs + pos_encoding\\n\\n if self.is_training:\\n encoder_inputs = tf.nn.dropout(\\n encoder_inputs, 1 - self.options[\\\"layer_postprocess_dropout\\\"])\\n\\n return self.encoder_stack(encoder_inputs, self.attention_bias, inputs_padding)\",\n \"def build_encoder(tparams, options):\\n\\t# word embedding (source)\\n\\tembedding = tensor.tensor3('embedding', dtype='float32')\\n\\tx_mask = tensor.matrix('x_mask', dtype='float32')\\n\\n\\t# encoder\\n\\tproj = get_layer(options['encoder'])[1](tparams, embedding, options,\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\tprefix='encoder',\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\tmask=x_mask)\\n\\tctx = proj[0][-1]\\n\\n\\treturn embedding, x_mask, ctx\",\n \"def _create_encoder(self):\\n\\n def _init_weights(layer):\\n \\\"\\\"\\\"Initializes the weights of a layer based on type.\\\"\\\"\\\"\\n if isinstance(layer, (nn.Conv2d, nn.Linear)):\\n torch.nn.init.xavier_uniform_(layer.weight)\\n try:\\n # Some layers may not have biases, so catch the exception and pass.\\n layer.bias.data.fill_(0.0)\\n except AttributeError:\\n pass\\n\\n kernel_size = 5\\n pad = 2\\n input_channels = 1\\n first_conv_channels = 6\\n second_conv_channels = 16\\n max_pool_kernel = 2\\n linear_size = 120\\n n_pixels = 7\\n\\n encoder = nn.Sequential(\\n nn.Conv2d(\\n input_channels, first_conv_channels, kernel_size, padding=pad),\\n nn.BatchNorm2d(first_conv_channels),\\n nn.ReLU(),\\n nn.MaxPool2d(max_pool_kernel),\\n nn.Conv2d(\\n first_conv_channels, second_conv_channels, kernel_size,\\n padding=pad),\\n nn.BatchNorm2d(second_conv_channels),\\n nn.ReLU(),\\n nn.MaxPool2d(max_pool_kernel),\\n utils.Flatten(),\\n nn.Linear(n_pixels * n_pixels * self.n_digits * second_conv_channels,\\n linear_size),\\n nn.BatchNorm1d(linear_size),\\n nn.ReLU(),\\n nn.Linear(linear_size, self.embedding_dim),\\n nn.Linear(self.embedding_dim, self.n_classes, bias=False),\\n )\\n\\n encoder.apply(_init_weights)\\n\\n # This is the empirical approximation for initialization the vMF\\n # distributions for each class in the final layer.\\n if self.use_vmf:\\n utils.vmf_class_weight_init(encoder[-1].weight, self.kappa_confidence,\\n self.embedding_dim)\\n\\n return encoder\",\n \"def encode(data, encoder):\\n # Get the list of hidden depths\\n\\thd = encoder.hidden_depths\\n # Find the middle hidden layer\\n\\tmiddle_layer_index = (len(hd)-1)/2\\n # Initialize empty container for the encoded data\\n\\tdata_encoded = np.zeros((data.shape[0],hd[middle_layer_index]))\\n\\tfor i, d_ in enumerate(data):\\n # feed forward, get all the activations, and just keep\\n # the middle layer, which is the encoding\\n\\t\\tx, z_container, x_container = encoder.ff(d_,True,True)\\n\\t\\tx_encoded = x_container[1+middle_layer_index]\\n\\t\\tdata_encoded[i] = x_encoded\\n\\t#\\n\\treturn data_encoded\",\n \"def make_encoder(self, input_size: int, latent_size: int) -> nn.Module:\\n pass\",\n \"def __init__(self,\\n input_dim,\\n dec_seq_len,\\n out_seq_len,\\n d_model=512,\\n nhead=8,\\n num_encoder_layers=6,\\n num_decoder_layers=6,\\n dim_feedforward=2048,\\n dropout=0.1,\\n activation='relu',\\n custom_encoder=None,\\n custom_decoder=None):\\n super(TransformerTS, self).__init__()\\n self.transform = nn.Transformer(\\n d_model=d_model,\\n nhead=nhead,\\n num_encoder_layers=num_encoder_layers,\\n num_decoder_layers=num_decoder_layers,\\n dim_feedforward=dim_feedforward,\\n dropout=dropout,\\n activation=activation,\\n custom_encoder=custom_encoder,\\n custom_decoder=custom_decoder\\n )\\n self.pos = PositionalEncoding(d_model)\\n self.enc_input_fc = nn.Linear(input_dim, d_model)\\n self.dec_input_fc = nn.Linear(input_dim, d_model)\\n self.out_fc = nn.Linear(dec_seq_len * d_model, out_seq_len)\\n self.dec_seq_len = dec_seq_len\",\n \"def TransformerEncoderLayer(self, name, stride=1, first_n=None,\\n ff_hidden_dim=None, num_heads=None,\\n use_moe=False):\\n p = self.params\\n if ff_hidden_dim is None:\\n ff_hidden_dim = p.ff_hidden_dim\\n if num_heads is None:\\n num_heads = p.num_heads\\n if use_moe:\\n ffw_p = self.MoE('moe', ff_hidden_dim=ff_hidden_dim)\\n else:\\n ffw_p = self.Feedforward('ff', ff_hidden_dim=ff_hidden_dim)\\n return self._Seq(name, self._Seq(\\n 'block',\\n self._StridedAttention('self_atten', stride=stride,\\n first_n=first_n, num_heads=num_heads),\\n ffw_p))\",\n \"def beam_init(self, encoder_outputs, K):\\n encoder_outputs = TransformerEncoderOutput(\\n repeat_batch(encoder_outputs.output, K),\\n repeat_batch(encoder_outputs.src_mask, K)\\n )\\n return encoder_outputs\",\n \"def transformer_prepare_encoder(inputs, hparams, features=None):\\n ishape_static = inputs.shape.as_list()\\n encoder_input = inputs\\n\\n # Usual case - not a packed dataset.\\n encoder_padding = common_attention.embedding_to_padding(encoder_input)\\n ignore_padding = common_attention.attention_bias_ignore_padding(\\n encoder_padding)\\n encoder_self_attention_bias = ignore_padding\\n encoder_decoder_attention_bias = ignore_padding\\n inputs_position = None\\n if hparams.proximity_bias:\\n encoder_self_attention_bias += common_attention.attention_bias_proximal(\\n common_layers.shape_list(inputs)[1])\\n\\n if inputs_position is not None:\\n encoder_input = common_attention.add_timing_signal_1d_given_position(\\n encoder_input, inputs_position)\\n else:\\n encoder_input = common_attention.add_timing_signal_1d(\\n encoder_input)\\n return (encoder_input, encoder_self_attention_bias,\\n encoder_decoder_attention_bias)\",\n \"def encode(self, img):\\n with tf.variable_scope('encoder'):\\n #conv1 = self.conv_layer(\\n # img, [5, 5], [3, 32], stride=2, initializer_type=1, name='conv1')\\n #conv2 = self.conv_layer(\\n # conv1, [5, 5], [32, 32], stride=2, initializer_type=1, name='conv2')\\n conv3 = self.conv_layer(\\n img, [5, 5], [self.in_channels, 64], stride=2, initializer_type=1, name='conv3')\\n #conv4 = self.conv_bn_layer(\\n conv4 = self.conv_layer(\\n conv3, [5, 5], [64, 128], stride=2, #is_training=self.is_training,\\n initializer_type=1, name='conv4')\\n #conv5 = self.conv_bn_layer(\\n conv5 = self.conv_layer(\\n conv4, [5, 5], [128, 256], stride=2, #is_training=self.is_training,\\n initializer_type=1, name='conv5')\\n shape = conv5.get_shape().as_list()\\n feature_map_size = shape[1]*shape[2]*shape[3]\\n conv5_flat = tf.reshape(\\n conv5, [-1, feature_map_size], 'conv5_flat')\\n #fc6 = self.fc_bn_layer(conv5_flat, 1024, is_training=self.is_training,\\n fc6 = self.fc_layer(conv5_flat, 1024,\\n initializer_type=1, name='fc6')\\n #fc7 = self.fc_layer(fc6, 1024, initializer_type=1, name='fc7')\\n return fc6, shape\",\n \"def encode(self, inputs, target_space, hparams, features=None, losses=None):\\n del losses\\n\\n inputs = common_layers.flatten4d3d(inputs)\\n\\n encoder_input, self_attention_bias, encoder_decoder_attention_bias = (\\n transformer.transformer_prepare_encoder(\\n inputs, target_space, hparams, features=features))\\n\\n encoder_input = tf.nn.dropout(encoder_input,\\n 1.0 - hparams.layer_prepostprocess_dropout)\\n\\n (encoder_output, encoder_extra_output) = (\\n my_universal_transformer_util.universal_transformer_encoder(\\n encoder_input,\\n self_attention_bias,\\n hparams,\\n nonpadding=transformer.features_to_nonpadding(features, \\\"inputs\\\"),\\n save_weights_to=self.attention_weights))\\n\\n return encoder_output, encoder_decoder_attention_bias, encoder_extra_output\",\n \"def forward(self, src, mask):\\n bs = src.shape[0]\\n src = src.permute(2, 0, 1)\\n m = src \\n enc_embed = self.enc_embed.weight.unsqueeze(1).repeat(1, bs, 1)\\n for layer in self.encoder_layers:\\n m = layer(m,\\n pos=enc_embed,\\n src_mask = mask\\n )\\n return m.permute(1, 2, 0), enc_embed.permute(1, 2, 0)\",\n \"def encode_all(self, inputs, encoder):\\n input_shape = tf.shape(inputs)\\n num_timesteps, batch_size = input_shape[0], input_shape[1]\\n reshaped_inputs = tf.reshape(inputs, [-1, inputs.shape[-1]])\\n inputs_encoded = encoder(reshaped_inputs)\\n inputs_encoded = tf.reshape(inputs_encoded, [num_timesteps, batch_size, encoder.output_size])\\n return inputs_encoded\",\n \"def encoder_model(self, input_shape, layers, filters, latent_dims):\\n\\n l = []\\n l.append(Input(shape=input_shape, name='encoder_input'))\\n\\n for i in range(0,layers):\\n l.append(Conv2D(filters[i], (5,5), padding='same',\\n data_format='channels_last', name='Conv_'+str(i),\\n activation='relu')(l[i*2]))\\n l.append(MaxPooling2D(pool_size=(2, 2),\\n data_format='channels_last',\\n name='Pool_'+str(i))(l[i*2+1]))\\n\\n conv_out = Flatten(name='Flatten_dims')(l[-1])\\n\\n self.z_mean = Dense(latent_dims, name='z_mean')(conv_out)\\n self.z_sigma = Dense(latent_dims, name='z_sigma')(conv_out)\\n self.z = Lambda(VAE_sampling, output_shape=(latent_dims,), name='z')([self.z_mean,self.z_sigma])\\n encoder = Model(l[0], [self.z_mean,self.z_sigma,self.z], name='encoder')\\n\\n encoder.summary()\\n\\n return encoder\",\n \"def __init__(self, args):\\n super(Transformer, self).__init__()\\n\\n self.name = 'Transformer'\\n if len(args.max_relative_pos) != args.nlayers:\\n assert len(args.max_relative_pos) == 1\\n args.max_relative_pos = args.max_relative_pos * args.nlayers\\n\\n self.embedder = Embedder(args)\\n self.encoder = Encoder(args, self.embedder.enc_input_size)\\n self.decoder_loc = nn.Linear(self.embedder.enc_input_size, 1)\\n self.decoder_fix = nn.Linear(self.embedder.enc_input_size, 1)\\n self.nobug_embedding = nn.Parameter(torch.zeros(1))\\n self.criterion = nn.CrossEntropyLoss(reduction='none')\",\n \"def build_encoder_bi(tparams, options):\\n\\t# word embedding (source)\\n\\tembedding = tensor.tensor3('embedding', dtype='float32')\\n\\tembeddingr = embedding[::-1]\\n\\tx_mask = tensor.matrix('x_mask', dtype='float32')\\n\\txr_mask = x_mask[::-1]\\n\\n\\t# encoder\\n\\tproj = get_layer(options['encoder'])[1](tparams, embedding, options,\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\tprefix='encoder',\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\tmask=x_mask)\\n\\tprojr = get_layer(options['encoder'])[1](tparams, embeddingr, options,\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t prefix='encoder_r',\\n\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t\\t mask=xr_mask)\\n\\n\\tctx = tensor.concatenate([proj[0][-1], projr[0][-1]], axis=1)\\n\\n\\treturn embedding, x_mask, ctx\",\n \"def build_encoder(img_shape):\\n input_img = Input(shape=(img_shape)) \\n x = Conv2D(16, (3, 3), activation='tanh', padding='same')(input_img)\\n x = AveragePooling2D((2, 2), padding='same')(x)\\n x = Conv2D(8, (3, 3), padding='same')(x)\\n x = LeakyReLU()(x)\\n x = AveragePooling2D((2, 2), padding='same')(x)\\n x = Conv2D(8, (3, 3), padding='same')(x)\\n x = LeakyReLU()(x)\\n x = Dropout(0.5)(x)\\n x = AveragePooling2D((2, 2), padding='same')(x)\\n x = Conv2D(1, (3, 3), padding='same')(x)\\n x = LeakyReLU()(x)\\n x = AveragePooling2D((3, 3), padding='same')(x)\\n x = Flatten()(x)\\n encoded = Dense(540, activation='tanh')(x)\\n Encoder=Model(input_img,encoded,name='encoder')\\n return input_img,encoded,Encoder\",\n \"def encoder(self, x):\\n # Input (batch_size, Channels=1, Width=28, Height=28)\\n x = self.enc_conv(x) # (batch_size, lin_before_latent_dim)\\n mean = self.enc_mean(x) # (batch_size, latent_dim)\\n logvar = self.enc_logvar(x) # (batch_size, latent_dim)\\n return mean, logvar\",\n \"def forward(self, x):\\n # Encoder1 --block1\\n encode_block1 = self.conv_encoder1(x)\\n if self.residus[0] == 1:\\n encode_block1 += self.residual_shortcut1(x)\\n encode_pool1 = self.max_pool_encoder1(encode_block1)\\n\\n # Encoder2 --block2\\n encode_block2 = self.conv_encoder2(encode_pool1)\\n if self.residus[1] == 1:\\n encode_block2 += self.residual_shortcut2(encode_pool1)\\n encode_pool2 = self.max_pool_encoder2(encode_block2)\\n \\n # Encoder3 --block3\\n encode_block3 = self.conv_encoder3(encode_pool2)\\n if self.residus[2] == 1:\\n encode_block3 += self.residual_shortcut3(encode_pool2) ##\\n encode_pool3 = self.max_pool_encoder3(encode_block3)\\n \\n # Encoder4 --block4\\n encode_block4 = self.conv_encoder4(encode_pool3) \\n if self.residus[3] == 1:\\n encode_block4 += self.residual_shortcut4(encode_pool3) ##\\n encode_pool4 = self.max_pool_encoder4(encode_block4)\\n\\n # Transitional block \\n encode_block_trans_1 = self.conv_encoder_trans_1(encode_pool4)\\n encode_block_trans_1 = self.conv_encoder_trans_2(encode_block_trans_1)\\n encode_block_trans_1 = self.conv_encoder_trans_3(encode_block_trans_1)\\n \\n encode_block_trans_1 = torch.cat((encode_pool4, encode_block_trans_1), 1) # Concatenation\\n \\n encode_block_trans_2 = self.conv_encoder_trans_4(encode_block_trans_1)\\n encode_block_trans_2 = self.conv_encoder_trans_5(encode_block_trans_2)\\n encode_block_trans_2 = self.conv_encoder_trans_6(encode_block_trans_2)\\n\\n middle_block = torch.cat((encode_block_trans_1, encode_block_trans_2), 1) # Concatenation\\n \\n convTranspose_transitional = self.convTranspose_transitional(middle_block) \\n # Decoder4 --block5\\n decode_block4 = torch.cat((convTranspose_transitional, encode_block4), 1) \\n if self.residus[4] == 1:\\n decode_block4 += self.residual_shortcut_encoder_decoder4(encode_block4)\\n\\n #--block6\\n cat_layer3 = self.conv_decoder4(decode_block4) \\n if self.residus[5] == 1:\\n cat_layer3 += self.residual_shortcut_decoder4(decode_block4)\\n convTranspose_decoder4 = self.convTranspose_decoder4(cat_layer3)\\n \\n \\n # Decoder3 --block7\\n decode_block3 = torch.cat((convTranspose_decoder4, encode_block3), 1)\\n if self.residus[6] == 1:\\n decode_block3 += self.residual_shortcut_encoder_decoder3(encode_block3)\\n \\n #--block8\\n cat_layer2 = self.conv_decoder3(decode_block3) \\n if self.residus[7] == 1:\\n cat_layer2 += self.residual_shortcut_decoder3(decode_block3)\\n convTranspose_decoder3 = self.convTranspose_decoder3(cat_layer2)\\n \\n # Decoder2 --block9\\n decode_block2 = torch.cat((convTranspose_decoder3, encode_block2), 1) \\n if self.residus[8] == 1:\\n decode_block2 += self.residual_shortcut_encoder_decoder2(encode_block2)\\n \\n #--block10\\n cat_layer1 = self.conv_decoder2(decode_block2)\\n if self.residus[9] == 1:\\n cat_layer1 += self.residual_shortcut_decoder2(decode_block2)\\n convTranspose_decoder2 = self.convTranspose_decoder2(cat_layer1)\\n \\n # Decoder1 --block11\\n decode_block1 = torch.cat((convTranspose_decoder2, encode_block1), 1) \\n if self.residus[10] == 1:\\n decode_block1 += self.residual_shortcut_encoder_decoder1(encode_block1)\\n \\n #--block12\\n final_layer = self.final_layer(decode_block1)\\n if self.residus[11] == 1:\\n final_layer += self.residual_shortcut_final_layer(decode_block1)\\n \\n \\n return final_layer\",\n \"def encode_input(self, x_tensor, inp_lens_tensor):\\r\\n input_emb = self.input_emb.forward(x_tensor)\\r\\n enc_output_each_word, enc_context_mask, enc_final_states = self.encoder(input_emb, inp_lens_tensor)\\r\\n enc_final_states_reshaped = (enc_final_states[0].unsqueeze(0), enc_final_states[1].unsqueeze(0))\\r\\n # print('lest go', enc_final_states_reshaped[1].shape)\\r\\n return enc_output_each_word, enc_context_mask, enc_final_states_reshaped\",\n \"def build_label_transform():\\n\\n return NALabelEncoder()\",\n \"def transform(self, X):\\n for i,f in enumerate(self.features):\\n X[f] = self._label_encoders_[i].transform(X[f])\\n return X\",\n \"def _define_encoder(self):\\n raise NotImplementedError\",\n \"def __init__(self, momentum: float = .5):\\n super(VanillaEncoder, self).__init__()\\n self.conv1 = PointNetConv2Layer(64, momentum)\\n self.conv2 = PointNetConv2Layer(64, momentum)\\n self.conv3 = PointNetConv2Layer(64, momentum)\\n self.conv4 = PointNetConv2Layer(128, momentum)\\n self.conv5 = PointNetConv2Layer(1024, momentum)\",\n \"def encoder_inference(self, features, states):\\n with tf.name_scope(f\\\"{self.name}_encoder\\\"):\\n outputs = tf.expand_dims(features, axis=0)\\n outputs, new_states = self.encoder.recognize(outputs, states)\\n return tf.squeeze(outputs, axis=0), new_states\",\n \"def __call__(self, encoder_inputs, attention_bias, inputs_padding):\\n for n, layer in enumerate(self.layers):\\n # Run inputs through the sublayers.\\n self_attention_layer = layer[0]\\n feed_forward_network = layer[1]\\n\\n with tf.variable_scope(\\\"encoder_layer_%d\\\" % n):\\n with tf.variable_scope(\\\"self_attention\\\"):\\n encoder_inputs = self_attention_layer(encoder_inputs, attention_bias)\\n with tf.variable_scope(\\\"ffn\\\"):\\n encoder_inputs = feed_forward_network(encoder_inputs, inputs_padding)\\n\\n with tf.variable_scope(\\\"encoder_output\\\"):\\n output = self.output_normalization(encoder_inputs)\\n\\n return output # self.output_normalization(encoder_inputs)\",\n \"def pretrain_forward(self, inp):\\n return self.encoder(inp)\",\n \"def encoder(self, features=[8], name=\\\"encoder\\\") -> KM.Model:\\n input_tensor = KL.Input(\\n shape=(32, 32, 3)\\n ) # shape of images for cifar10 dataset\\n encoded = KL.Conv2D(\\n features[0],\\n 3,\\n strides=(2, 2),\\n padding=\\\"same\\\",\\n use_bias=False,\\n name=name + f\\\"_conv_{1}\\\",\\n )(input_tensor)\\n encoded = KL.Activation(\\\"relu\\\")(KL.BatchNormalization()(encoded))\\n encoded_list = [encoded]\\n\\n # Prepare the skip tensor from input\\n skip_input_tensor = KL.Activation(\\\"relu\\\")(\\n KL.BatchNormalization()(\\n KL.Conv2D(features[0], 1, strides=1, use_bias=False)(input_tensor)\\n )\\n )\\n skip_input_tensor = KL.SpatialDropout2D(rate=0.2)(skip_input_tensor)\\n skip_input_tensor = KL.AveragePooling2D(pool_size=(2, 2), strides=2)(\\n skip_input_tensor\\n )\\n skip_tensors = tf.concat(\\n [\\n skip_input_tensor, # Routing info from input tensor to next levels\\n encoded, # Routes info from second level to next levels\\n ],\\n axis=-1,\\n )\\n for i, feature_num in enumerate(features[1:], start=2):\\n encoded, skip_tensors = conv_block(\\n encoded,\\n skip_tensors,\\n features_in=features[i - 2],\\n features_out=feature_num,\\n name=name + f\\\"_conv_{i}\\\",\\n )\\n encoded_list.append(encoded)\\n return KM.Model(inputs=input_tensor, outputs=encoded_list, name=name)\",\n \"def __call__(self, *args, **kwargs):\\n return self.encoder_model(args[0])\",\n \"def __call__(self, *args, **kwargs):\\n return self.encoder_model(args[0])\",\n \"def encode(self, x):\\n self.eval()\\n x = torch.as_tensor(x).unsqueeze(0)\\n if self.do_mt:\\n enc_output, _ = self.encoder_mt(x, None)\\n else:\\n enc_output, _ = self.encoder(x, None)\\n return enc_output.squeeze(0)\",\n \"def forward(self, x):\\n if self.training:\\n x = self.input_pert(x)\\n x = self.encoder(x)\\n x = self.decoder(x)\\n return x\",\n \"def encoder(x, h_dim, reuse=False, reorder_layers=True, training=False):\\n if(reorder_layers):\\n block = conv_block_reordered\\n else:\\n block = conv_block\\n with tf.variable_scope('encoder', reuse=reuse):\\n net = block(x, h_dim, name='conv_1', training=training, block_idx=0)\\n net = block(net, h_dim, name='conv_2', training=training, block_idx=1)\\n net = block(net, h_dim, name='conv_3', training=training, block_idx=2)\\n net = block(net, h_dim, name='conv_4', training=training, block_idx=3)\\n embeddings = tf.contrib.layers.flatten(net)\\n return embeddings\",\n \"def complex_encoder(self, x, reuse=True): \\n\\t\\tenc_layers = self.enc_layers\\n\\t\\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\\n\\t\\twith tf.variable_scope('model', reuse=reuse):\\n\\t\\t\\twith tf.variable_scope('encoder', reuse=reuse):\\n\\t\\t\\t\\tconv1 = tf.layers.conv2d(x, 32, 4, strides=(2,2), padding=\\\"SAME\\\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv1')\\n\\t\\t\\t\\tconv2 = tf.layers.conv2d(conv1, 32, 4, strides=(2,2), padding=\\\"SAME\\\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv2')\\n\\t\\t\\t\\tconv3 = tf.layers.conv2d(conv2, 64, 4, strides=(2,2), padding=\\\"SAME\\\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv3')\\n\\t\\t\\t\\tconv4 = tf.layers.conv2d(conv3, 64, 4, strides=(2,2), padding=\\\"SAME\\\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv4')\\n\\t\\t\\t\\tconv5 = tf.layers.conv2d(conv4, 256, 4, padding=\\\"VALID\\\", activation=tf.nn.elu, kernel_regularizer=regularizer, reuse=reuse, name='conv5')\\n\\t\\t\\t\\tflattened = tf.reshape(conv5, (-1, 256*1*1))\\n\\t\\t\\t\\tz_mean = tf.layers.dense(flattened, enc_layers[-1], activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-final')\\n\\t\\treturn z_mean\",\n \"def call(self, inputs, *args, **kwargs):\\n outputs = self.encoder(inputs, *args, **kwargs)\\n return outputs\",\n \"def forward(self, x):\\n # Encode\\n encode_block1 = self.conv_encoder1(x)\\n encode_pool1 = self.max_pool_encoder1(encode_block1)\\n encode_block2 = self.conv_encoder2(encode_pool1)\\n encode_pool2 = self.max_pool_encoder2(encode_block2)\\n encode_block3 = self.conv_encoder3(encode_pool2)\\n encode_pool3 = self.max_pool_encoder3(encode_block3)\\n encode_block4 = self.conv_encoder4(encode_pool3)\\n encode_pool4 = self.max_pool_encoder4(encode_block4)\\n\\n # Transitional block\\n middle_block = self.transitional_block(encode_pool4)\\n\\n # Decode\\n decode_block4 = torch.cat((middle_block, encode_block4), 1)\\n cat_layer3 = self.conv_decoder4(decode_block4)\\n decode_block3 = torch.cat((cat_layer3, encode_block3), 1)\\n cat_layer2 = self.conv_decoder3(decode_block3)\\n decode_block2 = torch.cat((cat_layer2, encode_block2), 1)\\n cat_layer1 = self.conv_decoder2(decode_block2)\\n decode_block1 = torch.cat((cat_layer1, encode_block1), 1)\\n final_layer = self.final_layer(decode_block1)\\n return final_layer\",\n \"def encode(self,x,x_len):\\n\\n ## Check to see if batch_size parameter is fixed or base on input batch\\n cur_batch_size = x.size()[1]\\n encode_init_state = self.encoder.initialize_hidden_state(cur_batch_size)\\n encoder_state, encoder_outputs = self.encoder.forward(x, encode_init_state, x_len)\\n\\n return encoder_outputs, encoder_state\",\n \"def forward(self, x):\\n # Get results of encoder network\\n h1 = self.encode_nn(x)\\n\\n # latent space\\n mu = self.encode_mu(h1)\\n log_var = self.encode_log_var(h1)\\n\\n # Reparameterize\\n z = self.reparameterize(mu, log_var)\\n return z, mu, log_var\",\n \"def forward(self, x):\\n # Get results of encoder network\\n h1 = self.encode_nn(x)\\n\\n # latent space\\n mu = self.encode_mu(h1)\\n log_var = self.encode_log_var(h1)\\n\\n # Reparameterize\\n z = self.reparameterize(mu, log_var)\\n return z, mu, log_var\",\n \"def base_encoder(cls, data, init_encoder, downsize_encoder, input_encoder):\\n #todo: maybe do positional encoding before passing to init_encoder\\n data = downsize_encoder(init_encoder(data))\\n return input_encoder(data)\",\n \"def transform(self, X):\\n \\n # Check is fit had been called\\n check_is_fitted(self, ['encoder_dict_'])\\n \\n # Check that the input is of the same shape as the one passed\\n # during fit.\\n if X.shape[1] != self.input_shape_[1]:\\n raise ValueError('Number of columns in dataset is different from training set used to fit the encoder')\\n \\n X = X.copy()\\n for feature in self.variables:\\n for category in self.encoder_dict_[feature]:\\n X[str(feature) + '_' + str(category)] = np.where(X[feature] == category, 1, 0)\\n \\n # drop the original non-encoded variables.\\n X.drop(labels=self.variables, axis=1, inplace=True)\\n \\n return X\",\n \"def encode(self, x):\\n mu = []\\n logvar = []\\n for i in range(self.n_views):\\n mu_, logvar_ = self.encoders[i](x[i])\\n mu.append(mu_)\\n logvar.append(logvar_)\\n mu = torch.stack(mu)\\n logvar = torch.stack(logvar)\\n mu_out, logvar_out = self.join_z(mu, logvar)\\n qz_x = hydra.utils.instantiate( \\n self.cfg.encoder.default.enc_dist, loc=mu_out, scale=logvar_out.exp().pow(0.5)\\n )\\n return [qz_x]\",\n \"def forward(self, *args):\\r\\n enc_src, src_mask, src_inp = self.unified_encoder(*args)\\r\\n batch_size, _, hid_dim = src_inp.shape\\r\\n device = src_inp.device\\r\\n # enc_src = [batch size, src len, hid dim]\\r\\n # src_mask = [batch size, src len]\\r\\n\\r\\n trg_inp = torch.cat((torch.zeros((batch_size, 1, hid_dim), device=device), src_inp[:, :-1, :]), 1)\\r\\n decoder_output, attention = self.decoder(trg_inp, enc_src, src_mask)\\r\\n\\r\\n # decoder_output = [batch size, trg len, hid dim]\\r\\n # attention = [batch size, n heads, trg len, src len]\\r\\n\\r\\n outputs = self.output_layer(decoder_output)\\r\\n\\r\\n return outputs, attention\",\n \"def _apply_encoder(self, frame, prop, encoder, encoder_type=\\\"category\\\"):\\n pass\",\n \"def forward(self, x, mask):\\n x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, mask))\\n #print('encoder')\\n #print(x.shape)\\n return self.sublayer[1](x, self.feed_forward)\",\n \"def autoencoder3(ds, compression_factor=16, input_noise=0.2, dropout_p=0.1, activ='tanh', final_activ='tanh'):\\n # compression_factor=20\\n print('DS shape: {}'.format(ds.shape))\\n in_dims = np.prod(ds.shape[1:])\\n encoding_dim = int(in_dims // compression_factor)\\n in_shape = ds[0].shape\\n print('Input Dims: {}, input shape: {}, encoding dims: {}'.format(in_dims, in_shape, encoding_dim))\\n\\n # this is our input placeholder\\n input_img = Input(shape=(in_dims,))\\n encoded = GaussianNoise(input_noise)(input_img)\\n\\n encoded = Dense(encoding_dim * 4, activation=activ, activity_regularizer=regularizers.activity_l1(10e-5))(encoded)\\n # encoded = Dense(encoding_dim*4, activation='sigmoid')(input_img)\\n encoded = BatchNormalization()(encoded)\\n encoded = Dropout(dropout_p)(encoded) # batch norm before dropout\\n # encoded = Dense(encoding_dim*3, activation=activ)(encoded)\\n # encoded = Dropout(dropout_p)(encoded)\\n encoded = Dense(encoding_dim * 2, activation=activ)(encoded)\\n encoded = Dropout(dropout_p)(encoded)\\n\\n encoded = Dense(encoding_dim, activation=activ)(encoded)\\n # Middle Noise\\n encoded = GaussianNoise(0.02)(encoded)\\n\\n # DECODED LAYER\\n # \\\"decoded\\\" is the lossy reconstruction of the input\\n decoded = Dense(encoding_dim * 2, activation=activ)(encoded)\\n # decoded = Dropout(dropout_p)(decoded)\\n decoded = Dense(encoding_dim * 4, activation=activ)(decoded)\\n decoded = Dropout(dropout_p)(decoded)\\n decoded = Dense(in_dims, activation=final_activ)(decoded)\\n\\n # MODEL\\n autoencoder = Model(input=input_img, output=decoded)\\n\\n # SEPERATE ENCODER MODEL\\n encoder = Model(input=input_img, output=encoded)\\n\\n # create a placeholder for an encoded (32-dimensional) input\\n encoded_input = Input(shape=(encoding_dim,))\\n\\n # retrieve the last layer of the autoencoder model\\n decoder_layer0 = autoencoder.layers[-4]\\n decoder_layer1 = autoencoder.layers[-3]\\n decoder_layer2 = autoencoder.layers[-2]\\n decoder_layer3 = autoencoder.layers[-1]\\n # todo: make this into a dedicated unrolling function\\n\\n # create the decoder model - unrolling the model as we go\\n decoder = Model(input=encoded_input, output=decoder_layer3(decoder_layer2(\\n decoder_layer1(decoder_layer0(encoded_input)))))\\n\\n # model.add(GaussianNoise(0.1), input_shape=(n_input_len,))\\n autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')\\n\\n autoencoder.model_name = 'Autoencoder 1'\\n return autoencoder, encoder, decoder\",\n \"def encoder(input_layer, num_filters, size=(4, 4), instance_norm=True):\\n # Initialize the weights\\n init = tf.random_normal_initializer(0.0, 0.02)\\n\\n # Convolutional Layer\\n encoder_layer = Conv2D(filters=num_filters, kernel_size=size, strides=(2, 2), padding='same',\\n kernel_initializer=init, use_bias=False)(input_layer)\\n\\n # Instance normalization\\n if instance_norm:\\n encoder_layer = tfa.layers.InstanceNormalization()(encoder_layer)\\n\\n # Leaky ReLU\\n encoder_layer = LeakyReLU(alpha=0.2)(encoder_layer)\\n return encoder_layer\",\n \"def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\\n latent_dimension):\\n super(VAEEncoder, self).__init__()\\n self.latent_dimension = latent_dimension\\n\\n # Reduce dimension up to second last layer of Encoder\\n self.encode_nn = nn.Sequential(\\n nn.Linear(in_dimension, layer_1d),\\n nn.ReLU(),\\n nn.Linear(layer_1d, layer_2d),\\n nn.ReLU(),\\n nn.Linear(layer_2d, layer_3d),\\n nn.ReLU()\\n )\\n\\n # Latent space mean\\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\\n\\n # Latent space variance\\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)\",\n \"def __init__(self, in_dimension, layer_1d, layer_2d, layer_3d,\\n latent_dimension):\\n super(VAEEncoder, self).__init__()\\n self.latent_dimension = latent_dimension\\n\\n # Reduce dimension up to second last layer of Encoder\\n self.encode_nn = nn.Sequential(\\n nn.Linear(in_dimension, layer_1d),\\n nn.ReLU(),\\n nn.Linear(layer_1d, layer_2d),\\n nn.ReLU(),\\n nn.Linear(layer_2d, layer_3d),\\n nn.ReLU()\\n )\\n\\n # Latent space mean\\n self.encode_mu = nn.Linear(layer_3d, latent_dimension)\\n\\n # Latent space variance\\n self.encode_log_var = nn.Linear(layer_3d, latent_dimension)\",\n \"def transformer_layers(self):\\n return self._transformer_layers\",\n \"def common_layers_with_encoder(self):\\n return [\\n self.self_attention, self.self_attention_layer_norm,\\n self.intermediate_dense, self.output_dense, self.output_layer_norm\\n ]\",\n \"def autoencoder(self, data):\\n with tf.variable_scope(\\\"autoencoder\\\"):\\n latent = self.encoder(data)\\n _, output = self.decoder(latent)\\n\\n return output, latent\",\n \"def encoder(self, x, reuse=True):\\n\\t\\te = x\\n\\t\\tenc_layers = self.enc_layers\\n\\t\\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\\n\\t\\twith tf.variable_scope('model', reuse=reuse):\\n\\t\\t\\twith tf.variable_scope('encoder', reuse=reuse):\\n\\t\\t\\t\\tfor layer_idx, layer_dim in enumerate(enc_layers[:-1]):\\n\\t\\t\\t\\t\\te = tf.layers.dense(e, layer_dim, activation=tf.nn.leaky_relu, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(layer_idx))\\n\\t\\t\\t\\tif self.channel_model == 'bsc':\\n\\t\\t\\t\\t\\tz_mean = tf.layers.dense(e, self.z_dim, activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(len(enc_layers)-1))\\n\\t\\t\\t\\telse:\\n\\t\\t\\t\\t\\t# N x D x 2 for erasure channel\\n\\t\\t\\t\\t\\tz_mean = tf.layers.dense(e, self.z_dim * 2, activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-'+str(len(enc_layers)-1))\\n\\t\\t\\t\\t\\tz_mean = tf.reshape(z_mean, (-1, self.z_dim, 2))\\n\\t\\treturn z_mean\",\n \"def build_encoder(shift):\\n ### TODO.\",\n \"def autoencoder(input_dims, filters, latent_dims):\\n e_inputs = keras.Input(input_dims)\\n d_inputs = keras.Input(latent_dims)\\n\\n encoder = e_inputs\\n for f in filters:\\n encoder = keras.layers.Conv2D(\\n f, (3, 3), activation='relu', padding='same')(encoder)\\n encoder = keras.layers.MaxPooling2D((2, 2), padding='same')(encoder)\\n\\n decoder = d_inputs\\n for i in reversed(range(1, len(filters))):\\n decoder = keras.layers.Conv2D(\\n filters[i], (3, 3), activation='relu', padding='same')(decoder)\\n decoder = keras.layers.UpSampling2D((2, 2))(decoder)\\n\\n decoder = keras.layers.Conv2D(\\n filters[0], (3, 3), activation='relu', padding='valid')(decoder)\\n decoder = keras.layers.UpSampling2D((2, 2))(decoder)\\n decoder = keras.layers.Conv2D(input_dims[-1], (3, 3),\\n activation='sigmoid',\\n padding='same')(decoder)\\n\\n encoder = keras.Model(e_inputs, encoder)\\n decoder = keras.Model(d_inputs, decoder)\\n\\n auto = keras.Model(e_inputs, decoder(encoder(e_inputs)))\\n auto.compile(optimizer=\\\"adam\\\", loss=\\\"binary_crossentropy\\\")\\n\\n return encoder, decoder, auto\",\n \"def encoder(x, reuse=False):\\n if reuse:\\n tf.get_variable_scope().reuse_variables()\\n with tf.name_scope('Encoder'):\\n e_dense_1 = tf.nn.relu(dense(x, input_dim, n_l1, 'e_dense_1'))\\n e_dense_2 = tf.nn.relu(dense(e_dense_1, n_l1, n_l2, 'e_dense_2'))\\n latent_variable = dense(e_dense_2, n_l2, z_dim, 'e_latent_variable')\\n return latent_variable\",\n \"def convert(encoder, bert_model):\\n num_layers = encoder._config[\\\"num_layers\\\"]\\n num_attention_heads = encoder._config[\\\"num_attention_heads\\\"]\\n hidden_size = encoder._config[\\\"hidden_size\\\"]\\n head_size = hidden_size // num_attention_heads\\n assert head_size * num_attention_heads == hidden_size\\n encoder._embedding_layer.set_weights(\\n [bert_model[\\\"embeddings.word_embeddings.weight\\\"]])\\n encoder._embedding_norm_layer.set_weights([\\n bert_model[\\\"embeddings.LayerNorm.weight\\\"],\\n bert_model[\\\"embeddings.LayerNorm.bias\\\"]\\n ])\\n encoder._type_embedding_layer.set_weights(\\n [bert_model[\\\"embeddings.token_type_embeddings.weight\\\"]])\\n encoder._position_embedding_layer.set_weights(\\n [bert_model[\\\"embeddings.position_embeddings.weight\\\"]])\\n for layer_num in range(num_layers):\\n encoder._transformer_layers[\\n layer_num]._attention_layer._key_dense.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.key.weight\\\"].T\\n .reshape((hidden_size, num_attention_heads, head_size)),\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.key.bias\\\"]\\n .reshape((num_attention_heads, head_size))\\n ])\\n encoder._transformer_layers[\\n layer_num]._attention_layer._query_dense.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.query.weight\\\"]\\n .T.reshape((hidden_size, num_attention_heads, head_size)),\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.query.bias\\\"]\\n .reshape((num_attention_heads, head_size))\\n ])\\n encoder._transformer_layers[\\n layer_num]._attention_layer._value_dense.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.value.weight\\\"]\\n .T.reshape((hidden_size, num_attention_heads, head_size)),\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.self.value.bias\\\"]\\n .reshape((num_attention_heads, head_size))\\n ])\\n encoder._transformer_layers[\\n layer_num]._attention_layer._output_dense.set_weights([\\n bert_model[\\n f\\\"encoder.layer.{layer_num}.attention.output.dense.weight\\\"].T\\n .reshape((num_attention_heads, head_size, hidden_size)),\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.output.dense.bias\\\"]\\n ])\\n encoder._transformer_layers[layer_num]._attention_layer_norm.set_weights([\\n bert_model[\\n f\\\"encoder.layer.{layer_num}.attention.output.LayerNorm.weight\\\"],\\n bert_model[f\\\"encoder.layer.{layer_num}.attention.output.LayerNorm.bias\\\"]\\n ])\\n encoder._transformer_layers[layer_num]._intermediate_dense.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.intermediate.dense.weight\\\"].T,\\n bert_model[f\\\"encoder.layer.{layer_num}.intermediate.dense.bias\\\"]\\n ])\\n encoder._transformer_layers[layer_num]._output_dense.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.output.dense.weight\\\"].T,\\n bert_model[f\\\"encoder.layer.{layer_num}.output.dense.bias\\\"]\\n ])\\n encoder._transformer_layers[layer_num]._output_layer_norm.set_weights([\\n bert_model[f\\\"encoder.layer.{layer_num}.output.LayerNorm.weight\\\"],\\n bert_model[f\\\"encoder.layer.{layer_num}.output.LayerNorm.bias\\\"]\\n ])\",\n \"def init_encoder(self):\\n\\n vgg = models.vgg16(pretrained=True)\\n\\n blocks = [self.layer_1,\\n self.layer_2,\\n self.layer_3,\\n self.layer_4,\\n self.layer_5]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\\n\\n blocks = [self.layer_11,\\n self.layer_12,\\n self.layer_13,\\n self.layer_14,\\n self.layer_15]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\",\n \"def transform():\",\n \"def init_encoder(self):\\n\\n vgg = models.vgg16(pretrained=True)\\n\\n blocks = [self.layer_1,\\n self.layer_2,\\n self.layer_3,\\n self.layer_4,\\n self.layer_5]\\n\\n ranges = [[0, 4], [5, 9], [10, 16], [17, 23], [24, 29]]\\n features = list(vgg.features.children())\\n\\n vgg_layers = []\\n for _layer in features:\\n if isinstance(_layer, nn.Conv2d):\\n vgg_layers.append(_layer)\\n\\n merged_layers = []\\n for idx, conv_block in enumerate(blocks):\\n if idx < 2:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit]\\n else:\\n units = [conv_block.conv1.cbr_unit,\\n conv_block.conv2.cbr_unit,\\n conv_block.conv3.cbr_unit]\\n for _unit in units:\\n for _layer in _unit:\\n if isinstance(_layer, nn.Conv2d):\\n merged_layers.append(_layer)\\n\\n assert len(vgg_layers) == len(merged_layers)\\n\\n for l1, l2 in zip(vgg_layers, merged_layers):\\n if isinstance(l1, nn.Conv2d) and isinstance(l2, nn.Conv2d):\\n assert l1.weight.size() == l2.weight.size()\\n assert l1.bias.size() == l2.bias.size()\\n l2.weight.data = l1.weight.data\\n l2.bias.data = l1.bias.data\",\n \"def forward(self, x):\\n # Get results of encoder network\\n q = self.encode_nn(x)\\n\\n return q\",\n \"def conv_encoder(input_tensor, num_latent, is_training=True):\\n del is_training\\n\\n e1 = tf.layers.conv2d(\\n inputs=input_tensor,\\n filters=32,\\n kernel_size=4,\\n strides=2,\\n activation=tf.nn.relu,\\n padding=\\\"same\\\",\\n name=\\\"e1\\\",\\n )\\n e2 = tf.layers.conv2d(\\n inputs=e1,\\n filters=32,\\n kernel_size=4,\\n strides=2,\\n activation=tf.nn.relu,\\n padding=\\\"same\\\",\\n name=\\\"e2\\\",\\n )\\n e3 = tf.layers.conv2d(\\n inputs=e2,\\n filters=64,\\n kernel_size=2,\\n strides=2,\\n activation=tf.nn.relu,\\n padding=\\\"same\\\",\\n name=\\\"e3\\\",\\n )\\n e4 = tf.layers.conv2d(\\n inputs=e3,\\n filters=64,\\n kernel_size=2,\\n strides=2,\\n activation=tf.nn.relu,\\n padding=\\\"same\\\",\\n name=\\\"e4\\\",\\n )\\n flat_e4 = tf.layers.flatten(e4)\\n e5 = tf.layers.dense(flat_e4, 256, activation=tf.nn.relu, name=\\\"e5\\\")\\n means = tf.layers.dense(e5, num_latent, activation=None, name=\\\"means\\\")\\n log_var = tf.layers.dense(e5, num_latent, activation=None, name=\\\"log_var\\\")\\n return means, log_var\",\n \"def forward(self, x):\\n # encode\\n encode_block1 = self.conv_encode1(x)\\n encode_pool1 = self.conv_maxpool1(encode_block1)\\n encode_block2 = self.conv_encode2(encode_pool1)\\n encode_pool2 = self.conv_maxpool2(encode_block2)\\n encode_block3 = self.conv_encode3(encode_pool2)\\n encode_pool3 = self.conv_maxpool3(encode_block3)\\n # Bottleneck\\n bottleneck1 = self.bottleneck(encode_pool3)\\n # Decode\\n decode_block3 = crop_and_concat(\\n bottleneck1, encode_block3, crop=True)\\n cat_layer2 = self.conv_decode3(decode_block3)\\n decode_block2 = crop_and_concat(\\n cat_layer2, encode_block2, crop=True)\\n cat_layer1 = self.conv_decode2(decode_block2)\\n decode_block1 = crop_and_concat(\\n cat_layer1, encode_block1, crop=True)\\n final_layer = self.final_layer(decode_block1)\\n return final_layer\",\n \"def transform(self):\\n\\n # Gets a serialized dict representation of the model, containing all its\\n # layers, their connections and configuration. This is the main structure\\n # which is used to understand model structure, and also manipulate it.\\n #\\n # config = {\\n # 'input_layers': [ ... ],\\n # 'layers': [{\\n # 'inbound_nodes': [INPUT CONFIG OF LAYER],\\n # 'name': 'LAYER_NAME',\\n # 'config': { LAYER_CONFIG }\\n # }, {\\n # ...\\n # }],\\n # 'output_layers': [ ... ],\\n # 'name': 'MODEL_NAME',\\n #\\n\\n # Ensures old Keras serialization format\\n self.model.use_legacy_config = True\\n self._config = self.model.get_config()\\n\\n # Stores map of Transform -> List of layer names matched by transform.\\n # Same transform should not match+replace the same layer more than once\\n # to prevent infinite loops.\\n self._transform_matched_layers_map = {}\\n self._layer_weights_map = {}\\n self._layer_names_and_weights_map = {}\\n\\n for layer in self.model.layers:\\n self._layer_weights_map[layer.name] = self._get_keras_layer_weights(layer)\\n self._layer_names_and_weights_map[\\n layer.name] = self._get_keras_layer_names_and_weights(layer)\\n\\n # Maintains a current mutable copy of the metadata through transformation.\\n self._layer_metadata_map = copy.deepcopy(self.layer_metadata)\\n\\n # We run an infinite loop and keep applying transformations as long as\\n # patterns are found. This allows recursive pattern matching where a\\n # modification by one transform may lead to another match.\\n #\\n # TODO(pulkitb): This leads to infinite loops with poor patterns which may\\n # match their replacement. Add counters with limits to fix it.\\n while True:\\n match_found = False\\n for transform in self.transforms:\\n # A transform may find multiple instances of a pattern in the model.\\n # Keep finding and replacing till done.\\n while True:\\n match_layer_node = self._find_pattern(\\n transform.pattern(), self._get_matched_layers(transform))\\n\\n # Pattern did not match any layer. Move to next transform.\\n if not match_layer_node:\\n break\\n\\n self._store_successful_match(transform, match_layer_node)\\n\\n # Copying the match_layer_node ensures the replacement code can\\n # freely modify the match.\\n replacement_layer_node = transform.replacement(\\n copy.deepcopy(match_layer_node))\\n\\n # If equal, the matched layers are being replaced with exactly the\\n # same set of layers that were matched with the same config.\\n # For Transforms, that may inadvertently do this we can end up in\\n # an infinite loop. Move on if no meaningful change has been made.\\n if match_layer_node == replacement_layer_node:\\n continue\\n\\n match_found = True\\n self._replace(match_layer_node, replacement_layer_node)\\n\\n # None of the transforms found a pattern. We can stop now.\\n if not match_found:\\n break\\n\\n custom_objects = {}\\n for transform in self.transforms:\\n custom_objects.update(transform.custom_objects())\\n\\n # Reconstruct model from the config, using the cloned layers.\\n if self._is_functional_model(self.model):\\n transformed_model = keras.Model.from_config(self._config, custom_objects)\\n else:\\n transformed_model = keras.Sequential.from_config(self._config,\\n custom_objects)\\n\\n for layer in transformed_model.layers:\\n weights = self._layer_weights_map.get(layer.name)\\n if weights:\\n self._set_layer_weights(layer, weights)\\n else:\\n names_and_weights = self._layer_names_and_weights_map.get(layer.name)\\n if names_and_weights:\\n self._set_layer_names_and_weights(layer, names_and_weights)\\n\\n # Ensures old Keras serialization format\\n transformed_model.use_legacy_config = True\\n return transformed_model, copy.deepcopy(self._layer_metadata_map)\",\n \"def SemiAutoencoder(ds, compression_factor=16, input_noise=0.2, dropout_p=0.1, activ='tanh', final_activ='tanh'):\\n # compression_factor=20\\n print('DS shape: {}'.format(ds.shape))\\n in_dims = np.prod(ds.shape[1:])\\n encoding_dim = int(in_dims // compression_factor)\\n in_shape = ds[0].shape\\n print('Input Dims: {}, input shape: {}, encoding dims: {}'.format(in_dims, in_shape, encoding_dim))\\n\\n # this is our input placeholder\\n input_img = Input(shape=(in_dims,))\\n encoded = GaussianNoise(input_noise)(input_img)\\n\\n encoded = Dense(encoding_dim * 4, activation=activ, activity_regularizer=regularizers.activity_l1(10e-5))(encoded)\\n # encoded = Dense(encoding_dim*4, activation='sigmoid')(input_img)\\n encoded = BatchNormalization()(encoded)\\n encoded = Dropout(dropout_p)(encoded) # batch norm before dropout\\n # encoded = Dense(encoding_dim*3, activation=activ)(encoded)\\n # encoded = Dropout(dropout_p)(encoded)\\n encoded = Dense(encoding_dim * 2, activation=activ)(encoded)\\n encoded = Dropout(dropout_p)(encoded)\\n\\n encoded = Dense(encoding_dim, activation=activ)(encoded)\\n # Middle Noise\\n encoded = GaussianNoise(0.02)(encoded)\\n\\n # DECODED LAYER\\n # \\\"decoded\\\" is the lossy reconstruction of the input\\n decoded = Dense(encoding_dim * 2, activation=activ)(encoded)\\n # decoded = Dropout(dropout_p)(decoded)\\n decoded = Dense(encoding_dim * 4, activation=activ)(decoded)\\n decoded = Dropout(dropout_p)(decoded)\\n decoded = Dense(in_dims, activation=final_activ)(decoded)\\n\\n # MODEL\\n autoencoder = Model(input=input_img, output=decoded)\\n\\n # SEPERATE ENCODER MODEL\\n encoder = Model(input=input_img, output=encoded)\\n\\n # create a placeholder for an encoded (32-dimensional) input\\n encoded_input = Input(shape=(encoding_dim,))\\n\\n # retrieve the last layer of the autoencoder model\\n decoder_layer0 = autoencoder.layers[-4]\\n decoder_layer1 = autoencoder.layers[-3]\\n decoder_layer2 = autoencoder.layers[-2]\\n decoder_layer3 = autoencoder.layers[-1]\\n # todo: make this into a dedicated unrolling function\\n\\n # create the decoder model - unrolling the model as we go\\n decoder = Model(input=encoded_input, output=decoder_layer3(decoder_layer2(\\n decoder_layer1(decoder_layer0(encoded_input)))))\\n\\n # model.add(GaussianNoise(0.1), input_shape=(n_input_len,))\\n autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')\\n\\n autoencoder.model_name = 'Autoencoder 1'\\n return autoencoder, encoder, decoder\",\n \"def make_encoder(self, input_size, latent_size):\\n return nn.Sequential(\\n nn.Linear(input_size, 500),\\n nn.ReLU(),\\n nn.Linear(500, 200),\\n nn.ReLU(),\\n nn.Linear(200, latent_size * 2)\\n # times 2 because this is the concatenated vector of latent mean and variance\\n )\",\n \"def _encoder(self, z_dimension, mcd):\\n inputs = Input(shape=(2 * self.n_dims,), name=\\\"encoder_input\\\")\\n x = inputs\\n x = Dense(\\n self.hidden_size,\\n activation=self.activation_func,\\n kernel_initializer=self.weight_init,\\n )(x)\\n x = Dropout(self.dropout_probability)(x, training=mcd)\\n x = Dense(\\n self.hidden_size // 2,\\n activation=self.activation_func,\\n kernel_initializer=self.weight_init,\\n )(x)\\n x = Dropout(self.dropout_probability)(x, training=mcd)\\n\\n x = Dense(\\n self.hidden_size // 3,\\n activation=self.activation_func,\\n kernel_initializer=self.weight_init,\\n )(x)\\n x = Dropout(self.dropout_probability)(x, training=mcd)\\n\\n x = Dense(\\n self.hidden_size // 4,\\n activation=self.activation_func,\\n kernel_initializer=self.weight_init,\\n )(x)\\n x = Dropout(self.dropout_probability)(x, training=mcd)\\n\\n # Latent mean\\n z_mean = Dense(z_dimension, name=\\\"z_mean\\\")(x)\\n \\n # latent variance\\n z_var = Dense(z_dimension, name=\\\"z_var\\\")(x)\\n self.encoder = Model(inputs, [z_mean, z_var], name=\\\"encoder\\\")\\n\\n return self.encoder, inputs, z_mean, z_var\",\n \"def convolutional_32_encoder(self, x, reuse=True): \\n\\t\\tenc_layers = self.enc_layers\\n\\t\\tregularizer = tf.contrib.layers.l2_regularizer(scale=self.reg_param)\\n\\t\\twith tf.variable_scope('model', reuse=reuse):\\n\\t\\t\\twith tf.variable_scope('encoder', reuse=reuse):\\n\\t\\t\\t\\tconv1 = tf.layers.conv2d(x, 128, 2, strides=(2,2), padding=\\\"VALID\\\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv1')\\n\\t\\t\\t\\tconv1 = tf.layers.batch_normalization(conv1)\\n\\n\\t\\t\\t\\tconv2 = tf.layers.conv2d(conv1, 256, 2, strides=(2,2), padding=\\\"VALID\\\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv2')\\n\\t\\t\\t\\tconv2 = tf.layers.batch_normalization(conv2)\\n\\n\\t\\t\\t\\tconv3 = tf.layers.conv2d(conv2, 512, 2, strides=(2,2), padding=\\\"VALID\\\", activation=tf.nn.relu, kernel_regularizer=regularizer, reuse=reuse, name='conv3')\\n\\t\\t\\t\\tconv3 = tf.layers.batch_normalization(conv3)\\n\\n\\t\\t\\t\\tflattened = tf.contrib.layers.flatten(conv3)\\n\\t\\t\\t\\tz_mean = tf.layers.dense(flattened, enc_layers[-1], activation=None, use_bias=False, kernel_regularizer=regularizer, reuse=reuse, name='fc-final')\\n\\t\\treturn z_mean\",\n \"def _encode(self):\\n with tf.variable_scope('encoding'):\\n self.sep_p_encodes, _ = bilstm_layer(self.p_emb, self.p_length, self.hidden_size)\\n tf.get_variable_scope().reuse_variables()\\n self.sep_q_encodes, _ = bilstm_layer(self.q_emb, self.q_length, self.hidden_size)\\n if self.use_dropout:\\n self.sep_p_encodes = tf.nn.dropout(self.sep_p_encodes, 1-self.dropout)\\n self.sep_q_encodes = tf.nn.dropout(self.sep_q_encodes, 1-self.dropout)\",\n \"def encode_input_for_decoder(x_tensor, inp_lens_tensor, model_input_emb: EmbeddingLayer, model_enc: RNNEncoder):\\n input_emb = model_input_emb.forward(x_tensor)\\n (enc_output_each_word, enc_context_mask, enc_final_states) = model_enc.forward(input_emb, inp_lens_tensor)\\n enc_final_states_reshaped = (enc_final_states[0].unsqueeze(0), enc_final_states[1].unsqueeze(0))\\n return (enc_output_each_word, enc_context_mask, enc_final_states_reshaped)\",\n \"def forward(self, inputs_encoder, inputs_decoder):\\n states_encoder = self.encoder(inputs_encoder)\\n outputs_decoder, states_decoder = self.decoder(inputs_decoder, states_encoder)\\n return outputs_decoder, states_decoder\",\n \"def forward(self, *args):\\r\\n enc_src, _, _ = self.unified_encoder(*args)\\r\\n enc_src = enc_src.view(enc_src.shape[0], -1)\\r\\n y_pred = self.mlp(enc_src)\\r\\n return y_pred\",\n \"def vae_encoder(self):\\n return stax.serial(\\n stax.Dense(self.hidden_dims[0], W_init=stax.randn()),\\n stax.Relu,\\n stax.Dense(self.hidden_dims[1], W_init=stax.randn()),\\n stax.Relu,\\n stax.FanOut(2),\\n stax.parallel(\\n stax.Dense(self.z_dim, W_init=stax.randn()), # mean\\n stax.serial(stax.Dense(self.z_dim, W_init=stax.randn()), stax.Exp), # std -- i.e. diagonal covariance\\n ),\\n )\",\n \"def setup_layers(self, params):\\n\\n #\\n # Input layer\\n #\\n\\n # Input and encoding layers\\n layer_stack = [\\n tf.placeholder(\\n tf.int32,\\n name=\\\"input\\\",\\n shape=[None, params.in_size]\\n )\\n ]\\n\\n in_layer = layer_stack[0]\\n\\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\\n character_depth = len(good_chars) + 2\\n\\n layer_stack.append(\\n self.make_onehot_encode_layer(\\n layer_stack[-1],\\n character_depth\\n )\\n )\\n\\n # Add conv layers\\n for idx, num_nodes in enumerate(params.conv_nodes):\\n layer_stack.append(\\n self.make_conv_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'conv_{}'.format(idx),\\n params.conv_params\\n )\\n )\\n\\n # Flatten convolutional layers\\n layer_stack.append(\\n tf.contrib.layers.flatten(\\n layer_stack[-1]\\n )\\n )\\n\\n # Add dense layers\\n for idx, num_nodes in enumerate(params.dense_nodes):\\n layer_stack.append(\\n self.make_dense_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'dense_{}'.format(idx),\\n params.dense_params\\n )\\n )\\n\\n # Output layer is broken into two pieces. The pre/post the application\\n # of the activation function. That's because the loss functions\\n # for cross-entropy rely on the pre-activation scores in preact\\n\\n preact_out_layer, out_layer = self.make_dense_output_layer(\\n layer_stack[-1],\\n params.out_size,\\n params.output_params\\n )\\n\\n target_layer = tf.placeholder(\\n \\\"float\\\",\\n name=\\\"target\\\",\\n shape=[None, params.out_size]\\n )\\n\\n if params.output_params.activation in ['sigmoid']:\\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\\n else:\\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\\n\\n return in_layer, out_layer, target_layer, loss\",\n \"def _fit_encoder(self, frame, prop, encoder_type=\\\"category\\\"):\\n pass\",\n \"def __init__(self):\\n super(Encoder, self).__init__()\\n\\n self.lconv1 = nn.Sequential(\\n nn.Conv1d(4, 64, kernel_size=9, padding=4),\\n nn.BatchNorm1d(64),\\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\\n nn.BatchNorm1d(64),\\n )\\n\\n self.conv1 = nn.Sequential(\\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\\n nn.BatchNorm1d(64),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(64, 64, kernel_size=9, padding=4),\\n nn.BatchNorm1d(64),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv2 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=4, stride=4),\\n nn.Conv1d(64, 96, kernel_size=9, padding=4),\\n nn.BatchNorm1d(96),\\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\\n nn.BatchNorm1d(96),\\n )\\n\\n self.conv2 = nn.Sequential(\\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\\n nn.BatchNorm1d(96),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(96, 96, kernel_size=9, padding=4),\\n nn.BatchNorm1d(96),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv3 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=4, stride=4),\\n nn.Conv1d(96, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n )\\n\\n self.conv3 = nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv4 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=5, stride=5),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n )\\n\\n self.conv4 = nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv5 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=5, stride=5),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n )\\n\\n self.conv5 = nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv6 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=5, stride=5),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n )\\n\\n self.conv6 = nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n )\\n\\n self.lconv7 = nn.Sequential(\\n nn.MaxPool1d(kernel_size=2, stride=2),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n )\\n\\n self.conv7 = nn.Sequential(\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n nn.Conv1d(128, 128, kernel_size=9, padding=4),\\n nn.BatchNorm1d(128),\\n nn.ReLU(inplace=True),\\n )\",\n \"def build_autoencoder(input_dim):\\r\\n input_layer = Input(shape=(input_dim, 1))\\r\\n enc = Conv1D(filters=16, kernel_size=2, padding='same', activation='relu')(input_layer)\\r\\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\\r\\n enc = Conv1D(filters=32, kernel_size=2, padding='same', activation='relu')(enc)\\r\\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\\r\\n enc = Conv1D(filters=64, kernel_size=2, padding='same', activation='relu')(enc)\\r\\n enc = MaxPooling1D(pool_size=2, padding='same')(enc)\\r\\n enc = Flatten()(enc)\\r\\n enc = Dense(64)(enc)\\r\\n\\r\\n dec = Dense(200704)(enc)\\r\\n dec = Reshape((3136, 64))(dec)\\r\\n dec = Conv1D(filters=64, kernel_size=2, padding='same', activation='relu')(dec)\\r\\n dec = UpSampling1D(2)(dec)\\r\\n dec = Conv1D(filters=32, kernel_size=2, padding='same', activation='relu')(dec)\\r\\n dec = UpSampling1D(2)(dec)\\r\\n dec = Conv1D(filters=16, kernel_size=2, padding='same', activation='relu')(dec)\\r\\n dec = UpSampling1D(2)(dec)\\r\\n dec = Conv1D(filters=1, kernel_size=2, padding='same', activation='relu')(dec)\\r\\n\\r\\n autoencoder = Model(input_layer, dec)\\r\\n autoencoder.compile(optimizer='adam', loss='mean_squared_error', metrics=['accuracy'])\\r\\n autoencoder.summary()\\r\\n encoder = Model(input_layer, enc)\\r\\n return autoencoder, encoder\",\n \"def base_encoder(cls, data, data_mask, init_encoder, downsize_encoder, input_encoder):\\n data = downsize_encoder(init_encoder(data))\\n return input_encoder(data, data_mask)\",\n \"def __init__(self, AEs):\\n \\\"\\\"\\\" the default view of the stacked autoencoders\\\"\\\"\\\"\\n sa = AEs\\n \\\"\\\"\\\" the encoder view of the stacked autoencoders \\\"\\\"\\\"\\n ec = Cat([a.ec for a in sa])\\n \\\"\\\"\\\" the decoder view of the stacked autoencoders \\\"\\\"\\\"\\n dc = Cat([a.dc for a in reversed(sa)])\\n\\n self.sa = sa # default view\\n self.ec = ec # encoder view\\n self.dc = dc # decoder view\\n\\n nts = []\\n nts.extend(ec)\\n nts.extend(dc)\\n super(SAE, self).__init__(nts)\",\n \"def encode(self, x):\\n h1 = F.relu(self.conv1(x))\\n h2 = F.relu(self.conv2(h1))\\n h3 = F.relu(self.conv3(h2))\\n h4 = F.relu(self.conv4(h3))\\n h4 = h4.reshape(-1)\\n mu = self.fc1(h4)\\n logvar = self.fc2(h4)\\n\\n # we take log of var as sigma cannot be negative \\n # while neural network can output negative values\\n sigma = torch.exp(logvar * 0.5)\\n\\n normal = torch.randn_like(sigma)\\n\\n return mu + sigma * normal, mu, logvar\",\n \"def encoder(self, value):\\n self._tensor.encoder = value\",\n \"def transform(self, inputs: list, stage: str) -> datapack.DataPack:\",\n \"def forward(self, x, neg=None):\\n if neg is None:\\n return self._target_encoder(tf.reshape(x, [-1]))\\n else:\\n return self._target_encoder(tf.reshape(x, [-1])), \\\\\\n self._context_encoder(tf.reshape(x, [-1])), \\\\\\n self._context_encoder(tf.reshape(neg, [-1]))\",\n \"def transform(self):\",\n \"def fc_encoder(input_tensor, num_latent, is_training=True):\\n del is_training\\n\\n flattened = tf.layers.flatten(input_tensor)\\n e1 = tf.layers.dense(flattened, 1200, activation=tf.nn.relu, name=\\\"e1\\\")\\n e2 = tf.layers.dense(e1, 1200, activation=tf.nn.relu, name=\\\"e2\\\")\\n means = tf.layers.dense(e2, num_latent, activation=None)\\n log_var = tf.layers.dense(e2, num_latent, activation=None)\\n return means, log_var\",\n \"def encode(self, x, encode_params, name=None):\\n values = [x] + list(encode_params)\\n with tf.compat.v1.name_scope(name, 'gather_encoder_encode', values):\\n x = tf.convert_to_tensor(x)\\n encode_params = tf.nest.map_structure(tf.convert_to_tensor, encode_params)\\n return self._encode_fn(x, encode_params)\",\n \"def encoder_decoder_archi_gan(inputs, is_train):\\n\\n encoder_layers = []\\n\\n encoded = inputs\\n\\n encoder_layers.append(encoded)\\n\\n for i in range(config.encoder_layers):\\n encoded = encoder_conv_block_gan(encoded, i, is_train)\\n encoder_layers.append(encoded)\\n \\n encoder_layers.reverse()\\n\\n\\n\\n decoded = encoder_layers[0]\\n\\n for i in range(config.encoder_layers):\\n decoded = decoder_conv_block_gan(decoded, encoder_layers[i+1], i, is_train)\\n\\n return decoded\",\n \"def encoder(inputs):\\n out = tf.layers.flatten(inputs)\\n out = tf.layers.dense(out, 400, activation=tf.nn.relu)\\n mean = tf.layers.dense(out, LATENT_SIZE)\\n logstd = tf.layers.dense(out, LATENT_SIZE)\\n return tf.distributions.Normal(loc=mean, scale=tf.exp(logstd))\",\n \"def _build_encoder(self, hparams, is_training):\\n with tf.variable_scope(\\\"trajectory_encoder\\\"):\\n with tf.name_scope(\\\"source_placeholder\\\"):\\n input_phs = list_ops.list_placeholder(self.num_gpu, (None, self.input_length, self.input_dims), tf.float32)\\n for ph in input_phs:\\n tf.add_to_collection('placeholder', ph)\\n \\n if hparams.encoder_type == \\\"rnn\\\":\\n net = input_phs\\n with tf.variable_scope(\\\"projection\\\"):\\n net = self._build_input_projection(hparams, net, is_training)\\n\\n with tf.name_scope(\\\"batch_time_transpose\\\"):\\n net = list_ops.list_transpose(net, perm=[1, 0, 2])\\n\\n with tf.variable_scope(\\\"rnn\\\"):\\n net, state = self._build_rnn_encoder(hparams, net, is_training)\\n\\n if hparams.relu_reconfiguration:\\n with tf.variable_scope(\\\"reconfiguration\\\"):\\n net = list_ops.list_dense_with_bn(net,\\n hparams.cnn_input_projector_filters[-1],\\n is_training,\\n self.bn_decay,\\n seed=self.random_seed)\\n\\n elif hparams.encoder_type == \\\"cnn\\\":\\n net = self._build_cnn_encoder(hparams, input_phs, is_training)\\n state = None\\n \\n else:\\n raise ValueError(\\\"Unknown encoder type {:s}.\\\".format(hparams.encoder_type))\\n\\n return net, state\",\n \"def transform(self, X):\\n ...\",\n \"def transform(self, X):\\n ...\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.72634786","0.7005841","0.68940455","0.68260175","0.68099016","0.6762681","0.6706989","0.66525126","0.66428787","0.6633862","0.6610087","0.6552235","0.65452635","0.6533876","0.6485411","0.64490217","0.6430559","0.64192617","0.64069134","0.638147","0.63809276","0.637174","0.6362806","0.63602066","0.634091","0.6331195","0.63092864","0.62918246","0.6281844","0.62611014","0.6259239","0.6255259","0.62396604","0.62396306","0.62254727","0.6224256","0.6219025","0.6203083","0.6203083","0.61865187","0.6178467","0.617287","0.6145369","0.6123169","0.6118295","0.6109964","0.61019146","0.61019146","0.60954696","0.60860276","0.6084581","0.607489","0.60729146","0.6072385","0.60638124","0.6056762","0.6054114","0.6054114","0.60317206","0.60311425","0.599379","0.5985512","0.5979341","0.59535587","0.59485507","0.59481734","0.59455323","0.59406847","0.5935502","0.5933471","0.59277666","0.59214735","0.59171563","0.5914759","0.59085745","0.5908045","0.59001976","0.5888749","0.5872039","0.5867673","0.58620715","0.5859339","0.5848321","0.58449906","0.5844849","0.58350295","0.5832193","0.58157647","0.5812399","0.58075714","0.58042175","0.5798156","0.5773544","0.57731426","0.57693774","0.57680917","0.5762211","0.5758109","0.57567674","0.57567674"],"string":"[\n \"0.72634786\",\n \"0.7005841\",\n \"0.68940455\",\n \"0.68260175\",\n \"0.68099016\",\n \"0.6762681\",\n \"0.6706989\",\n \"0.66525126\",\n \"0.66428787\",\n \"0.6633862\",\n \"0.6610087\",\n \"0.6552235\",\n \"0.65452635\",\n \"0.6533876\",\n \"0.6485411\",\n \"0.64490217\",\n \"0.6430559\",\n \"0.64192617\",\n \"0.64069134\",\n \"0.638147\",\n \"0.63809276\",\n \"0.637174\",\n \"0.6362806\",\n \"0.63602066\",\n \"0.634091\",\n \"0.6331195\",\n \"0.63092864\",\n \"0.62918246\",\n \"0.6281844\",\n \"0.62611014\",\n \"0.6259239\",\n \"0.6255259\",\n \"0.62396604\",\n \"0.62396306\",\n \"0.62254727\",\n \"0.6224256\",\n \"0.6219025\",\n \"0.6203083\",\n \"0.6203083\",\n \"0.61865187\",\n \"0.6178467\",\n \"0.617287\",\n \"0.6145369\",\n \"0.6123169\",\n \"0.6118295\",\n \"0.6109964\",\n \"0.61019146\",\n \"0.61019146\",\n \"0.60954696\",\n \"0.60860276\",\n \"0.6084581\",\n \"0.607489\",\n \"0.60729146\",\n \"0.6072385\",\n \"0.60638124\",\n \"0.6056762\",\n \"0.6054114\",\n \"0.6054114\",\n \"0.60317206\",\n \"0.60311425\",\n \"0.599379\",\n \"0.5985512\",\n \"0.5979341\",\n \"0.59535587\",\n \"0.59485507\",\n \"0.59481734\",\n \"0.59455323\",\n \"0.59406847\",\n \"0.5935502\",\n \"0.5933471\",\n \"0.59277666\",\n \"0.59214735\",\n \"0.59171563\",\n \"0.5914759\",\n \"0.59085745\",\n \"0.5908045\",\n \"0.59001976\",\n \"0.5888749\",\n \"0.5872039\",\n \"0.5867673\",\n \"0.58620715\",\n \"0.5859339\",\n \"0.5848321\",\n \"0.58449906\",\n \"0.5844849\",\n \"0.58350295\",\n \"0.5832193\",\n \"0.58157647\",\n \"0.5812399\",\n \"0.58075714\",\n \"0.58042175\",\n \"0.5798156\",\n \"0.5773544\",\n \"0.57731426\",\n \"0.57693774\",\n \"0.57680917\",\n \"0.5762211\",\n \"0.5758109\",\n \"0.57567674\",\n \"0.57567674\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.63730127"},"document_rank":{"kind":"string","value":"21"}}},{"rowIdx":90882,"cells":{"query":{"kind":"string","value":"Pass Embedded Patches through the layers"},"document":{"kind":"string","value":"def call(self, inputs, *args, **kwargs):\n outputs = self.encoder(inputs, *args, **kwargs)\n return outputs"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def run(layers):","def build_layers(self):\n raise NotImplementedError","def layers(self): # -> LayerView:\n ...","def UpdateLayers(self):\n pass","def layers(self, layers):\n self._layers = layers\n prev = None\n for layer in layers:\n if not layer.inputs and prev is not None:\n layer.inputs = [prev]\n prev = layer","def layers(self, layers):\n\n self._layers = layers","def __init__(self, layers):\n\n\t\tself.layers = layers","def handle_layers(context, model, toplayer, layerids, materials, update, import_hidden=False):\n #setup main container to hold all layer collections\n layer_col_id=\"Layers\"\n if not layer_col_id in context.blend_data.collections:\n layer_col = context.blend_data.collections.new(name=layer_col_id)\n try:\n toplayer.children.link(layer_col)\n except Exception:\n pass\n else:\n #If \"Layers\" collection is in place, we assume the plugin had imported 3dm before\n layer_col = context.blend_data.collections[layer_col_id]\n\n # build lookup table for LayerTable index\n # from GUID, create collection for each\n # layer\n for lid, l in enumerate(model.Layers):\n if not l.Visible and not import_hidden:\n continue\n lcol = utils.get_iddata(context.blend_data.collections, l.Id, l.Name, None)\n layerids[str(l.Id)] = (lid, lcol)\n utils.tag_data(layerids[str(l.Id)][1], l.Id, l.Name)\n '''\n matname = l.Name + \"+\" + str(l.Id)\n if matname not in materials:\n laymat = utils.get_iddata(context.blend_data.materials, l.Id, l.Name, None)\n if update:\n\t laymat.use_nodes = True\n\t r, g, b, _ = l.Color\n\t principled = PrincipledBSDFWrapper(laymat, is_readonly=False)\n\t principled.base_color = (r/255.0, g/255.0, b/255.0)\n materials[matname] = laymat\n '''\n # second pass so we can link layers to each other\n for l in model.Layers:\n # link up layers to their parent layers\n if str(l.ParentLayerId) in layerids:\n parentlayer = layerids[str(l.ParentLayerId)][1]\n try:\n parentlayer.children.link(layerids[str(l.Id)][1])\n except Exception:\n pass\n # or to the top collection if no parent layer was found\n else:\n try:\n layer_col.children.link(layerids[str(l.Id)][1])\n except Exception:\n pass","def consume_layer(self, reports):\n layer_list = []\n layer_count = 1\n for report in reports:\n layer = create_image_layer(report)\n layer.layer_index = layer_count\n layer_list.append(layer)\n layer_count += 1\n return layer_list","def __call__(cls, *args, **kwargs):\n layer = super(LayerAspect, cls).__call__(*args, **kwargs)\n\n if Job.Current:\n Job.Current.addLayer(layer)\n \n layer.afterInit()\n return layer","def layers(self, layers):\n self._layers = layers\n self.thetas = []\n prev = None\n for layer in layers:\n if not layer.inputs and prev is not None:\n layer.inputs = [prev]\n prev = layer\n self.thetas.extend(layer.thetas())","def layers(self, x):\n raise NotImplementedError","def setup_layer_structure(self):\n self.page_rank_convolution_1 = self.layer(self.feature_number, self.args.layers[0], self.args.iterations, self.args.alpha)\n self.page_rank_convolution_2 = self.layer(self.args.layers[0], self.args.layers[1], self.args.iterations, self.args.alpha)\n self.page_rank_convolution_3 = self.layer(self.args.layers[1], self.class_number, self.args.iterations, self.args.alpha)","def build_models(config):\n inputs = Input(shape=(config['patch_height'], config['patch_width'], config['depth'], config['channel']),\n name='patchimg')\n\n kernelinitfun = keras.initializers.RandomNormal(mean=0.0, stddev=0.1, seed=None)\n activationfun = 'relu'\n # kernelinitfun = 'glorot_normal'\n\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_1')(inputs)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_2')(x)\n # x = Dropout(0.3)(x)\n x = BatchNormalization(name='bn1')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp1', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_1')(x)\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_2')(x)\n # x = Dropout(0.2)(x)\n x = BatchNormalization(name='bn2')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp2', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_1')(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_2')(x)\n # x = Dropout(0.5)(x)\n x = BatchNormalization(name='bn3')(x)\n x = Activation(activationfun)(x)\n x = MaxPooling3D(name='mp3', strides=(2, 2, 1))(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_1')(x)\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_2')(x)\n # x = Dropout(0.5)(x)\n x = BatchNormalization(name='bn4')(x)\n x = Activation(activationfun)(x)\n\n x4 = Flatten(name='aux_fx')(x)\n\n source_classifier = Dropout(0.5)(x4)\n source_classifier = Dense(512, activation='softmax', name=\"mo1\")(source_classifier)\n source_classifier = Dropout(0.5)(source_classifier)\n source_classifier = Dense(128, activation='softmax', name=\"mo2\")(source_classifier)\n # source_classifier = Dropout(0.3)(source_classifier)\n source_classifier = Dense(1, name=\"mo\")(source_classifier)\n\n domain_classifier = Dense(32, activation='linear', name=\"do4\")(x4)\n domain_classifier = BatchNormalization(name=\"do5\")(domain_classifier)\n domain_classifier = Activation(\"elu\", name=\"do6\")(domain_classifier)\n domain_classifier = Dropout(0.5)(domain_classifier)\n\n domain_classifier = Dense(2, activation='softmax', name=\"do\")(domain_classifier)\n\n adamop = keras.optimizers.Adam(learning_rate=1e-5, beta_1=0.9, beta_2=0.999, amsgrad=False)\n comb_model = Model(inputs=inputs, outputs=[source_classifier, domain_classifier])\n comb_model.compile(optimizer=adamop,\n loss={'mo': 'mae', 'do': 'categorical_crossentropy'},\n loss_weights={'mo': 1, 'do': 2}, metrics=['accuracy'], )\n\n source_classification_model = Model(inputs=inputs, outputs=[source_classifier])\n source_classification_model.compile(optimizer=adamop,\n loss={'mo': 'mae'}, metrics=['accuracy'], )\n\n domain_classification_model = Model(inputs=inputs, outputs=[domain_classifier])\n domain_classification_model.compile(optimizer=adamop,\n loss={'do': 'categorical_crossentropy'}, metrics=['accuracy'])\n\n embeddings_model = Model(inputs=inputs, outputs=[x4])\n embeddings_model.compile(optimizer=adamop, loss='categorical_crossentropy', metrics=['accuracy'])\n\n return comb_model, source_classification_model, domain_classification_model, embeddings_model","def __init__(self, kernel_size, filters, stage, block):\n super().__init__(name='identity' + str(stage) + block)\n filters1, filters2, filters3 = filters\n if K.image_data_format() == 'channels_last':\n bn_axis = 3\n else:\n bn_axis = 1\n\n conv_name_base = 'res' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n self.conv1 = layers.Conv2D(\n filters1, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2a')\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\n self.act1 = layers.Activation('relu')\n\n self.conv2 = layers.Conv2D(\n filters2,\n kernel_size,\n padding='same',\n kernel_initializer='he_normal',\n name=conv_name_base + '2b')\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\n self.act2 = layers.Activation('relu')\n\n self.conv3 = layers.Conv2D(\n filters3, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2c')\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\n\n self.add = layers.Add()\n self.act = layers.Activation('relu')","def __init__(self, parent, layer):\n pass","def setup_layers(self):\n if self.args.model == \"exact\":\n self.layer = PPNPLayer\n else:\n self.layer = APPNPLayer\n self.setup_layer_structure()","def set_hooks(self):\n for layer_idx, layer in enumerate(self.vgg_features):\n if layer_idx in self.layers_to_watch:\n layer.register_forward_hook(self.layer_watch_hooks(layer_idx))","def __init__(self, kernel_size, filters, stage, block, strides=(2, 2)):\n super().__init__(name='conv_block' + str(stage) + block)\n filters1, filters2, filters3 = filters\n if K.image_data_format() == 'channels_last':\n bn_axis = 3\n else:\n bn_axis = 1\n conv_name_base = 'res' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n self.conv1 = layers.Conv2D(\n filters1, (1, 1),\n strides=strides,\n kernel_initializer='he_normal',\n name=conv_name_base + '2a')\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\n self.act1 = layers.Activation('relu')\n\n self.conv2 = layers.Conv2D(\n filters2,\n kernel_size,\n padding='same',\n kernel_initializer='he_normal',\n name=conv_name_base + '2b')\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\n self.act2 = layers.Activation('relu')\n\n self.conv3 = layers.Conv2D(\n filters3, (1, 1),\n kernel_initializer='he_normal',\n name=conv_name_base + '2c')\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\n\n self.shortcut_conv = layers.Conv2D(\n filters3, (1, 1),\n strides=strides,\n kernel_initializer='he_normal',\n name=conv_name_base + '1')\n self.shortcut_bn = layers.BatchNormalization(\n axis=bn_axis, name=bn_name_base + '1')\n\n self.add = layers.Add()\n self.act = layers.Activation('relu')","def _build(self):\n with tf.variable_scope (self.name + '_architecutre') as scope:\n images_square = unflatten_layer ( self.images )\n visualize_images(images_square)\n\n # Conv Layer 1\n conv1_out, params = conv_2d_layer ( input = images_square,\n neurons = CONV_1_N,\n filter_size = CONV_1_FILT,\n name = 'enc_conv_1',\n visualize = True )\n process_params(params, name = self.name)\n e1_params = params\n pool1_out = max_pool_2d_layer ( input = conv1_out, name = 'enc_pool_1')\n # lrn1_out = local_response_normalization_layer (pool1_out, name = 'lrn_1' )\n\n # Conv Layer 2\n conv2_out, params = conv_2d_layer ( input = pool1_out,\n neurons = CONV_2_N,\n filter_size = CONV_2_FILT,\n name = 'enc_conv_2' )\n process_params(params, name = self.name)\n e2_params = params\n pool2_out = max_pool_2d_layer ( input = conv2_out, name = 'enc_pool_2')\n # lrn2_out = local_response_normalization_layer (pool2_out, name = 'lrn_2' )\n\n flattened = flatten_layer(pool2_out)\n\n # Dropout Layer 1 \n flattened_dropout = dropout_layer ( input = flattened,\n prob = self.dropout_prob,\n name = 'enc_dropout_1') \n\n # Dot Product Layer 1\n fc1_out, params = dot_product_layer ( input = flattened_dropout,\n neurons = HIDDEN_1,\n name = 'enc_dot_1')\n process_params(params, name = self.name)\n e3_params = params \n\n # Dropout Layer 2 \n fc1_out_dropout = dropout_layer ( input = fc1_out,\n prob = self.dropout_prob,\n name = 'enc_dropout_2')\n # Dot Product Layer 2\n fc2_out, params = dot_product_layer ( input = fc1_out_dropout, \n neurons = HIDDEN_2,\n name = 'enc_dot_2')\n process_params(params, name = self.name)\n e4_params = params \n\n # Dropout Layer 3 \n fc2_out_dropout = dropout_layer ( input = fc2_out,\n prob = self.dropout_prob,\n name = 'enc_dropout_3')\n \n # Dot Product Layer 2\n self.codeword, params = dot_product_layer ( input = fc2_out_dropout, \n neurons = CODEWORD_LENGTH,\n activation = CODE_ACTIVATION,\n name = 'enc_dot_2')\n process_params(params, name = self.name)\n process_codeword_normalization_regularizer(self.codeword, \n coeff = AUTOENCODER_CODEWORD_COEFF,\n name = self.name)\n e5_params = params \n # tf.summary.histogram('codewords', self.codeword)\n # self.hash = threshold_layer ( input = self.codeword,\n # name = 'hash')\n # process_hash_regularizer(self.codeword, coeff = AUTOENCODER_HASH_COEFF,\n # name = self.name)\n\n # Decoder ... \n decoder_1_out, params = dot_product_layer ( input = self.codeword, \n neurons = HIDDEN_2,\n params = [tf.transpose(e5_params[0]), None],\n name = 'decoder_dot_1')\n d1_params = params\n process_params([params[1]], name = self.name)\n \n dec_1_out_dropout = dropout_layer ( input = decoder_1_out,\n prob = self.dropout_prob,\n name = 'dec_dropout_1')\n\n decoder_2_out, params = dot_product_layer ( input = dec_1_out_dropout, \n neurons = HIDDEN_1,\n params = [tf.transpose(e4_params[0]), None],\n name = 'decoder_dot_2')\n d2_params = params\n process_params([params[1]], name = self.name)\n \n # dropout 2\n dec_2_out_dropout = dropout_layer ( input = decoder_2_out,\n prob = self.dropout_prob,\n name = 'dec_dropout_2')\n\n decoder_3_out, params = dot_product_layer ( input = dec_2_out_dropout, \n neurons = 1250,\n params = [tf.transpose(e3_params[0]), None],\n name = 'decoder_dot_3')\n d3_params = params\n process_params([params[1]], name = self.name)\n\n # DeConv Layer 1\n # The output shapes need to be changed according to architecture.\n\n dec_3_square = unflatten_layer ( decoder_3_out, channels = CONV_2_N )\n upsample_1 = upsampling_layer (dec_3_square, size = (10,10), name = 'dec_upsampling_1')\n\n deconv1_out, params = deconv_2d_layer ( input = upsample_1,\n neurons = CONV_1_N,\n filter_size = CONV_2_FILT,\n output_shape = (12,12),\n # n_outs = MINI_BATCH_SIZE,\n stride = (1,1,1,1), \n params = [e2_params[0], None], \n name = 'dec_deconv_1' )\n\n process_params([params[1]], name = self.name)\n d4_params = params\n\n # DeConv Layer 2\n upsample_2 = upsampling_layer (deconv1_out, size = (24,24), name = 'dec_upsampling_2')\n decoded_images_square, params = deconv_2d_layer ( input = upsample_2,\n neurons = 1,\n filter_size = CONV_1_FILT,\n stride = (1,1,1,1),\n output_shape = (28,28),\n # n_outs = MINI_BATCH_SIZE, \n params = [e1_params[0], None], \n activation = 'tanh', \n name = 'dec_deconv_2' )\n \n process_params([params[1]], name = self.name)\n d5_params = params \n \n self.decoded = flatten_layer (decoded_images_square, in_shp = [-1, 28, 28, 1])\n visualize_images(decoded_images_square, name = 'decoded')\n # This is because transpose don't initialize.\n self.params = [ [e5_params[0], d1_params[1] ],\n [e4_params[0], d2_params[1] ],\n [e3_params[0], d3_params[1] ],\n [e2_params[0], d4_params[1] ],\n [e1_params[0], d5_params[1] ] ]\n\n with tf.variable_scope (self.name + '_objectives') as scope: \n with tf.variable_scope( self.name + '_decoder_error') as scope:\n reconstruction_error = rmse(self.images, self.decoded) \n tf.add_to_collection( self.name + '_objectives', reconstruction_error ) \n tf.summary.scalar('reconstruction_error', reconstruction_error)\n\n self._cook_optimizer( \n lr = AUTOENCODER_LR, \n optimizer = AUTOENCODER_OPTIMIZER,\n l1_coeff = AUTOENCODER_L1_COEFF,\n l2_coeff = AUTOENCODER_WEIGHT_DECAY_COEFF)","def setup_layers(self, params):\n\n #\n # Input layer\n #\n\n # Input and encoding layers\n layer_stack = [\n tf.placeholder(\n tf.int32,\n name=\"input\",\n shape=[None, params.in_size]\n )\n ]\n\n in_layer = layer_stack[0]\n\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\n character_depth = len(good_chars) + 2\n\n layer_stack.append(\n self.make_onehot_encode_layer(\n layer_stack[-1],\n character_depth\n )\n )\n\n # Add conv layers\n for idx, num_nodes in enumerate(params.conv_nodes):\n layer_stack.append(\n self.make_conv_layer(\n layer_stack[-1],\n num_nodes,\n 'conv_{}'.format(idx),\n params.conv_params\n )\n )\n\n # Flatten convolutional layers\n layer_stack.append(\n tf.contrib.layers.flatten(\n layer_stack[-1]\n )\n )\n\n # Add dense layers\n for idx, num_nodes in enumerate(params.dense_nodes):\n layer_stack.append(\n self.make_dense_layer(\n layer_stack[-1],\n num_nodes,\n 'dense_{}'.format(idx),\n params.dense_params\n )\n )\n\n # Output layer is broken into two pieces. The pre/post the application\n # of the activation function. That's because the loss functions\n # for cross-entropy rely on the pre-activation scores in preact\n\n preact_out_layer, out_layer = self.make_dense_output_layer(\n layer_stack[-1],\n params.out_size,\n params.output_params\n )\n\n target_layer = tf.placeholder(\n \"float\",\n name=\"target\",\n shape=[None, params.out_size]\n )\n\n if params.output_params.activation in ['sigmoid']:\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\n else:\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\n\n return in_layer, out_layer, target_layer, loss","def __call__(self, layers, inputs, name=None):\r\n def call_layer(l, x):\r\n if hasattr(l, \"call_by_computation_context\"):\r\n return l.call_by_computation_context(x, *self.args, **self.kwargs)\r\n else:\r\n return l(x, *self.args, **self.kwargs)\r\n\r\n if not isinstance(layers, (list, tuple)):\r\n # Single layer, just call\r\n return call_layer(layers, inputs)\r\n else:\r\n # Flatten the layers\r\n layers = tf.nest.flatten(layers)\r\n outputs = inputs\r\n\r\n # Add a name scope for wrapping\r\n if name is not None:\r\n scope_name = name\r\n else:\r\n scope_name = commonprefix([layer.name for layer in layers]).strip(\"-/\\\\?_ \\t\\n\") or \"ComputationContext\"\r\n with tf.name_scope(scope_name):\r\n for l in layers:\r\n outputs = call_layer(l, outputs)\r\n\r\n return outputs","def generate_legacy_layers(self, images_map, content_retriever):\n pass","def generate_legacy_layers(self, images_map, content_retriever):\n pass","def __init__(self, layer_nest):\n self._layer_nest = layer_nest\n super().__init__()","def preview_layerset(ls, size = 100, spacing = 100):\n D = Device()\n scale = size/100\n num_layers = len(ls._layers)\n matrix_size = int(np.ceil(np.sqrt(num_layers)))\n sorted_layers = sorted(ls._layers.values(),\n key = lambda x: (x.gds_layer, x.gds_datatype))\n for n, layer in enumerate(sorted_layers):\n R = rectangle(size = (100*scale, 100*scale), layer = layer)\n T = text(text = '%s\\n%s / %s'\\\n % (layer.name, layer.gds_layer, layer.gds_datatype),\n size = 20*scale,\n justify = 'center',\n layer = layer)\n\n T.move((50*scale, -20*scale))\n xloc = n % matrix_size\n yloc = int(n // matrix_size)\n D.add_ref(R)\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\n D.add_ref(T)\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\n return D","def add_layers(self, layers):\n\n existing_layers = self.layers\n assert len(existing_layers) > 0\n for layer in layers:\n assert layer.get_mlp() is None\n layer.set_mlp(self)\n layer.set_input_space(existing_layers[-1].get_output_space())\n existing_layers.append(layer)\n assert layer.layer_name not in self.layer_names\n self.layer_names.add(layer.layer_name)","def build(self):\n\n layers = GiraffeLayer.get_all_structural()\n \n for layer in layers:\n\n self.add_objects_from_layer(layer)\n\n return self","def init_nnpatches(self):\n # nnpatches = self._generate_nnpatches()\n\n # Register `nnpatches` to this `nnmodel`\n # Refer: NNFramework._init_model_params(...)\n\n # # Register this `nnmodel` to `nnpatch`\n # for nnpatch in nnpatches:\n # nnpatch.add_nnmodels(self)","def set_up_rlayer(self, rlname, rlname_other=None, include_layers=None,\n exclude_layers=None, mask_layers=None):\n scene = self.set_as_active()\n layer_numbers = constants.layer_numbers\n w_var.rlname = rlname\n\n if include_layers is None:\n include_layers = w_var.layer_numbers_all_used\n\n if exclude_layers is None:\n exclude_layers = []\n\n if mask_layers is None:\n mask_layers = []\n\n if w_var.cb_clear_rlayers:\n for layer in scene.render.layers[:-1]:\n scene.render.layers.remove(layer)\n\n scene.render.layers.active.name = rlname\n scene.render.layers.active.use = True\n\n new_rlayer = scene.render.layers.active\n\n # if not clearing render layers: creates new one\n else:\n new_rlayer = scene.render.layers.new(rlname)\n scene.render.layers.active = new_rlayer\n\n # there needs to be two render layers in the same scene for freestyle compositing\n if w_var.cb_composited:\n w_var.rlname_other = rlname_other\n other_rlayer = scene.render.layers.new(rlname_other)\n other_rlayer.layers[19] = True\n scene.render.layers[rlname_other].layers_zmask = (False,) * 20\n\n if w_var.cb_ao:\n scene.render.layers[rlname].use_pass_ambient_occlusion = True\n\n if w_var.cb_composited:\n scene.render.layers[rlname_other].use_pass_ambient_occlusion = True\n\n # because I can't deactivate a layer if it is the only active one\n new_rlayer.layers[19] = True\n \n scene.render.layers[rlname].layers_exclude = (False,) * 20\n scene.render.layers[rlname].layers_zmask = (False,) * 20\n\n for i in layer_numbers:\n if w_var.cb_composited:\n if i in w_var.layer_numbers_affected:\n scene.render.layers[rlname].layers[i] = True\n scene.render.layers[rlname_other].layers_zmask[i] = True\n\n else:\n scene.render.layers[rlname].layers[i] = False\n\n if i in w_var.layer_numbers_other:\n scene.render.layers[rlname_other].layers[i] = True\n\n else:\n scene.render.layers[rlname_other].layers[i] = False\n\n else:\n if i in include_layers:\n scene.render.layers[rlname].layers[i] = True\n\n else:\n scene.render.layers[rlname].layers[i] = False\n\n if i in mask_layers:\n scene.render.layers[rlname].layers_zmask[i] = True\n\n if i in exclude_layers:\n scene.render.layers[rlname].layers_exclude[i] = True","def draw_layers(self):\n\t\tfor z in xrange(0,16):\n\t\t\t#create surface for this layer\n\t\t\tsrf = pygame.Surface((16,128))\n\t\t\tfor x in xrange(0,16):\n\t\t\t\tfor y in xrange(0,128):\n\t\t\t\t\tv = self.data[ self.xyz_to_offset( x,y,z) ]\n\t\t\t\t\tif v != 0:\n\t\t\t\t\t\tsrf.fill( BLOCKS.get(v, [0,0])[1], \t(x, 127 -y, 1, 1 ))\n\t\t\t#save layer to dict for this chunk\n\t\t\tself.layers[z] = srf","def process_layers(size: Sequence[int],\n layers: Union[Source, Layer, Sequence[Layer]],\n a: Union[None, np.ndarray] = None,\n status: Union[None, Queue] = None) -> np.ndarray:\n # If no destination array was sent, it means we are either at the\n # start of the layer processing or we are starting the processing\n # of a layer group. If we are starting processing, then this will\n # contain the final image data. If we are starting the processing\n # of a layer group, this allows the layers in the group to only\n # blend with the other layers in the group before they blend with\n # the layers outside the group.\n if a is None:\n a = np.zeros(size, dtype=float)\n\n # If we got a sequence of layers, we process them recursively and\n # return the result.\n if isinstance(layers, Sequence):\n for layer in layers:\n a = process_layers(size, layer, a, status)\n return a\n\n # If we got a source layer, process it.\n if isinstance(layers.source, Source):\n if status is not None:\n src_name = layers.source.__class__.__name__\n status.put((sw.MSG, f'Rendering {src_name}...'))\n kwargs = {\n 'source': layers.source,\n 'size': size,\n 'location': layers.location,\n 'filters': layers.filters,\n }\n b = render_source(**kwargs)\n if status is not None:\n src_name = layers.source.__class__.__name__\n status.put((sw.MSG, f'Rendered {src_name}.'))\n status.put((sw.PROG,))\n\n # Otherwise we got a container layer, process its source and run\n # any filters that are set on the layer.\n else:\n new_size = f.preprocess(size, layers.filters)\n b = process_layers(new_size, layers.source, None, status)\n b = f.process(b, layers.filters)\n b = f.postprocess(b, layers.filters)\n\n # There are two possibilities for how the layers should be\n # blended: masked or unmasked. Masked blends will have a\n # Source in the mask attribute, which needs to be sent\n # to the blending operation.\n if layers.mask is not None:\n if status is not None:\n src_name = layers.mask.__class__.__name__\n status.put((sw.MSG, f'Rendering {src_name}...'))\n kwargs = {\n 'source': layers.mask,\n 'size': size,\n 'filters': layers.mask_filters,\n }\n mask = render_source(**kwargs)\n if status is not None:\n src_name = layers.mask.__class__.__name__\n status.put((sw.MSG, f'Rendered {src_name}.'))\n status.put((sw.PROG,))\n a, b, mask = _normalize_color_space(a, b, mask)\n return layers.blend(a, b, layers.blend_amount, mask)\n\n # Unmasked layers have no mask, so Blend the image data and return.\n a, b = _normalize_color_space(a, b)\n return layers.blend(a, b, layers.blend_amount)","def make_layers(self):\n w, h = self.image.get_size()\n shrink = pg.transform.smoothscale(self.image, (w//2, h//2))\n self.mid_image = tools.tile_surface((w,h), shrink, True)\n shrink = pg.transform.smoothscale(self.image, (w//4, h//4))\n self.base = tools.tile_surface(prepare.SCREEN_SIZE, shrink, True)","def __initAvailableLayerTypes(self):\n from backend.caffe.path_loader import PathLoader\n caffe = PathLoader().importCaffe()\n layerNameMainParts = list(caffe.layer_type_list())\n\n res = {}\n paramsPerLayerType = {}\n\n # calculate common parameters of all layer types\n # by removing all which will be used for one specific layer type only\n # also keep in mind which ones have been removed to readd them to specific layers\n commonParams = self._availableParameterGroupDescriptors[\"LayerParameter\"].parameter() #use .parameter() on purpose\n layerSpecificParameters = set()\n for nameMainPart in layerNameMainParts:\n specificParamsName = [nameMainPart + \"Parameter\"]\n if moreLayerNameParameter.has_key(nameMainPart):\n specificParamsName.append( moreLayerNameParameter[nameMainPart])\n paramsPerLayerType[nameMainPart] = {}\n for key, value in commonParams.items():\n if value.isParameterGroup() and value.parameterName() in specificParamsName:\n paramsPerLayerType[nameMainPart][key] = value\n layerSpecificParameters.add(key)\n\n\n # special case: shared params for loss layers\n key = \"loss_param\"\n value = commonParams[key]\n del commonParams[key]\n for nameMainPart in layerNameMainParts:\n if LayerType.getCategoryByName(nameMainPart) == LayerType.CATEGORY_LOSS:\n paramsPerLayerType[nameMainPart][key] = value\n\n # TODO is there a special case for the TransformationParameter?\n\n # create each layer type after one another\n for nameMainPart in layerNameMainParts:\n\n # add common params to the specific ones\n layerTypeParam = paramsPerLayerType[nameMainPart].keys()\n paramsPerLayerType[nameMainPart].update(commonParams)\n\n irrelevant = layerSpecificParameters.difference(layerTypeParam)\n res[nameMainPart] = LayerType(nameMainPart, paramsPerLayerType[nameMainPart], layerTypeParam, irrelevant)\n\n self._commonParams = commonParams\n self._availableLayerTypes = res","def EmbeddingLayers_pooling(pretrained=False, progress=True, **kwargs):\n kwargs['groups'] = 32\n kwargs['width_per_group'] = 4\n # return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,\n # **kwargs) \n return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],\n pretrained, progress, **kwargs)","def __init__(self,layer_def,input,input_shape,rs,clone_from=None):\n batch_size = int(layer_def.find(\"batchsize\").text)\n image_size = int(layer_def.find(\"imagesize\").text)\n image_channels = int(layer_def.find(\"imagechannels\").text)\n self.layer_name = layer_def.attrib[\"name\"]\n self.init(input, input_shape, batch_size,image_size,image_channels)","def feed(self, *args):\n\t\tprint(\"args: \", args)\n\t\tassert len(args) != 0\n\t\tself.terminals = []\n\t\tfor fed_layer in args:#\n\t\t\ttry:\n\t\t\t\tis_str = isinstance(fed_layer, basestring)\n\t\t\texcept NameError:\n\t\t\t\tis_str = isinstance(fed_layer, str)\n\t\t\tif is_str:\n\t\t\t\ttry:\n\t\t\t\t\tfed_layer = self.layers[fed_layer]\n\t\t\t\t\tprint(fed_layer)\n\t\t\t\texcept KeyError:\n\t\t\t\t\traise KeyError('Unknown layer name fed: %s' % fed_layer)\n\t\t\tself.terminals.append(fed_layer)\n\t\treturn self","def append_layer(self, *args, **kwargs) :\n \n self.insert_layer(len(self._layers), *args, **kwargs)","def get_data_from_layers(vk4_container, layers, step, is_image=False):\n log.debug(\"Entering get_data_from_layers()\")\n holder = []\n for x in range(0, len(layers), step):\n lay = layers[x:x + step]\n log.debug(\"In get_data_from_layers()\\n\\tCurrent layer to append: %s\" % lay)\n if lay not in layer_dict.keys():\n log.error(\"In get_data_from_layers()\\n\\tLayer {}: {} from param \" \\\n \"layers: {} is not a valid layer. See documentation.\"\n .format(x, lay, layers))\n holder.append(vk4_container.get_single_color_values(layer_dict[lay][0], layer_dict[lay][1]))\n\n if is_image:\n log.debug(\"Exiting get_data_from_layers() where is_image is {}\".format(is_image))\n return create_array_from_rgb_layers(vk4_container, holder)\n else:\n log.debug(\"Exiting get_data_from_layers() where is_image is {}\".format(is_image))\n return create_composite_rgb_values(vk4_container, holder)\n # return create_separate_rgb_values(vk4_container, holder)","def initialize_layers(self, layers_config: dict, inputs=None):\n layers_config = layers_config.copy()\n input_lyrs = []\n initiated_layers = OrderedDict()\n wrp_layer = None # indicator for wrapper layers\n first_layer = True\n\n for lyr, lyr_args in layers_config.items():\n\n lyr_config, lyr_inputs, named_outs, call_args = self.deconstruct_lyr_args(lyr, lyr_args)\n\n lyr_name, args, lyr_config, activation = self.check_lyr_config(lyr, lyr_config)\n\n if K.BACKEND == 'pytorch':\n\n if first_layer:\n first_layer = False\n\n if callable(lyr_config):\n lyr_initiated = lyr_config\n else:\n lyr_initiated = TORCH_LAYERS[lyr_name](**lyr_config)\n setattr(self, lyr, lyr_initiated)\n initiated_layers[lyr] = {\"layer\": lyr_initiated, \"named_outs\": named_outs, 'call_args': call_args,\n 'inputs': lyr_inputs}\n\n else:\n # may be user has defined layers without input layer, in this case add Input layer as first layer\n if first_layer:\n if inputs is not None: # This method was called by providing it inputs.\n assert isinstance(inputs, tf.Tensor)\n # since inputs have been defined, all the layers that will be added will be next to first layer\n first_layer = False\n layer_outputs = inputs\n initiated_layers[layer_outputs.name] = {'layer': layer_outputs, 'tf_name': lyr_name}\n\n elif lyr_name != \"Input\":\n if 'input_shape' in lyr_config: # input_shape is given in the first layer so make input layer\n initialized_layer = LAYERS[\"Input\"](shape=lyr_config['input_shape'])\n else:\n # for simple dense layer based models, lookback will not be used\n def_shape = (self.num_ins,) if self.lookback == 1 else (self.lookback, self.num_ins)\n initialized_layer = LAYERS[\"Input\"](shape=def_shape)\n\n # first layer is built so next iterations will not be for first layer\n first_layer = False\n # put the first layer in memory to be used for model compilation\n # add th layer which the user had specified as first layer\n initiated_layers[initialized_layer.name] = {'layer': initialized_layer,\n 'tf_name': lyr_name}\n input_lyrs.append(initialized_layer)\n\n # The inputs to the layer have not been specified, so either it is an Input layer\n if lyr_inputs is None:\n # or it uses the previous outputs as inputs\n if lyr_name == \"Input\":\n # it is an Input layer, hence should not be called\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'tf_name': lyr_name}\n input_lyrs.append(initialized_layer)\n else:\n # it is executable and uses previous outputs as inputs\n if lyr_name in ACTIVATION_LAYERS:\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\n wrp_layer = LAYERS[lyr_name]\n # because wrapper layer name is property\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\n 'tf_name': lyr_name}\n continue\n elif \"LAMBDA\" in lyr_name.upper():\n # lyr_config is serialized lambda layer, which needs to be deserialized\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\n # layers_config['lambda']['config'] still contails lambda, so we need to replace the python\n # object (lambda) with the serialized version (lyr_config) so that it can be saved as json file.\n layers_config[lyr]['config'] = lyr_config\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n else:\n if wrp_layer is not None:\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n wrp_layer = None\n else:\n if lyr_name == \"TemporalFusionTransformer\":\n lyr_config['return_attention_components'] = True\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n\n else: # The inputs to this layer have been specified so they must exist in lyr_cache.\n # it is an executable\n if lyr_name in ACTIVATION_LAYERS:\n\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\n wrp_layer = LAYERS[lyr_name]\n # because wrapper layer name is property\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\n 'tf_name': lyr_name}\n continue\n elif \"LAMBDA\" in lyr_name.upper():\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\n layers_config[lyr]['config'] = lyr_config\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n else:\n if wrp_layer is not None:\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n wrp_layer = None\n else:\n layer_initialized = LAYERS[lyr_name](*args, **lyr_config)\n initiated_layers[lyr_config['name']] = {'layer': layer_initialized,\n 'named_outs': named_outs,\n 'call_args': call_args,\n 'inputs': lyr_inputs,\n 'tf_name': lyr_name}\n\n if activation is not None: # put the string back to dictionary to be saved in config file\n lyr_config['activation'] = activation\n\n first_layer = False\n\n self.jsonize_lyr_config(lyr_config)\n\n # inputs = [] todo, indentify input layers\n # for k,v in lyr_cache.items():\n # since the model is not build yet and we have access to only output tensors of each list, this is probably\n # # the only way to know that how many `Input` layers were encountered during the run of this method. Each\n # tensor (except TimeDistributed) has .op.inputs attribute, which is empty if a tensor represents output of Input layer.\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\n # if k.upper() != \"TIMEDISTRIBUTED\" and hasattr(v, 'op'):\n # if hasattr(v.op, 'inputs'):\n # _ins = v.op.inputs\n # if len(_ins) == 0:\n # inputs.append(v)\n # else: # not sure if this is the proper way of checking if a layer receives an input or not!\n # if hasattr(v, '_keras_mask'):\n # inputs.append(v)\n\n setattr(self, 'initiated_layers', initiated_layers)\n setattr(self, 'input_lyrs', input_lyrs)\n\n\n # todo,\n # # for case when {Input -> Dense, Input_1}, this method wrongly makes Input_1 as output so in such case use\n # # {Input_1, Input -> Dense }, thus it makes Dense as output and first 2 as inputs, so throwing warning\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\n # if len(layer_outputs.op.inputs) < 1:\n # print(\"Warning: the output is of Input tensor class type\")\n # else:\n # if 'op' not in dir(layer_outputs): # layer_outputs does not have `op`, which means it has no incoming node\n # print(\"Warning: the output is of Input tensor class type\")\n\n # outs = None\n #if BACKEND == 'tensorflow':\n # outs = self.call(input_lyrs)\n # setattr(self, 'output_lyrs', outs)\n # if BACKEND == 'tensorflow':\n # ## Reinitial\n # super(Model, self).__init__(\n # inputs=input_lyrs,\n # outputs=outs)\n #MODEL.__init__(self, inputs=inputs, outputs=outs)\n\n return input_lyrs # , outs","def build(width, height, depth, classes, stages, filters, include_top, pooling,\n reg=1e-3, bnEps=2e-5, bnMom=0.0):\n inputShape = (height, width, depth)\n chanDim = -1\n\n if K.image_data_format() == \"channels_first\":\n inputShape = (depth, height, width)\n chanDim = 1\n\n inputs = Input(shape=inputShape)\n\n\n # block 1 (initial conv block)\n x = ZeroPadding2D(padding=((3, 3), (3, 3)), name='conv1_pad')(inputs)\n x = Conv2D(64, (7,7), use_bias=False, strides=(2,2),\n kernel_initializer=\"he_normal\", kernel_regularizer=l2(reg))(x)\n x = BatchNormalization(axis=chanDim, name=\"bn_conv1\")(x)\n x = Activation(\"relu\")(x)\n x = ZeroPadding2D(padding=((1,1), (1,1)), name=\"pool1_pad\")(x)\n x = MaxPooling2D(3, strides=2)(x)\n\n for i in range(0, len(stages)):\n stride = (1,1) if i == 0 else (2,2) # block 2 (projection block) w stride(1,1)\n\n print(\"Stage {}, Stride={}\".format(i, stride))\n x = SEResNet.residual_module(x, filters[i+1], stride,\n chanDim=chanDim, red=True, bnEps=bnEps, bnMom=bnMom)\n for j in range(0, stages[i] + 1): #stacking res block to each depth layer\n x = SEResNet.residual_module(x, filters[i+1], stride=(1,1),\n chanDim=chanDim, bnEps=bnEps,\n bnMom=bnMom)\n x = BatchNormalization(axis=chanDim, epsilon=bnEps,\n momentum=bnMom)(x)\n x = Activation(\"relu\")(x)\n\n if include_top:\n x = GlobalAveragePooling2D()(x)\n x = Dense(classes, use_bias=False, kernel_regularizer=l2(reg),\n activation='softmax')(x)\n else:\n if pooling == 'avg':\n print(\"Adding average pool\")\n x = GlobalAveragePooling2D()(x)\n elif pooling == 'max':\n x = GlobalMaxPooling2D()(x)\n\n model = Model(inputs=inputs, outputs=x, name=\"SEResNet\")\n return model","def convolutional_block(input_tensor, kernel_size, filters, stage, block, strides = (2, 2)):\n chanDim = 3\n eps = 1.1e-5\n \n if K.image_data_format() == \"channels_first\":\n inputShape = (depth, height, width)\n chanDim = 1\n\n # defining name basis\n conv_name_base = 'res152' + str(stage) + block + '_branch'\n bn_name_base = 'bn' + str(stage) + block + '_branch'\n\n # Retrieve Filters\n filters1, filters2, filters3 = filters\n \n ##### MAIN PATH #####\n # First component of main path \n X = Conv2D(filters = filters1, kernel_size = (1, 1), strides = strides, name = conv_name_base + '2a')(input_tensor)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2a')(X)\n X = Activation('relu', name = conv_name_base + '2a_relu')(X)\n \n # Second component of main path \n X = ZeroPadding2D(padding = (1, 1), name = conv_name_base + '2b_zeropadding')(X)\n X = Conv2D(filters = filters2, kernel_size = (kernel_size, kernel_size), name = conv_name_base + '2b')(X)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2b')(X)\n X = Activation('relu', name = conv_name_base + \"2b_relu\")(X)\n \n # Third component of main path \n X = Conv2D(filters = filters3, kernel_size = (1, 1), name = conv_name_base + '2c')(X)\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2c')(X)\n \n ##### SHORTCUT PATH #### \n X_shortcut = Conv2D(filters = filters3, kernel_size = (1, 1), strides = strides, name = conv_name_base + '1')(input_tensor)\n X_shortcut = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '1')(X_shortcut)\n \n # Final step: Add shortcut value to main path, and pass it through a RELU activation\n X = Add()([X, X_shortcut])\n X = Activation('relu', name = \"res152\" + str(stage) + block + \"_relu\")(X) \n \n return X","def init_layers(self):\n\n # get caching layers activated\n caching_layers = G3WCachingLayer.objects.all()\n for caching_layer in caching_layers:\n self.add_layer(str(caching_layer), caching_layer)","def _init_layers(self):\n cls_branch = []\n for _ in range(self.num_reg_fcs):\n cls_branch.append(Linear(self.embed_dims, self.embed_dims))\n cls_branch.append(nn.LayerNorm(self.embed_dims))\n cls_branch.append(nn.ReLU(inplace=True))\n cls_branch.append(Linear(self.embed_dims, self.cls_out_channels))\n fc_cls = nn.Sequential(*cls_branch)\n\n reg_branch = []\n for _ in range(self.num_reg_fcs):\n reg_branch.append(Linear(self.embed_dims, self.embed_dims))\n reg_branch.append(nn.ReLU())\n reg_branch.append(Linear(self.embed_dims, self.code_size))\n reg_branch = nn.Sequential(*reg_branch)\n\n past_traj_reg_branch = []\n for _ in range(self.num_reg_fcs):\n past_traj_reg_branch.append(\n Linear(self.embed_dims, self.embed_dims))\n past_traj_reg_branch.append(nn.ReLU())\n past_traj_reg_branch.append(\n Linear(self.embed_dims, (self.past_steps + self.fut_steps)*2))\n past_traj_reg_branch = nn.Sequential(*past_traj_reg_branch)\n\n def _get_clones(module, N):\n return nn.ModuleList([copy.deepcopy(module) for i in range(N)])\n\n # last reg_branch is used to generate proposal from\n # encode feature map when as_two_stage is True.\n num_pred = (self.transformer.decoder.num_layers + 1) if \\\n self.as_two_stage else self.transformer.decoder.num_layers\n\n if self.with_box_refine:\n self.cls_branches = _get_clones(fc_cls, num_pred)\n self.reg_branches = _get_clones(reg_branch, num_pred)\n self.past_traj_reg_branches = _get_clones(\n past_traj_reg_branch, num_pred)\n else:\n self.cls_branches = nn.ModuleList(\n [fc_cls for _ in range(num_pred)])\n self.reg_branches = nn.ModuleList(\n [reg_branch for _ in range(num_pred)])\n self.past_traj_reg_branches = nn.ModuleList(\n [past_traj_reg_branch for _ in range(num_pred)])\n if not self.as_two_stage:\n self.bev_embedding = nn.Embedding(\n self.bev_h * self.bev_w, self.embed_dims)","def __call__(\n self, inputs, training, get_all_layers=None, skip_final_dense=False,\n skip_final_dense_with_pool=False, final_pooling=None):\n global ENDING_POINTS\n ENDING_POINTS = []\n\n with self._model_variable_scope():\n inputs = self._preprocess_data(inputs)\n\n inputs = conv3d_fixed_padding(\n inputs=inputs, filters=self.num_filters, \n kernel_size=self.kernel_size, time_kernel_size=self.time_kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = tf.identity(inputs, 'initial_conv')\n\n # We do not include batch normalization or activation functions in V2\n # for the initial conv1 because the first ResNet unit will perform these\n # for both the shortcut and non-shortcut paths as part of the first\n # block's projection. Cf. Appendix of [2].\n if self.resnet_version == 1:\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n\n _fp_size = self.first_pool_size\n _fp_stride = self.first_pool_stride\n if self.first_pool_size:\n inputs = tf.layers.max_pooling3d(\n inputs=inputs, \n pool_size=[_fp_size, _fp_size, _fp_size],\n strides=[_fp_stride, _fp_stride, _fp_stride], \n padding='SAME',\n data_format=self.data_format)\n inputs = tf.identity(inputs, 'initial_max_pool')\n\n ENDING_POINTS.append(inputs)\n\n for i, num_blocks in enumerate(self.block_sizes):\n num_filters = self.num_filters * (2**i)\n inputs = block_layer(\n inputs=inputs, filters=num_filters, bottleneck=self.bottleneck,\n block_fn=self.block_fn, blocks=num_blocks,\n strides=self.block_strides[i], training=training,\n name='block_layer{}'.format(i + 1), data_format=self.data_format)\n curr_res_name = 'res{}'.format(i + 1)\n\n # Only apply the BN and ReLU for model that does pre_activation in each\n # building/bottleneck block, eg resnet V2.\n if self.pre_activation:\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n ENDING_POINTS.append(inputs)\n\n if skip_final_dense:\n bs = inputs.get_shape().as_list()[0]\n # Along the temporal dimension\n inputs = tf.reduce_mean(inputs, axis=2)\n if final_pooling is None:\n inputs = tf.reshape(inputs, [bs, -1])\n return inputs\n if final_pooling == 1:\n axes = [2, 3] if self.data_format == 'channels_first' else [1, 2]\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\n return tf.reshape(inputs, [bs, 1 * 1 * self.final_size])\n \n # The current top layer has shape\n # `batch_size x pool_size x pool_size x final_size`.\n # ResNet does an Average Pooling layer over pool_size,\n # but that is the same as doing a reduce_mean. We do a reduce_mean\n # here because it performs better than AveragePooling2D.\n axes = [2, 3, 4] if self.data_format == 'channels_first' else [1, 2, 3]\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\n inputs = tf.identity(inputs, 'final_reduce_mean')\n\n if skip_final_dense_with_pool:\n return inputs\n\n inputs = self._get_final_dense(inputs)\n if not get_all_layers:\n return inputs\n else:\n return inputs, ENDING_POINTS","def make_feature_layers(self, config):\n raise NotImplementedError","def add_layer_data(prj_id, expl_id, exploration):\n layers = tbl_explsoilboring[\n (tbl_explsoilboring.lng_KeyProject == prj_id) &\n (tbl_explsoilboring.txt_KeyExplorationName == expl_id)]\n\n # add new column with layer numbers\n layers['layer'] = [i for i in range(1, len(layers.index) + 1)]\n\n # get SPT data\n spt_data = tbl_explboring[\n (tbl_explboring.lng_KeyProject == prj_id) &\n (tbl_explboring.txt_KeyExplorationName == expl_id)]\n\n if not spt_data.empty and not layers.empty:\n spt_depths = spt_data.dbl_KeyDepth\n layer_i = []\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\n for d in spt_depths:\n if d <= max_d:\n layr = layers[\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\n else:\n layr = None\n layer_i.append(layr)\n spt_data['layer'] = layer_i\n avg_spt = spt_data[\n ['layer', 'dbl_FieldBlowCount']].groupby('layer').agg('mean')\n else:\n avg_spt = pd.DataFrame()\n\n # get lab data\n lab_data = tbl_expllab[\n (tbl_expllab.lng_KeyProject == prj_id) &\n (tbl_expllab.txt_KeyExplorationName == expl_id)]\n\n if not lab_data.empty and not layers.empty:\n lab_depths = lab_data.dbl_KeyDepth\n layer_i = []\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\n for d in lab_depths:\n if d <= max_d:\n layr = layers[\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\n else:\n layr = None\n layer_i.append(layr)\n lab_data['layer'] = layer_i\n avg_lab = lab_data[\n ['layer', 'dbl_MoistureContent', 'dbl_TotalUnitWeight',\n 'dbl_LiquidLimit', 'dbl_PlasticityIndex', 'dbl_Cohesion']\n ].groupby('layer').agg('mean')\n else:\n avg_lab = pd.DataFrame()\n\n for i in layers.index:\n index = int(layers.layer[i])\n if index == 1:\n height = tbl_explsoilboring['dbl_KeyDepthToBottom'][i] * 0.00328084\n else:\n height = (tbl_explsoilboring['dbl_KeyDepthToBottom'][i]\n - tbl_explsoilboring['dbl_KeyDepthToBottom'][i-1]) \\\n * 0.00328084\n soil_type = tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i] \\\n if not pd.isna(\n tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i]) else None\n soil_type = primary_explsoil[soil_type] if soil_type else None\n uscs_type = tbl_explsoilboring['txt_USCSCode'][i] if not \\\n pd.isna(tbl_explsoilboring['txt_USCSCode'][i]) else None\n if not soil_type and uscs_type:\n soil_type = uscs_type\n description = tbl_explsoilboring['txt_LayerDescription'][i] \\\n if not pd.isna(tbl_explsoilboring['txt_LayerDescription'][i]) \\\n else None\n nval = tbl_explsoilboring['dbl_BlowCountInterpreted'][i] \\\n if not pd.isna(tbl_explsoilboring['dbl_BlowCountInterpreted'][i]) \\\n else None\n nval_from_avg_spt = avg_spt[avg_spt.index == index]\n if not nval and not nval_from_avg_spt.empty:\n nval = nval_from_avg_spt.values[0][0] if not pd.isna(\n nval_from_avg_spt.values[0][0]) else None\n angle = tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i] \\\n if not pd.isna(\n tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i]) else None\n data_from_avg_lab = avg_lab[avg_lab.index == index]\n cohesion = tbl_explsoilboring['dbl_CohesionInterpreted'][i] \\\n * 2.0885434273e-05 if not pd.isna(\n tbl_explsoilboring['dbl_CohesionInterpreted'][i]) else None\n if not cohesion and not data_from_avg_lab.empty:\n cohesion = data_from_avg_lab.values[0][4] * 2.0885434273e-05 \\\n if not pd.isna(data_from_avg_lab.values[0][4]) else None\n if cohesion:\n cohesion = cohesion if cohesion < 99 else None\n tuw = tbl_explsoilboring['dbl_UnitWeightInterpreted'][i] * 0.062428 \\\n if not pd.isna(tbl_explsoilboring['dbl_UnitWeightInterpreted'][i]) \\\n else None\n if not tuw and not data_from_avg_lab.empty:\n tuw = data_from_avg_lab.values[0][1] * 0.062428 if not pd.isna(\n data_from_avg_lab.values[0][1]) else None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][0]):\n water_content = int(data_from_avg_lab.values[0][0])\n else:\n water_content = None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][2]):\n liquid_limit = int(data_from_avg_lab.values[0][2])\n else:\n liquid_limit = None\n if not data_from_avg_lab.empty and not pd.isna(\n data_from_avg_lab.values[0][3]):\n plasticity = int(data_from_avg_lab.values[0][3])\n else:\n plasticity = None\n layer = Layers(\n boring=exploration,\n index=index,\n soil_type=soil_type,\n height=height,\n description=description,\n nval=nval,\n ssuu=cohesion,\n friction_angle=angle,\n tuw=tuw,\n water_content=water_content,\n liquid_limit=liquid_limit,\n plasticity=plasticity\n )\n db.session.add(layer)","def layers(self):\n return self['layers']","def expandWithoutMutex(self, previousLayer):\n previousLayerProposition = previousLayer.getPropositionLayer()\n \"*** YOUR CODE HERE ***\"","def __init__(self, layers) -> None:\n super(FeaturizerPhi, self).__init__()\n self.layers = layers","def filesystem_layers(self):\n pass","def renderLayerPostProcess(*args, keepImages: bool=True, sceneName: Union[AnyStr, bool]=\"\",\n q=True, query=True, **kwargs)->Union[None, Any]:\n pass","def update(self):\r\n\r\n # Update the vision frames in the system\r\n self._system.update()\r\n\r\n # Create blank PIL images to hold the video streams\r\n layered = PIL.Image.new('RGBA', (400, 400))\r\n stacked = PIL.Image.new('RGBA', (200, 800))\r\n control = PIL.Image.new('RGBA', (600, 800))\r\n\r\n focalpoint = self._system[self._appString[\"device\"].get()].focalpoint()\r\n # print(focalpoint)\r\n\r\n # Get each vision key and vision for the selected device\r\n visionList = [(visionKey, vision) for visionKey, vision in self._system[self._appString[\"device\"].get()]]\r\n\r\n # Loop through each vision in the vision list\r\n for i, (visionKey, vision) in enumerate(visionList):\r\n\r\n # Grab the frames from the vision when it is \"curr\"\r\n frameList = [frame for frameKey, frame in vision if frameKey==self._appString[\"frame\"].get()]\r\n\r\n # Loop through each frame in the frame list\r\n for frame in frameList:\r\n\r\n # Get the properties and turn the image into RGBA\r\n ratio, size = vision.properties()\r\n rgbFrame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGBA)\r\n\r\n # print(rgbFrame.shape)\r\n width, height, channels = rgbFrame.shape\r\n\r\n # Paste the images together in layered\r\n\r\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (int(400 * ratio), int(400 * ratio))))\r\n layered.paste(imgFrame, (int(200 * (1 - ratio)), int(200 * (1 - ratio))))\r\n\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 // width)), int(200 * (1 - ratio) - focalpoint[1] * (200 // height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1)), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200/width) / ratio), int(200 * (1 - ratio) - focalpoint[1] * (200/height) / ratio)))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1) / 200), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1) / 200)))\r\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (400//width * (1- ratio))), int(200 * (1 - ratio) - focalpoint[1] * (400//height * (1 - ratio)))))\r\n\r\n # Paste the images together in stacked\r\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (200, 200)))\r\n stacked.paste(imgFrame, (0, 200 * i))\r\n\r\n # Add the stacked image to the canvas\r\n self._pilFrames[\"stacked\"] = PIL.ImageTk.PhotoImage(image=stacked)\r\n self._appCanvas[\"stacked\"].create_image(100, 0, image=self._pilFrames[\"stacked\"], anchor=tkinter.NW)\r\n\r\n # Add the layered image to the canvas\r\n self._pilFrames[\"layered\"] = PIL.ImageTk.PhotoImage(image=layered)\r\n self._appCanvas[\"layered\"].create_image(0, 0, image=self._pilFrames[\"layered\"], anchor=tkinter.NW)\r\n\r\n # Add the control image to the canvas\r\n imgFrame = cv2.cvtColor(self._system[self._appString[\"device\"].get()][self._appString[\"vision\"].get()][self._appString[\"frame\"].get()], cv2.COLOR_BGR2RGBA)\r\n control = PIL.Image.fromarray(cv2.resize(imgFrame, (600, 600)))\r\n self._pilFrames[\"control\"] = PIL.ImageTk.PhotoImage(image=control)\r\n self._appCanvas[\"control\"].create_image(100, 90, image=self._pilFrames[\"control\"], anchor=tkinter.NW)\r\n\r\n # Continue to update with a delay of 15\r\n self.after(15, self.update)","def _get_layers(self) :\n \n return self._layers","def initLayers(self):\n self.addRow(self.currentLayerData)\n self.selected = self.jobRow.child(0)\n self.setSelectedFromItem(self.selected)","def layers(self):\r\n return self._flc.layers","def num_layers(self): # -> int:\n ...","def get_patches(rimage, gimage, mimage, num_patches=48, patch_size=80, patch_stride=80):\n num_FSpatches = 16\n num_RApatches = 32\n rpatches = []\n gpatches = []\n mpatches = []\n #R_imgs = ((rimage+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'rainy.jpg', R_imgs[0,:,:,:])\n for i in range(int(math.sqrt(num_FSpatches))):\n for j in range(int(math.sqrt(num_FSpatches))):\n point_x = patch_stride*i\n point_y = patch_stride*j\n rpatch = rimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n #print(point_x)\n #print(point_y)\n #print(point_y+patch_size)\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'patch_%d_%d.jpg'%(i,j), P_imgs[0,:,:,:])\n #print(rpatch.shape)\n rpatches.append(rpatch)\n #print(np.array(rpatches).shape)\n gpatch = gimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n gpatches.append(gpatch)\n mpatch = mimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\n mpatches.append(mpatch)\n\n for k in range(num_RApatches):\n point1 = random.randint(0,240) # 116 comes from the image source size (320) - the patch dimension (80)\n point2 = random.randint(0,240)\n #rpatch = tf.image.crop_to_bounding_box(rimage, point1, point2, patch_size, patch_size)\n rpatch = rimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\n #scipy.misc.imsave('results'+'/' + 'patch_%d.jpg'%i, P_imgs[0,:,:,:])\n #print(rpatch.shape)\n rpatches.append(rpatch)\n #print(np.array(rpatches).shape)\n gpatch = gimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n gpatches.append(gpatch)\n mpatch = mimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\n mpatches.append(mpatch)\n\n rpatches = np.array(rpatches)\n rpatches = np.squeeze(rpatches)\n #print(rpatches.shape)\n gpatches = np.array(gpatches)\n gpatches = np.squeeze(gpatches)\n mpatches = np.array(mpatches)\n mpatches = np.squeeze(mpatches)\n #assert rpatches.get_shape().dims == [num_patches, patch_size, patch_size, 3]\n assert rpatches.shape == (num_patches, patch_size, patch_size, 3)\n return rpatches, gpatches, mpatches","def _init_layers(self):\n self._init_predictor()\n if self.use_edge_fusion:\n self._init_edge_module()","def lightened_v2(inputs, is_training=True,\n dropout_keep_prob=0.8,\n reuse=None,\n scope='LightenedV1'):\n end_points = {}\n \n with tf.variable_scope(scope, 'LightenedV1', [inputs], reuse=reuse):\n with slim.arg_scope([slim.batch_norm, slim.dropout],\n is_training=is_training):\n with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],\n stride=1, padding='SAME'):\n \n endpoints = {}\n \n net=conv(inputs, 48, 9, stride=1, padding='VALID', scope='conv1_9x9')\n end_points['conv1_9x9'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool1')\n end_points['pool1'] = net\n \n net=conv(net, 96, 5, stride=1, padding='VALID', scope='conv2_5x5')\n end_points['conv2_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool2')\n end_points['pool2'] = net\n\n net=conv(net, 128, 5, stride=1, padding='VALID', scope='conv3_5x5')\n end_points['conv3_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool3')\n end_points['pool3'] = net\n\n net=conv(net, 192, 4, stride=1, padding='VALID', scope='conv4_4x4')\n end_points['conv4_5x5'] = net\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool4')\n end_points['pool4'] = net\n\n with tf.variable_scope('Logits'):\n net = slim.flatten(net)\n net = slim.dropout(net, dropout_keep_prob, is_training=is_training,\n scope='Dropout') \n endpoints['PreLogitsFlatten'] = net\n \n return net, end_points","def layerListObject(self, parent):\n pass","def __init__(self, batch_env):\n\n dims = [84, 84]\n nature_transform = lambda o: tf.image.rgb_to_grayscale( # pylint: disable=g-long-lambda\n tf.image.resize_images(o, dims))\n\n super(WarpFrameWrapper, self).__init__(batch_env, transform_observation=(\n nature_transform, dims, tf.float32))","def update_layer(LayerId=None, Name=None, Shortname=None, Attributes=None, CloudWatchLogsConfiguration=None, CustomInstanceProfileArn=None, CustomJson=None, CustomSecurityGroupIds=None, Packages=None, VolumeConfigurations=None, EnableAutoHealing=None, AutoAssignElasticIps=None, AutoAssignPublicIps=None, CustomRecipes=None, InstallUpdatesOnBoot=None, UseEbsOptimizedInstances=None, LifecycleEventConfiguration=None):\n pass","def addExportLayerToCoreml(builder):\n outputNames = [output.name for output in builder.spec.description.output]\n\n for i, outputName in enumerate(outputNames):\n # formulas: https://github.com/ultralytics/yolov5/issues/471\n builder.add_activation(\n name=f\"sigmoid_{outputName}\",\n non_linearity=\"SIGMOID\",\n input_name=outputName,\n output_name=f\"{outputName}_sigmoid\",\n )\n\n ### Coordinates calculation ###\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2) -> nC = 640 / strides[i]\n builder.add_slice(\n name=f\"slice_coordinates_xy_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_sliced_coordinates_xy\",\n axis=\"width\",\n start_index=0,\n end_index=2,\n )\n # x,y * 2\n builder.add_elementwise(\n name=f\"multiply_xy_by_two_{outputName}\",\n input_names=[f\"{outputName}_sliced_coordinates_xy\"],\n output_name=f\"{outputName}_multiplied_xy_by_two\",\n mode=\"MULTIPLY\",\n alpha=2,\n )\n # x,y * 2 - 0.5\n builder.add_elementwise(\n name=f\"subtract_0_5_from_xy_{outputName}\",\n input_names=[f\"{outputName}_multiplied_xy_by_two\"],\n output_name=f\"{outputName}_subtracted_0_5_from_xy\",\n mode=\"ADD\",\n alpha=-0.5,\n )\n grid = make_grid(featureMapDimensions[i], featureMapDimensions[i]).numpy()\n # x,y * 2 - 0.5 + grid[i]\n builder.add_bias(\n name=f\"add_grid_from_xy_{outputName}\",\n input_name=f\"{outputName}_subtracted_0_5_from_xy\",\n output_name=f\"{outputName}_added_grid_xy\",\n b=grid,\n shape_bias=grid.shape,\n )\n # (x,y * 2 - 0.5 + grid[i]) * stride[i]\n builder.add_elementwise(\n name=f\"multiply_xy_by_stride_{outputName}\",\n input_names=[f\"{outputName}_added_grid_xy\"],\n output_name=f\"{outputName}_calculated_xy\",\n mode=\"MULTIPLY\",\n alpha=strides[i],\n )\n\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2)\n builder.add_slice(\n name=f\"slice_coordinates_wh_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_sliced_coordinates_wh\",\n axis=\"width\",\n start_index=2,\n end_index=4,\n )\n # w,h * 2\n builder.add_elementwise(\n name=f\"multiply_wh_by_two_{outputName}\",\n input_names=[f\"{outputName}_sliced_coordinates_wh\"],\n output_name=f\"{outputName}_multiplied_wh_by_two\",\n mode=\"MULTIPLY\",\n alpha=2,\n )\n # (w,h * 2) ** 2\n builder.add_unary(\n name=f\"power_wh_{outputName}\",\n input_name=f\"{outputName}_multiplied_wh_by_two\",\n output_name=f\"{outputName}_power_wh\",\n mode=\"power\",\n alpha=2,\n )\n # (w,h * 2) ** 2 * anchor_grid[i]\n anchor = (\n anchorGrid[i]\n .expand(-1, featureMapDimensions[i], featureMapDimensions[i], -1)\n .numpy()\n )\n builder.add_load_constant_nd(\n name=f\"anchors_{outputName}\",\n output_name=f\"{outputName}_anchors\",\n constant_value=anchor,\n shape=anchor.shape,\n )\n builder.add_elementwise(\n name=f\"multiply_wh_with_achors_{outputName}\",\n input_names=[f\"{outputName}_power_wh\", f\"{outputName}_anchors\"],\n output_name=f\"{outputName}_calculated_wh\",\n mode=\"MULTIPLY\",\n )\n\n builder.add_concat_nd(\n name=f\"concat_coordinates_{outputName}\",\n input_names=[f\"{outputName}_calculated_xy\", f\"{outputName}_calculated_wh\"],\n output_name=f\"{outputName}_raw_coordinates\",\n axis=-1,\n )\n builder.add_scale(\n name=f\"normalize_coordinates_{outputName}\",\n input_name=f\"{outputName}_raw_coordinates\",\n output_name=f\"{outputName}_raw_normalized_coordinates\",\n W=torch.tensor([1 / 640]).numpy(),\n b=0,\n has_bias=False,\n )\n\n ### Confidence calculation ###\n builder.add_slice(\n name=f\"slice_object_confidence_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_object_confidence\",\n axis=\"width\",\n start_index=4,\n end_index=5,\n )\n builder.add_slice(\n name=f\"slice_label_confidence_{outputName}\",\n input_name=f\"{outputName}_sigmoid\",\n output_name=f\"{outputName}_label_confidence\",\n axis=\"width\",\n start_index=5,\n end_index=0,\n )\n # confidence = object_confidence * label_confidence\n builder.add_multiply_broadcastable(\n name=f\"multiply_object_label_confidence_{outputName}\",\n input_names=[\n f\"{outputName}_label_confidence\",\n f\"{outputName}_object_confidence\",\n ],\n output_name=f\"{outputName}_raw_confidence\",\n )\n\n # input: (1, 3, nC, nC, 85), output: (3 * nc^2, 85)\n builder.add_flatten_to_2d(\n name=f\"flatten_confidence_{outputName}\",\n input_name=f\"{outputName}_raw_confidence\",\n output_name=f\"{outputName}_flatten_raw_confidence\",\n axis=-1,\n )\n builder.add_flatten_to_2d(\n name=f\"flatten_coordinates_{outputName}\",\n input_name=f\"{outputName}_raw_normalized_coordinates\",\n output_name=f\"{outputName}_flatten_raw_coordinates\",\n axis=-1,\n )\n\n builder.add_concat_nd(\n name=\"concat_confidence\",\n input_names=[\n f\"{outputName}_flatten_raw_confidence\" for outputName in outputNames\n ],\n output_name=\"raw_confidence\",\n axis=-2,\n )\n builder.add_concat_nd(\n name=\"concat_coordinates\",\n input_names=[\n f\"{outputName}_flatten_raw_coordinates\" for outputName in outputNames\n ],\n output_name=\"raw_coordinates\",\n axis=-2,\n )\n\n builder.set_output(\n output_names=[\"raw_confidence\", \"raw_coordinates\"],\n output_dims=[(25200, numberOfClassLabels), (25200, 4)],\n )","def stage(self, stage):\n self._stage = stage\n self._layer = Sdf.Layer.CreateAnonymous()\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)","def __init__(self,\n decoder_options,\n panoptic_deeplab_options,\n bn_layer=tf.keras.layers.BatchNormalization):\n super(PanopticDeepLab, self).__init__(name='PanopticDeepLab')\n\n low_level_feature_keys = [\n item.feature_key for item in panoptic_deeplab_options.low_level\n ]\n low_level_channels_project = [\n item.channels_project for item in panoptic_deeplab_options.low_level\n ]\n\n self._semantic_decoder = PanopticDeepLabSingleDecoder(\n high_level_feature_name=decoder_options.feature_key,\n low_level_feature_names=low_level_feature_keys,\n low_level_channels_project=low_level_channels_project,\n aspp_output_channels=decoder_options.aspp_channels,\n decoder_output_channels=decoder_options.decoder_channels,\n atrous_rates=decoder_options.atrous_rates,\n name='semantic_decoder',\n aspp_use_only_1x1_proj_conv=decoder_options.aspp_use_only_1x1_proj_conv,\n decoder_conv_type=decoder_options.decoder_conv_type,\n bn_layer=bn_layer)\n self._semantic_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.semantic_head.head_channels,\n panoptic_deeplab_options.semantic_head.output_channels,\n common.PRED_SEMANTIC_LOGITS_KEY,\n name='semantic_head',\n conv_type=panoptic_deeplab_options.semantic_head.head_conv_type,\n bn_layer=bn_layer)\n\n self._instance_decoder = None\n self._instance_center_head = None\n self._instance_regression_head = None\n\n if panoptic_deeplab_options.instance.enable:\n if panoptic_deeplab_options.instance.low_level_override:\n low_level_options = panoptic_deeplab_options.instance.low_level_override\n else:\n low_level_options = panoptic_deeplab_options.low_level\n\n # If instance_decoder is set, use those options; otherwise reuse the\n # architecture as defined for the semantic decoder.\n if panoptic_deeplab_options.instance.HasField(\n 'instance_decoder_override'):\n decoder_options = (panoptic_deeplab_options.instance\n .instance_decoder_override)\n\n low_level_feature_keys = [item.feature_key for item in low_level_options]\n low_level_channels_project = [\n item.channels_project for item in low_level_options\n ]\n\n self._instance_decoder = PanopticDeepLabSingleDecoder(\n high_level_feature_name=decoder_options.feature_key,\n low_level_feature_names=low_level_feature_keys,\n low_level_channels_project=low_level_channels_project,\n aspp_output_channels=decoder_options.aspp_channels,\n decoder_output_channels=decoder_options.decoder_channels,\n atrous_rates=decoder_options.atrous_rates,\n name='instance_decoder',\n aspp_use_only_1x1_proj_conv=(\n decoder_options.aspp_use_only_1x1_proj_conv),\n decoder_conv_type=decoder_options.decoder_conv_type,\n bn_layer=bn_layer)\n self._instance_center_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.instance.center_head.head_channels,\n panoptic_deeplab_options.instance.center_head.output_channels,\n common.PRED_CENTER_HEATMAP_KEY,\n name='instance_center_head',\n conv_type=(\n panoptic_deeplab_options.instance.center_head.head_conv_type),\n bn_layer=bn_layer)\n self._instance_regression_head = PanopticDeepLabSingleHead(\n panoptic_deeplab_options.instance.regression_head.head_channels,\n panoptic_deeplab_options.instance.regression_head.output_channels,\n common.PRED_OFFSET_MAP_KEY,\n name='instance_regression_head',\n conv_type=(\n panoptic_deeplab_options.instance.regression_head.head_conv_type),\n bn_layer=bn_layer)","def develop_image(pipeline, camera=None, batch=None, image=None, patch_size=0, patches=2, root_dir='./data', pipeline_args=None):\n\n if camera is not None:\n supported_cameras = fsutil.listdir(os.path.join(root_dir, 'models', 'nip'), '.*')\n if camera not in supported_cameras:\n raise ValueError('Camera data not found ({})! Available cameras: {}'.format(camera, ', '.join(supported_cameras)))\n root_dirname = os.path.join(root_dir, 'models', 'nip', camera)\n data_dirname = os.path.join(root_dir, 'raw', 'training_data', camera)\n\n if patch_size != 0 and (patch_size < 4 or patch_size > 2048):\n raise ValueError('Patch size seems to be invalid!')\n\n # Lazy imports to minimize delay for invalid command line parameters\n import numpy as np\n import imageio as io\n import matplotlib.pyplot as plt\n import tensorflow as tf\n from models import pipelines\n\n # Construct the NIP model ---------------------------------------------------------------------\n \n if os.path.isdir(pipeline):\n # Restore a NIP model from a training log\n model = tfmodel.restore(pipeline, pipelines)\n else:\n # Construct the NIP model from class name (and optional arguments)\n if pipeline_args is None:\n model = getattr(pipelines, pipeline)()\n else:\n model = getattr(pipelines, pipeline)(**pipeline_args)\n\n loaded_model = False\n candidate_dirs = [os.path.join(root_dirname, model.model_code), os.path.join(root_dirname)]\n for candidate in candidate_dirs:\n if os.path.isdir(candidate):\n model.load_model(candidate)\n loaded_model = True\n break\n\n if not loaded_model:\n raise FileNotFoundError(f'Could not find the corresponding model: {candidate_dirs}')\n\n # Load image(s) -------------------------------------------------------------------------------\n \n if image is None and batch is not None:\n print('Loading a batch of {} images'.format(batch))\n data = dataset.Dataset(data_dirname, n_images=0, v_images=batch, val_rgb_patch_size=patch_size or 256, val_n_patches=patches)\n sample_x, sample_y = data.next_validation_batch(0, data.count_validation)\n\n with open('config/cameras.json') as f:\n cameras = json.load(f)\n cfa, srgb = cameras[camera]['cfa'], np.array(cameras[camera]['srgb'])\n\n elif image is not None:\n print('Loading a RAW image {}'.format(image))\n sample_x, cfa, srgb, _ = raw.unpack(image, expand=True)\n sample_y = raw.process(image, brightness=None, expand=True)\n\n if isinstance(model, pipelines.ClassicISP):\n print('Configuring ISP to CFA: {} & sRGB {}'.format(cfa, srgb.round(2).tolist()))\n model.set_cfa_pattern(cfa)\n model.set_srgb_conversion(srgb)\n\n sample_Y = model.process(sample_x).numpy()\n\n if patch_size > 0:\n xx = (sample_y.shape[2] - patch_size) // 2\n yy = (sample_y.shape[1] - patch_size) // 2\n sample_y = sample_y[:, yy:yy+patch_size, xx:xx+patch_size, :]\n sample_Y = sample_Y[:, yy:yy+patch_size, xx:xx+patch_size, :]\n\n psnrs = metrics.psnr(sample_y, sample_Y)\n ssims = metrics.ssim(sample_y, sample_Y)\n\n print('sample x: {}'.format(sample_x.shape))\n print('sample y: {}'.format(sample_y.shape))\n print('sample Y: {}'.format(sample_Y.shape))\n\n # Plot images ---------------------------------------------------------------------------------\n if len(sample_y) > 1:\n sample_y = plots.thumbnails(sample_y, batch, True)\n sample_Y = plots.thumbnails(sample_Y, batch, True)\n else:\n sample_y = sample_y.squeeze()\n sample_Y = sample_Y.squeeze()\n\n print('thumbnails: {}'.format(sample_y.shape))\n\n ncols = 1 if sample_y.shape[1] > sample_y.shape[0] else 2\n nrows = 2 if ncols == 1 else 1\n fig, axes = plt.subplots(nrows, ncols)\n\n plots.image(sample_Y, '{}, PSNR={:.1f} dB, SSIM={:.2f} : {{}}'.format(model.model_code, float(psnrs.mean()), float(ssims.mean())), axes=axes[0])\n plots.image(sample_y, 'Target RGB images () : {}', axes=axes[1])\n\n plt.show()\n plt.close()","def __init__(self,\n high_level_feature_name,\n low_level_feature_names,\n low_level_channels_project,\n aspp_output_channels,\n decoder_output_channels,\n atrous_rates,\n name,\n aspp_use_only_1x1_proj_conv=False,\n decoder_conv_type='depthwise_separable_conv',\n bn_layer=tf.keras.layers.BatchNormalization):\n super(PanopticDeepLabSingleDecoder, self).__init__(name=name)\n self._channel_axis = 3\n\n self._aspp = aspp.ASPP(\n aspp_output_channels,\n atrous_rates,\n aspp_use_only_1x1_proj_conv=aspp_use_only_1x1_proj_conv,\n name='aspp',\n bn_layer=bn_layer)\n self._high_level_feature_name = high_level_feature_name\n\n if len(low_level_feature_names) != len(low_level_channels_project):\n raise ValueError('The Panoptic-DeepLab decoder requires the same number '\n 'of low-level features as the number of low-level '\n 'projection channels. But got %d and %d.'\n % (len(low_level_feature_names),\n len(low_level_channels_project)))\n\n self._low_level_feature_names = low_level_feature_names\n\n for i, channels_project in enumerate(low_level_channels_project):\n # Check if channel sizes increases and issue a warning.\n if i > 0 and low_level_channels_project[i - 1] < channels_project:\n logging.warning(\n 'The low level projection channels usually do not '\n 'increase for features with higher spatial resolution. '\n 'Please make sure, this behavior is intended.')\n current_low_level_conv_name, current_fusion_conv_name = (\n utils.get_low_level_conv_fusion_conv_current_names(i))\n utils.safe_setattr(\n self, current_low_level_conv_name, convolutions.Conv2DSame(\n channels_project,\n kernel_size=1,\n name=utils.get_layer_name(current_low_level_conv_name),\n use_bias=False,\n use_bn=True,\n bn_layer=bn_layer,\n activation='relu'))\n\n utils.safe_setattr(\n self, current_fusion_conv_name, convolutions.StackedConv2DSame(\n conv_type=decoder_conv_type,\n num_layers=1,\n output_channels=decoder_output_channels,\n kernel_size=5,\n name=utils.get_layer_name(current_fusion_conv_name),\n use_bias=False,\n use_bn=True,\n bn_layer=bn_layer,\n activation='relu'))","def inception_network():\n input = K.Input(shape=(224, 224, 3))\n\n conv_1 = K.layers.Conv2D(\n filters=64,\n kernel_size=7,\n strides=2,\n padding='same',\n activation='relu'\n )(input)\n\n pool_1 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(conv_1)\n\n conv_2 = K.layers.Conv2D(\n filters=64,\n kernel_size=1,\n strides=1,\n padding='valid',\n activation='relu'\n )(pool_1)\n\n conv_3 = K.layers.Conv2D(\n filters=192,\n kernel_size=3,\n strides=1,\n padding='same',\n activation='relu'\n )(conv_2)\n\n pool_2 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(conv_3)\n\n inception_1 = inception_block(pool_2, (64, 96, 128, 16, 32, 32))\n inception_2 = inception_block(inception_1, (128, 128, 192, 32, 96, 64))\n\n pool_3 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same'\n )(inception_2)\n\n inception_3 = inception_block(pool_3, (192, 96, 208, 16, 48, 64))\n inception_4 = inception_block(inception_3, (160, 112, 224, 24, 64, 64))\n inception_5 = inception_block(inception_4, (128, 128, 256, 24, 64, 64))\n inception_6 = inception_block(inception_5, (112, 144, 288, 32, 64, 64))\n inception_7 = inception_block(inception_6, (256, 160, 320, 32, 128, 128))\n\n pool_4 = K.layers.MaxPooling2D(\n pool_size=3,\n strides=2,\n padding='same',\n )(inception_7)\n\n inception_8 = inception_block(pool_4, (256, 160, 320, 32, 128, 128))\n inception_9 = inception_block(inception_8, (384, 192, 384, 48, 128, 128))\n\n pool_5 = K.layers.AveragePooling2D(\n pool_size=7,\n strides=1,\n padding='valid'\n )(inception_9)\n\n dropout = K.layers.Dropout(rate=0.4)(pool_5)\n linear = K.layers.Dense(units=1000, activation='softmax')(dropout)\n\n return K.Model(inputs=input, outputs=linear)","def __init__(self):\n super(GatherLastLayer, self).__init__()","def SetUpLayerManager(self):\n pass","def _save_data(self, outputs, latents):\n list(map(lambda e, v: e.append(v), self.curr_patches, outputs))\n list(map(lambda e, v: e.append(v), self.curr_latents, latents))\n self.count_patches += len(self.curr_patches[0][-1])\n\n while self.count_patches >= self.patches_ref:\n aux_patches = list(map(lambda e: tf.concat(e, axis=0),\n self.curr_patches))\n aux_latents = list(map(lambda e: tf.concat(e, axis=0),\n self.curr_latents))\n\n data_name = str((self.out_folder / str(Folders.RAW_DATA))\n / self.curr_img.stem)\n for lvl, (aux_p, aux_l) in enumerate(zip(aux_patches, aux_latents)):\n patches, remaining = tf.split(\n aux_p, [self.patches_ref, len(aux_p) - self.patches_ref],\n axis=0)\n self.curr_patches[lvl] = [remaining]\n latent, remaining = tf.split(\n aux_l, [self.patches_ref, len(aux_l) - self.patches_ref],\n axis=0)\n self.curr_latents[lvl] = [remaining]\n\n patch_size = len(patches[0])\n with Image.open(self.curr_img) as img:\n # width, height -> height, width (to be compatible with\n # arrays)\n img_size = np.array(img.size)[::-1]\n p_size = np.ceil(img_size / patch_size).astype(int)\n\n img = tf.reshape(\n patches, [*p_size, *self.m_cfg['configs']['patch']])\n img = tf.transpose(img, [0, 2, 1, 3, 4])\n img = tf.reshape(img, [*patch_size * p_size, 3])\n img = tf.image.convert_image_dtype(img, tf.uint8)\n img = img[:img_size[0], :img_size[1], :]\n\n latent_name = data_name + '_latent' + str(lvl)\n serialized = tf.io.serialize_tensor(latent)\n tf.io.write_file(latent_name, serialized)\n\n img_name = data_name + '_' + str(lvl) + self.curr_img.suffix\n tf.io.write_file(img_name, tf.image.encode_png(\n img, compression=0))\n\n with open(Path(data_name).with_suffix('.txt'), 'w') as file:\n file.write(str(self.curr_img))\n\n self.count_patches = len(self.curr_patches[0][0])\n if len(self.img_paths):\n self.curr_img = Path(self.img_paths.pop(0))\n self.patches_ref = self.calc_n_patches(\n self.curr_img, self.m_cfg['configs']['patch'][0])","def registerPatches(self,pl):\n self.set('patchmesh.patches',pl)","def multi_run_wrapper(args):\n\treturn img_preprocessing(*args)","def _set_freeze_layers(self):\n for layer in self.encoder.layers[:self.freeze_layers]:\n layer.trainable = False","def __call__(self, inputs, training):\n\n with self._model_variable_scope():\n if self.data_format == 'channels_first':\n # Convert the inputs from channels_last (NHWC) to channels_first (NCHW).\n # This provides a large performance boost on GPU. See\n # https://www.tensorflow.org/performance/performance_guide#data_formats\n inputs = tf.transpose(a=inputs, perm=[0, 3, 1, 2])\n\n with tf.variable_scope('stem'):\n inputs = conv2d_fixed_padding(\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n inputs = tf.identity(inputs, 'stem_conv1')\n\n inputs = conv2d_fixed_padding(\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\n strides=self.conv_stride, data_format=self.data_format)\n inputs = batch_norm(inputs, training, self.data_format)\n inputs = tf.nn.relu(inputs)\n inputs = tf.identity(inputs, 'stem_conv2')\n\n # stage 1\n with tf.variable_scope('stage1'):\n stage1_out = block_layer(\n inputs=inputs, inp_filters=self.num_filters,\n filters=self.num_filters, bottleneck=True,\n block_fn=_bottleneck_block_v1, blocks=self.block_sizes[0],\n strides=1, training=training,\n name='stage1', data_format=self.data_format)\n\n pre_stage_channels = self.num_filters * 4\n stagex_input = [stage1_out]\n num_inchannels = [pre_stage_channels]\n\n # stage 2 to last stage\n for i, num_modules in enumerate(self.module_sizes[1:], 2):\n with tf.variable_scope('stage{}'.format(i)):\n num_channels = [self.hrnet_size * 2 ** n for n in range(i)]\n transn = Transition(\n stagex_input[-1], num_outchannels=num_channels[-1], \n training=training, data_format=self.data_format)\n stagex_input.append(transn)\n num_inchannels.append(num_inchannels[-1])\n\n for j in range(num_modules):\n if i == len(self.module_sizes) and j == num_modules - 1:\n multi_scale_output = False\n else:\n multi_scale_output = True\n\n stagex_input = HighResolutionModule(\n stagex_input, num_branches=i, \n num_inchannels=num_inchannels, num_channels=num_channels,\n bottleneck=self.bottleneck, block_fn=self.block_fn,\n num_blocks=self.block_sizes[i - 1], training=training,\n name='stage{}_block{}'.format(i, j), data_format=self.data_format,\n multi_scale_output=multi_scale_output)\n num_inchannels = num_channels\n\n return stagex_input","def _augment_pipeline_cfg(self):","def __call__(self, **kwargs):\n segname = 'block_{}_expand_relu'\n blocks = [13, 6, 3, 1]\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\n backbone_out = self._backbone.get_layer('block_16_project')\n\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\n p4 = self._fpn_block(p5, skips[1].output)\n p3 = self._fpn_block(p4, skips[2].output)\n p2 = self._fpn_block(p3, skips[3].output)\n\n s5 = self._conv_block(p5, 128)\n s4 = self._conv_block(p4, 128)\n s3 = self._conv_block(p3, 128)\n s2 = self._conv_block(p2, 128)\n\n s5 = tf.keras.layers.UpSampling2D(\n size=(8, 8),\n interpolation='nearest'\n )(s5)\n\n s4 = tf.keras.layers.UpSampling2D(\n size=(4, 4),\n interpolation='nearest'\n )(s4)\n\n s3 = tf.keras.layers.UpSampling2D(\n size=(2, 2),\n interpolation='nearest'\n )(s3)\n\n concat = [s5, s4, s3, s2]\n x = tf.keras.layers.Concatenate()(concat)\n x = tf.keras.layers.Conv2D(\n 64,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n x = tf.keras.layers.BatchNormalization()(x)\n x = tf.keras.layers.Activation('relu')(x)\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\n\n x = tf.keras.layers.Conv2D(\n 1,\n kernel_size=3,\n padding='same',\n kernel_initializer='he_uniform'\n )(x)\n\n out = tf.keras.layers.Activation('sigmoid')(x)\n model = tf.keras.models.Model(\n inputs=self._backbone.input,\n outputs=out\n )\n\n return model","def add_bluprint_layers(self, models_dict, graph, class_names, dimensions,\n show_activation, show_constant_input):\n top_model = models_dict['top-model']\n # Get the top-model color list containing\n # the base color and the layer shades\n top_model_color = top_model[1]\n # get the model\n top_model = top_model[0]\n\n # Get the layers of the model\n layers = top_model[\"config\"][\"layers\"]\n # Loop through the layers\n for layer in layers:\n # If the layer is not a model\n if layer[\"class_name\"] != \"Model\":\n # Get the layer name\n layer_name = layer[\"name\"]\n # If label only layer's class name\n if class_names:\n # Get the layer's information\n layer_info = self.viz_utils.query_layer(layer_name,\n models_dict)\n # Get the layer's class name\n layer_class = layer_info['class_name']\n # If the layer is a a constant input layer,\n # manually specify the class name\n if layer_name.find('constant_input') != -1:\n layer_class = 'Constant Input'\n # Depending on the class name\n # find the the layer shade\n # If the layer is a constant_input layer\n # the color is black\n model_color = top_model_color[1].get(layer_class, \"black\")\n else:\n # If don't use class names for layers\n # then use the layer name from the JSON\n layer_class = layer_name\n model_color = top_model_color[0]\n\n\n # Add the node to the graph\n graph = self.viz_utils.add_nodes(layer_name, graph,\n layer_class, model_color,\n dimensions, show_constant_input)\n\n # Add Blueprint Inbound Edges\n graph = self.connect_blueprint_inbounds(models_dict,\n layer, graph,\n class_names, dimensions,\n show_activation, show_constant_input)\n else:\n # Add Softmod\n graph = self.connect_softmod_in_blueprint(models_dict,\n layer, graph, class_names,\n dimensions, show_activation, show_constant_input)\n\n return graph","def load_registration_as_layers() -> None:\n viewer = getattr(widget, \"viewer\").value\n registration_directory = pathlib.Path(\n getattr(widget, \"registration_output_folder\").value\n )\n add_registered_image_layers(\n viewer, registration_directory=registration_directory\n )","def intermediate_layer(layer, filters, kernel_size):\n layer = Conv2D(filters=filters, kernel_size=kernel_size,\n strides=(2, 2), padding=\"same\")(layer)\n layer = LeakyReLU(alpha=LEAKY_RELU_ALPHA)(layer)\n return layer","def deep_dream_of_extreme_control(FLAGS,model,input_images=[],num_iterations=10,step_size=0.1):\n if len(input_images) == 0:\n # use predefined images\n img_dir='/esat/opal/kkelchte/docker_home/pilot_data/visualization_images'\n input_images=sorted([img_dir+'/'+f for f in os.listdir(img_dir)])\n\n print(\"[tools.py]: extracting deep dream maps of {0} in {1}\".format([os.path.basename(i) for i in input_images], os.path.dirname(input_images[0])))\n \n # experts=np.asarray([[k]*(FLAGS.action_quantity if FLAGS.discrete else 1) for v in sorted(model.factor_offsets.values()) for k in model.factor_offsets.keys() if model.factor_offsets[k]==v]).flatten()\n\n inputs = load_images(input_images, model.input_size[1:])\n \n # collect gradients for output endpoint of evaluation model\n grads={}\n with tf.device('/cpu:0'):\n output_tensor = model.endpoints['eval']['outputs']\n for i in range(output_tensor.shape[1].value):\n layer_loss = output_tensor[:,i]\n gradients = tf.gradients(layer_loss, model.inputs)[0]\n gradients /= (tf.sqrt(tf.reduce_mean(tf.square(gradients))) + 1e-5)\n grads[output_tensor.name+'_'+str(i)]=gradients\n\n\n # apply gradient ascent for all outputs and each input image\n # if number of outputs ==1 apply gradient descent for contrast\n if len(grads.keys())== 1:\n opposite_results={}\n else:\n opposite_results=None\n\n import copy\n results = {}\n for gk in grads.keys(): \n results[gk]=copy.deepcopy(inputs)\n if isinstance(opposite_results,dict): opposite_results[gk]=copy.deepcopy(inputs)\n\n for step in range(num_iterations):\n if step%10==0: print \"{0} step: {1}\".format(time.ctime(), step)\n for i,gk in enumerate(sorted(grads.keys())):\n results[gk] += step_size * model.sess.run(grads[gk], {model.inputs: results[gk]})\n if isinstance(opposite_results,dict):\n opposite_results[gk] -= step_size * model.sess.run(grads[gk], {model.inputs: opposite_results[gk]})\n\n # Normalize results within 0:1 range\n clean_results={}\n for gk in results.keys():\n clean_results[gk]=[]\n for i in range(results[gk].shape[0]):\n clean_results[gk].append(deprocess_image(results[gk][i], one_channel=True))\n # results[gk][i]=deprocess_image(results[gk][i], one_channel=True)\n if isinstance(opposite_results,dict):\n opposite_results[gk][i]=deprocess_image(opposite_results[gk][i])\n\n # combine adjust input images in one overview image\n # one column for each input image\n # one row with each extreme control for separate and difference images\n num_rows=1+len(results.keys())\n fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\n # fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\n # add original images in first row\n for i in range(axes.shape[1]):\n axes[0, i].set_title(os.path.basename(input_images[i]).split('.')[0])\n axes[0, i].imshow(matplotlibprove(inputs[i]), cmap='inferno')\n axes[0, i].axis('off')\n\n # add for each filter the modified input\n row_index=1\n for gk in sorted(results.keys()):\n for i in range(axes.shape[1]):\n # print gk\n # axes[row_index, i].set_title('Grad Asc: '+gk.split('/')[1]+'/'+gk[-1]) \n axes[row_index, i].set_title('Grad Asc: '+gk)\n # axes[row_index, i].set_title(experts[row_index-1])\n\n axes[row_index, i].imshow(np.concatenate((inputs[i],np.expand_dims(clean_results[gk][i],axis=2)), axis=2), cmap='inferno')\n # axes[row_index, i].imshow(matplotlibprove(results[gk][i]), cmap='inferno')\n axes[row_index, i].axis('off')\n row_index+=1\n # In cas of continouos controls: visualize the gradient descent and difference\n # if isinstance(opposite_results,dict):\n # for gk in opposite_results.keys():\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Grad Desc: '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Grad Desc: '+gk) \n # axes[row_index, i].imshow(matplotlibprove(opposite_results[gk][i]), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n \n # # add difference\n # for gk in opposite_results.keys():\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Diff: '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Diff: '+gk) \n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((opposite_results[gk][i]-results[gk][i])**2)), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n # else:\n # # add difference between 2 exteme actions\n # gk_left=sorted(results.keys())[0]\n # gk_right=sorted(results.keys())[-1]\n # for i in range(axes.shape[1]):\n # # axes[row_index, i].set_title('Diff : '+gk.split('/')[1]) \n # axes[row_index, i].set_title('Diff : '+gk) \n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((results[gk_left][i]-results[gk_right][i])**2)), cmap='inferno')\n # axes[row_index, i].axis('off')\n # row_index+=1\n \n \n plt.savefig(FLAGS.summary_dir+FLAGS.log_tag+'/control_dream_maps.jpg',bbox_inches='tight')\n # plt.show()","def build(self, unused_input_shapes):\n self.layers = []\n for i in range(self.num_hidden_layers):\n self.layers.append(\n TransformerDecoderBlock(\n hidden_size=self.hidden_size,\n num_attention_heads=self.num_attention_heads,\n intermediate_size=self.intermediate_size,\n intermediate_activation=self.intermediate_activation,\n hidden_dropout_prob=self.hidden_dropout_prob,\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\n initializer_range=self.initializer_range,\n multi_channel_cross_attention=self.multi_channel_cross_attention,\n name=(\"layer_%d\" % i)))\n super(TransformerDecoder, self).build(unused_input_shapes)","def GetShapesOfLayer(self, *args):\n return _XCAFDoc.XCAFDoc_LayerTool_GetShapesOfLayer(self, *args)","def __init__(self, layer_list_info):\n self.layer_list_info = layer_list_info","def depth_rendering(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\n\n x_indices = np.arange(lf_size[1])\n y_indices = np.arange(lf_size[2])\n b_indices = np.arange(lf_size[0])\n\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\n\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\n stacked = np.stack(grid, 2)\n positions = stacked.reshape(-1, 2)\n\n # Compute the distance from each lf position from the reference view\n # Repeat the elements of this to match the size of the disparity map\n ref_pos = np.array(\n [lf_one_way // 2, lf_one_way // 2])\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\n\n\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\n\n # Compute the shifted pixels\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\n\n #indices for linear interpolation in a square around the central point\n x_low = np.around(x_shifted).astype(int)\n #x_high = x_low + 1\n\n y_low = np.around(y_shifted).astype(int)\n #y_high = y_low + 1\n\n #Place co-ordinates outside the image back into the image\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\n #x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\n #y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\n\n #Gather the interpolation points\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\n #interp_pts_2 = np.stack((x_low_clip, y_high_clip))\n #interp_pts_3 = np.stack((x_high_clip, y_low_clip))\n #interp_pts_4 = np.stack((x_high_clip, y_high_clip))\n\n #Index into the images\n desired_shape = lf_size\n res_1 = torch_big_sample(ref_view, interp_pts_1, desired_shape)\n return res_1\n res_2 = torch_big_sample(ref_view, interp_pts_2, desired_shape)\n res_3 = torch_big_sample(ref_view, interp_pts_3, desired_shape)\n res_4 = torch_big_sample(ref_view, interp_pts_4, desired_shape)\n\n #Compute interpolation weights\n x_low_f = x_low.astype(np.float32)\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\n d_x_high = 1.0 - d_x_low\n y_low_f = y_low.astype(np.float32)\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\n d_y_high = 1.0 - d_y_low\n\n w1 = torch.from_numpy(d_x_low * d_y_low)\n w2 = torch.from_numpy(d_x_low * d_y_high)\n w3 = torch.from_numpy(d_x_high * d_y_low)\n w4 = torch.from_numpy(d_x_high * d_y_high)\n\n #THEY AGREE AT THIS POINT\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape), res_1)\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape), res_2)\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape), res_3)\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape), res_4)\n\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\n torch.add(novel_view, weighted_4, out=novel_view)\n return novel_view","def process_layer(layer_data):\n layer_name = layer_data['name']\n\n # Instantiate layer.\n from tensorflow.python.keras._impl.keras.layers import deserialize as deserialize_layer # pylint: disable=g-import-not-at-top\n\n layer = deserialize_layer(layer_data, custom_objects=custom_objects)\n created_layers[layer_name] = layer\n\n # Gather layer inputs.\n inbound_nodes_data = layer_data['inbound_nodes']\n for node_data in inbound_nodes_data:\n # We don't process nodes (i.e. make layer calls)\n # on the fly because the inbound node may not yet exist,\n # in case of layer shared at different topological depths\n # (e.g. a model such as A(B(A(B(x)))))\n add_unprocessed_node(layer, node_data)","def __init__(self, stage, layer=None, proxyPrimPrefix=\"\", renderPrimPrefix=\"\"):\n\n self._stage = stage\n\n # Layer\n if layer is None:\n self._layer = Sdf.Layer.CreateAnonymous()\n elif isinstance(layer, Sdf.Layer):\n self._layer = layer\n elif isinstance(layer, str):\n self._layer = Sdf.Layer.Open(layer)\n\n self._layer.SetMuted(True)\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\n self._proxyPrimPrefix = proxyPrimPrefix\n self._renderPrimPrefix = renderPrimPrefix","def __reslayer(self, inputs, in_filters, out_filters, stride=1):\n with tf.variable_scope('sub1'):\n kernel = tf.get_variable('weights', [3, 3, in_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n conv = tf.nn.conv2d(inputs, kernel, [1, stride, stride, 1],\n padding='SAME',\n name='conv')\n batch_norm = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\n conv = tf.nn.elu(batch_norm, 'elu')\n\n with tf.variable_scope('sub2'):\n kernel = tf.get_variable('weights',\n [3, 3, out_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n conv = tf.nn.conv2d(conv, kernel, [1, 1, 1, 1], padding='SAME',\n name='conv1')\n bias = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\n\n with tf.variable_scope('subadd'):\n if in_filters != out_filters:\n kernel = tf.get_variable('weights', [1, 1, in_filters, out_filters],\n initializer=xavier_initializer(\n dtype=tf.float32),\n dtype=tf.float32)\n inputs = tf.nn.conv2d(\n inputs, kernel, [1, stride, stride, 1], padding='SAME')\n bias += inputs\n conv = tf.nn.elu(bias, 'elu')\n\n num = np.power(2, np.floor(np.log2(out_filters) / 2))\n\n grid = self.__put_activations_on_grid(conv, (int(num),\n int(out_filters /\n num)))\n tf.summary.image('sub2/activations', grid, max_outputs=1)\n\n return conv","def depth_rendering_gpu(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\n\n x_indices = np.arange(lf_size[1])\n y_indices = np.arange(lf_size[2])\n b_indices = np.arange(lf_size[0])\n\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\n\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\n stacked = np.stack(grid, 2)\n positions = stacked.reshape(-1, 2)\n\n # Compute the distance from each lf position from the reference view\n # Repeat the elements of this to match the size of the disparity map\n ref_pos = np.array(\n [lf_one_way // 2, lf_one_way // 2])\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\n\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\n\n # Compute the shifted pixels\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\n\n #indices for linear interpolation in a square around the central point\n x_low = np.floor(x_shifted).astype(int)\n x_high = x_low + 1\n\n y_low = np.floor(y_shifted).astype(int)\n y_high = y_low + 1\n\n #Place co-ordinates outside the image back into the image\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\n x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\n y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\n\n #Gather the interpolation points\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\n interp_pts_2 = np.stack((x_low_clip, y_high_clip))\n interp_pts_3 = np.stack((x_high_clip, y_low_clip))\n interp_pts_4 = np.stack((x_high_clip, y_high_clip))\n\n #Index into the images\n desired_shape = lf_size\n ref_view = torch.tensor(ref_view, dtype=torch.float32).cuda()\n res_1 = torch_tensor_sample(ref_view, interp_pts_1, desired_shape)\n res_2 = torch_tensor_sample(ref_view, interp_pts_2, desired_shape)\n res_3 = torch_tensor_sample(ref_view, interp_pts_3, desired_shape)\n res_4 = torch_tensor_sample(ref_view, interp_pts_4, desired_shape)\n\n #Compute interpolation weights\n x_low_f = x_low.astype(np.float32)\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\n d_x_high = 1.0 - d_x_low\n y_low_f = y_low.astype(np.float32)\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\n d_y_high = 1.0 - d_y_low\n\n w1 = torch.from_numpy(d_x_low * d_y_low)\n w2 = torch.from_numpy(d_x_low * d_y_high)\n w3 = torch.from_numpy(d_x_high * d_y_low)\n w4 = torch.from_numpy(d_x_high * d_y_high)\n\n #THEY AGREE AT THIS POINT\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape).cuda(), res_1)\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape).cuda(), res_2)\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape).cuda(), res_3)\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape).cuda(), res_4)\n\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\n torch.add(novel_view, weighted_4, out=novel_view)\n return novel_view","def __call__(self, inputs, *args, **kwargs):\n out = inputs\n for i, layer in enumerate(self.layers):\n if i == 0:\n out = layer(out, *args, **kwargs)\n else:\n out = layer(out)\n return out","def add_convolution_layers(self):\n with tf.variable_scope(\"conv_layer_1\"):\n # our input to this layer is from word embeddings. which would be of\n # shape (batch_size, number of words in the sentence, dimension of\n # embedding). For us dimension of embedding is the number of input\n # channels, self.out_channels_1 is the output channels.\n\n # First we have to define a filter. The shape of the filter would be [\n # self.filter_size_1, dimension of embedding, self.out_channels_1]\n filter1 = tf.get_variable(\"filter1\", dtype=tf.float32,\n shape=[self.filter_size_1,\n self.word_embeddings.shape[2],\n self.out_channels_1])\n conv_layer_1 = tf.nn.conv1d(self.word_embeddings, filters=filter1,\n stride=self.strides_1, padding=\"SAME\")\n conv_layer_1 = tf.nn.tanh(conv_layer_1)\n # now add pooling layer\n print(\"adding pooling to layer1\")\n\n if self.max_pooling:\n print(\"Adding max pooling layer\")\n pooling_layer = tf.layers.max_pooling1d\n else:\n print(\"Adding average pooling layer\")\n pooling_layer = tf.layers.average_pooling1d\n conv_layer_1 = pooling_layer(conv_layer_1,\n pool_size=self.pool_size_1,\n strides=self.strides_1,\n padding=\"SAME\")\n\n with tf.variable_scope(\"conv_layer_2\"):\n print(\"adding convolution layer 2\")\n # Input to this layer is conv_layer_1 from above.\n # get another filter now. Shape of this filter would be [\n # self.filter_size_2, self.out_channels_1, self.out_channels_2]\n filter2 = tf.get_variable(\"filter2\", dtype=tf.float32,\n shape=[self.filter_size_2,\n self.out_channels_1,\n self.out_channels_2])\n conv_layer_2 = tf.nn.conv1d(conv_layer_1, filters=filter2,\n stride=self.strides_2, padding=\"SAME\")\n conv_layer_2 = tf.nn.tanh(conv_layer_2)\n print(\"adding pooling to layer2\")\n conv_layer_2 = pooling_layer(conv_layer_2,\n pool_size=self.pool_size_2,\n strides=self.strides_2,\n padding=\"SAME\")\n\n self.cnn_layer = tf.nn.dropout(conv_layer_2, self.dropout_holder)\n self.word_reps = self.cnn_layer\n if self.use_additional and self.hybrid:\n self.cnn_layer = tf.concat([self.word_reps, self.word_embeddings_2], \n axis=-1)\n self.word_reps = self.cnn_layer","def _init_layers(self) -> None:\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\n self.embed_dims = self.self_attn.embed_dims\n self.ffn = FFN(**self.ffn_cfg)\n norms_list = [\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\n for _ in range(3)\n ]\n self.norms = ModuleList(norms_list)","def _embed_composite_feature(self, features, embedding_layers):\n embedding_list = []\n for i in range(len(embedding_layers)):\n embedding_list.append(embedding_layers[i](features[:, :, i]))\n embedding = tf.add_n(embedding_list)\n return embedding","def _add_layer(self, layer_dict, layer_name, input_layers, merge_mode=None, share_params_with=None):\n util.colorprint(layer_name, 'teal')\n \n layer_dict = dict(layer_dict)\n util.colorprint(layer_dict, 'red')\n \n if share_params_with is not None:\n print \"Warning: ignoring share_params_with\"\n \n layer_options = layer_dict[\"options\"]\n layer=None\n if layer_dict[\"type\"]==\"conv2d\":\n #TODO: remove below\n nb_filter, nb_row, nb_col = 3,3,3\n layer = keras.layers.convolutional.Convolution2D(nb_filter, nb_row, nb_col, **layer_options)\n elif layer_dict[\"type\"]==\"dense\":\n dim = layer_dict[\"output_dim\"]\n # del layer_options[\"output_dim\"]\n layer = keras.layers.core.Dense(dim, **layer_options) \n else:\n print \"Ursol Major\"\n RaiseError()\n # TODO: one of the layers is a string\n if isinstance(input_layers, list):\n #this means that there is input from a loop to this layer\n self.model.add_node(layer, name=layer_name, inputs=input_layers, merge_mode=merge_mode)\n else:\n self.model.add_node(layer, name=layer_name, input=input_layers)\n\n return layer_name","def test_2d_local_convolution(batch_exp, input_height, input_width, in_channels, filters, drawer, is_first):\n batch_size = 2 ** batch_exp\n if is_first:\n data_format = \"channels_first\"\n else:\n data_format = \"channels_last\"\n kernel_height = drawer.draw(integers(min_value=1, max_value=input_height-1))\n kernel_width = drawer.draw(integers(min_value=1, max_value=input_width-1))\n\n if data_format == \"channels_last\":\n input_shape = (input_height, input_width, in_channels)\n else:\n input_shape = (in_channels, input_height, input_width)\n\n # Create keras model\n keras_model = keras.Sequential()\n keras_local = keras.layers.LocallyConnected2D(filters, (kernel_height, kernel_width),\n use_bias=True,\n bias_initializer=\"random_uniform\",\n input_shape=input_shape,\n data_format=data_format)\n keras_model.add(keras_local)\n\n # Setup weights\n weights = drawer.draw(arrays(shape=keras_local.kernel.shape.as_list(), dtype=np.float32, elements=floats(min_value=-1e6, max_value=1e6)))\n keras_model.set_weights([weights])\n\n # Translate keras to PyTorch model\n torch_model = translate.translate_layer(keras_local)[0].to(device)\n\n keras_input = np.random.uniform(-10, 10, (batch_size,) + input_shape).astype(np.float32)\n keras_output = keras_model.predict(keras_input)\n\n torch_input = torch.Tensor(keras_input).to(device)\n torch_output = torch_model(torch_input).cpu().data.numpy()\n nan_fill = np.random.uniform(-5, 5)\n assert (np.isnan(keras_output) == np.isnan(torch_output)).all()\n keras_output[np.isnan(keras_output)] = nan_fill\n torch_output[np.isnan(torch_output)] = nan_fill\n assert np.isclose(keras_output, torch_output, atol=1e-3, rtol=1e-3).all()","def __init__(self, config_data, in_dims, layer_num):\n self.layer_num = layer_num\n self.in_dims = in_dims\n self.out_dims = in_dims[\"width\"] * in_dims[\"height\"] * in_dims[\"channels\"]\n self.layer_type = config_data[\"type\"]\n self.name = config_data[\"name\"]\n self.params = []","def multiply(self, layer):\n pass","def forward(self,bottom,top):\n # imgmaps = self.batch_loader.mixup_gen()\n # print(\"DataLayer forward!!\")\n trainX, trainY = self.batch_loader.batch_imgs()\n # print(\"trainX:\",trainX.shape)\n # print(\"trainY:\",trainY.shape)\n # print(\"trainY:\", trainY)\n # print(\"top[0].data.shape:\",top[0].data.shape)\n # print(\"top[1].data.shape:\", top[1].data.shape)\n top[0].data[:, ...] = trainX\n top[1].data[:, ...] = trainY\n # print(\"DataLayer forward!!\")","def apply_net_tiled(net,img):\n\n # borders = [8,20,20] ## border width within each patch that is thrown away after prediction\n # patchshape_padded = [32,240,240] ## the size of the patch that we feed into the net. must be divisible by 8 or net fails.\n # patchshape = [16,200,200] ## must be divisible by 8 to avoid artifacts.\n # stride = [16,200,200] ## same as patchshape in this case\n def f(n,m): return (ceil(n/m)*m)-n ## f(n,m) gives padding needed for n to be divisible by m\n def g(n,m): return (floor(n/m)*m)-n ## f(n,m) gives un-padding needed for n to be divisible by m\n\n a,b,c = img.shape[1:]\n q,r,s = f(a,8),f(b,8),f(c,8) ## calculate extra border needed for stride % 8 = 0\n\n ZPAD,YPAD,XPAD = 8,24,24\n\n img_padded = np.pad(img,[(0,0),(ZPAD,ZPAD+q),(YPAD,YPAD+r),(XPAD,XPAD+s)],mode='constant') ## pad for patch borders\n output = np.zeros(img.shape)\n\n zs = np.r_[:a:16]\n ys = np.r_[:b:200]\n xs = np.r_[:c:200]\n\n for x,y,z in itertools.product(xs,ys,zs):\n qe,re,se = min(z+16,a+q),min(y+200,b+r),min(x+200,c+s)\n ae,be,ce = min(z+16,a),min(y+200,b),min(x+200,c)\n patch = img_padded[:,z:qe+2*ZPAD,y:re+2*YPAD,x:se+2*XPAD]\n patch = torch.from_numpy(patch).cuda().float()\n patch = net(patch[None])[0,:,ZPAD:-ZPAD,YPAD:-YPAD,XPAD:-XPAD].detach().cpu().numpy()\n output[:,z:ae,y:be,x:ce] = patch[:,:ae-z,:be-y,:ce-x]\n\n return output","def layeredTexturePort(*args, annotation: Union[AnyStr, bool]=\"\", backgroundColor:\n Union[List[float, float, float], bool]=None, defineTemplate: AnyStr=\"\",\n docTag: Union[AnyStr, bool]=\"\", dragCallback: Script=None, dropCallback:\n Script=None, enable: bool=True, enableBackground: bool=True,\n enableKeyboardFocus: bool=True, exists: bool=True, fullPathName:\n bool=True, height: Union[int, bool]=0, highlightColor: Union[List[float,\n float, float], bool]=None, isObscured: bool=True, manage: bool=True,\n noBackground: bool=True, node: name=None, numberOfPopupMenus: bool=True,\n parent: Union[AnyStr, bool]=\"\", popupMenuArray: bool=True,\n preventOverride: bool=True, selectedAlphaControl: AnyStr=\"\",\n selectedBlendModeControl: AnyStr=\"\", selectedColorControl: AnyStr=\"\",\n selectedIsVisibleControl: AnyStr=\"\", statusBarMessage: AnyStr=\"\",\n useTemplate: AnyStr=\"\", visible: bool=True, visibleChangeCommand:\n Union[Script, bool]=None, width: Union[int, bool]=0, q=True, query=True,\n e=True, edit=True, **kwargs)->Union[AnyStr, Any]:\n pass"],"string":"[\n \"def run(layers):\",\n \"def build_layers(self):\\n raise NotImplementedError\",\n \"def layers(self): # -> LayerView:\\n ...\",\n \"def UpdateLayers(self):\\n pass\",\n \"def layers(self, layers):\\n self._layers = layers\\n prev = None\\n for layer in layers:\\n if not layer.inputs and prev is not None:\\n layer.inputs = [prev]\\n prev = layer\",\n \"def layers(self, layers):\\n\\n self._layers = layers\",\n \"def __init__(self, layers):\\n\\n\\t\\tself.layers = layers\",\n \"def handle_layers(context, model, toplayer, layerids, materials, update, import_hidden=False):\\n #setup main container to hold all layer collections\\n layer_col_id=\\\"Layers\\\"\\n if not layer_col_id in context.blend_data.collections:\\n layer_col = context.blend_data.collections.new(name=layer_col_id)\\n try:\\n toplayer.children.link(layer_col)\\n except Exception:\\n pass\\n else:\\n #If \\\"Layers\\\" collection is in place, we assume the plugin had imported 3dm before\\n layer_col = context.blend_data.collections[layer_col_id]\\n\\n # build lookup table for LayerTable index\\n # from GUID, create collection for each\\n # layer\\n for lid, l in enumerate(model.Layers):\\n if not l.Visible and not import_hidden:\\n continue\\n lcol = utils.get_iddata(context.blend_data.collections, l.Id, l.Name, None)\\n layerids[str(l.Id)] = (lid, lcol)\\n utils.tag_data(layerids[str(l.Id)][1], l.Id, l.Name)\\n '''\\n matname = l.Name + \\\"+\\\" + str(l.Id)\\n if matname not in materials:\\n laymat = utils.get_iddata(context.blend_data.materials, l.Id, l.Name, None)\\n if update:\\n\\t laymat.use_nodes = True\\n\\t r, g, b, _ = l.Color\\n\\t principled = PrincipledBSDFWrapper(laymat, is_readonly=False)\\n\\t principled.base_color = (r/255.0, g/255.0, b/255.0)\\n materials[matname] = laymat\\n '''\\n # second pass so we can link layers to each other\\n for l in model.Layers:\\n # link up layers to their parent layers\\n if str(l.ParentLayerId) in layerids:\\n parentlayer = layerids[str(l.ParentLayerId)][1]\\n try:\\n parentlayer.children.link(layerids[str(l.Id)][1])\\n except Exception:\\n pass\\n # or to the top collection if no parent layer was found\\n else:\\n try:\\n layer_col.children.link(layerids[str(l.Id)][1])\\n except Exception:\\n pass\",\n \"def consume_layer(self, reports):\\n layer_list = []\\n layer_count = 1\\n for report in reports:\\n layer = create_image_layer(report)\\n layer.layer_index = layer_count\\n layer_list.append(layer)\\n layer_count += 1\\n return layer_list\",\n \"def __call__(cls, *args, **kwargs):\\n layer = super(LayerAspect, cls).__call__(*args, **kwargs)\\n\\n if Job.Current:\\n Job.Current.addLayer(layer)\\n \\n layer.afterInit()\\n return layer\",\n \"def layers(self, layers):\\n self._layers = layers\\n self.thetas = []\\n prev = None\\n for layer in layers:\\n if not layer.inputs and prev is not None:\\n layer.inputs = [prev]\\n prev = layer\\n self.thetas.extend(layer.thetas())\",\n \"def layers(self, x):\\n raise NotImplementedError\",\n \"def setup_layer_structure(self):\\n self.page_rank_convolution_1 = self.layer(self.feature_number, self.args.layers[0], self.args.iterations, self.args.alpha)\\n self.page_rank_convolution_2 = self.layer(self.args.layers[0], self.args.layers[1], self.args.iterations, self.args.alpha)\\n self.page_rank_convolution_3 = self.layer(self.args.layers[1], self.class_number, self.args.iterations, self.args.alpha)\",\n \"def build_models(config):\\n inputs = Input(shape=(config['patch_height'], config['patch_width'], config['depth'], config['channel']),\\n name='patchimg')\\n\\n kernelinitfun = keras.initializers.RandomNormal(mean=0.0, stddev=0.1, seed=None)\\n activationfun = 'relu'\\n # kernelinitfun = 'glorot_normal'\\n\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_1')(inputs)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv1_2')(x)\\n # x = Dropout(0.3)(x)\\n x = BatchNormalization(name='bn1')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp1', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_1')(x)\\n x = Conv3D(32, (3, 3, 3), padding='same', kernel_initializer=kernelinitfun, name='conv2_2')(x)\\n # x = Dropout(0.2)(x)\\n x = BatchNormalization(name='bn2')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp2', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_1')(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv3_2')(x)\\n # x = Dropout(0.5)(x)\\n x = BatchNormalization(name='bn3')(x)\\n x = Activation(activationfun)(x)\\n x = MaxPooling3D(name='mp3', strides=(2, 2, 1))(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_1')(x)\\n x = Conv3D(32, (3, 3, 3), strides=(1, 1, 1), padding='same', kernel_initializer=kernelinitfun, name='conv4_2')(x)\\n # x = Dropout(0.5)(x)\\n x = BatchNormalization(name='bn4')(x)\\n x = Activation(activationfun)(x)\\n\\n x4 = Flatten(name='aux_fx')(x)\\n\\n source_classifier = Dropout(0.5)(x4)\\n source_classifier = Dense(512, activation='softmax', name=\\\"mo1\\\")(source_classifier)\\n source_classifier = Dropout(0.5)(source_classifier)\\n source_classifier = Dense(128, activation='softmax', name=\\\"mo2\\\")(source_classifier)\\n # source_classifier = Dropout(0.3)(source_classifier)\\n source_classifier = Dense(1, name=\\\"mo\\\")(source_classifier)\\n\\n domain_classifier = Dense(32, activation='linear', name=\\\"do4\\\")(x4)\\n domain_classifier = BatchNormalization(name=\\\"do5\\\")(domain_classifier)\\n domain_classifier = Activation(\\\"elu\\\", name=\\\"do6\\\")(domain_classifier)\\n domain_classifier = Dropout(0.5)(domain_classifier)\\n\\n domain_classifier = Dense(2, activation='softmax', name=\\\"do\\\")(domain_classifier)\\n\\n adamop = keras.optimizers.Adam(learning_rate=1e-5, beta_1=0.9, beta_2=0.999, amsgrad=False)\\n comb_model = Model(inputs=inputs, outputs=[source_classifier, domain_classifier])\\n comb_model.compile(optimizer=adamop,\\n loss={'mo': 'mae', 'do': 'categorical_crossentropy'},\\n loss_weights={'mo': 1, 'do': 2}, metrics=['accuracy'], )\\n\\n source_classification_model = Model(inputs=inputs, outputs=[source_classifier])\\n source_classification_model.compile(optimizer=adamop,\\n loss={'mo': 'mae'}, metrics=['accuracy'], )\\n\\n domain_classification_model = Model(inputs=inputs, outputs=[domain_classifier])\\n domain_classification_model.compile(optimizer=adamop,\\n loss={'do': 'categorical_crossentropy'}, metrics=['accuracy'])\\n\\n embeddings_model = Model(inputs=inputs, outputs=[x4])\\n embeddings_model.compile(optimizer=adamop, loss='categorical_crossentropy', metrics=['accuracy'])\\n\\n return comb_model, source_classification_model, domain_classification_model, embeddings_model\",\n \"def __init__(self, kernel_size, filters, stage, block):\\n super().__init__(name='identity' + str(stage) + block)\\n filters1, filters2, filters3 = filters\\n if K.image_data_format() == 'channels_last':\\n bn_axis = 3\\n else:\\n bn_axis = 1\\n\\n conv_name_base = 'res' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n self.conv1 = layers.Conv2D(\\n filters1, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2a')\\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\\n self.act1 = layers.Activation('relu')\\n\\n self.conv2 = layers.Conv2D(\\n filters2,\\n kernel_size,\\n padding='same',\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2b')\\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\\n self.act2 = layers.Activation('relu')\\n\\n self.conv3 = layers.Conv2D(\\n filters3, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2c')\\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\\n\\n self.add = layers.Add()\\n self.act = layers.Activation('relu')\",\n \"def __init__(self, parent, layer):\\n pass\",\n \"def setup_layers(self):\\n if self.args.model == \\\"exact\\\":\\n self.layer = PPNPLayer\\n else:\\n self.layer = APPNPLayer\\n self.setup_layer_structure()\",\n \"def set_hooks(self):\\n for layer_idx, layer in enumerate(self.vgg_features):\\n if layer_idx in self.layers_to_watch:\\n layer.register_forward_hook(self.layer_watch_hooks(layer_idx))\",\n \"def __init__(self, kernel_size, filters, stage, block, strides=(2, 2)):\\n super().__init__(name='conv_block' + str(stage) + block)\\n filters1, filters2, filters3 = filters\\n if K.image_data_format() == 'channels_last':\\n bn_axis = 3\\n else:\\n bn_axis = 1\\n conv_name_base = 'res' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n self.conv1 = layers.Conv2D(\\n filters1, (1, 1),\\n strides=strides,\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2a')\\n self.bn1 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')\\n self.act1 = layers.Activation('relu')\\n\\n self.conv2 = layers.Conv2D(\\n filters2,\\n kernel_size,\\n padding='same',\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2b')\\n self.bn2 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')\\n self.act2 = layers.Activation('relu')\\n\\n self.conv3 = layers.Conv2D(\\n filters3, (1, 1),\\n kernel_initializer='he_normal',\\n name=conv_name_base + '2c')\\n self.bn3 = layers.BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')\\n\\n self.shortcut_conv = layers.Conv2D(\\n filters3, (1, 1),\\n strides=strides,\\n kernel_initializer='he_normal',\\n name=conv_name_base + '1')\\n self.shortcut_bn = layers.BatchNormalization(\\n axis=bn_axis, name=bn_name_base + '1')\\n\\n self.add = layers.Add()\\n self.act = layers.Activation('relu')\",\n \"def _build(self):\\n with tf.variable_scope (self.name + '_architecutre') as scope:\\n images_square = unflatten_layer ( self.images )\\n visualize_images(images_square)\\n\\n # Conv Layer 1\\n conv1_out, params = conv_2d_layer ( input = images_square,\\n neurons = CONV_1_N,\\n filter_size = CONV_1_FILT,\\n name = 'enc_conv_1',\\n visualize = True )\\n process_params(params, name = self.name)\\n e1_params = params\\n pool1_out = max_pool_2d_layer ( input = conv1_out, name = 'enc_pool_1')\\n # lrn1_out = local_response_normalization_layer (pool1_out, name = 'lrn_1' )\\n\\n # Conv Layer 2\\n conv2_out, params = conv_2d_layer ( input = pool1_out,\\n neurons = CONV_2_N,\\n filter_size = CONV_2_FILT,\\n name = 'enc_conv_2' )\\n process_params(params, name = self.name)\\n e2_params = params\\n pool2_out = max_pool_2d_layer ( input = conv2_out, name = 'enc_pool_2')\\n # lrn2_out = local_response_normalization_layer (pool2_out, name = 'lrn_2' )\\n\\n flattened = flatten_layer(pool2_out)\\n\\n # Dropout Layer 1 \\n flattened_dropout = dropout_layer ( input = flattened,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_1') \\n\\n # Dot Product Layer 1\\n fc1_out, params = dot_product_layer ( input = flattened_dropout,\\n neurons = HIDDEN_1,\\n name = 'enc_dot_1')\\n process_params(params, name = self.name)\\n e3_params = params \\n\\n # Dropout Layer 2 \\n fc1_out_dropout = dropout_layer ( input = fc1_out,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_2')\\n # Dot Product Layer 2\\n fc2_out, params = dot_product_layer ( input = fc1_out_dropout, \\n neurons = HIDDEN_2,\\n name = 'enc_dot_2')\\n process_params(params, name = self.name)\\n e4_params = params \\n\\n # Dropout Layer 3 \\n fc2_out_dropout = dropout_layer ( input = fc2_out,\\n prob = self.dropout_prob,\\n name = 'enc_dropout_3')\\n \\n # Dot Product Layer 2\\n self.codeword, params = dot_product_layer ( input = fc2_out_dropout, \\n neurons = CODEWORD_LENGTH,\\n activation = CODE_ACTIVATION,\\n name = 'enc_dot_2')\\n process_params(params, name = self.name)\\n process_codeword_normalization_regularizer(self.codeword, \\n coeff = AUTOENCODER_CODEWORD_COEFF,\\n name = self.name)\\n e5_params = params \\n # tf.summary.histogram('codewords', self.codeword)\\n # self.hash = threshold_layer ( input = self.codeword,\\n # name = 'hash')\\n # process_hash_regularizer(self.codeword, coeff = AUTOENCODER_HASH_COEFF,\\n # name = self.name)\\n\\n # Decoder ... \\n decoder_1_out, params = dot_product_layer ( input = self.codeword, \\n neurons = HIDDEN_2,\\n params = [tf.transpose(e5_params[0]), None],\\n name = 'decoder_dot_1')\\n d1_params = params\\n process_params([params[1]], name = self.name)\\n \\n dec_1_out_dropout = dropout_layer ( input = decoder_1_out,\\n prob = self.dropout_prob,\\n name = 'dec_dropout_1')\\n\\n decoder_2_out, params = dot_product_layer ( input = dec_1_out_dropout, \\n neurons = HIDDEN_1,\\n params = [tf.transpose(e4_params[0]), None],\\n name = 'decoder_dot_2')\\n d2_params = params\\n process_params([params[1]], name = self.name)\\n \\n # dropout 2\\n dec_2_out_dropout = dropout_layer ( input = decoder_2_out,\\n prob = self.dropout_prob,\\n name = 'dec_dropout_2')\\n\\n decoder_3_out, params = dot_product_layer ( input = dec_2_out_dropout, \\n neurons = 1250,\\n params = [tf.transpose(e3_params[0]), None],\\n name = 'decoder_dot_3')\\n d3_params = params\\n process_params([params[1]], name = self.name)\\n\\n # DeConv Layer 1\\n # The output shapes need to be changed according to architecture.\\n\\n dec_3_square = unflatten_layer ( decoder_3_out, channels = CONV_2_N )\\n upsample_1 = upsampling_layer (dec_3_square, size = (10,10), name = 'dec_upsampling_1')\\n\\n deconv1_out, params = deconv_2d_layer ( input = upsample_1,\\n neurons = CONV_1_N,\\n filter_size = CONV_2_FILT,\\n output_shape = (12,12),\\n # n_outs = MINI_BATCH_SIZE,\\n stride = (1,1,1,1), \\n params = [e2_params[0], None], \\n name = 'dec_deconv_1' )\\n\\n process_params([params[1]], name = self.name)\\n d4_params = params\\n\\n # DeConv Layer 2\\n upsample_2 = upsampling_layer (deconv1_out, size = (24,24), name = 'dec_upsampling_2')\\n decoded_images_square, params = deconv_2d_layer ( input = upsample_2,\\n neurons = 1,\\n filter_size = CONV_1_FILT,\\n stride = (1,1,1,1),\\n output_shape = (28,28),\\n # n_outs = MINI_BATCH_SIZE, \\n params = [e1_params[0], None], \\n activation = 'tanh', \\n name = 'dec_deconv_2' )\\n \\n process_params([params[1]], name = self.name)\\n d5_params = params \\n \\n self.decoded = flatten_layer (decoded_images_square, in_shp = [-1, 28, 28, 1])\\n visualize_images(decoded_images_square, name = 'decoded')\\n # This is because transpose don't initialize.\\n self.params = [ [e5_params[0], d1_params[1] ],\\n [e4_params[0], d2_params[1] ],\\n [e3_params[0], d3_params[1] ],\\n [e2_params[0], d4_params[1] ],\\n [e1_params[0], d5_params[1] ] ]\\n\\n with tf.variable_scope (self.name + '_objectives') as scope: \\n with tf.variable_scope( self.name + '_decoder_error') as scope:\\n reconstruction_error = rmse(self.images, self.decoded) \\n tf.add_to_collection( self.name + '_objectives', reconstruction_error ) \\n tf.summary.scalar('reconstruction_error', reconstruction_error)\\n\\n self._cook_optimizer( \\n lr = AUTOENCODER_LR, \\n optimizer = AUTOENCODER_OPTIMIZER,\\n l1_coeff = AUTOENCODER_L1_COEFF,\\n l2_coeff = AUTOENCODER_WEIGHT_DECAY_COEFF)\",\n \"def setup_layers(self, params):\\n\\n #\\n # Input layer\\n #\\n\\n # Input and encoding layers\\n layer_stack = [\\n tf.placeholder(\\n tf.int32,\\n name=\\\"input\\\",\\n shape=[None, params.in_size]\\n )\\n ]\\n\\n in_layer = layer_stack[0]\\n\\n good_chars = getattr(params, 'good_chars', tops.TextProcessor.DEFAULT_CHARS)\\n character_depth = len(good_chars) + 2\\n\\n layer_stack.append(\\n self.make_onehot_encode_layer(\\n layer_stack[-1],\\n character_depth\\n )\\n )\\n\\n # Add conv layers\\n for idx, num_nodes in enumerate(params.conv_nodes):\\n layer_stack.append(\\n self.make_conv_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'conv_{}'.format(idx),\\n params.conv_params\\n )\\n )\\n\\n # Flatten convolutional layers\\n layer_stack.append(\\n tf.contrib.layers.flatten(\\n layer_stack[-1]\\n )\\n )\\n\\n # Add dense layers\\n for idx, num_nodes in enumerate(params.dense_nodes):\\n layer_stack.append(\\n self.make_dense_layer(\\n layer_stack[-1],\\n num_nodes,\\n 'dense_{}'.format(idx),\\n params.dense_params\\n )\\n )\\n\\n # Output layer is broken into two pieces. The pre/post the application\\n # of the activation function. That's because the loss functions\\n # for cross-entropy rely on the pre-activation scores in preact\\n\\n preact_out_layer, out_layer = self.make_dense_output_layer(\\n layer_stack[-1],\\n params.out_size,\\n params.output_params\\n )\\n\\n target_layer = tf.placeholder(\\n \\\"float\\\",\\n name=\\\"target\\\",\\n shape=[None, params.out_size]\\n )\\n\\n if params.output_params.activation in ['sigmoid']:\\n loss = tops.loss_sigmoid_ce(preact_out_layer, target_layer)\\n else:\\n loss = tops.loss_softmax_ce(preact_out_layer, target_layer)\\n\\n return in_layer, out_layer, target_layer, loss\",\n \"def __call__(self, layers, inputs, name=None):\\r\\n def call_layer(l, x):\\r\\n if hasattr(l, \\\"call_by_computation_context\\\"):\\r\\n return l.call_by_computation_context(x, *self.args, **self.kwargs)\\r\\n else:\\r\\n return l(x, *self.args, **self.kwargs)\\r\\n\\r\\n if not isinstance(layers, (list, tuple)):\\r\\n # Single layer, just call\\r\\n return call_layer(layers, inputs)\\r\\n else:\\r\\n # Flatten the layers\\r\\n layers = tf.nest.flatten(layers)\\r\\n outputs = inputs\\r\\n\\r\\n # Add a name scope for wrapping\\r\\n if name is not None:\\r\\n scope_name = name\\r\\n else:\\r\\n scope_name = commonprefix([layer.name for layer in layers]).strip(\\\"-/\\\\\\\\?_ \\\\t\\\\n\\\") or \\\"ComputationContext\\\"\\r\\n with tf.name_scope(scope_name):\\r\\n for l in layers:\\r\\n outputs = call_layer(l, outputs)\\r\\n\\r\\n return outputs\",\n \"def generate_legacy_layers(self, images_map, content_retriever):\\n pass\",\n \"def generate_legacy_layers(self, images_map, content_retriever):\\n pass\",\n \"def __init__(self, layer_nest):\\n self._layer_nest = layer_nest\\n super().__init__()\",\n \"def preview_layerset(ls, size = 100, spacing = 100):\\n D = Device()\\n scale = size/100\\n num_layers = len(ls._layers)\\n matrix_size = int(np.ceil(np.sqrt(num_layers)))\\n sorted_layers = sorted(ls._layers.values(),\\n key = lambda x: (x.gds_layer, x.gds_datatype))\\n for n, layer in enumerate(sorted_layers):\\n R = rectangle(size = (100*scale, 100*scale), layer = layer)\\n T = text(text = '%s\\\\n%s / %s'\\\\\\n % (layer.name, layer.gds_layer, layer.gds_datatype),\\n size = 20*scale,\\n justify = 'center',\\n layer = layer)\\n\\n T.move((50*scale, -20*scale))\\n xloc = n % matrix_size\\n yloc = int(n // matrix_size)\\n D.add_ref(R)\\\\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\\n D.add_ref(T)\\\\\\n .movex((100+spacing) * xloc*scale).movey(-(100+spacing) * yloc*scale)\\n return D\",\n \"def add_layers(self, layers):\\n\\n existing_layers = self.layers\\n assert len(existing_layers) > 0\\n for layer in layers:\\n assert layer.get_mlp() is None\\n layer.set_mlp(self)\\n layer.set_input_space(existing_layers[-1].get_output_space())\\n existing_layers.append(layer)\\n assert layer.layer_name not in self.layer_names\\n self.layer_names.add(layer.layer_name)\",\n \"def build(self):\\n\\n layers = GiraffeLayer.get_all_structural()\\n \\n for layer in layers:\\n\\n self.add_objects_from_layer(layer)\\n\\n return self\",\n \"def init_nnpatches(self):\\n # nnpatches = self._generate_nnpatches()\\n\\n # Register `nnpatches` to this `nnmodel`\\n # Refer: NNFramework._init_model_params(...)\\n\\n # # Register this `nnmodel` to `nnpatch`\\n # for nnpatch in nnpatches:\\n # nnpatch.add_nnmodels(self)\",\n \"def set_up_rlayer(self, rlname, rlname_other=None, include_layers=None,\\n exclude_layers=None, mask_layers=None):\\n scene = self.set_as_active()\\n layer_numbers = constants.layer_numbers\\n w_var.rlname = rlname\\n\\n if include_layers is None:\\n include_layers = w_var.layer_numbers_all_used\\n\\n if exclude_layers is None:\\n exclude_layers = []\\n\\n if mask_layers is None:\\n mask_layers = []\\n\\n if w_var.cb_clear_rlayers:\\n for layer in scene.render.layers[:-1]:\\n scene.render.layers.remove(layer)\\n\\n scene.render.layers.active.name = rlname\\n scene.render.layers.active.use = True\\n\\n new_rlayer = scene.render.layers.active\\n\\n # if not clearing render layers: creates new one\\n else:\\n new_rlayer = scene.render.layers.new(rlname)\\n scene.render.layers.active = new_rlayer\\n\\n # there needs to be two render layers in the same scene for freestyle compositing\\n if w_var.cb_composited:\\n w_var.rlname_other = rlname_other\\n other_rlayer = scene.render.layers.new(rlname_other)\\n other_rlayer.layers[19] = True\\n scene.render.layers[rlname_other].layers_zmask = (False,) * 20\\n\\n if w_var.cb_ao:\\n scene.render.layers[rlname].use_pass_ambient_occlusion = True\\n\\n if w_var.cb_composited:\\n scene.render.layers[rlname_other].use_pass_ambient_occlusion = True\\n\\n # because I can't deactivate a layer if it is the only active one\\n new_rlayer.layers[19] = True\\n \\n scene.render.layers[rlname].layers_exclude = (False,) * 20\\n scene.render.layers[rlname].layers_zmask = (False,) * 20\\n\\n for i in layer_numbers:\\n if w_var.cb_composited:\\n if i in w_var.layer_numbers_affected:\\n scene.render.layers[rlname].layers[i] = True\\n scene.render.layers[rlname_other].layers_zmask[i] = True\\n\\n else:\\n scene.render.layers[rlname].layers[i] = False\\n\\n if i in w_var.layer_numbers_other:\\n scene.render.layers[rlname_other].layers[i] = True\\n\\n else:\\n scene.render.layers[rlname_other].layers[i] = False\\n\\n else:\\n if i in include_layers:\\n scene.render.layers[rlname].layers[i] = True\\n\\n else:\\n scene.render.layers[rlname].layers[i] = False\\n\\n if i in mask_layers:\\n scene.render.layers[rlname].layers_zmask[i] = True\\n\\n if i in exclude_layers:\\n scene.render.layers[rlname].layers_exclude[i] = True\",\n \"def draw_layers(self):\\n\\t\\tfor z in xrange(0,16):\\n\\t\\t\\t#create surface for this layer\\n\\t\\t\\tsrf = pygame.Surface((16,128))\\n\\t\\t\\tfor x in xrange(0,16):\\n\\t\\t\\t\\tfor y in xrange(0,128):\\n\\t\\t\\t\\t\\tv = self.data[ self.xyz_to_offset( x,y,z) ]\\n\\t\\t\\t\\t\\tif v != 0:\\n\\t\\t\\t\\t\\t\\tsrf.fill( BLOCKS.get(v, [0,0])[1], \\t(x, 127 -y, 1, 1 ))\\n\\t\\t\\t#save layer to dict for this chunk\\n\\t\\t\\tself.layers[z] = srf\",\n \"def process_layers(size: Sequence[int],\\n layers: Union[Source, Layer, Sequence[Layer]],\\n a: Union[None, np.ndarray] = None,\\n status: Union[None, Queue] = None) -> np.ndarray:\\n # If no destination array was sent, it means we are either at the\\n # start of the layer processing or we are starting the processing\\n # of a layer group. If we are starting processing, then this will\\n # contain the final image data. If we are starting the processing\\n # of a layer group, this allows the layers in the group to only\\n # blend with the other layers in the group before they blend with\\n # the layers outside the group.\\n if a is None:\\n a = np.zeros(size, dtype=float)\\n\\n # If we got a sequence of layers, we process them recursively and\\n # return the result.\\n if isinstance(layers, Sequence):\\n for layer in layers:\\n a = process_layers(size, layer, a, status)\\n return a\\n\\n # If we got a source layer, process it.\\n if isinstance(layers.source, Source):\\n if status is not None:\\n src_name = layers.source.__class__.__name__\\n status.put((sw.MSG, f'Rendering {src_name}...'))\\n kwargs = {\\n 'source': layers.source,\\n 'size': size,\\n 'location': layers.location,\\n 'filters': layers.filters,\\n }\\n b = render_source(**kwargs)\\n if status is not None:\\n src_name = layers.source.__class__.__name__\\n status.put((sw.MSG, f'Rendered {src_name}.'))\\n status.put((sw.PROG,))\\n\\n # Otherwise we got a container layer, process its source and run\\n # any filters that are set on the layer.\\n else:\\n new_size = f.preprocess(size, layers.filters)\\n b = process_layers(new_size, layers.source, None, status)\\n b = f.process(b, layers.filters)\\n b = f.postprocess(b, layers.filters)\\n\\n # There are two possibilities for how the layers should be\\n # blended: masked or unmasked. Masked blends will have a\\n # Source in the mask attribute, which needs to be sent\\n # to the blending operation.\\n if layers.mask is not None:\\n if status is not None:\\n src_name = layers.mask.__class__.__name__\\n status.put((sw.MSG, f'Rendering {src_name}...'))\\n kwargs = {\\n 'source': layers.mask,\\n 'size': size,\\n 'filters': layers.mask_filters,\\n }\\n mask = render_source(**kwargs)\\n if status is not None:\\n src_name = layers.mask.__class__.__name__\\n status.put((sw.MSG, f'Rendered {src_name}.'))\\n status.put((sw.PROG,))\\n a, b, mask = _normalize_color_space(a, b, mask)\\n return layers.blend(a, b, layers.blend_amount, mask)\\n\\n # Unmasked layers have no mask, so Blend the image data and return.\\n a, b = _normalize_color_space(a, b)\\n return layers.blend(a, b, layers.blend_amount)\",\n \"def make_layers(self):\\n w, h = self.image.get_size()\\n shrink = pg.transform.smoothscale(self.image, (w//2, h//2))\\n self.mid_image = tools.tile_surface((w,h), shrink, True)\\n shrink = pg.transform.smoothscale(self.image, (w//4, h//4))\\n self.base = tools.tile_surface(prepare.SCREEN_SIZE, shrink, True)\",\n \"def __initAvailableLayerTypes(self):\\n from backend.caffe.path_loader import PathLoader\\n caffe = PathLoader().importCaffe()\\n layerNameMainParts = list(caffe.layer_type_list())\\n\\n res = {}\\n paramsPerLayerType = {}\\n\\n # calculate common parameters of all layer types\\n # by removing all which will be used for one specific layer type only\\n # also keep in mind which ones have been removed to readd them to specific layers\\n commonParams = self._availableParameterGroupDescriptors[\\\"LayerParameter\\\"].parameter() #use .parameter() on purpose\\n layerSpecificParameters = set()\\n for nameMainPart in layerNameMainParts:\\n specificParamsName = [nameMainPart + \\\"Parameter\\\"]\\n if moreLayerNameParameter.has_key(nameMainPart):\\n specificParamsName.append( moreLayerNameParameter[nameMainPart])\\n paramsPerLayerType[nameMainPart] = {}\\n for key, value in commonParams.items():\\n if value.isParameterGroup() and value.parameterName() in specificParamsName:\\n paramsPerLayerType[nameMainPart][key] = value\\n layerSpecificParameters.add(key)\\n\\n\\n # special case: shared params for loss layers\\n key = \\\"loss_param\\\"\\n value = commonParams[key]\\n del commonParams[key]\\n for nameMainPart in layerNameMainParts:\\n if LayerType.getCategoryByName(nameMainPart) == LayerType.CATEGORY_LOSS:\\n paramsPerLayerType[nameMainPart][key] = value\\n\\n # TODO is there a special case for the TransformationParameter?\\n\\n # create each layer type after one another\\n for nameMainPart in layerNameMainParts:\\n\\n # add common params to the specific ones\\n layerTypeParam = paramsPerLayerType[nameMainPart].keys()\\n paramsPerLayerType[nameMainPart].update(commonParams)\\n\\n irrelevant = layerSpecificParameters.difference(layerTypeParam)\\n res[nameMainPart] = LayerType(nameMainPart, paramsPerLayerType[nameMainPart], layerTypeParam, irrelevant)\\n\\n self._commonParams = commonParams\\n self._availableLayerTypes = res\",\n \"def EmbeddingLayers_pooling(pretrained=False, progress=True, **kwargs):\\n kwargs['groups'] = 32\\n kwargs['width_per_group'] = 4\\n # return _resnet('resnet18', BasicBlock, [2, 2, 2, 2], pretrained, progress,\\n # **kwargs) \\n return _resnet('resnext50_32x4d', Bottleneck, [3, 4, 6, 3],\\n pretrained, progress, **kwargs)\",\n \"def __init__(self,layer_def,input,input_shape,rs,clone_from=None):\\n batch_size = int(layer_def.find(\\\"batchsize\\\").text)\\n image_size = int(layer_def.find(\\\"imagesize\\\").text)\\n image_channels = int(layer_def.find(\\\"imagechannels\\\").text)\\n self.layer_name = layer_def.attrib[\\\"name\\\"]\\n self.init(input, input_shape, batch_size,image_size,image_channels)\",\n \"def feed(self, *args):\\n\\t\\tprint(\\\"args: \\\", args)\\n\\t\\tassert len(args) != 0\\n\\t\\tself.terminals = []\\n\\t\\tfor fed_layer in args:#\\n\\t\\t\\ttry:\\n\\t\\t\\t\\tis_str = isinstance(fed_layer, basestring)\\n\\t\\t\\texcept NameError:\\n\\t\\t\\t\\tis_str = isinstance(fed_layer, str)\\n\\t\\t\\tif is_str:\\n\\t\\t\\t\\ttry:\\n\\t\\t\\t\\t\\tfed_layer = self.layers[fed_layer]\\n\\t\\t\\t\\t\\tprint(fed_layer)\\n\\t\\t\\t\\texcept KeyError:\\n\\t\\t\\t\\t\\traise KeyError('Unknown layer name fed: %s' % fed_layer)\\n\\t\\t\\tself.terminals.append(fed_layer)\\n\\t\\treturn self\",\n \"def append_layer(self, *args, **kwargs) :\\n \\n self.insert_layer(len(self._layers), *args, **kwargs)\",\n \"def get_data_from_layers(vk4_container, layers, step, is_image=False):\\n log.debug(\\\"Entering get_data_from_layers()\\\")\\n holder = []\\n for x in range(0, len(layers), step):\\n lay = layers[x:x + step]\\n log.debug(\\\"In get_data_from_layers()\\\\n\\\\tCurrent layer to append: %s\\\" % lay)\\n if lay not in layer_dict.keys():\\n log.error(\\\"In get_data_from_layers()\\\\n\\\\tLayer {}: {} from param \\\" \\\\\\n \\\"layers: {} is not a valid layer. See documentation.\\\"\\n .format(x, lay, layers))\\n holder.append(vk4_container.get_single_color_values(layer_dict[lay][0], layer_dict[lay][1]))\\n\\n if is_image:\\n log.debug(\\\"Exiting get_data_from_layers() where is_image is {}\\\".format(is_image))\\n return create_array_from_rgb_layers(vk4_container, holder)\\n else:\\n log.debug(\\\"Exiting get_data_from_layers() where is_image is {}\\\".format(is_image))\\n return create_composite_rgb_values(vk4_container, holder)\\n # return create_separate_rgb_values(vk4_container, holder)\",\n \"def initialize_layers(self, layers_config: dict, inputs=None):\\n layers_config = layers_config.copy()\\n input_lyrs = []\\n initiated_layers = OrderedDict()\\n wrp_layer = None # indicator for wrapper layers\\n first_layer = True\\n\\n for lyr, lyr_args in layers_config.items():\\n\\n lyr_config, lyr_inputs, named_outs, call_args = self.deconstruct_lyr_args(lyr, lyr_args)\\n\\n lyr_name, args, lyr_config, activation = self.check_lyr_config(lyr, lyr_config)\\n\\n if K.BACKEND == 'pytorch':\\n\\n if first_layer:\\n first_layer = False\\n\\n if callable(lyr_config):\\n lyr_initiated = lyr_config\\n else:\\n lyr_initiated = TORCH_LAYERS[lyr_name](**lyr_config)\\n setattr(self, lyr, lyr_initiated)\\n initiated_layers[lyr] = {\\\"layer\\\": lyr_initiated, \\\"named_outs\\\": named_outs, 'call_args': call_args,\\n 'inputs': lyr_inputs}\\n\\n else:\\n # may be user has defined layers without input layer, in this case add Input layer as first layer\\n if first_layer:\\n if inputs is not None: # This method was called by providing it inputs.\\n assert isinstance(inputs, tf.Tensor)\\n # since inputs have been defined, all the layers that will be added will be next to first layer\\n first_layer = False\\n layer_outputs = inputs\\n initiated_layers[layer_outputs.name] = {'layer': layer_outputs, 'tf_name': lyr_name}\\n\\n elif lyr_name != \\\"Input\\\":\\n if 'input_shape' in lyr_config: # input_shape is given in the first layer so make input layer\\n initialized_layer = LAYERS[\\\"Input\\\"](shape=lyr_config['input_shape'])\\n else:\\n # for simple dense layer based models, lookback will not be used\\n def_shape = (self.num_ins,) if self.lookback == 1 else (self.lookback, self.num_ins)\\n initialized_layer = LAYERS[\\\"Input\\\"](shape=def_shape)\\n\\n # first layer is built so next iterations will not be for first layer\\n first_layer = False\\n # put the first layer in memory to be used for model compilation\\n # add th layer which the user had specified as first layer\\n initiated_layers[initialized_layer.name] = {'layer': initialized_layer,\\n 'tf_name': lyr_name}\\n input_lyrs.append(initialized_layer)\\n\\n # The inputs to the layer have not been specified, so either it is an Input layer\\n if lyr_inputs is None:\\n # or it uses the previous outputs as inputs\\n if lyr_name == \\\"Input\\\":\\n # it is an Input layer, hence should not be called\\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'tf_name': lyr_name}\\n input_lyrs.append(initialized_layer)\\n else:\\n # it is executable and uses previous outputs as inputs\\n if lyr_name in ACTIVATION_LAYERS:\\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\\n wrp_layer = LAYERS[lyr_name]\\n # because wrapper layer name is property\\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\\n 'tf_name': lyr_name}\\n continue\\n elif \\\"LAMBDA\\\" in lyr_name.upper():\\n # lyr_config is serialized lambda layer, which needs to be deserialized\\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\\n # layers_config['lambda']['config'] still contails lambda, so we need to replace the python\\n # object (lambda) with the serialized version (lyr_config) so that it can be saved as json file.\\n layers_config[lyr]['config'] = lyr_config\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n else:\\n if wrp_layer is not None:\\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n wrp_layer = None\\n else:\\n if lyr_name == \\\"TemporalFusionTransformer\\\":\\n lyr_config['return_attention_components'] = True\\n initialized_layer = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n\\n else: # The inputs to this layer have been specified so they must exist in lyr_cache.\\n # it is an executable\\n if lyr_name in ACTIVATION_LAYERS:\\n\\n layer_outputs = ACTIVATION_LAYERS[lyr_name](name=lyr_config['name'])\\n initiated_layers[lyr_config['name']] = {'layer': layer_outputs,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n elif lyr_name in ['TimeDistributed', 'Bidirectional']:\\n wrp_layer = LAYERS[lyr_name]\\n # because wrapper layer name is property\\n initiated_layers[lyr_config['name']] = {'layer': wrp_layer,\\n 'tf_name': lyr_name}\\n continue\\n elif \\\"LAMBDA\\\" in lyr_name.upper():\\n initialized_layer = tf.keras.layers.deserialize(lyr_config)\\n layers_config[lyr]['config'] = lyr_config\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n else:\\n if wrp_layer is not None:\\n initialized_layer = wrp_layer(LAYERS[lyr_name](*args, **lyr_config))\\n initiated_layers[lyr_config['name']] = {'layer': initialized_layer,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n wrp_layer = None\\n else:\\n layer_initialized = LAYERS[lyr_name](*args, **lyr_config)\\n initiated_layers[lyr_config['name']] = {'layer': layer_initialized,\\n 'named_outs': named_outs,\\n 'call_args': call_args,\\n 'inputs': lyr_inputs,\\n 'tf_name': lyr_name}\\n\\n if activation is not None: # put the string back to dictionary to be saved in config file\\n lyr_config['activation'] = activation\\n\\n first_layer = False\\n\\n self.jsonize_lyr_config(lyr_config)\\n\\n # inputs = [] todo, indentify input layers\\n # for k,v in lyr_cache.items():\\n # since the model is not build yet and we have access to only output tensors of each list, this is probably\\n # # the only way to know that how many `Input` layers were encountered during the run of this method. Each\\n # tensor (except TimeDistributed) has .op.inputs attribute, which is empty if a tensor represents output of Input layer.\\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\\n # if k.upper() != \\\"TIMEDISTRIBUTED\\\" and hasattr(v, 'op'):\\n # if hasattr(v.op, 'inputs'):\\n # _ins = v.op.inputs\\n # if len(_ins) == 0:\\n # inputs.append(v)\\n # else: # not sure if this is the proper way of checking if a layer receives an input or not!\\n # if hasattr(v, '_keras_mask'):\\n # inputs.append(v)\\n\\n setattr(self, 'initiated_layers', initiated_layers)\\n setattr(self, 'input_lyrs', input_lyrs)\\n\\n\\n # todo,\\n # # for case when {Input -> Dense, Input_1}, this method wrongly makes Input_1 as output so in such case use\\n # # {Input_1, Input -> Dense }, thus it makes Dense as output and first 2 as inputs, so throwing warning\\n # if int(''.join(tf.__version__.split('.')[0:2]).ljust(3, '0')) < 240:\\n # if len(layer_outputs.op.inputs) < 1:\\n # print(\\\"Warning: the output is of Input tensor class type\\\")\\n # else:\\n # if 'op' not in dir(layer_outputs): # layer_outputs does not have `op`, which means it has no incoming node\\n # print(\\\"Warning: the output is of Input tensor class type\\\")\\n\\n # outs = None\\n #if BACKEND == 'tensorflow':\\n # outs = self.call(input_lyrs)\\n # setattr(self, 'output_lyrs', outs)\\n # if BACKEND == 'tensorflow':\\n # ## Reinitial\\n # super(Model, self).__init__(\\n # inputs=input_lyrs,\\n # outputs=outs)\\n #MODEL.__init__(self, inputs=inputs, outputs=outs)\\n\\n return input_lyrs # , outs\",\n \"def build(width, height, depth, classes, stages, filters, include_top, pooling,\\n reg=1e-3, bnEps=2e-5, bnMom=0.0):\\n inputShape = (height, width, depth)\\n chanDim = -1\\n\\n if K.image_data_format() == \\\"channels_first\\\":\\n inputShape = (depth, height, width)\\n chanDim = 1\\n\\n inputs = Input(shape=inputShape)\\n\\n\\n # block 1 (initial conv block)\\n x = ZeroPadding2D(padding=((3, 3), (3, 3)), name='conv1_pad')(inputs)\\n x = Conv2D(64, (7,7), use_bias=False, strides=(2,2),\\n kernel_initializer=\\\"he_normal\\\", kernel_regularizer=l2(reg))(x)\\n x = BatchNormalization(axis=chanDim, name=\\\"bn_conv1\\\")(x)\\n x = Activation(\\\"relu\\\")(x)\\n x = ZeroPadding2D(padding=((1,1), (1,1)), name=\\\"pool1_pad\\\")(x)\\n x = MaxPooling2D(3, strides=2)(x)\\n\\n for i in range(0, len(stages)):\\n stride = (1,1) if i == 0 else (2,2) # block 2 (projection block) w stride(1,1)\\n\\n print(\\\"Stage {}, Stride={}\\\".format(i, stride))\\n x = SEResNet.residual_module(x, filters[i+1], stride,\\n chanDim=chanDim, red=True, bnEps=bnEps, bnMom=bnMom)\\n for j in range(0, stages[i] + 1): #stacking res block to each depth layer\\n x = SEResNet.residual_module(x, filters[i+1], stride=(1,1),\\n chanDim=chanDim, bnEps=bnEps,\\n bnMom=bnMom)\\n x = BatchNormalization(axis=chanDim, epsilon=bnEps,\\n momentum=bnMom)(x)\\n x = Activation(\\\"relu\\\")(x)\\n\\n if include_top:\\n x = GlobalAveragePooling2D()(x)\\n x = Dense(classes, use_bias=False, kernel_regularizer=l2(reg),\\n activation='softmax')(x)\\n else:\\n if pooling == 'avg':\\n print(\\\"Adding average pool\\\")\\n x = GlobalAveragePooling2D()(x)\\n elif pooling == 'max':\\n x = GlobalMaxPooling2D()(x)\\n\\n model = Model(inputs=inputs, outputs=x, name=\\\"SEResNet\\\")\\n return model\",\n \"def convolutional_block(input_tensor, kernel_size, filters, stage, block, strides = (2, 2)):\\n chanDim = 3\\n eps = 1.1e-5\\n \\n if K.image_data_format() == \\\"channels_first\\\":\\n inputShape = (depth, height, width)\\n chanDim = 1\\n\\n # defining name basis\\n conv_name_base = 'res152' + str(stage) + block + '_branch'\\n bn_name_base = 'bn' + str(stage) + block + '_branch'\\n\\n # Retrieve Filters\\n filters1, filters2, filters3 = filters\\n \\n ##### MAIN PATH #####\\n # First component of main path \\n X = Conv2D(filters = filters1, kernel_size = (1, 1), strides = strides, name = conv_name_base + '2a')(input_tensor)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2a')(X)\\n X = Activation('relu', name = conv_name_base + '2a_relu')(X)\\n \\n # Second component of main path \\n X = ZeroPadding2D(padding = (1, 1), name = conv_name_base + '2b_zeropadding')(X)\\n X = Conv2D(filters = filters2, kernel_size = (kernel_size, kernel_size), name = conv_name_base + '2b')(X)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2b')(X)\\n X = Activation('relu', name = conv_name_base + \\\"2b_relu\\\")(X)\\n \\n # Third component of main path \\n X = Conv2D(filters = filters3, kernel_size = (1, 1), name = conv_name_base + '2c')(X)\\n X = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '2c')(X)\\n \\n ##### SHORTCUT PATH #### \\n X_shortcut = Conv2D(filters = filters3, kernel_size = (1, 1), strides = strides, name = conv_name_base + '1')(input_tensor)\\n X_shortcut = BatchNormalization(epsilon = eps, axis = chanDim, name = bn_name_base + '1')(X_shortcut)\\n \\n # Final step: Add shortcut value to main path, and pass it through a RELU activation\\n X = Add()([X, X_shortcut])\\n X = Activation('relu', name = \\\"res152\\\" + str(stage) + block + \\\"_relu\\\")(X) \\n \\n return X\",\n \"def init_layers(self):\\n\\n # get caching layers activated\\n caching_layers = G3WCachingLayer.objects.all()\\n for caching_layer in caching_layers:\\n self.add_layer(str(caching_layer), caching_layer)\",\n \"def _init_layers(self):\\n cls_branch = []\\n for _ in range(self.num_reg_fcs):\\n cls_branch.append(Linear(self.embed_dims, self.embed_dims))\\n cls_branch.append(nn.LayerNorm(self.embed_dims))\\n cls_branch.append(nn.ReLU(inplace=True))\\n cls_branch.append(Linear(self.embed_dims, self.cls_out_channels))\\n fc_cls = nn.Sequential(*cls_branch)\\n\\n reg_branch = []\\n for _ in range(self.num_reg_fcs):\\n reg_branch.append(Linear(self.embed_dims, self.embed_dims))\\n reg_branch.append(nn.ReLU())\\n reg_branch.append(Linear(self.embed_dims, self.code_size))\\n reg_branch = nn.Sequential(*reg_branch)\\n\\n past_traj_reg_branch = []\\n for _ in range(self.num_reg_fcs):\\n past_traj_reg_branch.append(\\n Linear(self.embed_dims, self.embed_dims))\\n past_traj_reg_branch.append(nn.ReLU())\\n past_traj_reg_branch.append(\\n Linear(self.embed_dims, (self.past_steps + self.fut_steps)*2))\\n past_traj_reg_branch = nn.Sequential(*past_traj_reg_branch)\\n\\n def _get_clones(module, N):\\n return nn.ModuleList([copy.deepcopy(module) for i in range(N)])\\n\\n # last reg_branch is used to generate proposal from\\n # encode feature map when as_two_stage is True.\\n num_pred = (self.transformer.decoder.num_layers + 1) if \\\\\\n self.as_two_stage else self.transformer.decoder.num_layers\\n\\n if self.with_box_refine:\\n self.cls_branches = _get_clones(fc_cls, num_pred)\\n self.reg_branches = _get_clones(reg_branch, num_pred)\\n self.past_traj_reg_branches = _get_clones(\\n past_traj_reg_branch, num_pred)\\n else:\\n self.cls_branches = nn.ModuleList(\\n [fc_cls for _ in range(num_pred)])\\n self.reg_branches = nn.ModuleList(\\n [reg_branch for _ in range(num_pred)])\\n self.past_traj_reg_branches = nn.ModuleList(\\n [past_traj_reg_branch for _ in range(num_pred)])\\n if not self.as_two_stage:\\n self.bev_embedding = nn.Embedding(\\n self.bev_h * self.bev_w, self.embed_dims)\",\n \"def __call__(\\n self, inputs, training, get_all_layers=None, skip_final_dense=False,\\n skip_final_dense_with_pool=False, final_pooling=None):\\n global ENDING_POINTS\\n ENDING_POINTS = []\\n\\n with self._model_variable_scope():\\n inputs = self._preprocess_data(inputs)\\n\\n inputs = conv3d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, \\n kernel_size=self.kernel_size, time_kernel_size=self.time_kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = tf.identity(inputs, 'initial_conv')\\n\\n # We do not include batch normalization or activation functions in V2\\n # for the initial conv1 because the first ResNet unit will perform these\\n # for both the shortcut and non-shortcut paths as part of the first\\n # block's projection. Cf. Appendix of [2].\\n if self.resnet_version == 1:\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n\\n _fp_size = self.first_pool_size\\n _fp_stride = self.first_pool_stride\\n if self.first_pool_size:\\n inputs = tf.layers.max_pooling3d(\\n inputs=inputs, \\n pool_size=[_fp_size, _fp_size, _fp_size],\\n strides=[_fp_stride, _fp_stride, _fp_stride], \\n padding='SAME',\\n data_format=self.data_format)\\n inputs = tf.identity(inputs, 'initial_max_pool')\\n\\n ENDING_POINTS.append(inputs)\\n\\n for i, num_blocks in enumerate(self.block_sizes):\\n num_filters = self.num_filters * (2**i)\\n inputs = block_layer(\\n inputs=inputs, filters=num_filters, bottleneck=self.bottleneck,\\n block_fn=self.block_fn, blocks=num_blocks,\\n strides=self.block_strides[i], training=training,\\n name='block_layer{}'.format(i + 1), data_format=self.data_format)\\n curr_res_name = 'res{}'.format(i + 1)\\n\\n # Only apply the BN and ReLU for model that does pre_activation in each\\n # building/bottleneck block, eg resnet V2.\\n if self.pre_activation:\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n ENDING_POINTS.append(inputs)\\n\\n if skip_final_dense:\\n bs = inputs.get_shape().as_list()[0]\\n # Along the temporal dimension\\n inputs = tf.reduce_mean(inputs, axis=2)\\n if final_pooling is None:\\n inputs = tf.reshape(inputs, [bs, -1])\\n return inputs\\n if final_pooling == 1:\\n axes = [2, 3] if self.data_format == 'channels_first' else [1, 2]\\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\\n return tf.reshape(inputs, [bs, 1 * 1 * self.final_size])\\n \\n # The current top layer has shape\\n # `batch_size x pool_size x pool_size x final_size`.\\n # ResNet does an Average Pooling layer over pool_size,\\n # but that is the same as doing a reduce_mean. We do a reduce_mean\\n # here because it performs better than AveragePooling2D.\\n axes = [2, 3, 4] if self.data_format == 'channels_first' else [1, 2, 3]\\n inputs = tf.reduce_mean(inputs, axes, keepdims=False)\\n inputs = tf.identity(inputs, 'final_reduce_mean')\\n\\n if skip_final_dense_with_pool:\\n return inputs\\n\\n inputs = self._get_final_dense(inputs)\\n if not get_all_layers:\\n return inputs\\n else:\\n return inputs, ENDING_POINTS\",\n \"def make_feature_layers(self, config):\\n raise NotImplementedError\",\n \"def add_layer_data(prj_id, expl_id, exploration):\\n layers = tbl_explsoilboring[\\n (tbl_explsoilboring.lng_KeyProject == prj_id) &\\n (tbl_explsoilboring.txt_KeyExplorationName == expl_id)]\\n\\n # add new column with layer numbers\\n layers['layer'] = [i for i in range(1, len(layers.index) + 1)]\\n\\n # get SPT data\\n spt_data = tbl_explboring[\\n (tbl_explboring.lng_KeyProject == prj_id) &\\n (tbl_explboring.txt_KeyExplorationName == expl_id)]\\n\\n if not spt_data.empty and not layers.empty:\\n spt_depths = spt_data.dbl_KeyDepth\\n layer_i = []\\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\\n for d in spt_depths:\\n if d <= max_d:\\n layr = layers[\\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\\n else:\\n layr = None\\n layer_i.append(layr)\\n spt_data['layer'] = layer_i\\n avg_spt = spt_data[\\n ['layer', 'dbl_FieldBlowCount']].groupby('layer').agg('mean')\\n else:\\n avg_spt = pd.DataFrame()\\n\\n # get lab data\\n lab_data = tbl_expllab[\\n (tbl_expllab.lng_KeyProject == prj_id) &\\n (tbl_expllab.txt_KeyExplorationName == expl_id)]\\n\\n if not lab_data.empty and not layers.empty:\\n lab_depths = lab_data.dbl_KeyDepth\\n layer_i = []\\n max_d = layers['dbl_KeyDepthToBottom'].values[-1]\\n for d in lab_depths:\\n if d <= max_d:\\n layr = layers[\\n layers.dbl_KeyDepthToBottom >= d]['layer'].values[0]\\n else:\\n layr = None\\n layer_i.append(layr)\\n lab_data['layer'] = layer_i\\n avg_lab = lab_data[\\n ['layer', 'dbl_MoistureContent', 'dbl_TotalUnitWeight',\\n 'dbl_LiquidLimit', 'dbl_PlasticityIndex', 'dbl_Cohesion']\\n ].groupby('layer').agg('mean')\\n else:\\n avg_lab = pd.DataFrame()\\n\\n for i in layers.index:\\n index = int(layers.layer[i])\\n if index == 1:\\n height = tbl_explsoilboring['dbl_KeyDepthToBottom'][i] * 0.00328084\\n else:\\n height = (tbl_explsoilboring['dbl_KeyDepthToBottom'][i]\\n - tbl_explsoilboring['dbl_KeyDepthToBottom'][i-1]) \\\\\\n * 0.00328084\\n soil_type = tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i] \\\\\\n if not pd.isna(\\n tbl_explsoilboring['txt_PrimarySoilDescriptionCode'][i]) else None\\n soil_type = primary_explsoil[soil_type] if soil_type else None\\n uscs_type = tbl_explsoilboring['txt_USCSCode'][i] if not \\\\\\n pd.isna(tbl_explsoilboring['txt_USCSCode'][i]) else None\\n if not soil_type and uscs_type:\\n soil_type = uscs_type\\n description = tbl_explsoilboring['txt_LayerDescription'][i] \\\\\\n if not pd.isna(tbl_explsoilboring['txt_LayerDescription'][i]) \\\\\\n else None\\n nval = tbl_explsoilboring['dbl_BlowCountInterpreted'][i] \\\\\\n if not pd.isna(tbl_explsoilboring['dbl_BlowCountInterpreted'][i]) \\\\\\n else None\\n nval_from_avg_spt = avg_spt[avg_spt.index == index]\\n if not nval and not nval_from_avg_spt.empty:\\n nval = nval_from_avg_spt.values[0][0] if not pd.isna(\\n nval_from_avg_spt.values[0][0]) else None\\n angle = tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i] \\\\\\n if not pd.isna(\\n tbl_explsoilboring['dbl_FrictionAngleInterpreted'][i]) else None\\n data_from_avg_lab = avg_lab[avg_lab.index == index]\\n cohesion = tbl_explsoilboring['dbl_CohesionInterpreted'][i] \\\\\\n * 2.0885434273e-05 if not pd.isna(\\n tbl_explsoilboring['dbl_CohesionInterpreted'][i]) else None\\n if not cohesion and not data_from_avg_lab.empty:\\n cohesion = data_from_avg_lab.values[0][4] * 2.0885434273e-05 \\\\\\n if not pd.isna(data_from_avg_lab.values[0][4]) else None\\n if cohesion:\\n cohesion = cohesion if cohesion < 99 else None\\n tuw = tbl_explsoilboring['dbl_UnitWeightInterpreted'][i] * 0.062428 \\\\\\n if not pd.isna(tbl_explsoilboring['dbl_UnitWeightInterpreted'][i]) \\\\\\n else None\\n if not tuw and not data_from_avg_lab.empty:\\n tuw = data_from_avg_lab.values[0][1] * 0.062428 if not pd.isna(\\n data_from_avg_lab.values[0][1]) else None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][0]):\\n water_content = int(data_from_avg_lab.values[0][0])\\n else:\\n water_content = None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][2]):\\n liquid_limit = int(data_from_avg_lab.values[0][2])\\n else:\\n liquid_limit = None\\n if not data_from_avg_lab.empty and not pd.isna(\\n data_from_avg_lab.values[0][3]):\\n plasticity = int(data_from_avg_lab.values[0][3])\\n else:\\n plasticity = None\\n layer = Layers(\\n boring=exploration,\\n index=index,\\n soil_type=soil_type,\\n height=height,\\n description=description,\\n nval=nval,\\n ssuu=cohesion,\\n friction_angle=angle,\\n tuw=tuw,\\n water_content=water_content,\\n liquid_limit=liquid_limit,\\n plasticity=plasticity\\n )\\n db.session.add(layer)\",\n \"def layers(self):\\n return self['layers']\",\n \"def expandWithoutMutex(self, previousLayer):\\n previousLayerProposition = previousLayer.getPropositionLayer()\\n \\\"*** YOUR CODE HERE ***\\\"\",\n \"def __init__(self, layers) -> None:\\n super(FeaturizerPhi, self).__init__()\\n self.layers = layers\",\n \"def filesystem_layers(self):\\n pass\",\n \"def renderLayerPostProcess(*args, keepImages: bool=True, sceneName: Union[AnyStr, bool]=\\\"\\\",\\n q=True, query=True, **kwargs)->Union[None, Any]:\\n pass\",\n \"def update(self):\\r\\n\\r\\n # Update the vision frames in the system\\r\\n self._system.update()\\r\\n\\r\\n # Create blank PIL images to hold the video streams\\r\\n layered = PIL.Image.new('RGBA', (400, 400))\\r\\n stacked = PIL.Image.new('RGBA', (200, 800))\\r\\n control = PIL.Image.new('RGBA', (600, 800))\\r\\n\\r\\n focalpoint = self._system[self._appString[\\\"device\\\"].get()].focalpoint()\\r\\n # print(focalpoint)\\r\\n\\r\\n # Get each vision key and vision for the selected device\\r\\n visionList = [(visionKey, vision) for visionKey, vision in self._system[self._appString[\\\"device\\\"].get()]]\\r\\n\\r\\n # Loop through each vision in the vision list\\r\\n for i, (visionKey, vision) in enumerate(visionList):\\r\\n\\r\\n # Grab the frames from the vision when it is \\\"curr\\\"\\r\\n frameList = [frame for frameKey, frame in vision if frameKey==self._appString[\\\"frame\\\"].get()]\\r\\n\\r\\n # Loop through each frame in the frame list\\r\\n for frame in frameList:\\r\\n\\r\\n # Get the properties and turn the image into RGBA\\r\\n ratio, size = vision.properties()\\r\\n rgbFrame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGBA)\\r\\n\\r\\n # print(rgbFrame.shape)\\r\\n width, height, channels = rgbFrame.shape\\r\\n\\r\\n # Paste the images together in layered\\r\\n\\r\\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (int(400 * ratio), int(400 * ratio))))\\r\\n layered.paste(imgFrame, (int(200 * (1 - ratio)), int(200 * (1 - ratio))))\\r\\n\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 // width)), int(200 * (1 - ratio) - focalpoint[1] * (200 // height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1)), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200/width) / ratio), int(200 * (1 - ratio) - focalpoint[1] * (200/height) / ratio)))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (200 / width)), int(200 * (1 - ratio) - focalpoint[1] * (200 / height))))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (ratio ** -1) / 200), int(200 * (1 - ratio) - focalpoint[1] * (ratio ** -1) / 200)))\\r\\n # layered.paste(imgFrame, (int(200 * (1 - ratio) + focalpoint[0] * (400//width * (1- ratio))), int(200 * (1 - ratio) - focalpoint[1] * (400//height * (1 - ratio)))))\\r\\n\\r\\n # Paste the images together in stacked\\r\\n imgFrame = PIL.Image.fromarray(cv2.resize(rgbFrame, (200, 200)))\\r\\n stacked.paste(imgFrame, (0, 200 * i))\\r\\n\\r\\n # Add the stacked image to the canvas\\r\\n self._pilFrames[\\\"stacked\\\"] = PIL.ImageTk.PhotoImage(image=stacked)\\r\\n self._appCanvas[\\\"stacked\\\"].create_image(100, 0, image=self._pilFrames[\\\"stacked\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Add the layered image to the canvas\\r\\n self._pilFrames[\\\"layered\\\"] = PIL.ImageTk.PhotoImage(image=layered)\\r\\n self._appCanvas[\\\"layered\\\"].create_image(0, 0, image=self._pilFrames[\\\"layered\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Add the control image to the canvas\\r\\n imgFrame = cv2.cvtColor(self._system[self._appString[\\\"device\\\"].get()][self._appString[\\\"vision\\\"].get()][self._appString[\\\"frame\\\"].get()], cv2.COLOR_BGR2RGBA)\\r\\n control = PIL.Image.fromarray(cv2.resize(imgFrame, (600, 600)))\\r\\n self._pilFrames[\\\"control\\\"] = PIL.ImageTk.PhotoImage(image=control)\\r\\n self._appCanvas[\\\"control\\\"].create_image(100, 90, image=self._pilFrames[\\\"control\\\"], anchor=tkinter.NW)\\r\\n\\r\\n # Continue to update with a delay of 15\\r\\n self.after(15, self.update)\",\n \"def _get_layers(self) :\\n \\n return self._layers\",\n \"def initLayers(self):\\n self.addRow(self.currentLayerData)\\n self.selected = self.jobRow.child(0)\\n self.setSelectedFromItem(self.selected)\",\n \"def layers(self):\\r\\n return self._flc.layers\",\n \"def num_layers(self): # -> int:\\n ...\",\n \"def get_patches(rimage, gimage, mimage, num_patches=48, patch_size=80, patch_stride=80):\\n num_FSpatches = 16\\n num_RApatches = 32\\n rpatches = []\\n gpatches = []\\n mpatches = []\\n #R_imgs = ((rimage+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'rainy.jpg', R_imgs[0,:,:,:])\\n for i in range(int(math.sqrt(num_FSpatches))):\\n for j in range(int(math.sqrt(num_FSpatches))):\\n point_x = patch_stride*i\\n point_y = patch_stride*j\\n rpatch = rimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n #print(point_x)\\n #print(point_y)\\n #print(point_y+patch_size)\\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'patch_%d_%d.jpg'%(i,j), P_imgs[0,:,:,:])\\n #print(rpatch.shape)\\n rpatches.append(rpatch)\\n #print(np.array(rpatches).shape)\\n gpatch = gimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n gpatches.append(gpatch)\\n mpatch = mimage[:,(point_x):(point_x+patch_size), (point_y):(point_y+patch_size),:]\\n mpatches.append(mpatch)\\n\\n for k in range(num_RApatches):\\n point1 = random.randint(0,240) # 116 comes from the image source size (320) - the patch dimension (80)\\n point2 = random.randint(0,240)\\n #rpatch = tf.image.crop_to_bounding_box(rimage, point1, point2, patch_size, patch_size)\\n rpatch = rimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n #P_imgs = ((rpatch+1)*127.5).astype(np.uint8)\\n #scipy.misc.imsave('results'+'/' + 'patch_%d.jpg'%i, P_imgs[0,:,:,:])\\n #print(rpatch.shape)\\n rpatches.append(rpatch)\\n #print(np.array(rpatches).shape)\\n gpatch = gimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n gpatches.append(gpatch)\\n mpatch = mimage[:,(point1):(point1+patch_size), (point2):(point2+patch_size),:]\\n mpatches.append(mpatch)\\n\\n rpatches = np.array(rpatches)\\n rpatches = np.squeeze(rpatches)\\n #print(rpatches.shape)\\n gpatches = np.array(gpatches)\\n gpatches = np.squeeze(gpatches)\\n mpatches = np.array(mpatches)\\n mpatches = np.squeeze(mpatches)\\n #assert rpatches.get_shape().dims == [num_patches, patch_size, patch_size, 3]\\n assert rpatches.shape == (num_patches, patch_size, patch_size, 3)\\n return rpatches, gpatches, mpatches\",\n \"def _init_layers(self):\\n self._init_predictor()\\n if self.use_edge_fusion:\\n self._init_edge_module()\",\n \"def lightened_v2(inputs, is_training=True,\\n dropout_keep_prob=0.8,\\n reuse=None,\\n scope='LightenedV1'):\\n end_points = {}\\n \\n with tf.variable_scope(scope, 'LightenedV1', [inputs], reuse=reuse):\\n with slim.arg_scope([slim.batch_norm, slim.dropout],\\n is_training=is_training):\\n with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],\\n stride=1, padding='SAME'):\\n \\n endpoints = {}\\n \\n net=conv(inputs, 48, 9, stride=1, padding='VALID', scope='conv1_9x9')\\n end_points['conv1_9x9'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool1')\\n end_points['pool1'] = net\\n \\n net=conv(net, 96, 5, stride=1, padding='VALID', scope='conv2_5x5')\\n end_points['conv2_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool2')\\n end_points['pool2'] = net\\n\\n net=conv(net, 128, 5, stride=1, padding='VALID', scope='conv3_5x5')\\n end_points['conv3_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool3')\\n end_points['pool3'] = net\\n\\n net=conv(net, 192, 4, stride=1, padding='VALID', scope='conv4_4x4')\\n end_points['conv4_5x5'] = net\\n net = slim.max_pool2d(net, 2, stride=2, padding='SAME', scope='pool4')\\n end_points['pool4'] = net\\n\\n with tf.variable_scope('Logits'):\\n net = slim.flatten(net)\\n net = slim.dropout(net, dropout_keep_prob, is_training=is_training,\\n scope='Dropout') \\n endpoints['PreLogitsFlatten'] = net\\n \\n return net, end_points\",\n \"def layerListObject(self, parent):\\n pass\",\n \"def __init__(self, batch_env):\\n\\n dims = [84, 84]\\n nature_transform = lambda o: tf.image.rgb_to_grayscale( # pylint: disable=g-long-lambda\\n tf.image.resize_images(o, dims))\\n\\n super(WarpFrameWrapper, self).__init__(batch_env, transform_observation=(\\n nature_transform, dims, tf.float32))\",\n \"def update_layer(LayerId=None, Name=None, Shortname=None, Attributes=None, CloudWatchLogsConfiguration=None, CustomInstanceProfileArn=None, CustomJson=None, CustomSecurityGroupIds=None, Packages=None, VolumeConfigurations=None, EnableAutoHealing=None, AutoAssignElasticIps=None, AutoAssignPublicIps=None, CustomRecipes=None, InstallUpdatesOnBoot=None, UseEbsOptimizedInstances=None, LifecycleEventConfiguration=None):\\n pass\",\n \"def addExportLayerToCoreml(builder):\\n outputNames = [output.name for output in builder.spec.description.output]\\n\\n for i, outputName in enumerate(outputNames):\\n # formulas: https://github.com/ultralytics/yolov5/issues/471\\n builder.add_activation(\\n name=f\\\"sigmoid_{outputName}\\\",\\n non_linearity=\\\"SIGMOID\\\",\\n input_name=outputName,\\n output_name=f\\\"{outputName}_sigmoid\\\",\\n )\\n\\n ### Coordinates calculation ###\\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2) -> nC = 640 / strides[i]\\n builder.add_slice(\\n name=f\\\"slice_coordinates_xy_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_sliced_coordinates_xy\\\",\\n axis=\\\"width\\\",\\n start_index=0,\\n end_index=2,\\n )\\n # x,y * 2\\n builder.add_elementwise(\\n name=f\\\"multiply_xy_by_two_{outputName}\\\",\\n input_names=[f\\\"{outputName}_sliced_coordinates_xy\\\"],\\n output_name=f\\\"{outputName}_multiplied_xy_by_two\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=2,\\n )\\n # x,y * 2 - 0.5\\n builder.add_elementwise(\\n name=f\\\"subtract_0_5_from_xy_{outputName}\\\",\\n input_names=[f\\\"{outputName}_multiplied_xy_by_two\\\"],\\n output_name=f\\\"{outputName}_subtracted_0_5_from_xy\\\",\\n mode=\\\"ADD\\\",\\n alpha=-0.5,\\n )\\n grid = make_grid(featureMapDimensions[i], featureMapDimensions[i]).numpy()\\n # x,y * 2 - 0.5 + grid[i]\\n builder.add_bias(\\n name=f\\\"add_grid_from_xy_{outputName}\\\",\\n input_name=f\\\"{outputName}_subtracted_0_5_from_xy\\\",\\n output_name=f\\\"{outputName}_added_grid_xy\\\",\\n b=grid,\\n shape_bias=grid.shape,\\n )\\n # (x,y * 2 - 0.5 + grid[i]) * stride[i]\\n builder.add_elementwise(\\n name=f\\\"multiply_xy_by_stride_{outputName}\\\",\\n input_names=[f\\\"{outputName}_added_grid_xy\\\"],\\n output_name=f\\\"{outputName}_calculated_xy\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=strides[i],\\n )\\n\\n # input (1, 3, nC, nC, 85), output (1, 3, nC, nC, 2)\\n builder.add_slice(\\n name=f\\\"slice_coordinates_wh_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_sliced_coordinates_wh\\\",\\n axis=\\\"width\\\",\\n start_index=2,\\n end_index=4,\\n )\\n # w,h * 2\\n builder.add_elementwise(\\n name=f\\\"multiply_wh_by_two_{outputName}\\\",\\n input_names=[f\\\"{outputName}_sliced_coordinates_wh\\\"],\\n output_name=f\\\"{outputName}_multiplied_wh_by_two\\\",\\n mode=\\\"MULTIPLY\\\",\\n alpha=2,\\n )\\n # (w,h * 2) ** 2\\n builder.add_unary(\\n name=f\\\"power_wh_{outputName}\\\",\\n input_name=f\\\"{outputName}_multiplied_wh_by_two\\\",\\n output_name=f\\\"{outputName}_power_wh\\\",\\n mode=\\\"power\\\",\\n alpha=2,\\n )\\n # (w,h * 2) ** 2 * anchor_grid[i]\\n anchor = (\\n anchorGrid[i]\\n .expand(-1, featureMapDimensions[i], featureMapDimensions[i], -1)\\n .numpy()\\n )\\n builder.add_load_constant_nd(\\n name=f\\\"anchors_{outputName}\\\",\\n output_name=f\\\"{outputName}_anchors\\\",\\n constant_value=anchor,\\n shape=anchor.shape,\\n )\\n builder.add_elementwise(\\n name=f\\\"multiply_wh_with_achors_{outputName}\\\",\\n input_names=[f\\\"{outputName}_power_wh\\\", f\\\"{outputName}_anchors\\\"],\\n output_name=f\\\"{outputName}_calculated_wh\\\",\\n mode=\\\"MULTIPLY\\\",\\n )\\n\\n builder.add_concat_nd(\\n name=f\\\"concat_coordinates_{outputName}\\\",\\n input_names=[f\\\"{outputName}_calculated_xy\\\", f\\\"{outputName}_calculated_wh\\\"],\\n output_name=f\\\"{outputName}_raw_coordinates\\\",\\n axis=-1,\\n )\\n builder.add_scale(\\n name=f\\\"normalize_coordinates_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_coordinates\\\",\\n output_name=f\\\"{outputName}_raw_normalized_coordinates\\\",\\n W=torch.tensor([1 / 640]).numpy(),\\n b=0,\\n has_bias=False,\\n )\\n\\n ### Confidence calculation ###\\n builder.add_slice(\\n name=f\\\"slice_object_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_object_confidence\\\",\\n axis=\\\"width\\\",\\n start_index=4,\\n end_index=5,\\n )\\n builder.add_slice(\\n name=f\\\"slice_label_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_sigmoid\\\",\\n output_name=f\\\"{outputName}_label_confidence\\\",\\n axis=\\\"width\\\",\\n start_index=5,\\n end_index=0,\\n )\\n # confidence = object_confidence * label_confidence\\n builder.add_multiply_broadcastable(\\n name=f\\\"multiply_object_label_confidence_{outputName}\\\",\\n input_names=[\\n f\\\"{outputName}_label_confidence\\\",\\n f\\\"{outputName}_object_confidence\\\",\\n ],\\n output_name=f\\\"{outputName}_raw_confidence\\\",\\n )\\n\\n # input: (1, 3, nC, nC, 85), output: (3 * nc^2, 85)\\n builder.add_flatten_to_2d(\\n name=f\\\"flatten_confidence_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_confidence\\\",\\n output_name=f\\\"{outputName}_flatten_raw_confidence\\\",\\n axis=-1,\\n )\\n builder.add_flatten_to_2d(\\n name=f\\\"flatten_coordinates_{outputName}\\\",\\n input_name=f\\\"{outputName}_raw_normalized_coordinates\\\",\\n output_name=f\\\"{outputName}_flatten_raw_coordinates\\\",\\n axis=-1,\\n )\\n\\n builder.add_concat_nd(\\n name=\\\"concat_confidence\\\",\\n input_names=[\\n f\\\"{outputName}_flatten_raw_confidence\\\" for outputName in outputNames\\n ],\\n output_name=\\\"raw_confidence\\\",\\n axis=-2,\\n )\\n builder.add_concat_nd(\\n name=\\\"concat_coordinates\\\",\\n input_names=[\\n f\\\"{outputName}_flatten_raw_coordinates\\\" for outputName in outputNames\\n ],\\n output_name=\\\"raw_coordinates\\\",\\n axis=-2,\\n )\\n\\n builder.set_output(\\n output_names=[\\\"raw_confidence\\\", \\\"raw_coordinates\\\"],\\n output_dims=[(25200, numberOfClassLabels), (25200, 4)],\\n )\",\n \"def stage(self, stage):\\n self._stage = stage\\n self._layer = Sdf.Layer.CreateAnonymous()\\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\",\n \"def __init__(self,\\n decoder_options,\\n panoptic_deeplab_options,\\n bn_layer=tf.keras.layers.BatchNormalization):\\n super(PanopticDeepLab, self).__init__(name='PanopticDeepLab')\\n\\n low_level_feature_keys = [\\n item.feature_key for item in panoptic_deeplab_options.low_level\\n ]\\n low_level_channels_project = [\\n item.channels_project for item in panoptic_deeplab_options.low_level\\n ]\\n\\n self._semantic_decoder = PanopticDeepLabSingleDecoder(\\n high_level_feature_name=decoder_options.feature_key,\\n low_level_feature_names=low_level_feature_keys,\\n low_level_channels_project=low_level_channels_project,\\n aspp_output_channels=decoder_options.aspp_channels,\\n decoder_output_channels=decoder_options.decoder_channels,\\n atrous_rates=decoder_options.atrous_rates,\\n name='semantic_decoder',\\n aspp_use_only_1x1_proj_conv=decoder_options.aspp_use_only_1x1_proj_conv,\\n decoder_conv_type=decoder_options.decoder_conv_type,\\n bn_layer=bn_layer)\\n self._semantic_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.semantic_head.head_channels,\\n panoptic_deeplab_options.semantic_head.output_channels,\\n common.PRED_SEMANTIC_LOGITS_KEY,\\n name='semantic_head',\\n conv_type=panoptic_deeplab_options.semantic_head.head_conv_type,\\n bn_layer=bn_layer)\\n\\n self._instance_decoder = None\\n self._instance_center_head = None\\n self._instance_regression_head = None\\n\\n if panoptic_deeplab_options.instance.enable:\\n if panoptic_deeplab_options.instance.low_level_override:\\n low_level_options = panoptic_deeplab_options.instance.low_level_override\\n else:\\n low_level_options = panoptic_deeplab_options.low_level\\n\\n # If instance_decoder is set, use those options; otherwise reuse the\\n # architecture as defined for the semantic decoder.\\n if panoptic_deeplab_options.instance.HasField(\\n 'instance_decoder_override'):\\n decoder_options = (panoptic_deeplab_options.instance\\n .instance_decoder_override)\\n\\n low_level_feature_keys = [item.feature_key for item in low_level_options]\\n low_level_channels_project = [\\n item.channels_project for item in low_level_options\\n ]\\n\\n self._instance_decoder = PanopticDeepLabSingleDecoder(\\n high_level_feature_name=decoder_options.feature_key,\\n low_level_feature_names=low_level_feature_keys,\\n low_level_channels_project=low_level_channels_project,\\n aspp_output_channels=decoder_options.aspp_channels,\\n decoder_output_channels=decoder_options.decoder_channels,\\n atrous_rates=decoder_options.atrous_rates,\\n name='instance_decoder',\\n aspp_use_only_1x1_proj_conv=(\\n decoder_options.aspp_use_only_1x1_proj_conv),\\n decoder_conv_type=decoder_options.decoder_conv_type,\\n bn_layer=bn_layer)\\n self._instance_center_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.instance.center_head.head_channels,\\n panoptic_deeplab_options.instance.center_head.output_channels,\\n common.PRED_CENTER_HEATMAP_KEY,\\n name='instance_center_head',\\n conv_type=(\\n panoptic_deeplab_options.instance.center_head.head_conv_type),\\n bn_layer=bn_layer)\\n self._instance_regression_head = PanopticDeepLabSingleHead(\\n panoptic_deeplab_options.instance.regression_head.head_channels,\\n panoptic_deeplab_options.instance.regression_head.output_channels,\\n common.PRED_OFFSET_MAP_KEY,\\n name='instance_regression_head',\\n conv_type=(\\n panoptic_deeplab_options.instance.regression_head.head_conv_type),\\n bn_layer=bn_layer)\",\n \"def develop_image(pipeline, camera=None, batch=None, image=None, patch_size=0, patches=2, root_dir='./data', pipeline_args=None):\\n\\n if camera is not None:\\n supported_cameras = fsutil.listdir(os.path.join(root_dir, 'models', 'nip'), '.*')\\n if camera not in supported_cameras:\\n raise ValueError('Camera data not found ({})! Available cameras: {}'.format(camera, ', '.join(supported_cameras)))\\n root_dirname = os.path.join(root_dir, 'models', 'nip', camera)\\n data_dirname = os.path.join(root_dir, 'raw', 'training_data', camera)\\n\\n if patch_size != 0 and (patch_size < 4 or patch_size > 2048):\\n raise ValueError('Patch size seems to be invalid!')\\n\\n # Lazy imports to minimize delay for invalid command line parameters\\n import numpy as np\\n import imageio as io\\n import matplotlib.pyplot as plt\\n import tensorflow as tf\\n from models import pipelines\\n\\n # Construct the NIP model ---------------------------------------------------------------------\\n \\n if os.path.isdir(pipeline):\\n # Restore a NIP model from a training log\\n model = tfmodel.restore(pipeline, pipelines)\\n else:\\n # Construct the NIP model from class name (and optional arguments)\\n if pipeline_args is None:\\n model = getattr(pipelines, pipeline)()\\n else:\\n model = getattr(pipelines, pipeline)(**pipeline_args)\\n\\n loaded_model = False\\n candidate_dirs = [os.path.join(root_dirname, model.model_code), os.path.join(root_dirname)]\\n for candidate in candidate_dirs:\\n if os.path.isdir(candidate):\\n model.load_model(candidate)\\n loaded_model = True\\n break\\n\\n if not loaded_model:\\n raise FileNotFoundError(f'Could not find the corresponding model: {candidate_dirs}')\\n\\n # Load image(s) -------------------------------------------------------------------------------\\n \\n if image is None and batch is not None:\\n print('Loading a batch of {} images'.format(batch))\\n data = dataset.Dataset(data_dirname, n_images=0, v_images=batch, val_rgb_patch_size=patch_size or 256, val_n_patches=patches)\\n sample_x, sample_y = data.next_validation_batch(0, data.count_validation)\\n\\n with open('config/cameras.json') as f:\\n cameras = json.load(f)\\n cfa, srgb = cameras[camera]['cfa'], np.array(cameras[camera]['srgb'])\\n\\n elif image is not None:\\n print('Loading a RAW image {}'.format(image))\\n sample_x, cfa, srgb, _ = raw.unpack(image, expand=True)\\n sample_y = raw.process(image, brightness=None, expand=True)\\n\\n if isinstance(model, pipelines.ClassicISP):\\n print('Configuring ISP to CFA: {} & sRGB {}'.format(cfa, srgb.round(2).tolist()))\\n model.set_cfa_pattern(cfa)\\n model.set_srgb_conversion(srgb)\\n\\n sample_Y = model.process(sample_x).numpy()\\n\\n if patch_size > 0:\\n xx = (sample_y.shape[2] - patch_size) // 2\\n yy = (sample_y.shape[1] - patch_size) // 2\\n sample_y = sample_y[:, yy:yy+patch_size, xx:xx+patch_size, :]\\n sample_Y = sample_Y[:, yy:yy+patch_size, xx:xx+patch_size, :]\\n\\n psnrs = metrics.psnr(sample_y, sample_Y)\\n ssims = metrics.ssim(sample_y, sample_Y)\\n\\n print('sample x: {}'.format(sample_x.shape))\\n print('sample y: {}'.format(sample_y.shape))\\n print('sample Y: {}'.format(sample_Y.shape))\\n\\n # Plot images ---------------------------------------------------------------------------------\\n if len(sample_y) > 1:\\n sample_y = plots.thumbnails(sample_y, batch, True)\\n sample_Y = plots.thumbnails(sample_Y, batch, True)\\n else:\\n sample_y = sample_y.squeeze()\\n sample_Y = sample_Y.squeeze()\\n\\n print('thumbnails: {}'.format(sample_y.shape))\\n\\n ncols = 1 if sample_y.shape[1] > sample_y.shape[0] else 2\\n nrows = 2 if ncols == 1 else 1\\n fig, axes = plt.subplots(nrows, ncols)\\n\\n plots.image(sample_Y, '{}, PSNR={:.1f} dB, SSIM={:.2f} : {{}}'.format(model.model_code, float(psnrs.mean()), float(ssims.mean())), axes=axes[0])\\n plots.image(sample_y, 'Target RGB images () : {}', axes=axes[1])\\n\\n plt.show()\\n plt.close()\",\n \"def __init__(self,\\n high_level_feature_name,\\n low_level_feature_names,\\n low_level_channels_project,\\n aspp_output_channels,\\n decoder_output_channels,\\n atrous_rates,\\n name,\\n aspp_use_only_1x1_proj_conv=False,\\n decoder_conv_type='depthwise_separable_conv',\\n bn_layer=tf.keras.layers.BatchNormalization):\\n super(PanopticDeepLabSingleDecoder, self).__init__(name=name)\\n self._channel_axis = 3\\n\\n self._aspp = aspp.ASPP(\\n aspp_output_channels,\\n atrous_rates,\\n aspp_use_only_1x1_proj_conv=aspp_use_only_1x1_proj_conv,\\n name='aspp',\\n bn_layer=bn_layer)\\n self._high_level_feature_name = high_level_feature_name\\n\\n if len(low_level_feature_names) != len(low_level_channels_project):\\n raise ValueError('The Panoptic-DeepLab decoder requires the same number '\\n 'of low-level features as the number of low-level '\\n 'projection channels. But got %d and %d.'\\n % (len(low_level_feature_names),\\n len(low_level_channels_project)))\\n\\n self._low_level_feature_names = low_level_feature_names\\n\\n for i, channels_project in enumerate(low_level_channels_project):\\n # Check if channel sizes increases and issue a warning.\\n if i > 0 and low_level_channels_project[i - 1] < channels_project:\\n logging.warning(\\n 'The low level projection channels usually do not '\\n 'increase for features with higher spatial resolution. '\\n 'Please make sure, this behavior is intended.')\\n current_low_level_conv_name, current_fusion_conv_name = (\\n utils.get_low_level_conv_fusion_conv_current_names(i))\\n utils.safe_setattr(\\n self, current_low_level_conv_name, convolutions.Conv2DSame(\\n channels_project,\\n kernel_size=1,\\n name=utils.get_layer_name(current_low_level_conv_name),\\n use_bias=False,\\n use_bn=True,\\n bn_layer=bn_layer,\\n activation='relu'))\\n\\n utils.safe_setattr(\\n self, current_fusion_conv_name, convolutions.StackedConv2DSame(\\n conv_type=decoder_conv_type,\\n num_layers=1,\\n output_channels=decoder_output_channels,\\n kernel_size=5,\\n name=utils.get_layer_name(current_fusion_conv_name),\\n use_bias=False,\\n use_bn=True,\\n bn_layer=bn_layer,\\n activation='relu'))\",\n \"def inception_network():\\n input = K.Input(shape=(224, 224, 3))\\n\\n conv_1 = K.layers.Conv2D(\\n filters=64,\\n kernel_size=7,\\n strides=2,\\n padding='same',\\n activation='relu'\\n )(input)\\n\\n pool_1 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(conv_1)\\n\\n conv_2 = K.layers.Conv2D(\\n filters=64,\\n kernel_size=1,\\n strides=1,\\n padding='valid',\\n activation='relu'\\n )(pool_1)\\n\\n conv_3 = K.layers.Conv2D(\\n filters=192,\\n kernel_size=3,\\n strides=1,\\n padding='same',\\n activation='relu'\\n )(conv_2)\\n\\n pool_2 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(conv_3)\\n\\n inception_1 = inception_block(pool_2, (64, 96, 128, 16, 32, 32))\\n inception_2 = inception_block(inception_1, (128, 128, 192, 32, 96, 64))\\n\\n pool_3 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same'\\n )(inception_2)\\n\\n inception_3 = inception_block(pool_3, (192, 96, 208, 16, 48, 64))\\n inception_4 = inception_block(inception_3, (160, 112, 224, 24, 64, 64))\\n inception_5 = inception_block(inception_4, (128, 128, 256, 24, 64, 64))\\n inception_6 = inception_block(inception_5, (112, 144, 288, 32, 64, 64))\\n inception_7 = inception_block(inception_6, (256, 160, 320, 32, 128, 128))\\n\\n pool_4 = K.layers.MaxPooling2D(\\n pool_size=3,\\n strides=2,\\n padding='same',\\n )(inception_7)\\n\\n inception_8 = inception_block(pool_4, (256, 160, 320, 32, 128, 128))\\n inception_9 = inception_block(inception_8, (384, 192, 384, 48, 128, 128))\\n\\n pool_5 = K.layers.AveragePooling2D(\\n pool_size=7,\\n strides=1,\\n padding='valid'\\n )(inception_9)\\n\\n dropout = K.layers.Dropout(rate=0.4)(pool_5)\\n linear = K.layers.Dense(units=1000, activation='softmax')(dropout)\\n\\n return K.Model(inputs=input, outputs=linear)\",\n \"def __init__(self):\\n super(GatherLastLayer, self).__init__()\",\n \"def SetUpLayerManager(self):\\n pass\",\n \"def _save_data(self, outputs, latents):\\n list(map(lambda e, v: e.append(v), self.curr_patches, outputs))\\n list(map(lambda e, v: e.append(v), self.curr_latents, latents))\\n self.count_patches += len(self.curr_patches[0][-1])\\n\\n while self.count_patches >= self.patches_ref:\\n aux_patches = list(map(lambda e: tf.concat(e, axis=0),\\n self.curr_patches))\\n aux_latents = list(map(lambda e: tf.concat(e, axis=0),\\n self.curr_latents))\\n\\n data_name = str((self.out_folder / str(Folders.RAW_DATA))\\n / self.curr_img.stem)\\n for lvl, (aux_p, aux_l) in enumerate(zip(aux_patches, aux_latents)):\\n patches, remaining = tf.split(\\n aux_p, [self.patches_ref, len(aux_p) - self.patches_ref],\\n axis=0)\\n self.curr_patches[lvl] = [remaining]\\n latent, remaining = tf.split(\\n aux_l, [self.patches_ref, len(aux_l) - self.patches_ref],\\n axis=0)\\n self.curr_latents[lvl] = [remaining]\\n\\n patch_size = len(patches[0])\\n with Image.open(self.curr_img) as img:\\n # width, height -> height, width (to be compatible with\\n # arrays)\\n img_size = np.array(img.size)[::-1]\\n p_size = np.ceil(img_size / patch_size).astype(int)\\n\\n img = tf.reshape(\\n patches, [*p_size, *self.m_cfg['configs']['patch']])\\n img = tf.transpose(img, [0, 2, 1, 3, 4])\\n img = tf.reshape(img, [*patch_size * p_size, 3])\\n img = tf.image.convert_image_dtype(img, tf.uint8)\\n img = img[:img_size[0], :img_size[1], :]\\n\\n latent_name = data_name + '_latent' + str(lvl)\\n serialized = tf.io.serialize_tensor(latent)\\n tf.io.write_file(latent_name, serialized)\\n\\n img_name = data_name + '_' + str(lvl) + self.curr_img.suffix\\n tf.io.write_file(img_name, tf.image.encode_png(\\n img, compression=0))\\n\\n with open(Path(data_name).with_suffix('.txt'), 'w') as file:\\n file.write(str(self.curr_img))\\n\\n self.count_patches = len(self.curr_patches[0][0])\\n if len(self.img_paths):\\n self.curr_img = Path(self.img_paths.pop(0))\\n self.patches_ref = self.calc_n_patches(\\n self.curr_img, self.m_cfg['configs']['patch'][0])\",\n \"def registerPatches(self,pl):\\n self.set('patchmesh.patches',pl)\",\n \"def multi_run_wrapper(args):\\n\\treturn img_preprocessing(*args)\",\n \"def _set_freeze_layers(self):\\n for layer in self.encoder.layers[:self.freeze_layers]:\\n layer.trainable = False\",\n \"def __call__(self, inputs, training):\\n\\n with self._model_variable_scope():\\n if self.data_format == 'channels_first':\\n # Convert the inputs from channels_last (NHWC) to channels_first (NCHW).\\n # This provides a large performance boost on GPU. See\\n # https://www.tensorflow.org/performance/performance_guide#data_formats\\n inputs = tf.transpose(a=inputs, perm=[0, 3, 1, 2])\\n\\n with tf.variable_scope('stem'):\\n inputs = conv2d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n inputs = tf.identity(inputs, 'stem_conv1')\\n\\n inputs = conv2d_fixed_padding(\\n inputs=inputs, filters=self.num_filters, kernel_size=self.kernel_size,\\n strides=self.conv_stride, data_format=self.data_format)\\n inputs = batch_norm(inputs, training, self.data_format)\\n inputs = tf.nn.relu(inputs)\\n inputs = tf.identity(inputs, 'stem_conv2')\\n\\n # stage 1\\n with tf.variable_scope('stage1'):\\n stage1_out = block_layer(\\n inputs=inputs, inp_filters=self.num_filters,\\n filters=self.num_filters, bottleneck=True,\\n block_fn=_bottleneck_block_v1, blocks=self.block_sizes[0],\\n strides=1, training=training,\\n name='stage1', data_format=self.data_format)\\n\\n pre_stage_channels = self.num_filters * 4\\n stagex_input = [stage1_out]\\n num_inchannels = [pre_stage_channels]\\n\\n # stage 2 to last stage\\n for i, num_modules in enumerate(self.module_sizes[1:], 2):\\n with tf.variable_scope('stage{}'.format(i)):\\n num_channels = [self.hrnet_size * 2 ** n for n in range(i)]\\n transn = Transition(\\n stagex_input[-1], num_outchannels=num_channels[-1], \\n training=training, data_format=self.data_format)\\n stagex_input.append(transn)\\n num_inchannels.append(num_inchannels[-1])\\n\\n for j in range(num_modules):\\n if i == len(self.module_sizes) and j == num_modules - 1:\\n multi_scale_output = False\\n else:\\n multi_scale_output = True\\n\\n stagex_input = HighResolutionModule(\\n stagex_input, num_branches=i, \\n num_inchannels=num_inchannels, num_channels=num_channels,\\n bottleneck=self.bottleneck, block_fn=self.block_fn,\\n num_blocks=self.block_sizes[i - 1], training=training,\\n name='stage{}_block{}'.format(i, j), data_format=self.data_format,\\n multi_scale_output=multi_scale_output)\\n num_inchannels = num_channels\\n\\n return stagex_input\",\n \"def _augment_pipeline_cfg(self):\",\n \"def __call__(self, **kwargs):\\n segname = 'block_{}_expand_relu'\\n blocks = [13, 6, 3, 1]\\n skips = [self._backbone.get_layer(segname.format(i)) for i in blocks]\\n backbone_out = self._backbone.get_layer('block_16_project')\\n\\n p5 = self._fpn_block(backbone_out.output, skips[0].output)\\n p4 = self._fpn_block(p5, skips[1].output)\\n p3 = self._fpn_block(p4, skips[2].output)\\n p2 = self._fpn_block(p3, skips[3].output)\\n\\n s5 = self._conv_block(p5, 128)\\n s4 = self._conv_block(p4, 128)\\n s3 = self._conv_block(p3, 128)\\n s2 = self._conv_block(p2, 128)\\n\\n s5 = tf.keras.layers.UpSampling2D(\\n size=(8, 8),\\n interpolation='nearest'\\n )(s5)\\n\\n s4 = tf.keras.layers.UpSampling2D(\\n size=(4, 4),\\n interpolation='nearest'\\n )(s4)\\n\\n s3 = tf.keras.layers.UpSampling2D(\\n size=(2, 2),\\n interpolation='nearest'\\n )(s3)\\n\\n concat = [s5, s4, s3, s2]\\n x = tf.keras.layers.Concatenate()(concat)\\n x = tf.keras.layers.Conv2D(\\n 64,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n x = tf.keras.layers.BatchNormalization()(x)\\n x = tf.keras.layers.Activation('relu')(x)\\n x = tf.keras.layers.UpSampling2D((2, 2))(x)\\n\\n x = tf.keras.layers.Conv2D(\\n 1,\\n kernel_size=3,\\n padding='same',\\n kernel_initializer='he_uniform'\\n )(x)\\n\\n out = tf.keras.layers.Activation('sigmoid')(x)\\n model = tf.keras.models.Model(\\n inputs=self._backbone.input,\\n outputs=out\\n )\\n\\n return model\",\n \"def add_bluprint_layers(self, models_dict, graph, class_names, dimensions,\\n show_activation, show_constant_input):\\n top_model = models_dict['top-model']\\n # Get the top-model color list containing\\n # the base color and the layer shades\\n top_model_color = top_model[1]\\n # get the model\\n top_model = top_model[0]\\n\\n # Get the layers of the model\\n layers = top_model[\\\"config\\\"][\\\"layers\\\"]\\n # Loop through the layers\\n for layer in layers:\\n # If the layer is not a model\\n if layer[\\\"class_name\\\"] != \\\"Model\\\":\\n # Get the layer name\\n layer_name = layer[\\\"name\\\"]\\n # If label only layer's class name\\n if class_names:\\n # Get the layer's information\\n layer_info = self.viz_utils.query_layer(layer_name,\\n models_dict)\\n # Get the layer's class name\\n layer_class = layer_info['class_name']\\n # If the layer is a a constant input layer,\\n # manually specify the class name\\n if layer_name.find('constant_input') != -1:\\n layer_class = 'Constant Input'\\n # Depending on the class name\\n # find the the layer shade\\n # If the layer is a constant_input layer\\n # the color is black\\n model_color = top_model_color[1].get(layer_class, \\\"black\\\")\\n else:\\n # If don't use class names for layers\\n # then use the layer name from the JSON\\n layer_class = layer_name\\n model_color = top_model_color[0]\\n\\n\\n # Add the node to the graph\\n graph = self.viz_utils.add_nodes(layer_name, graph,\\n layer_class, model_color,\\n dimensions, show_constant_input)\\n\\n # Add Blueprint Inbound Edges\\n graph = self.connect_blueprint_inbounds(models_dict,\\n layer, graph,\\n class_names, dimensions,\\n show_activation, show_constant_input)\\n else:\\n # Add Softmod\\n graph = self.connect_softmod_in_blueprint(models_dict,\\n layer, graph, class_names,\\n dimensions, show_activation, show_constant_input)\\n\\n return graph\",\n \"def load_registration_as_layers() -> None:\\n viewer = getattr(widget, \\\"viewer\\\").value\\n registration_directory = pathlib.Path(\\n getattr(widget, \\\"registration_output_folder\\\").value\\n )\\n add_registered_image_layers(\\n viewer, registration_directory=registration_directory\\n )\",\n \"def intermediate_layer(layer, filters, kernel_size):\\n layer = Conv2D(filters=filters, kernel_size=kernel_size,\\n strides=(2, 2), padding=\\\"same\\\")(layer)\\n layer = LeakyReLU(alpha=LEAKY_RELU_ALPHA)(layer)\\n return layer\",\n \"def deep_dream_of_extreme_control(FLAGS,model,input_images=[],num_iterations=10,step_size=0.1):\\n if len(input_images) == 0:\\n # use predefined images\\n img_dir='/esat/opal/kkelchte/docker_home/pilot_data/visualization_images'\\n input_images=sorted([img_dir+'/'+f for f in os.listdir(img_dir)])\\n\\n print(\\\"[tools.py]: extracting deep dream maps of {0} in {1}\\\".format([os.path.basename(i) for i in input_images], os.path.dirname(input_images[0])))\\n \\n # experts=np.asarray([[k]*(FLAGS.action_quantity if FLAGS.discrete else 1) for v in sorted(model.factor_offsets.values()) for k in model.factor_offsets.keys() if model.factor_offsets[k]==v]).flatten()\\n\\n inputs = load_images(input_images, model.input_size[1:])\\n \\n # collect gradients for output endpoint of evaluation model\\n grads={}\\n with tf.device('/cpu:0'):\\n output_tensor = model.endpoints['eval']['outputs']\\n for i in range(output_tensor.shape[1].value):\\n layer_loss = output_tensor[:,i]\\n gradients = tf.gradients(layer_loss, model.inputs)[0]\\n gradients /= (tf.sqrt(tf.reduce_mean(tf.square(gradients))) + 1e-5)\\n grads[output_tensor.name+'_'+str(i)]=gradients\\n\\n\\n # apply gradient ascent for all outputs and each input image\\n # if number of outputs ==1 apply gradient descent for contrast\\n if len(grads.keys())== 1:\\n opposite_results={}\\n else:\\n opposite_results=None\\n\\n import copy\\n results = {}\\n for gk in grads.keys(): \\n results[gk]=copy.deepcopy(inputs)\\n if isinstance(opposite_results,dict): opposite_results[gk]=copy.deepcopy(inputs)\\n\\n for step in range(num_iterations):\\n if step%10==0: print \\\"{0} step: {1}\\\".format(time.ctime(), step)\\n for i,gk in enumerate(sorted(grads.keys())):\\n results[gk] += step_size * model.sess.run(grads[gk], {model.inputs: results[gk]})\\n if isinstance(opposite_results,dict):\\n opposite_results[gk] -= step_size * model.sess.run(grads[gk], {model.inputs: opposite_results[gk]})\\n\\n # Normalize results within 0:1 range\\n clean_results={}\\n for gk in results.keys():\\n clean_results[gk]=[]\\n for i in range(results[gk].shape[0]):\\n clean_results[gk].append(deprocess_image(results[gk][i], one_channel=True))\\n # results[gk][i]=deprocess_image(results[gk][i], one_channel=True)\\n if isinstance(opposite_results,dict):\\n opposite_results[gk][i]=deprocess_image(opposite_results[gk][i])\\n\\n # combine adjust input images in one overview image\\n # one column for each input image\\n # one row with each extreme control for separate and difference images\\n num_rows=1+len(results.keys())\\n fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\\n # fig, axes = plt.subplots(num_rows ,min(len(input_images),5),figsize=(23, 4*(len(grads.keys())+1)))\\n # add original images in first row\\n for i in range(axes.shape[1]):\\n axes[0, i].set_title(os.path.basename(input_images[i]).split('.')[0])\\n axes[0, i].imshow(matplotlibprove(inputs[i]), cmap='inferno')\\n axes[0, i].axis('off')\\n\\n # add for each filter the modified input\\n row_index=1\\n for gk in sorted(results.keys()):\\n for i in range(axes.shape[1]):\\n # print gk\\n # axes[row_index, i].set_title('Grad Asc: '+gk.split('/')[1]+'/'+gk[-1]) \\n axes[row_index, i].set_title('Grad Asc: '+gk)\\n # axes[row_index, i].set_title(experts[row_index-1])\\n\\n axes[row_index, i].imshow(np.concatenate((inputs[i],np.expand_dims(clean_results[gk][i],axis=2)), axis=2), cmap='inferno')\\n # axes[row_index, i].imshow(matplotlibprove(results[gk][i]), cmap='inferno')\\n axes[row_index, i].axis('off')\\n row_index+=1\\n # In cas of continouos controls: visualize the gradient descent and difference\\n # if isinstance(opposite_results,dict):\\n # for gk in opposite_results.keys():\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Grad Desc: '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Grad Desc: '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(opposite_results[gk][i]), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n \\n # # add difference\\n # for gk in opposite_results.keys():\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Diff: '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Diff: '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((opposite_results[gk][i]-results[gk][i])**2)), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n # else:\\n # # add difference between 2 exteme actions\\n # gk_left=sorted(results.keys())[0]\\n # gk_right=sorted(results.keys())[-1]\\n # for i in range(axes.shape[1]):\\n # # axes[row_index, i].set_title('Diff : '+gk.split('/')[1]) \\n # axes[row_index, i].set_title('Diff : '+gk) \\n # axes[row_index, i].imshow(matplotlibprove(deprocess_image((results[gk_left][i]-results[gk_right][i])**2)), cmap='inferno')\\n # axes[row_index, i].axis('off')\\n # row_index+=1\\n \\n \\n plt.savefig(FLAGS.summary_dir+FLAGS.log_tag+'/control_dream_maps.jpg',bbox_inches='tight')\\n # plt.show()\",\n \"def build(self, unused_input_shapes):\\n self.layers = []\\n for i in range(self.num_hidden_layers):\\n self.layers.append(\\n TransformerDecoderBlock(\\n hidden_size=self.hidden_size,\\n num_attention_heads=self.num_attention_heads,\\n intermediate_size=self.intermediate_size,\\n intermediate_activation=self.intermediate_activation,\\n hidden_dropout_prob=self.hidden_dropout_prob,\\n attention_probs_dropout_prob=self.attention_probs_dropout_prob,\\n initializer_range=self.initializer_range,\\n multi_channel_cross_attention=self.multi_channel_cross_attention,\\n name=(\\\"layer_%d\\\" % i)))\\n super(TransformerDecoder, self).build(unused_input_shapes)\",\n \"def GetShapesOfLayer(self, *args):\\n return _XCAFDoc.XCAFDoc_LayerTool_GetShapesOfLayer(self, *args)\",\n \"def __init__(self, layer_list_info):\\n self.layer_list_info = layer_list_info\",\n \"def depth_rendering(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\\n\\n x_indices = np.arange(lf_size[1])\\n y_indices = np.arange(lf_size[2])\\n b_indices = np.arange(lf_size[0])\\n\\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\\n\\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\\n stacked = np.stack(grid, 2)\\n positions = stacked.reshape(-1, 2)\\n\\n # Compute the distance from each lf position from the reference view\\n # Repeat the elements of this to match the size of the disparity map\\n ref_pos = np.array(\\n [lf_one_way // 2, lf_one_way // 2])\\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\\n\\n\\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\\n\\n # Compute the shifted pixels\\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\\n\\n #indices for linear interpolation in a square around the central point\\n x_low = np.around(x_shifted).astype(int)\\n #x_high = x_low + 1\\n\\n y_low = np.around(y_shifted).astype(int)\\n #y_high = y_low + 1\\n\\n #Place co-ordinates outside the image back into the image\\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\\n #x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\\n #y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\\n\\n #Gather the interpolation points\\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\\n #interp_pts_2 = np.stack((x_low_clip, y_high_clip))\\n #interp_pts_3 = np.stack((x_high_clip, y_low_clip))\\n #interp_pts_4 = np.stack((x_high_clip, y_high_clip))\\n\\n #Index into the images\\n desired_shape = lf_size\\n res_1 = torch_big_sample(ref_view, interp_pts_1, desired_shape)\\n return res_1\\n res_2 = torch_big_sample(ref_view, interp_pts_2, desired_shape)\\n res_3 = torch_big_sample(ref_view, interp_pts_3, desired_shape)\\n res_4 = torch_big_sample(ref_view, interp_pts_4, desired_shape)\\n\\n #Compute interpolation weights\\n x_low_f = x_low.astype(np.float32)\\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\\n d_x_high = 1.0 - d_x_low\\n y_low_f = y_low.astype(np.float32)\\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\\n d_y_high = 1.0 - d_y_low\\n\\n w1 = torch.from_numpy(d_x_low * d_y_low)\\n w2 = torch.from_numpy(d_x_low * d_y_high)\\n w3 = torch.from_numpy(d_x_high * d_y_low)\\n w4 = torch.from_numpy(d_x_high * d_y_high)\\n\\n #THEY AGREE AT THIS POINT\\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape), res_1)\\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape), res_2)\\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape), res_3)\\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape), res_4)\\n\\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\\n torch.add(novel_view, weighted_4, out=novel_view)\\n return novel_view\",\n \"def process_layer(layer_data):\\n layer_name = layer_data['name']\\n\\n # Instantiate layer.\\n from tensorflow.python.keras._impl.keras.layers import deserialize as deserialize_layer # pylint: disable=g-import-not-at-top\\n\\n layer = deserialize_layer(layer_data, custom_objects=custom_objects)\\n created_layers[layer_name] = layer\\n\\n # Gather layer inputs.\\n inbound_nodes_data = layer_data['inbound_nodes']\\n for node_data in inbound_nodes_data:\\n # We don't process nodes (i.e. make layer calls)\\n # on the fly because the inbound node may not yet exist,\\n # in case of layer shared at different topological depths\\n # (e.g. a model such as A(B(A(B(x)))))\\n add_unprocessed_node(layer, node_data)\",\n \"def __init__(self, stage, layer=None, proxyPrimPrefix=\\\"\\\", renderPrimPrefix=\\\"\\\"):\\n\\n self._stage = stage\\n\\n # Layer\\n if layer is None:\\n self._layer = Sdf.Layer.CreateAnonymous()\\n elif isinstance(layer, Sdf.Layer):\\n self._layer = layer\\n elif isinstance(layer, str):\\n self._layer = Sdf.Layer.Open(layer)\\n\\n self._layer.SetMuted(True)\\n self._stage.GetSessionLayer().subLayerPaths.append(self._layer.identifier)\\n self._proxyPrimPrefix = proxyPrimPrefix\\n self._renderPrimPrefix = renderPrimPrefix\",\n \"def __reslayer(self, inputs, in_filters, out_filters, stride=1):\\n with tf.variable_scope('sub1'):\\n kernel = tf.get_variable('weights', [3, 3, in_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n conv = tf.nn.conv2d(inputs, kernel, [1, stride, stride, 1],\\n padding='SAME',\\n name='conv')\\n batch_norm = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\\n conv = tf.nn.elu(batch_norm, 'elu')\\n\\n with tf.variable_scope('sub2'):\\n kernel = tf.get_variable('weights',\\n [3, 3, out_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n conv = tf.nn.conv2d(conv, kernel, [1, 1, 1, 1], padding='SAME',\\n name='conv1')\\n bias = self.__batch_norm_wrapper(conv, decay=0.9999, shape=[0, 1, 2])\\n\\n with tf.variable_scope('subadd'):\\n if in_filters != out_filters:\\n kernel = tf.get_variable('weights', [1, 1, in_filters, out_filters],\\n initializer=xavier_initializer(\\n dtype=tf.float32),\\n dtype=tf.float32)\\n inputs = tf.nn.conv2d(\\n inputs, kernel, [1, stride, stride, 1], padding='SAME')\\n bias += inputs\\n conv = tf.nn.elu(bias, 'elu')\\n\\n num = np.power(2, np.floor(np.log2(out_filters) / 2))\\n\\n grid = self.__put_activations_on_grid(conv, (int(num),\\n int(out_filters /\\n num)))\\n tf.summary.image('sub2/activations', grid, max_outputs=1)\\n\\n return conv\",\n \"def depth_rendering_gpu(ref_view, disparity_map, lf_size = (64, 512, 512, 3)):\\n lf_one_way = int(math.floor(math.sqrt(lf_size[0])))\\n\\n x_indices = np.arange(lf_size[1])\\n y_indices = np.arange(lf_size[2])\\n b_indices = np.arange(lf_size[0])\\n\\n #Create a grid of size lf_size[:3] consisting of the pixel co ordinates of each image\\n _, x, y = np.meshgrid(b_indices, x_indices, y_indices, indexing= 'ij')\\n\\n # Create a grid of size (lf_size[0], 2) consiting of the row, col lf positions\\n grid = np.meshgrid(np.arange(lf_one_way), np.arange(lf_one_way), indexing= 'ij')\\n stacked = np.stack(grid, 2)\\n positions = stacked.reshape(-1, 2)\\n\\n # Compute the distance from each lf position from the reference view\\n # Repeat the elements of this to match the size of the disparity map\\n ref_pos = np.array(\\n [lf_one_way // 2, lf_one_way // 2])\\n distance = (np.tile(ref_pos, (lf_size[0], 1)) - positions).T\\n dis_repeated = np.repeat(distance, lf_size[1] * lf_size[2], axis = 1)\\n dis_repeated = dis_repeated.reshape(2, lf_size[0], lf_size[1], lf_size[2])\\n\\n # Tile the disparity map so that there is one for each lf_position - lf_size[0]\\n tiled_map = np.tile(disparity_map, (lf_size[0], 1, 1))\\n\\n # Compute the shifted pixels\\n x_shifted = (x.astype(np.float32) - tiled_map * dis_repeated[0]).flatten()\\n y_shifted = (y.astype(np.float32) - tiled_map * dis_repeated[1]).flatten()\\n\\n #indices for linear interpolation in a square around the central point\\n x_low = np.floor(x_shifted).astype(int)\\n x_high = x_low + 1\\n\\n y_low = np.floor(y_shifted).astype(int)\\n y_high = y_low + 1\\n\\n #Place co-ordinates outside the image back into the image\\n x_low_clip = np.clip(x_low, 0, ref_view.shape[0] - 1)\\n x_high_clip = np.clip(x_high, 0, ref_view.shape[0] - 1)\\n y_low_clip = np.clip(y_low, 0, ref_view.shape[1] - 1)\\n y_high_clip = np.clip(y_high, 0, ref_view.shape[1] - 1)\\n\\n #Gather the interpolation points\\n interp_pts_1 = np.stack((x_low_clip, y_low_clip))\\n interp_pts_2 = np.stack((x_low_clip, y_high_clip))\\n interp_pts_3 = np.stack((x_high_clip, y_low_clip))\\n interp_pts_4 = np.stack((x_high_clip, y_high_clip))\\n\\n #Index into the images\\n desired_shape = lf_size\\n ref_view = torch.tensor(ref_view, dtype=torch.float32).cuda()\\n res_1 = torch_tensor_sample(ref_view, interp_pts_1, desired_shape)\\n res_2 = torch_tensor_sample(ref_view, interp_pts_2, desired_shape)\\n res_3 = torch_tensor_sample(ref_view, interp_pts_3, desired_shape)\\n res_4 = torch_tensor_sample(ref_view, interp_pts_4, desired_shape)\\n\\n #Compute interpolation weights\\n x_low_f = x_low.astype(np.float32)\\n d_x_low = 1.0 - (x_shifted.astype(np.float32) - x_low_f)\\n d_x_high = 1.0 - d_x_low\\n y_low_f = y_low.astype(np.float32)\\n d_y_low = 1.0 - (y_shifted.astype(np.float32) - y_low_f)\\n d_y_high = 1.0 - d_y_low\\n\\n w1 = torch.from_numpy(d_x_low * d_y_low)\\n w2 = torch.from_numpy(d_x_low * d_y_high)\\n w3 = torch.from_numpy(d_x_high * d_y_low)\\n w4 = torch.from_numpy(d_x_high * d_y_high)\\n\\n #THEY AGREE AT THIS POINT\\n weighted_1 = torch.mul(repeat_weights(w1, desired_shape).cuda(), res_1)\\n weighted_2 = torch.mul(repeat_weights(w2, desired_shape).cuda(), res_2)\\n weighted_3 = torch.mul(repeat_weights(w3, desired_shape).cuda(), res_3)\\n weighted_4 = torch.mul(repeat_weights(w4, desired_shape).cuda(), res_4)\\n\\n novel_view = torch.add(torch.add(weighted_1, weighted_2), weighted_3)\\n torch.add(novel_view, weighted_4, out=novel_view)\\n return novel_view\",\n \"def __call__(self, inputs, *args, **kwargs):\\n out = inputs\\n for i, layer in enumerate(self.layers):\\n if i == 0:\\n out = layer(out, *args, **kwargs)\\n else:\\n out = layer(out)\\n return out\",\n \"def add_convolution_layers(self):\\n with tf.variable_scope(\\\"conv_layer_1\\\"):\\n # our input to this layer is from word embeddings. which would be of\\n # shape (batch_size, number of words in the sentence, dimension of\\n # embedding). For us dimension of embedding is the number of input\\n # channels, self.out_channels_1 is the output channels.\\n\\n # First we have to define a filter. The shape of the filter would be [\\n # self.filter_size_1, dimension of embedding, self.out_channels_1]\\n filter1 = tf.get_variable(\\\"filter1\\\", dtype=tf.float32,\\n shape=[self.filter_size_1,\\n self.word_embeddings.shape[2],\\n self.out_channels_1])\\n conv_layer_1 = tf.nn.conv1d(self.word_embeddings, filters=filter1,\\n stride=self.strides_1, padding=\\\"SAME\\\")\\n conv_layer_1 = tf.nn.tanh(conv_layer_1)\\n # now add pooling layer\\n print(\\\"adding pooling to layer1\\\")\\n\\n if self.max_pooling:\\n print(\\\"Adding max pooling layer\\\")\\n pooling_layer = tf.layers.max_pooling1d\\n else:\\n print(\\\"Adding average pooling layer\\\")\\n pooling_layer = tf.layers.average_pooling1d\\n conv_layer_1 = pooling_layer(conv_layer_1,\\n pool_size=self.pool_size_1,\\n strides=self.strides_1,\\n padding=\\\"SAME\\\")\\n\\n with tf.variable_scope(\\\"conv_layer_2\\\"):\\n print(\\\"adding convolution layer 2\\\")\\n # Input to this layer is conv_layer_1 from above.\\n # get another filter now. Shape of this filter would be [\\n # self.filter_size_2, self.out_channels_1, self.out_channels_2]\\n filter2 = tf.get_variable(\\\"filter2\\\", dtype=tf.float32,\\n shape=[self.filter_size_2,\\n self.out_channels_1,\\n self.out_channels_2])\\n conv_layer_2 = tf.nn.conv1d(conv_layer_1, filters=filter2,\\n stride=self.strides_2, padding=\\\"SAME\\\")\\n conv_layer_2 = tf.nn.tanh(conv_layer_2)\\n print(\\\"adding pooling to layer2\\\")\\n conv_layer_2 = pooling_layer(conv_layer_2,\\n pool_size=self.pool_size_2,\\n strides=self.strides_2,\\n padding=\\\"SAME\\\")\\n\\n self.cnn_layer = tf.nn.dropout(conv_layer_2, self.dropout_holder)\\n self.word_reps = self.cnn_layer\\n if self.use_additional and self.hybrid:\\n self.cnn_layer = tf.concat([self.word_reps, self.word_embeddings_2], \\n axis=-1)\\n self.word_reps = self.cnn_layer\",\n \"def _init_layers(self) -> None:\\n self.self_attn = MultiheadAttention(**self.self_attn_cfg)\\n self.cross_attn = MultiheadAttention(**self.cross_attn_cfg)\\n self.embed_dims = self.self_attn.embed_dims\\n self.ffn = FFN(**self.ffn_cfg)\\n norms_list = [\\n build_norm_layer(self.norm_cfg, self.embed_dims)[1]\\n for _ in range(3)\\n ]\\n self.norms = ModuleList(norms_list)\",\n \"def _embed_composite_feature(self, features, embedding_layers):\\n embedding_list = []\\n for i in range(len(embedding_layers)):\\n embedding_list.append(embedding_layers[i](features[:, :, i]))\\n embedding = tf.add_n(embedding_list)\\n return embedding\",\n \"def _add_layer(self, layer_dict, layer_name, input_layers, merge_mode=None, share_params_with=None):\\n util.colorprint(layer_name, 'teal')\\n \\n layer_dict = dict(layer_dict)\\n util.colorprint(layer_dict, 'red')\\n \\n if share_params_with is not None:\\n print \\\"Warning: ignoring share_params_with\\\"\\n \\n layer_options = layer_dict[\\\"options\\\"]\\n layer=None\\n if layer_dict[\\\"type\\\"]==\\\"conv2d\\\":\\n #TODO: remove below\\n nb_filter, nb_row, nb_col = 3,3,3\\n layer = keras.layers.convolutional.Convolution2D(nb_filter, nb_row, nb_col, **layer_options)\\n elif layer_dict[\\\"type\\\"]==\\\"dense\\\":\\n dim = layer_dict[\\\"output_dim\\\"]\\n # del layer_options[\\\"output_dim\\\"]\\n layer = keras.layers.core.Dense(dim, **layer_options) \\n else:\\n print \\\"Ursol Major\\\"\\n RaiseError()\\n # TODO: one of the layers is a string\\n if isinstance(input_layers, list):\\n #this means that there is input from a loop to this layer\\n self.model.add_node(layer, name=layer_name, inputs=input_layers, merge_mode=merge_mode)\\n else:\\n self.model.add_node(layer, name=layer_name, input=input_layers)\\n\\n return layer_name\",\n \"def test_2d_local_convolution(batch_exp, input_height, input_width, in_channels, filters, drawer, is_first):\\n batch_size = 2 ** batch_exp\\n if is_first:\\n data_format = \\\"channels_first\\\"\\n else:\\n data_format = \\\"channels_last\\\"\\n kernel_height = drawer.draw(integers(min_value=1, max_value=input_height-1))\\n kernel_width = drawer.draw(integers(min_value=1, max_value=input_width-1))\\n\\n if data_format == \\\"channels_last\\\":\\n input_shape = (input_height, input_width, in_channels)\\n else:\\n input_shape = (in_channels, input_height, input_width)\\n\\n # Create keras model\\n keras_model = keras.Sequential()\\n keras_local = keras.layers.LocallyConnected2D(filters, (kernel_height, kernel_width),\\n use_bias=True,\\n bias_initializer=\\\"random_uniform\\\",\\n input_shape=input_shape,\\n data_format=data_format)\\n keras_model.add(keras_local)\\n\\n # Setup weights\\n weights = drawer.draw(arrays(shape=keras_local.kernel.shape.as_list(), dtype=np.float32, elements=floats(min_value=-1e6, max_value=1e6)))\\n keras_model.set_weights([weights])\\n\\n # Translate keras to PyTorch model\\n torch_model = translate.translate_layer(keras_local)[0].to(device)\\n\\n keras_input = np.random.uniform(-10, 10, (batch_size,) + input_shape).astype(np.float32)\\n keras_output = keras_model.predict(keras_input)\\n\\n torch_input = torch.Tensor(keras_input).to(device)\\n torch_output = torch_model(torch_input).cpu().data.numpy()\\n nan_fill = np.random.uniform(-5, 5)\\n assert (np.isnan(keras_output) == np.isnan(torch_output)).all()\\n keras_output[np.isnan(keras_output)] = nan_fill\\n torch_output[np.isnan(torch_output)] = nan_fill\\n assert np.isclose(keras_output, torch_output, atol=1e-3, rtol=1e-3).all()\",\n \"def __init__(self, config_data, in_dims, layer_num):\\n self.layer_num = layer_num\\n self.in_dims = in_dims\\n self.out_dims = in_dims[\\\"width\\\"] * in_dims[\\\"height\\\"] * in_dims[\\\"channels\\\"]\\n self.layer_type = config_data[\\\"type\\\"]\\n self.name = config_data[\\\"name\\\"]\\n self.params = []\",\n \"def multiply(self, layer):\\n pass\",\n \"def forward(self,bottom,top):\\n # imgmaps = self.batch_loader.mixup_gen()\\n # print(\\\"DataLayer forward!!\\\")\\n trainX, trainY = self.batch_loader.batch_imgs()\\n # print(\\\"trainX:\\\",trainX.shape)\\n # print(\\\"trainY:\\\",trainY.shape)\\n # print(\\\"trainY:\\\", trainY)\\n # print(\\\"top[0].data.shape:\\\",top[0].data.shape)\\n # print(\\\"top[1].data.shape:\\\", top[1].data.shape)\\n top[0].data[:, ...] = trainX\\n top[1].data[:, ...] = trainY\\n # print(\\\"DataLayer forward!!\\\")\",\n \"def apply_net_tiled(net,img):\\n\\n # borders = [8,20,20] ## border width within each patch that is thrown away after prediction\\n # patchshape_padded = [32,240,240] ## the size of the patch that we feed into the net. must be divisible by 8 or net fails.\\n # patchshape = [16,200,200] ## must be divisible by 8 to avoid artifacts.\\n # stride = [16,200,200] ## same as patchshape in this case\\n def f(n,m): return (ceil(n/m)*m)-n ## f(n,m) gives padding needed for n to be divisible by m\\n def g(n,m): return (floor(n/m)*m)-n ## f(n,m) gives un-padding needed for n to be divisible by m\\n\\n a,b,c = img.shape[1:]\\n q,r,s = f(a,8),f(b,8),f(c,8) ## calculate extra border needed for stride % 8 = 0\\n\\n ZPAD,YPAD,XPAD = 8,24,24\\n\\n img_padded = np.pad(img,[(0,0),(ZPAD,ZPAD+q),(YPAD,YPAD+r),(XPAD,XPAD+s)],mode='constant') ## pad for patch borders\\n output = np.zeros(img.shape)\\n\\n zs = np.r_[:a:16]\\n ys = np.r_[:b:200]\\n xs = np.r_[:c:200]\\n\\n for x,y,z in itertools.product(xs,ys,zs):\\n qe,re,se = min(z+16,a+q),min(y+200,b+r),min(x+200,c+s)\\n ae,be,ce = min(z+16,a),min(y+200,b),min(x+200,c)\\n patch = img_padded[:,z:qe+2*ZPAD,y:re+2*YPAD,x:se+2*XPAD]\\n patch = torch.from_numpy(patch).cuda().float()\\n patch = net(patch[None])[0,:,ZPAD:-ZPAD,YPAD:-YPAD,XPAD:-XPAD].detach().cpu().numpy()\\n output[:,z:ae,y:be,x:ce] = patch[:,:ae-z,:be-y,:ce-x]\\n\\n return output\",\n \"def layeredTexturePort(*args, annotation: Union[AnyStr, bool]=\\\"\\\", backgroundColor:\\n Union[List[float, float, float], bool]=None, defineTemplate: AnyStr=\\\"\\\",\\n docTag: Union[AnyStr, bool]=\\\"\\\", dragCallback: Script=None, dropCallback:\\n Script=None, enable: bool=True, enableBackground: bool=True,\\n enableKeyboardFocus: bool=True, exists: bool=True, fullPathName:\\n bool=True, height: Union[int, bool]=0, highlightColor: Union[List[float,\\n float, float], bool]=None, isObscured: bool=True, manage: bool=True,\\n noBackground: bool=True, node: name=None, numberOfPopupMenus: bool=True,\\n parent: Union[AnyStr, bool]=\\\"\\\", popupMenuArray: bool=True,\\n preventOverride: bool=True, selectedAlphaControl: AnyStr=\\\"\\\",\\n selectedBlendModeControl: AnyStr=\\\"\\\", selectedColorControl: AnyStr=\\\"\\\",\\n selectedIsVisibleControl: AnyStr=\\\"\\\", statusBarMessage: AnyStr=\\\"\\\",\\n useTemplate: AnyStr=\\\"\\\", visible: bool=True, visibleChangeCommand:\\n Union[Script, bool]=None, width: Union[int, bool]=0, q=True, query=True,\\n e=True, edit=True, **kwargs)->Union[AnyStr, Any]:\\n pass\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.6448038","0.6187745","0.6163194","0.5998497","0.5929416","0.5873209","0.56773156","0.5597431","0.5518665","0.5488085","0.5382779","0.53090936","0.53070503","0.52928275","0.5291836","0.5291313","0.5284668","0.5253536","0.5242199","0.523257","0.51847124","0.5180517","0.51741755","0.51741755","0.51732266","0.5167482","0.5118782","0.51178277","0.5111082","0.51026183","0.51008844","0.50771827","0.50711435","0.50676274","0.5054252","0.5045026","0.5039009","0.5035538","0.50333977","0.5026929","0.50123966","0.5009639","0.5000623","0.49903697","0.49874657","0.49824077","0.49795514","0.4976205","0.49664342","0.49619997","0.4959416","0.4958737","0.49574858","0.49445528","0.4934589","0.49344352","0.49250442","0.49227726","0.49037787","0.48979825","0.4891321","0.4888596","0.48880768","0.48879242","0.48799154","0.487743","0.48731208","0.48723292","0.4866994","0.48663637","0.4866091","0.48644984","0.4859447","0.48593128","0.48487973","0.4842857","0.4840985","0.48402968","0.48269343","0.4824775","0.48236597","0.48159477","0.4796426","0.47948","0.4794728","0.47944185","0.4786368","0.47800967","0.477725","0.47771776","0.47703788","0.47682023","0.4764765","0.47631317","0.47626153","0.4761117","0.47603813","0.47597635","0.47540298","0.47537652","0.47516534"],"string":"[\n \"0.6448038\",\n \"0.6187745\",\n \"0.6163194\",\n \"0.5998497\",\n \"0.5929416\",\n \"0.5873209\",\n \"0.56773156\",\n \"0.5597431\",\n \"0.5518665\",\n \"0.5488085\",\n \"0.5382779\",\n \"0.53090936\",\n \"0.53070503\",\n \"0.52928275\",\n \"0.5291836\",\n \"0.5291313\",\n \"0.5284668\",\n \"0.5253536\",\n \"0.5242199\",\n \"0.523257\",\n \"0.51847124\",\n \"0.5180517\",\n \"0.51741755\",\n \"0.51741755\",\n \"0.51732266\",\n \"0.5167482\",\n \"0.5118782\",\n \"0.51178277\",\n \"0.5111082\",\n \"0.51026183\",\n \"0.51008844\",\n \"0.50771827\",\n \"0.50711435\",\n \"0.50676274\",\n \"0.5054252\",\n \"0.5045026\",\n \"0.5039009\",\n \"0.5035538\",\n \"0.50333977\",\n \"0.5026929\",\n \"0.50123966\",\n \"0.5009639\",\n \"0.5000623\",\n \"0.49903697\",\n \"0.49874657\",\n \"0.49824077\",\n \"0.49795514\",\n \"0.4976205\",\n \"0.49664342\",\n \"0.49619997\",\n \"0.4959416\",\n \"0.4958737\",\n \"0.49574858\",\n \"0.49445528\",\n \"0.4934589\",\n \"0.49344352\",\n \"0.49250442\",\n \"0.49227726\",\n \"0.49037787\",\n \"0.48979825\",\n \"0.4891321\",\n \"0.4888596\",\n \"0.48880768\",\n \"0.48879242\",\n \"0.48799154\",\n \"0.487743\",\n \"0.48731208\",\n \"0.48723292\",\n \"0.4866994\",\n \"0.48663637\",\n \"0.4866091\",\n \"0.48644984\",\n \"0.4859447\",\n \"0.48593128\",\n \"0.48487973\",\n \"0.4842857\",\n \"0.4840985\",\n \"0.48402968\",\n \"0.48269343\",\n \"0.4824775\",\n \"0.48236597\",\n \"0.48159477\",\n \"0.4796426\",\n \"0.47948\",\n \"0.4794728\",\n \"0.47944185\",\n \"0.4786368\",\n \"0.47800967\",\n \"0.477725\",\n \"0.47771776\",\n \"0.47703788\",\n \"0.47682023\",\n \"0.4764765\",\n \"0.47631317\",\n \"0.47626153\",\n \"0.4761117\",\n \"0.47603813\",\n \"0.47597635\",\n \"0.47540298\",\n \"0.47537652\",\n \"0.47516534\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90883,"cells":{"query":{"kind":"string","value":"Creates an upper confidence tree by simulating N games starting from the given state in order to select the move leading to the best results"},"document":{"kind":"string","value":"def get_uct(self, state):\n root = ut.UCTree(state)\n for _ in range(self.N):\n moves = []\n leaf = root.select(moves, state.courant)\n win = leaf.simulate(moves)\n root.back_propagate(leaf, moves, win)\n return root"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def __call__(self, s, n=1000):\n\n root = StateNode(None, s, self.game)\n \n if root.parent is not None:\n raise ValueError(\"Root's parent must be None.\")\n \n for _ in range(n):\n #selection\n node = _get_next_node(root, self.tree_policy)\n #simulation\n node.reward = self.default_policy(node)\n #print(node.reward)\n #back\n self.backup(node)\n \n root.reset(copy.deepcopy(self.game_bak))\n \n #for i in root.children:\n # print(root.children[i].__dict__)\n # for j in root.children[i].children:\n # print(root.children[i].children[j].__dict__)\n # print(\"=======\")\n return rand_max(root.children.values(), key=lambda x: x.q).action, rand_max(root.children.values(), key=lambda x: x.q).q","def run_simulation(self, state):\n \"*** YOUR CODE HERE ***\"\n player = 0\n visited_states = [(player, state)]\n depth_limited = self.depth != -1\n depth = self.depth\n expand = True\n while not visited_states[-1][1].isWin() and not visited_states[-1][1].isLose():\n if depth_limited and depth == 0: break\n state = self.UCB1(state, player) # Selection & Simulation\n if expand and state not in self.plays: # Expansion\n expand = False\n self.plays[state] = 0\n self.wins[state] = 0\n visited_states.append((player, state))\n player = (player + 1) % state.getNumAgents()\n if not expand and depth_limited and player == 0: depth -= 1\n \n for player, state in visited_states:\n if state in self.plays: # Not simulated nodes\n self.plays[state] += 1\n eval = self.evaluationFunction(visited_states[-1][1])\n if depth_limited:\n if player == 0: self.wins[state] += eval\n if player != 0: self.wins[state] -= eval\n else:\n if player == 0: self.wins[state] += eval\n if player != 0: self.wins[state] += (1 - eval)","def find_best_move(state) -> None:\n C: float = 2 ** 0.5 # Start with sqrt(2)\n idx: int = -1\n win_ratio: float = 0\n\n # First create the root node for the game\n root = State(None, None, state)\n root = expand(root)\n root = run_each_move(root, state)\n print(\"New Move UCBs: \")\n for m in range(len(root.move_ucbs)):\n print(\" \", root.move_ucbs[m])\n \n #while True:\n for _ in range(10000):\n _, win_ratio = MCTS(root, state, True, win_ratio) \n #print(\"\\n\\n\\nRoot Attempts: \", root.attempts)\n #print(\"Back in main\\n\\n\\n\")\n \n # Pick one of the highest ranked by UCB's","def chooseAction(self, epsilon, state):\n if random.uniform(0, 1) < epsilon:\n return random.randrange(9)\n\n cur_best_val = -float('inf')\n cur_best_action = 0\n\n data = env.getAllNextStates(state)\n\n with torch.no_grad():\n for action, next_state, done in data:\n if next_state != state:\n value = self.NN(self.RBF[next_state]).item() if not done else 0\n if value > cur_best_val:\n cur_best_val = value\n cur_best_action = action\n #print(data)\n return cur_best_action","def MCTS(parent: State, state: GameState, root: int, win_ratio: float) -> int:\n #print(\"\\nPrior Move UCBs: \")\n #for m in range(len(parent.move_ucbs)):\n # print(\" \", parent.move_ucbs[m])\n # Find the maximum UCB value to explore\n idx = parent.move_ucbs.index(max(parent.move_ucbs))\n # TODO: May want to make this a random selection instead of first\n # Select the child with the highest UCB value\n child = parent.moves[idx]\n #print(\"Child Selected: \", child.idx)\n # Check if the child has moves expanded already\n if child.attempts == 0 or len(child.moves) == 0: # Expand if no moves\n child = expand(child)\n # Pick child from expanded set\n sim_idx = random.randint(0, len(child.moves)-1)\n simulate_child = child.moves[sim_idx]\n # Perform Simulation\n outcome = simulate(simulate_child.board)\n else: # Recurse if there are moves\n #print(\"Recursing.................................\")\n outcome, _ = MCTS(child, state, False, 0) \n # Update the wins/attempts\n #print(\"Outcome: \", outcome)\n #print(\"Player: \", state.player)\n if outcome == state.player:\n #print(\"********** Winner **********\")\n #print(\"Index in list for ucb: \", idx)\n child.winner()\n if root:\n win_ratio = update_selected(state, child, win_ratio)\n else:\n #print(\"********** Loser **********\")\n #print(\"Index in list for ucb: \", idx)\n child.loser()\n parent.attempts += 1\n update_ucbs(parent)\n #print(\"New Move UCBs: \")\n #for m in range(len(parent.move_ucbs)):\n # print(\" \", parent.move_ucbs[m])\n \n\n return outcome, win_ratio","def ismcts(rootstate, itermax, verbose=False, belief=[]):\n\tprint(belief)\n\trootnode = Node()\n\n\tfor i in range(itermax):\n\t\tnode = rootnode\n\n\t\t# Determinize\n\t\tstate = rootstate.clone_and_randomize(rootstate.player_to_move)\n\n\t\t# Select\n\t\twhile state.get_moves() != [] and node.get_untried_moves(state.get_moves()) == []:\n\t\t\t# node is fully expanded and non-terminal\n\t\t\tnode = node.ucb_select_child(state.get_moves())\n\t\t\tstate.do_move(node.move)\n\n\t\t# Expand\n\t\tuntried_moves = node.get_untried_moves(state.get_moves())\n\t\tif untried_moves: # if we can expand (i.e. state/node is non-terminal)\n\t\t\tm = random.choice(untried_moves)\n\t\t\tplayer = state.player_to_move\n\t\t\tstate.do_move(m)\n\t\t\tnode = node.add_child(m, player) # add child and descend tree\n\n\t\t# Simulate\n\t\twhile state.get_moves(): # while state is non-terminal\n\t\t\tstate.do_move(random.choice(state.get_moves()))\n\n\t\t# Backpropagate\n\t\twhile node is not None: # backpropagate from the expanded node and work back to the root node\n\t\t\tnode.update(state)\n\t\t\tnode = node.parentNode\n\n\t# Output some information about the tree - can be omitted\n\tif verbose:\n\t\tprint(rootnode.tree_to_string(0))\n\telse:\n\t\tprint(rootnode.children_to_string())\n\n\treturn max(rootnode.child_nodes, key=lambda c: c.visits).move # return the move that was most visited","def mcts_search(self, state):\n assert state.current_player() == self.player\n root = SearchNode(None, 1)\n for _ in range(self.max_simulations):\n visit_path, working_state = self._apply_tree_policy(root, state)\n if working_state.is_terminal():\n node_value = working_state.player_return(self.player)\n else:\n node_value = self.evaluator.evaluate(\n working_state, self.player, self._random_state)\n\n for node in visit_path:\n node.total_reward += node_value * node.player_sign\n node.explore_count += 1\n\n most_visited = root.most_visited_child()\n\n if self.verbose:\n print(\"Root:\", root.to_str())\n print(\"Children:\")\n print(root.children_str(working_state))\n print(\"Children of chosen:\")\n chosen_state = state.clone()\n chosen_state.apply_action(most_visited.action)\n print(most_visited.children_str(chosen_state))\n\n return most_visited.action","def test_state(\n size: Union[int, tuple],\n num_berries: int,\n number_steps: int,\n state_sizes: List[int] = [3, 5],\n) -> None:\n for state_size in state_sizes:\n game = Game(\n size,\n [0, 0],\n -1,\n 5,\n -5,\n 10,\n num_berries,\n berry_movement_probabilities=[0.5] * num_berries,\n state_size=state_size,\n )\n done = False\n i = 1\n print(f\"Beginning full board\\n{game.get_state(full=True)}\")\n print(f\"And the state\\n{game.get_state(state_size)}\")\n while not done and i < number_steps:\n action = random.choice(MOVEMENTS)\n print(f\"Action taken {action}\")\n state, reward, done = game.step(action)\n print(f\"Full board\\n{game.get_state(full=True)}\")\n print(f\"The state\\n{game.get_state(state_size)}\")\n i += 1","def move(self, state):\n result = None\n self.currentDepthLimit = 0\t\n\tself.transposition = {}\n\tself.counter = 0\n\n\twhile True:\n u = float(\"inf\")\n\t v = float(\"-inf\")\n\t self.counter = 0\n\t result = None\n\t self.transposition = {}\n\t for a in state.actions():\n new = self.min_value(state.result(a), float(\"-inf\"), float(\"inf\"),self.currentDepthLimit)\n\t if new > v:\n\t v = new\n\t result = a\n\n\t elif new == v:\n\t if a.index < result.index:\n\t result = a\n\t if self.is_time_up():\n\t return result\n\t \n\t self.currentDepthLimit += 1\n\t \"\"\"If we never use evaluate function, it means all state are terminated, so return whatever the result is\"\"\"\n\t if self.counter == 0:\n\t break\n\t if self.is_time_up():\n \t return result\n\treturn result","def play_game(start_state=GoState(),\n policy_value=NaivePolicyValue(),\n max_game_length=1e99,\n opponent=None,\n n_simulations=N_SIMULATIONS):\n\n # TODO: This will set .move_number = 0, should maybe track whose turn it is instead:\n tree_root = TreeStructure(start_state)\n policy, value = policy_value.predict(tree_root.state)\n tree_root.prior_policy = policy[tree_root.state.valid_actions]\n tree_root.add_noise_to_prior()\n game_history = []\n\n while not tree_root.state.game_over and tree_root.move_number < max_game_length:\n\n mcts(tree_root, policy_value, n_simulations)\n\n # print_tree(tree_root,0)\n # Store the state and distribution before we prune the tree:\n # TODO: Refactor this\n\n game_history.append(tree_root.history_sample())\n\n choice = choice_to_play(tree_root, bool(opponent))\n tree_root = tree_root.children[choice]\n tree_root.parent = None\n tree_root.add_noise_to_prior()\n\n if opponent:\n game_history.append(tree_root.history_sample())\n choice = opponent(tree_root.state)\n if choice in tree_root.children:\n tree_root = tree_root.children[choice]\n else:\n new_state = step(tree_root.state, choice)\n tree_root = TreeStructure(new_state)\n policy, value = policy_value.predict(tree_root.state)\n tree_root.prior_policy = policy[tree_root.state.valid_actions]\n tree_root.parent = None\n\n if tree_root.move_number >= max_game_length:\n tree_root.state.winner = tree_root.state._compute_winner()\n\n return game_history, tree_root.state.winner","def gen_game_tree(state_init):\n current_path = [state_init]\n game_tree = {}\n while current_path:\n cur_state = current_path[-1]\n if cur_state not in game_tree:\n ttt = TicTacToe(cur_state)\n game_tree[cur_state] = {\n \"unexplored\": ttt.next_states,\n \"explored\": [],\n \"status\": ttt.game_status,\n }\n\n if game_tree[cur_state][\"unexplored\"]:\n current_path.append(game_tree[cur_state][\"unexplored\"].pop(0))\n else:\n explored = current_path.pop(-1)\n if explored != state_init:\n game_tree[current_path[-1]][\"explored\"].append(explored)\n\n status, player = game_tree[cur_state][\"status\"]\n if status == \"tie\":\n value = 0\n outcomes = {0: 1}\n elif status == \"win\":\n value = -1 if player == \"1\" else 1\n outcomes = {value: 1}\n else:\n value = (min if player == \"1\" else max)(\n [\n game_tree[state][\"value\"]\n for state in game_tree[cur_state][\"explored\"]\n ]\n )\n outcomes = {}\n for state in game_tree[cur_state][\"explored\"]:\n for res, res_ct in game_tree[state][\"outcomes\"].items():\n outcomes[res] = outcomes.get(res, 0) + res_ct\n\n game_tree[cur_state][\"value\"] = value\n game_tree[cur_state][\"outcomes\"] = outcomes\n\n return game_tree","def optimize(seqlen, trees, nstates=2):\n num_params = nstates * (nstates + 1)\n guess = np.log(np.random.rand(num_params))\n # guess = np.log(np.array([0.5, 0.5, 0.13, 0.37, 0.37, 0.13,\n # 0.32, 0.18, 0.18, 0.32, 0.95, 0.05, 0.05, 0.95]))\n i, t, s = unfold_params(guess, nstates=nstates)\n i = np.log(np.ones(nstates) / nstates)\n i, t = normalize(i), normalize(t)\n\n i = np.log(np.array([0.5, 0.5]))\n # t = np.log(np.array([\n # [1 - 0.01, 0.01],\n # [0.01, 1 - 0.01]\n # ]))\n t = np.log(np.array([\n [0.5, 0.5],\n [0.5, 0.5]\n ]))\n s = 1.0 / trees[0].tot_branch_len()\n\n guess = np.concatenate((i, np.ndarray.flatten(t), [1]))\n res = minimize(getProb, guess, args=(nstates, seqlen, trees), method=\"L-BFGS-B\",\n options={\"maxiter\": 250, \"disp\": True},\n bounds=(([(np.NINF, 0)] * num_params) + [(0, np.Inf)]))\n i, t, s = unfold_params(res.x, nstates=nstates)\n i, t = normalize(i), normalize(t)\n print(forward(seqlen, t, i, (trees * s)))\n return np.exp(i), np.exp(t), s","def dfs_maximizing(state) :\n #print state.describe_previous_move()\n global state_evals, path, _path, _score, level, _state;\n\n level+=1\n path.append(state)\n for stt in state.generate_next_states():\n score=0\n agenda.append((stt, level))\n \n if stt.is_game_over():\n state_evals+=1\n score=stt.get_endgame_score()\n if score>_score:\n _score=score\n _path = path[0:]\n _state = stt\n if not agenda:\n\n _path.append(_state)\n return [_path, _score, state_evals];\n else:\n new_state, level=agenda.pop()\n path=path[0:level]\n level-=1\n return dfs_maximizing(new_state)","def alphabeta_search(state):\r\n \r\n '''\r\n Terminates when game.actions is empty\r\n Class Game needs the following functions:\r\n - game.result(state, a) -- successor\r\n - game.actions(state) -- possible moves\r\n - game.utility -- returns the state of the game (win/lose or tie, when game is terminal)\r\n \r\n '''\r\n #sort state.actions in increasing or decreasing based on max or min (alpha or beta)\r\n #use heuristics fn to get a value for each move (move is in format (x,y) where x and y are ints\r\n \r\n d = depthset[0] #this is the cutoff test depth value. if we exceed this value, stop\r\n cutoff_test=None\r\n sort_fn = [vitalpoint, eyeHeur]\r\n eval_fn = survivalheur \r\n #randnumheuristics \r\n player = state.to_move()\r\n prune = 0\r\n pruned = {} #this will store the depth of the prune\r\n totaldepth = [0]\r\n visited = {}\r\n heuristicInd = 0\r\n \r\n def max_value(state, alpha, beta, depth, heuristicInd):\r\n branches = len(state.actions())\r\n onbranch = 0\r\n \r\n if totaldepth[0] < depth:\r\n totaldepth[0] = depth\r\n if cutoff_test(state, depth):\r\n return eval_fn(state)\r\n v = -infinity\r\n \r\n #sort state.actions based on heuristics before calling\r\n #max wants decreasing\r\n #sorted(state.actions(), key = eval_sort, reverse = True)\r\n \r\n #sort by favorites first, returns a list of actions\r\n # for sorts in sort_fn:\r\n tempher = heuristicInd\r\n\r\n sorts = sort_fn[heuristicInd]\r\n sortedactions, heuristicInd = sorts(state)\r\n #if heuristicInd != tempher:\r\n # print 's',\r\n ''''''\r\n for a in sortedactions:\r\n if visited.get(depth) == None:\r\n visited[depth] = [a]\r\n else:\r\n visited[depth].append(a)\r\n \r\n onbranch += 1\r\n v = max(v, min_value(state.result(a),\r\n alpha, beta, depth+1, heuristicInd)) #+ vitscore.count(a)\r\n if v >= beta: #pruning happens here, but in branches\r\n if pruned.get(depth) == None:\r\n pruned[depth] = branches - onbranch\r\n else:\r\n pruned[depth] += (branches - onbranch)\r\n #print \"prune\", depth, \" \", state.actions()\r\n #state.display()\r\n return v\r\n alpha = max(alpha, v)\r\n \r\n #print depth, \" \", state.actions()\r\n #state.display()\r\n \r\n return v\r\n\r\n def min_value(state, alpha, beta, depth, heuristicInd):\r\n branches = len(state.actions())\r\n onbranch = 0\r\n \r\n if totaldepth[0] < depth:\r\n totaldepth[0] = depth\r\n if cutoff_test(state, depth):\r\n return eval_fn(state)\r\n v = infinity\r\n \r\n #sort state.actions based on heuristics before calling\r\n #min wants increasing\r\n #sorted(state.actions(), key = eval_sort)\r\n #Shayne\r\n tempher = heuristicInd\r\n sorts = sort_fn[heuristicInd]\r\n sortedactions, heuristicInd = sorts(state, 1)\r\n #if heuristicInd != tempher:\r\n # print 's',\r\n for a in sortedactions: #state.actions():\r\n onbranch += 1\r\n if visited.get(depth) == None:\r\n visited[depth] = [a]\r\n else:\r\n visited[depth].append(a)\r\n v = min(v, max_value(state.result(a),\r\n alpha, beta, depth+1, heuristicInd))\r\n if v <= alpha: #pruning happens here, but in branches\r\n if pruned.get(depth) == None:\r\n pruned[depth] = branches - onbranch\r\n else:\r\n pruned[depth] += (branches - onbranch)\r\n #print \"prune\", depth, \" \", state.actions()\r\n #state.display()\r\n return v\r\n beta = min(beta, v)\r\n #print depth, \" \", state.actions()\r\n #state.display()\r\n return v\r\n\r\n # Body of alphabeta_search starts here:\r\n #def cutoff_test and eval_fn \r\n cutoff_test = (cutoff_test or\r\n (lambda state,depth: depth>d or state.terminal_test()))\r\n eval_fn = eval_fn or (lambda state: state.utility(player))\r\n #by default, utility score is used\r\n \r\n \r\n #argmax goes through all the possible actions and \r\n # applies the alphabeta search onto all of them\r\n # and returns the move with the best score \r\n #print state.actions()\r\n heuristicInd = 0\r\n sorts = sort_fn[heuristicInd]\r\n sortedact, heuristicInd = sorts(state)\r\n abmove = argmax(sortedact,\r\n lambda a: min_value(state.result(a),\r\n -infinity, infinity, 0, heuristicInd))\r\n\r\n print 'problem,', problemno[0], ', total tree depth,', totaldepth[0]\r\n for i in range(1, len(visited)):\r\n if len(pruned) < i:\r\n pruned[i] = 0\r\n print i, \",\", len(visited[i]), \",\", pruned[i]\r\n \r\n return abmove","def simulate(state: GameState) -> int:\n moves = list(state.moves)\n #print(\" moves available: \", moves)\n for i in range(len(state.moves)):\n move = random.choice(moves)\n #print(\" move making: \", move)\n move_idx = moves.index(move)\n #print(\" index of move: \", move_idx)\n moves.pop(move_idx)\n #print(\" new moves available: \", moves)\n state = state.traverse(move)\n #print(\" Winner: \", state.util)\n #print(\" New Board: \", state.display)\n return state.util","def ai_strategy3(gstate: TicTacToe, game_tree):\n status, player = gstate.game_status\n\n if status != \"turn\":\n logging.warning(\"Game status = %s. No move needed.\", status)\n return None\n\n win, lose = (-1, 1) if player == \"1\" else (1, -1)\n\n move_vals = {\n move: win * game_tree[state][\"value\"] for move, state in gstate.next_moves.items()\n }\n max_val = max(move_vals.values())\n\n if max_val == 1:\n # ai_strategy2 can handle \"won\" states\n return ai_strategy2(gstate, game_tree)\n\n else:\n ok_moves = [move for move, val in move_vals.items() if val == max_val]\n\n move_vals = {\n move: collections.Counter(\n [\n game_tree[state2][\"value\"] for state2 in game_tree[gstate.next_moves[move]][\"explored\"]\n ]\n )\n for move in ok_moves\n }\n\n move_eval = {\n move: val_ctr.get(win, 0) / val_ctr.get(lose, 0.5)\n for move, val_ctr in move_vals.items()\n }\n max_win_pct = max(move_eval.values())\n good_moves = [move for move, win_pct in move_eval.items() if win_pct ==\n max_win_pct]\n\n move = random.choice(good_moves)\n\n all_ct = sum(move_vals[move].values())\n win_ct = move_vals[move].get(win, 0)\n logging.debug(\n \"move: %i; value: %i, win %%: %.1f%%, moves: %s\\n\",\n move, win * game_tree[gstate.state][\"value\"],\n win_ct / max(all_ct, 0.1) * 100, move_vals\n )\n\n return move","def decide(self, state: OthelloState, actions: list):\n # -------- TASK 2 ------------------------------------------------------\n # Your task is to implement an algorithm to choose an action form the\n # given `actions` list. You can implement any algorithm you want.\n # However, you should keep in mind that the execution time of this\n # function is limited. So, instead of choosing just one action, you can\n # generate a sequence of increasing good action.\n # This function is a generator. So, you should use `yield` statement\n # rather than `return` statement. To find more information about\n # generator functions, you can take a look at:\n # https://www.geeksforgeeks.org/generators-in-python/\n #\n # If you generate multiple actions, the last action will be used in the\n # game.\n #\n # Tips\n # ====\n # 1. During development of your algorithm, you may want to find the next\n # state after applying an action to the current state; in this case,\n # you can use the following patterns:\n # `next_state = current_state.successor(action)`\n #\n # 2. If you need to simulate a game from a specific state to find the\n # the winner, you can use the following pattern:\n # ```\n # simulator = Game(FirstAgent(), SecondAgent())\n # winner = simulator.play(starting_state=specified_state)\n # ```\n # The `MarkovAgent` has illustrated a concrete example of this\n # pattern.\n #\n # 3. You are free to choose what kind of game-playing agent you\n # implement. Some of the obvious approaches are the following:\n # 3.1 Implement alpha-beta (and investigate its potential for searching deeper\n # than what is possible with Minimax). Also, the order in which the actions\n # are tried in a given node impacts the effectiveness of alpha-beta: you could\n # investigate different ways of ordering the actions/successor states.\n # 3.2 Try out better heuristics, e.g. ones that take into account the higher\n # importance of edge and corner cells. Find material on this in the Internet.\n # 3.3 You could try out more advanced Monte Carlo search methods (however, we do\n # not know whether MCTS is competitive because of the high cost of the full\n # gameplays.)\n # 3.4 You could of course try something completely different if you are willing to\n # invest more time.\n #\n # GL HF :)\n # ----------------------------------------------------------------------\n\n # Replace the following lines with your algorithm\n best_action = actions[0]\n yield best_action","def get_move(self, state):\n for n in range(self._n_playout):\n state_copy = copy.deepcopy(state)\n self._playout(state_copy)\n print(n)\n print(\"Judge done\")\n return max(self._root._children.items(),\n key=lambda act_node: act_node[1]._n_visits)[0]","def policies(self, QTable, epsilon, state, next_states, action_to_do): # Inspiration from https://www.geeksforgeeks.org/q-learning-in-python/?fbclid=IwAR1UXR88IuJBhhTakjxNq_gcf3nCmJB0puuoA46J8mZnEan_qx9hhoFzhK8\r\n num_actions = 5 # 5 actions-value, [moved_out, into_goal, send_opp_home, send_self_home, move_token] \r\n def epsilonGreedyPolicy(): \r\n tmp_state = str(state.state[0])\r\n valid_actions = np.append(action_to_do, True) # the True appended is move_token\r\n valid_act_len = len(np.where(valid_actions==True)[0])\r\n\r\n Action_probabilities = np.ones(num_actions, dtype = float) * epsilon / valid_act_len # divides probability based on number of valid actions and epsilon (each 0.025 if 4 actions) \r\n Action_probabilities = np.multiply(Action_probabilities, valid_actions)\r\n\r\n # If same values in QTable choose random valid action \r\n best_action = np.argmax(QTable[tmp_state]) # Find index of action which gives highest QValue\r\n # Check if valid action else find new best action\r\n if not valid_actions[best_action]:\r\n actions = np.argsort(-QTable[tmp_state]) # descending order of action values\r\n for i in range(len(valid_actions)):\r\n if valid_actions[actions[i]]:\r\n best_action = actions[i]\r\n break\r\n\r\n Action_probabilities[best_action] += (1.0 - epsilon) # Assigns rest probability to best action so probability sums to 1\r\n\r\n return Action_probabilities \r\n\r\n def greedyPolicy():\r\n tmp_state = str(state.state[0])\r\n valid_actions = np.append(action_to_do, True) # the True appended is move_token\r\n\r\n Action_probabilities = np.zeros(num_actions, dtype = float)\r\n\r\n best_action = np.argmax(QTable[tmp_state]) # Find index of action which gives highest QValue\r\n # Check if valid action else find new best action\r\n if not valid_actions[best_action]:\r\n actions = np.argsort(-QTable[tmp_state]) # descending order of action values\r\n for i in range(len(valid_actions)):\r\n if valid_actions[actions[i]]:\r\n best_action = actions[i]\r\n break\r\n\r\n\r\n Action_probabilities[best_action] += 1.0\r\n return Action_probabilities\r\n\r\n\r\n if(self.__chosenPolicy == \"epsilon greedy\"):\r\n return epsilonGreedyPolicy \r\n if(self.__chosenPolicy == \"greedy\"):\r\n return greedyPolicy","def next_state_generation(state, target, operators):\n current_swap_state = ''\n current_swap_res = sys.maxsize\n current_change_state = ''\n current_change_res = sys.maxsize\n # for swapping\n for i in range(0, len(state), 2):\n for j in range(2, len(state), 2):\n new_swap_state, swap_res = swap(i, j, state, target)\n if current_swap_res > swap_res:\n current_swap_res = swap_res\n current_swap_state = new_swap_state\n # for changing\n for i in range(1, len(state) - 1, 2):\n new_change_state, change_res = change(i, random.choice(operators), state, target)\n if current_change_res > change_res:\n current_change_res = change_res\n current_change_state = new_change_state\n # return the lowest of the 2 value and state amongst swapping and changing operations\n print(\"Swap res:\", current_swap_res)\n print(\"Change res:\", current_change_res)\n if current_swap_res < current_change_res:\n print(\"Best State \", current_swap_state)\n print(\"Distance from target: \", current_swap_res)\n print()\n return current_swap_state, current_swap_res\n else:\n print(\"Best State \", current_change_state)\n print(\"Distance from target: \", current_change_res)\n print()\n return current_change_state, current_change_res","def select_move(self, game_state, return_visit_counts=False):\n \n # Start with a tree consisting of a root node only. The root node\n # is associated with the given board position.\n root = self.create_node(game_state)\n \n # If no legal moves can be made from the given board position, pass \n # the turn. This happens when all of the players pieces are surrounded,\n # if the player has no pieces left or if the game is over. \n if not root.branches:\n if return_visit_counts:\n return Act.pass_turn(), {}\n return Act.pass_turn()\n \n for i in range(self.num_rounds):\n # On each iteration, walk down the tree to a leaf node and select\n # a move to make from the corresponding leaf game state.\n node = root\n next_move = self.select_branch(node)\n while node.has_child(next_move):\n node = node.get_child(next_move)\n next_move = self.select_branch(node)\n \n # Create a new tree node for the selected move and add it to\n # the tree. If the leaf node corresponds to a finished game\n # then don't create a new node and assign a value to the node\n # based on who won.\n if node.state.is_not_over():\n if next_move:\n new_state = copy.deepcopy(node.state)\n new_state.take_turn_with_no_checks(Act.play(next_move))\n child_node = self.create_node(new_state, \n move=next_move, parent=node)\n move = next_move\n value = -1 * child_node.value \n else:\n # If the current player can't make any moves from the\n # selected gamestate then next_move will be 'None' meaning\n # the player passes the turn.\n new_state = copy.deepcopy(node.state)\n new_state.take_turn_with_no_checks(Act.pass_turn())\n child_node = self.create_node(new_state, \n move=next_move, parent=node)\n move = next_move\n value = -1 * child_node.value\n else:\n # If the game in the current state is over, then the last\n # player must have won the game. Thus the value/reward for the\n # other player is 1. The current node is not updated with\n # the new reward as no branches can stem from a finished game\n # state.\n move = node.last_move\n node = node.parent\n value = 1\n \n # Update the nodes traversed to get to the leaf node with the \n # new value for the new move.\n while node is not None:\n node.record_visit(move, value)\n move = node.last_move\n node = node.parent\n value *= -1\n \n # Get the visit counts of the branches if they were requested.\n if return_visit_counts:\n visit_counts = {}\n for move in root.branches.keys():\n visit_counts[move] = root.branches[move].visit_count\n \n # Get a list of possible moves sorted according to visit count,\n # the move with the highest visit count should be first in the list.\n moves = [move for move in root.moves()]\n moves = sorted(moves, key=root.visit_count, reverse=True)\n \n # Loop through the sorted moves and return the first legal one.\n for move in moves:\n if not game_state.is_move_illegal(move):\n if return_visit_counts:\n return Act.play(move), visit_counts\n return Act.play(move)\n \n # If no legal move is found then pass the turn.\n if return_visit_counts:\n return Act.pass_turn(), visit_counts\n return Act.pass_turn()","def greedy(self, state, timestep, epsilon=0):\n\n counts = np.bincount(self.call_locs, minlength=self.num_nodes)\n # print(self.lengths)\n # print(counts)\n score = self.lengths @ counts\n action = []\n for _ in range(self.num_ambulance):\n node = np.argmin(score)\n action.append(node)\n score[node] = 99999999\n return action","def move(self, state):\n \n self.depth_limit=1\n self.best_utility=-2\n action=None\n while not self.is_time_up():\n self.terminal=True\n self.cache={}\n action=self.alpha_beta_search(state,0)\n if self.terminal==True:\n break\n self.depth_limit=self.depth_limit+1\n \n return action","def marcovNuc (i = random.choice(stateSpace), step = 100):\n # matrix of transition probabilities\n #matrix = [[0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25]] \n matrix = [[0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1]] \n step += 1 # add one to the range because we remove it at the end\n sims = [] # List to hold the results of the Marcov chain\n sims.append(i) # append the seed value to the sims list\n for x in range(step):\n \n if sims[-1] == 'A':\n w = np.random.random() # Random number generator\n # the next set of if statements determine where the random number \n # sits on the number line of probabilities\n if matrix[0][0] > w:\n sims.append('A')\n elif matrix[0][1] + matrix[0][0] > w:\n sims.append('C')\n elif matrix[0][2] + matrix[0][1] + matrix[0][0] > w:\n sims.append('G')\n else:\n sims.append('T')\n elif sims[-1] == 'C':\n x = np.random.random()\n if matrix[1][0] > x:\n sims.append('A')\n elif matrix[1][1] + matrix[1][0] > x:\n sims.append('C')\n elif matrix[1][2] + matrix[1][1] + matrix[1][0] > x:\n sims.append('G')\n else:\n sims.append('T')\n \n elif sims[-1] == 'G':\n y = np.random.random()\n if matrix[2][0] > y:\n sims.append('A')\n elif matrix[2][1] + matrix[2][0] > y:\n sims.append('C')\n elif matrix[2][2] + matrix[2][1] + matrix[2][0] > y:\n sims.append('G')\n else:\n sims.append('T')\n\n else:\n z = np.random.random()\n if matrix[3][0] > z:\n sims.append('A')\n elif matrix[3][1] + matrix[3][0] > z:\n sims.append('C')\n elif matrix[3][2] + matrix[3][1] + matrix[3][0] > z:\n sims.append('G')\n else:\n sims.append('T')\n\n return sims[1:-1] # remove the initial value (the seed)","def solution(self) -> State:","def result(self, state, action):\n state_after_act = [[0 for i in range(self.col)] for j in range(self.row)]\n for k in action:\n x = k[1][0]\n y = k[1][1]\n if k[0] == \"vaccinate\":\n state_after_act[x][y] = ('I', 1)\n else:\n state_after_act[x][y] = ('Q', 1)\n\n for i in range(self.row):\n for j in range(self.col):\n if state_after_act[i][j] == 0:\n if state[i][j][0] == 'U' or state[i][j][0] == 'I':\n state_after_act[i][j] = state[i][j]\n\n elif state[i][j][0] == 'S':\n if state[i][j][1] == 3:\n state_after_act[i][j] = ('H', 1)\n else:\n if state[i][j][1] == 1:\n state_after_act[i][j] = ('S', 2)\n elif state[i][j][1] == 2:\n state_after_act[i][j] = ('S', 3)\n\n elif state[i][j][0] == 'Q':\n if state[i][j][1] == 2:\n state_after_act[i][j] = ('H', 1)\n else:\n state_after_act[i][j] = ('Q', 2)\n\n elif state[i][j][0] == 'H':\n state_after_act[i][j] = self.healthy(i, j, state,state_after_act)\n state_after_act[i] = tuple(state_after_act[i])\n return tuple(state_after_act)","def propagate(self, game_state, maximizing, max_hitherto, depth, max_depth):\n \n self.moves += 1\n \n\n if depth >= max_depth:\n rating = self.rate_game_state(game_state)\n return (rating, None, None)\n \n else:\n \n color = self.color if maximizing else (1 if self.color==0 else 0)\n\n self.timer.start(\"\")\n choices = []\n for line in self.get_all_lines(game_state, color):\n for vector in LEGAL_VECTORS:\n\n new_game_state = self.get_new_game_state(game_state, line, vector, color)\n if new_game_state is not None:\n choices.append( (new_game_state, self.rate_game_state(new_game_state), line, vector) )\n\n \n\n choices = sorted(choices, key=lambda x: x[1], reverse=maximizing)\n \n final_choice = choices[0][1:]\n del choices[0]\n\n for choice in choices:\n\n new_game_state, estimated_rating, line, vector = choice\n actual_rating, _, _ = self.propagate(new_game_state, not maximizing, final_choice[0], depth+1, max_depth)\n \n if maximizing:\n if final_choice[0] < actual_rating:\n final_choice = (actual_rating, line, vector)\n\n else: #minimizing\n if (max_hitherto is not None) and (actual_rating < max_hitherto): return (actual_rating, line, vector)\n\n if actual_rating < final_choice[0]:\n final_choice = (actual_rating, line, vector)\n\n return final_choice","def train(n):\n\n ai = NimAI()\n\n print(f\"Play {n} training games\")\n for _ in range(n): \n game = Nim()\n\n # Keep track of last move made by either player\n last = {\n 0: {\"state\": None, \"action\": None},\n 1: {\"state\": None, \"action\": None}\n }\n\n # Game loop\n while True:\n\n # Keep track of current state and action\n state = game.piles.copy()\n action = ai.chooseAction(game.piles)\n\n # Keep track of last state and action\n last[game.player][\"state\"] = state\n last[game.player][\"action\"] = action\n\n # Make move and switch players\n game.move(action)\n new_state = game.piles.copy()\n\n # When game is over, update Q values with rewards\n if game.winner is not None:\n # The game is over when a player just made a move that lost him the game.\n # The move from the previous player was therefore game winning.\n # Both events are used to update the AI.\n # new_state is [0, 0, 0, 0] here and its used to update the AI, because\n # future rewards should not be considered in the Q-learning formula.\n ai.update(state, action, new_state, -1)\n ai.update(\n last[game.player][\"state\"],\n last[game.player][\"action\"],\n new_state,\n 1\n )\n break\n\n # If game is continuing, no rewards yet\n elif last[game.player][\"state\"] is not None:\n ai.update(\n last[game.player][\"state\"],\n last[game.player][\"action\"],\n new_state,\n 0\n )\n\n print(\"Done training\")\n\n # Return the trained AI\n return ai","def greedy_next_action(self, state):\n max_val = float('-inf')\n if self.verbose:\n cells = []\n max_candidates = {}\n for i in range(3):\n for j in range(3):\n if state[i][j] == VALUES.EMPTY:\n val = self.q_value((state, (i, j)))\n if val >= max_val:\n max_val = val\n max_move = (i, j)\n max_candidates[max_move] = val\n if self.verbose:\n cells.append('{0:.3f}'.format(val).center(6))\n elif self.verbose:\n cells.append(state[i][j].center(6))\n if self.verbose:\n self.logger.info(BOARD.format(*cells))\n possible_actions = [k for k, v in max_candidates.items() if v == max_val]\n action = random.choice(possible_actions) if len(possible_actions) > 0 else None\n return action","def next_state(s_curr, action, params):\n P_dist = params['P_dist']\n R = params['R']\n n_rows = params['n_rows']\n n_cols = params['n_cols']\n occ_grid = params['occ_grid']\n\n rnd = np.random.uniform()\n\n s_next = s_curr\n\n # Actions - ['left','right','up','down']\n\n if rnd <= P_dist:\n if action == 0:\n move = 2\n elif action == 1:\n move = 2\n elif action == 2:\n move = 1\n else:\n move = 0\n elif rnd < 2*P_dist:\n if action == 0:\n move = 3\n elif action == 1:\n move = 3\n elif action == 2:\n move = 1\n else:\n move = 1\n else:\n move = action\n\n # Move left\n if move == 0:\n row_next = s_curr[0]\n col_next = s_curr[1] - 1\n if col_next >= 0 and occ_grid[row_next, col_next] == 0:\n s_next = [row_next, col_next]\n\n # Move right\n if move == 1:\n row_next = s_curr[0]\n col_next = s_curr[1] + 1\n if col_next < n_cols and occ_grid[row_next, col_next] == 0:\n s_next = [row_next, col_next]\n\n # Move up\n if move == 2:\n row_next = s_curr[0] - 1\n col_next = s_curr[1]\n if row_next >= 0 and occ_grid[row_next, col_next] == 0:\n s_next = [row_next, col_next]\n\n # Move down\n if move == 3:\n row_next = s_curr[0] + 1\n col_next = s_curr[1]\n if row_next < n_rows and occ_grid[row_next, col_next] == 0:\n s_next = [row_next, col_next]\n\n r = R[s_next[0], s_next[1]]\n return s_next, r","def simulation():\n # initialize action set\n action_set = np.zeros(int((s.MAX_INSPECT - s.MIN_INSPECT) / s.DELTA) + 3)\n x, i = s.MIN_INSPECT, 1\n while x <= s.MAX_INSPECT:\n action_set[i] = x\n x += s.DELTA\n i += 1\n action_set[-1] = np.inf\n action_number = len(action_set)\n\n # initialize current state\n current_state = math.floor(np.random.rand(1) * s.NUM_STATES)\n\n # initialize action index\n if current_state == 0:\n action_index = 0\n elif current_state == s.NUM_STATES - 1:\n action_index = action_number - 1\n\n if current_state != 0 and current_state != s.NUM_STATES - 1:\n action_index = action_number - 2\n\n # initialize policy set\n greedy_policy = np.zeros(s.NUM_STATES)\n greedy_policy[-1] = np.inf\n for i in range(1, s.NUM_STATES - 1):\n greedy_policy[i] = s.MAX_INSPECT\n\n visit_times = np.zeros([s.NUM_STATES, action_number])\n\n # initialization for simulation\n falpha, Aalpha, delay_T, uni_parameter = equivalent_markov(greedy_policy)\n stable_prob, potential = stable_potential(falpha, Aalpha, uni_parameter)\n last_value = falpha + np.matmul(Aalpha, potential)\n dis_value = last_value\n # ave_vector = np.matmul(stable_prob, falpha)\n # ave_estimate = ave_vector.tolist()\n each_transit_cost, each_transit_time, total_reward = (0 for i in range(3))\n\n # initialize DQN model if selected\n dqn = DQN() if MODEL == 1 else None\n # initialize Q-table if Q-learning selected\n q_factor = ql.init_q_factor(action_number) if MODEL == 2 else None\n\n for out_step in range(s.EPOCH):\n epsilon = s.EPSILON_1 if MODEL == 1 else s.EPSILON_2\n\n for inner_step in range(s.EPOCH_LEARN):\n\n visit_times[current_state, action_index] += 1\n current_action = greedy_policy[current_state]\n\n inspect_cost = 0 if current_state == s.NUM_STATES - 1 else s.K5 * current_action\n\n flag, sojourn_T, service_T, next_state = state_transition(current_state, current_action)\n each_transit_time = s.DISCOUNT * each_transit_time + (sojourn_T - each_transit_time) / pow(\n out_step * s.EPOCH_LEARN + (inner_step + 1), s.Q_AVE_STEP)\n end_sojourn_T = math.exp(- s.ALPHA * sojourn_T)\n end_serve_T = math.exp(- s.ALPHA * service_T)\n\n if s.ALPHA == 0:\n dis_T, dis_serve_T, dis_wait_T = sojourn_T, service_T, sojourn_T - service_T\n else:\n dis_T, dis_serve_T = (1 - end_sojourn_T) / s.ALPHA, (1 - end_serve_T) / s.ALPHA\n dis_wait_T = (end_serve_T - end_sojourn_T) / s.ALPHA\n\n if flag == 0: # no processing, waiting\n cost_real = (s.K1 * (s.NUM_STATES - current_state) + s.K3) * sojourn_T + inspect_cost\n cost_purt = (s.K1 * (s.NUM_STATES - current_state) + s.K3) * dis_T + inspect_cost\n else: # no waiting, processing\n cost_real = s.K1 * (s.NUM_STATES - current_state - 1) * sojourn_T + s.K2 * service_T + s.K3 * (\n sojourn_T - service_T) + s.K4 + inspect_cost\n cost_purt = s.K1 * (s.NUM_STATES - current_state - 1) * dis_T + s.K2 * dis_serve_T + s.K3 * dis_wait_T \\\n + s.K4 * end_serve_T + inspect_cost\n\n each_transit_cost = s.DISCOUNT * each_transit_cost + (cost_real - each_transit_cost) / (\n pow(out_step * s.EPOCH_LEARN + (inner_step + 1), s.Q_AVE_STEP))\n\n ave_q_cost = each_transit_cost / each_transit_time\n # ave_estimate.append(ave_q_cost)\n cost_dis = cost_purt - ave_q_cost * dis_T\n\n if MODEL == 1:\n reward = - cost_dis\n dqn.store_transition(current_state, action_index, reward, next_state)\n if dqn.memory_counter >= s.MEMORY_CAPACITY:\n dqn.learn(s.EPOCH_LEARN, inner_step, PS)\n else:\n difference = cost_dis + end_sojourn_T * min(q_factor[next_state, :]) \\\n - q_factor[current_state, action_index]\n q_factor = ql.update_q_factor(q_factor, current_state, action_index, difference,\n visit_times, inner_step, PS)\n current_state = next_state # transit to next state\n\n if current_state == 0:\n action_index = 0\n elif current_state == s.NUM_STATES - 1:\n action_index = action_number - 1\n else:\n if MODEL == 1:\n action_index = int(dqn.choose_action(current_state, epsilon))\n if action_set[action_index] <= 1:\n greedy_policy[current_state] = action_set[action_index]\n else:\n greedy_policy[current_state] = 1\n else:\n if np.random.rand(1) < epsilon:\n action_index = int(np.floor(np.random.rand(1) * (action_number - 2)) + 1)\n else:\n # minimal_q_value = np.min(q_factor[current_state, :])\n action_index = np.argmin(q_factor[current_state, :])\n greedy_policy[current_state] = action_set[action_index]\n\n # store the policy learned from the iterations\n optimal_policy = greedy_policy\n\n if MODEL != 1:\n for i in range(1, s.NUM_STATES - 1):\n # minimal_q_value_temp = np.min(q_factor[i, :])\n action_index_temp = np.argmin(q_factor[i, :])\n optimal_policy[i] = action_set[action_index_temp]\n\n falpha, Aalpha, delay_T, uni_parameter = equivalent_markov(optimal_policy)\n stable_prob, potential = stable_potential(falpha, Aalpha, uni_parameter)\n\n last_value = falpha + np.matmul(Aalpha, potential)\n dis_value = np.concatenate((dis_value, last_value), axis=1)\n total_reward += - np.ndarray.item(last_value[0])\n # new_ave_cost = np.matmul(stable_prob, falpha)\n # ave_vector = np.concatenate((ave_vector, new_ave_cost))\n print(\"epoch: {} , the epoch reward is {}\".format(out_step, round(- np.ndarray.item(last_value[0]), 2)))\n\n # result = np.asarray(dis_value)\n print(\"total reward:\", total_reward)\n\n return dis_value, total_reward","def next_state(self):\n \n self.state = np.random.choice(['checkout', 'dairy', 'drinks', 'fruit', 'spices'], p=self.tr_array_dict[f'{self.state}'])","def breadthfirst(initial_game):\n # Creat dict for solved case\n solved_cases = {}\n\n # Track states.\n seen_states = []\n\n queue = q.Queue()\n\n # Add begin state to queue.\n initial_state = initial_game.current_state\n queue.put((initial_state, []))\n seen_states.append(initial_state)\n board_size = initial_game.board.length\n\n while not queue.empty():\n # Get first from queue.\n parent_node = copy.deepcopy(queue.get())\n\n # Create a game instance from the state\n parent_state = parent_node[0]\n parent_game = create_game_from_state(parent_state, board_size)\n\n # Loop over the possible moves in the state.\n possible_moves = parent_game.possible_moves\n for move in possible_moves:\n # Get the move specifics.\n vehicle, steps = move\n new_position = vehicle.speculate_new_position(steps)\n\n # Execute the move to create a child node.\n child_game = copy.deepcopy(parent_game)\n child_vehicle = child_game.vehicles.get(vehicle.id)\n child_game.move(child_vehicle, new_position)\n\n # Heuristic: Check if the state is unique.\n child_state = child_game.current_state\n if not is_state_unique(seen_states, child_state):\n continue\n\n # Otherwise, remember the state.\n seen_states.append(child_state)\n \n if len(seen_states) % 100 == 0:\n print(f\"Seen states: {len(seen_states)}\")\n\n # Add new child node to queue.\n moves_until_state = copy.deepcopy(parent_node[1])\n moves_until_state.append((vehicle.id, steps))\n child_node = (child_state, moves_until_state)\n queue.put(child_node)\n\n # Heuristic: check if the boxes from the red car up until the exit are free.\n if is_exit_reachable(child_game):\n last_move = finish_game(child_game)\n moves = child_node[1]\n moves.append(last_move)\n solved_cases[0] = moves\n\n if child_game.is_finished():\n print(f\"Seen states: {len(seen_states)}\")\n return solved_cases\n\n return \"No solved cases have been found :(\"","def solveOneStep(self):\n ### Student code goes here\n # Mark this move as explored\n self.visited[self.currentState] = True\n self.visited_states.append(self.currentState.state)\n\n # Get move to make\n movables = self.gm.getMovables()\n # print(\"EXPLORING GAME STATE \" + str(self.gm.getGameState()) + \"---------------------------------------------------------\")\n to_move = self.currentState.nextChildToVisit # movables index\n # print(\"depth \", self.currentState.depth)\n\n # Return if done\n if self.currentState.state == self.victoryCondition:\n # print(\"DONE\")\n return True\n\n # If current state has no children, make children\n if not self.currentState.children:\n for movable_statement in movables:\n # Make the move\n # print(\"implementing move \", movable_statement)\n self.gm.makeMove(movable_statement)\n\n # Create a new state with this move made\n new_state = self.gm.getGameState()\n # print (\"new state \", new_state)\n\n # If the new state hasn't been visited and isn't in the queue then add it as a child and to the queue\n if (new_state not in self.visited_states):\n new_gs = GameState(new_state, self.currentState.depth + 1, movable_statement)\n new_gs.parent = self.currentState\n self.currentState.children.append(new_gs)\n self.currentState.nextChildToVisit = to_move + 1\n self.visited[new_gs] = True\n self.visited_states.append(new_state)\n self.gs_queue.append(new_gs)\n\n self.gm.reverseMove(movable_statement)\n\n # Return false if no more to explore\n if not self.gs_queue:\n return False\n\n # Revert to state at when current and next start to change\n root_curr = self.currentState\n self.currentState = self.gs_queue.popleft()\n root_new = self.currentState\n\n # Backtrack to when current node and new node start to diverge\n if root_new.depth == root_curr.depth:\n while root_curr.state != root_new.state:\n self.gm.reverseMove(root_curr.requiredMovable)\n root_curr = root_curr.parent\n root_new = root_new.parent\n else:\n while root_curr.requiredMovable:\n self.gm.reverseMove(root_curr.requiredMovable)\n root_curr = root_curr.parent\n\n # Return game master to state that we are exploring\n # Find path between root and current state\n path = []\n currNode = self.currentState\n while currNode != root_curr:\n path.append(currNode.requiredMovable)\n currNode = currNode.parent\n\n # Created backwards path, now make moves from root to current state\n path.reverse()\n for movable_statement in path:\n self.gm.makeMove(movable_statement)\n\n return False","def UCT(player, rootstate, iter, prev_board, verbose=False):\n rootnode = Node(player=player, state=rootstate, preState=prev_board)\n\n for i in range(iter):\n node = rootnode\n state = deepcopy(rootstate)\n pre_state = deepcopy(state)\n currentPlayer = -player # turn of the player x\n\n # Select\n while (\n node.untriedMoves == [] and node.childNodes != []\n ): # node is fully expanded and non-terminal, loop until have untried Child or it is leaf node\n # choose the next child\n node = node.UCTChooseChild()\n # update the state of the board ((x,y),(x,y))\n pre_state = deepcopy(state)\n state = board_after_move_and_capturing(node.move[0], node.move[1], state)\n currentPlayer = -currentPlayer\n\n # Expand\n if (\n node.untriedMoves != []\n ): # if we can expand (i.e. state/node is non-terminal)\n m = random.choice(node.untriedMoves)\n # update the state of the board\n pre_state = deepcopy(state)\n state = board_after_move_and_capturing(m[0], m[1], state)\n currentPlayer = -currentPlayer\n\n node = node.addChild(m, state, pre_state) # add child and descend tree\n\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\n height = 0\n while get_winner(state) == 0: # while state is non-terminal\n if height > 30:\n break\n choice = random.choice(get_all_legal_moves(pre_state, state, currentPlayer))\n # update the state of the board\n pre_state = deepcopy(state)\n state = board_after_move_and_capturing(choice[0], choice[1], state)\n # print_board(state)\n currentPlayer = -currentPlayer\n height += 1\n\n # cout board\n # print(f\"The board becomes {board_to_string(state)}\")\n\n # Backpropagate\n while (\n node != None\n ): # backpropagate from the expanded node and work back to the root node\n if height > 30:\n winner = player\n else:\n winner = get_winner(state)\n\n if winner == -player:\n node.updateNode(1)\n else:\n node.updateNode(0)\n\n # state is terminal. Update node with result from POV of node.playerJustMoved\n node = node.parentNode\n\n # cout root node\n # for i in rootnode.childNodes:\n # print(i.move)\n # print(i.wins)\n # print(i.visits)\n\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[\n -1\n ].move # return the move that was most visited","def do(self, state):\r\n\r\n # action is [0,1,2,3]: left, right, up, down\r\n _state = json.loads(state)\r\n if _state[0] == 0:\r\n self.Q[state][2] = 0\r\n if _state[0] == self.max_row:\r\n self.Q[state][3] = 0\r\n if _state[1] == 0:\r\n self.Q[state][0] = 0\r\n if _state[1] == self.max_col:\r\n self.Q[state][1] = 0\r\n action_map = {0: [0, -1], # left\r\n 1: [0, 1], # right\r\n 2: [-1, 0], # up\r\n 3: [1, 0], # down\r\n }\r\n chance = np.random.rand()\r\n if self.eps > chance:\r\n while True:\r\n action = np.random.randint(0,4)\r\n next = np.array(_state) + np.array(action_map[action])\r\n if all(next>=0) and next[0]<=self.max_row and next[1]<=self.max_col:\r\n break\r\n else:\r\n action = np.argmax(self.Q[state])\r\n #print(_state,action,self.Q[state])\r\n\r\n \"\"\"Please Fill Your Code Here.\r\n \"\"\"\r\n return action","def leafEvaluation(self, state):\n\n \"\"\"\n Use random generated values for now\n \"\"\"\n z = np.random.randint(2)\n v = random.uniform(0, 1) \n return (1-LAMBDA) * v + LAMBDA * z","def __call__(self, state):\n if random.random() > self._epsilon:\n return self._max_policy(state)\n return random.choice(np.arange(self._action_size))","def uct(root_state, iter_max, verbose=False):\n\n root_node = Node(state=root_state)\n\n for i in range(iter_max):\n node = root_node\n state = root_state.clone()\n\n # Select\n # node is fully expanded and non-terminal\n while not node.untried_moves and node.child_nodes:\n node = node.uct_select_child()\n state.do_move(node.move)\n\n # Expand\n # if we can expand (i.e. state/node is non-terminal)\n if node.untried_moves:\n m = random.choice(node.untried_moves)\n state.do_move(m)\n # add child and descend tree\n node = node.add_child(m, state)\n\n # Rollout - this can often be made orders of magnitude quicker\n # using a state.get_random_move() function\n # while state is non-terminal\n while state.get_moves():\n state.do_move(random.choice(state.get_moves()))\n\n # Backpropagate\n # backpropagate from the expanded node and work back to the root node\n while node is not None:\n # state is terminal. Update node with result\n # from POV of node.playerJustMoved\n node.update(state.get_result(node.player_just_moved))\n node = node.parent_node\n\n # Output some information about the tree - can be omitted\n if verbose:\n print(root_node.tree_to_string(0))\n else:\n print(root_node.children_to_string())\n\n # return the move that was most visited\n return sorted(root_node.child_nodes,\n key=lambda c: c.Q)[-1].move","def get_successors(state): \n \n child_states = []\n \n size = len(state)\n i = 0\n j = 0\n for i in range (size):\n if 0 in state[i]:\n for j in range (size):\n if state[i][j] == 0:\n break \n break\n\n if j != size-1:\n child_states.append ((\"Left\", swap_cells(state, i, j, i, j+1)))\n if j != 0:\n child_states.append ((\"Right\", swap_cells(state, i, j, i, j-1)))\n if i != size-1:\n child_states.append ((\"Up\", swap_cells(state, i, j, i+1, j)))\n if i != 0:\n child_states.append ((\"Down\", swap_cells(state, i, j, i-1, j)))\n \n return child_states","def find_best_move(state: GameState) -> None:","def evaluate(self, state):\n\n fitness = 0\n\n for i in range(1, len(state)):\n if state[i] != state[i - 1]:\n fitness += 1\n\n return fitness","def build(self, w=1):\n states = self.a.states.values()\n reachedStates = set()\n remainingStates = set()\n\n _open = Heap()\n\n for s in states:\n if s.acceptingFunction is not None:\n s.h = 0\n _open.push(0, s)\n reachedStates.add(s)\n else:\n remainingStates.add(s)\n\n if not reachedStates:\n Color.RED.print(\"No states reached!\")\n Color.RED.print(\"States were not reached: %s\" % remainingStates)\n raise Exception(\"Query has no goal, please review it!\")\n\n while _open:\n (h, s) = _open.pop()\n\n for cS in s._prev():\n if h+w < cS.h:\n reachedStates.add(cS)\n _open.push(h+w, cS)\n\n cS.h = h+w\n if cS in remainingStates:\n remainingStates.remove(cS)\n\n if remainingStates:\n Color.RED.print(\"Some states were not reached: %s\" % remainingStates)\n Color.YELLOW.print(\"States reached: %s\" % reachedStates)\n raise Exception(\"Query is probably unfinished, please review it\")\n\n reached = [(s.h, s) for s in reachedStates]\n reached.sort()\n\n if __debug__:\n Color.GREEN.print(\"Heuristic:\")\n for (h, s) in reached:\n Color.GREEN.print(\" * %d: %s\" % (h, s))\n\n for s in states:\n s.prepare()\n\n return self.a","def select_action(self, state):\r\n policy_s = self.epsilon_greedy_probs(self.nA, self.Q[state], self.count, self.epsilon)\r\n return np.random.choice(np.arange(self.nA), p=policy_s)","def myHeuristic3(state, problem=None):\n #canto =[(1,1), (1,2), (2,1), (2,2), (36,36), (35,36), (35,35), (36,35), (1,36),(1,35),(2,36),(2,35),(36,1),(36,2),(35,1),(35,2)]\n canto = []\n for l in range(2):\n for c in range(2):\n canto.append((l,c))\n heru = abs(state[0] - 1) + abs(state[1] - 1)\n #heru=10\n if state in canto:\n #print(\"sim\")\n heru = heru * 0.5\n return heru","def astar(model, start, end):\n\n # Create start and end node\n start_node = Node(None,0.0, start)\n start_node.g = 0\n start_node.h = calculate_h(start_node,end)\n start_node.f = start_node.g + start_node.h \n \n end_node = Node(None,0.0, end)\n end_node.g = end_node.h = end_node.f = 0\n\n # Initialize both open and closed list\n open_list = []\n closed_list = []\n\n # Add the start node\n open_list.append(start_node)\n\n\n\n iterations = 0\n # Loop until you find the end\n while len(open_list) > 0:\n iterations += 1\n # print(iterations)\n print('openlist={},closed_list={}'.format(len(open_list),len(closed_list)))\n\n # Get the current node\n current_node = open_list[0]\n current_index = 0\n for index, item in enumerate(open_list):\n if item.f < current_node.f:\n current_node = item\n current_index = index\n\n # Pop current off open list, add to closed list\n open_list.pop(current_index)\n closed_list.append(current_node)\n\n print('currentstate={}, parentaction={}, f={}'.format(current_node.state, current_node.parent_action, current_node.f))\n\n # run_sim(current_node.parent_action)\n\n # Found the goal\n if current_node == end_node: \n actions = []\t\n current = current_node\n while current is not None:\n actions.append(current.parent_action)\n current = current.parent_state\n # print(current.state)\n return actions[::-1] # Return reversed path\n\n # Generate children states\n children = []\n\n for action in discretized_action: # all possible actions\n\n # Get new state node (closest to discretized state space)\n \n # new_state = get_next_state_from_model(model,current_node,action)\n new_state = get_next_state_from_eqn(current_node,action)\n # new_state = run_sim(action)\n # print(new_state)\n\n # # Make sure within range\n # if node_position[0] > (len(maze) - 1) or node_position[0] < 0 or node_position[1] > (len(maze[len(maze)-1]) -1) or node_position[1] < 0:\n # continue\n\n # # Make sure walkable terrain\n # if maze[node_position[0]][node_position[1]] != 0:\n # continue\n\n #############Make sure to wrap theta \n \n # Create new node\n new_node = Node(current_node, action, new_state)\n\n # Append\n children.append(new_node)\n\n # Loop through children\n for child in children:\n\n # Child is on the closed list\n for closed_child in closed_list:\n if child == closed_child:\t\n continue\n\n # Create the f, g, and h values\n child.g = calculate_g(child,current_node) #current_node.g + 1\n child.h = calculate_h(child,end_node)\n child.f = child.g + child.h\n \n # \tprint(child.state, child.parent_action, child.f) \n \n # Child is already in the open list\n for open_node in open_list:\n if child == open_node and child.g > open_node.g:\n continue\n\n # Add the child to the open list\n open_list.append(child)","def UCT(rootstate, itermax, verbose=False):\n\n rootnode = Node(state=rootstate)\n\n for i in range(itermax):\n node = rootnode\n state = rootstate.Clone()\n\n # Select\n while node.untriedMoves == [] and node.childNodes != []: # node is fully expanded and non-terminal\n node = node.UCTSelectChild()\n state.DoMove(node.move)\n\n # Expand\n expand = True\n while expand and node.untriedMoves != []: # if we can expand (i.e. state/node is non-terminal)\n m = random.choice(node.untriedMoves)\n # print(\"[Expand] Untried move %s, %s, %s\" % (m[0], m[1], m[2]))\n expand = not state.DoMove(m)\n node = node.AddChild(m, state) # add child and descend tree\n\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\n while state.GetMoves() != []: # while state is non-terminal\n state.DoMove(random.choice(state.GetMoves()))\n\n # Backpropagate\n while node != None: # backpropagate from the expanded node and work back to the root node\n node.Update(state.GetResult()) # state is terminal. Update node with result from POV of node.playerJustMoved\n node = node.parentNode\n\n # Output some information about the tree - can be omitted\n if (verbose):\n print(rootnode.TreeToString(0))\n else:\n print(rootnode.ChildrenToString())\n\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[-1].move # return the move that was most visited","def next_possible_states(path, check_dict, check):\r\n \r\n current_state_tuple = path[-1]\r\n state_container = []\r\n x = current_state_tuple[1][0]\r\n y = current_state_tuple[1][1]\r\n current_state = current_state_tuple[0]\r\n\r\n # Down\r\n if y < 3:\r\n new_state = copy.deepcopy(current_state)\r\n new_state[y][x] = new_state[y + 1][x]\r\n new_state[y + 1][x] = 0\r\n if not been_there(new_state, check_dict, check):\r\n new_index = (x, y + 1)\r\n h1 = euclidean_dist(new_state, path)\r\n new_state_tuple = (new_state, new_index, h1)\r\n state_container.append(new_state_tuple)\r\n\r\n # Up\r\n if y > 0:\r\n new_state = copy.deepcopy(current_state)\r\n if y == 1 and x == 0:\r\n new_state[y][x] = new_state[y - 1][x]\r\n new_state[y - 1][x] = 0\r\n if is_goal(new_state):\r\n new_index = (x, y - 1)\r\n h1 = euclidean_dist(new_state, path)\r\n new_state_tuple = (new_state, new_index, h1)\r\n state_container.append(new_state_tuple)\r\n elif y > 1:\r\n new_state[y][x] = new_state[y - 1][x]\r\n new_state[y - 1][x] = 0\r\n if not been_there(new_state, check_dict, check):\r\n new_index = (x, y - 1)\r\n h1 = euclidean_dist(new_state, path)\r\n new_state_tuple = (new_state, new_index, h1)\r\n state_container.append(new_state_tuple)\r\n\r\n # Left\r\n if x > 0 and y > 0:\r\n new_state = copy.deepcopy(current_state)\r\n new_state[y][x] = new_state[y][x - 1]\r\n new_state[y][x - 1] = 0\r\n if not been_there(new_state, check_dict, check):\r\n new_index = (x - 1, y)\r\n h1 = euclidean_dist(new_state, path)\r\n new_state_tuple = (new_state, new_index, h1)\r\n state_container.append(new_state_tuple)\r\n\r\n # Right\r\n if x < 2 and y > 0:\r\n new_state = copy.deepcopy(current_state)\r\n new_state[y][x] = new_state[y][x + 1]\r\n new_state[y][x + 1] = 0\r\n if not been_there(new_state, check_dict, check):\r\n new_index = (x + 1, y)\r\n h1 = euclidean_dist(new_state, path)\r\n new_state_tuple = (new_state, new_index, h1)\r\n state_container.append(new_state_tuple)\r\n\r\n return state_container","def choose_action(self, state):\n prob = [] # Probability distribution\n for i in range(len(ACTIONS)):\n prob.append(self.epsilon/4)\n Q_func = self.policy.predict(process_state(state))\n Q_vals = Q_func[0]\n max_index = []\n Qmax = np.amax(Q_vals)\n for i in range(len(prob)):\n if Q_vals[i] == Qmax:\n # max_index.append(i)\n prob[i] = 1 - self.epsilon + self.epsilon/4\n break\n # ind = np.random.choice(max_index)\n # prob[ind] = 1 - self.epsilon + self.epsilon/4\n action = np.random.choice(ACTIONS, p = prob)\n return action","def epsilon_greedy(Q, epsilon, state):\n random_number = random.random()\n if (random_number < epsilon) and (state not in critical_states):\n return env.action_space.sample()\n\n else:\n return np.argmax(Q[state])","def ai_strategy2(gstate: TicTacToe, game_tree):\n status, player = gstate.game_status\n\n if status != \"turn\":\n logging.warning(\"Game status = %s. No move needed.\", status)\n return None\n\n win, lose = (-1, 1) if player == \"1\" else (1, -1)\n next_move_vals = {\n idx: win * game_tree[state][\"value\"] for idx, state in gstate.next_moves.items()\n }\n max_val = max(next_move_vals.values())\n moves = {\n idx: (\n game_tree[gstate.next_moves[idx]][\"outcomes\"].get(str(win), 0),\n game_tree[gstate.next_moves[idx]][\"outcomes\"].get(str(0), 0),\n game_tree[gstate.next_moves[idx]][\"outcomes\"].get(str(lose), 0)\n )\n for idx, val in next_move_vals.items() if val == max_val\n }\n\n win_ct = {idx: vals[0] for idx, vals in moves.items()}\n win_pct = {idx: vals[0] / sum(vals) for idx, vals in moves.items()}\n lose_pct = {idx: vals[2] / sum(vals) for idx, vals in moves.items()}\n wl_ratio = {idx: vals[0] / max(vals[2], 0.5)\n for idx, vals in moves.items()}\n\n # criteria, agg_func = lose_pct, min\n # criteria, agg_func = win_pct, max\n criteria, agg_func = wl_ratio, max\n\n if max_val == 1 and 1 in win_ct.values():\n move = [idx for idx, val in win_ct.items() if val == 1][0]\n else:\n move = random.choice(\n [idx for idx, val in criteria.items() if val ==\n agg_func(criteria.values())]\n )\n\n logging.debug(\"move: %i; value: %i, win paths %%: %.1f%%, lose paths %%: %.1f%%, moves: %s\\n\",\n move, max_val, win_pct[move] * 100, lose_pct[move] * 100, moves)\n\n # trash talk\n if max_val == 1:\n print(\n \"*beep* *boop* *beep*\"\n \" -=[ I calculate chances of winning to be 100% ]=- \"\n \"*beep* *boop* *beep*\"\n )\n\n return move","def test_create_state(self):\n\n # States should be abled to be identified by numbers or by strings I suppose.\n # I don't imagine that strings will ever be used.\n mdp = MDP()\n mdp.add_state(0)\n mdp.add_state(1)\n mdp.add_state(2)\n mdp.add_state(3)\n mdp.add_state(4)\n mdp.add_state(5, terminal=True)\n self.assertEqual(mdp.num_states(), 6)","def learn(state=\"_________\"):\n game_tree = gen_game_tree(state)\n with open(GAME_TREE_FILE, \"w\") as gt_file:\n json.dump(game_tree, gt_file, indent=4)","def __init__(self, state: State):\n self._state = state\n self._parent = None\n self._untried_edges = self._state.possible_actions\n self.children = {} # {AbstractAction: AbstractState}\n self.tot_reward = 0\n self.num_samples = 0","def max_diffs(state):\n # your code here\n return best_action(state, pig_actions, Q_pig, win_diff)","def act(self, state):\n\n afterstates = state.afterstates()\n\n if self.choose_to_explore(): # explore\n action_id = self._generator.integers(0, len(afterstates))\n return afterstates[action_id]\n\n estimates = self.act_with_minimax(afterstates, self._minimax_depth,\n maximizing=False, best=None,\n initial=True)\n min = np.amin(estimates)\n argmin = np.argwhere(estimates == min).flatten()\n action_id = argmin[self._generator.integers(0, len(argmin))]\n return afterstates[action_id]","def getAction(self, gameState):\n \"*** YOUR CODE HERE ***\"\n self.NumAgents = gameState.getNumAgents()\n self.root = gameState\n # alpha-beta-search\n # one call of Max_Value is enough to iterate the whole tree and keep the optimal move\n self.Max_Value(gameState, -10000, 10000, self.index, self.depth)\n return self.move\n util.raiseNotDefined()","def simulate(params,n_states,n_trials,env = \"rich\", policy=\"softmax\",\\\n D=0.5, mod = \"constant\",thresh = 0, k=1,rnd_seeds = None, V0=0.0, full=False,\n rmag = 1, lmag = 0):\n\n\tdef calc_D(state):\n\t\t\"\"\"\n\t\tcalculates D for the current trial and returns\n\t\tthe updated state tracker for D and respective betas\n\n\t\tD represents dopamine levels (equivalent of rho in OpAL)\n\t\tScales between 0 and 1, with 1 high level of DA\n\t\t\"\"\"\n\t\tif t < thresh:\n\t\t\tstate.D_g[t] = 0.5\n\t\t\tstate.D_n[t] = 0.5\n\t\telse:\n\t\t\tif mod == \"constant\":\n\t\t\t\tstate.D_g[t] = D\n\t\t\t\tstate.D_n[t] = 1-D\n\t\t\tif mod == \"value\":\n\t\t\t\t# NOTE: if rmag and lmag is 1/0, can just use V\n\t\t\t\t# average of two actions\n\t\t\t\tV = np.mean(1/2*(state.QG[t,:] - state.QN[t,:])) # state average(?) \n\t\t\t\tV = 1/(1 + np.exp(-V*k)) # translate between 0 and 1\n\t\t\t\tstate.D_g[t] = V \n\t\t\t\tstate.D_n[t] = 1 - V\n\t\treturn state\n\n\n\tdef generate_state():\n\t\t\"\"\"\n\t\tGet appropriate reward probabilities and magnitudes\n\t\tfor the specified environment type\n\t\t\"\"\"\n\n\t\tprobs = calc_probs(env)\n\t\tn_options = len(probs)\n\n\t\t# feedback for agent\n\t\tr_mag = np.zeros(n_options) + rmag\n\t\tl_mag = np.zeros(n_options) + lmag\n\n\t\tnew_state = Bogacz(n_trials, n_options, probs, r_mag, l_mag, V0=V0)\n\t\treturn new_state\n\n\n\t# learning rate, damping, decay, softmax temp\n\talpha_a, epsilon, lbda, beta = params\n\tstates = []\n\n\t# do the thing\n\tfor s in np.arange(n_states):\n\n\t\t# check if random seed provided\n\t\tif rnd_seeds is not None:\n\t\t\trandom.seed(rnd_seeds[s])\n\t\t\tnp.random.seed(rnd_seeds[s])\n\n\t\tstate = generate_state()\n\t\tfor t in range(n_trials):\n\n\t\t\tstate.idx = t\n\t\t\tstate=calc_D(state)\t\t\t\t\t# get D\n\t\t\tstate.policy_softmax(beta)\n\t\t\tstate.act(alpha_a, epsilon, lbda)\t# update \n\n\t\t\tif full:\n\t\t\t\tstate.update_other_actions(alpha_a, epsilon, lbda)\n\n\t\tstates.append(state)\t\t\t\t\t# save sim\n\n\treturn states","def generate_cubes(num_states, distance, startstate='RRRRRRRRRBBBBBBBBBOOOOOOOOOGGGGGGGGGWWWWWWWWWYYYYYYYYY'):\n terminal_states = {'RRRRRRRRRYYYYYYYYYOOOOOOOOOWWWWWWWWWBBBBBBBBBGGGGGGGGG',\n 'BBBBBBBBBYYYYYYYYYGGGGGGGGGWWWWWWWWWOOOOOOOOORRRRRRRRR',\n 'YYYYYYYYYBBBBBBBBBWWWWWWWWWGGGGGGGGGRRRRRRRRROOOOOOOOO',\n 'BBBBBBBBBWWWWWWWWWGGGGGGGGGYYYYYYYYYRRRRRRRRROOOOOOOOO',\n 'OOOOOOOOOGGGGGGGGGRRRRRRRRRBBBBBBBBBWWWWWWWWWYYYYYYYYY',\n 'GGGGGGGGGOOOOOOOOOBBBBBBBBBRRRRRRRRRYYYYYYYYYWWWWWWWWW',\n 'GGGGGGGGGRRRRRRRRRBBBBBBBBBOOOOOOOOOWWWWWWWWWYYYYYYYYY',\n 'GGGGGGGGGYYYYYYYYYBBBBBBBBBWWWWWWWWWRRRRRRRRROOOOOOOOO',\n 'RRRRRRRRRWWWWWWWWWOOOOOOOOOYYYYYYYYYGGGGGGGGGBBBBBBBBB',\n 'OOOOOOOOOWWWWWWWWWRRRRRRRRRYYYYYYYYYBBBBBBBBBGGGGGGGGG',\n 'WWWWWWWWWRRRRRRRRRYYYYYYYYYOOOOOOOOOBBBBBBBBBGGGGGGGGG',\n 'BBBBBBBBBOOOOOOOOOGGGGGGGGGRRRRRRRRRWWWWWWWWWYYYYYYYYY',\n 'BBBBBBBBBRRRRRRRRRGGGGGGGGGOOOOOOOOOYYYYYYYYYWWWWWWWWW',\n 'RRRRRRRRRGGGGGGGGGOOOOOOOOOBBBBBBBBBYYYYYYYYYWWWWWWWWW',\n 'YYYYYYYYYRRRRRRRRRWWWWWWWWWOOOOOOOOOGGGGGGGGGBBBBBBBBB',\n 'YYYYYYYYYOOOOOOOOOWWWWWWWWWRRRRRRRRRBBBBBBBBBGGGGGGGGG',\n 'GGGGGGGGGWWWWWWWWWBBBBBBBBBYYYYYYYYYOOOOOOOOORRRRRRRRR',\n 'WWWWWWWWWGGGGGGGGGYYYYYYYYYBBBBBBBBBRRRRRRRRROOOOOOOOO',\n 'OOOOOOOOOYYYYYYYYYRRRRRRRRRWWWWWWWWWGGGGGGGGGBBBBBBBBB',\n 'RRRRRRRRRBBBBBBBBBOOOOOOOOOGGGGGGGGGWWWWWWWWWYYYYYYYYY',\n 'WWWWWWWWWBBBBBBBBBYYYYYYYYYGGGGGGGGGOOOOOOOOORRRRRRRRR',\n 'WWWWWWWWWOOOOOOOOOYYYYYYYYYRRRRRRRRRGGGGGGGGGBBBBBBBBB',\n 'OOOOOOOOOBBBBBBBBBRRRRRRRRRGGGGGGGGGYYYYYYYYYWWWWWWWWW',\n 'YYYYYYYYYGGGGGGGGGWWWWWWWWWBBBBBBBBBOOOOOOOOORRRRRRRRR'}\n states = []\n while len(states) < num_states:\n x = RubiksCubeOld(startstate)\n for j in range(distance):\n x.apply_move(np.random.randint(0,18))\n newstate = x.get_state()\n if newstate not in terminal_states: states.append(newstate)\n states = list(set(states))\n\n return states","def solveOneStep(self):\n ### Student code goes here\n if (self.currentState.state == self.victoryCondition) or (self.currentState not in self.visited):\n self.visited[self.currentState] = True\n win_or_not = self.currentState.state == self.victoryCondition\n return win_or_not\n\n if not self.currentState.nextChildToVisit: \n its = 0\n for movable in self.gm.getMovables():\n its += 1\n # time test\n # too long \n if its == \"too long\":\n return \"too long\"\n #make every move in movable\n self.gm.makeMove(movable)\n new = self.gm.getGameState()\n new_gs = GameState(new, self.currentState.depth+1, movable)\n \n if new_gs not in self.visited:\n new_gs.parent = self.currentState\n self.currentState.children.append(new_gs)\n self.gm.reverseMove(movable) \n \n num_children = len(self.currentState.children)\n if self.currentState.nextChildToVisit < num_children:\n new = self.currentState.children[self.currentState.nextChildToVisit]\n self.currentState.nextChildToVisit = self.currentState.nextChildToVisit + 1\n self.gm.makeMove(new.requiredMovable)\n self.currentState = new\n #recurse\n return self.solveOneStep()\n else:\n self.currentState.nextChildToVisit = self.currentState.nextChildToVisit + 1\n self.gm.reverseMove(self.currentState.requiredMovable)\n self.currentState = self.currentState.parent\n #recurse\n return self.solveOneStep()","def evaluateBoardState(self, board):\n\n \"\"\"\n These are the variables and functions for board objects which may be helpful when creating your Agent.\n Look into board.py for more information/descriptions of each, or to look for any other definitions which may help you.\n\n Board Variables:\n board.width \n board.height\n board.last_move\n board.num_to_connect\n board.winning_zones\n board.score_array \n board.current_player_score\n\n Board Functions:\n get_cell_value(row, col)\n try_move(col)\n valid_move(row, col)\n valid_moves()\n terminal(self)\n legal_moves()\n next_state(turn)\n winner()\n \"\"\"\n if self.id == 1:\n opponent_id = 2\n else:\n opponent_id = 1\n\n maxvalue = 100000\n minvalue = -maxvalue\n winner = board.winner()\n if winner == self.id:\n return maxvalue\n elif winner == opponent_id:\n return minvalue\n size_y = board.height\n size_x = board.width\n map_ = []\n num_to_connect = board.num_to_connect\n total_points = 0\n\n multiply_reachable = 1\n multiply_oddeven = 1\n # basically this function is calculating all the possible win positions\n # more pieces in a possible win position will be counted with more weights\n # a win position with X pieces in it will be counted as X^2 points\n # initialise the zones maps\n for i in range(size_y):\n map_.append([])\n for j in range(size_x):\n map_[i].append([])\n\n # Fill in the horizontal win positions\n for i in range(size_y):\n for j in range(size_x - num_to_connect + 1):\n points = 0\n self_pieces_count = 0\n opponent_pieces_count = 0\n for k in range(num_to_connect):\n if board.board[i][j + k] == opponent_id:\n opponent_pieces_count += 1\n elif board.board[i][j + k] == self.id:\n points += len(board.winning_zones[j+k][i])\n if (self.id == 1 and i % 2 == 1) or (self.id == 2 and i%2 == 0):\n points *= multiply_oddeven\n self_pieces_count += 1\n if self_pieces_count == 3 and opponent_pieces_count == 0:\n if j - 1 >= 0 and board.board[i][j + 3] == 0 and board.board[i][j - 1] == 0 \\\n and board.try_move(j + 3) == i and board.try_move(j - 1) == i:\n return maxvalue\n elif j + 4 < size_y and board.board[i][j + 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j + 4) == i and board.try_move(j) == i:\n return maxvalue\n else:\n for k in range(num_to_connect):\n if board.board[i][j + k] == 0 and board.try_move(j + k) == i:\n points *= multiply_reachable\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\n if j - 1 >= 0 and board.board[i][j + 3] == 0 and board.board[i][j - 1] == 0 \\\n and board.try_move(j + 3) == i and board.try_move(j - 1) == i:\n return minvalue\n elif j + 4 < size_y and board.board[i][j + 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j + 4) == i and board.try_move(j) == i:\n return minvalue\n # else:\n # for k in range(num_to_connect):\n # if board.board[i][j + k] == 0 and board.try_move(j + k) == i:\n # points *= -multiply_reachable\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\n total_points += points\n\n # Fill in the vertical win positions\n for i in range(size_x):\n for j in range(size_y - num_to_connect + 1):\n points = 0\n self_pieces_count = 0\n for k in range(num_to_connect):\n if board.board[j + k][i] == opponent_id:\n opponent_pieces_count += 1\n elif board.board[j + k][i] == self.id:\n points += len(board.winning_zones[i][j+k])\n if (self.id == 1 and (j+k) % 2 == 1) or (self.id == 2 and (j+k)%2 == 0):\n points *= multiply_oddeven\n self_pieces_count += 1\n points *= multiply_reachable\n # if opponent_pieces_count == 3 and self_pieces_count == 0:\n # points *= -1\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\n total_points += points\n\n # Fill in the forward diagonal win positions\n for i in range(size_y - num_to_connect + 1):\n for j in range(size_x - num_to_connect + 1):\n points = 0\n self_pieces_count = 0\n opponent_pieces_count = 0\n for k in range(num_to_connect):\n if board.board[i + k][j + k] == opponent_id:\n opponent_pieces_count += 1\n elif board.board[i + k][j + k] == self.id:\n points += len(board.winning_zones[j+k][i+k])\n if (self.id == 1 and (i+k) % 2 == 1) or (self.id == 2 and (i+k)%2 == 0):\n points *= multiply_oddeven\n self_pieces_count += 1\n if self_pieces_count == 3 and opponent_pieces_count == 0:\n if i - 1 >= 0 and j - 1 >= 0 and board.board[i + 3][j + 3] == 0 and board.board[i - 1][j - 1] == 0 \\\n and board.try_move(j + 3) == i + 3 and board.try_move(j - 1) == i - 1:\n return maxvalue\n elif i + 4 < size_y and j + 4 < size_x and board.board[i + 4][j + 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j + 4) == i + 4 and board.try_move(j) == i:\n return maxvalue\n else:\n for k in range(num_to_connect):\n if board.board[i + k][j + k] == 0 and board.try_move(j + k) == i + k:\n points *= multiply_reachable\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\n if i - 1 >= 0 and j - 1 >= 0 and board.board[i + 3][j + 3] == 0 and board.board[i - 1][j - 1] == 0 \\\n and board.try_move(j + 3) == i + 3 and board.try_move(j - 1) == i - 1:\n return minvalue\n elif i + 4 < size_y and j + 4 < size_x and board.board[i + 4][j + 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j + 4) == i + 4 and board.try_move(j) == i:\n return minvalue\n # else:\n # for k in range(num_to_connect):\n # if board.board[i + k][j + k] == 0 and board.try_move(j + k) == i + k:\n # points *= -multiply_reachable\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\n total_points += points\n\n # Fill in the backward diagonal win positions\n for i in range(size_y - num_to_connect + 1):\n for j in range(size_x - 1, num_to_connect - 1 - 1, -1):\n points = 0\n self_pieces_count = 0\n opponent_pieces_count = 0\n for k in range(num_to_connect):\n if board.board[i + k][j - k] == opponent_id:\n opponent_pieces_count += 1\n elif board.board[i + k][j - k] == self.id:\n points += len(board.winning_zones[j-k][i+k])\n if (self.id == 1 and (i+k) % 2 == 1) or (self.id == 2 and (i+k)%2 == 0):\n points *= multiply_oddeven\n self_pieces_count += 1\n if self_pieces_count == 3 and self_pieces_count == 0:\n if board.board[i + 3][j - 3] == 0 and board.board[i - 1][j + 1] == 0 \\\n and board.try_move(j - 3) == i + 3 and board.try_move(j + 1) == i - 1:\n return maxvalue\n elif i + 4 < size_y and j - 4 >= 0 and board.board[i + 4][j - 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j - 4) == i + 4 and board.try_move(j) == i:\n return maxvalue\n else:\n for k in range(num_to_connect):\n if board.board[i + k][j - k] == 0 and board.try_move(j - k) == i + k:\n points *= multiply_reachable\n\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\n if board.board[i + 3][j - 3] == 0 and board.board[i - 1][j + 1] == 0 \\\n and board.try_move(j - 3) == i + 3 and board.try_move(j + 1) == i - 1:\n return minvalue\n elif i + 4 < size_y and j - 4 >= 0 and board.board[i + 4][j - 4] == 0 and board.board[i][j] == 0 \\\n and board.try_move(j - 4) == i + 4 and board.try_move(j) == i:\n return minvalue\n # else:\n # for k in range(num_to_connect):\n # if board.board[i + k][j - k] == 0 and board.try_move(j - k) == i + k:\n # points *= -multiply_reachable\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\n total_points += points\n return total_points","def make_move(self, state, actions):\n\n # Here we are doing an exploration so don't update scores\n # for selected state\n if random.random() < self._exploration:\n action = random.choice(actions)\n self._last_state = self._get_state_for_action(state, action)\n return action\n\n # So we are not doing exploration, here we want to update it\n max_selection = None\n for a, s in self._states_from_actions(state, actions).items():\n state_value = self._state_values[s]\n if max_selection is None or state_value > self._state_values[max_selection[1]]:\n max_selection = (a, s)\n\n if self._last_state is not None:\n self._update_state_value(self._last_state, max_selection[1])\n\n self._last_state = max_selection[1]\n return max_selection[0]","def DFS(self, nDepth, treenode, state):\n \n visited = []\n visited.insert(0, (state, treenode))\n \n for index in range(0, nDepth-1): \n actions = self.priorProb(state)\n treenode.expansion(actions)\n treenode.updateU_value(actions)\n treenode, action = treenode.selection() \n state = state.do_move(action).copy()\n visited.insert(0, (state, treenode)) \n \n for index in range(0, len(visited)-1): \n if(visited[index][1].isLeaf() == True):\n value = self.leafEvaluation(visited[index][0])\n else: \n value = visited[index][1].backUp(value)\n visited[-1][1].updateQ_value(value)\n visited[-1][1].updateVisits()\n return visited[-1][1]","def main():\n\n N, k = map(int, sys.stdin.readline().split())\n initial_state = tuple( map(int, sys.stdin.readline().split()) )\n final_state = tuple( map(int, sys.stdin.readline().split()) )\n\n init_path = [initial_state]\n all_paths = [init_path]\n\n visited_states = set()\n visited_states.add(initial_state)\n\n while all_paths:\n popped_path = all_paths.pop(0)\n current_state = popped_path[-1]\n\n if current_state == final_state:\n solution = popped_path\n break\n\n for state, action in generate_moves(current_state, N, k).items():\n if state not in visited_states:\n visited_states.add(state)\n new_path = popped_path + [action,state]\n all_paths.append(new_path)\n\n display_solution(solution)","def choose_epsilon_greedy(self, state: Tuple[int, ...], valid_actions: Tuple[int, ...]) -> int:\n if random.random() < self.__epsilon:\n return self.choose_uniform(valid_actions)\n return self.choose_greedy(state, valid_actions)","def get_every_nth_state(n=10):\n loops = int(50/n)\n state_list = []\n for i in range(loops):\n state_list.append(states[(i+1)*n-1])\n return state_list","def get_action(self,state):\n \n q_values = self.__network.predict(state[None])[0]\n \n ###YOUR CODE\n if np.random.rand()<self.epsilon:\n return np.random.choice(self.n_actions)\n return np.argmax(q_values)","def get_action(self, state):\n\n best_action = None\n best_value = -np.inf\n actions = [0, 1, 2, 3] # left, down, right, up\n for a in actions:\n row = state // self.edge\n col = state % self.edge\n # print (row, col)\n if a == 0:\n col = max(col-1, 0)\n elif a == 1:\n row = min(row+1, self.edge-1)\n elif a == 2:\n col = min(col+1, self.edge-1)\n elif a == 3:\n row = max(row-1, 0)\n # print (row, col)\n\n new_state = row * self.edge + col\n # print (new_state)\n if (self.values[new_state] > best_value or new_state == self.num_states-1): #goal\n best_value = 1.0 if new_state == self.num_states-1 else self.values[new_state]\n best_action = a\n return best_action","def solveOneStep(self):\n ### Student code goes here\n # Mark this move as explored\n self.visited[self.currentState] = True\n\n # Get move to make\n movables = self.gm.getMovables()\n # print(\"EXPLORING GAME STATE \" + str(self.gm.getGameState()) + \"---------------------------------------------------------\")\n to_move = self.currentState.nextChildToVisit # movables index\n # print(\"depth \", self.currentState.depth)\n\n # Return if done\n if self.currentState.state == self.victoryCondition:\n # print(\"DONE\")\n return True\n\n while to_move < len(movables):\n # Make the move\n movable_statement = movables[to_move]\n # print(\"implementing move \", movable_statement)\n self.gm.makeMove(movable_statement)\n\n # Create a new state with this move made\n new_state = self.gm.getGameState()\n\n # Find out if this state has already been explored\n visited = False\n for visited_state in self.visited.keys():\n if visited_state.state == new_state:\n visited = True\n\n # If the new state hasn't been visited then add it as a child then move down to this child\n if not visited:\n new_gs = GameState(new_state, self.currentState.depth + 1, movable_statement)\n new_gs.parent = self.currentState\n self.currentState.children.append(new_gs)\n self.currentState.nextChildToVisit = to_move + 1\n self.currentState = new_gs\n break\n\n # Else skip this state and try going to the next movable statement\n else:\n # print(\"SKIP THIS STATE\")\n self.gm.reverseMove(movable_statement)\n to_move += 1\n\n # Went all the way down to a leaf, backtrack\n if (to_move >= len(movables)):\n self.gm.reverseMove(self.currentState.requiredMovable)\n self.currentState = self.currentState.parent\n\n return False","def get_random_depth_sample(n=8, depths=list(range(2,26,2)), num_samples=100):\n\n def get_states(start):\n frontier = [start]\n frontier_set = {start}\n explored = set()\n\n states = [False for _ in range(len(depths))]\n while not all(states):\n node = frontier.pop(0)\n frontier_set.remove(node)\n explored.add(node)\n\n children = node.get_children()\n\n # It's necessary to shuffle children to get a truly random sample; otherwise, the first child (always\n # produced from the parent by the same action) produced at a certain depth will always be selected,\n # and children produced by other actions will never be selected\n shuffle(children)\n\n for child in children:\n if child not in frontier_set and child not in explored:\n frontier_set.add(child)\n frontier.append(child)\n child.path_cost = node.path_cost+1\n index = depths.index(child.path_cost) if child.path_cost in depths else None\n if index is not None and not states[index]:\n states[index] = {'start': start.sequence, 'end': child.sequence}\n\n return states\n\n depth_sample = [[] for depth in range(len(depths))]\n\n for _ in range(num_samples):\n start = list(range(1,n+2))\n shuffle(start)\n start = PuzzleState(start, path_cost=0)\n\n states = get_states(start)\n print('\\rSet ' + str(_+1) + ' of ' + str(num_samples) + ' complete', end='', flush=True)\n list(map(list.append, depth_sample, states))\n\n return depth_sample","def simulate_tree(self, current_state):\n sim_trace = [(current_state, None)]\n while not self.state_manager.game_over():\n performed_action = self.make_move(current_state, is_rollout=False)\n current_state, subtree_existed_prev = self.get_child(current_state, performed_action)\n sim_trace.append((current_state, performed_action))\n #the case where one has reached a leaf node\n if subtree_existed_prev == False:\n break\n return sim_trace, current_state","def run_each_move(root: State, state: GameState) -> State:\n for m, base_move in enumerate(root.moves):\n for _ in range(50):\n base_move = expand(base_move)\n outcome, _ = MCTS(base_move, state, False, 0)\n if outcome == state.player:\n base_move.winner()\n else:\n base_move.loser()\n root.attempts += 1\n root.moves[m] = base_move\n print(\"Move: \", base_move.idx)\n print(\" Wins: \", base_move.wins)\n update_ucbs(root)\n return root","def choose_action(self, d, name):\n # some initializations\n current_state = d.state\n caution = False\n confidence = False\n buy_rules = [0,0,0,0]\n next_vec = self.transition_matrix[name][int(current_state)]\n num_undesirable_states = (self.trigger_states[name]+1)\n num_desirable_states = (next_vec.size-num_undesirable_states)\n if num_undesirable_states<5:\n left_basket_max = 2\n else:\n left_basket_max = num_undesirable_states//3\n if num_desirable_states<5:\n right_basket_min = next_vec.size-2\n else:\n right_basket_min = next_vec.size-num_undesirable_states//3\n # check if rules are satisfied\n # rule-1\n m1 = np.max(next_vec[:self.trigger_states[name]+1])\n m1_idx = np.argmax(next_vec[:self.trigger_states[name]+1])\n m2 = np.max(next_vec[self.trigger_states[name]+1:])\n m2_idx = np.argmax(next_vec[self.trigger_states[name]+1:])+\\\n next_vec[:self.trigger_states[name]+1].size\n if m2-m1>=0.1: # threshold\n #print('Rule #1 satisfied.')\n buy_rules[0]=1\n # rule-2\n if np.sum(next_vec[self.trigger_states[name]+1:])-\\\n np.sum(next_vec[:self.trigger_states[name]+1])>=0.25: # threshold\n #print('Rule #2 satisfied.')\n buy_rules[1]=1\n # rule-3 \n if m1_idx<left_basket_max: \n if buy_rules[0]!=1:\n caution=True\n #print('Predicted state is very undesirable.')\n # rule-3\n if m2_idx>=right_basket_min:\n if buy_rules[0]==1:\n confidence=True\n #print('Predicted state is very desirable.')\n if d.MACD>d.signal_line:\n #print('Rule #3 satisfied.')\n buy_rules[2] = True\n # sum of k most undesirable vs k most desirable\n temp_1 = np.sort(next_vec[self.trigger_states[name]+1:])\n temp_2 = np.sort(next_vec[:self.trigger_states[name]+1])\n size = 3\n if temp_1.size<size or temp_2.size<size:\n size = min(temp_1.size, temp_2.size)\n k1 = np.sum(temp_1[::-size])\n k2 = np.sum(temp_2[::-size])\n if k1-k2>0.25:\n #print('Rule #4 satisfied.')\n buy_rules[3] = True\n # finally, make a call using the rules\n if confidence or sum(buy_rules)>=3:\n return 'buy'\n elif caution or (buy_rules[0]==0 and sum(buy_rules)<=2 and m1-m2>0.05):\n return 'sell'\n else:\n return 'hold'","def getAction(self, gameState):\n \"*** YOUR CODE HERE ***\"\n \n\n def max_value(gameState,depth):\n if depth==0 or gameState.isWin() or gameState.isLose():\n return (self.evaluationFunction(gameState), None)\n\n moves=gameState.getLegalActions()\n score_list = []\n for i in moves:\n current_state= self.index\n next_state = gameState.generateSuccessor(current_state,i) \n value, action= min_value(next_state,1, depth) \n score_list.append(value)\n\n max_score=max(score_list)\n score_len= len(score_list)\n for i in range(score_len):\n if score_list[i] == max_score:\n best_state = [i]\n chosenIndex = random.choice(best_state)\n return max_score,moves[chosenIndex]\n\n def min_value(gameState,agent, depth): \n if depth==0 or gameState.isWin() or gameState.isLose():\n return (self.evaluationFunction(gameState), None)\n \n moves=gameState.getLegalActions(agent) \n score_list=[]\n \n for i in moves:\n next_state = gameState.generateSuccessor(agent,i)\n ghost_no = gameState.getNumAgents()\n if(agent == ghost_no-1):\n value, action = max_value(next_state,(depth-1))\n score_list.append(value)\n else:\n \n value, action = min_value(next_state,agent+1,depth)\n score_list.append(value)\n\n min_score=min(score_list)\n score_len = len(score_list)\n for j in range(score_len):\n if score_list[j] == min_score:\n worst_state = [j]\n chosenIndex = random.choice(worst_state)\n return min_score, moves[chosenIndex]\n \n final_score,final_action= max_value(gameState,self.depth)\n \n return final_action\n util.raiseNotDefined()","def choose(self, state: State) -> State:","def UCT(rootstate, itermax, verbose = False):\n\n rootnode = Node(state=rootstate)\n\n for i in range(itermax):\n node = rootnode\n state = rootstate.Clone()\n\n # Select\n while node.untriedMoves == [] and node.childNodes != []: # node is fully expanded and non-terminal\n\n node = node.UCTSelectChild()\n state.DoMove(node.move)\n\n\n # Expand\n if node.untriedMoves != []: # if we can expand (i.e. state/node is non-terminal)\n m = random.choice(node.untriedMoves)\n state.DoMove(m)\n node = node.AddChild(m, state) # add child and descend tree\n\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\n while state.GetMoves() != []: # while state is non-terminal\n state.DoMove(random.choice(state.GetMoves()))\n\n # Backpropagate\n while node != None: # backpropagate from the expanded node and work back to the root node\n node.Update(state.GetResult(node.whosemove))\n node = node.parentNode\n\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[-1].move # return the move that was most visited","def change(a, b, state, target):\n # a = random.randrange(1, 199, 2)\n # b = random.randrange(1, 199, 2)\n new_state = ''\n for i in range(len(state)):\n if i == a:\n new_state += b\n continue\n new_state += state[i]\n # print(new_state)\n res = solving(int(new_state[0]), int(new_state[2]), new_state[1])\n for i in range(2, len(new_state) - 2, 2):\n res = solving(res, int(new_state[i + 2]), new_state[i + 1])\n # print(\"Distance from target: \", target - res)\n return new_state, abs(target - res)","def calc_nearest_state(self): # TODO: Check if we need here state, instead of self.state\n self.stateC = self.toConceptual(self.state)\n CTP, winners = self.find_winner()\n\n state_name = self.find_TPname(filleridx=winners)\n binding = self.find_symBinding(filleridx=winners)\n state_num = self.find_TPnum(stateName=state_name)\n TP_state = self.TP.matmul(fortran_reshape(CTP, (torch.numel(CTP), 1)))\n Cdist = torch.norm(CTP - self.stateC, p='fro') # Frobenius Norm\n Sdist = self.L2norm(TP_state - self.state)\n TP_h = self.calc_harmony(state=TP_state)\n\n return TP_state, winners, state_name, binding, Cdist, Sdist, state_num, TP_h","def update_state(state, Y_next):\n if max(Y_next) > state.best_value + 1e-3 * fabs(state.best_value):\n state.success_counter += 1\n state.failure_counter = 0\n else:\n state.success_counter = 0\n state.failure_counter += 1\n\n if state.success_counter == state.success_tolerance: # Expand trust region\n state.length = min(2.0 * state.length, state.length_max)\n state.success_counter = 0\n elif state.failure_counter == state.failure_tolerance: # Shrink trust region\n state.length /= 2.0\n state.failure_counter = 0\n\n state.best_value = max(state.best_value, max(Y_next).item())\n if state.length < state.length_min:\n state.restart_triggered = True\n\n return state","def move_length_5(self, move, new_state):\n # First consider the move on 6 corners\n new1_state = new_state\n if move in [\"A\", \"B\", \"O\", \"U\", \"T\", \"Y\"]:\n corners = [[\"A\", 1, 2, 2, 5, 9, 14, 0, 3, 7, 12, 18, 24, 17],\n [\"B\", 0, 2, 4, 8, 13, 19, 12, 3, 6, 10, 15, 20, 1],\n [\"O\", 20, 13, 9, 5, 2, 0, 0, 15, 16, 17, 18, 19, 10],\n [\"U\", 14, 13, 21, 22, 23, 24, 11, 1, 3, 6, 10, 15, 1],\n [\"T\", 24, 9, 1, 4, 8, 13, 12, 14, 15, 16, 17, 18, 10],\n [\"Y\", 19, 9, 20, 21, 22, 23, 11, 0, 3, 7, 12, 18, 17]]\n for i in corners:\n new1_state = self.loop51(move, new_state, i)\n\n if move in [\"C\", \"E\", \"J\", \"N\", \"V\", \"X\"]:\n # then consider the move on the middle1 of each side\n middle1 = [[\"C\", 3, 4, 4, 0, 5, 9, 14, 0, 6, 11, 17, 23, 16],\n [\"E\", 2, 3, 4, 1, 8, 13, 19, 12, 7, 11, 16, 21, 3],\n [\"J\", 15, 21, 14, 10, 11, 12, 13, 8, 0, 2, 5, 14, 0],\n [\"N\", 18, 23, 7, 1, 4, 8, 19, 12, 9, 10, 11, 12, 8],\n [\"V\", 9, 15, 14, 20, 22, 23, 24, 11, 4, 7, 11, 16, 3],\n [\"X\", 13, 18, 7, 20, 21, 22, 24, 11, 2, 6, 11, 17, 16]]\n for i in middle1:\n new1_state = self.loop52(move, new_state, i)\n return StonehengeState(not self.p1_turn, new1_state.length,\n new1_state.letters, new1_state.claim)","def chooseAction(self,state):\r\n #generate float btwn 0-1\r\n choice = random.random()\r\n \r\n #choose according to that number\r\n if choice > self.epsilon:\r\n return(self.maxQ(state)[1])\r\n else:\r\n #choose randomly\r\n return(self.actions[random.randrange(0,len(self.actions))])","def chooseAction(self,state):\r\n #generate float btwn 0-1\r\n choice = random.random()\r\n \r\n #choose according to that number\r\n if choice > self.epsilon:\r\n return(self.maxQ(state)[1])\r\n else:\r\n #choose randomly\r\n return(self.actions[random.randrange(0,len(self.actions))])","def miniMax(self,state,depth=0):\n \n #print(\"NextState (depth \"+str(depth)+\"):\")\n #print(\"Action: \"+state.get_action())\n \n if state in self.__explored:\n return self.__explored[state.get_hashable_state()]\n \n if state.is_end_state() or depth >= (self.__max_depth - 1):\n self.__explored[state.get_hashable_state()] = state.get_utility_value()\n return state.get_utility_value() #Return terminal state's utility value\n \n is_max_turn = state.get_max_turn()\n childList = state.get_successors()\n \n if is_max_turn:\n utility = float(\"-inf\")\n for c in childList:\n utility = max(utility,self.miniMax(c, depth+1))\n self.__explored[state.get_hashable_state()] = utility\n return utility\n else:\n utility = float(\"inf\")\n for c in childList:\n utility = min(utility,self.miniMax(c, depth+1))\n self.__explored[state.get_hashable_state()] = utility\n return utility","def do(num_range,i, sub):\n\t\n\t#Converting the vector in all_states_explored to a matrix. \n\tmatrix = pMatrix.make_matrix(all_states_explored[i])\n\t\n\t#Tree for currrent iteration.\n\ttree = pMatrix.create_tree(matrix, num_range, sub) \n\t\n\t#Adding the tree to all_trees.\n\tall_trees.append(tree)\n\t\n\t#Calculating the number of states in current iteration.\n\tnum_states = tree.get_num_states()\n\t\t\n\t#Adding total number of states to all_total_states. \n\tall_total_states.append(num_states) \n\t\n\t#Finding results for each iteration.\n\tr = pMatrix.main(matrix, num_range, sub)\n\t\t\n\t#Adding results for current iteration to all_results. \n\tall_results.append(r)\n\t\n\t#Finding any new, previously unseen state and adding it to all_states_explored list.\n\tfind_new_states(tree)\n\t\n\t#Adding super_states\n\tupdate_super_states(tree)","def __call__(self, state, q_values):\n\n if self.policy_type == \"greedy\":\n is_greedy = True\n else:\n is_greedy = random.uniform(0, 1) > self.epsilon\n\n if is_greedy :\n # choose greedy action\n index_action = np.argmax(q_values[state])\n else:\n # get a random action\n index_action = random.randint(0,3)\n\n return actions_dict[index_action]","def __init__(self, mdp, discount = 0.9, iterations = 100):\n self.mdp = mdp\n self.discount = discount\n self.iterations = iterations\n self.values = util.Counter() # A Counter is a dict with default 0\n \n \"*** YOUR CODE HERE ***\"\n # OUR CODE HERE\n #Note: I think we should use the util.Counter thing?\n for times in range(0, iterations):\n #values from previous iteration so we don't update over them while iterating\n prevVals = self.values.copy()\n #iterate through all states\n for state in mdp.getStates():\n #will store the action-value for the iteration\n value = util.Counter()\n for action in mdp.getPossibleActions(state):\n for transitionState, probability in mdp.getTransitionStatesAndProbs(state, action):\n #expected value, probability * reward for the state with the discount * reward\n value[action] += probability * (mdp.getReward( state, action, transitionState) + discount * prevVals[transitionState])\n #update the values to the new value from the iteration\n #the .argMax() function returns the one with the largest value\n self.values[state] = value[value.argMax()]\n # END OUR CODE","def __init__(self,\n game,\n exploration_strategy='uniform-exhaustive',\n confidence_method='ucb-standard',\n delta=0.01,\n ucb_eps=0,\n per_payoff_confidence=True,\n time_dependent_delta=False):\n self.exploration_strategy = exploration_strategy\n self.confidence_method = confidence_method\n self.ucb_eps = ucb_eps\n self.G = game # pylint: disable=invalid-name\n self.per_payoff_confidence = per_payoff_confidence\n self.time_dependent_delta = time_dependent_delta\n if self.per_payoff_confidence:\n self._delta = delta\n else:\n self._delta = delta / (\n self.G.n_players *\n functools.reduce(operator.mul, self.G.strategy_spaces, 1))\n\n # Compute the graph\n self.V = list( # pylint: disable=invalid-name\n itertools.product(*[range(smax) for smax in self.G.strategy_spaces]))\n self.E = [] # pylint: disable=invalid-name\n for v in self.V:\n adj_strats = [\n list(range(v[k] + 1, self.G.strategy_spaces[k]))\n for k in range(self.G.n_players)\n ]\n for k in range(self.G.n_players):\n for new_s in adj_strats[k]:\n second_vertex = list(v)\n second_vertex[k] = new_s\n second_vertex = tuple(second_vertex)\n self.E.append((v, second_vertex))\n self.count_history = {v: [] for v in self.V}\n self.total_interactions = 0","def predict(self, state):\n if self.phase is None or self.device is None:\n raise AttributeError('Phase, device attributes have to be set!')\n if self.phase == 'train' and self.epsilon is None:\n raise AttributeError('Epsilon attribute has to be set in training phase')\n\n if self.reach_destination(state):\n return ActionXY(0, 0) if self.kinematics == 'holonomic' else ActionRot(0, 0)\n if self.action_space is None:\n self.build_action_space(state.robot_state.v_pref)\n\n probability = np.random.random()\n if self.phase == 'train' and probability < self.epsilon:\n max_action = self.action_space[np.random.choice(len(self.action_space))]\n else:\n max_action = None\n max_value = float('-inf')\n max_traj = None\n\n if self.do_action_clip:\n state_tensor = state.to_tensor(add_batch_size=True, device=self.device)\n action_space_clipped = self.action_clip(state_tensor, self.action_space, self.planning_width)\n else:\n action_space_clipped = self.action_space\n\n for action in action_space_clipped:\n state_tensor = state.to_tensor(add_batch_size=True, device=self.device)\n next_state = self.state_predictor(state_tensor, action)\n max_next_return, max_next_traj = self.V_planning(next_state, self.planning_depth, self.planning_width)\n reward_est = self.estimate_reward(state, action)\n value = reward_est + self.get_normalized_gamma() * max_next_return\n if value > max_value:\n max_value = value\n max_action = action\n max_traj = [(state_tensor, action, reward_est)] + max_next_traj\n if max_action is None:\n raise ValueError('Value network is not well trained.')\n\n if self.phase == 'train':\n self.last_state = self.transform(state)\n else:\n self.traj = max_traj\n\n return max_action","def N_states_for_learner(self):\n idx_max = []\n limits = 50, 2*_math.pi, 50, 50, 50, 50, 50, 50, 50\n for idx, limit in enumerate(limits):\n test = [0 for i in xrange(len(limits))]\n check = _arange(-limit,limit,limit/1000.)\n maxi = 0\n for v in check:\n test[idx]=v\n ret = self._state_index(*test)\n maxi = max((maxi, ret[idx]))\n idx_max.append(maxi)\n\n return tuple([idx+1 for idx in idx_max])","def get_move(self, state):\n # this method should only be called when self is real root.,so that's here where we can should use mutiprocess\n if self._root.is_leaf(): # no expanded children yet\n action_probs, _ = self._policy(state)\n self._root.expand(action_probs)\n\n the_children = self._root._children\n i = 0\n sorted_children = sorted(the_children.items(), key=lambda act_node: act_node[1].get_value(self._c_puct))\n for child_node in sorted_children:\n i += 1\n child_tree = MCTS(policy_value_fn,root=child_node[1])\n state_copy = copy.deepcopy(state)\n state_copy.do_move(child_node[0])\n visits_count = 0\n for j in range(0,relu(1200-i*20),10): # at least run one time\n child_tree._playout(copy.deepcopy(state_copy))\n visits_count += 1\n self._root.update(-child_tree.get_root_node().last_leafvalue,visits_count=visits_count) # update real root\n child_tree.get_root_node().set_parent(self._root) # to link the sub tree\n\n '''\n for n in range(self._n_playout):\n # get top n (assumed to be 6) nodes from children\n # step1 let all children of root have chance to run in parallel\n # adjust the round count of children by value\n if n%6 == 0:\n the_children = self._root._children\n top_n = sorted(the_children.items(),key=lambda act_node: act_node[1].get_value(self._c_puct))[:6]\n for child_node in top_n:\n # child_tree = MCTS(policy_value_fn,copy.deepcopy(child_node)) # use copy because we will use it in multiprocess\n child_tree = MCTS(policy_value_fn,\n child_node) \n state_copy = copy.deepcopy(state)\n state_copy.do_move(child_node[0])\n child_tree._playout(state_copy)\n self._root.update(-child_tree.get_root_node().last_leafvalue) # update real root\n child_tree.get_root_node().set_parent(self._root) # to link the sub tree\n # self._root.get_children()[child_node[0]] = child_tree.get_root_node() # copy sub tree\n '''\n\n '''\n return max(self._root._children.items(),\n # key=lambda act_node: act_node[1].get_visits())[0]\n key=lambda act_node: act_node[1].get_value(self._c_puct))[0]\n '''\n\n for n in range(300):\n state_copy = copy.deepcopy(state)\n self._playout(state_copy)\n return max(self._root._children.items(),\n key=lambda act_node: act_node[1].get_value(self._c_puct))[0]","def solve(self, state, times):","def _greedy(self, state):\n\n node = self.mcts_head\n if self.verbose > 1:\n logger.debug(f\"Starting greedy algorithm.\")\n\n while not node.terminal:\n # Parse current state\n this_state, total_reward, terminal = self._parse_path(state, node.path)\n node.set_terminal(terminal)\n if self.verbose > 1:\n logger.debug(f\" Analyzing node {node.path}\")\n\n # Expand\n if not node.terminal and not node.children:\n actions = self._find_legal_actions(this_state)\n step_rewards = [self._parse_action(action, from_which_env=\"sim\") for action in actions]\n if self.verbose > 1:\n logger.debug(f\" Expanding: {len(actions)} legal actions\")\n node.expand(actions, step_rewards=step_rewards)\n\n # If terminal, backup reward\n if node.terminal:\n if self.verbose > 1:\n logger.debug(f\" Node is terminal\")\n if self.verbose > 1:\n logger.debug(f\" Backing up total reward {total_reward}\")\n node.give_reward(self.episode_reward + total_reward, backup=True)\n\n # Debugging -- this should not happen\n if not node.terminal and not node.children:\n logger.warning(\n f\"Unexpected lack of children! Path: {node.path}, children: {node.children.keys()}, legal actions: {self._find_legal_actions(this_state)}, terminal: {node.terminal}\"\n )\n node.set_terminal(True)\n\n # Greedily select next action\n if not node.terminal:\n action = node.select_greedy()\n node = node.children[action]\n\n if self.verbose > 0:\n choice = self.mcts_head.select_best(mode=\"max\")\n self._report_decision(choice, state, \"Greedy\")","def mcts_uct(ei, iterations, approach):\n clear_states()\n root = generateState(None, None, ei, approach)\n for i in xrange(iterations):\n child = root\n while not child.terminal() and (not child.experimentInstance.AllTestsReached()):\n if not child.fully_expanded():\n child = child.expand()\n break\n else:\n child = child.best_child()\n cost = child.simulation()\n update_parents(child, cost)\n return root.best_action(c=0)","def act(self, state):\n if np.random.rand() <= self.epsilon:\n return random.randrange(self.n_actions)\n act_values = self.model.predict(state)\n return np.argmax(act_values[0])","def priorProb(self, state):\n actions = []\n for i in range(0, 10):\n actions.append(((i, i+1), random.uniform(0, 1))) \n \n return actions","def getTransitionStatesAndProbs(self, state, action = None):\n # may move ball\n ball = state[0][:2]\n if action == None:\n ballVelocity = state[0][2:]\n else:\n ballVelocity = (0, 0)\n\n keepers = list(self.getKeepers(state))\n takers = list(self.getTakers(state))\n\n chasers = sorted(keepers, key=lambda keeper: util.getPointVectorDistance(keeper, ball, ballVelocity))\n # most closest agent, possess the ball, or go to the ball \n if self.weHaveBall(state):\n # j has the ball, its transition depends on the action\n if action[0] == 'hold':\n pass\n elif action[0] == 'pass':\n # pass the ball to a teammate\n rand = util.randomVector(0.1)\n target = keepers[action[1]]\n diff = util.getDirection(keepers[0], (target[0] + rand[0], target[1] + rand[1]))\n ballVelocity = (self.ballSpeed * diff[0], self.ballSpeed * diff[1])\n else:\n raise Exception('Unknown action')\n else:\n # j should go to the ball\n chasers[0] = self.moveTowards(chasers[0], ball)\n\n # other agents get open for a pass\n for i in xrange(1, len(chasers)):\n # concretely, this agent goes to a least congested place\n chasers[i] = self.moveTowards(chasers[i], self.getLeastCongestedLoc(state, chasers[i]))\n keepers = sorted(chasers, key=lambda keeper: util.getDistance(keeper, ball))\n \n for i in xrange(2):\n takers[i] = self.moveTowards(takers[i], ball)\n for i in xrange(2, len(takers)):\n takers[i] = self.moveTowards(takers[i], keepers[1])\n takers = sorted(takers, key=lambda taker: util.getDistance(taker, keepers[0]))\n \n newBall = (ball[0] + ballVelocity[0], ball[1] + ballVelocity[1],\\\n ballVelocity[0], ballVelocity[1])\n newState = [newBall] + keepers + takers\n return [(tuple(newState), 1)]","def predict_next_state_gt(self, states, actions):\n # TODO: write your code here\n\n # return [self.env.get_nxt_state(states[i], actions) for i in range(self.num_particles)]\n return np.array([[self.env.get_nxt_state(states[j][i], actions[j]) for i in range(self.num_particles)] for j in range(self.popsize)])","def solveOneStep(self):\n ### Student code goes here\n if self.first_step == False:\n self.first_step = True\n if self.solveOneStep():\n return True\n if self.queue:\n self.gm_init()\n ele = self.queue.get()\n #print (len(ele))\n state = ele[0]\n premoves = ele[1]\n\n for m in premoves:\n self.gm.makeMove(m)\n if state.state == self.victoryCondition:\n return True\n self.visited[state] = True\n print(\"CURRENTSTATE:\")\n print(self.gm.getGameState())\n print(\"*******\")\n moves = self.gm.getMovables()\n for m in moves:\n self.gm.makeMove(m)\n if (((state.parent is not None) and (self.gm.getGameState() == state.parent.state))) or GameState(self.gm.getGameState(), 0, None) in self.visited:\n self.gm.reverseMove(m)\n continue\n self.visited[GameState(self.gm.getGameState(), 0, None)] = True\n new_pmv = [i for i in premoves]\n new_pmv.append(m)\n next_state = GameState(self.gm.getGameState(), state.depth+1, m)\n next_state.parent = state\n state.children.append(next_state)\n self.queue.put([next_state, new_pmv])\n self.gm.reverseMove(m)\n self.currentState = state\n\n #for i in range(len(premoves)-1, -1, -1):\n # mv = premoves[i]\n # self.gm.reverseMove(mv)\n return False","def _playout(self, state):\n node = self._root\n while (1):\n if node.is_leaf():\n break\n # Greedily select next move.\n action, node = node.select(self._c_puct)\n state.do_move(action)\n\n # Evaluate the leaf using a network which outputs a list of\n # (action, probability) tuples p and also a score v in [-1, 1]\n # for the current player.\n action_probs, leaf_value = self._policy(state)\n # Check for end of game.\n end, winner = state.game_end()\n if not end:\n node.expand(action_probs)\n else:\n # for end state，return the \"true\" leaf_value\n if winner == -1: # tie\n leaf_value = 0.0\n else:\n leaf_value = (\n 1.0 if winner == state.get_current_player() else -1.0\n )\n\n # Update value and visit count of nodes in this traversal.\n node.update_recursive(-leaf_value)","def getAction(self, gameState):\n \"*** YOUR CODE HERE ***\"\n # Again, we use the fundamental foundation built in Q2 for Q4, however here we modify our minimizer function\n # to serve the purpose of finding the expected value\n actionList = gameState.getLegalActions(0)\n pacmanAgentIndex = 0\n ghostAgentIndices = list(range(1,gameState.getNumAgents())) # List of each agent index for looping\n count = util.Counter()\n agentEnd = gameState.getNumAgents()-1 # Last agent in the list\n def maximizer(curState, agentIndex, depth):\n\n ghostActions = curState.getLegalActions(agentIndex)\n maxDepth = self.depth # Quantifying the end of the tree so we know when we reached a leaf node\n weight = -99999999 # Worst case starting value to be changed in the code\n if depth == maxDepth: # If we are at a leaf node\n return self.evaluationFunction(curState) # evaluate the state of this leaf node\n # Otherwise, we progress the tree until the above condition is reached\n if len(ghostActions) != 0:\n for x in ghostActions:\n if weight >= minimizer(curState.generateSuccessor(agentIndex, x), agentIndex + 1, depth):\n weight = weight\n else:\n weight = minimizer(curState.generateSuccessor(agentIndex, x), agentIndex + 1, depth)\n return weight\n else:\n # if there are no legal actions left then evaluate at the last known state\n return self.evaluationFunction(curState)\n\n def minimizer(curState, agentIndex, depth):\n ghostActions = curState.getLegalActions(agentIndex)\n weight = 0 # Starting value of zero to be incremented below\n if len(ghostActions) != 0:\n if agentIndex == agentEnd: # If we've reached the last ghost, we maximise\n for x in ghostActions: # For each legal action in the current position\n temp = (float(1.0) / len(ghostActions))*maximizer(curState.generateSuccessor(agentIndex, x), pacmanAgentIndex, depth+1)\n weight = weight + temp\n else: # Otherwise, we continue to minimize\n for x in ghostActions: # For each legal action in the current position\n temp = (float(1.0) / len(ghostActions))*minimizer(curState.generateSuccessor(agentIndex, x), agentIndex+1, depth)\n weight = weight + temp\n return weight\n else:\n # if there are no legal actions left then evaluate at the last known state\n return self.evaluationFunction(curState)\n\n # Executing the minimizer for all possible actions\n for x in actionList:\n tempState = gameState.generateSuccessor(pacmanAgentIndex,x)\n count[x] = minimizer(tempState,1,0)\n # print('HELLO THERE')\n # print(count)\n return count.argMax()"],"string":"[\n \"def __call__(self, s, n=1000):\\n\\n root = StateNode(None, s, self.game)\\n \\n if root.parent is not None:\\n raise ValueError(\\\"Root's parent must be None.\\\")\\n \\n for _ in range(n):\\n #selection\\n node = _get_next_node(root, self.tree_policy)\\n #simulation\\n node.reward = self.default_policy(node)\\n #print(node.reward)\\n #back\\n self.backup(node)\\n \\n root.reset(copy.deepcopy(self.game_bak))\\n \\n #for i in root.children:\\n # print(root.children[i].__dict__)\\n # for j in root.children[i].children:\\n # print(root.children[i].children[j].__dict__)\\n # print(\\\"=======\\\")\\n return rand_max(root.children.values(), key=lambda x: x.q).action, rand_max(root.children.values(), key=lambda x: x.q).q\",\n \"def run_simulation(self, state):\\n \\\"*** YOUR CODE HERE ***\\\"\\n player = 0\\n visited_states = [(player, state)]\\n depth_limited = self.depth != -1\\n depth = self.depth\\n expand = True\\n while not visited_states[-1][1].isWin() and not visited_states[-1][1].isLose():\\n if depth_limited and depth == 0: break\\n state = self.UCB1(state, player) # Selection & Simulation\\n if expand and state not in self.plays: # Expansion\\n expand = False\\n self.plays[state] = 0\\n self.wins[state] = 0\\n visited_states.append((player, state))\\n player = (player + 1) % state.getNumAgents()\\n if not expand and depth_limited and player == 0: depth -= 1\\n \\n for player, state in visited_states:\\n if state in self.plays: # Not simulated nodes\\n self.plays[state] += 1\\n eval = self.evaluationFunction(visited_states[-1][1])\\n if depth_limited:\\n if player == 0: self.wins[state] += eval\\n if player != 0: self.wins[state] -= eval\\n else:\\n if player == 0: self.wins[state] += eval\\n if player != 0: self.wins[state] += (1 - eval)\",\n \"def find_best_move(state) -> None:\\n C: float = 2 ** 0.5 # Start with sqrt(2)\\n idx: int = -1\\n win_ratio: float = 0\\n\\n # First create the root node for the game\\n root = State(None, None, state)\\n root = expand(root)\\n root = run_each_move(root, state)\\n print(\\\"New Move UCBs: \\\")\\n for m in range(len(root.move_ucbs)):\\n print(\\\" \\\", root.move_ucbs[m])\\n \\n #while True:\\n for _ in range(10000):\\n _, win_ratio = MCTS(root, state, True, win_ratio) \\n #print(\\\"\\\\n\\\\n\\\\nRoot Attempts: \\\", root.attempts)\\n #print(\\\"Back in main\\\\n\\\\n\\\\n\\\")\\n \\n # Pick one of the highest ranked by UCB's\",\n \"def chooseAction(self, epsilon, state):\\n if random.uniform(0, 1) < epsilon:\\n return random.randrange(9)\\n\\n cur_best_val = -float('inf')\\n cur_best_action = 0\\n\\n data = env.getAllNextStates(state)\\n\\n with torch.no_grad():\\n for action, next_state, done in data:\\n if next_state != state:\\n value = self.NN(self.RBF[next_state]).item() if not done else 0\\n if value > cur_best_val:\\n cur_best_val = value\\n cur_best_action = action\\n #print(data)\\n return cur_best_action\",\n \"def MCTS(parent: State, state: GameState, root: int, win_ratio: float) -> int:\\n #print(\\\"\\\\nPrior Move UCBs: \\\")\\n #for m in range(len(parent.move_ucbs)):\\n # print(\\\" \\\", parent.move_ucbs[m])\\n # Find the maximum UCB value to explore\\n idx = parent.move_ucbs.index(max(parent.move_ucbs))\\n # TODO: May want to make this a random selection instead of first\\n # Select the child with the highest UCB value\\n child = parent.moves[idx]\\n #print(\\\"Child Selected: \\\", child.idx)\\n # Check if the child has moves expanded already\\n if child.attempts == 0 or len(child.moves) == 0: # Expand if no moves\\n child = expand(child)\\n # Pick child from expanded set\\n sim_idx = random.randint(0, len(child.moves)-1)\\n simulate_child = child.moves[sim_idx]\\n # Perform Simulation\\n outcome = simulate(simulate_child.board)\\n else: # Recurse if there are moves\\n #print(\\\"Recursing.................................\\\")\\n outcome, _ = MCTS(child, state, False, 0) \\n # Update the wins/attempts\\n #print(\\\"Outcome: \\\", outcome)\\n #print(\\\"Player: \\\", state.player)\\n if outcome == state.player:\\n #print(\\\"********** Winner **********\\\")\\n #print(\\\"Index in list for ucb: \\\", idx)\\n child.winner()\\n if root:\\n win_ratio = update_selected(state, child, win_ratio)\\n else:\\n #print(\\\"********** Loser **********\\\")\\n #print(\\\"Index in list for ucb: \\\", idx)\\n child.loser()\\n parent.attempts += 1\\n update_ucbs(parent)\\n #print(\\\"New Move UCBs: \\\")\\n #for m in range(len(parent.move_ucbs)):\\n # print(\\\" \\\", parent.move_ucbs[m])\\n \\n\\n return outcome, win_ratio\",\n \"def ismcts(rootstate, itermax, verbose=False, belief=[]):\\n\\tprint(belief)\\n\\trootnode = Node()\\n\\n\\tfor i in range(itermax):\\n\\t\\tnode = rootnode\\n\\n\\t\\t# Determinize\\n\\t\\tstate = rootstate.clone_and_randomize(rootstate.player_to_move)\\n\\n\\t\\t# Select\\n\\t\\twhile state.get_moves() != [] and node.get_untried_moves(state.get_moves()) == []:\\n\\t\\t\\t# node is fully expanded and non-terminal\\n\\t\\t\\tnode = node.ucb_select_child(state.get_moves())\\n\\t\\t\\tstate.do_move(node.move)\\n\\n\\t\\t# Expand\\n\\t\\tuntried_moves = node.get_untried_moves(state.get_moves())\\n\\t\\tif untried_moves: # if we can expand (i.e. state/node is non-terminal)\\n\\t\\t\\tm = random.choice(untried_moves)\\n\\t\\t\\tplayer = state.player_to_move\\n\\t\\t\\tstate.do_move(m)\\n\\t\\t\\tnode = node.add_child(m, player) # add child and descend tree\\n\\n\\t\\t# Simulate\\n\\t\\twhile state.get_moves(): # while state is non-terminal\\n\\t\\t\\tstate.do_move(random.choice(state.get_moves()))\\n\\n\\t\\t# Backpropagate\\n\\t\\twhile node is not None: # backpropagate from the expanded node and work back to the root node\\n\\t\\t\\tnode.update(state)\\n\\t\\t\\tnode = node.parentNode\\n\\n\\t# Output some information about the tree - can be omitted\\n\\tif verbose:\\n\\t\\tprint(rootnode.tree_to_string(0))\\n\\telse:\\n\\t\\tprint(rootnode.children_to_string())\\n\\n\\treturn max(rootnode.child_nodes, key=lambda c: c.visits).move # return the move that was most visited\",\n \"def mcts_search(self, state):\\n assert state.current_player() == self.player\\n root = SearchNode(None, 1)\\n for _ in range(self.max_simulations):\\n visit_path, working_state = self._apply_tree_policy(root, state)\\n if working_state.is_terminal():\\n node_value = working_state.player_return(self.player)\\n else:\\n node_value = self.evaluator.evaluate(\\n working_state, self.player, self._random_state)\\n\\n for node in visit_path:\\n node.total_reward += node_value * node.player_sign\\n node.explore_count += 1\\n\\n most_visited = root.most_visited_child()\\n\\n if self.verbose:\\n print(\\\"Root:\\\", root.to_str())\\n print(\\\"Children:\\\")\\n print(root.children_str(working_state))\\n print(\\\"Children of chosen:\\\")\\n chosen_state = state.clone()\\n chosen_state.apply_action(most_visited.action)\\n print(most_visited.children_str(chosen_state))\\n\\n return most_visited.action\",\n \"def test_state(\\n size: Union[int, tuple],\\n num_berries: int,\\n number_steps: int,\\n state_sizes: List[int] = [3, 5],\\n) -> None:\\n for state_size in state_sizes:\\n game = Game(\\n size,\\n [0, 0],\\n -1,\\n 5,\\n -5,\\n 10,\\n num_berries,\\n berry_movement_probabilities=[0.5] * num_berries,\\n state_size=state_size,\\n )\\n done = False\\n i = 1\\n print(f\\\"Beginning full board\\\\n{game.get_state(full=True)}\\\")\\n print(f\\\"And the state\\\\n{game.get_state(state_size)}\\\")\\n while not done and i < number_steps:\\n action = random.choice(MOVEMENTS)\\n print(f\\\"Action taken {action}\\\")\\n state, reward, done = game.step(action)\\n print(f\\\"Full board\\\\n{game.get_state(full=True)}\\\")\\n print(f\\\"The state\\\\n{game.get_state(state_size)}\\\")\\n i += 1\",\n \"def move(self, state):\\n result = None\\n self.currentDepthLimit = 0\\t\\n\\tself.transposition = {}\\n\\tself.counter = 0\\n\\n\\twhile True:\\n u = float(\\\"inf\\\")\\n\\t v = float(\\\"-inf\\\")\\n\\t self.counter = 0\\n\\t result = None\\n\\t self.transposition = {}\\n\\t for a in state.actions():\\n new = self.min_value(state.result(a), float(\\\"-inf\\\"), float(\\\"inf\\\"),self.currentDepthLimit)\\n\\t if new > v:\\n\\t v = new\\n\\t result = a\\n\\n\\t elif new == v:\\n\\t if a.index < result.index:\\n\\t result = a\\n\\t if self.is_time_up():\\n\\t return result\\n\\t \\n\\t self.currentDepthLimit += 1\\n\\t \\\"\\\"\\\"If we never use evaluate function, it means all state are terminated, so return whatever the result is\\\"\\\"\\\"\\n\\t if self.counter == 0:\\n\\t break\\n\\t if self.is_time_up():\\n \\t return result\\n\\treturn result\",\n \"def play_game(start_state=GoState(),\\n policy_value=NaivePolicyValue(),\\n max_game_length=1e99,\\n opponent=None,\\n n_simulations=N_SIMULATIONS):\\n\\n # TODO: This will set .move_number = 0, should maybe track whose turn it is instead:\\n tree_root = TreeStructure(start_state)\\n policy, value = policy_value.predict(tree_root.state)\\n tree_root.prior_policy = policy[tree_root.state.valid_actions]\\n tree_root.add_noise_to_prior()\\n game_history = []\\n\\n while not tree_root.state.game_over and tree_root.move_number < max_game_length:\\n\\n mcts(tree_root, policy_value, n_simulations)\\n\\n # print_tree(tree_root,0)\\n # Store the state and distribution before we prune the tree:\\n # TODO: Refactor this\\n\\n game_history.append(tree_root.history_sample())\\n\\n choice = choice_to_play(tree_root, bool(opponent))\\n tree_root = tree_root.children[choice]\\n tree_root.parent = None\\n tree_root.add_noise_to_prior()\\n\\n if opponent:\\n game_history.append(tree_root.history_sample())\\n choice = opponent(tree_root.state)\\n if choice in tree_root.children:\\n tree_root = tree_root.children[choice]\\n else:\\n new_state = step(tree_root.state, choice)\\n tree_root = TreeStructure(new_state)\\n policy, value = policy_value.predict(tree_root.state)\\n tree_root.prior_policy = policy[tree_root.state.valid_actions]\\n tree_root.parent = None\\n\\n if tree_root.move_number >= max_game_length:\\n tree_root.state.winner = tree_root.state._compute_winner()\\n\\n return game_history, tree_root.state.winner\",\n \"def gen_game_tree(state_init):\\n current_path = [state_init]\\n game_tree = {}\\n while current_path:\\n cur_state = current_path[-1]\\n if cur_state not in game_tree:\\n ttt = TicTacToe(cur_state)\\n game_tree[cur_state] = {\\n \\\"unexplored\\\": ttt.next_states,\\n \\\"explored\\\": [],\\n \\\"status\\\": ttt.game_status,\\n }\\n\\n if game_tree[cur_state][\\\"unexplored\\\"]:\\n current_path.append(game_tree[cur_state][\\\"unexplored\\\"].pop(0))\\n else:\\n explored = current_path.pop(-1)\\n if explored != state_init:\\n game_tree[current_path[-1]][\\\"explored\\\"].append(explored)\\n\\n status, player = game_tree[cur_state][\\\"status\\\"]\\n if status == \\\"tie\\\":\\n value = 0\\n outcomes = {0: 1}\\n elif status == \\\"win\\\":\\n value = -1 if player == \\\"1\\\" else 1\\n outcomes = {value: 1}\\n else:\\n value = (min if player == \\\"1\\\" else max)(\\n [\\n game_tree[state][\\\"value\\\"]\\n for state in game_tree[cur_state][\\\"explored\\\"]\\n ]\\n )\\n outcomes = {}\\n for state in game_tree[cur_state][\\\"explored\\\"]:\\n for res, res_ct in game_tree[state][\\\"outcomes\\\"].items():\\n outcomes[res] = outcomes.get(res, 0) + res_ct\\n\\n game_tree[cur_state][\\\"value\\\"] = value\\n game_tree[cur_state][\\\"outcomes\\\"] = outcomes\\n\\n return game_tree\",\n \"def optimize(seqlen, trees, nstates=2):\\n num_params = nstates * (nstates + 1)\\n guess = np.log(np.random.rand(num_params))\\n # guess = np.log(np.array([0.5, 0.5, 0.13, 0.37, 0.37, 0.13,\\n # 0.32, 0.18, 0.18, 0.32, 0.95, 0.05, 0.05, 0.95]))\\n i, t, s = unfold_params(guess, nstates=nstates)\\n i = np.log(np.ones(nstates) / nstates)\\n i, t = normalize(i), normalize(t)\\n\\n i = np.log(np.array([0.5, 0.5]))\\n # t = np.log(np.array([\\n # [1 - 0.01, 0.01],\\n # [0.01, 1 - 0.01]\\n # ]))\\n t = np.log(np.array([\\n [0.5, 0.5],\\n [0.5, 0.5]\\n ]))\\n s = 1.0 / trees[0].tot_branch_len()\\n\\n guess = np.concatenate((i, np.ndarray.flatten(t), [1]))\\n res = minimize(getProb, guess, args=(nstates, seqlen, trees), method=\\\"L-BFGS-B\\\",\\n options={\\\"maxiter\\\": 250, \\\"disp\\\": True},\\n bounds=(([(np.NINF, 0)] * num_params) + [(0, np.Inf)]))\\n i, t, s = unfold_params(res.x, nstates=nstates)\\n i, t = normalize(i), normalize(t)\\n print(forward(seqlen, t, i, (trees * s)))\\n return np.exp(i), np.exp(t), s\",\n \"def dfs_maximizing(state) :\\n #print state.describe_previous_move()\\n global state_evals, path, _path, _score, level, _state;\\n\\n level+=1\\n path.append(state)\\n for stt in state.generate_next_states():\\n score=0\\n agenda.append((stt, level))\\n \\n if stt.is_game_over():\\n state_evals+=1\\n score=stt.get_endgame_score()\\n if score>_score:\\n _score=score\\n _path = path[0:]\\n _state = stt\\n if not agenda:\\n\\n _path.append(_state)\\n return [_path, _score, state_evals];\\n else:\\n new_state, level=agenda.pop()\\n path=path[0:level]\\n level-=1\\n return dfs_maximizing(new_state)\",\n \"def alphabeta_search(state):\\r\\n \\r\\n '''\\r\\n Terminates when game.actions is empty\\r\\n Class Game needs the following functions:\\r\\n - game.result(state, a) -- successor\\r\\n - game.actions(state) -- possible moves\\r\\n - game.utility -- returns the state of the game (win/lose or tie, when game is terminal)\\r\\n \\r\\n '''\\r\\n #sort state.actions in increasing or decreasing based on max or min (alpha or beta)\\r\\n #use heuristics fn to get a value for each move (move is in format (x,y) where x and y are ints\\r\\n \\r\\n d = depthset[0] #this is the cutoff test depth value. if we exceed this value, stop\\r\\n cutoff_test=None\\r\\n sort_fn = [vitalpoint, eyeHeur]\\r\\n eval_fn = survivalheur \\r\\n #randnumheuristics \\r\\n player = state.to_move()\\r\\n prune = 0\\r\\n pruned = {} #this will store the depth of the prune\\r\\n totaldepth = [0]\\r\\n visited = {}\\r\\n heuristicInd = 0\\r\\n \\r\\n def max_value(state, alpha, beta, depth, heuristicInd):\\r\\n branches = len(state.actions())\\r\\n onbranch = 0\\r\\n \\r\\n if totaldepth[0] < depth:\\r\\n totaldepth[0] = depth\\r\\n if cutoff_test(state, depth):\\r\\n return eval_fn(state)\\r\\n v = -infinity\\r\\n \\r\\n #sort state.actions based on heuristics before calling\\r\\n #max wants decreasing\\r\\n #sorted(state.actions(), key = eval_sort, reverse = True)\\r\\n \\r\\n #sort by favorites first, returns a list of actions\\r\\n # for sorts in sort_fn:\\r\\n tempher = heuristicInd\\r\\n\\r\\n sorts = sort_fn[heuristicInd]\\r\\n sortedactions, heuristicInd = sorts(state)\\r\\n #if heuristicInd != tempher:\\r\\n # print 's',\\r\\n ''''''\\r\\n for a in sortedactions:\\r\\n if visited.get(depth) == None:\\r\\n visited[depth] = [a]\\r\\n else:\\r\\n visited[depth].append(a)\\r\\n \\r\\n onbranch += 1\\r\\n v = max(v, min_value(state.result(a),\\r\\n alpha, beta, depth+1, heuristicInd)) #+ vitscore.count(a)\\r\\n if v >= beta: #pruning happens here, but in branches\\r\\n if pruned.get(depth) == None:\\r\\n pruned[depth] = branches - onbranch\\r\\n else:\\r\\n pruned[depth] += (branches - onbranch)\\r\\n #print \\\"prune\\\", depth, \\\" \\\", state.actions()\\r\\n #state.display()\\r\\n return v\\r\\n alpha = max(alpha, v)\\r\\n \\r\\n #print depth, \\\" \\\", state.actions()\\r\\n #state.display()\\r\\n \\r\\n return v\\r\\n\\r\\n def min_value(state, alpha, beta, depth, heuristicInd):\\r\\n branches = len(state.actions())\\r\\n onbranch = 0\\r\\n \\r\\n if totaldepth[0] < depth:\\r\\n totaldepth[0] = depth\\r\\n if cutoff_test(state, depth):\\r\\n return eval_fn(state)\\r\\n v = infinity\\r\\n \\r\\n #sort state.actions based on heuristics before calling\\r\\n #min wants increasing\\r\\n #sorted(state.actions(), key = eval_sort)\\r\\n #Shayne\\r\\n tempher = heuristicInd\\r\\n sorts = sort_fn[heuristicInd]\\r\\n sortedactions, heuristicInd = sorts(state, 1)\\r\\n #if heuristicInd != tempher:\\r\\n # print 's',\\r\\n for a in sortedactions: #state.actions():\\r\\n onbranch += 1\\r\\n if visited.get(depth) == None:\\r\\n visited[depth] = [a]\\r\\n else:\\r\\n visited[depth].append(a)\\r\\n v = min(v, max_value(state.result(a),\\r\\n alpha, beta, depth+1, heuristicInd))\\r\\n if v <= alpha: #pruning happens here, but in branches\\r\\n if pruned.get(depth) == None:\\r\\n pruned[depth] = branches - onbranch\\r\\n else:\\r\\n pruned[depth] += (branches - onbranch)\\r\\n #print \\\"prune\\\", depth, \\\" \\\", state.actions()\\r\\n #state.display()\\r\\n return v\\r\\n beta = min(beta, v)\\r\\n #print depth, \\\" \\\", state.actions()\\r\\n #state.display()\\r\\n return v\\r\\n\\r\\n # Body of alphabeta_search starts here:\\r\\n #def cutoff_test and eval_fn \\r\\n cutoff_test = (cutoff_test or\\r\\n (lambda state,depth: depth>d or state.terminal_test()))\\r\\n eval_fn = eval_fn or (lambda state: state.utility(player))\\r\\n #by default, utility score is used\\r\\n \\r\\n \\r\\n #argmax goes through all the possible actions and \\r\\n # applies the alphabeta search onto all of them\\r\\n # and returns the move with the best score \\r\\n #print state.actions()\\r\\n heuristicInd = 0\\r\\n sorts = sort_fn[heuristicInd]\\r\\n sortedact, heuristicInd = sorts(state)\\r\\n abmove = argmax(sortedact,\\r\\n lambda a: min_value(state.result(a),\\r\\n -infinity, infinity, 0, heuristicInd))\\r\\n\\r\\n print 'problem,', problemno[0], ', total tree depth,', totaldepth[0]\\r\\n for i in range(1, len(visited)):\\r\\n if len(pruned) < i:\\r\\n pruned[i] = 0\\r\\n print i, \\\",\\\", len(visited[i]), \\\",\\\", pruned[i]\\r\\n \\r\\n return abmove\",\n \"def simulate(state: GameState) -> int:\\n moves = list(state.moves)\\n #print(\\\" moves available: \\\", moves)\\n for i in range(len(state.moves)):\\n move = random.choice(moves)\\n #print(\\\" move making: \\\", move)\\n move_idx = moves.index(move)\\n #print(\\\" index of move: \\\", move_idx)\\n moves.pop(move_idx)\\n #print(\\\" new moves available: \\\", moves)\\n state = state.traverse(move)\\n #print(\\\" Winner: \\\", state.util)\\n #print(\\\" New Board: \\\", state.display)\\n return state.util\",\n \"def ai_strategy3(gstate: TicTacToe, game_tree):\\n status, player = gstate.game_status\\n\\n if status != \\\"turn\\\":\\n logging.warning(\\\"Game status = %s. No move needed.\\\", status)\\n return None\\n\\n win, lose = (-1, 1) if player == \\\"1\\\" else (1, -1)\\n\\n move_vals = {\\n move: win * game_tree[state][\\\"value\\\"] for move, state in gstate.next_moves.items()\\n }\\n max_val = max(move_vals.values())\\n\\n if max_val == 1:\\n # ai_strategy2 can handle \\\"won\\\" states\\n return ai_strategy2(gstate, game_tree)\\n\\n else:\\n ok_moves = [move for move, val in move_vals.items() if val == max_val]\\n\\n move_vals = {\\n move: collections.Counter(\\n [\\n game_tree[state2][\\\"value\\\"] for state2 in game_tree[gstate.next_moves[move]][\\\"explored\\\"]\\n ]\\n )\\n for move in ok_moves\\n }\\n\\n move_eval = {\\n move: val_ctr.get(win, 0) / val_ctr.get(lose, 0.5)\\n for move, val_ctr in move_vals.items()\\n }\\n max_win_pct = max(move_eval.values())\\n good_moves = [move for move, win_pct in move_eval.items() if win_pct ==\\n max_win_pct]\\n\\n move = random.choice(good_moves)\\n\\n all_ct = sum(move_vals[move].values())\\n win_ct = move_vals[move].get(win, 0)\\n logging.debug(\\n \\\"move: %i; value: %i, win %%: %.1f%%, moves: %s\\\\n\\\",\\n move, win * game_tree[gstate.state][\\\"value\\\"],\\n win_ct / max(all_ct, 0.1) * 100, move_vals\\n )\\n\\n return move\",\n \"def decide(self, state: OthelloState, actions: list):\\n # -------- TASK 2 ------------------------------------------------------\\n # Your task is to implement an algorithm to choose an action form the\\n # given `actions` list. You can implement any algorithm you want.\\n # However, you should keep in mind that the execution time of this\\n # function is limited. So, instead of choosing just one action, you can\\n # generate a sequence of increasing good action.\\n # This function is a generator. So, you should use `yield` statement\\n # rather than `return` statement. To find more information about\\n # generator functions, you can take a look at:\\n # https://www.geeksforgeeks.org/generators-in-python/\\n #\\n # If you generate multiple actions, the last action will be used in the\\n # game.\\n #\\n # Tips\\n # ====\\n # 1. During development of your algorithm, you may want to find the next\\n # state after applying an action to the current state; in this case,\\n # you can use the following patterns:\\n # `next_state = current_state.successor(action)`\\n #\\n # 2. If you need to simulate a game from a specific state to find the\\n # the winner, you can use the following pattern:\\n # ```\\n # simulator = Game(FirstAgent(), SecondAgent())\\n # winner = simulator.play(starting_state=specified_state)\\n # ```\\n # The `MarkovAgent` has illustrated a concrete example of this\\n # pattern.\\n #\\n # 3. You are free to choose what kind of game-playing agent you\\n # implement. Some of the obvious approaches are the following:\\n # 3.1 Implement alpha-beta (and investigate its potential for searching deeper\\n # than what is possible with Minimax). Also, the order in which the actions\\n # are tried in a given node impacts the effectiveness of alpha-beta: you could\\n # investigate different ways of ordering the actions/successor states.\\n # 3.2 Try out better heuristics, e.g. ones that take into account the higher\\n # importance of edge and corner cells. Find material on this in the Internet.\\n # 3.3 You could try out more advanced Monte Carlo search methods (however, we do\\n # not know whether MCTS is competitive because of the high cost of the full\\n # gameplays.)\\n # 3.4 You could of course try something completely different if you are willing to\\n # invest more time.\\n #\\n # GL HF :)\\n # ----------------------------------------------------------------------\\n\\n # Replace the following lines with your algorithm\\n best_action = actions[0]\\n yield best_action\",\n \"def get_move(self, state):\\n for n in range(self._n_playout):\\n state_copy = copy.deepcopy(state)\\n self._playout(state_copy)\\n print(n)\\n print(\\\"Judge done\\\")\\n return max(self._root._children.items(),\\n key=lambda act_node: act_node[1]._n_visits)[0]\",\n \"def policies(self, QTable, epsilon, state, next_states, action_to_do): # Inspiration from https://www.geeksforgeeks.org/q-learning-in-python/?fbclid=IwAR1UXR88IuJBhhTakjxNq_gcf3nCmJB0puuoA46J8mZnEan_qx9hhoFzhK8\\r\\n num_actions = 5 # 5 actions-value, [moved_out, into_goal, send_opp_home, send_self_home, move_token] \\r\\n def epsilonGreedyPolicy(): \\r\\n tmp_state = str(state.state[0])\\r\\n valid_actions = np.append(action_to_do, True) # the True appended is move_token\\r\\n valid_act_len = len(np.where(valid_actions==True)[0])\\r\\n\\r\\n Action_probabilities = np.ones(num_actions, dtype = float) * epsilon / valid_act_len # divides probability based on number of valid actions and epsilon (each 0.025 if 4 actions) \\r\\n Action_probabilities = np.multiply(Action_probabilities, valid_actions)\\r\\n\\r\\n # If same values in QTable choose random valid action \\r\\n best_action = np.argmax(QTable[tmp_state]) # Find index of action which gives highest QValue\\r\\n # Check if valid action else find new best action\\r\\n if not valid_actions[best_action]:\\r\\n actions = np.argsort(-QTable[tmp_state]) # descending order of action values\\r\\n for i in range(len(valid_actions)):\\r\\n if valid_actions[actions[i]]:\\r\\n best_action = actions[i]\\r\\n break\\r\\n\\r\\n Action_probabilities[best_action] += (1.0 - epsilon) # Assigns rest probability to best action so probability sums to 1\\r\\n\\r\\n return Action_probabilities \\r\\n\\r\\n def greedyPolicy():\\r\\n tmp_state = str(state.state[0])\\r\\n valid_actions = np.append(action_to_do, True) # the True appended is move_token\\r\\n\\r\\n Action_probabilities = np.zeros(num_actions, dtype = float)\\r\\n\\r\\n best_action = np.argmax(QTable[tmp_state]) # Find index of action which gives highest QValue\\r\\n # Check if valid action else find new best action\\r\\n if not valid_actions[best_action]:\\r\\n actions = np.argsort(-QTable[tmp_state]) # descending order of action values\\r\\n for i in range(len(valid_actions)):\\r\\n if valid_actions[actions[i]]:\\r\\n best_action = actions[i]\\r\\n break\\r\\n\\r\\n\\r\\n Action_probabilities[best_action] += 1.0\\r\\n return Action_probabilities\\r\\n\\r\\n\\r\\n if(self.__chosenPolicy == \\\"epsilon greedy\\\"):\\r\\n return epsilonGreedyPolicy \\r\\n if(self.__chosenPolicy == \\\"greedy\\\"):\\r\\n return greedyPolicy\",\n \"def next_state_generation(state, target, operators):\\n current_swap_state = ''\\n current_swap_res = sys.maxsize\\n current_change_state = ''\\n current_change_res = sys.maxsize\\n # for swapping\\n for i in range(0, len(state), 2):\\n for j in range(2, len(state), 2):\\n new_swap_state, swap_res = swap(i, j, state, target)\\n if current_swap_res > swap_res:\\n current_swap_res = swap_res\\n current_swap_state = new_swap_state\\n # for changing\\n for i in range(1, len(state) - 1, 2):\\n new_change_state, change_res = change(i, random.choice(operators), state, target)\\n if current_change_res > change_res:\\n current_change_res = change_res\\n current_change_state = new_change_state\\n # return the lowest of the 2 value and state amongst swapping and changing operations\\n print(\\\"Swap res:\\\", current_swap_res)\\n print(\\\"Change res:\\\", current_change_res)\\n if current_swap_res < current_change_res:\\n print(\\\"Best State \\\", current_swap_state)\\n print(\\\"Distance from target: \\\", current_swap_res)\\n print()\\n return current_swap_state, current_swap_res\\n else:\\n print(\\\"Best State \\\", current_change_state)\\n print(\\\"Distance from target: \\\", current_change_res)\\n print()\\n return current_change_state, current_change_res\",\n \"def select_move(self, game_state, return_visit_counts=False):\\n \\n # Start with a tree consisting of a root node only. The root node\\n # is associated with the given board position.\\n root = self.create_node(game_state)\\n \\n # If no legal moves can be made from the given board position, pass \\n # the turn. This happens when all of the players pieces are surrounded,\\n # if the player has no pieces left or if the game is over. \\n if not root.branches:\\n if return_visit_counts:\\n return Act.pass_turn(), {}\\n return Act.pass_turn()\\n \\n for i in range(self.num_rounds):\\n # On each iteration, walk down the tree to a leaf node and select\\n # a move to make from the corresponding leaf game state.\\n node = root\\n next_move = self.select_branch(node)\\n while node.has_child(next_move):\\n node = node.get_child(next_move)\\n next_move = self.select_branch(node)\\n \\n # Create a new tree node for the selected move and add it to\\n # the tree. If the leaf node corresponds to a finished game\\n # then don't create a new node and assign a value to the node\\n # based on who won.\\n if node.state.is_not_over():\\n if next_move:\\n new_state = copy.deepcopy(node.state)\\n new_state.take_turn_with_no_checks(Act.play(next_move))\\n child_node = self.create_node(new_state, \\n move=next_move, parent=node)\\n move = next_move\\n value = -1 * child_node.value \\n else:\\n # If the current player can't make any moves from the\\n # selected gamestate then next_move will be 'None' meaning\\n # the player passes the turn.\\n new_state = copy.deepcopy(node.state)\\n new_state.take_turn_with_no_checks(Act.pass_turn())\\n child_node = self.create_node(new_state, \\n move=next_move, parent=node)\\n move = next_move\\n value = -1 * child_node.value\\n else:\\n # If the game in the current state is over, then the last\\n # player must have won the game. Thus the value/reward for the\\n # other player is 1. The current node is not updated with\\n # the new reward as no branches can stem from a finished game\\n # state.\\n move = node.last_move\\n node = node.parent\\n value = 1\\n \\n # Update the nodes traversed to get to the leaf node with the \\n # new value for the new move.\\n while node is not None:\\n node.record_visit(move, value)\\n move = node.last_move\\n node = node.parent\\n value *= -1\\n \\n # Get the visit counts of the branches if they were requested.\\n if return_visit_counts:\\n visit_counts = {}\\n for move in root.branches.keys():\\n visit_counts[move] = root.branches[move].visit_count\\n \\n # Get a list of possible moves sorted according to visit count,\\n # the move with the highest visit count should be first in the list.\\n moves = [move for move in root.moves()]\\n moves = sorted(moves, key=root.visit_count, reverse=True)\\n \\n # Loop through the sorted moves and return the first legal one.\\n for move in moves:\\n if not game_state.is_move_illegal(move):\\n if return_visit_counts:\\n return Act.play(move), visit_counts\\n return Act.play(move)\\n \\n # If no legal move is found then pass the turn.\\n if return_visit_counts:\\n return Act.pass_turn(), visit_counts\\n return Act.pass_turn()\",\n \"def greedy(self, state, timestep, epsilon=0):\\n\\n counts = np.bincount(self.call_locs, minlength=self.num_nodes)\\n # print(self.lengths)\\n # print(counts)\\n score = self.lengths @ counts\\n action = []\\n for _ in range(self.num_ambulance):\\n node = np.argmin(score)\\n action.append(node)\\n score[node] = 99999999\\n return action\",\n \"def move(self, state):\\n \\n self.depth_limit=1\\n self.best_utility=-2\\n action=None\\n while not self.is_time_up():\\n self.terminal=True\\n self.cache={}\\n action=self.alpha_beta_search(state,0)\\n if self.terminal==True:\\n break\\n self.depth_limit=self.depth_limit+1\\n \\n return action\",\n \"def marcovNuc (i = random.choice(stateSpace), step = 100):\\n # matrix of transition probabilities\\n #matrix = [[0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25]] \\n matrix = [[0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1], [0.4, 0.3, 0.2, 0.1]] \\n step += 1 # add one to the range because we remove it at the end\\n sims = [] # List to hold the results of the Marcov chain\\n sims.append(i) # append the seed value to the sims list\\n for x in range(step):\\n \\n if sims[-1] == 'A':\\n w = np.random.random() # Random number generator\\n # the next set of if statements determine where the random number \\n # sits on the number line of probabilities\\n if matrix[0][0] > w:\\n sims.append('A')\\n elif matrix[0][1] + matrix[0][0] > w:\\n sims.append('C')\\n elif matrix[0][2] + matrix[0][1] + matrix[0][0] > w:\\n sims.append('G')\\n else:\\n sims.append('T')\\n elif sims[-1] == 'C':\\n x = np.random.random()\\n if matrix[1][0] > x:\\n sims.append('A')\\n elif matrix[1][1] + matrix[1][0] > x:\\n sims.append('C')\\n elif matrix[1][2] + matrix[1][1] + matrix[1][0] > x:\\n sims.append('G')\\n else:\\n sims.append('T')\\n \\n elif sims[-1] == 'G':\\n y = np.random.random()\\n if matrix[2][0] > y:\\n sims.append('A')\\n elif matrix[2][1] + matrix[2][0] > y:\\n sims.append('C')\\n elif matrix[2][2] + matrix[2][1] + matrix[2][0] > y:\\n sims.append('G')\\n else:\\n sims.append('T')\\n\\n else:\\n z = np.random.random()\\n if matrix[3][0] > z:\\n sims.append('A')\\n elif matrix[3][1] + matrix[3][0] > z:\\n sims.append('C')\\n elif matrix[3][2] + matrix[3][1] + matrix[3][0] > z:\\n sims.append('G')\\n else:\\n sims.append('T')\\n\\n return sims[1:-1] # remove the initial value (the seed)\",\n \"def solution(self) -> State:\",\n \"def result(self, state, action):\\n state_after_act = [[0 for i in range(self.col)] for j in range(self.row)]\\n for k in action:\\n x = k[1][0]\\n y = k[1][1]\\n if k[0] == \\\"vaccinate\\\":\\n state_after_act[x][y] = ('I', 1)\\n else:\\n state_after_act[x][y] = ('Q', 1)\\n\\n for i in range(self.row):\\n for j in range(self.col):\\n if state_after_act[i][j] == 0:\\n if state[i][j][0] == 'U' or state[i][j][0] == 'I':\\n state_after_act[i][j] = state[i][j]\\n\\n elif state[i][j][0] == 'S':\\n if state[i][j][1] == 3:\\n state_after_act[i][j] = ('H', 1)\\n else:\\n if state[i][j][1] == 1:\\n state_after_act[i][j] = ('S', 2)\\n elif state[i][j][1] == 2:\\n state_after_act[i][j] = ('S', 3)\\n\\n elif state[i][j][0] == 'Q':\\n if state[i][j][1] == 2:\\n state_after_act[i][j] = ('H', 1)\\n else:\\n state_after_act[i][j] = ('Q', 2)\\n\\n elif state[i][j][0] == 'H':\\n state_after_act[i][j] = self.healthy(i, j, state,state_after_act)\\n state_after_act[i] = tuple(state_after_act[i])\\n return tuple(state_after_act)\",\n \"def propagate(self, game_state, maximizing, max_hitherto, depth, max_depth):\\n \\n self.moves += 1\\n \\n\\n if depth >= max_depth:\\n rating = self.rate_game_state(game_state)\\n return (rating, None, None)\\n \\n else:\\n \\n color = self.color if maximizing else (1 if self.color==0 else 0)\\n\\n self.timer.start(\\\"\\\")\\n choices = []\\n for line in self.get_all_lines(game_state, color):\\n for vector in LEGAL_VECTORS:\\n\\n new_game_state = self.get_new_game_state(game_state, line, vector, color)\\n if new_game_state is not None:\\n choices.append( (new_game_state, self.rate_game_state(new_game_state), line, vector) )\\n\\n \\n\\n choices = sorted(choices, key=lambda x: x[1], reverse=maximizing)\\n \\n final_choice = choices[0][1:]\\n del choices[0]\\n\\n for choice in choices:\\n\\n new_game_state, estimated_rating, line, vector = choice\\n actual_rating, _, _ = self.propagate(new_game_state, not maximizing, final_choice[0], depth+1, max_depth)\\n \\n if maximizing:\\n if final_choice[0] < actual_rating:\\n final_choice = (actual_rating, line, vector)\\n\\n else: #minimizing\\n if (max_hitherto is not None) and (actual_rating < max_hitherto): return (actual_rating, line, vector)\\n\\n if actual_rating < final_choice[0]:\\n final_choice = (actual_rating, line, vector)\\n\\n return final_choice\",\n \"def train(n):\\n\\n ai = NimAI()\\n\\n print(f\\\"Play {n} training games\\\")\\n for _ in range(n): \\n game = Nim()\\n\\n # Keep track of last move made by either player\\n last = {\\n 0: {\\\"state\\\": None, \\\"action\\\": None},\\n 1: {\\\"state\\\": None, \\\"action\\\": None}\\n }\\n\\n # Game loop\\n while True:\\n\\n # Keep track of current state and action\\n state = game.piles.copy()\\n action = ai.chooseAction(game.piles)\\n\\n # Keep track of last state and action\\n last[game.player][\\\"state\\\"] = state\\n last[game.player][\\\"action\\\"] = action\\n\\n # Make move and switch players\\n game.move(action)\\n new_state = game.piles.copy()\\n\\n # When game is over, update Q values with rewards\\n if game.winner is not None:\\n # The game is over when a player just made a move that lost him the game.\\n # The move from the previous player was therefore game winning.\\n # Both events are used to update the AI.\\n # new_state is [0, 0, 0, 0] here and its used to update the AI, because\\n # future rewards should not be considered in the Q-learning formula.\\n ai.update(state, action, new_state, -1)\\n ai.update(\\n last[game.player][\\\"state\\\"],\\n last[game.player][\\\"action\\\"],\\n new_state,\\n 1\\n )\\n break\\n\\n # If game is continuing, no rewards yet\\n elif last[game.player][\\\"state\\\"] is not None:\\n ai.update(\\n last[game.player][\\\"state\\\"],\\n last[game.player][\\\"action\\\"],\\n new_state,\\n 0\\n )\\n\\n print(\\\"Done training\\\")\\n\\n # Return the trained AI\\n return ai\",\n \"def greedy_next_action(self, state):\\n max_val = float('-inf')\\n if self.verbose:\\n cells = []\\n max_candidates = {}\\n for i in range(3):\\n for j in range(3):\\n if state[i][j] == VALUES.EMPTY:\\n val = self.q_value((state, (i, j)))\\n if val >= max_val:\\n max_val = val\\n max_move = (i, j)\\n max_candidates[max_move] = val\\n if self.verbose:\\n cells.append('{0:.3f}'.format(val).center(6))\\n elif self.verbose:\\n cells.append(state[i][j].center(6))\\n if self.verbose:\\n self.logger.info(BOARD.format(*cells))\\n possible_actions = [k for k, v in max_candidates.items() if v == max_val]\\n action = random.choice(possible_actions) if len(possible_actions) > 0 else None\\n return action\",\n \"def next_state(s_curr, action, params):\\n P_dist = params['P_dist']\\n R = params['R']\\n n_rows = params['n_rows']\\n n_cols = params['n_cols']\\n occ_grid = params['occ_grid']\\n\\n rnd = np.random.uniform()\\n\\n s_next = s_curr\\n\\n # Actions - ['left','right','up','down']\\n\\n if rnd <= P_dist:\\n if action == 0:\\n move = 2\\n elif action == 1:\\n move = 2\\n elif action == 2:\\n move = 1\\n else:\\n move = 0\\n elif rnd < 2*P_dist:\\n if action == 0:\\n move = 3\\n elif action == 1:\\n move = 3\\n elif action == 2:\\n move = 1\\n else:\\n move = 1\\n else:\\n move = action\\n\\n # Move left\\n if move == 0:\\n row_next = s_curr[0]\\n col_next = s_curr[1] - 1\\n if col_next >= 0 and occ_grid[row_next, col_next] == 0:\\n s_next = [row_next, col_next]\\n\\n # Move right\\n if move == 1:\\n row_next = s_curr[0]\\n col_next = s_curr[1] + 1\\n if col_next < n_cols and occ_grid[row_next, col_next] == 0:\\n s_next = [row_next, col_next]\\n\\n # Move up\\n if move == 2:\\n row_next = s_curr[0] - 1\\n col_next = s_curr[1]\\n if row_next >= 0 and occ_grid[row_next, col_next] == 0:\\n s_next = [row_next, col_next]\\n\\n # Move down\\n if move == 3:\\n row_next = s_curr[0] + 1\\n col_next = s_curr[1]\\n if row_next < n_rows and occ_grid[row_next, col_next] == 0:\\n s_next = [row_next, col_next]\\n\\n r = R[s_next[0], s_next[1]]\\n return s_next, r\",\n \"def simulation():\\n # initialize action set\\n action_set = np.zeros(int((s.MAX_INSPECT - s.MIN_INSPECT) / s.DELTA) + 3)\\n x, i = s.MIN_INSPECT, 1\\n while x <= s.MAX_INSPECT:\\n action_set[i] = x\\n x += s.DELTA\\n i += 1\\n action_set[-1] = np.inf\\n action_number = len(action_set)\\n\\n # initialize current state\\n current_state = math.floor(np.random.rand(1) * s.NUM_STATES)\\n\\n # initialize action index\\n if current_state == 0:\\n action_index = 0\\n elif current_state == s.NUM_STATES - 1:\\n action_index = action_number - 1\\n\\n if current_state != 0 and current_state != s.NUM_STATES - 1:\\n action_index = action_number - 2\\n\\n # initialize policy set\\n greedy_policy = np.zeros(s.NUM_STATES)\\n greedy_policy[-1] = np.inf\\n for i in range(1, s.NUM_STATES - 1):\\n greedy_policy[i] = s.MAX_INSPECT\\n\\n visit_times = np.zeros([s.NUM_STATES, action_number])\\n\\n # initialization for simulation\\n falpha, Aalpha, delay_T, uni_parameter = equivalent_markov(greedy_policy)\\n stable_prob, potential = stable_potential(falpha, Aalpha, uni_parameter)\\n last_value = falpha + np.matmul(Aalpha, potential)\\n dis_value = last_value\\n # ave_vector = np.matmul(stable_prob, falpha)\\n # ave_estimate = ave_vector.tolist()\\n each_transit_cost, each_transit_time, total_reward = (0 for i in range(3))\\n\\n # initialize DQN model if selected\\n dqn = DQN() if MODEL == 1 else None\\n # initialize Q-table if Q-learning selected\\n q_factor = ql.init_q_factor(action_number) if MODEL == 2 else None\\n\\n for out_step in range(s.EPOCH):\\n epsilon = s.EPSILON_1 if MODEL == 1 else s.EPSILON_2\\n\\n for inner_step in range(s.EPOCH_LEARN):\\n\\n visit_times[current_state, action_index] += 1\\n current_action = greedy_policy[current_state]\\n\\n inspect_cost = 0 if current_state == s.NUM_STATES - 1 else s.K5 * current_action\\n\\n flag, sojourn_T, service_T, next_state = state_transition(current_state, current_action)\\n each_transit_time = s.DISCOUNT * each_transit_time + (sojourn_T - each_transit_time) / pow(\\n out_step * s.EPOCH_LEARN + (inner_step + 1), s.Q_AVE_STEP)\\n end_sojourn_T = math.exp(- s.ALPHA * sojourn_T)\\n end_serve_T = math.exp(- s.ALPHA * service_T)\\n\\n if s.ALPHA == 0:\\n dis_T, dis_serve_T, dis_wait_T = sojourn_T, service_T, sojourn_T - service_T\\n else:\\n dis_T, dis_serve_T = (1 - end_sojourn_T) / s.ALPHA, (1 - end_serve_T) / s.ALPHA\\n dis_wait_T = (end_serve_T - end_sojourn_T) / s.ALPHA\\n\\n if flag == 0: # no processing, waiting\\n cost_real = (s.K1 * (s.NUM_STATES - current_state) + s.K3) * sojourn_T + inspect_cost\\n cost_purt = (s.K1 * (s.NUM_STATES - current_state) + s.K3) * dis_T + inspect_cost\\n else: # no waiting, processing\\n cost_real = s.K1 * (s.NUM_STATES - current_state - 1) * sojourn_T + s.K2 * service_T + s.K3 * (\\n sojourn_T - service_T) + s.K4 + inspect_cost\\n cost_purt = s.K1 * (s.NUM_STATES - current_state - 1) * dis_T + s.K2 * dis_serve_T + s.K3 * dis_wait_T \\\\\\n + s.K4 * end_serve_T + inspect_cost\\n\\n each_transit_cost = s.DISCOUNT * each_transit_cost + (cost_real - each_transit_cost) / (\\n pow(out_step * s.EPOCH_LEARN + (inner_step + 1), s.Q_AVE_STEP))\\n\\n ave_q_cost = each_transit_cost / each_transit_time\\n # ave_estimate.append(ave_q_cost)\\n cost_dis = cost_purt - ave_q_cost * dis_T\\n\\n if MODEL == 1:\\n reward = - cost_dis\\n dqn.store_transition(current_state, action_index, reward, next_state)\\n if dqn.memory_counter >= s.MEMORY_CAPACITY:\\n dqn.learn(s.EPOCH_LEARN, inner_step, PS)\\n else:\\n difference = cost_dis + end_sojourn_T * min(q_factor[next_state, :]) \\\\\\n - q_factor[current_state, action_index]\\n q_factor = ql.update_q_factor(q_factor, current_state, action_index, difference,\\n visit_times, inner_step, PS)\\n current_state = next_state # transit to next state\\n\\n if current_state == 0:\\n action_index = 0\\n elif current_state == s.NUM_STATES - 1:\\n action_index = action_number - 1\\n else:\\n if MODEL == 1:\\n action_index = int(dqn.choose_action(current_state, epsilon))\\n if action_set[action_index] <= 1:\\n greedy_policy[current_state] = action_set[action_index]\\n else:\\n greedy_policy[current_state] = 1\\n else:\\n if np.random.rand(1) < epsilon:\\n action_index = int(np.floor(np.random.rand(1) * (action_number - 2)) + 1)\\n else:\\n # minimal_q_value = np.min(q_factor[current_state, :])\\n action_index = np.argmin(q_factor[current_state, :])\\n greedy_policy[current_state] = action_set[action_index]\\n\\n # store the policy learned from the iterations\\n optimal_policy = greedy_policy\\n\\n if MODEL != 1:\\n for i in range(1, s.NUM_STATES - 1):\\n # minimal_q_value_temp = np.min(q_factor[i, :])\\n action_index_temp = np.argmin(q_factor[i, :])\\n optimal_policy[i] = action_set[action_index_temp]\\n\\n falpha, Aalpha, delay_T, uni_parameter = equivalent_markov(optimal_policy)\\n stable_prob, potential = stable_potential(falpha, Aalpha, uni_parameter)\\n\\n last_value = falpha + np.matmul(Aalpha, potential)\\n dis_value = np.concatenate((dis_value, last_value), axis=1)\\n total_reward += - np.ndarray.item(last_value[0])\\n # new_ave_cost = np.matmul(stable_prob, falpha)\\n # ave_vector = np.concatenate((ave_vector, new_ave_cost))\\n print(\\\"epoch: {} , the epoch reward is {}\\\".format(out_step, round(- np.ndarray.item(last_value[0]), 2)))\\n\\n # result = np.asarray(dis_value)\\n print(\\\"total reward:\\\", total_reward)\\n\\n return dis_value, total_reward\",\n \"def next_state(self):\\n \\n self.state = np.random.choice(['checkout', 'dairy', 'drinks', 'fruit', 'spices'], p=self.tr_array_dict[f'{self.state}'])\",\n \"def breadthfirst(initial_game):\\n # Creat dict for solved case\\n solved_cases = {}\\n\\n # Track states.\\n seen_states = []\\n\\n queue = q.Queue()\\n\\n # Add begin state to queue.\\n initial_state = initial_game.current_state\\n queue.put((initial_state, []))\\n seen_states.append(initial_state)\\n board_size = initial_game.board.length\\n\\n while not queue.empty():\\n # Get first from queue.\\n parent_node = copy.deepcopy(queue.get())\\n\\n # Create a game instance from the state\\n parent_state = parent_node[0]\\n parent_game = create_game_from_state(parent_state, board_size)\\n\\n # Loop over the possible moves in the state.\\n possible_moves = parent_game.possible_moves\\n for move in possible_moves:\\n # Get the move specifics.\\n vehicle, steps = move\\n new_position = vehicle.speculate_new_position(steps)\\n\\n # Execute the move to create a child node.\\n child_game = copy.deepcopy(parent_game)\\n child_vehicle = child_game.vehicles.get(vehicle.id)\\n child_game.move(child_vehicle, new_position)\\n\\n # Heuristic: Check if the state is unique.\\n child_state = child_game.current_state\\n if not is_state_unique(seen_states, child_state):\\n continue\\n\\n # Otherwise, remember the state.\\n seen_states.append(child_state)\\n \\n if len(seen_states) % 100 == 0:\\n print(f\\\"Seen states: {len(seen_states)}\\\")\\n\\n # Add new child node to queue.\\n moves_until_state = copy.deepcopy(parent_node[1])\\n moves_until_state.append((vehicle.id, steps))\\n child_node = (child_state, moves_until_state)\\n queue.put(child_node)\\n\\n # Heuristic: check if the boxes from the red car up until the exit are free.\\n if is_exit_reachable(child_game):\\n last_move = finish_game(child_game)\\n moves = child_node[1]\\n moves.append(last_move)\\n solved_cases[0] = moves\\n\\n if child_game.is_finished():\\n print(f\\\"Seen states: {len(seen_states)}\\\")\\n return solved_cases\\n\\n return \\\"No solved cases have been found :(\\\"\",\n \"def solveOneStep(self):\\n ### Student code goes here\\n # Mark this move as explored\\n self.visited[self.currentState] = True\\n self.visited_states.append(self.currentState.state)\\n\\n # Get move to make\\n movables = self.gm.getMovables()\\n # print(\\\"EXPLORING GAME STATE \\\" + str(self.gm.getGameState()) + \\\"---------------------------------------------------------\\\")\\n to_move = self.currentState.nextChildToVisit # movables index\\n # print(\\\"depth \\\", self.currentState.depth)\\n\\n # Return if done\\n if self.currentState.state == self.victoryCondition:\\n # print(\\\"DONE\\\")\\n return True\\n\\n # If current state has no children, make children\\n if not self.currentState.children:\\n for movable_statement in movables:\\n # Make the move\\n # print(\\\"implementing move \\\", movable_statement)\\n self.gm.makeMove(movable_statement)\\n\\n # Create a new state with this move made\\n new_state = self.gm.getGameState()\\n # print (\\\"new state \\\", new_state)\\n\\n # If the new state hasn't been visited and isn't in the queue then add it as a child and to the queue\\n if (new_state not in self.visited_states):\\n new_gs = GameState(new_state, self.currentState.depth + 1, movable_statement)\\n new_gs.parent = self.currentState\\n self.currentState.children.append(new_gs)\\n self.currentState.nextChildToVisit = to_move + 1\\n self.visited[new_gs] = True\\n self.visited_states.append(new_state)\\n self.gs_queue.append(new_gs)\\n\\n self.gm.reverseMove(movable_statement)\\n\\n # Return false if no more to explore\\n if not self.gs_queue:\\n return False\\n\\n # Revert to state at when current and next start to change\\n root_curr = self.currentState\\n self.currentState = self.gs_queue.popleft()\\n root_new = self.currentState\\n\\n # Backtrack to when current node and new node start to diverge\\n if root_new.depth == root_curr.depth:\\n while root_curr.state != root_new.state:\\n self.gm.reverseMove(root_curr.requiredMovable)\\n root_curr = root_curr.parent\\n root_new = root_new.parent\\n else:\\n while root_curr.requiredMovable:\\n self.gm.reverseMove(root_curr.requiredMovable)\\n root_curr = root_curr.parent\\n\\n # Return game master to state that we are exploring\\n # Find path between root and current state\\n path = []\\n currNode = self.currentState\\n while currNode != root_curr:\\n path.append(currNode.requiredMovable)\\n currNode = currNode.parent\\n\\n # Created backwards path, now make moves from root to current state\\n path.reverse()\\n for movable_statement in path:\\n self.gm.makeMove(movable_statement)\\n\\n return False\",\n \"def UCT(player, rootstate, iter, prev_board, verbose=False):\\n rootnode = Node(player=player, state=rootstate, preState=prev_board)\\n\\n for i in range(iter):\\n node = rootnode\\n state = deepcopy(rootstate)\\n pre_state = deepcopy(state)\\n currentPlayer = -player # turn of the player x\\n\\n # Select\\n while (\\n node.untriedMoves == [] and node.childNodes != []\\n ): # node is fully expanded and non-terminal, loop until have untried Child or it is leaf node\\n # choose the next child\\n node = node.UCTChooseChild()\\n # update the state of the board ((x,y),(x,y))\\n pre_state = deepcopy(state)\\n state = board_after_move_and_capturing(node.move[0], node.move[1], state)\\n currentPlayer = -currentPlayer\\n\\n # Expand\\n if (\\n node.untriedMoves != []\\n ): # if we can expand (i.e. state/node is non-terminal)\\n m = random.choice(node.untriedMoves)\\n # update the state of the board\\n pre_state = deepcopy(state)\\n state = board_after_move_and_capturing(m[0], m[1], state)\\n currentPlayer = -currentPlayer\\n\\n node = node.addChild(m, state, pre_state) # add child and descend tree\\n\\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\\n height = 0\\n while get_winner(state) == 0: # while state is non-terminal\\n if height > 30:\\n break\\n choice = random.choice(get_all_legal_moves(pre_state, state, currentPlayer))\\n # update the state of the board\\n pre_state = deepcopy(state)\\n state = board_after_move_and_capturing(choice[0], choice[1], state)\\n # print_board(state)\\n currentPlayer = -currentPlayer\\n height += 1\\n\\n # cout board\\n # print(f\\\"The board becomes {board_to_string(state)}\\\")\\n\\n # Backpropagate\\n while (\\n node != None\\n ): # backpropagate from the expanded node and work back to the root node\\n if height > 30:\\n winner = player\\n else:\\n winner = get_winner(state)\\n\\n if winner == -player:\\n node.updateNode(1)\\n else:\\n node.updateNode(0)\\n\\n # state is terminal. Update node with result from POV of node.playerJustMoved\\n node = node.parentNode\\n\\n # cout root node\\n # for i in rootnode.childNodes:\\n # print(i.move)\\n # print(i.wins)\\n # print(i.visits)\\n\\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[\\n -1\\n ].move # return the move that was most visited\",\n \"def do(self, state):\\r\\n\\r\\n # action is [0,1,2,3]: left, right, up, down\\r\\n _state = json.loads(state)\\r\\n if _state[0] == 0:\\r\\n self.Q[state][2] = 0\\r\\n if _state[0] == self.max_row:\\r\\n self.Q[state][3] = 0\\r\\n if _state[1] == 0:\\r\\n self.Q[state][0] = 0\\r\\n if _state[1] == self.max_col:\\r\\n self.Q[state][1] = 0\\r\\n action_map = {0: [0, -1], # left\\r\\n 1: [0, 1], # right\\r\\n 2: [-1, 0], # up\\r\\n 3: [1, 0], # down\\r\\n }\\r\\n chance = np.random.rand()\\r\\n if self.eps > chance:\\r\\n while True:\\r\\n action = np.random.randint(0,4)\\r\\n next = np.array(_state) + np.array(action_map[action])\\r\\n if all(next>=0) and next[0]<=self.max_row and next[1]<=self.max_col:\\r\\n break\\r\\n else:\\r\\n action = np.argmax(self.Q[state])\\r\\n #print(_state,action,self.Q[state])\\r\\n\\r\\n \\\"\\\"\\\"Please Fill Your Code Here.\\r\\n \\\"\\\"\\\"\\r\\n return action\",\n \"def leafEvaluation(self, state):\\n\\n \\\"\\\"\\\"\\n Use random generated values for now\\n \\\"\\\"\\\"\\n z = np.random.randint(2)\\n v = random.uniform(0, 1) \\n return (1-LAMBDA) * v + LAMBDA * z\",\n \"def __call__(self, state):\\n if random.random() > self._epsilon:\\n return self._max_policy(state)\\n return random.choice(np.arange(self._action_size))\",\n \"def uct(root_state, iter_max, verbose=False):\\n\\n root_node = Node(state=root_state)\\n\\n for i in range(iter_max):\\n node = root_node\\n state = root_state.clone()\\n\\n # Select\\n # node is fully expanded and non-terminal\\n while not node.untried_moves and node.child_nodes:\\n node = node.uct_select_child()\\n state.do_move(node.move)\\n\\n # Expand\\n # if we can expand (i.e. state/node is non-terminal)\\n if node.untried_moves:\\n m = random.choice(node.untried_moves)\\n state.do_move(m)\\n # add child and descend tree\\n node = node.add_child(m, state)\\n\\n # Rollout - this can often be made orders of magnitude quicker\\n # using a state.get_random_move() function\\n # while state is non-terminal\\n while state.get_moves():\\n state.do_move(random.choice(state.get_moves()))\\n\\n # Backpropagate\\n # backpropagate from the expanded node and work back to the root node\\n while node is not None:\\n # state is terminal. Update node with result\\n # from POV of node.playerJustMoved\\n node.update(state.get_result(node.player_just_moved))\\n node = node.parent_node\\n\\n # Output some information about the tree - can be omitted\\n if verbose:\\n print(root_node.tree_to_string(0))\\n else:\\n print(root_node.children_to_string())\\n\\n # return the move that was most visited\\n return sorted(root_node.child_nodes,\\n key=lambda c: c.Q)[-1].move\",\n \"def get_successors(state): \\n \\n child_states = []\\n \\n size = len(state)\\n i = 0\\n j = 0\\n for i in range (size):\\n if 0 in state[i]:\\n for j in range (size):\\n if state[i][j] == 0:\\n break \\n break\\n\\n if j != size-1:\\n child_states.append ((\\\"Left\\\", swap_cells(state, i, j, i, j+1)))\\n if j != 0:\\n child_states.append ((\\\"Right\\\", swap_cells(state, i, j, i, j-1)))\\n if i != size-1:\\n child_states.append ((\\\"Up\\\", swap_cells(state, i, j, i+1, j)))\\n if i != 0:\\n child_states.append ((\\\"Down\\\", swap_cells(state, i, j, i-1, j)))\\n \\n return child_states\",\n \"def find_best_move(state: GameState) -> None:\",\n \"def evaluate(self, state):\\n\\n fitness = 0\\n\\n for i in range(1, len(state)):\\n if state[i] != state[i - 1]:\\n fitness += 1\\n\\n return fitness\",\n \"def build(self, w=1):\\n states = self.a.states.values()\\n reachedStates = set()\\n remainingStates = set()\\n\\n _open = Heap()\\n\\n for s in states:\\n if s.acceptingFunction is not None:\\n s.h = 0\\n _open.push(0, s)\\n reachedStates.add(s)\\n else:\\n remainingStates.add(s)\\n\\n if not reachedStates:\\n Color.RED.print(\\\"No states reached!\\\")\\n Color.RED.print(\\\"States were not reached: %s\\\" % remainingStates)\\n raise Exception(\\\"Query has no goal, please review it!\\\")\\n\\n while _open:\\n (h, s) = _open.pop()\\n\\n for cS in s._prev():\\n if h+w < cS.h:\\n reachedStates.add(cS)\\n _open.push(h+w, cS)\\n\\n cS.h = h+w\\n if cS in remainingStates:\\n remainingStates.remove(cS)\\n\\n if remainingStates:\\n Color.RED.print(\\\"Some states were not reached: %s\\\" % remainingStates)\\n Color.YELLOW.print(\\\"States reached: %s\\\" % reachedStates)\\n raise Exception(\\\"Query is probably unfinished, please review it\\\")\\n\\n reached = [(s.h, s) for s in reachedStates]\\n reached.sort()\\n\\n if __debug__:\\n Color.GREEN.print(\\\"Heuristic:\\\")\\n for (h, s) in reached:\\n Color.GREEN.print(\\\" * %d: %s\\\" % (h, s))\\n\\n for s in states:\\n s.prepare()\\n\\n return self.a\",\n \"def select_action(self, state):\\r\\n policy_s = self.epsilon_greedy_probs(self.nA, self.Q[state], self.count, self.epsilon)\\r\\n return np.random.choice(np.arange(self.nA), p=policy_s)\",\n \"def myHeuristic3(state, problem=None):\\n #canto =[(1,1), (1,2), (2,1), (2,2), (36,36), (35,36), (35,35), (36,35), (1,36),(1,35),(2,36),(2,35),(36,1),(36,2),(35,1),(35,2)]\\n canto = []\\n for l in range(2):\\n for c in range(2):\\n canto.append((l,c))\\n heru = abs(state[0] - 1) + abs(state[1] - 1)\\n #heru=10\\n if state in canto:\\n #print(\\\"sim\\\")\\n heru = heru * 0.5\\n return heru\",\n \"def astar(model, start, end):\\n\\n # Create start and end node\\n start_node = Node(None,0.0, start)\\n start_node.g = 0\\n start_node.h = calculate_h(start_node,end)\\n start_node.f = start_node.g + start_node.h \\n \\n end_node = Node(None,0.0, end)\\n end_node.g = end_node.h = end_node.f = 0\\n\\n # Initialize both open and closed list\\n open_list = []\\n closed_list = []\\n\\n # Add the start node\\n open_list.append(start_node)\\n\\n\\n\\n iterations = 0\\n # Loop until you find the end\\n while len(open_list) > 0:\\n iterations += 1\\n # print(iterations)\\n print('openlist={},closed_list={}'.format(len(open_list),len(closed_list)))\\n\\n # Get the current node\\n current_node = open_list[0]\\n current_index = 0\\n for index, item in enumerate(open_list):\\n if item.f < current_node.f:\\n current_node = item\\n current_index = index\\n\\n # Pop current off open list, add to closed list\\n open_list.pop(current_index)\\n closed_list.append(current_node)\\n\\n print('currentstate={}, parentaction={}, f={}'.format(current_node.state, current_node.parent_action, current_node.f))\\n\\n # run_sim(current_node.parent_action)\\n\\n # Found the goal\\n if current_node == end_node: \\n actions = []\\t\\n current = current_node\\n while current is not None:\\n actions.append(current.parent_action)\\n current = current.parent_state\\n # print(current.state)\\n return actions[::-1] # Return reversed path\\n\\n # Generate children states\\n children = []\\n\\n for action in discretized_action: # all possible actions\\n\\n # Get new state node (closest to discretized state space)\\n \\n # new_state = get_next_state_from_model(model,current_node,action)\\n new_state = get_next_state_from_eqn(current_node,action)\\n # new_state = run_sim(action)\\n # print(new_state)\\n\\n # # Make sure within range\\n # if node_position[0] > (len(maze) - 1) or node_position[0] < 0 or node_position[1] > (len(maze[len(maze)-1]) -1) or node_position[1] < 0:\\n # continue\\n\\n # # Make sure walkable terrain\\n # if maze[node_position[0]][node_position[1]] != 0:\\n # continue\\n\\n #############Make sure to wrap theta \\n \\n # Create new node\\n new_node = Node(current_node, action, new_state)\\n\\n # Append\\n children.append(new_node)\\n\\n # Loop through children\\n for child in children:\\n\\n # Child is on the closed list\\n for closed_child in closed_list:\\n if child == closed_child:\\t\\n continue\\n\\n # Create the f, g, and h values\\n child.g = calculate_g(child,current_node) #current_node.g + 1\\n child.h = calculate_h(child,end_node)\\n child.f = child.g + child.h\\n \\n # \\tprint(child.state, child.parent_action, child.f) \\n \\n # Child is already in the open list\\n for open_node in open_list:\\n if child == open_node and child.g > open_node.g:\\n continue\\n\\n # Add the child to the open list\\n open_list.append(child)\",\n \"def UCT(rootstate, itermax, verbose=False):\\n\\n rootnode = Node(state=rootstate)\\n\\n for i in range(itermax):\\n node = rootnode\\n state = rootstate.Clone()\\n\\n # Select\\n while node.untriedMoves == [] and node.childNodes != []: # node is fully expanded and non-terminal\\n node = node.UCTSelectChild()\\n state.DoMove(node.move)\\n\\n # Expand\\n expand = True\\n while expand and node.untriedMoves != []: # if we can expand (i.e. state/node is non-terminal)\\n m = random.choice(node.untriedMoves)\\n # print(\\\"[Expand] Untried move %s, %s, %s\\\" % (m[0], m[1], m[2]))\\n expand = not state.DoMove(m)\\n node = node.AddChild(m, state) # add child and descend tree\\n\\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\\n while state.GetMoves() != []: # while state is non-terminal\\n state.DoMove(random.choice(state.GetMoves()))\\n\\n # Backpropagate\\n while node != None: # backpropagate from the expanded node and work back to the root node\\n node.Update(state.GetResult()) # state is terminal. Update node with result from POV of node.playerJustMoved\\n node = node.parentNode\\n\\n # Output some information about the tree - can be omitted\\n if (verbose):\\n print(rootnode.TreeToString(0))\\n else:\\n print(rootnode.ChildrenToString())\\n\\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[-1].move # return the move that was most visited\",\n \"def next_possible_states(path, check_dict, check):\\r\\n \\r\\n current_state_tuple = path[-1]\\r\\n state_container = []\\r\\n x = current_state_tuple[1][0]\\r\\n y = current_state_tuple[1][1]\\r\\n current_state = current_state_tuple[0]\\r\\n\\r\\n # Down\\r\\n if y < 3:\\r\\n new_state = copy.deepcopy(current_state)\\r\\n new_state[y][x] = new_state[y + 1][x]\\r\\n new_state[y + 1][x] = 0\\r\\n if not been_there(new_state, check_dict, check):\\r\\n new_index = (x, y + 1)\\r\\n h1 = euclidean_dist(new_state, path)\\r\\n new_state_tuple = (new_state, new_index, h1)\\r\\n state_container.append(new_state_tuple)\\r\\n\\r\\n # Up\\r\\n if y > 0:\\r\\n new_state = copy.deepcopy(current_state)\\r\\n if y == 1 and x == 0:\\r\\n new_state[y][x] = new_state[y - 1][x]\\r\\n new_state[y - 1][x] = 0\\r\\n if is_goal(new_state):\\r\\n new_index = (x, y - 1)\\r\\n h1 = euclidean_dist(new_state, path)\\r\\n new_state_tuple = (new_state, new_index, h1)\\r\\n state_container.append(new_state_tuple)\\r\\n elif y > 1:\\r\\n new_state[y][x] = new_state[y - 1][x]\\r\\n new_state[y - 1][x] = 0\\r\\n if not been_there(new_state, check_dict, check):\\r\\n new_index = (x, y - 1)\\r\\n h1 = euclidean_dist(new_state, path)\\r\\n new_state_tuple = (new_state, new_index, h1)\\r\\n state_container.append(new_state_tuple)\\r\\n\\r\\n # Left\\r\\n if x > 0 and y > 0:\\r\\n new_state = copy.deepcopy(current_state)\\r\\n new_state[y][x] = new_state[y][x - 1]\\r\\n new_state[y][x - 1] = 0\\r\\n if not been_there(new_state, check_dict, check):\\r\\n new_index = (x - 1, y)\\r\\n h1 = euclidean_dist(new_state, path)\\r\\n new_state_tuple = (new_state, new_index, h1)\\r\\n state_container.append(new_state_tuple)\\r\\n\\r\\n # Right\\r\\n if x < 2 and y > 0:\\r\\n new_state = copy.deepcopy(current_state)\\r\\n new_state[y][x] = new_state[y][x + 1]\\r\\n new_state[y][x + 1] = 0\\r\\n if not been_there(new_state, check_dict, check):\\r\\n new_index = (x + 1, y)\\r\\n h1 = euclidean_dist(new_state, path)\\r\\n new_state_tuple = (new_state, new_index, h1)\\r\\n state_container.append(new_state_tuple)\\r\\n\\r\\n return state_container\",\n \"def choose_action(self, state):\\n prob = [] # Probability distribution\\n for i in range(len(ACTIONS)):\\n prob.append(self.epsilon/4)\\n Q_func = self.policy.predict(process_state(state))\\n Q_vals = Q_func[0]\\n max_index = []\\n Qmax = np.amax(Q_vals)\\n for i in range(len(prob)):\\n if Q_vals[i] == Qmax:\\n # max_index.append(i)\\n prob[i] = 1 - self.epsilon + self.epsilon/4\\n break\\n # ind = np.random.choice(max_index)\\n # prob[ind] = 1 - self.epsilon + self.epsilon/4\\n action = np.random.choice(ACTIONS, p = prob)\\n return action\",\n \"def epsilon_greedy(Q, epsilon, state):\\n random_number = random.random()\\n if (random_number < epsilon) and (state not in critical_states):\\n return env.action_space.sample()\\n\\n else:\\n return np.argmax(Q[state])\",\n \"def ai_strategy2(gstate: TicTacToe, game_tree):\\n status, player = gstate.game_status\\n\\n if status != \\\"turn\\\":\\n logging.warning(\\\"Game status = %s. No move needed.\\\", status)\\n return None\\n\\n win, lose = (-1, 1) if player == \\\"1\\\" else (1, -1)\\n next_move_vals = {\\n idx: win * game_tree[state][\\\"value\\\"] for idx, state in gstate.next_moves.items()\\n }\\n max_val = max(next_move_vals.values())\\n moves = {\\n idx: (\\n game_tree[gstate.next_moves[idx]][\\\"outcomes\\\"].get(str(win), 0),\\n game_tree[gstate.next_moves[idx]][\\\"outcomes\\\"].get(str(0), 0),\\n game_tree[gstate.next_moves[idx]][\\\"outcomes\\\"].get(str(lose), 0)\\n )\\n for idx, val in next_move_vals.items() if val == max_val\\n }\\n\\n win_ct = {idx: vals[0] for idx, vals in moves.items()}\\n win_pct = {idx: vals[0] / sum(vals) for idx, vals in moves.items()}\\n lose_pct = {idx: vals[2] / sum(vals) for idx, vals in moves.items()}\\n wl_ratio = {idx: vals[0] / max(vals[2], 0.5)\\n for idx, vals in moves.items()}\\n\\n # criteria, agg_func = lose_pct, min\\n # criteria, agg_func = win_pct, max\\n criteria, agg_func = wl_ratio, max\\n\\n if max_val == 1 and 1 in win_ct.values():\\n move = [idx for idx, val in win_ct.items() if val == 1][0]\\n else:\\n move = random.choice(\\n [idx for idx, val in criteria.items() if val ==\\n agg_func(criteria.values())]\\n )\\n\\n logging.debug(\\\"move: %i; value: %i, win paths %%: %.1f%%, lose paths %%: %.1f%%, moves: %s\\\\n\\\",\\n move, max_val, win_pct[move] * 100, lose_pct[move] * 100, moves)\\n\\n # trash talk\\n if max_val == 1:\\n print(\\n \\\"*beep* *boop* *beep*\\\"\\n \\\" -=[ I calculate chances of winning to be 100% ]=- \\\"\\n \\\"*beep* *boop* *beep*\\\"\\n )\\n\\n return move\",\n \"def test_create_state(self):\\n\\n # States should be abled to be identified by numbers or by strings I suppose.\\n # I don't imagine that strings will ever be used.\\n mdp = MDP()\\n mdp.add_state(0)\\n mdp.add_state(1)\\n mdp.add_state(2)\\n mdp.add_state(3)\\n mdp.add_state(4)\\n mdp.add_state(5, terminal=True)\\n self.assertEqual(mdp.num_states(), 6)\",\n \"def learn(state=\\\"_________\\\"):\\n game_tree = gen_game_tree(state)\\n with open(GAME_TREE_FILE, \\\"w\\\") as gt_file:\\n json.dump(game_tree, gt_file, indent=4)\",\n \"def __init__(self, state: State):\\n self._state = state\\n self._parent = None\\n self._untried_edges = self._state.possible_actions\\n self.children = {} # {AbstractAction: AbstractState}\\n self.tot_reward = 0\\n self.num_samples = 0\",\n \"def max_diffs(state):\\n # your code here\\n return best_action(state, pig_actions, Q_pig, win_diff)\",\n \"def act(self, state):\\n\\n afterstates = state.afterstates()\\n\\n if self.choose_to_explore(): # explore\\n action_id = self._generator.integers(0, len(afterstates))\\n return afterstates[action_id]\\n\\n estimates = self.act_with_minimax(afterstates, self._minimax_depth,\\n maximizing=False, best=None,\\n initial=True)\\n min = np.amin(estimates)\\n argmin = np.argwhere(estimates == min).flatten()\\n action_id = argmin[self._generator.integers(0, len(argmin))]\\n return afterstates[action_id]\",\n \"def getAction(self, gameState):\\n \\\"*** YOUR CODE HERE ***\\\"\\n self.NumAgents = gameState.getNumAgents()\\n self.root = gameState\\n # alpha-beta-search\\n # one call of Max_Value is enough to iterate the whole tree and keep the optimal move\\n self.Max_Value(gameState, -10000, 10000, self.index, self.depth)\\n return self.move\\n util.raiseNotDefined()\",\n \"def simulate(params,n_states,n_trials,env = \\\"rich\\\", policy=\\\"softmax\\\",\\\\\\n D=0.5, mod = \\\"constant\\\",thresh = 0, k=1,rnd_seeds = None, V0=0.0, full=False,\\n rmag = 1, lmag = 0):\\n\\n\\tdef calc_D(state):\\n\\t\\t\\\"\\\"\\\"\\n\\t\\tcalculates D for the current trial and returns\\n\\t\\tthe updated state tracker for D and respective betas\\n\\n\\t\\tD represents dopamine levels (equivalent of rho in OpAL)\\n\\t\\tScales between 0 and 1, with 1 high level of DA\\n\\t\\t\\\"\\\"\\\"\\n\\t\\tif t < thresh:\\n\\t\\t\\tstate.D_g[t] = 0.5\\n\\t\\t\\tstate.D_n[t] = 0.5\\n\\t\\telse:\\n\\t\\t\\tif mod == \\\"constant\\\":\\n\\t\\t\\t\\tstate.D_g[t] = D\\n\\t\\t\\t\\tstate.D_n[t] = 1-D\\n\\t\\t\\tif mod == \\\"value\\\":\\n\\t\\t\\t\\t# NOTE: if rmag and lmag is 1/0, can just use V\\n\\t\\t\\t\\t# average of two actions\\n\\t\\t\\t\\tV = np.mean(1/2*(state.QG[t,:] - state.QN[t,:])) # state average(?) \\n\\t\\t\\t\\tV = 1/(1 + np.exp(-V*k)) # translate between 0 and 1\\n\\t\\t\\t\\tstate.D_g[t] = V \\n\\t\\t\\t\\tstate.D_n[t] = 1 - V\\n\\t\\treturn state\\n\\n\\n\\tdef generate_state():\\n\\t\\t\\\"\\\"\\\"\\n\\t\\tGet appropriate reward probabilities and magnitudes\\n\\t\\tfor the specified environment type\\n\\t\\t\\\"\\\"\\\"\\n\\n\\t\\tprobs = calc_probs(env)\\n\\t\\tn_options = len(probs)\\n\\n\\t\\t# feedback for agent\\n\\t\\tr_mag = np.zeros(n_options) + rmag\\n\\t\\tl_mag = np.zeros(n_options) + lmag\\n\\n\\t\\tnew_state = Bogacz(n_trials, n_options, probs, r_mag, l_mag, V0=V0)\\n\\t\\treturn new_state\\n\\n\\n\\t# learning rate, damping, decay, softmax temp\\n\\talpha_a, epsilon, lbda, beta = params\\n\\tstates = []\\n\\n\\t# do the thing\\n\\tfor s in np.arange(n_states):\\n\\n\\t\\t# check if random seed provided\\n\\t\\tif rnd_seeds is not None:\\n\\t\\t\\trandom.seed(rnd_seeds[s])\\n\\t\\t\\tnp.random.seed(rnd_seeds[s])\\n\\n\\t\\tstate = generate_state()\\n\\t\\tfor t in range(n_trials):\\n\\n\\t\\t\\tstate.idx = t\\n\\t\\t\\tstate=calc_D(state)\\t\\t\\t\\t\\t# get D\\n\\t\\t\\tstate.policy_softmax(beta)\\n\\t\\t\\tstate.act(alpha_a, epsilon, lbda)\\t# update \\n\\n\\t\\t\\tif full:\\n\\t\\t\\t\\tstate.update_other_actions(alpha_a, epsilon, lbda)\\n\\n\\t\\tstates.append(state)\\t\\t\\t\\t\\t# save sim\\n\\n\\treturn states\",\n \"def generate_cubes(num_states, distance, startstate='RRRRRRRRRBBBBBBBBBOOOOOOOOOGGGGGGGGGWWWWWWWWWYYYYYYYYY'):\\n terminal_states = {'RRRRRRRRRYYYYYYYYYOOOOOOOOOWWWWWWWWWBBBBBBBBBGGGGGGGGG',\\n 'BBBBBBBBBYYYYYYYYYGGGGGGGGGWWWWWWWWWOOOOOOOOORRRRRRRRR',\\n 'YYYYYYYYYBBBBBBBBBWWWWWWWWWGGGGGGGGGRRRRRRRRROOOOOOOOO',\\n 'BBBBBBBBBWWWWWWWWWGGGGGGGGGYYYYYYYYYRRRRRRRRROOOOOOOOO',\\n 'OOOOOOOOOGGGGGGGGGRRRRRRRRRBBBBBBBBBWWWWWWWWWYYYYYYYYY',\\n 'GGGGGGGGGOOOOOOOOOBBBBBBBBBRRRRRRRRRYYYYYYYYYWWWWWWWWW',\\n 'GGGGGGGGGRRRRRRRRRBBBBBBBBBOOOOOOOOOWWWWWWWWWYYYYYYYYY',\\n 'GGGGGGGGGYYYYYYYYYBBBBBBBBBWWWWWWWWWRRRRRRRRROOOOOOOOO',\\n 'RRRRRRRRRWWWWWWWWWOOOOOOOOOYYYYYYYYYGGGGGGGGGBBBBBBBBB',\\n 'OOOOOOOOOWWWWWWWWWRRRRRRRRRYYYYYYYYYBBBBBBBBBGGGGGGGGG',\\n 'WWWWWWWWWRRRRRRRRRYYYYYYYYYOOOOOOOOOBBBBBBBBBGGGGGGGGG',\\n 'BBBBBBBBBOOOOOOOOOGGGGGGGGGRRRRRRRRRWWWWWWWWWYYYYYYYYY',\\n 'BBBBBBBBBRRRRRRRRRGGGGGGGGGOOOOOOOOOYYYYYYYYYWWWWWWWWW',\\n 'RRRRRRRRRGGGGGGGGGOOOOOOOOOBBBBBBBBBYYYYYYYYYWWWWWWWWW',\\n 'YYYYYYYYYRRRRRRRRRWWWWWWWWWOOOOOOOOOGGGGGGGGGBBBBBBBBB',\\n 'YYYYYYYYYOOOOOOOOOWWWWWWWWWRRRRRRRRRBBBBBBBBBGGGGGGGGG',\\n 'GGGGGGGGGWWWWWWWWWBBBBBBBBBYYYYYYYYYOOOOOOOOORRRRRRRRR',\\n 'WWWWWWWWWGGGGGGGGGYYYYYYYYYBBBBBBBBBRRRRRRRRROOOOOOOOO',\\n 'OOOOOOOOOYYYYYYYYYRRRRRRRRRWWWWWWWWWGGGGGGGGGBBBBBBBBB',\\n 'RRRRRRRRRBBBBBBBBBOOOOOOOOOGGGGGGGGGWWWWWWWWWYYYYYYYYY',\\n 'WWWWWWWWWBBBBBBBBBYYYYYYYYYGGGGGGGGGOOOOOOOOORRRRRRRRR',\\n 'WWWWWWWWWOOOOOOOOOYYYYYYYYYRRRRRRRRRGGGGGGGGGBBBBBBBBB',\\n 'OOOOOOOOOBBBBBBBBBRRRRRRRRRGGGGGGGGGYYYYYYYYYWWWWWWWWW',\\n 'YYYYYYYYYGGGGGGGGGWWWWWWWWWBBBBBBBBBOOOOOOOOORRRRRRRRR'}\\n states = []\\n while len(states) < num_states:\\n x = RubiksCubeOld(startstate)\\n for j in range(distance):\\n x.apply_move(np.random.randint(0,18))\\n newstate = x.get_state()\\n if newstate not in terminal_states: states.append(newstate)\\n states = list(set(states))\\n\\n return states\",\n \"def solveOneStep(self):\\n ### Student code goes here\\n if (self.currentState.state == self.victoryCondition) or (self.currentState not in self.visited):\\n self.visited[self.currentState] = True\\n win_or_not = self.currentState.state == self.victoryCondition\\n return win_or_not\\n\\n if not self.currentState.nextChildToVisit: \\n its = 0\\n for movable in self.gm.getMovables():\\n its += 1\\n # time test\\n # too long \\n if its == \\\"too long\\\":\\n return \\\"too long\\\"\\n #make every move in movable\\n self.gm.makeMove(movable)\\n new = self.gm.getGameState()\\n new_gs = GameState(new, self.currentState.depth+1, movable)\\n \\n if new_gs not in self.visited:\\n new_gs.parent = self.currentState\\n self.currentState.children.append(new_gs)\\n self.gm.reverseMove(movable) \\n \\n num_children = len(self.currentState.children)\\n if self.currentState.nextChildToVisit < num_children:\\n new = self.currentState.children[self.currentState.nextChildToVisit]\\n self.currentState.nextChildToVisit = self.currentState.nextChildToVisit + 1\\n self.gm.makeMove(new.requiredMovable)\\n self.currentState = new\\n #recurse\\n return self.solveOneStep()\\n else:\\n self.currentState.nextChildToVisit = self.currentState.nextChildToVisit + 1\\n self.gm.reverseMove(self.currentState.requiredMovable)\\n self.currentState = self.currentState.parent\\n #recurse\\n return self.solveOneStep()\",\n \"def evaluateBoardState(self, board):\\n\\n \\\"\\\"\\\"\\n These are the variables and functions for board objects which may be helpful when creating your Agent.\\n Look into board.py for more information/descriptions of each, or to look for any other definitions which may help you.\\n\\n Board Variables:\\n board.width \\n board.height\\n board.last_move\\n board.num_to_connect\\n board.winning_zones\\n board.score_array \\n board.current_player_score\\n\\n Board Functions:\\n get_cell_value(row, col)\\n try_move(col)\\n valid_move(row, col)\\n valid_moves()\\n terminal(self)\\n legal_moves()\\n next_state(turn)\\n winner()\\n \\\"\\\"\\\"\\n if self.id == 1:\\n opponent_id = 2\\n else:\\n opponent_id = 1\\n\\n maxvalue = 100000\\n minvalue = -maxvalue\\n winner = board.winner()\\n if winner == self.id:\\n return maxvalue\\n elif winner == opponent_id:\\n return minvalue\\n size_y = board.height\\n size_x = board.width\\n map_ = []\\n num_to_connect = board.num_to_connect\\n total_points = 0\\n\\n multiply_reachable = 1\\n multiply_oddeven = 1\\n # basically this function is calculating all the possible win positions\\n # more pieces in a possible win position will be counted with more weights\\n # a win position with X pieces in it will be counted as X^2 points\\n # initialise the zones maps\\n for i in range(size_y):\\n map_.append([])\\n for j in range(size_x):\\n map_[i].append([])\\n\\n # Fill in the horizontal win positions\\n for i in range(size_y):\\n for j in range(size_x - num_to_connect + 1):\\n points = 0\\n self_pieces_count = 0\\n opponent_pieces_count = 0\\n for k in range(num_to_connect):\\n if board.board[i][j + k] == opponent_id:\\n opponent_pieces_count += 1\\n elif board.board[i][j + k] == self.id:\\n points += len(board.winning_zones[j+k][i])\\n if (self.id == 1 and i % 2 == 1) or (self.id == 2 and i%2 == 0):\\n points *= multiply_oddeven\\n self_pieces_count += 1\\n if self_pieces_count == 3 and opponent_pieces_count == 0:\\n if j - 1 >= 0 and board.board[i][j + 3] == 0 and board.board[i][j - 1] == 0 \\\\\\n and board.try_move(j + 3) == i and board.try_move(j - 1) == i:\\n return maxvalue\\n elif j + 4 < size_y and board.board[i][j + 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j + 4) == i and board.try_move(j) == i:\\n return maxvalue\\n else:\\n for k in range(num_to_connect):\\n if board.board[i][j + k] == 0 and board.try_move(j + k) == i:\\n points *= multiply_reachable\\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\\n if j - 1 >= 0 and board.board[i][j + 3] == 0 and board.board[i][j - 1] == 0 \\\\\\n and board.try_move(j + 3) == i and board.try_move(j - 1) == i:\\n return minvalue\\n elif j + 4 < size_y and board.board[i][j + 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j + 4) == i and board.try_move(j) == i:\\n return minvalue\\n # else:\\n # for k in range(num_to_connect):\\n # if board.board[i][j + k] == 0 and board.try_move(j + k) == i:\\n # points *= -multiply_reachable\\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\\n total_points += points\\n\\n # Fill in the vertical win positions\\n for i in range(size_x):\\n for j in range(size_y - num_to_connect + 1):\\n points = 0\\n self_pieces_count = 0\\n for k in range(num_to_connect):\\n if board.board[j + k][i] == opponent_id:\\n opponent_pieces_count += 1\\n elif board.board[j + k][i] == self.id:\\n points += len(board.winning_zones[i][j+k])\\n if (self.id == 1 and (j+k) % 2 == 1) or (self.id == 2 and (j+k)%2 == 0):\\n points *= multiply_oddeven\\n self_pieces_count += 1\\n points *= multiply_reachable\\n # if opponent_pieces_count == 3 and self_pieces_count == 0:\\n # points *= -1\\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\\n total_points += points\\n\\n # Fill in the forward diagonal win positions\\n for i in range(size_y - num_to_connect + 1):\\n for j in range(size_x - num_to_connect + 1):\\n points = 0\\n self_pieces_count = 0\\n opponent_pieces_count = 0\\n for k in range(num_to_connect):\\n if board.board[i + k][j + k] == opponent_id:\\n opponent_pieces_count += 1\\n elif board.board[i + k][j + k] == self.id:\\n points += len(board.winning_zones[j+k][i+k])\\n if (self.id == 1 and (i+k) % 2 == 1) or (self.id == 2 and (i+k)%2 == 0):\\n points *= multiply_oddeven\\n self_pieces_count += 1\\n if self_pieces_count == 3 and opponent_pieces_count == 0:\\n if i - 1 >= 0 and j - 1 >= 0 and board.board[i + 3][j + 3] == 0 and board.board[i - 1][j - 1] == 0 \\\\\\n and board.try_move(j + 3) == i + 3 and board.try_move(j - 1) == i - 1:\\n return maxvalue\\n elif i + 4 < size_y and j + 4 < size_x and board.board[i + 4][j + 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j + 4) == i + 4 and board.try_move(j) == i:\\n return maxvalue\\n else:\\n for k in range(num_to_connect):\\n if board.board[i + k][j + k] == 0 and board.try_move(j + k) == i + k:\\n points *= multiply_reachable\\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\\n if i - 1 >= 0 and j - 1 >= 0 and board.board[i + 3][j + 3] == 0 and board.board[i - 1][j - 1] == 0 \\\\\\n and board.try_move(j + 3) == i + 3 and board.try_move(j - 1) == i - 1:\\n return minvalue\\n elif i + 4 < size_y and j + 4 < size_x and board.board[i + 4][j + 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j + 4) == i + 4 and board.try_move(j) == i:\\n return minvalue\\n # else:\\n # for k in range(num_to_connect):\\n # if board.board[i + k][j + k] == 0 and board.try_move(j + k) == i + k:\\n # points *= -multiply_reachable\\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\\n total_points += points\\n\\n # Fill in the backward diagonal win positions\\n for i in range(size_y - num_to_connect + 1):\\n for j in range(size_x - 1, num_to_connect - 1 - 1, -1):\\n points = 0\\n self_pieces_count = 0\\n opponent_pieces_count = 0\\n for k in range(num_to_connect):\\n if board.board[i + k][j - k] == opponent_id:\\n opponent_pieces_count += 1\\n elif board.board[i + k][j - k] == self.id:\\n points += len(board.winning_zones[j-k][i+k])\\n if (self.id == 1 and (i+k) % 2 == 1) or (self.id == 2 and (i+k)%2 == 0):\\n points *= multiply_oddeven\\n self_pieces_count += 1\\n if self_pieces_count == 3 and self_pieces_count == 0:\\n if board.board[i + 3][j - 3] == 0 and board.board[i - 1][j + 1] == 0 \\\\\\n and board.try_move(j - 3) == i + 3 and board.try_move(j + 1) == i - 1:\\n return maxvalue\\n elif i + 4 < size_y and j - 4 >= 0 and board.board[i + 4][j - 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j - 4) == i + 4 and board.try_move(j) == i:\\n return maxvalue\\n else:\\n for k in range(num_to_connect):\\n if board.board[i + k][j - k] == 0 and board.try_move(j - k) == i + k:\\n points *= multiply_reachable\\n\\n elif opponent_pieces_count == 3 and self_pieces_count == 0:\\n if board.board[i + 3][j - 3] == 0 and board.board[i - 1][j + 1] == 0 \\\\\\n and board.try_move(j - 3) == i + 3 and board.try_move(j + 1) == i - 1:\\n return minvalue\\n elif i + 4 < size_y and j - 4 >= 0 and board.board[i + 4][j - 4] == 0 and board.board[i][j] == 0 \\\\\\n and board.try_move(j - 4) == i + 4 and board.try_move(j) == i:\\n return minvalue\\n # else:\\n # for k in range(num_to_connect):\\n # if board.board[i + k][j - k] == 0 and board.try_move(j - k) == i + k:\\n # points *= -multiply_reachable\\n if (opponent_pieces_count == 3 and self_pieces_count == 0) or opponent_pieces_count == 0:\\n total_points += points\\n return total_points\",\n \"def make_move(self, state, actions):\\n\\n # Here we are doing an exploration so don't update scores\\n # for selected state\\n if random.random() < self._exploration:\\n action = random.choice(actions)\\n self._last_state = self._get_state_for_action(state, action)\\n return action\\n\\n # So we are not doing exploration, here we want to update it\\n max_selection = None\\n for a, s in self._states_from_actions(state, actions).items():\\n state_value = self._state_values[s]\\n if max_selection is None or state_value > self._state_values[max_selection[1]]:\\n max_selection = (a, s)\\n\\n if self._last_state is not None:\\n self._update_state_value(self._last_state, max_selection[1])\\n\\n self._last_state = max_selection[1]\\n return max_selection[0]\",\n \"def DFS(self, nDepth, treenode, state):\\n \\n visited = []\\n visited.insert(0, (state, treenode))\\n \\n for index in range(0, nDepth-1): \\n actions = self.priorProb(state)\\n treenode.expansion(actions)\\n treenode.updateU_value(actions)\\n treenode, action = treenode.selection() \\n state = state.do_move(action).copy()\\n visited.insert(0, (state, treenode)) \\n \\n for index in range(0, len(visited)-1): \\n if(visited[index][1].isLeaf() == True):\\n value = self.leafEvaluation(visited[index][0])\\n else: \\n value = visited[index][1].backUp(value)\\n visited[-1][1].updateQ_value(value)\\n visited[-1][1].updateVisits()\\n return visited[-1][1]\",\n \"def main():\\n\\n N, k = map(int, sys.stdin.readline().split())\\n initial_state = tuple( map(int, sys.stdin.readline().split()) )\\n final_state = tuple( map(int, sys.stdin.readline().split()) )\\n\\n init_path = [initial_state]\\n all_paths = [init_path]\\n\\n visited_states = set()\\n visited_states.add(initial_state)\\n\\n while all_paths:\\n popped_path = all_paths.pop(0)\\n current_state = popped_path[-1]\\n\\n if current_state == final_state:\\n solution = popped_path\\n break\\n\\n for state, action in generate_moves(current_state, N, k).items():\\n if state not in visited_states:\\n visited_states.add(state)\\n new_path = popped_path + [action,state]\\n all_paths.append(new_path)\\n\\n display_solution(solution)\",\n \"def choose_epsilon_greedy(self, state: Tuple[int, ...], valid_actions: Tuple[int, ...]) -> int:\\n if random.random() < self.__epsilon:\\n return self.choose_uniform(valid_actions)\\n return self.choose_greedy(state, valid_actions)\",\n \"def get_every_nth_state(n=10):\\n loops = int(50/n)\\n state_list = []\\n for i in range(loops):\\n state_list.append(states[(i+1)*n-1])\\n return state_list\",\n \"def get_action(self,state):\\n \\n q_values = self.__network.predict(state[None])[0]\\n \\n ###YOUR CODE\\n if np.random.rand()<self.epsilon:\\n return np.random.choice(self.n_actions)\\n return np.argmax(q_values)\",\n \"def get_action(self, state):\\n\\n best_action = None\\n best_value = -np.inf\\n actions = [0, 1, 2, 3] # left, down, right, up\\n for a in actions:\\n row = state // self.edge\\n col = state % self.edge\\n # print (row, col)\\n if a == 0:\\n col = max(col-1, 0)\\n elif a == 1:\\n row = min(row+1, self.edge-1)\\n elif a == 2:\\n col = min(col+1, self.edge-1)\\n elif a == 3:\\n row = max(row-1, 0)\\n # print (row, col)\\n\\n new_state = row * self.edge + col\\n # print (new_state)\\n if (self.values[new_state] > best_value or new_state == self.num_states-1): #goal\\n best_value = 1.0 if new_state == self.num_states-1 else self.values[new_state]\\n best_action = a\\n return best_action\",\n \"def solveOneStep(self):\\n ### Student code goes here\\n # Mark this move as explored\\n self.visited[self.currentState] = True\\n\\n # Get move to make\\n movables = self.gm.getMovables()\\n # print(\\\"EXPLORING GAME STATE \\\" + str(self.gm.getGameState()) + \\\"---------------------------------------------------------\\\")\\n to_move = self.currentState.nextChildToVisit # movables index\\n # print(\\\"depth \\\", self.currentState.depth)\\n\\n # Return if done\\n if self.currentState.state == self.victoryCondition:\\n # print(\\\"DONE\\\")\\n return True\\n\\n while to_move < len(movables):\\n # Make the move\\n movable_statement = movables[to_move]\\n # print(\\\"implementing move \\\", movable_statement)\\n self.gm.makeMove(movable_statement)\\n\\n # Create a new state with this move made\\n new_state = self.gm.getGameState()\\n\\n # Find out if this state has already been explored\\n visited = False\\n for visited_state in self.visited.keys():\\n if visited_state.state == new_state:\\n visited = True\\n\\n # If the new state hasn't been visited then add it as a child then move down to this child\\n if not visited:\\n new_gs = GameState(new_state, self.currentState.depth + 1, movable_statement)\\n new_gs.parent = self.currentState\\n self.currentState.children.append(new_gs)\\n self.currentState.nextChildToVisit = to_move + 1\\n self.currentState = new_gs\\n break\\n\\n # Else skip this state and try going to the next movable statement\\n else:\\n # print(\\\"SKIP THIS STATE\\\")\\n self.gm.reverseMove(movable_statement)\\n to_move += 1\\n\\n # Went all the way down to a leaf, backtrack\\n if (to_move >= len(movables)):\\n self.gm.reverseMove(self.currentState.requiredMovable)\\n self.currentState = self.currentState.parent\\n\\n return False\",\n \"def get_random_depth_sample(n=8, depths=list(range(2,26,2)), num_samples=100):\\n\\n def get_states(start):\\n frontier = [start]\\n frontier_set = {start}\\n explored = set()\\n\\n states = [False for _ in range(len(depths))]\\n while not all(states):\\n node = frontier.pop(0)\\n frontier_set.remove(node)\\n explored.add(node)\\n\\n children = node.get_children()\\n\\n # It's necessary to shuffle children to get a truly random sample; otherwise, the first child (always\\n # produced from the parent by the same action) produced at a certain depth will always be selected,\\n # and children produced by other actions will never be selected\\n shuffle(children)\\n\\n for child in children:\\n if child not in frontier_set and child not in explored:\\n frontier_set.add(child)\\n frontier.append(child)\\n child.path_cost = node.path_cost+1\\n index = depths.index(child.path_cost) if child.path_cost in depths else None\\n if index is not None and not states[index]:\\n states[index] = {'start': start.sequence, 'end': child.sequence}\\n\\n return states\\n\\n depth_sample = [[] for depth in range(len(depths))]\\n\\n for _ in range(num_samples):\\n start = list(range(1,n+2))\\n shuffle(start)\\n start = PuzzleState(start, path_cost=0)\\n\\n states = get_states(start)\\n print('\\\\rSet ' + str(_+1) + ' of ' + str(num_samples) + ' complete', end='', flush=True)\\n list(map(list.append, depth_sample, states))\\n\\n return depth_sample\",\n \"def simulate_tree(self, current_state):\\n sim_trace = [(current_state, None)]\\n while not self.state_manager.game_over():\\n performed_action = self.make_move(current_state, is_rollout=False)\\n current_state, subtree_existed_prev = self.get_child(current_state, performed_action)\\n sim_trace.append((current_state, performed_action))\\n #the case where one has reached a leaf node\\n if subtree_existed_prev == False:\\n break\\n return sim_trace, current_state\",\n \"def run_each_move(root: State, state: GameState) -> State:\\n for m, base_move in enumerate(root.moves):\\n for _ in range(50):\\n base_move = expand(base_move)\\n outcome, _ = MCTS(base_move, state, False, 0)\\n if outcome == state.player:\\n base_move.winner()\\n else:\\n base_move.loser()\\n root.attempts += 1\\n root.moves[m] = base_move\\n print(\\\"Move: \\\", base_move.idx)\\n print(\\\" Wins: \\\", base_move.wins)\\n update_ucbs(root)\\n return root\",\n \"def choose_action(self, d, name):\\n # some initializations\\n current_state = d.state\\n caution = False\\n confidence = False\\n buy_rules = [0,0,0,0]\\n next_vec = self.transition_matrix[name][int(current_state)]\\n num_undesirable_states = (self.trigger_states[name]+1)\\n num_desirable_states = (next_vec.size-num_undesirable_states)\\n if num_undesirable_states<5:\\n left_basket_max = 2\\n else:\\n left_basket_max = num_undesirable_states//3\\n if num_desirable_states<5:\\n right_basket_min = next_vec.size-2\\n else:\\n right_basket_min = next_vec.size-num_undesirable_states//3\\n # check if rules are satisfied\\n # rule-1\\n m1 = np.max(next_vec[:self.trigger_states[name]+1])\\n m1_idx = np.argmax(next_vec[:self.trigger_states[name]+1])\\n m2 = np.max(next_vec[self.trigger_states[name]+1:])\\n m2_idx = np.argmax(next_vec[self.trigger_states[name]+1:])+\\\\\\n next_vec[:self.trigger_states[name]+1].size\\n if m2-m1>=0.1: # threshold\\n #print('Rule #1 satisfied.')\\n buy_rules[0]=1\\n # rule-2\\n if np.sum(next_vec[self.trigger_states[name]+1:])-\\\\\\n np.sum(next_vec[:self.trigger_states[name]+1])>=0.25: # threshold\\n #print('Rule #2 satisfied.')\\n buy_rules[1]=1\\n # rule-3 \\n if m1_idx<left_basket_max: \\n if buy_rules[0]!=1:\\n caution=True\\n #print('Predicted state is very undesirable.')\\n # rule-3\\n if m2_idx>=right_basket_min:\\n if buy_rules[0]==1:\\n confidence=True\\n #print('Predicted state is very desirable.')\\n if d.MACD>d.signal_line:\\n #print('Rule #3 satisfied.')\\n buy_rules[2] = True\\n # sum of k most undesirable vs k most desirable\\n temp_1 = np.sort(next_vec[self.trigger_states[name]+1:])\\n temp_2 = np.sort(next_vec[:self.trigger_states[name]+1])\\n size = 3\\n if temp_1.size<size or temp_2.size<size:\\n size = min(temp_1.size, temp_2.size)\\n k1 = np.sum(temp_1[::-size])\\n k2 = np.sum(temp_2[::-size])\\n if k1-k2>0.25:\\n #print('Rule #4 satisfied.')\\n buy_rules[3] = True\\n # finally, make a call using the rules\\n if confidence or sum(buy_rules)>=3:\\n return 'buy'\\n elif caution or (buy_rules[0]==0 and sum(buy_rules)<=2 and m1-m2>0.05):\\n return 'sell'\\n else:\\n return 'hold'\",\n \"def getAction(self, gameState):\\n \\\"*** YOUR CODE HERE ***\\\"\\n \\n\\n def max_value(gameState,depth):\\n if depth==0 or gameState.isWin() or gameState.isLose():\\n return (self.evaluationFunction(gameState), None)\\n\\n moves=gameState.getLegalActions()\\n score_list = []\\n for i in moves:\\n current_state= self.index\\n next_state = gameState.generateSuccessor(current_state,i) \\n value, action= min_value(next_state,1, depth) \\n score_list.append(value)\\n\\n max_score=max(score_list)\\n score_len= len(score_list)\\n for i in range(score_len):\\n if score_list[i] == max_score:\\n best_state = [i]\\n chosenIndex = random.choice(best_state)\\n return max_score,moves[chosenIndex]\\n\\n def min_value(gameState,agent, depth): \\n if depth==0 or gameState.isWin() or gameState.isLose():\\n return (self.evaluationFunction(gameState), None)\\n \\n moves=gameState.getLegalActions(agent) \\n score_list=[]\\n \\n for i in moves:\\n next_state = gameState.generateSuccessor(agent,i)\\n ghost_no = gameState.getNumAgents()\\n if(agent == ghost_no-1):\\n value, action = max_value(next_state,(depth-1))\\n score_list.append(value)\\n else:\\n \\n value, action = min_value(next_state,agent+1,depth)\\n score_list.append(value)\\n\\n min_score=min(score_list)\\n score_len = len(score_list)\\n for j in range(score_len):\\n if score_list[j] == min_score:\\n worst_state = [j]\\n chosenIndex = random.choice(worst_state)\\n return min_score, moves[chosenIndex]\\n \\n final_score,final_action= max_value(gameState,self.depth)\\n \\n return final_action\\n util.raiseNotDefined()\",\n \"def choose(self, state: State) -> State:\",\n \"def UCT(rootstate, itermax, verbose = False):\\n\\n rootnode = Node(state=rootstate)\\n\\n for i in range(itermax):\\n node = rootnode\\n state = rootstate.Clone()\\n\\n # Select\\n while node.untriedMoves == [] and node.childNodes != []: # node is fully expanded and non-terminal\\n\\n node = node.UCTSelectChild()\\n state.DoMove(node.move)\\n\\n\\n # Expand\\n if node.untriedMoves != []: # if we can expand (i.e. state/node is non-terminal)\\n m = random.choice(node.untriedMoves)\\n state.DoMove(m)\\n node = node.AddChild(m, state) # add child and descend tree\\n\\n # Rollout - this can often be made orders of magnitude quicker using a state.GetRandomMove() function\\n while state.GetMoves() != []: # while state is non-terminal\\n state.DoMove(random.choice(state.GetMoves()))\\n\\n # Backpropagate\\n while node != None: # backpropagate from the expanded node and work back to the root node\\n node.Update(state.GetResult(node.whosemove))\\n node = node.parentNode\\n\\n return sorted(rootnode.childNodes, key=lambda c: c.visits)[-1].move # return the move that was most visited\",\n \"def change(a, b, state, target):\\n # a = random.randrange(1, 199, 2)\\n # b = random.randrange(1, 199, 2)\\n new_state = ''\\n for i in range(len(state)):\\n if i == a:\\n new_state += b\\n continue\\n new_state += state[i]\\n # print(new_state)\\n res = solving(int(new_state[0]), int(new_state[2]), new_state[1])\\n for i in range(2, len(new_state) - 2, 2):\\n res = solving(res, int(new_state[i + 2]), new_state[i + 1])\\n # print(\\\"Distance from target: \\\", target - res)\\n return new_state, abs(target - res)\",\n \"def calc_nearest_state(self): # TODO: Check if we need here state, instead of self.state\\n self.stateC = self.toConceptual(self.state)\\n CTP, winners = self.find_winner()\\n\\n state_name = self.find_TPname(filleridx=winners)\\n binding = self.find_symBinding(filleridx=winners)\\n state_num = self.find_TPnum(stateName=state_name)\\n TP_state = self.TP.matmul(fortran_reshape(CTP, (torch.numel(CTP), 1)))\\n Cdist = torch.norm(CTP - self.stateC, p='fro') # Frobenius Norm\\n Sdist = self.L2norm(TP_state - self.state)\\n TP_h = self.calc_harmony(state=TP_state)\\n\\n return TP_state, winners, state_name, binding, Cdist, Sdist, state_num, TP_h\",\n \"def update_state(state, Y_next):\\n if max(Y_next) > state.best_value + 1e-3 * fabs(state.best_value):\\n state.success_counter += 1\\n state.failure_counter = 0\\n else:\\n state.success_counter = 0\\n state.failure_counter += 1\\n\\n if state.success_counter == state.success_tolerance: # Expand trust region\\n state.length = min(2.0 * state.length, state.length_max)\\n state.success_counter = 0\\n elif state.failure_counter == state.failure_tolerance: # Shrink trust region\\n state.length /= 2.0\\n state.failure_counter = 0\\n\\n state.best_value = max(state.best_value, max(Y_next).item())\\n if state.length < state.length_min:\\n state.restart_triggered = True\\n\\n return state\",\n \"def move_length_5(self, move, new_state):\\n # First consider the move on 6 corners\\n new1_state = new_state\\n if move in [\\\"A\\\", \\\"B\\\", \\\"O\\\", \\\"U\\\", \\\"T\\\", \\\"Y\\\"]:\\n corners = [[\\\"A\\\", 1, 2, 2, 5, 9, 14, 0, 3, 7, 12, 18, 24, 17],\\n [\\\"B\\\", 0, 2, 4, 8, 13, 19, 12, 3, 6, 10, 15, 20, 1],\\n [\\\"O\\\", 20, 13, 9, 5, 2, 0, 0, 15, 16, 17, 18, 19, 10],\\n [\\\"U\\\", 14, 13, 21, 22, 23, 24, 11, 1, 3, 6, 10, 15, 1],\\n [\\\"T\\\", 24, 9, 1, 4, 8, 13, 12, 14, 15, 16, 17, 18, 10],\\n [\\\"Y\\\", 19, 9, 20, 21, 22, 23, 11, 0, 3, 7, 12, 18, 17]]\\n for i in corners:\\n new1_state = self.loop51(move, new_state, i)\\n\\n if move in [\\\"C\\\", \\\"E\\\", \\\"J\\\", \\\"N\\\", \\\"V\\\", \\\"X\\\"]:\\n # then consider the move on the middle1 of each side\\n middle1 = [[\\\"C\\\", 3, 4, 4, 0, 5, 9, 14, 0, 6, 11, 17, 23, 16],\\n [\\\"E\\\", 2, 3, 4, 1, 8, 13, 19, 12, 7, 11, 16, 21, 3],\\n [\\\"J\\\", 15, 21, 14, 10, 11, 12, 13, 8, 0, 2, 5, 14, 0],\\n [\\\"N\\\", 18, 23, 7, 1, 4, 8, 19, 12, 9, 10, 11, 12, 8],\\n [\\\"V\\\", 9, 15, 14, 20, 22, 23, 24, 11, 4, 7, 11, 16, 3],\\n [\\\"X\\\", 13, 18, 7, 20, 21, 22, 24, 11, 2, 6, 11, 17, 16]]\\n for i in middle1:\\n new1_state = self.loop52(move, new_state, i)\\n return StonehengeState(not self.p1_turn, new1_state.length,\\n new1_state.letters, new1_state.claim)\",\n \"def chooseAction(self,state):\\r\\n #generate float btwn 0-1\\r\\n choice = random.random()\\r\\n \\r\\n #choose according to that number\\r\\n if choice > self.epsilon:\\r\\n return(self.maxQ(state)[1])\\r\\n else:\\r\\n #choose randomly\\r\\n return(self.actions[random.randrange(0,len(self.actions))])\",\n \"def chooseAction(self,state):\\r\\n #generate float btwn 0-1\\r\\n choice = random.random()\\r\\n \\r\\n #choose according to that number\\r\\n if choice > self.epsilon:\\r\\n return(self.maxQ(state)[1])\\r\\n else:\\r\\n #choose randomly\\r\\n return(self.actions[random.randrange(0,len(self.actions))])\",\n \"def miniMax(self,state,depth=0):\\n \\n #print(\\\"NextState (depth \\\"+str(depth)+\\\"):\\\")\\n #print(\\\"Action: \\\"+state.get_action())\\n \\n if state in self.__explored:\\n return self.__explored[state.get_hashable_state()]\\n \\n if state.is_end_state() or depth >= (self.__max_depth - 1):\\n self.__explored[state.get_hashable_state()] = state.get_utility_value()\\n return state.get_utility_value() #Return terminal state's utility value\\n \\n is_max_turn = state.get_max_turn()\\n childList = state.get_successors()\\n \\n if is_max_turn:\\n utility = float(\\\"-inf\\\")\\n for c in childList:\\n utility = max(utility,self.miniMax(c, depth+1))\\n self.__explored[state.get_hashable_state()] = utility\\n return utility\\n else:\\n utility = float(\\\"inf\\\")\\n for c in childList:\\n utility = min(utility,self.miniMax(c, depth+1))\\n self.__explored[state.get_hashable_state()] = utility\\n return utility\",\n \"def do(num_range,i, sub):\\n\\t\\n\\t#Converting the vector in all_states_explored to a matrix. \\n\\tmatrix = pMatrix.make_matrix(all_states_explored[i])\\n\\t\\n\\t#Tree for currrent iteration.\\n\\ttree = pMatrix.create_tree(matrix, num_range, sub) \\n\\t\\n\\t#Adding the tree to all_trees.\\n\\tall_trees.append(tree)\\n\\t\\n\\t#Calculating the number of states in current iteration.\\n\\tnum_states = tree.get_num_states()\\n\\t\\t\\n\\t#Adding total number of states to all_total_states. \\n\\tall_total_states.append(num_states) \\n\\t\\n\\t#Finding results for each iteration.\\n\\tr = pMatrix.main(matrix, num_range, sub)\\n\\t\\t\\n\\t#Adding results for current iteration to all_results. \\n\\tall_results.append(r)\\n\\t\\n\\t#Finding any new, previously unseen state and adding it to all_states_explored list.\\n\\tfind_new_states(tree)\\n\\t\\n\\t#Adding super_states\\n\\tupdate_super_states(tree)\",\n \"def __call__(self, state, q_values):\\n\\n if self.policy_type == \\\"greedy\\\":\\n is_greedy = True\\n else:\\n is_greedy = random.uniform(0, 1) > self.epsilon\\n\\n if is_greedy :\\n # choose greedy action\\n index_action = np.argmax(q_values[state])\\n else:\\n # get a random action\\n index_action = random.randint(0,3)\\n\\n return actions_dict[index_action]\",\n \"def __init__(self, mdp, discount = 0.9, iterations = 100):\\n self.mdp = mdp\\n self.discount = discount\\n self.iterations = iterations\\n self.values = util.Counter() # A Counter is a dict with default 0\\n \\n \\\"*** YOUR CODE HERE ***\\\"\\n # OUR CODE HERE\\n #Note: I think we should use the util.Counter thing?\\n for times in range(0, iterations):\\n #values from previous iteration so we don't update over them while iterating\\n prevVals = self.values.copy()\\n #iterate through all states\\n for state in mdp.getStates():\\n #will store the action-value for the iteration\\n value = util.Counter()\\n for action in mdp.getPossibleActions(state):\\n for transitionState, probability in mdp.getTransitionStatesAndProbs(state, action):\\n #expected value, probability * reward for the state with the discount * reward\\n value[action] += probability * (mdp.getReward( state, action, transitionState) + discount * prevVals[transitionState])\\n #update the values to the new value from the iteration\\n #the .argMax() function returns the one with the largest value\\n self.values[state] = value[value.argMax()]\\n # END OUR CODE\",\n \"def __init__(self,\\n game,\\n exploration_strategy='uniform-exhaustive',\\n confidence_method='ucb-standard',\\n delta=0.01,\\n ucb_eps=0,\\n per_payoff_confidence=True,\\n time_dependent_delta=False):\\n self.exploration_strategy = exploration_strategy\\n self.confidence_method = confidence_method\\n self.ucb_eps = ucb_eps\\n self.G = game # pylint: disable=invalid-name\\n self.per_payoff_confidence = per_payoff_confidence\\n self.time_dependent_delta = time_dependent_delta\\n if self.per_payoff_confidence:\\n self._delta = delta\\n else:\\n self._delta = delta / (\\n self.G.n_players *\\n functools.reduce(operator.mul, self.G.strategy_spaces, 1))\\n\\n # Compute the graph\\n self.V = list( # pylint: disable=invalid-name\\n itertools.product(*[range(smax) for smax in self.G.strategy_spaces]))\\n self.E = [] # pylint: disable=invalid-name\\n for v in self.V:\\n adj_strats = [\\n list(range(v[k] + 1, self.G.strategy_spaces[k]))\\n for k in range(self.G.n_players)\\n ]\\n for k in range(self.G.n_players):\\n for new_s in adj_strats[k]:\\n second_vertex = list(v)\\n second_vertex[k] = new_s\\n second_vertex = tuple(second_vertex)\\n self.E.append((v, second_vertex))\\n self.count_history = {v: [] for v in self.V}\\n self.total_interactions = 0\",\n \"def predict(self, state):\\n if self.phase is None or self.device is None:\\n raise AttributeError('Phase, device attributes have to be set!')\\n if self.phase == 'train' and self.epsilon is None:\\n raise AttributeError('Epsilon attribute has to be set in training phase')\\n\\n if self.reach_destination(state):\\n return ActionXY(0, 0) if self.kinematics == 'holonomic' else ActionRot(0, 0)\\n if self.action_space is None:\\n self.build_action_space(state.robot_state.v_pref)\\n\\n probability = np.random.random()\\n if self.phase == 'train' and probability < self.epsilon:\\n max_action = self.action_space[np.random.choice(len(self.action_space))]\\n else:\\n max_action = None\\n max_value = float('-inf')\\n max_traj = None\\n\\n if self.do_action_clip:\\n state_tensor = state.to_tensor(add_batch_size=True, device=self.device)\\n action_space_clipped = self.action_clip(state_tensor, self.action_space, self.planning_width)\\n else:\\n action_space_clipped = self.action_space\\n\\n for action in action_space_clipped:\\n state_tensor = state.to_tensor(add_batch_size=True, device=self.device)\\n next_state = self.state_predictor(state_tensor, action)\\n max_next_return, max_next_traj = self.V_planning(next_state, self.planning_depth, self.planning_width)\\n reward_est = self.estimate_reward(state, action)\\n value = reward_est + self.get_normalized_gamma() * max_next_return\\n if value > max_value:\\n max_value = value\\n max_action = action\\n max_traj = [(state_tensor, action, reward_est)] + max_next_traj\\n if max_action is None:\\n raise ValueError('Value network is not well trained.')\\n\\n if self.phase == 'train':\\n self.last_state = self.transform(state)\\n else:\\n self.traj = max_traj\\n\\n return max_action\",\n \"def N_states_for_learner(self):\\n idx_max = []\\n limits = 50, 2*_math.pi, 50, 50, 50, 50, 50, 50, 50\\n for idx, limit in enumerate(limits):\\n test = [0 for i in xrange(len(limits))]\\n check = _arange(-limit,limit,limit/1000.)\\n maxi = 0\\n for v in check:\\n test[idx]=v\\n ret = self._state_index(*test)\\n maxi = max((maxi, ret[idx]))\\n idx_max.append(maxi)\\n\\n return tuple([idx+1 for idx in idx_max])\",\n \"def get_move(self, state):\\n # this method should only be called when self is real root.,so that's here where we can should use mutiprocess\\n if self._root.is_leaf(): # no expanded children yet\\n action_probs, _ = self._policy(state)\\n self._root.expand(action_probs)\\n\\n the_children = self._root._children\\n i = 0\\n sorted_children = sorted(the_children.items(), key=lambda act_node: act_node[1].get_value(self._c_puct))\\n for child_node in sorted_children:\\n i += 1\\n child_tree = MCTS(policy_value_fn,root=child_node[1])\\n state_copy = copy.deepcopy(state)\\n state_copy.do_move(child_node[0])\\n visits_count = 0\\n for j in range(0,relu(1200-i*20),10): # at least run one time\\n child_tree._playout(copy.deepcopy(state_copy))\\n visits_count += 1\\n self._root.update(-child_tree.get_root_node().last_leafvalue,visits_count=visits_count) # update real root\\n child_tree.get_root_node().set_parent(self._root) # to link the sub tree\\n\\n '''\\n for n in range(self._n_playout):\\n # get top n (assumed to be 6) nodes from children\\n # step1 let all children of root have chance to run in parallel\\n # adjust the round count of children by value\\n if n%6 == 0:\\n the_children = self._root._children\\n top_n = sorted(the_children.items(),key=lambda act_node: act_node[1].get_value(self._c_puct))[:6]\\n for child_node in top_n:\\n # child_tree = MCTS(policy_value_fn,copy.deepcopy(child_node)) # use copy because we will use it in multiprocess\\n child_tree = MCTS(policy_value_fn,\\n child_node) \\n state_copy = copy.deepcopy(state)\\n state_copy.do_move(child_node[0])\\n child_tree._playout(state_copy)\\n self._root.update(-child_tree.get_root_node().last_leafvalue) # update real root\\n child_tree.get_root_node().set_parent(self._root) # to link the sub tree\\n # self._root.get_children()[child_node[0]] = child_tree.get_root_node() # copy sub tree\\n '''\\n\\n '''\\n return max(self._root._children.items(),\\n # key=lambda act_node: act_node[1].get_visits())[0]\\n key=lambda act_node: act_node[1].get_value(self._c_puct))[0]\\n '''\\n\\n for n in range(300):\\n state_copy = copy.deepcopy(state)\\n self._playout(state_copy)\\n return max(self._root._children.items(),\\n key=lambda act_node: act_node[1].get_value(self._c_puct))[0]\",\n \"def solve(self, state, times):\",\n \"def _greedy(self, state):\\n\\n node = self.mcts_head\\n if self.verbose > 1:\\n logger.debug(f\\\"Starting greedy algorithm.\\\")\\n\\n while not node.terminal:\\n # Parse current state\\n this_state, total_reward, terminal = self._parse_path(state, node.path)\\n node.set_terminal(terminal)\\n if self.verbose > 1:\\n logger.debug(f\\\" Analyzing node {node.path}\\\")\\n\\n # Expand\\n if not node.terminal and not node.children:\\n actions = self._find_legal_actions(this_state)\\n step_rewards = [self._parse_action(action, from_which_env=\\\"sim\\\") for action in actions]\\n if self.verbose > 1:\\n logger.debug(f\\\" Expanding: {len(actions)} legal actions\\\")\\n node.expand(actions, step_rewards=step_rewards)\\n\\n # If terminal, backup reward\\n if node.terminal:\\n if self.verbose > 1:\\n logger.debug(f\\\" Node is terminal\\\")\\n if self.verbose > 1:\\n logger.debug(f\\\" Backing up total reward {total_reward}\\\")\\n node.give_reward(self.episode_reward + total_reward, backup=True)\\n\\n # Debugging -- this should not happen\\n if not node.terminal and not node.children:\\n logger.warning(\\n f\\\"Unexpected lack of children! Path: {node.path}, children: {node.children.keys()}, legal actions: {self._find_legal_actions(this_state)}, terminal: {node.terminal}\\\"\\n )\\n node.set_terminal(True)\\n\\n # Greedily select next action\\n if not node.terminal:\\n action = node.select_greedy()\\n node = node.children[action]\\n\\n if self.verbose > 0:\\n choice = self.mcts_head.select_best(mode=\\\"max\\\")\\n self._report_decision(choice, state, \\\"Greedy\\\")\",\n \"def mcts_uct(ei, iterations, approach):\\n clear_states()\\n root = generateState(None, None, ei, approach)\\n for i in xrange(iterations):\\n child = root\\n while not child.terminal() and (not child.experimentInstance.AllTestsReached()):\\n if not child.fully_expanded():\\n child = child.expand()\\n break\\n else:\\n child = child.best_child()\\n cost = child.simulation()\\n update_parents(child, cost)\\n return root.best_action(c=0)\",\n \"def act(self, state):\\n if np.random.rand() <= self.epsilon:\\n return random.randrange(self.n_actions)\\n act_values = self.model.predict(state)\\n return np.argmax(act_values[0])\",\n \"def priorProb(self, state):\\n actions = []\\n for i in range(0, 10):\\n actions.append(((i, i+1), random.uniform(0, 1))) \\n \\n return actions\",\n \"def getTransitionStatesAndProbs(self, state, action = None):\\n # may move ball\\n ball = state[0][:2]\\n if action == None:\\n ballVelocity = state[0][2:]\\n else:\\n ballVelocity = (0, 0)\\n\\n keepers = list(self.getKeepers(state))\\n takers = list(self.getTakers(state))\\n\\n chasers = sorted(keepers, key=lambda keeper: util.getPointVectorDistance(keeper, ball, ballVelocity))\\n # most closest agent, possess the ball, or go to the ball \\n if self.weHaveBall(state):\\n # j has the ball, its transition depends on the action\\n if action[0] == 'hold':\\n pass\\n elif action[0] == 'pass':\\n # pass the ball to a teammate\\n rand = util.randomVector(0.1)\\n target = keepers[action[1]]\\n diff = util.getDirection(keepers[0], (target[0] + rand[0], target[1] + rand[1]))\\n ballVelocity = (self.ballSpeed * diff[0], self.ballSpeed * diff[1])\\n else:\\n raise Exception('Unknown action')\\n else:\\n # j should go to the ball\\n chasers[0] = self.moveTowards(chasers[0], ball)\\n\\n # other agents get open for a pass\\n for i in xrange(1, len(chasers)):\\n # concretely, this agent goes to a least congested place\\n chasers[i] = self.moveTowards(chasers[i], self.getLeastCongestedLoc(state, chasers[i]))\\n keepers = sorted(chasers, key=lambda keeper: util.getDistance(keeper, ball))\\n \\n for i in xrange(2):\\n takers[i] = self.moveTowards(takers[i], ball)\\n for i in xrange(2, len(takers)):\\n takers[i] = self.moveTowards(takers[i], keepers[1])\\n takers = sorted(takers, key=lambda taker: util.getDistance(taker, keepers[0]))\\n \\n newBall = (ball[0] + ballVelocity[0], ball[1] + ballVelocity[1],\\\\\\n ballVelocity[0], ballVelocity[1])\\n newState = [newBall] + keepers + takers\\n return [(tuple(newState), 1)]\",\n \"def predict_next_state_gt(self, states, actions):\\n # TODO: write your code here\\n\\n # return [self.env.get_nxt_state(states[i], actions) for i in range(self.num_particles)]\\n return np.array([[self.env.get_nxt_state(states[j][i], actions[j]) for i in range(self.num_particles)] for j in range(self.popsize)])\",\n \"def solveOneStep(self):\\n ### Student code goes here\\n if self.first_step == False:\\n self.first_step = True\\n if self.solveOneStep():\\n return True\\n if self.queue:\\n self.gm_init()\\n ele = self.queue.get()\\n #print (len(ele))\\n state = ele[0]\\n premoves = ele[1]\\n\\n for m in premoves:\\n self.gm.makeMove(m)\\n if state.state == self.victoryCondition:\\n return True\\n self.visited[state] = True\\n print(\\\"CURRENTSTATE:\\\")\\n print(self.gm.getGameState())\\n print(\\\"*******\\\")\\n moves = self.gm.getMovables()\\n for m in moves:\\n self.gm.makeMove(m)\\n if (((state.parent is not None) and (self.gm.getGameState() == state.parent.state))) or GameState(self.gm.getGameState(), 0, None) in self.visited:\\n self.gm.reverseMove(m)\\n continue\\n self.visited[GameState(self.gm.getGameState(), 0, None)] = True\\n new_pmv = [i for i in premoves]\\n new_pmv.append(m)\\n next_state = GameState(self.gm.getGameState(), state.depth+1, m)\\n next_state.parent = state\\n state.children.append(next_state)\\n self.queue.put([next_state, new_pmv])\\n self.gm.reverseMove(m)\\n self.currentState = state\\n\\n #for i in range(len(premoves)-1, -1, -1):\\n # mv = premoves[i]\\n # self.gm.reverseMove(mv)\\n return False\",\n \"def _playout(self, state):\\n node = self._root\\n while (1):\\n if node.is_leaf():\\n break\\n # Greedily select next move.\\n action, node = node.select(self._c_puct)\\n state.do_move(action)\\n\\n # Evaluate the leaf using a network which outputs a list of\\n # (action, probability) tuples p and also a score v in [-1, 1]\\n # for the current player.\\n action_probs, leaf_value = self._policy(state)\\n # Check for end of game.\\n end, winner = state.game_end()\\n if not end:\\n node.expand(action_probs)\\n else:\\n # for end state，return the \\\"true\\\" leaf_value\\n if winner == -1: # tie\\n leaf_value = 0.0\\n else:\\n leaf_value = (\\n 1.0 if winner == state.get_current_player() else -1.0\\n )\\n\\n # Update value and visit count of nodes in this traversal.\\n node.update_recursive(-leaf_value)\",\n \"def getAction(self, gameState):\\n \\\"*** YOUR CODE HERE ***\\\"\\n # Again, we use the fundamental foundation built in Q2 for Q4, however here we modify our minimizer function\\n # to serve the purpose of finding the expected value\\n actionList = gameState.getLegalActions(0)\\n pacmanAgentIndex = 0\\n ghostAgentIndices = list(range(1,gameState.getNumAgents())) # List of each agent index for looping\\n count = util.Counter()\\n agentEnd = gameState.getNumAgents()-1 # Last agent in the list\\n def maximizer(curState, agentIndex, depth):\\n\\n ghostActions = curState.getLegalActions(agentIndex)\\n maxDepth = self.depth # Quantifying the end of the tree so we know when we reached a leaf node\\n weight = -99999999 # Worst case starting value to be changed in the code\\n if depth == maxDepth: # If we are at a leaf node\\n return self.evaluationFunction(curState) # evaluate the state of this leaf node\\n # Otherwise, we progress the tree until the above condition is reached\\n if len(ghostActions) != 0:\\n for x in ghostActions:\\n if weight >= minimizer(curState.generateSuccessor(agentIndex, x), agentIndex + 1, depth):\\n weight = weight\\n else:\\n weight = minimizer(curState.generateSuccessor(agentIndex, x), agentIndex + 1, depth)\\n return weight\\n else:\\n # if there are no legal actions left then evaluate at the last known state\\n return self.evaluationFunction(curState)\\n\\n def minimizer(curState, agentIndex, depth):\\n ghostActions = curState.getLegalActions(agentIndex)\\n weight = 0 # Starting value of zero to be incremented below\\n if len(ghostActions) != 0:\\n if agentIndex == agentEnd: # If we've reached the last ghost, we maximise\\n for x in ghostActions: # For each legal action in the current position\\n temp = (float(1.0) / len(ghostActions))*maximizer(curState.generateSuccessor(agentIndex, x), pacmanAgentIndex, depth+1)\\n weight = weight + temp\\n else: # Otherwise, we continue to minimize\\n for x in ghostActions: # For each legal action in the current position\\n temp = (float(1.0) / len(ghostActions))*minimizer(curState.generateSuccessor(agentIndex, x), agentIndex+1, depth)\\n weight = weight + temp\\n return weight\\n else:\\n # if there are no legal actions left then evaluate at the last known state\\n return self.evaluationFunction(curState)\\n\\n # Executing the minimizer for all possible actions\\n for x in actionList:\\n tempState = gameState.generateSuccessor(pacmanAgentIndex,x)\\n count[x] = minimizer(tempState,1,0)\\n # print('HELLO THERE')\\n # print(count)\\n return count.argMax()\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.69618815","0.65259063","0.6470664","0.6314135","0.62374127","0.6210219","0.61834335","0.6150433","0.6120203","0.6094854","0.6084462","0.60547906","0.6046755","0.6046614","0.6031796","0.5984574","0.5967185","0.59669054","0.595718","0.595556","0.5952896","0.59379166","0.5914684","0.59108835","0.59044576","0.5884407","0.5880219","0.58754003","0.5874352","0.58675843","0.58646303","0.58545166","0.58303577","0.5811021","0.57970166","0.57896155","0.5783645","0.57541436","0.57452583","0.57360363","0.5734928","0.57326055","0.57319903","0.571486","0.5713003","0.5711298","0.5705586","0.56738836","0.5668272","0.56658745","0.5665105","0.5663275","0.5660759","0.5656775","0.56555384","0.5648616","0.5646942","0.56425714","0.5640016","0.5624263","0.5621697","0.56061006","0.558721","0.55807865","0.55756724","0.5574964","0.55733","0.5569893","0.55618596","0.55615056","0.5547527","0.5546858","0.5546807","0.55419356","0.55386716","0.55238837","0.5522879","0.552119","0.5518739","0.5513169","0.5508691","0.5508691","0.5502739","0.54981494","0.5497056","0.54783165","0.5477994","0.5477481","0.5475727","0.5473579","0.5466497","0.54662436","0.54655206","0.54564023","0.5455785","0.5454358","0.5454237","0.54533947","0.54387367","0.5438344"],"string":"[\n \"0.69618815\",\n \"0.65259063\",\n \"0.6470664\",\n \"0.6314135\",\n \"0.62374127\",\n \"0.6210219\",\n \"0.61834335\",\n \"0.6150433\",\n \"0.6120203\",\n \"0.6094854\",\n \"0.6084462\",\n \"0.60547906\",\n \"0.6046755\",\n \"0.6046614\",\n \"0.6031796\",\n \"0.5984574\",\n \"0.5967185\",\n \"0.59669054\",\n \"0.595718\",\n \"0.595556\",\n \"0.5952896\",\n \"0.59379166\",\n \"0.5914684\",\n \"0.59108835\",\n \"0.59044576\",\n \"0.5884407\",\n \"0.5880219\",\n \"0.58754003\",\n \"0.5874352\",\n \"0.58675843\",\n \"0.58646303\",\n \"0.58545166\",\n \"0.58303577\",\n \"0.5811021\",\n \"0.57970166\",\n \"0.57896155\",\n \"0.5783645\",\n \"0.57541436\",\n \"0.57452583\",\n \"0.57360363\",\n \"0.5734928\",\n \"0.57326055\",\n \"0.57319903\",\n \"0.571486\",\n \"0.5713003\",\n \"0.5711298\",\n \"0.5705586\",\n \"0.56738836\",\n \"0.5668272\",\n \"0.56658745\",\n \"0.5665105\",\n \"0.5663275\",\n \"0.5660759\",\n \"0.5656775\",\n \"0.56555384\",\n \"0.5648616\",\n \"0.5646942\",\n \"0.56425714\",\n \"0.5640016\",\n \"0.5624263\",\n \"0.5621697\",\n \"0.56061006\",\n \"0.558721\",\n \"0.55807865\",\n \"0.55756724\",\n \"0.5574964\",\n \"0.55733\",\n \"0.5569893\",\n \"0.55618596\",\n \"0.55615056\",\n \"0.5547527\",\n \"0.5546858\",\n \"0.5546807\",\n \"0.55419356\",\n \"0.55386716\",\n \"0.55238837\",\n \"0.5522879\",\n \"0.552119\",\n \"0.5518739\",\n \"0.5513169\",\n \"0.5508691\",\n \"0.5508691\",\n \"0.5502739\",\n \"0.54981494\",\n \"0.5497056\",\n \"0.54783165\",\n \"0.5477994\",\n \"0.5477481\",\n \"0.5475727\",\n \"0.5473579\",\n \"0.5466497\",\n \"0.54662436\",\n \"0.54655206\",\n \"0.54564023\",\n \"0.5455785\",\n \"0.5454358\",\n \"0.5454237\",\n \"0.54533947\",\n \"0.54387367\",\n \"0.5438344\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6509543"},"document_rank":{"kind":"string","value":"2"}}},{"rowIdx":90884,"cells":{"query":{"kind":"string","value":"Return the last bytes of a file containing utf8 formatted text rather than the whole contents. The last bytes is determined by size_in_bytes"},"document":{"kind":"string","value":"def read_last_bytes_of_file(file_path: str, size_in_bytes: int) -> str:\n with open(file_path, \"rb\") as file:\n # Note the minus sign\n file.seek(-size_in_bytes, os.SEEK_END)\n content = file.read().decode(\"utf-8\")\n return content"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def text_file_last_line(f, encoding=\"utf-8\"):\n f.seek(0, os.SEEK_END) # start at the second to last character, since text files should end in a newline\n if f.tell() < 2: return \"\"\n f.seek(-2, os.SEEK_END) # start at the second to last character, since text files should end in a newline\n while f.read(1) != b\"\\n\":\n if f.tell() < 2: # arrived at beginning of file, so there is only one line\n f.seek(0, os.SEEK_SET) # only one line; move to beginning so we read the whole line\n break\n f.seek(-2, os.SEEK_CUR) # go backward from the end of the file until we find a full line\n return f.readline().decode(encoding)","def getLastLine(f, blockSize=3072): \n \n f.seek(0,os.SEEK_END) \n \n totalBytes = f.tell() \n \n if totalBytes > blockSize: \n f.seek(-blockSize,os.SEEK_END) \n else: \n f.seek(0) \n \n lastLines = f.readlines() \n lastLine = lastLines[-1] \n \n if lastLine =='': \n lastLine = lastLines[-2] \n \n return lastLine","def injected_file_content_bytes(self):\n return int(self.get('injected_file_content_bytes'))","def last_chars(fh):\n output = ''\n with open(fh.name) as fh:\n for line in fh:\n output += line[-2]\n return output","def get_unknown_bytes(filename):\n d_file = open(os.path.join(u_path, filename), \"rb\")\n b = bytearray(d_file.read())\n d_file.close()\n return b","def get_file_content(full_path):\n print(full_path)\n bytes = min(32, os.path.getsize(full_path))\n raw = open(full_path, 'rb').read(bytes)\n if '\\\\xff\\\\xfe' in str(raw):\n print(\"utf-16\")\n the_file = open(full_path, encoding=\"utf-16\",\n errors=\"backslashreplace\")\n data = the_file.read()\n else:\n print(\"latin-1\")\n the_file = open(full_path, encoding=\"latin-1\",\n errors=\"backslashreplace\")\n data = the_file.read()\n return data","def read_file_bytes(read_file_path):\n file_string = \"\"\n with open(read_file_path, \"r\") as f:\n for line in f:\n file_string += line\n return bytes(file_string, \"UTF-8\")","def get_file_size(file):\n\tfile.seek(0, os.SEEK_END)\n\treturn file.tell()","def get_file_encoding_bomsize(filename):\n bom_encodings = ((codecs.BOM_UTF8, 'utf-8-sig'),\n (codecs.BOM_UTF16_LE, 'utf-16le'),\n (codecs.BOM_UTF16_BE, 'utf-16be'),\n (codecs.BOM_UTF32_LE, 'utf-32be'),\n (codecs.BOM_UTF32_BE, 'utf-32be'))\n\n firstbytes = open(filename, 'rb').read(4)\n for bom, encoding in bom_encodings:\n if firstbytes.startswith(bom):\n file_encoding, size = encoding, len(bom)\n break\n else:\n file_encoding, size = \"utf-8\", 0\n\n return file_encoding, size","def max_offset(self):\n return self.offset + self.filesize - 1","def chars(count):\n\n global offset\n\n bytes=midifile[offset:offset+count]\n offset+=count\n return bytes","def read(self, size: int = -1) -> bytes:\n if self.size_read >= self.chunksize:\n return b''\n if size < 0:\n size = self.chunksize - self.size_read\n if size > self.chunksize - self.size_read:\n size = self.chunksize - self.size_read\n data = self.file.read(size)\n self.size_read = self.size_read + len(data)\n if self.size_read == self.chunksize and (self.chunksize & 1):\n dummy = self.file.read(1)\n self.size_read = self.size_read + len(dummy)\n return data","def get_file_size(fname, size_length):\n size = os.path.getsize(fname)\n return hex_encode(size, size_length)","def get_file_size(file_path):\n with open(file_path, 'rb') as infile:\n infile.seek(0, 2)\n return infile.tell()","def get_char(self):\n for i in range(1,5):\n c = self.file[self.index:self.index+i] \n try:\n u = c.decode(self.encoding)\n break\n except UnicodeDecodeError:\n if i==4:\n raise \n\n self.index+=i\n\n if self.index > len(self.file):\n self.eof=True\n return None\n\n if u =='\\n':\n self.line +=1\n self.column = 0\n else:\n self.column+=1\n\n return u","def get_bytes(self):\n return self.bytes[:-len(ImageStream.END_SEQUENCE)]","def get_file_size(file: Path) -> str:\n return (\n humanize.naturalsize(file.stat().st_size) if file.stat().st_size else \"unknown\"\n )","def num_long_term_bytes(self) -> str:\n return pulumi.get(self, \"num_long_term_bytes\")","def get_file_contents(file: str) -> tuple[bytes, str]:\n\n with open(file, \"rb\") as fp:\n data = fp.read()\n\n if data[:4] == b\"\\x00\\x00\\xfe\\xff\":\n return data, data.decode(\"utf-32be\")\n elif data[:4] == b\"\\xff\\xfe\\x00\\x00\":\n return data, data.decode(\"utf-32le\")\n elif data[:4] == b\"\\x00\\x00\\x00\\x3c\":\n return data, data.decode(\"utf-32be\")\n elif data[:4] == b\"\\x3c\\x00\\x00\\x00\":\n return data, data.decode(\"utf-32le\")\n elif data[:4] == b\"\\x00\\x3c\\x00\\x3f\":\n return data, data.decode(\"utf-16be\")\n elif data[:4] == b\"\\x3c\\x00\\x3f\\x00\":\n return data, data.decode(\"utf-16le\")\n elif (data[:2] == b\"\\xfe\\xff\") and (data[2:4] != b\"\\x00\\x00\"):\n return data, data[2:].decode(\"utf-16be\")\n elif (data[:2] == b\"\\xff\\xfe\") and (data[2:4] != b\"\\x00\\x00\"):\n return data, data[2:].decode(\"utf-16le\")\n elif data[:3] == b\"\\xef\\xbb\\xbf\":\n return data, data[3:].decode(\"utf-8\", errors=\"ignore\")\n\n # No byte order mark was found. Assume UTF-8, and ignore errors.\n return data, data.decode(\"utf-8\", errors=\"ignore\")","def reads(self, n):\n return self.file.read(n).decode('iso-8859-1')","def read_file_bytes(filepath: str):\n with open(filepath, \"rb\") as reader:\n return reader.read()","def get_data_block_contents_bytes(self):\n bb = self.volume.blkdev.block_bytes\n if self.volume.is_ffs:\n return bb\n else:\n return bb - 24","def get_content(self):\r\n try:\r\n with open(self.path, \"rb\") as f:\r\n raw_data = f.read()\r\n result = chardet.detect(raw_data)\r\n encoding = result['encoding']\r\n if encoding:\r\n with open(self.path, 'r', encoding=encoding) as f:\r\n content = f.readlines()\r\n else:\r\n try:\r\n with open(self.path, 'r', encoding='ansi') as f:\r\n content = f.readlines()\r\n except UnicodeDecodeError:\r\n print('Jakis blad kodowania pliku')\r\n return -1\r\n return content\r\n except FileNotFoundError:\r\n print(f'Wskazany plik {self.path} nie istnieje')\r\n exit()","def UTF8Size(utf):\n ret = libxml2mod.xmlUTF8Size(utf)\n return ret","def get_filesize(inputfile) -> int:\n with open(inputfile, \"rb\") as f:\n lines = 0\n buf_size = 1024 * 1024\n read_f = f.raw.read\n\n buf = read_f(buf_size)\n while buf:\n lines += buf.count(b\"\\n\")\n buf = read_f(buf_size)\n\n return lines","async def collect_file(self, file):\n with open(file, \"wb\") as f: # Opening file as writable in bytes\n\n total_bytes = 0\n while True:\n self.reader._eof = False # Force to read\n data = await self.reader.read(1024)\n\n # Get buffer using BytesIO\n chunk = io.BytesIO(data)\n\n total_bytes += chunk.getbuffer().nbytes\n\n # last_four = data[:-4].decode(\"utf-8\")\n # last_four = chunk.getvalue()[-1:]\n last_four = chunk.getvalue()\n # print((last_four))\n\n check = \"end\"\n\n # print(check.encode())\n\n if last_four == check.encode():\n print(\"Not Data\")\n break\n\n f.write(chunk.getvalue())\n print(f\"Collected: {total_bytes!r} bytes\")","def get_file_size(input_file):\n old_file_position = input_file.tell()\n input_file.seek(0, os.SEEK_END)\n size = input_file.tell()\n input_file.seek(old_file_position, os.SEEK_SET)\n return size","def _extract_bytes(l):\n # input should consist of a list of strings similar to ['123456', 'bytes', 'read']\n if len(l) != 3 or l[1] != 'bytes' or l[2] != 'read':\n raise UBootError('received unexpected reply when reading file')\n\n try:\n return int(l[0])\n except ValueError:\n raise UBootError('received unexpected bytes value when reading file')","def get_most_common_non_ascii_char(file_path: str) -> str:\n with open(file_path, encoding=\"unicode-escape\") as f:\n text = f.read()\n chars = set(text) - set(\n string.punctuation + string.ascii_letters + string.digits + \"\\n \"\n )\n char_dict = {char: text.count(char) for char in chars}\n return sorted(char_dict.items(), key=lambda x: x[1])[-1][0]","def content_length(self):\n if isinstance(self.raw_content, bytes):\n return len(self.raw_content)\n if isinstance(self.raw_content, str):\n return len(self.raw_content.encode('utf-8'))\n pos = self.raw_content.tell()\n try:\n self.raw_content.seek(0, os.SEEK_END)\n return self.raw_content.tell()\n finally:\n self.raw_content.seek(pos)","def read_bytes(self, size):\n return self.read('bytes:'+str(size))","def get_data(filename=filename, maxlen=None):\n print('reading {} with maxlen={}'.format(filename, maxlen))\n if filename.startswith('s3'):\n client = boto3.client('s3')\n obj = client.get_object(Bucket='misc-data-ml-cottrell', Key='mangled_data.txt.gz')\n fin = gzip.open(obj['Body'])\n else:\n fin = gzip.open(filename)\n data = fin.read().decode('utf-8')\n print('data len is {}'.format(len(data)))\n data = data[:(maxlen+1)]\n print('using data len {}'.format(len(data)))\n chars = sorted(list(set(data))) # is int if data is bytes, is str if data is str\n VOCAB_SIZE = len(chars)\n return data, chars, VOCAB_SIZE","def read_utf8_string(self, length):\n return self.read(length).decode(\"utf-8\")","def readBytes(self, size=1):\n return self.bytes","def get_file_size_in_bytes(file_path):\n size = os.path.getsize(file_path)\n return size","def _read_until(self, c, chunk_size=96):\n s = io.BytesIO()\n fp = self._fp\n eof = False\n\n while True:\n chunk = fp.read(chunk_size)\n\n if not chunk:\n # The end of the file was reached. We'll bail out of the loop\n # and return everything we've read so far.\n eof = True\n break\n\n i = chunk.find(c)\n\n if i == -1:\n # We didn't find the character. Store the entire chunk.\n s.write(chunk)\n else:\n # We found the character. Store everything up to and including\n # it, and then go back in the stream for the next read.\n s.write(chunk[:i + 1])\n fp.seek(i + 1 - len(chunk), os.SEEK_CUR)\n break\n\n result = s.getvalue()\n s.close()\n\n return result, eof","def test_utf8_bytes_at_end(self):\n # Python3 doesn't support bytestrings, don't run this test\n if str is unicode:\n return\n input = \"A r\\xc3\\xa9sum\\xc3\\xa9, also spelled resum\\xc3\\xa9 or resume\"\n output = input.split(\" \")\n output[1] = output[1][0:-1]\n for (itmO, itmV) in zip(output, tokenize_en(input)):\n self.assertEqual(itmO, itmV[0])","def unpack_utf8(self):\n num_bytes_consumed, s = decode_utf8(self.view)\n\n self.num_bytes_consumed += num_bytes_consumed\n self.view = self.view[num_bytes_consumed:]\n\n return num_bytes_consumed, s","async def readinto(self, b):\n # get the size of the file\n size = await self._getsize()\n b[:size] = await self.read(size)\n return size","def get_position(self):\n\n return (self._fileobj.tell() - self._pos) * 8 - self._bits","def get_file_content(filename):\n fp = open(filename, \"r\")\n file_content = fp.read()\n fp.close()\n\n # if the last character is an end of line, remove it\n if file_content[-1] == '\\n':\n file_content = file_content[:-1]\n return file_content","def get_file_encoding(file_path):\n with open(file_path, 'r', encoding='utf-8') as f:\n try:\n for line in f:\n pass\n except UnicodeDecodeError:\n return 'ISO-8859-1'\n else:\n return 'utf-8'","def _decode_end(_fp):\n return 0","def download_bytes(self) -> str:\n return pulumi.get(self, \"download_bytes\")","def read_text_file(fpath, encoding, read_size=-1, force_unix_linebreaks=True):\n with codecs.open(fpath, encoding=encoding) as f:\n contents = f.read(read_size)\n\n if read_size > 0:\n contents = contents[:read_size]\n\n if force_unix_linebreaks:\n contents = linebreaks_win2unix(contents)\n\n return contents","def _getsize(path, transport_params):\n with smart_open.open(path, 'rb', ignore_ext=True, transport_params=transport_params) as fin:\n fin.seek(0, io.SEEK_END)\n return fin.tell()","def _read_bytes(self, start, num_bytes):\n with self._fp_lock:\n self._fp.seek(start)\n return self._fp.read(num_bytes)","def byte_truncate(text):\n\ts = text.encode(ENCODING)\n\tif len(s) <= NAME_MAX_BYTES:\n\t\treturn s\n\n\tif ENCODING == \"utf-8\":\n\t\tlead_byte = utf8_lead_byte\n\telse:\n\t\traise NotImplementedError()\n\n\ti = NAME_MAX_BYTES\n\twhile i > 0 and not lead_byte(s[i]):\n\t\ti -= 1\n\treturn s[:i]","def get_size(self, fileobject):\n # move the cursor to the end of the file\n fileobject.seek(0, 2)\n size = fileobject.tell()\n # move the cursor to the begin of the file\n fileobject.seek(0)\n return size","def get_last_uploaded_byte(self, check_uri):\n upload_check_headers = {'Content-Range': 'bytes */*'}\n request_headers = dict(upload_check_headers.items() + self.standard_headers.items())\n try:\n HTTPClient().fetch(check_uri, method=\"PUT\", body='', headers=request_headers)\n except HTTPError as e:\n log.info(\"Upload check: status code %s\" % e.code)\n if e.code == 308:\n _range = int(e.response.headers['Range'].split('-')[1])\n log.info(\"Last uploaded byte: %d\" % _range)\n return _range\n else: raise\n raise ValueError(\"Upload check unsuccessful\")","def read_gzip_bytes(self):\n with gzip.open(self, 'rb') as f:\n return f.read()","def read(self, size=None):\n if not size:\n rlen = self.end - self.pos\n else:\n rlen = size\n if rlen == 0:\n return '' # Fake EOF\n # Clamp rlen to window end\n if self.pos + rlen > self.end:\n rlen = self.end - self.pos\n data = self.file.read(rlen)\n self.pos += len(data)\n return data","def read_txt(txtfile):\n with open(txtfile, \"r\", encoding=\"utf8\") as infile: \n text = infile.read()\n #print(text[0:100])\n return text","def read_bytes(self) -> bytes:\n t = self.pc\n while self.data[self.pc] != 0:\n self.pc += 1\n result = self.data[t:self.pc]\n self.pc += 1 # jump '\\0'\n return result","def get_encoding_file(fname):\n with io.open(fname, 'rt', encoding='ascii', errors='ignore') as f:\n for unused in range(2):\n line = f.readline()\n match = get_encoding_re.search(line)\n if match:\n return match.group(1)\n return 'ascii'","def ra_unicode_read(fh, start, end):\n tmp_pos = fh.tell()\n fh.seek(0)\n\n buf = u\"\"\n iter_fh = (line.decode('utf-8') for line in fh)\n\n for line in iter_fh:\n line_len = len(line)\n # .. recalibrate start/end ...\n if line_len < start:\n start = start - line_len\n if end is not None:\n end = end - line_len\n else:\n # ...until start is met\n buf = buf + line\n\n # cutoff start\n buf = buf[start:]\n if end is not None:\n end = end - start\n\n break\n\n if end is not None:\n # read enough lines so the end is met\n while len(buf) < end:\n try:\n buf = buf + iter_fh.next()\n except StopIteration:\n # end-of-file: end will not be met\n break\n buf = buf[:end]\n\n fh.seek(tmp_pos) # restore the file position\n\n return buf","def get_size(fileobj):\n\n old_pos = fileobj.tell()\n try:\n fileobj.seek(0, 2)\n return fileobj.tell()\n finally:\n fileobj.seek(old_pos, 0)","def _filelength(self):\r\n with open(self.fileName, 'rb') as f:\r\n f.seek(0, 2) # move to end of file\r\n length = f.tell() # get current position\r\n return length","def get_uncompressed_size(self):\n return self.__uncompressed_size","def read_block(file, block_size):\n block = b\"\"\n for i in range(block_size):\n this_byte = file.read(1)\n # If the last block consumed the last char in file:\n if this_byte == b\"\" and i == 0:\n return (-1, False)\n # If we reach EOF prematurely:\n elif this_byte == b\"\":\n block += chr(0).encode()*(block_size - i)\n return (block, False)\n else:\n block += this_byte\n return (block, True)","def test_file_bin_read_unicode_as_bin(self):\n FileWriter(self.unicode_path).write_utf8(self.unicode_string)\n bin_data = FileReader(self.unicode_path).read_bin() #read unicode file as binary\n uni_text = bin_data.decode(\"utf-8\") #decode to utf-8\n self.assertEqual(uni_text, self.unicode_string)","def GetEncoding(self): \n return self.file.GetEncoding()","def FileLen(filename):\n return os.stat(str(filename))[6]","def tailFile(filename, offset, length):\n\n overflow = False\n try:\n f = open(filename, 'rb')\n f.seek(0, 2)\n sz = f.tell()\n\n if sz > (offset + length):\n overflow = True\n offset = sz - 1\n\n if (offset + length) > sz:\n if offset > (sz - 1):\n length = 0\n offset = sz - length\n\n if offset < 0:\n offset = 0\n if length < 0:\n length = 0\n\n if length == 0:\n data = ''\n else:\n f.seek(offset)\n data = f.read(length)\n\n offset = sz\n return [data, offset, overflow]\n\n except (OSError, IOError):\n return ['', offset, False]","def find_file_end(chunks, pos):\n\n\tpos = pos + 1\n\twhile pos < len(chunks)-1:\n\n\t\tif chunks[pos][0] != 0x100 and chunks[pos][0] != 0x102:\n\n\t\t\t# This is not a block\n\t\t\treturn pos\n\n\t\telse:\n\t\t\tpos = pos + 1\n\n\treturn pos","def biggestLine(file_name, start, end):\r\n\r\n with open(file_name, \"r\") as file:\r\n bg_line = 0\r\n\r\n bg_line = len(file.readline())\r\n\r\n for line in islice(file, start, end):\r\n\r\n if bg_line < len(line):\r\n bg_line = len(line)\r\n\r\n return bg_line","def read(self, size: int=-1) -> bytes:\n ...","def read(self, size: int=-1) -> bytes:\n ...","def file_size(file_path):\n if os.path.isfile(file_path):\n file_info = os.stat(file_path)\n return convert_bytes(file_info.st_size)","def file_size(file_path):\n if os.path.isfile(file_path):\n file_info = os.stat(file_path)\n return convert_bytes(file_info.st_size)","def file_size(file_path):\n if os.path.isfile(file_path):\n file_info = os.stat(file_path)\n return convert_bytes(file_info.st_size)","def get_string_length(self):\n return int(self.read('H')[0])","def get_file_chunk(self):\n chunk = self.infile.read(self.CHUNK_SIZE)\n if len(chunk) == 0:\n return None\n return chunk","def file_size():\n return os.path.getsize(FILE_NAME)","def count_deleted_bytes(self): # FileObj.count_deleted_bytes\n if self.deleted:\n return self.bytes \n else:\n return 0","def read(num_bytes):\n # If file cannot be mmap'd (e.g. is stdin, or a fifo), fall back\n # to doing an actual read from the file.\n if not can_be_mmapd(fd):\n return fd.read(num_bytes)\n\n bytes_available = max(file_size - offset, 0)\n if bytes_available == 0:\n return b\"\"\n\n return mmap.mmap(fd.fileno(), min(num_bytes, bytes_available), offset=offset, access=mmap.ACCESS_READ)","def _get_correct_string_size(self, fd):\n # store the lenght value bits\n size_bits = []\n # the value can have max 5 bytes\n for _ in range(5):\n size = self.read_bytes_to_int(fd, 1)\n # if the length value is less then 128 it means its the last\n # byte with the lenght value information and we can end.\n if size < 128:\n size_bits.append(bin(size)[2:])\n break\n else:\n # when its bigger than 128 it means that the length value\n # is using more bytes. We strip also the most significant\n # bit\n size_bits.append(bin(size)[3:])\n\n size_bits.reverse()\n correct_size = int(\"\".join(size_bits), 2) \n return correct_size","def tailFile(filename, offset, length):\r\n\r\n overflow = False\r\n f = None\r\n try:\r\n f = open(filename, 'rb')\r\n f.seek(0, 2)\r\n sz = f.tell()\r\n\r\n if sz > (offset + length):\r\n overflow = True\r\n offset = sz - 1\r\n\r\n if (offset + length) > sz:\r\n if offset > (sz - 1):\r\n length = 0\r\n offset = sz - length\r\n\r\n if offset < 0: offset = 0\r\n if length < 0: length = 0\r\n\r\n if length == 0:\r\n data = ''\r\n else:\r\n f.seek(offset)\r\n data = f.read(length)\r\n\r\n offset = sz\r\n return [as_string(data), offset, overflow]\r\n\r\n except (OSError, IOError):\r\n return ['', offset, False]\r\n finally:\r\n if f:\r\n f.close()","def getFileContent(fileName, encoding = \"UTF-8\"):\n file = io.open(fileName, mode = \"r\", encoding = encoding)\n text = file.read()\n file.close()\n return text","def last_line(filename):\n if not os.path.exists(filename):\n return None\n\n try:\n with open(filename, 'rb') as f:\n f.seek(-2, os.SEEK_END)\n while f.read(1) != b\"\\n\":\n f.seek(-2, os.SEEK_CUR)\n\n last = f.readline()\n return last\n except OSError:\n return None","def UTF8Strsize(utf, len):\n ret = libxml2mod.xmlUTF8Strsize(utf, len)\n return ret","def get_cur_file_size(fp, position_to_eof=False):\r\n if not position_to_eof:\r\n cur_pos = fp.tell()\r\n fp.seek(0, os.SEEK_END)\r\n cur_file_size = fp.tell()\r\n if not position_to_eof:\r\n fp.seek(cur_pos, os.SEEK_SET)\r\n return cur_file_size","def test_file_utf8_readwrite(self):\n FileWriter(self.unicode_path).write_utf8(self.unicode_string)\n unicode_text = FileReader(self.unicode_path).read_utf8()\n self.assertEqual(unicode_text, self.unicode_string)","def get_uncompressed_size(self):\n if not self.is_compressed():\n return self.size\n elif self.data is None:\n raise ValueError('Data not read from record')\n else:\n return unpack('<L', self.data[:4])[0]","def GetBytes(byte, size):\n if sys.version_info[0] >= 3:\n data = bytes([byte]) * size\n else:\n data = chr(byte) * size\n return data","def my_size(self):\n return u\"{} bytes\".format(self.size)","def read_full_file(filename, options=\"rb+\"):\n with open(filename, options) as f:\n text = f.read()\n return text","def __read(self, i: int) -> bytes:\r\n b = self.data[self.idx: self.idx + i]\r\n self.idx += i\r\n if len(b) != i:\r\n raise DecodingError(\r\n \"Incorrect byte length returned between indexes of {0} and {1}. Possible unexpected End of File.\"\r\n .format(str(self.idx), str(self.idx - i)))\r\n return b","def getTRRFileSize(dataBlock):\n print(dataBlock[0:70])\n if dataBlock[0:4].decode() == \"TRQ \":\n if dataBlock[4:6].decode() == \"f \":\n for index in range(7, len(dataBlock)):\n if dataBlock[index:index+1].decode() == \" \":\n break\n filesize = \"\"\n for byte in range(index+1, len(dataBlock)):\n if dataBlock[byte:byte+1].decode() == \" \":\n return [int(filesize), byte+1]\n filesize += dataBlock[byte:byte+1].decode()\n return \"random return string\"","def file_tail(filename, n):\n result = ''\n with open(filename, 'r') as f:\n for line in (f.readlines()[-n:]):\n result += line\n\n return result","def createFileBytes(self, filename: unicode, offset: long, size: long, is_: java.io.InputStream, monitor: ghidra.util.task.TaskMonitor) -> ghidra.program.database.mem.FileBytes:\n ...","def len(self):\n return self._fsize - self._fp.tell()","def get_line_length(file_path):\n with open(file_path, 'rb+') as f:\n return len(f.readline())","def current_bytes(self):\n return self._current_bytes","def get_unknown_text(filename):\n d_file = codecs.open(os.path.join(u_path, filename), \"r\", \"utf-8\")\n s = d_file.read()\n d_file.close()\n return s","def get_file_mine(file, size_to_read=(5 * (1024 * 1024))):\n return magic.from_buffer(file.read(size_to_read), mime=True)","def trunc_utf8(string, num, etc=\"...\"):\n gb = string.decode(\"utf8\", \"ignore\").encode(\"gb18030\", \"ignore\")\n if num >= len(gb):\n return string\n ret = gb[:num].decode(\"gb18030\", \"ignore\").encode(\"utf8\")\n if etc:\n ret += etc\n return ret","def read(self) -> Optional[bytes]:","def _get_file_size(self):\n return self.s3_file.size","def get_size(fname):\n return os.path.getsize(fname)"],"string":"[\n \"def text_file_last_line(f, encoding=\\\"utf-8\\\"):\\n f.seek(0, os.SEEK_END) # start at the second to last character, since text files should end in a newline\\n if f.tell() < 2: return \\\"\\\"\\n f.seek(-2, os.SEEK_END) # start at the second to last character, since text files should end in a newline\\n while f.read(1) != b\\\"\\\\n\\\":\\n if f.tell() < 2: # arrived at beginning of file, so there is only one line\\n f.seek(0, os.SEEK_SET) # only one line; move to beginning so we read the whole line\\n break\\n f.seek(-2, os.SEEK_CUR) # go backward from the end of the file until we find a full line\\n return f.readline().decode(encoding)\",\n \"def getLastLine(f, blockSize=3072): \\n \\n f.seek(0,os.SEEK_END) \\n \\n totalBytes = f.tell() \\n \\n if totalBytes > blockSize: \\n f.seek(-blockSize,os.SEEK_END) \\n else: \\n f.seek(0) \\n \\n lastLines = f.readlines() \\n lastLine = lastLines[-1] \\n \\n if lastLine =='': \\n lastLine = lastLines[-2] \\n \\n return lastLine\",\n \"def injected_file_content_bytes(self):\\n return int(self.get('injected_file_content_bytes'))\",\n \"def last_chars(fh):\\n output = ''\\n with open(fh.name) as fh:\\n for line in fh:\\n output += line[-2]\\n return output\",\n \"def get_unknown_bytes(filename):\\n d_file = open(os.path.join(u_path, filename), \\\"rb\\\")\\n b = bytearray(d_file.read())\\n d_file.close()\\n return b\",\n \"def get_file_content(full_path):\\n print(full_path)\\n bytes = min(32, os.path.getsize(full_path))\\n raw = open(full_path, 'rb').read(bytes)\\n if '\\\\\\\\xff\\\\\\\\xfe' in str(raw):\\n print(\\\"utf-16\\\")\\n the_file = open(full_path, encoding=\\\"utf-16\\\",\\n errors=\\\"backslashreplace\\\")\\n data = the_file.read()\\n else:\\n print(\\\"latin-1\\\")\\n the_file = open(full_path, encoding=\\\"latin-1\\\",\\n errors=\\\"backslashreplace\\\")\\n data = the_file.read()\\n return data\",\n \"def read_file_bytes(read_file_path):\\n file_string = \\\"\\\"\\n with open(read_file_path, \\\"r\\\") as f:\\n for line in f:\\n file_string += line\\n return bytes(file_string, \\\"UTF-8\\\")\",\n \"def get_file_size(file):\\n\\tfile.seek(0, os.SEEK_END)\\n\\treturn file.tell()\",\n \"def get_file_encoding_bomsize(filename):\\n bom_encodings = ((codecs.BOM_UTF8, 'utf-8-sig'),\\n (codecs.BOM_UTF16_LE, 'utf-16le'),\\n (codecs.BOM_UTF16_BE, 'utf-16be'),\\n (codecs.BOM_UTF32_LE, 'utf-32be'),\\n (codecs.BOM_UTF32_BE, 'utf-32be'))\\n\\n firstbytes = open(filename, 'rb').read(4)\\n for bom, encoding in bom_encodings:\\n if firstbytes.startswith(bom):\\n file_encoding, size = encoding, len(bom)\\n break\\n else:\\n file_encoding, size = \\\"utf-8\\\", 0\\n\\n return file_encoding, size\",\n \"def max_offset(self):\\n return self.offset + self.filesize - 1\",\n \"def chars(count):\\n\\n global offset\\n\\n bytes=midifile[offset:offset+count]\\n offset+=count\\n return bytes\",\n \"def read(self, size: int = -1) -> bytes:\\n if self.size_read >= self.chunksize:\\n return b''\\n if size < 0:\\n size = self.chunksize - self.size_read\\n if size > self.chunksize - self.size_read:\\n size = self.chunksize - self.size_read\\n data = self.file.read(size)\\n self.size_read = self.size_read + len(data)\\n if self.size_read == self.chunksize and (self.chunksize & 1):\\n dummy = self.file.read(1)\\n self.size_read = self.size_read + len(dummy)\\n return data\",\n \"def get_file_size(fname, size_length):\\n size = os.path.getsize(fname)\\n return hex_encode(size, size_length)\",\n \"def get_file_size(file_path):\\n with open(file_path, 'rb') as infile:\\n infile.seek(0, 2)\\n return infile.tell()\",\n \"def get_char(self):\\n for i in range(1,5):\\n c = self.file[self.index:self.index+i] \\n try:\\n u = c.decode(self.encoding)\\n break\\n except UnicodeDecodeError:\\n if i==4:\\n raise \\n\\n self.index+=i\\n\\n if self.index > len(self.file):\\n self.eof=True\\n return None\\n\\n if u =='\\\\n':\\n self.line +=1\\n self.column = 0\\n else:\\n self.column+=1\\n\\n return u\",\n \"def get_bytes(self):\\n return self.bytes[:-len(ImageStream.END_SEQUENCE)]\",\n \"def get_file_size(file: Path) -> str:\\n return (\\n humanize.naturalsize(file.stat().st_size) if file.stat().st_size else \\\"unknown\\\"\\n )\",\n \"def num_long_term_bytes(self) -> str:\\n return pulumi.get(self, \\\"num_long_term_bytes\\\")\",\n \"def get_file_contents(file: str) -> tuple[bytes, str]:\\n\\n with open(file, \\\"rb\\\") as fp:\\n data = fp.read()\\n\\n if data[:4] == b\\\"\\\\x00\\\\x00\\\\xfe\\\\xff\\\":\\n return data, data.decode(\\\"utf-32be\\\")\\n elif data[:4] == b\\\"\\\\xff\\\\xfe\\\\x00\\\\x00\\\":\\n return data, data.decode(\\\"utf-32le\\\")\\n elif data[:4] == b\\\"\\\\x00\\\\x00\\\\x00\\\\x3c\\\":\\n return data, data.decode(\\\"utf-32be\\\")\\n elif data[:4] == b\\\"\\\\x3c\\\\x00\\\\x00\\\\x00\\\":\\n return data, data.decode(\\\"utf-32le\\\")\\n elif data[:4] == b\\\"\\\\x00\\\\x3c\\\\x00\\\\x3f\\\":\\n return data, data.decode(\\\"utf-16be\\\")\\n elif data[:4] == b\\\"\\\\x3c\\\\x00\\\\x3f\\\\x00\\\":\\n return data, data.decode(\\\"utf-16le\\\")\\n elif (data[:2] == b\\\"\\\\xfe\\\\xff\\\") and (data[2:4] != b\\\"\\\\x00\\\\x00\\\"):\\n return data, data[2:].decode(\\\"utf-16be\\\")\\n elif (data[:2] == b\\\"\\\\xff\\\\xfe\\\") and (data[2:4] != b\\\"\\\\x00\\\\x00\\\"):\\n return data, data[2:].decode(\\\"utf-16le\\\")\\n elif data[:3] == b\\\"\\\\xef\\\\xbb\\\\xbf\\\":\\n return data, data[3:].decode(\\\"utf-8\\\", errors=\\\"ignore\\\")\\n\\n # No byte order mark was found. Assume UTF-8, and ignore errors.\\n return data, data.decode(\\\"utf-8\\\", errors=\\\"ignore\\\")\",\n \"def reads(self, n):\\n return self.file.read(n).decode('iso-8859-1')\",\n \"def read_file_bytes(filepath: str):\\n with open(filepath, \\\"rb\\\") as reader:\\n return reader.read()\",\n \"def get_data_block_contents_bytes(self):\\n bb = self.volume.blkdev.block_bytes\\n if self.volume.is_ffs:\\n return bb\\n else:\\n return bb - 24\",\n \"def get_content(self):\\r\\n try:\\r\\n with open(self.path, \\\"rb\\\") as f:\\r\\n raw_data = f.read()\\r\\n result = chardet.detect(raw_data)\\r\\n encoding = result['encoding']\\r\\n if encoding:\\r\\n with open(self.path, 'r', encoding=encoding) as f:\\r\\n content = f.readlines()\\r\\n else:\\r\\n try:\\r\\n with open(self.path, 'r', encoding='ansi') as f:\\r\\n content = f.readlines()\\r\\n except UnicodeDecodeError:\\r\\n print('Jakis blad kodowania pliku')\\r\\n return -1\\r\\n return content\\r\\n except FileNotFoundError:\\r\\n print(f'Wskazany plik {self.path} nie istnieje')\\r\\n exit()\",\n \"def UTF8Size(utf):\\n ret = libxml2mod.xmlUTF8Size(utf)\\n return ret\",\n \"def get_filesize(inputfile) -> int:\\n with open(inputfile, \\\"rb\\\") as f:\\n lines = 0\\n buf_size = 1024 * 1024\\n read_f = f.raw.read\\n\\n buf = read_f(buf_size)\\n while buf:\\n lines += buf.count(b\\\"\\\\n\\\")\\n buf = read_f(buf_size)\\n\\n return lines\",\n \"async def collect_file(self, file):\\n with open(file, \\\"wb\\\") as f: # Opening file as writable in bytes\\n\\n total_bytes = 0\\n while True:\\n self.reader._eof = False # Force to read\\n data = await self.reader.read(1024)\\n\\n # Get buffer using BytesIO\\n chunk = io.BytesIO(data)\\n\\n total_bytes += chunk.getbuffer().nbytes\\n\\n # last_four = data[:-4].decode(\\\"utf-8\\\")\\n # last_four = chunk.getvalue()[-1:]\\n last_four = chunk.getvalue()\\n # print((last_four))\\n\\n check = \\\"end\\\"\\n\\n # print(check.encode())\\n\\n if last_four == check.encode():\\n print(\\\"Not Data\\\")\\n break\\n\\n f.write(chunk.getvalue())\\n print(f\\\"Collected: {total_bytes!r} bytes\\\")\",\n \"def get_file_size(input_file):\\n old_file_position = input_file.tell()\\n input_file.seek(0, os.SEEK_END)\\n size = input_file.tell()\\n input_file.seek(old_file_position, os.SEEK_SET)\\n return size\",\n \"def _extract_bytes(l):\\n # input should consist of a list of strings similar to ['123456', 'bytes', 'read']\\n if len(l) != 3 or l[1] != 'bytes' or l[2] != 'read':\\n raise UBootError('received unexpected reply when reading file')\\n\\n try:\\n return int(l[0])\\n except ValueError:\\n raise UBootError('received unexpected bytes value when reading file')\",\n \"def get_most_common_non_ascii_char(file_path: str) -> str:\\n with open(file_path, encoding=\\\"unicode-escape\\\") as f:\\n text = f.read()\\n chars = set(text) - set(\\n string.punctuation + string.ascii_letters + string.digits + \\\"\\\\n \\\"\\n )\\n char_dict = {char: text.count(char) for char in chars}\\n return sorted(char_dict.items(), key=lambda x: x[1])[-1][0]\",\n \"def content_length(self):\\n if isinstance(self.raw_content, bytes):\\n return len(self.raw_content)\\n if isinstance(self.raw_content, str):\\n return len(self.raw_content.encode('utf-8'))\\n pos = self.raw_content.tell()\\n try:\\n self.raw_content.seek(0, os.SEEK_END)\\n return self.raw_content.tell()\\n finally:\\n self.raw_content.seek(pos)\",\n \"def read_bytes(self, size):\\n return self.read('bytes:'+str(size))\",\n \"def get_data(filename=filename, maxlen=None):\\n print('reading {} with maxlen={}'.format(filename, maxlen))\\n if filename.startswith('s3'):\\n client = boto3.client('s3')\\n obj = client.get_object(Bucket='misc-data-ml-cottrell', Key='mangled_data.txt.gz')\\n fin = gzip.open(obj['Body'])\\n else:\\n fin = gzip.open(filename)\\n data = fin.read().decode('utf-8')\\n print('data len is {}'.format(len(data)))\\n data = data[:(maxlen+1)]\\n print('using data len {}'.format(len(data)))\\n chars = sorted(list(set(data))) # is int if data is bytes, is str if data is str\\n VOCAB_SIZE = len(chars)\\n return data, chars, VOCAB_SIZE\",\n \"def read_utf8_string(self, length):\\n return self.read(length).decode(\\\"utf-8\\\")\",\n \"def readBytes(self, size=1):\\n return self.bytes\",\n \"def get_file_size_in_bytes(file_path):\\n size = os.path.getsize(file_path)\\n return size\",\n \"def _read_until(self, c, chunk_size=96):\\n s = io.BytesIO()\\n fp = self._fp\\n eof = False\\n\\n while True:\\n chunk = fp.read(chunk_size)\\n\\n if not chunk:\\n # The end of the file was reached. We'll bail out of the loop\\n # and return everything we've read so far.\\n eof = True\\n break\\n\\n i = chunk.find(c)\\n\\n if i == -1:\\n # We didn't find the character. Store the entire chunk.\\n s.write(chunk)\\n else:\\n # We found the character. Store everything up to and including\\n # it, and then go back in the stream for the next read.\\n s.write(chunk[:i + 1])\\n fp.seek(i + 1 - len(chunk), os.SEEK_CUR)\\n break\\n\\n result = s.getvalue()\\n s.close()\\n\\n return result, eof\",\n \"def test_utf8_bytes_at_end(self):\\n # Python3 doesn't support bytestrings, don't run this test\\n if str is unicode:\\n return\\n input = \\\"A r\\\\xc3\\\\xa9sum\\\\xc3\\\\xa9, also spelled resum\\\\xc3\\\\xa9 or resume\\\"\\n output = input.split(\\\" \\\")\\n output[1] = output[1][0:-1]\\n for (itmO, itmV) in zip(output, tokenize_en(input)):\\n self.assertEqual(itmO, itmV[0])\",\n \"def unpack_utf8(self):\\n num_bytes_consumed, s = decode_utf8(self.view)\\n\\n self.num_bytes_consumed += num_bytes_consumed\\n self.view = self.view[num_bytes_consumed:]\\n\\n return num_bytes_consumed, s\",\n \"async def readinto(self, b):\\n # get the size of the file\\n size = await self._getsize()\\n b[:size] = await self.read(size)\\n return size\",\n \"def get_position(self):\\n\\n return (self._fileobj.tell() - self._pos) * 8 - self._bits\",\n \"def get_file_content(filename):\\n fp = open(filename, \\\"r\\\")\\n file_content = fp.read()\\n fp.close()\\n\\n # if the last character is an end of line, remove it\\n if file_content[-1] == '\\\\n':\\n file_content = file_content[:-1]\\n return file_content\",\n \"def get_file_encoding(file_path):\\n with open(file_path, 'r', encoding='utf-8') as f:\\n try:\\n for line in f:\\n pass\\n except UnicodeDecodeError:\\n return 'ISO-8859-1'\\n else:\\n return 'utf-8'\",\n \"def _decode_end(_fp):\\n return 0\",\n \"def download_bytes(self) -> str:\\n return pulumi.get(self, \\\"download_bytes\\\")\",\n \"def read_text_file(fpath, encoding, read_size=-1, force_unix_linebreaks=True):\\n with codecs.open(fpath, encoding=encoding) as f:\\n contents = f.read(read_size)\\n\\n if read_size > 0:\\n contents = contents[:read_size]\\n\\n if force_unix_linebreaks:\\n contents = linebreaks_win2unix(contents)\\n\\n return contents\",\n \"def _getsize(path, transport_params):\\n with smart_open.open(path, 'rb', ignore_ext=True, transport_params=transport_params) as fin:\\n fin.seek(0, io.SEEK_END)\\n return fin.tell()\",\n \"def _read_bytes(self, start, num_bytes):\\n with self._fp_lock:\\n self._fp.seek(start)\\n return self._fp.read(num_bytes)\",\n \"def byte_truncate(text):\\n\\ts = text.encode(ENCODING)\\n\\tif len(s) <= NAME_MAX_BYTES:\\n\\t\\treturn s\\n\\n\\tif ENCODING == \\\"utf-8\\\":\\n\\t\\tlead_byte = utf8_lead_byte\\n\\telse:\\n\\t\\traise NotImplementedError()\\n\\n\\ti = NAME_MAX_BYTES\\n\\twhile i > 0 and not lead_byte(s[i]):\\n\\t\\ti -= 1\\n\\treturn s[:i]\",\n \"def get_size(self, fileobject):\\n # move the cursor to the end of the file\\n fileobject.seek(0, 2)\\n size = fileobject.tell()\\n # move the cursor to the begin of the file\\n fileobject.seek(0)\\n return size\",\n \"def get_last_uploaded_byte(self, check_uri):\\n upload_check_headers = {'Content-Range': 'bytes */*'}\\n request_headers = dict(upload_check_headers.items() + self.standard_headers.items())\\n try:\\n HTTPClient().fetch(check_uri, method=\\\"PUT\\\", body='', headers=request_headers)\\n except HTTPError as e:\\n log.info(\\\"Upload check: status code %s\\\" % e.code)\\n if e.code == 308:\\n _range = int(e.response.headers['Range'].split('-')[1])\\n log.info(\\\"Last uploaded byte: %d\\\" % _range)\\n return _range\\n else: raise\\n raise ValueError(\\\"Upload check unsuccessful\\\")\",\n \"def read_gzip_bytes(self):\\n with gzip.open(self, 'rb') as f:\\n return f.read()\",\n \"def read(self, size=None):\\n if not size:\\n rlen = self.end - self.pos\\n else:\\n rlen = size\\n if rlen == 0:\\n return '' # Fake EOF\\n # Clamp rlen to window end\\n if self.pos + rlen > self.end:\\n rlen = self.end - self.pos\\n data = self.file.read(rlen)\\n self.pos += len(data)\\n return data\",\n \"def read_txt(txtfile):\\n with open(txtfile, \\\"r\\\", encoding=\\\"utf8\\\") as infile: \\n text = infile.read()\\n #print(text[0:100])\\n return text\",\n \"def read_bytes(self) -> bytes:\\n t = self.pc\\n while self.data[self.pc] != 0:\\n self.pc += 1\\n result = self.data[t:self.pc]\\n self.pc += 1 # jump '\\\\0'\\n return result\",\n \"def get_encoding_file(fname):\\n with io.open(fname, 'rt', encoding='ascii', errors='ignore') as f:\\n for unused in range(2):\\n line = f.readline()\\n match = get_encoding_re.search(line)\\n if match:\\n return match.group(1)\\n return 'ascii'\",\n \"def ra_unicode_read(fh, start, end):\\n tmp_pos = fh.tell()\\n fh.seek(0)\\n\\n buf = u\\\"\\\"\\n iter_fh = (line.decode('utf-8') for line in fh)\\n\\n for line in iter_fh:\\n line_len = len(line)\\n # .. recalibrate start/end ...\\n if line_len < start:\\n start = start - line_len\\n if end is not None:\\n end = end - line_len\\n else:\\n # ...until start is met\\n buf = buf + line\\n\\n # cutoff start\\n buf = buf[start:]\\n if end is not None:\\n end = end - start\\n\\n break\\n\\n if end is not None:\\n # read enough lines so the end is met\\n while len(buf) < end:\\n try:\\n buf = buf + iter_fh.next()\\n except StopIteration:\\n # end-of-file: end will not be met\\n break\\n buf = buf[:end]\\n\\n fh.seek(tmp_pos) # restore the file position\\n\\n return buf\",\n \"def get_size(fileobj):\\n\\n old_pos = fileobj.tell()\\n try:\\n fileobj.seek(0, 2)\\n return fileobj.tell()\\n finally:\\n fileobj.seek(old_pos, 0)\",\n \"def _filelength(self):\\r\\n with open(self.fileName, 'rb') as f:\\r\\n f.seek(0, 2) # move to end of file\\r\\n length = f.tell() # get current position\\r\\n return length\",\n \"def get_uncompressed_size(self):\\n return self.__uncompressed_size\",\n \"def read_block(file, block_size):\\n block = b\\\"\\\"\\n for i in range(block_size):\\n this_byte = file.read(1)\\n # If the last block consumed the last char in file:\\n if this_byte == b\\\"\\\" and i == 0:\\n return (-1, False)\\n # If we reach EOF prematurely:\\n elif this_byte == b\\\"\\\":\\n block += chr(0).encode()*(block_size - i)\\n return (block, False)\\n else:\\n block += this_byte\\n return (block, True)\",\n \"def test_file_bin_read_unicode_as_bin(self):\\n FileWriter(self.unicode_path).write_utf8(self.unicode_string)\\n bin_data = FileReader(self.unicode_path).read_bin() #read unicode file as binary\\n uni_text = bin_data.decode(\\\"utf-8\\\") #decode to utf-8\\n self.assertEqual(uni_text, self.unicode_string)\",\n \"def GetEncoding(self): \\n return self.file.GetEncoding()\",\n \"def FileLen(filename):\\n return os.stat(str(filename))[6]\",\n \"def tailFile(filename, offset, length):\\n\\n overflow = False\\n try:\\n f = open(filename, 'rb')\\n f.seek(0, 2)\\n sz = f.tell()\\n\\n if sz > (offset + length):\\n overflow = True\\n offset = sz - 1\\n\\n if (offset + length) > sz:\\n if offset > (sz - 1):\\n length = 0\\n offset = sz - length\\n\\n if offset < 0:\\n offset = 0\\n if length < 0:\\n length = 0\\n\\n if length == 0:\\n data = ''\\n else:\\n f.seek(offset)\\n data = f.read(length)\\n\\n offset = sz\\n return [data, offset, overflow]\\n\\n except (OSError, IOError):\\n return ['', offset, False]\",\n \"def find_file_end(chunks, pos):\\n\\n\\tpos = pos + 1\\n\\twhile pos < len(chunks)-1:\\n\\n\\t\\tif chunks[pos][0] != 0x100 and chunks[pos][0] != 0x102:\\n\\n\\t\\t\\t# This is not a block\\n\\t\\t\\treturn pos\\n\\n\\t\\telse:\\n\\t\\t\\tpos = pos + 1\\n\\n\\treturn pos\",\n \"def biggestLine(file_name, start, end):\\r\\n\\r\\n with open(file_name, \\\"r\\\") as file:\\r\\n bg_line = 0\\r\\n\\r\\n bg_line = len(file.readline())\\r\\n\\r\\n for line in islice(file, start, end):\\r\\n\\r\\n if bg_line < len(line):\\r\\n bg_line = len(line)\\r\\n\\r\\n return bg_line\",\n \"def read(self, size: int=-1) -> bytes:\\n ...\",\n \"def read(self, size: int=-1) -> bytes:\\n ...\",\n \"def file_size(file_path):\\n if os.path.isfile(file_path):\\n file_info = os.stat(file_path)\\n return convert_bytes(file_info.st_size)\",\n \"def file_size(file_path):\\n if os.path.isfile(file_path):\\n file_info = os.stat(file_path)\\n return convert_bytes(file_info.st_size)\",\n \"def file_size(file_path):\\n if os.path.isfile(file_path):\\n file_info = os.stat(file_path)\\n return convert_bytes(file_info.st_size)\",\n \"def get_string_length(self):\\n return int(self.read('H')[0])\",\n \"def get_file_chunk(self):\\n chunk = self.infile.read(self.CHUNK_SIZE)\\n if len(chunk) == 0:\\n return None\\n return chunk\",\n \"def file_size():\\n return os.path.getsize(FILE_NAME)\",\n \"def count_deleted_bytes(self): # FileObj.count_deleted_bytes\\n if self.deleted:\\n return self.bytes \\n else:\\n return 0\",\n \"def read(num_bytes):\\n # If file cannot be mmap'd (e.g. is stdin, or a fifo), fall back\\n # to doing an actual read from the file.\\n if not can_be_mmapd(fd):\\n return fd.read(num_bytes)\\n\\n bytes_available = max(file_size - offset, 0)\\n if bytes_available == 0:\\n return b\\\"\\\"\\n\\n return mmap.mmap(fd.fileno(), min(num_bytes, bytes_available), offset=offset, access=mmap.ACCESS_READ)\",\n \"def _get_correct_string_size(self, fd):\\n # store the lenght value bits\\n size_bits = []\\n # the value can have max 5 bytes\\n for _ in range(5):\\n size = self.read_bytes_to_int(fd, 1)\\n # if the length value is less then 128 it means its the last\\n # byte with the lenght value information and we can end.\\n if size < 128:\\n size_bits.append(bin(size)[2:])\\n break\\n else:\\n # when its bigger than 128 it means that the length value\\n # is using more bytes. We strip also the most significant\\n # bit\\n size_bits.append(bin(size)[3:])\\n\\n size_bits.reverse()\\n correct_size = int(\\\"\\\".join(size_bits), 2) \\n return correct_size\",\n \"def tailFile(filename, offset, length):\\r\\n\\r\\n overflow = False\\r\\n f = None\\r\\n try:\\r\\n f = open(filename, 'rb')\\r\\n f.seek(0, 2)\\r\\n sz = f.tell()\\r\\n\\r\\n if sz > (offset + length):\\r\\n overflow = True\\r\\n offset = sz - 1\\r\\n\\r\\n if (offset + length) > sz:\\r\\n if offset > (sz - 1):\\r\\n length = 0\\r\\n offset = sz - length\\r\\n\\r\\n if offset < 0: offset = 0\\r\\n if length < 0: length = 0\\r\\n\\r\\n if length == 0:\\r\\n data = ''\\r\\n else:\\r\\n f.seek(offset)\\r\\n data = f.read(length)\\r\\n\\r\\n offset = sz\\r\\n return [as_string(data), offset, overflow]\\r\\n\\r\\n except (OSError, IOError):\\r\\n return ['', offset, False]\\r\\n finally:\\r\\n if f:\\r\\n f.close()\",\n \"def getFileContent(fileName, encoding = \\\"UTF-8\\\"):\\n file = io.open(fileName, mode = \\\"r\\\", encoding = encoding)\\n text = file.read()\\n file.close()\\n return text\",\n \"def last_line(filename):\\n if not os.path.exists(filename):\\n return None\\n\\n try:\\n with open(filename, 'rb') as f:\\n f.seek(-2, os.SEEK_END)\\n while f.read(1) != b\\\"\\\\n\\\":\\n f.seek(-2, os.SEEK_CUR)\\n\\n last = f.readline()\\n return last\\n except OSError:\\n return None\",\n \"def UTF8Strsize(utf, len):\\n ret = libxml2mod.xmlUTF8Strsize(utf, len)\\n return ret\",\n \"def get_cur_file_size(fp, position_to_eof=False):\\r\\n if not position_to_eof:\\r\\n cur_pos = fp.tell()\\r\\n fp.seek(0, os.SEEK_END)\\r\\n cur_file_size = fp.tell()\\r\\n if not position_to_eof:\\r\\n fp.seek(cur_pos, os.SEEK_SET)\\r\\n return cur_file_size\",\n \"def test_file_utf8_readwrite(self):\\n FileWriter(self.unicode_path).write_utf8(self.unicode_string)\\n unicode_text = FileReader(self.unicode_path).read_utf8()\\n self.assertEqual(unicode_text, self.unicode_string)\",\n \"def get_uncompressed_size(self):\\n if not self.is_compressed():\\n return self.size\\n elif self.data is None:\\n raise ValueError('Data not read from record')\\n else:\\n return unpack('<L', self.data[:4])[0]\",\n \"def GetBytes(byte, size):\\n if sys.version_info[0] >= 3:\\n data = bytes([byte]) * size\\n else:\\n data = chr(byte) * size\\n return data\",\n \"def my_size(self):\\n return u\\\"{} bytes\\\".format(self.size)\",\n \"def read_full_file(filename, options=\\\"rb+\\\"):\\n with open(filename, options) as f:\\n text = f.read()\\n return text\",\n \"def __read(self, i: int) -> bytes:\\r\\n b = self.data[self.idx: self.idx + i]\\r\\n self.idx += i\\r\\n if len(b) != i:\\r\\n raise DecodingError(\\r\\n \\\"Incorrect byte length returned between indexes of {0} and {1}. Possible unexpected End of File.\\\"\\r\\n .format(str(self.idx), str(self.idx - i)))\\r\\n return b\",\n \"def getTRRFileSize(dataBlock):\\n print(dataBlock[0:70])\\n if dataBlock[0:4].decode() == \\\"TRQ \\\":\\n if dataBlock[4:6].decode() == \\\"f \\\":\\n for index in range(7, len(dataBlock)):\\n if dataBlock[index:index+1].decode() == \\\" \\\":\\n break\\n filesize = \\\"\\\"\\n for byte in range(index+1, len(dataBlock)):\\n if dataBlock[byte:byte+1].decode() == \\\" \\\":\\n return [int(filesize), byte+1]\\n filesize += dataBlock[byte:byte+1].decode()\\n return \\\"random return string\\\"\",\n \"def file_tail(filename, n):\\n result = ''\\n with open(filename, 'r') as f:\\n for line in (f.readlines()[-n:]):\\n result += line\\n\\n return result\",\n \"def createFileBytes(self, filename: unicode, offset: long, size: long, is_: java.io.InputStream, monitor: ghidra.util.task.TaskMonitor) -> ghidra.program.database.mem.FileBytes:\\n ...\",\n \"def len(self):\\n return self._fsize - self._fp.tell()\",\n \"def get_line_length(file_path):\\n with open(file_path, 'rb+') as f:\\n return len(f.readline())\",\n \"def current_bytes(self):\\n return self._current_bytes\",\n \"def get_unknown_text(filename):\\n d_file = codecs.open(os.path.join(u_path, filename), \\\"r\\\", \\\"utf-8\\\")\\n s = d_file.read()\\n d_file.close()\\n return s\",\n \"def get_file_mine(file, size_to_read=(5 * (1024 * 1024))):\\n return magic.from_buffer(file.read(size_to_read), mime=True)\",\n \"def trunc_utf8(string, num, etc=\\\"...\\\"):\\n gb = string.decode(\\\"utf8\\\", \\\"ignore\\\").encode(\\\"gb18030\\\", \\\"ignore\\\")\\n if num >= len(gb):\\n return string\\n ret = gb[:num].decode(\\\"gb18030\\\", \\\"ignore\\\").encode(\\\"utf8\\\")\\n if etc:\\n ret += etc\\n return ret\",\n \"def read(self) -> Optional[bytes]:\",\n \"def _get_file_size(self):\\n return self.s3_file.size\",\n \"def get_size(fname):\\n return os.path.getsize(fname)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.61801714","0.5925","0.59065217","0.58524173","0.5808898","0.56951183","0.5602501","0.5586138","0.55584174","0.55470455","0.55171245","0.5497499","0.5476367","0.5474634","0.54546565","0.5451746","0.5424187","0.541994","0.5400577","0.53996587","0.53982","0.5373797","0.53469837","0.5346389","0.53294647","0.5326958","0.53194237","0.5297021","0.5294629","0.52917695","0.5291011","0.52806985","0.5280027","0.52791595","0.52740765","0.5245539","0.5242934","0.52234954","0.5222632","0.5201054","0.51988757","0.5196334","0.51933736","0.5185536","0.51826364","0.5178726","0.51740015","0.5172801","0.51677644","0.51551276","0.51484007","0.51425505","0.51270115","0.5116509","0.5112104","0.5110208","0.5109874","0.5107554","0.5107416","0.510292","0.50893587","0.5087791","0.5079252","0.50719756","0.50592476","0.50457215","0.50457186","0.50457186","0.5045126","0.5045126","0.5045126","0.5044449","0.50349396","0.5032005","0.502986","0.5017214","0.501388","0.50130427","0.5012574","0.5012195","0.50083905","0.5006031","0.49965233","0.49950078","0.49942607","0.49934858","0.4989784","0.49827966","0.49767438","0.49669212","0.49668485","0.4962458","0.49573925","0.49547744","0.4953094","0.4952851","0.49509415","0.49507475","0.49452808","0.49452734"],"string":"[\n \"0.61801714\",\n \"0.5925\",\n \"0.59065217\",\n \"0.58524173\",\n \"0.5808898\",\n \"0.56951183\",\n \"0.5602501\",\n \"0.5586138\",\n \"0.55584174\",\n \"0.55470455\",\n \"0.55171245\",\n \"0.5497499\",\n \"0.5476367\",\n \"0.5474634\",\n \"0.54546565\",\n \"0.5451746\",\n \"0.5424187\",\n \"0.541994\",\n \"0.5400577\",\n \"0.53996587\",\n \"0.53982\",\n \"0.5373797\",\n \"0.53469837\",\n \"0.5346389\",\n \"0.53294647\",\n \"0.5326958\",\n \"0.53194237\",\n \"0.5297021\",\n \"0.5294629\",\n \"0.52917695\",\n \"0.5291011\",\n \"0.52806985\",\n \"0.5280027\",\n \"0.52791595\",\n \"0.52740765\",\n \"0.5245539\",\n \"0.5242934\",\n \"0.52234954\",\n \"0.5222632\",\n \"0.5201054\",\n \"0.51988757\",\n \"0.5196334\",\n \"0.51933736\",\n \"0.5185536\",\n \"0.51826364\",\n \"0.5178726\",\n \"0.51740015\",\n \"0.5172801\",\n \"0.51677644\",\n \"0.51551276\",\n \"0.51484007\",\n \"0.51425505\",\n \"0.51270115\",\n \"0.5116509\",\n \"0.5112104\",\n \"0.5110208\",\n \"0.5109874\",\n \"0.5107554\",\n \"0.5107416\",\n \"0.510292\",\n \"0.50893587\",\n \"0.5087791\",\n \"0.5079252\",\n \"0.50719756\",\n \"0.50592476\",\n \"0.50457215\",\n \"0.50457186\",\n \"0.50457186\",\n \"0.5045126\",\n \"0.5045126\",\n \"0.5045126\",\n \"0.5044449\",\n \"0.50349396\",\n \"0.5032005\",\n \"0.502986\",\n \"0.5017214\",\n \"0.501388\",\n \"0.50130427\",\n \"0.5012574\",\n \"0.5012195\",\n \"0.50083905\",\n \"0.5006031\",\n \"0.49965233\",\n \"0.49950078\",\n \"0.49942607\",\n \"0.49934858\",\n \"0.4989784\",\n \"0.49827966\",\n \"0.49767438\",\n \"0.49669212\",\n \"0.49668485\",\n \"0.4962458\",\n \"0.49573925\",\n \"0.49547744\",\n \"0.4953094\",\n \"0.4952851\",\n \"0.49509415\",\n \"0.49507475\",\n \"0.49452808\",\n \"0.49452734\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.79039353"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90885,"cells":{"query":{"kind":"string","value":"Test basic ranking page."},"document":{"kind":"string","value":"def test_northamerica_population(self):\n\n self.driver.get(self.url_ +\n '/ranking/Count_Person/Country/northamerica')\n\n subtitle_present = EC.text_to_be_present_in_element(\n (By.TAG_NAME, 'h3'), 'All Countries in North America')\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(subtitle_present)\n\n self.assertEqual(\n self.driver.find_element_by_tag_name('h1').text,\n 'Ranking by Population')\n\n table = self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/table')\n headers = table.find_elements_by_xpath('.//thead/tr/th')\n self.assertEqual(headers[0].text, 'Rank')\n self.assertEqual(headers[1].text, 'Country')\n self.assertEqual(headers[2].text, 'Value')\n row = table.find_elements_by_xpath('.//tbody/tr[1]/td')\n self.assertEqual(row[0].text, '1')\n self.assertEqual(row[1].text, 'United States of America')\n\n chart = self.driver.find_element_by_id('ranking-chart')\n y_text = chart.find_elements_by_class_name(\n 'y')[0].find_elements_by_tag_name('text')\n self.assertEqual(y_text[0].text, '0')\n self.assertEqual(y_text[-1].text, '300M')\n\n x_text = chart.find_elements_by_class_name(\n 'x')[0].find_elements_by_tag_name('text')\n self.assertEqual(x_text[0].text, 'United States of America')\n self.assertEqual(x_text[-1].text, 'Montserrat')"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def test_job_rank(self):\n inv_search = 'rank:Postdoc'\n spi_search = 'find rank Postdoc'\n self._compare_searches(inv_search, spi_search)","def test_something(self):\n\n allure.dynamic.title(\"Testing compute_ranks\")\n allure.dynamic.severity(allure.severity_level.NORMAL)\n allure.dynamic.description_html('<h3>Codewars badge:</h3>'\n '<img src=\"https://www.codewars.com/users/myFirstCode'\n '/badges/large\">'\n '<h3>Test Description:</h3>'\n \"<p>Test the function taht organizes a sports league in a \"\n \"round-robin-system. Each team meets all other teams. \"\n \"In your league a win gives a team 2 points, a draw gives \"\n \"both teams 1 point. After some games you have to compute \"\n \"the order of the teams in your league. You use the following \"\n \"criteria to arrange the teams:</p>\"\n \"<ul><li>- Points</li>\"\n \"<li>- Scoring differential (the difference between goals \"\n \"scored and those conceded)</li>\"\n \"<li>- Goals scored</li></ul>\")\n\n test_data = [\n (6,\n [[0, 5, 2, 2],\n [1, 4, 0, 2],\n [2, 3, 1, 2],\n [1, 5, 2, 2],\n [2, 0, 1, 1],\n [3, 4, 1, 1],\n [2, 5, 0, 2],\n [3, 1, 1, 1],\n [4, 0, 2, 0]],\n [4, 4, 6, 3, 1, 2]),\n (6,\n [[0, 5, 2, 0],\n [1, 4, 2, 2],\n [2, 3, 1, 3],\n [1, 5, 0, 0],\n [2, 0, 2, 1],\n [3, 4, 3, 1]],\n [2, 3, 4, 1, 5, 6]),\n (4,\n [[0, 3, 1, 1],\n [1, 2, 2, 2],\n [1, 3, 2, 0],\n [2, 0, 2, 0]],\n [3, 1, 1, 3]),\n (10,\n [],\n [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]),\n (8,\n [[0, 7, 2, 0]],\n [1, 2, 2, 2, 2, 2, 2, 8])\n ]\n\n for data in test_data:\n number = data[0]\n games = data[1]\n expected = data[2]\n actual_result = compute_ranks(number, games)\n print_log(number=number,\n games=games,\n expected=expected,\n actual_result=actual_result)\n\n with allure.step(\"Enter a test data and verify the result:\"):\n self.assertEqual(expected, actual_result)","def test_rank(self):\n self.assertEqual(self.vectors.rank('dog.n.01', 'dog.n.01'), 1)\n self.assertEqual(self.vectors.rank('dog.n.01', 'carnivore.n.01'), 3)","def test_get_ranking(self):\n card = Card.objects.create(suit=Card.CLUB, rank=\"jack\")\n self.assertEqual(card.get_ranking(), 11)","def test_rank(self):\n card = self._card\n self.assertEqual(card.rank, self._rank)","def test_get_rank_future(client_challenge_future):\n retval = login(client_challenge_future,\n \"cwhqsam\",\n \"supersecurepassword\")\n assert retval.status_code == 200\n\n retval = client_challenge_future.get(\"/api/v1/questions/rank\")\n assert retval.status_code == 200\n assert retval.get_json()[\"rank\"] == -1","def which_ranking(self, request):\n options = [\"user ranking\", \"team ranking\"]\n anwser = main_menu(\"Ranking\", options,\n \"Which ranking you want to see\")\n if anwser == options[0]:\n self.generate_rank(request, True)\n else:\n self.generate_rank(request, False)","def test_post_rank(self):\n # Upvote last post\n post = self.posts[-1]\n url = reverse_lazy('post-vote', kwargs={'pk': post.pk})\n response = self.put_json(url)\n\n self.assertEqual(response.status_code, status.HTTP_200_OK)\n\n # Search first tag\n url = reverse_lazy('tag-list') + '?search={}'.format(self.tags[0])\n\n response = self.client.get(url)\n self.assertEqual(response.status_code, status.HTTP_200_OK)\n\n result = response.data['results'][0]\n posts = result['posts']\n\n self.assertEqual(posts[0]['id'], post.pk)","def rank():\n return 0","def test_ranking_achievements(self):\n client = Client()\n user = create_user('passwordmichu', 'michu')\n client.login(username='michu', password='passwordmichu')\n coll = create_collection('Coleccion de cartas')\n problem = create_select_problem(coll, 'Problema')\n create_an_achievement_of_each(coll)\n submit_select_url = reverse('judge:submit', args=[problem.pk])\n client.post(submit_select_url, {'code': problem.solution}, follow=True)\n ranking_url = reverse('judge:result', args=[coll.pk])\n group_a = create_group('1A')\n group_a.user_set.add(user)\n response = client.get(ranking_url, follow=True)\n self.assertIn('x5', response.content.decode('utf-8'))","def display_ranking(args: argparse.Namespace, cookie: str) -> None:\n number: int = args.number\n if not number:\n return\n\n if number >= 50:\n print(\"Number of users to display cannot exceed 49 for now. Exiting.\")\n return\n\n cookies = {\n \"spip_session\": cookie,\n }\n\n ranking = get_global_ranking(args.offset, cookies)\n\n for counter, entry in enumerate(ranking):\n\n if counter >= number:\n break\n\n userid = ranking[entry][\"id_auteur\"]\n userinfo = get_user_info(userid, cookies)\n if userinfo is None:\n continue\n\n print(\n f\"{Fore.YELLOW}{ranking[entry]['place']:>4}{Style.RESET_ALL}. \"\n f\"{userinfo['nom']:<25} \"\n f\"{userinfo['score']} \"\n )\n\n return","def test_get_rank_today(client_challenge_today):\n retval = register(client_challenge_today, \"sam@codewizardshq.com\",\n \"cwhqsam\", \"supersecurepassword\", \"Sam\", \"Hoffman\",\n \"samuelhoffman2@gmail.com\")\n assert retval.status_code == 200\n\n retval = login(client_challenge_today,\n \"cwhqsam\",\n \"supersecurepassword\")\n assert retval.status_code == 200\n\n retval = client_challenge_today.get(\"/api/v1/questions/rank\")\n assert retval.status_code == 200\n assert retval.get_json()[\"rank\"] == 1","def test_wikipedia_matches_ranked_higher_than_facebook(w_driver):\n #1.) Do a Kasner search for some word\n w_driver.get('localhost:8000')\n query='sports'\n element = w_driver.find_element_by_name('query')\n element.send_keys(query + Keys.RETURN)\n\n #2.) Verify that wikipedia page is ranked higher than the facebook page.\n results=w_driver.page_source\n #results should list testsitewithnumber5.com before wikipedia result\n text_wikip='wikipedia.org/wiki/'+query\n #find the character that begins the reference to wikipedia page\n matches_wikip = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)), results)]\n #there are 2 references for testsitewithnumber6.com; we get the first.\n matches_wikip=matches_wikip[0]\n\n text_fb='www.facebook.com/public/'+query\n #find the character that begins the reference to facebook page\n matches_fb = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_fb)), results)]\n #there are 2 references for testsitewithnumber6.com; we get the first.\n matches_fb=matches_fb[0]\n\n #wikipedia page should occur first; its reference number should be smaller\n assert (matches_wikip<matches_fb)","def browse_rank(posts, rank):\n \n if rank > len(posts):\n click.secho(('\\n\\tCannot get that rank. Currently %d products have been hunted '\n 'with Spear.\\n' % len(posts)), fg=\"red\")\n return\n\n url = BASE_URL + posts[int(rank)-1]['permalink']\n webbrowser.open(url, new=2)","def test_lsr_rankings():\n for case in iter_testcases('rankings'):\n n_items = case[\"n_items\"]\n data = case[\"data\"]\n assert np.allclose(\n case[\"lsr_est\"], lsr_rankings(n_items, data),\n atol=ATOL, rtol=RTOL)","def test_ilsr_rankings():\n for case in iter_testcases('rankings'):\n n_items = case[\"n_items\"]\n data = case[\"data\"]\n assert np.allclose(\n case[\"ml_est\"], ilsr_rankings(n_items, data),\n atol=ATOL, rtol=RTOL)","def test_is_rank_integer(self):\n self.assertIsInstance(cardutils.Card(10,1).rank, int)","def test_home_pg(self):\n\n with self.client as client:\n resp = client.get(\"/\")\n html = resp.get_data(as_text=True)\n\n self.assertEqual(resp.status_code, 200)\n # p1 should be top 3 in trending while p4 is not\n self.assertIn(f'<a href=\"/post/{self.p1_id}\">', html)\n self.assertNotIn(f'<a href=\"/post/{self.p4_id}\">', html)\n # p5 is oldest post and should not appear\n self.assertNotIn(f'<a href=\"/post/{self.p5_id}\">', html)","def rank(request):\n links = Link.objects.order_by('-visualization')[:5]\n context = {'links' : links}\n return render(request, 'encurtador/rank.html', context)","def test_variant_rank_scores(case_obj, variant_obj):\n\n # GIVEN a case with SNV variants with rank score model\n assert case_obj[\"rank_model_version\"]\n # GIVEN a snv variant\n assert variant_obj[\"category\"] == \"snv\"\n # GIVEN that the variant has rank scores:\n variant_obj[\"rank_score_results\"] = [\n {\"category\": \"Splicing\", \"score\": 0},\n {\"category\": \"Inheritance_Models\", \"score\": -12},\n {\"category\": \"Consequence\", \"score\": 1},\n ]\n\n # GIVEN a test app containing config params to retrieve a genetic model\n test_app = create_app(\n config=dict(\n TESTING=True,\n DEBUG=True,\n MONGO_DBNAME=\"TEST_DB\",\n DEBUG_TB_ENABLED=False,\n LOGIN_DISABLED=True,\n RANK_MODEL_LINK_PREFIX=\"https://raw.githubusercontent.com/Clinical-Genomics/reference-files/master/rare-disease/rank_model/rank_model_-v\",\n RANK_MODEL_LINK_POSTFIX=\"-.ini\",\n )\n )\n with test_app.app_context():\n # THEN the rank score results of the variant should be returned by the function\n rank_score_results = variant_rank_scores(store, case_obj, variant_obj)\n assert isinstance(rank_score_results, list)\n # WITH the relative model ranges values\n assert rank_score_results[0][\"model_ranges\"]","def calculate_page_rank(self, iterations=5):\n # clear out the current page rank tables\n self.con.execute('drop table if exists pagerank')\n self.con.execute('create table pagerank(urlid primary key,score)')\n\n # initialize every url with a page rank of 1\n for (urlid,) in self.con.execute('select rowid from urllist'):\n self.con.execute('insert into pagerank(urlid,score) values (%d,1.0)' % urlid)\n self.dbcommit()\n\n for i in range(iterations):\n # Need multiple iterations, as the page ranks of pages linked to this\n # one will be consistently updated on each iteration\n print(\"Iteration %d\" % i)\n for (urlid,) in self.con.execute('select rowid from urllist'):\n # Default page rank\n page_rank = 0.15\n\n # Loop through all the pages that link to this one\n for (linker,) in self.con.execute('select distinct fromid from link where toid=%d'\n % urlid):\n # Get the page rank of the linker\n linkingpr = self.con.execute('select score from pagerank where urlid=%d'\n % linker).fetchone()[0]\n\n # Get the total number of links from the linker\n linkingcount = self.con.execute('select count(*) from link where fromid=%d'\n % linker).fetchone()[0]\n # add to page rank, accounting for the link count\n page_rank += 0.85 * (linkingpr / linkingcount)\n self.con.execute('update pagerank set score=%f where urlid=%d'\n % (page_rank, urlid))\n self.dbcommit()","def test_display_review(self):\n\n result = self.client.get(\"/brand/P87985432\")\n self.assertIn(b\"ever ever\", result.data)","def print_player_rank_and_points(self):\r\n pass","def test_rank(rank_list, word, n, test='bw'):\n tests = ['bw', 'ks']\n if test not in test:\n raise ValueError('Invalid test used. Need to choose either {} or {}\\\n tests.'.format(*tests))\n\n _, base_check = zip(*sorted(rank_list[word]))\n # just subtract the mean of the dataset, to make the mean 0\n std_base_check = [data - np.mean(base_check) for data in base_check]\n \n print(std_base_check)\n\n words = []\n for key, values in rank_list.items():\n if key == word:\n continue\n\n _, test_check = zip(*sorted(values))\n # move the data to a mean of 0\n std_test_check = [data - np.mean(test_check) for data in test_check]\n if test == 'bw':\n score = brendan_whitney_test(std_base_check, std_test_check)\n else:\n score, _ = ks_2samp(std_base_check, std_test_check)\n \n words.append((score, key))\n \n words = sorted(words)\n print(words[:n])\n return words[:n]","def __ranking_function(self, doc, query_tokens):","def test_mirror_ranking(self):\n updater = AptMirrorUpdater()\n # Make sure that multiple discovered mirrors are available.\n assert sum(m.is_available for m in updater.ranked_mirrors) > 10","def print_scores(ranked_list, links, topN=10):\n n = topN\n for url, score in ranked_list.items():\n try:\n _ = links.index(url)\n print(\"Score: %f \\t URL: %s\" %(score, url))\n\n n -= 1\n\n if n <= 0:\n break\n except:\n pass","def test_get_game_boxscore(self):\n pass","def test_page_renders_with_correct_template(self):\n p = Player()\n p.save()\n\n c = Client()\n response = c.get(reverse(self.test_url))\n self.assertTemplateUsed(response, 'display_stats/display_stats.html')","def test_score():\n print(\"Tests for 'score' function\")\n test_suite = TestSuite()\n\n # Testing with empty hand\n result = score([])\n test_suite.run_test(result, 0, '0')\n # Testing with non-empty hand\n result = score([1, 3])\n test_suite.run_test(result, 3, '1')\n # Testing with non-empty hand\n result = score([1, 3, 1, 1])\n test_suite.run_test(result, 3, '2')\n # Testing with non-empty hand\n result = score([4, 3, 4, 3, 3])\n test_suite.run_test(result, 9, '3')\n\n # Show report\n test_suite.report_results()","def test_get_user_rankings(self):\n user = User(name=u'no win', email=u'generic@thingy.com')\n user.put()\n \n userone = User(name=u'one win', email=u'generic@thingy.com', total_played=1, wins=1)\n userone.put()\n \n usertwo = User(name=u'two wins', email=u'generic@thingy.com', total_played=2, wins=1)\n usertwo.put()\n \n user_rankings = users.get_user_rankings()\n self.assertEquals(len(user_rankings), 2)\n self.assertEquals(user_rankings.pop().key, usertwo.key)\n self.assertEquals(user_rankings.pop().key, userone.key)","def test_happy_display_points_board_index(self):\n\n rv = self.app.get(\"/\")\n\n assert rv.status_code in [200]\n assert b\"Registration Portal\" in rv.data\n\n assert b\"Points Board\" in rv.data\n for club in self.clubs:\n assert str.encode(club[\"name\"]) in rv.data\n assert str.encode(f\"Current Points: {club['points']}\") in rv.data","def test_name_matches_ranked_higher_than_wikipedia(w_driver):\n w_driver.get('localhost:8000/add_form')\n #1.) Add a test website into db\n name_element=w_driver.find_element_by_name('name')\n name_element.send_keys('test site')\n url_element=w_driver.find_element_by_name('url')\n url_element.send_keys('www.testsitewithnumber5.com')\n number_element=w_driver.find_element_by_name('number')\n number_element.send_keys(5)\n name_element.send_keys(Keys.RETURN)\n\n #2.) Do a Kasner search and verify that the website is ranked higher than \n #the wikipedia result for the query.\n w_driver.get('localhost:8000')\n element = w_driver.find_element_by_name('query')\n element.send_keys('test site' + Keys.RETURN)\n\n results=w_driver.page_source\n #results should list testsitewithnumber5.com before wikipedia result\n text5='www.testsitewithnumber5.com' \n #find the character that begins the reference to testsitewithnumber6.com\n matches5 = [matches.start() \n for matches in re.finditer(r'{}'.format(re.escape(text5)), results)]\n #there are 2 references for testsitewithnumber6.com; we get the first.\n matches5=matches5[0]\n\n text_wikip='wikipedia.org/wiki/test'\n #find the character that begins the reference to wikipedia site\n matches_wikip = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)),results)]\n #there are 2 references for the url; we get the first.\n matches_wikip=matches_wikip[0]\n\n #matches5 should occur first, so its reference character should be smaller\n assert (matches5<matches_wikip)","def print_end_of_round_rankings(self):\n print('\\nFINAL SCORES')\n print('-'*12)\n for k, v in enumerate(self.ranking_dict):\n print(f\"{k+1} {v[0]}: {v[1]}\")\n print('\\n')","def test_get_roles_rank(self):\n contrib_as = self.make_assignment(\n self.project, self.user_bob, self.role_contributor\n )\n guest_as = self.make_assignment(\n self.project, self.user_carol, self.role_guest\n )\n roles = self.project.get_roles()\n self.assertIn(contrib_as, roles)\n self.assertIn(guest_as, roles)\n roles = self.project.get_roles(max_rank=30)\n self.assertEqual(roles, [contrib_as, self.owner_as_cat])\n roles = self.project.get_roles(max_rank=40)\n self.assertEqual(roles, [contrib_as, guest_as, self.owner_as_cat])\n roles = self.project.get_roles(min_rank=30)\n self.assertEqual(roles, [contrib_as, guest_as])\n roles = self.project.get_roles(min_rank=40)\n self.assertEqual(roles, [guest_as])\n roles = self.project.get_roles(min_rank=30, max_rank=30)\n self.assertEqual(roles, [contrib_as])","def rank(args: argparse.Namespace, cookie: str) -> None:\n cookies = {\n \"spip_session\": cookie,\n }\n\n max_len_user = max((len(x) for x in args.user))\n\n for username in args.user:\n\n users: Dict[str, Searchresult] = search_user(username, cookies)\n\n userids = []\n # Parse API search results in order to find the correct users\n for user in users:\n if users[user][\"nom\"] == username:\n userids.append(users[user][\"id_auteur\"])\n\n if not userids:\n raise ValueError(\n (\n \"User %s not found in 50 first search terms.\"\n \" Did you mispell it? \"\n ),\n username,\n )\n\n for userid in userids:\n userinfo = get_user_info(userid, cookies)\n if userinfo is None:\n continue\n\n print(\n f\"[{Fore.GREEN}+{Style.RESET_ALL}] \"\n f\"{userinfo['nom']:<{max_len_user + 2}} \"\n f\"Score: {userinfo['score']:<6} \"\n f\"Position: {userinfo['position']}\"\n )","def diversified_ranking(self):\n self.error_throw('rank')\n instance = Instance(self.table_name)\n instance.addTable(Table(instance,False,'','')) # 'False'->transformed '',''->no describe yet\n if self.import_method == 'mysql': instance = self.mysql_handle(instance)\n elif self.import_method == 'csv': instance = self.csv_handle(instance)\n\n self.rank_partial(instance)\n\n self.rank_method = methods_of_ranking[3] # = 'diversified_ranking'","def get_num_hit_rank(boxes_truth, boxes_pred, rank):\n\n def is_hit(box_truth, box_pred):\n return is_label_match_rank(box_truth, box_pred, rank)\n\n return get_num_hit(boxes_truth, boxes_pred, is_hit)","def calc_rank(id=13197473):\r\n player_url = urllib.parse.urlparse(\"http://osu.ppy.sh/pages/include/profile-general.php?u=player_id&m=0\".replace('player_id', str(id)))\r\n page = urlopen(player_url.geturl())\r\n soup = BeautifulSoup(page, features=\"html.parser\")\r\n table_divs = soup.findAll('div', attrs={'class': 'profileStatLine'})\r\n\r\n import re\r\n pattern = '\\(#\\d*,*\\d+\\)'\r\n for div in table_divs:\r\n for childdiv in div.find_all('b'):\r\n result = re.search(pattern, str(childdiv.text))\r\n my_ranking = int(result.group(0).replace(',', '').replace(\"(#\", '').replace(\")\", ''))\r\n break\r\n break\r\n return my_ranking","def test(\n self, render=True, opponent=None, wormzero_player=None, num_tests=1, num_gpus=0\n ):\n opponent = opponent if opponent else self.config.opponent\n wormzero_player = wormzero_player if wormzero_player else self.config.wormzero_player\n self_play_worker = self_play.SelfPlay.options(\n num_cpus=0, num_gpus=num_gpus,\n ).remote(self.checkpoint, self.Game, self.config, np.random.randint(10000))\n results = []\n for i in range(num_tests):\n print(f\"Testing {i+1}/{num_tests}\")\n results.append(\n ray.get(\n self_play_worker.play_game.remote(\n 0, 0, render, opponent, wormzero_player,\n )\n )\n )\n self_play_worker.close_game.remote()\n\n if len(self.config.players) == 1:\n result = np.mean([sum(history.reward_history) for history in results])\n else:\n result = np.mean(\n [\n sum(\n reward\n for i, reward in enumerate(history.reward_history)\n if history.to_play_history[i - 1] == wormzero_player\n )\n for history in results\n ]\n )\n return result","def __rank__(self) -> int:","def get_ranks(d): \n raise NotImplementedError(\"Problem 3 Incomplete\")","def get_rank(self) -> int:\r\n return self.rank","def test_verify_non_existing_rank(self):\n data = {'rank_id': 5}\n res = self.post(url='/verify/4', data=data, role='admin')\n self.assertEqual(res.status_code, 401)\n self.assertException(res, exc.EntryNotFound)","def get_rank(self):\n details = self._tab.find(\"table\", class_=\"details\")\n rank, = self.rank_re.match(details.find(\"td\", class_=\"value\").get_text()).groups()\n return rank","def test_profile_page_returns_correct_html(self):\n self.add_testuser()\n response = self.client.get('/profile/testuser/')\n self.assertContains(response, 'Album Count')","def test_03_leaderboard(self):\r\n # As Anonymou user\r\n url = \"/leaderboard\"\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should not be shown to anonymous users\"\r\n assert dom.find(id='enforce_privacy') is not None, err_msg\r\n # As Authenticated user but NOT ADMIN\r\n self.signin()\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should not be shown to authenticated users\"\r\n assert dom.find(id='enforce_privacy') is not None, err_msg\r\n self.signout\r\n # As Authenticated user but ADMIN\r\n self.signin(email=self.root_addr, password=self.root_password)\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should be shown to admin users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout()","def test_figure1(self):\n\n searcher = SimpleSearcher.from_prebuilt_index('msmarco-passage')\n hits = searcher.search('what is a lobster roll?', 10)\n\n self.assertAlmostEqual(hits[0].score, 11.00830, delta=0.0001)\n self.assertEqual(hits[0].docid, '7157707')\n\n self.assertAlmostEqual(hits[9].score, 9.92200, delta=0.0001)\n self.assertEqual(hits[9].docid, '6234461')\n\n self.assertEqual(len(hits), 10)","def test_03_leaderboard(self):\r\n # As Anonymou user\r\n url = \"/leaderboard\"\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should be shown to anonymous users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n # As Authenticated user but NOT ADMIN\r\n self.signin()\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should be shown to authenticated users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout\r\n # As Authenticated user but ADMIN\r\n self.signin(email=self.root_addr, password=self.root_password)\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Leaderboard page should be shown to admin users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout()","def get_rank(self):\r\n return self.rank","def test_pages_with_same_name_ranked_by_number(w_driver):\n w_driver.get('localhost:8000/add_form')\n #1.) Add two websites with same name, but different number into db\n #First site, with number=5\n name_element=w_driver.find_element_by_name('name')\n name_element.send_keys('test site')\n url_element=w_driver.find_element_by_name('url')\n url_element.send_keys('www.testsitewithnumber5.com')\n number_element=w_driver.find_element_by_name('number')\n number_element.send_keys(5)\n name_element.send_keys(Keys.RETURN)\n\n #Second site, with number=6\n name_element=w_driver.find_element_by_name('name')\n name_element.send_keys('test site')\n url_element=w_driver.find_element_by_name('url')\n url_element.send_keys('www.testsitewithnumber6.com')\n number_element=w_driver.find_element_by_name('number')\n number_element.send_keys(6)\n name_element.send_keys(Keys.RETURN)\n\n #2.) Do a Kasner search and verify that the websites are sorted in \n # descending order in the search results.\n w_driver.get('localhost:8000')\n element = w_driver.find_element_by_name('query')\n element.send_keys('test site' + Keys.RETURN)\n\n results=w_driver.page_source\n #results should list testsitewithnumber6.com before testsitewithnumber5.com\n text6='www.testsitewithnumber6.com' \n #find the character that begins the reference to testsitewithnumber6.com\n matches6 = [matches.start() \n for matches in re.finditer(r'{}'.format(re.escape(text6)), results)]\n #there are 2 references for testsitewithnumber6.com; we get the first.\n matches6=matches6[0]\n\n #find the character that begins the reference to testsitewithnumber5.com\n text5='www.testsitewithnumber5.com' \n matches5 = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text5)), results)]\n #there are 2 references for testsitewithnumber5.com. we get the first.\n matches5=matches5[0]\n\n #matches6 should occur first, so its reference character should be smaller\n assert (matches6<matches5)","def article_rank():\n db, c = connect(DBNAME)\n c.execute(\"select title, count(title) as views from \\\"pathslug\\\" \"\n \"group by title order by views desc limit 3\")\n article_table = c.fetchall()\n db.close()\n print \"\\nThree Most Popular Articles All Time:\"\n for article in article_table:\n print str(article[0]) + \" - \" + str(article[1]) + \" views\"","def test_calculate_number_page(self):\n self.requestapi = RequestApi('snack')\n result = self.requestapi.calculate_number_page()\n self.assertEqual(result, 1)","def report(results, n_top=1):\n for i in range(1, n_top + 1):\n candidates = np.flatnonzero(results['rank_test_score'] == i)\n for candidate in candidates:\n print(f\"Model with rank: {i}\")\n print(f\"Mean validation score: {results['mean_test_score'][candidate]} (std: {results['std_test_score'][candidate]}\")\n print(f\"Parameters: {results['params'][candidate]}\")","def print_top_pageranks(topn):\n print('Top page ranks:\\n%s' % ('\\n'.join('%s\\t%.5f' % (u, v) for u, v in topn)))","def test_scrape(self):\n self.assertEqual(self.scraped.title, 'Heading!')\n self.assertEqual(self.scraped.link_text, 'Go to Google')\n self.assertEqual(self.scraped.link_url, 'http://Google.com')","def rank():\n return int(os.environ['RANK'])","def testHighscore(self):\n \n game = Game.objects.get(title='game0')\n gameplayeds = game.gameplayed_set\n ply_group = Group.objects.get(name='Player')\n for i in range(4):\n user = ply_group.user_set.get(username='ply{}'.format(i))\n gameplayed = gameplayeds.get(user=user)\n gameplayed.gameScore = i\n gameplayed.save()\n \n response = self.client.get(\n reverse('api:game-buyers', args=['v1', 'game0']),\n {'order_by': 'gameScore'},\n format='json'\n )\n self.assertEquals(response.status_code, 200)\n content = self.parser.parse(BytesIO(response.content))\n for i in range(4):\n self.assertEquals(content['results'][i]['user'], 'ply{}'.format(i))\n \n response = self.client.get(\n reverse('api:game-buyers', args=['v1', 'game0']),\n {'order_by': '-gameScore'},\n format='json'\n )\n self.assertEquals(response.status_code, 200)\n content = self.parser.parse(BytesIO(response.content))\n for i in range(4):\n self.assertEquals(content['results'][i]['user'], 'ply{}'.format(3 - i))","def test_page(self):\n result = self.test_client.page\n\n assert result == 1","def test_rank():\n keys = list(range(100))\n d = OrderedTreeDict((key, None) for key in keys)\n assert all(list(\n keys[d.rank(k)] == k for k in keys\n ))","def test_summary_with_results(self):\n add_constituency_result_line('X, 10, C')\n r = self.client.get('/summary')\n self.assertEqual(r.status_code, 200)\n soup = BeautifulSoup(r.data, 'html.parser')\n self.assertIs(soup.find(id='no-results'), None)\n self.assertIsNot(soup.find(id='results-table'), None)","def test_get_score(self):\r\n score_dict = self.combinedoe.get_score()\r\n self.assertEqual(score_dict['score'], 0)\r\n self.assertEqual(score_dict['total'], 1)","def evaluate_ranks(articles, rank_tuples):\n rank_tp=defaultdict(int)\n rank_fn=defaultdict(int)\n rank_fp=defaultdict(int)\n \n for article in articles:\n for mention in article.entity_mentions:\n form=mention.mention\n meaning=mention.gold_link\n sys_meaning=mention.sys_link\n t_gold=(form, meaning)\n t_sys=(form, sys_meaning)\n for rank, r_tuples in rank_tuples.items():\n if t_gold in r_tuples and t_sys in r_tuples:\n rank_tp[rank]+=1\n break\n elif t_gold in r_tuples:\n rank_fn[rank]+=1\n elif t_sys in r_tuples:\n rank_fp[rank]+=1\n print('tp', rank_tp)\n print('fp', rank_fp)\n print('fn', rank_fn)\n \n rank_prec={}\n rank_recall={}\n rank_f1={}\n \n for rank in range(1,13):\n if rank_tp[rank]+rank_fp[rank]>0:\n rank_prec[rank]=rank_tp[rank]/(rank_tp[rank]+rank_fp[rank])\n else:\n rank_prec[rank]=0.0\n if rank_tp[rank]+rank_fn[rank]>0:\n rank_recall[rank]=rank_tp[rank]/(rank_tp[rank]+rank_fn[rank])\n else:\n rank_recall[rank]=0.0\n if rank_prec[rank]+rank_recall[rank]>0:\n rank_f1[rank]=2*rank_prec[rank]*rank_recall[rank]/(rank_prec[rank]+rank_recall[rank])\n else:\n rank_f1[rank]=0.0\n print('precision', rank_prec)\n print()\n print('recall', rank_recall)\n print()\n print('f1', rank_f1)\n print()\n return rank_prec, rank_recall, rank_f1","def alexa_rank(url):\n return ALEXA_MAP.index(url)","def test_index_page(self):\n response = self.client.get(\"\")\n self.assertEqual(response.status_code, 200)\n self.assertTemplateUsed(response, \"search/base.html\")\n self.assertTemplateUsed(response, \"search/search_form.html\")\n self.assertTemplateUsed(response, \"search/index.html\")","def test_scores(self):\n pig = game.pig.Pig('PlayerA', 'PlayerB', 'PlayerC')\n self.assertEqual(\n pig.get_score(),\n {\n 'PlayerA': 0,\n 'PlayerB': 0,\n 'PlayerC': 0\n }\n )","def score(self, test_data):\n\n\t\tpass","def testGetsPresentRichness(self):\n self.assertEqual(37, self.coal.get_species_richness(1))\n self.assertEqual(53, self.coal.get_species_richness(3))","def rerank_candidates(s, pred2sub_rank, all_predictions, rerank_top=20):\n predicted_smiles = []\n model_input = []\n for (predict_smi, label), _ in Counter(all_predictions).most_common(rerank_top):\n if predict_smi == s:\n continue\n features = get_all_features(\n get_all_ranking_info(pred2sub_rank[predict_smi]))\n predicted_smiles.append((predict_smi, label))\n model_input.append(features)\n\n model = RankingModel()\n model.load_state_dict(torch.load('./models/ranker/rank_model.pt', map_location='cpu'))\n model.eval()\n\n test_loader = DataLoader(RankingTestDataset(\n model_input), batch_size=1000, shuffle=False, num_workers=2)\n ranking_scores = []\n for data in test_loader:\n outputs = model(data)[0]\n ranking_scores.extend(outputs.detach().cpu().numpy())\n\n assert len(predicted_smiles) == len(ranking_scores)\n pred_smi2score = {k: v[1]\n for k, v in zip(predicted_smiles, ranking_scores)}\n return pred_smi2score","def rank(self):\n return self.lib.calculate_rank()","def test_default_page_size(self):\n response = self.client.get('/')\n self.assertEqual(10, len(response.data['results']))","def test_wikipedia_matches_ranked_higher_than_keyword_matches(w_driver):\n keyword='test site'\n w_driver.get('localhost:8000/add_form')\n \n #1.) Add 2 websites into db with keyword=test site\n name_element1=w_driver.find_element_by_name('name')\n name_element1.send_keys('first one')\n url_element1=w_driver.find_element_by_name('url')\n url_element1.send_keys('www.firstsite5.com')\n number_element1=w_driver.find_element_by_name('number')\n number_element1.send_keys(5)\n words_element1=w_driver.find_element_by_name('words')\n words_element1.send_keys(keyword)\n words_element1.send_keys(Keys.RETURN)\n\n name_element2=w_driver.find_element_by_name('name')\n name_element2.send_keys('first one')\n url_element2=w_driver.find_element_by_name('url')\n url_element2.send_keys('www.secondsite5.com')\n number_element2=w_driver.find_element_by_name('number')\n number_element2.send_keys(5)\n words_element2=w_driver.find_element_by_name('words')\n words_element2.send_keys(keyword)\n words_element2.send_keys(Keys.RETURN)\n\n #2.) Do a Kasner search and verify that the websites with keyword=tests is\n #ranked higher than the wikipedia site.\n w_driver.get('localhost:8000')\n element = w_driver.find_element_by_name('query')\n element.send_keys(keyword + Keys.RETURN)\n\n results=w_driver.page_source\n #results should list wikipedia site before keyword match results\n text_wikip='wikipedia.org/wiki/'+keyword\n #find the character that begins the reference to wikipedia site \n matches_wikip = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)), results)]\n #there are 2 references for wikipedia site; we get the first.\n matches_wikip=matches_wikip[0]\n\n text_kw_1='www.firstsite5.com'\n #find the character that begins the reference to www.firstsite5.com\n matches_kw_1 = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_kw_1)),results)]\n #there are 2 references for the url; we get the first.\n matches_kw_1=matches_kw_1[0]\n\n text_kw_2='www.secondsite5.com'\n #find the character that begins the reference to www.firstsite5.com\n matches_kw_2 = [matches.start()\n for matches in re.finditer(r'{}'.format(re.escape(text_kw_2)),results)]\n #there are 2 references for the url; we get the first.\n matches_kw_2=matches_kw_2[0]\n \n assert (matches_wikip<matches_kw_1)\n assert (matches_wikip<matches_kw_2)","def index(self):\n\t\t# Return a rendered template\n\t\t# return render('/some/template.mako')\n\t\t# or, Return a response\n\t\tScoreRankHistogramType = model.Stock_250kDB.ScoreRankHistogramType\n\t\trows = ScoreRankHistogramType.query.order_by(ScoreRankHistogramType.results_type).\\\n\t\t\torder_by(ScoreRankHistogramType.null_distribution_type_id).\\\n\t\t\torder_by(ScoreRankHistogramType.min_distance).order_by(ScoreRankHistogramType.get_closest).\\\n\t\t\torder_by(ScoreRankHistogramType.min_MAF).order_by(ScoreRankHistogramType.allow_two_sample_overlapping).all()\n\t\tc.score_rank_hist_type_ls = rows\n\t\treturn render('/score_rank_histgram.html')","def test_quiz_statistics_view(self):\n target_url = url_for('quiz.quiz_statistics')\n redirect_url = url_for('users.login', next=target_url)\n response = self.client.get(target_url)\n self.assertEqual(response.status_code, 302)\n self.assertRedirects(response, redirect_url)","def test_github_explore(self):\n python_trending_url = \"https://github.com/trending/python\"\n driver = self.driver\n driver.get(self.base_url)\n search_box = driver.find_element_by_link_text(\"Explore\")\n search_box.send_keys(Keys.RETURN)\n time.sleep(2)\n trending_elem = driver.find_element_by_partial_link_text(\"Trending\")\n trending_elem.send_keys(Keys.ENTER)\n time.sleep(2)\n python_lang = driver.find_element_by_link_text(\"Python\")\n python_lang.send_keys(Keys.ENTER)\n time.sleep(2)\n assert python_trending_url == driver.current_url","def test_homepage(self):\n\n response = self.client.get(\"/\")\n self.assertIn(\"Books</title>\", response.data)\n self.assertIn(\"Goodreads ID\", response.data)","def get_ranking(self,\n query=\"\",\n number=10,\n use_SVD=True #if false: VSM based similarity used\n ):\n if self.is_valid:\n print(\"Ranking - Query: \", query)\n\n # Preprocess string: Convert into bag of words\n bow = fp.string_transformation(query)\n # bow = ['underground','underwater','wireless','communications']\n # Doc sci.electronics/52434\n print(\"Ranking - BoW: \", bow)\n\n # Convert into term vector\n term_vec = self.tfidf_handler.convertBoWToTermVector(bow)\n print(\"Ranking - Term Vector: Non-Zero positions:\", term_vec.nonzero())\n print(\"Ranking - Term Vector: Non-Zero values:\", term_vec[term_vec.nonzero()])\n print(\"Ranking - Term Vector: Size:\", term_vec.size)\n\n if(use_SVD):\n # Calc. LSI ranking (SVD)\n doc_ind, similarities = self.svd_handler.get_ranking(vec=term_vec, number=number)\n else:\n # Calc. VSM ranking\n train_dtm = self.tfidf_handler.getTDM().transpose().tocsr()\n norms = norm(train_dtm, axis=1) * np.linalg.norm(term_vec) # Normalize\n sims = np.array(train_dtm.dot(term_vec) / norms)\n\n # Derive top-n\n topn = np.argsort(sims)[::-1][:number]\n similarities = sims[topn]\n doc_ind = topn\n\n return self.tfidf_handler.getDocuments(doc_ind), similarities\n else:\n print(\"Ranking: Error - Instantiation of LSI failed\")\n return [\"\"], [0]","def test_index(self):\n result = self.client.get('/')\n self.assertEqual(result.status, '200 OK')\n self.assertIn(b'Game', result.data)","def rank(ctx, path, metric, revision, limit, desc, threshold, wrap):\n config = ctx.obj[\"CONFIG\"]\n\n if not exists(config):\n handle_no_cache(ctx)\n\n from wily.commands.rank import rank\n\n logger.debug(f\"Running rank on {path} for metric {metric} and revision {revision}\")\n rank(\n config=config,\n path=path,\n metric=metric,\n revision_index=revision,\n limit=limit,\n threshold=threshold,\n descending=desc,\n wrap=wrap,\n )","def test_04_global_stats_index(self):\r\n # As Anonymou user\r\n url = \"/stats\"\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should not be shown to anonymous users\"\r\n assert dom.find(id='enforce_privacy') is not None, err_msg\r\n # As Authenticated user but NOT ADMIN\r\n self.signin()\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should not be shown to authenticated users\"\r\n assert dom.find(id='enforce_privacy') is not None, err_msg\r\n self.signout\r\n # As Authenticated user but ADMIN\r\n self.signin(email=self.root_addr, password=self.root_password)\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should be shown to admin users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout()","def test_get_score(self):\r\n score_dict = self.combinedoe.get_score()\r\n self.assertEqual(score_dict['score'], 15.0)\r\n self.assertEqual(score_dict['total'], 5.0)","def assignRanks(self):\r\n\t\trank = 0\r\n\t\tscores = list(self._playerScores)\r\n\t\tscores.reverse()\r\n\t\tfor playerScore in scores:\r\n\t\t\tif not playerScore.has(NOT_MET) or not playerScore.value(NOT_MET):\r\n\t\t\t\trank += 1\r\n\t\t\t\tplayerScore.set(RANK, smallText(BugUtil.colorText(u\"%d\" % rank, ScoreOpt.getRankColor())))\r\n\t\tif rank > 0:\r\n\t\t\tself._anyHas[RANK] = True","def disp_score():","def getPageRank(elistPath, alpha, maxiter, tolerance):\n\n adjGraph = AdjGraph(elistPath, separator=\" \")\n graph = adjGraph.SNAPGraph\n\n preference_vector = []\n for node in graph.Nodes():\n id = node.GetId()\n if (id % 4) == 0:\n preference_vector.append(id)\n\n pageRank, convIter, time = biasedPageRank(\n adjGraph, preference_vector=preference_vector, alpha=alpha,\n max_iterations=maxiter, tolerance=tolerance)\n\n writeCentrality(\"pagerank.txt\", pageRank)\n return pageRank, convIter, time","def test_04_global_stats_index(self):\r\n # As Anonymou user\r\n url = \"/stats\"\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should be shown to anonymous users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n # As Authenticated user but NOT ADMIN\r\n self.signin()\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should be shown to authenticated users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout\r\n # As Authenticated user but ADMIN\r\n self.signin(email=self.root_addr, password=self.root_password)\r\n res = self.app.get(url, follow_redirects=True)\r\n dom = BeautifulSoup(res.data)\r\n err_msg = \"Stats page should be shown to admin users\"\r\n assert dom.find(id='enforce_privacy') is None, err_msg\r\n self.signout()","async def handle_rank_ups(self, user: discord.User, brawler: str):\n\n brawler_data = await self.get_player_stat(\n user, 'brawlers', is_iter=True, substat=brawler)\n\n pb = brawler_data['pb']\n rank = brawler_data['rank']\n\n rank_as_per_pb = self.get_rank(pb)\n\n if rank_as_per_pb <= rank:\n return False\n\n await self.update_player_stat(\n user, 'brawlers', rank_as_per_pb, brawler, 'rank')\n\n rank_up_tokens = self.RANKS[str(rank)][\"PrimaryLvlUpRewardCount\"]\n\n token_doubler = await self.get_player_stat(user, 'token_doubler')\n\n upd_td = token_doubler - rank_up_tokens\n if upd_td < 0:\n upd_td = 0\n\n if token_doubler > rank_up_tokens:\n rank_up_tokens *= 2\n else:\n rank_up_tokens += token_doubler\n\n rank_up_starpoints = self.RANKS[str(rank)][\"SecondaryLvlUpRewardCount\"]\n\n await self.update_player_stat(\n user, 'tokens', rank_up_tokens, add_self=True)\n await self.update_player_stat(\n user, 'starpoints', rank_up_starpoints, add_self=True)\n await self.update_player_stat(\n user, 'token_doubler', upd_td)\n\n embed = discord.Embed(\n color=EMBED_COLOR,\n title=f\"Brawler Rank Up! {rank} → {rank_as_per_pb}\"\n )\n embed.set_author(name=user.name, icon_url=user.avatar_url)\n embed.add_field(\n name=\"Brawler\", value=f\"{brawler_emojis[brawler]} {brawler}\")\n embed.add_field(\n name=\"Tokens\", value=f\"{emojis['token']} {rank_up_tokens}\")\n if rank_up_starpoints:\n embed.add_field(\n name=\"Star Points\",\n value=f\"{emojis['starpoints']} {rank_up_starpoints}\"\n )\n if token_doubler > 0:\n embed.add_field(\n name=\"Token Doubler\",\n value=f\"{emojis['tokendoubler']} x{upd_td} remaining!\",\n inline=False\n )\n return embed","def test_happy_display_points_board_welcome(self):\n\n rv = self.login(\"john@simplylift.co\")\n\n assert rv.status_code in [200]\n assert b\"Logout\" in rv.data\n\n assert b\"Points Board\" in rv.data\n for club in self.clubs:\n assert str.encode(club[\"name\"]) in rv.data\n assert str.encode(f\"Current Points: {club['points']}\") in rv.data","def page_rank_score(self, rows):\n pageranks = dict([(row[0], self.con.execute('select score from pagerank where urlid=%d'\n % row[0]).fetchone()[0]) for row in rows])\n maxrank = max(pageranks.values())\n normalizedscores = dict([(u, float(l) / maxrank) for (u, l) in pageranks.items()])\n return normalizedscores","def test_recommendation_evaluation_17(model, interactions_ds):\n fun = lambda x: 1\n try:\n recommendation_evaluation(model, interactions_ds[1], n_test_users=None, k=5, n_pos_interactions=None,\n novelty=False, metrics=[fun], verbose=False)\n assert False\n except Exception as e:\n assert str(e) == f'Expected metric {fun} to be an instance of type RankingMetricABC.'","def learning_to_rank(self):\n self.error_throw('rank')\n\n instance = Instance(self.table_name)\n instance.addTable(Table_LTR(instance,False,'',''))\n if self.import_method == 'mysql': instance = self.mysql_handle(instance)\n elif self.import_method == 'csv': instance = self.csv_handle(instance)\n\n self.rank_learning(instance)\n\n self.rank_method = methods_of_ranking[1] # = 'learn_to_rank'","def getRank(self):\r\n return self.rank","def test_get_grid_rank(flopy_dis_mf6):\n flopy_dis, mf6 = flopy_dis_mf6\n mf6.initialize()\n\n if flopy_dis.nlay == 1:\n prescribed_grid_rank = 2\n else:\n prescribed_grid_rank = 3\n\n # Getting the grid id from the model, requires specifying one variable\n k11_tag = mf6.get_var_address(\"K11\", flopy_dis.model_name, \"NPF\")\n grid_id = mf6.get_var_grid(k11_tag)\n\n assert prescribed_grid_rank == mf6.get_grid_rank(grid_id)","def test_posts_page(self):\n\n with self.client:\n result = self.client.get('/posts')\n self.assertEqual(result.status_code, 200)\n self.assertIn(b'<h5 class=\"card-title title-preview\">test 1 title</h5>', result.data)","def stats_page():\n import alltheitems.stats\n return alltheitems.stats.index()","def isranked(self):\n rank = self.console.getCvar('sv_ranked')\n return rank and rank.getInt() == 2","def test(self,name,curr,expected):\n rank = mpicom.rank()\n res = \"passed\"\n if curr != expected:\n res = \"failed\"\n self.failures += 1\n rtxt = \"%d:%s=%s:%s\\n\" % (rank,name,repr(str(curr)),res)\n self.results += rtxt","def test_default_score_post(self):\n post = Post.objects.get(body=\"123ABC Body\")\n self.assertEqual(post.score, DEFAULT_SCORE)","def result_page():\n results = session.get('results')\n points = 0\n if results:\n for k,p in results.items():\n points += p['points']\n return render_template('results.html', results=results, points=points),","def top_ten(subreddit):\n url = \"https://www.reddit.com/r/\" + subreddit + \"/top/.json\"\n r = requests.get(url,\n headers={'User-agent': 'norman'},\n params={\"limit\": 10},\n allow_redirects=False)\n if r.status_code == 200:\n for dic in r.json().get('data').get('children'):\n print(dic.get('data').get('title'))\n else:\n print(\"None\")","def get_rank() -> int:\n return collective.get_rank()"],"string":"[\n \"def test_job_rank(self):\\n inv_search = 'rank:Postdoc'\\n spi_search = 'find rank Postdoc'\\n self._compare_searches(inv_search, spi_search)\",\n \"def test_something(self):\\n\\n allure.dynamic.title(\\\"Testing compute_ranks\\\")\\n allure.dynamic.severity(allure.severity_level.NORMAL)\\n allure.dynamic.description_html('<h3>Codewars badge:</h3>'\\n '<img src=\\\"https://www.codewars.com/users/myFirstCode'\\n '/badges/large\\\">'\\n '<h3>Test Description:</h3>'\\n \\\"<p>Test the function taht organizes a sports league in a \\\"\\n \\\"round-robin-system. Each team meets all other teams. \\\"\\n \\\"In your league a win gives a team 2 points, a draw gives \\\"\\n \\\"both teams 1 point. After some games you have to compute \\\"\\n \\\"the order of the teams in your league. You use the following \\\"\\n \\\"criteria to arrange the teams:</p>\\\"\\n \\\"<ul><li>- Points</li>\\\"\\n \\\"<li>- Scoring differential (the difference between goals \\\"\\n \\\"scored and those conceded)</li>\\\"\\n \\\"<li>- Goals scored</li></ul>\\\")\\n\\n test_data = [\\n (6,\\n [[0, 5, 2, 2],\\n [1, 4, 0, 2],\\n [2, 3, 1, 2],\\n [1, 5, 2, 2],\\n [2, 0, 1, 1],\\n [3, 4, 1, 1],\\n [2, 5, 0, 2],\\n [3, 1, 1, 1],\\n [4, 0, 2, 0]],\\n [4, 4, 6, 3, 1, 2]),\\n (6,\\n [[0, 5, 2, 0],\\n [1, 4, 2, 2],\\n [2, 3, 1, 3],\\n [1, 5, 0, 0],\\n [2, 0, 2, 1],\\n [3, 4, 3, 1]],\\n [2, 3, 4, 1, 5, 6]),\\n (4,\\n [[0, 3, 1, 1],\\n [1, 2, 2, 2],\\n [1, 3, 2, 0],\\n [2, 0, 2, 0]],\\n [3, 1, 1, 3]),\\n (10,\\n [],\\n [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]),\\n (8,\\n [[0, 7, 2, 0]],\\n [1, 2, 2, 2, 2, 2, 2, 8])\\n ]\\n\\n for data in test_data:\\n number = data[0]\\n games = data[1]\\n expected = data[2]\\n actual_result = compute_ranks(number, games)\\n print_log(number=number,\\n games=games,\\n expected=expected,\\n actual_result=actual_result)\\n\\n with allure.step(\\\"Enter a test data and verify the result:\\\"):\\n self.assertEqual(expected, actual_result)\",\n \"def test_rank(self):\\n self.assertEqual(self.vectors.rank('dog.n.01', 'dog.n.01'), 1)\\n self.assertEqual(self.vectors.rank('dog.n.01', 'carnivore.n.01'), 3)\",\n \"def test_get_ranking(self):\\n card = Card.objects.create(suit=Card.CLUB, rank=\\\"jack\\\")\\n self.assertEqual(card.get_ranking(), 11)\",\n \"def test_rank(self):\\n card = self._card\\n self.assertEqual(card.rank, self._rank)\",\n \"def test_get_rank_future(client_challenge_future):\\n retval = login(client_challenge_future,\\n \\\"cwhqsam\\\",\\n \\\"supersecurepassword\\\")\\n assert retval.status_code == 200\\n\\n retval = client_challenge_future.get(\\\"/api/v1/questions/rank\\\")\\n assert retval.status_code == 200\\n assert retval.get_json()[\\\"rank\\\"] == -1\",\n \"def which_ranking(self, request):\\n options = [\\\"user ranking\\\", \\\"team ranking\\\"]\\n anwser = main_menu(\\\"Ranking\\\", options,\\n \\\"Which ranking you want to see\\\")\\n if anwser == options[0]:\\n self.generate_rank(request, True)\\n else:\\n self.generate_rank(request, False)\",\n \"def test_post_rank(self):\\n # Upvote last post\\n post = self.posts[-1]\\n url = reverse_lazy('post-vote', kwargs={'pk': post.pk})\\n response = self.put_json(url)\\n\\n self.assertEqual(response.status_code, status.HTTP_200_OK)\\n\\n # Search first tag\\n url = reverse_lazy('tag-list') + '?search={}'.format(self.tags[0])\\n\\n response = self.client.get(url)\\n self.assertEqual(response.status_code, status.HTTP_200_OK)\\n\\n result = response.data['results'][0]\\n posts = result['posts']\\n\\n self.assertEqual(posts[0]['id'], post.pk)\",\n \"def rank():\\n return 0\",\n \"def test_ranking_achievements(self):\\n client = Client()\\n user = create_user('passwordmichu', 'michu')\\n client.login(username='michu', password='passwordmichu')\\n coll = create_collection('Coleccion de cartas')\\n problem = create_select_problem(coll, 'Problema')\\n create_an_achievement_of_each(coll)\\n submit_select_url = reverse('judge:submit', args=[problem.pk])\\n client.post(submit_select_url, {'code': problem.solution}, follow=True)\\n ranking_url = reverse('judge:result', args=[coll.pk])\\n group_a = create_group('1A')\\n group_a.user_set.add(user)\\n response = client.get(ranking_url, follow=True)\\n self.assertIn('x5', response.content.decode('utf-8'))\",\n \"def display_ranking(args: argparse.Namespace, cookie: str) -> None:\\n number: int = args.number\\n if not number:\\n return\\n\\n if number >= 50:\\n print(\\\"Number of users to display cannot exceed 49 for now. Exiting.\\\")\\n return\\n\\n cookies = {\\n \\\"spip_session\\\": cookie,\\n }\\n\\n ranking = get_global_ranking(args.offset, cookies)\\n\\n for counter, entry in enumerate(ranking):\\n\\n if counter >= number:\\n break\\n\\n userid = ranking[entry][\\\"id_auteur\\\"]\\n userinfo = get_user_info(userid, cookies)\\n if userinfo is None:\\n continue\\n\\n print(\\n f\\\"{Fore.YELLOW}{ranking[entry]['place']:>4}{Style.RESET_ALL}. \\\"\\n f\\\"{userinfo['nom']:<25} \\\"\\n f\\\"{userinfo['score']} \\\"\\n )\\n\\n return\",\n \"def test_get_rank_today(client_challenge_today):\\n retval = register(client_challenge_today, \\\"sam@codewizardshq.com\\\",\\n \\\"cwhqsam\\\", \\\"supersecurepassword\\\", \\\"Sam\\\", \\\"Hoffman\\\",\\n \\\"samuelhoffman2@gmail.com\\\")\\n assert retval.status_code == 200\\n\\n retval = login(client_challenge_today,\\n \\\"cwhqsam\\\",\\n \\\"supersecurepassword\\\")\\n assert retval.status_code == 200\\n\\n retval = client_challenge_today.get(\\\"/api/v1/questions/rank\\\")\\n assert retval.status_code == 200\\n assert retval.get_json()[\\\"rank\\\"] == 1\",\n \"def test_wikipedia_matches_ranked_higher_than_facebook(w_driver):\\n #1.) Do a Kasner search for some word\\n w_driver.get('localhost:8000')\\n query='sports'\\n element = w_driver.find_element_by_name('query')\\n element.send_keys(query + Keys.RETURN)\\n\\n #2.) Verify that wikipedia page is ranked higher than the facebook page.\\n results=w_driver.page_source\\n #results should list testsitewithnumber5.com before wikipedia result\\n text_wikip='wikipedia.org/wiki/'+query\\n #find the character that begins the reference to wikipedia page\\n matches_wikip = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)), results)]\\n #there are 2 references for testsitewithnumber6.com; we get the first.\\n matches_wikip=matches_wikip[0]\\n\\n text_fb='www.facebook.com/public/'+query\\n #find the character that begins the reference to facebook page\\n matches_fb = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_fb)), results)]\\n #there are 2 references for testsitewithnumber6.com; we get the first.\\n matches_fb=matches_fb[0]\\n\\n #wikipedia page should occur first; its reference number should be smaller\\n assert (matches_wikip<matches_fb)\",\n \"def browse_rank(posts, rank):\\n \\n if rank > len(posts):\\n click.secho(('\\\\n\\\\tCannot get that rank. Currently %d products have been hunted '\\n 'with Spear.\\\\n' % len(posts)), fg=\\\"red\\\")\\n return\\n\\n url = BASE_URL + posts[int(rank)-1]['permalink']\\n webbrowser.open(url, new=2)\",\n \"def test_lsr_rankings():\\n for case in iter_testcases('rankings'):\\n n_items = case[\\\"n_items\\\"]\\n data = case[\\\"data\\\"]\\n assert np.allclose(\\n case[\\\"lsr_est\\\"], lsr_rankings(n_items, data),\\n atol=ATOL, rtol=RTOL)\",\n \"def test_ilsr_rankings():\\n for case in iter_testcases('rankings'):\\n n_items = case[\\\"n_items\\\"]\\n data = case[\\\"data\\\"]\\n assert np.allclose(\\n case[\\\"ml_est\\\"], ilsr_rankings(n_items, data),\\n atol=ATOL, rtol=RTOL)\",\n \"def test_is_rank_integer(self):\\n self.assertIsInstance(cardutils.Card(10,1).rank, int)\",\n \"def test_home_pg(self):\\n\\n with self.client as client:\\n resp = client.get(\\\"/\\\")\\n html = resp.get_data(as_text=True)\\n\\n self.assertEqual(resp.status_code, 200)\\n # p1 should be top 3 in trending while p4 is not\\n self.assertIn(f'<a href=\\\"/post/{self.p1_id}\\\">', html)\\n self.assertNotIn(f'<a href=\\\"/post/{self.p4_id}\\\">', html)\\n # p5 is oldest post and should not appear\\n self.assertNotIn(f'<a href=\\\"/post/{self.p5_id}\\\">', html)\",\n \"def rank(request):\\n links = Link.objects.order_by('-visualization')[:5]\\n context = {'links' : links}\\n return render(request, 'encurtador/rank.html', context)\",\n \"def test_variant_rank_scores(case_obj, variant_obj):\\n\\n # GIVEN a case with SNV variants with rank score model\\n assert case_obj[\\\"rank_model_version\\\"]\\n # GIVEN a snv variant\\n assert variant_obj[\\\"category\\\"] == \\\"snv\\\"\\n # GIVEN that the variant has rank scores:\\n variant_obj[\\\"rank_score_results\\\"] = [\\n {\\\"category\\\": \\\"Splicing\\\", \\\"score\\\": 0},\\n {\\\"category\\\": \\\"Inheritance_Models\\\", \\\"score\\\": -12},\\n {\\\"category\\\": \\\"Consequence\\\", \\\"score\\\": 1},\\n ]\\n\\n # GIVEN a test app containing config params to retrieve a genetic model\\n test_app = create_app(\\n config=dict(\\n TESTING=True,\\n DEBUG=True,\\n MONGO_DBNAME=\\\"TEST_DB\\\",\\n DEBUG_TB_ENABLED=False,\\n LOGIN_DISABLED=True,\\n RANK_MODEL_LINK_PREFIX=\\\"https://raw.githubusercontent.com/Clinical-Genomics/reference-files/master/rare-disease/rank_model/rank_model_-v\\\",\\n RANK_MODEL_LINK_POSTFIX=\\\"-.ini\\\",\\n )\\n )\\n with test_app.app_context():\\n # THEN the rank score results of the variant should be returned by the function\\n rank_score_results = variant_rank_scores(store, case_obj, variant_obj)\\n assert isinstance(rank_score_results, list)\\n # WITH the relative model ranges values\\n assert rank_score_results[0][\\\"model_ranges\\\"]\",\n \"def calculate_page_rank(self, iterations=5):\\n # clear out the current page rank tables\\n self.con.execute('drop table if exists pagerank')\\n self.con.execute('create table pagerank(urlid primary key,score)')\\n\\n # initialize every url with a page rank of 1\\n for (urlid,) in self.con.execute('select rowid from urllist'):\\n self.con.execute('insert into pagerank(urlid,score) values (%d,1.0)' % urlid)\\n self.dbcommit()\\n\\n for i in range(iterations):\\n # Need multiple iterations, as the page ranks of pages linked to this\\n # one will be consistently updated on each iteration\\n print(\\\"Iteration %d\\\" % i)\\n for (urlid,) in self.con.execute('select rowid from urllist'):\\n # Default page rank\\n page_rank = 0.15\\n\\n # Loop through all the pages that link to this one\\n for (linker,) in self.con.execute('select distinct fromid from link where toid=%d'\\n % urlid):\\n # Get the page rank of the linker\\n linkingpr = self.con.execute('select score from pagerank where urlid=%d'\\n % linker).fetchone()[0]\\n\\n # Get the total number of links from the linker\\n linkingcount = self.con.execute('select count(*) from link where fromid=%d'\\n % linker).fetchone()[0]\\n # add to page rank, accounting for the link count\\n page_rank += 0.85 * (linkingpr / linkingcount)\\n self.con.execute('update pagerank set score=%f where urlid=%d'\\n % (page_rank, urlid))\\n self.dbcommit()\",\n \"def test_display_review(self):\\n\\n result = self.client.get(\\\"/brand/P87985432\\\")\\n self.assertIn(b\\\"ever ever\\\", result.data)\",\n \"def print_player_rank_and_points(self):\\r\\n pass\",\n \"def test_rank(rank_list, word, n, test='bw'):\\n tests = ['bw', 'ks']\\n if test not in test:\\n raise ValueError('Invalid test used. Need to choose either {} or {}\\\\\\n tests.'.format(*tests))\\n\\n _, base_check = zip(*sorted(rank_list[word]))\\n # just subtract the mean of the dataset, to make the mean 0\\n std_base_check = [data - np.mean(base_check) for data in base_check]\\n \\n print(std_base_check)\\n\\n words = []\\n for key, values in rank_list.items():\\n if key == word:\\n continue\\n\\n _, test_check = zip(*sorted(values))\\n # move the data to a mean of 0\\n std_test_check = [data - np.mean(test_check) for data in test_check]\\n if test == 'bw':\\n score = brendan_whitney_test(std_base_check, std_test_check)\\n else:\\n score, _ = ks_2samp(std_base_check, std_test_check)\\n \\n words.append((score, key))\\n \\n words = sorted(words)\\n print(words[:n])\\n return words[:n]\",\n \"def __ranking_function(self, doc, query_tokens):\",\n \"def test_mirror_ranking(self):\\n updater = AptMirrorUpdater()\\n # Make sure that multiple discovered mirrors are available.\\n assert sum(m.is_available for m in updater.ranked_mirrors) > 10\",\n \"def print_scores(ranked_list, links, topN=10):\\n n = topN\\n for url, score in ranked_list.items():\\n try:\\n _ = links.index(url)\\n print(\\\"Score: %f \\\\t URL: %s\\\" %(score, url))\\n\\n n -= 1\\n\\n if n <= 0:\\n break\\n except:\\n pass\",\n \"def test_get_game_boxscore(self):\\n pass\",\n \"def test_page_renders_with_correct_template(self):\\n p = Player()\\n p.save()\\n\\n c = Client()\\n response = c.get(reverse(self.test_url))\\n self.assertTemplateUsed(response, 'display_stats/display_stats.html')\",\n \"def test_score():\\n print(\\\"Tests for 'score' function\\\")\\n test_suite = TestSuite()\\n\\n # Testing with empty hand\\n result = score([])\\n test_suite.run_test(result, 0, '0')\\n # Testing with non-empty hand\\n result = score([1, 3])\\n test_suite.run_test(result, 3, '1')\\n # Testing with non-empty hand\\n result = score([1, 3, 1, 1])\\n test_suite.run_test(result, 3, '2')\\n # Testing with non-empty hand\\n result = score([4, 3, 4, 3, 3])\\n test_suite.run_test(result, 9, '3')\\n\\n # Show report\\n test_suite.report_results()\",\n \"def test_get_user_rankings(self):\\n user = User(name=u'no win', email=u'generic@thingy.com')\\n user.put()\\n \\n userone = User(name=u'one win', email=u'generic@thingy.com', total_played=1, wins=1)\\n userone.put()\\n \\n usertwo = User(name=u'two wins', email=u'generic@thingy.com', total_played=2, wins=1)\\n usertwo.put()\\n \\n user_rankings = users.get_user_rankings()\\n self.assertEquals(len(user_rankings), 2)\\n self.assertEquals(user_rankings.pop().key, usertwo.key)\\n self.assertEquals(user_rankings.pop().key, userone.key)\",\n \"def test_happy_display_points_board_index(self):\\n\\n rv = self.app.get(\\\"/\\\")\\n\\n assert rv.status_code in [200]\\n assert b\\\"Registration Portal\\\" in rv.data\\n\\n assert b\\\"Points Board\\\" in rv.data\\n for club in self.clubs:\\n assert str.encode(club[\\\"name\\\"]) in rv.data\\n assert str.encode(f\\\"Current Points: {club['points']}\\\") in rv.data\",\n \"def test_name_matches_ranked_higher_than_wikipedia(w_driver):\\n w_driver.get('localhost:8000/add_form')\\n #1.) Add a test website into db\\n name_element=w_driver.find_element_by_name('name')\\n name_element.send_keys('test site')\\n url_element=w_driver.find_element_by_name('url')\\n url_element.send_keys('www.testsitewithnumber5.com')\\n number_element=w_driver.find_element_by_name('number')\\n number_element.send_keys(5)\\n name_element.send_keys(Keys.RETURN)\\n\\n #2.) Do a Kasner search and verify that the website is ranked higher than \\n #the wikipedia result for the query.\\n w_driver.get('localhost:8000')\\n element = w_driver.find_element_by_name('query')\\n element.send_keys('test site' + Keys.RETURN)\\n\\n results=w_driver.page_source\\n #results should list testsitewithnumber5.com before wikipedia result\\n text5='www.testsitewithnumber5.com' \\n #find the character that begins the reference to testsitewithnumber6.com\\n matches5 = [matches.start() \\n for matches in re.finditer(r'{}'.format(re.escape(text5)), results)]\\n #there are 2 references for testsitewithnumber6.com; we get the first.\\n matches5=matches5[0]\\n\\n text_wikip='wikipedia.org/wiki/test'\\n #find the character that begins the reference to wikipedia site\\n matches_wikip = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)),results)]\\n #there are 2 references for the url; we get the first.\\n matches_wikip=matches_wikip[0]\\n\\n #matches5 should occur first, so its reference character should be smaller\\n assert (matches5<matches_wikip)\",\n \"def print_end_of_round_rankings(self):\\n print('\\\\nFINAL SCORES')\\n print('-'*12)\\n for k, v in enumerate(self.ranking_dict):\\n print(f\\\"{k+1} {v[0]}: {v[1]}\\\")\\n print('\\\\n')\",\n \"def test_get_roles_rank(self):\\n contrib_as = self.make_assignment(\\n self.project, self.user_bob, self.role_contributor\\n )\\n guest_as = self.make_assignment(\\n self.project, self.user_carol, self.role_guest\\n )\\n roles = self.project.get_roles()\\n self.assertIn(contrib_as, roles)\\n self.assertIn(guest_as, roles)\\n roles = self.project.get_roles(max_rank=30)\\n self.assertEqual(roles, [contrib_as, self.owner_as_cat])\\n roles = self.project.get_roles(max_rank=40)\\n self.assertEqual(roles, [contrib_as, guest_as, self.owner_as_cat])\\n roles = self.project.get_roles(min_rank=30)\\n self.assertEqual(roles, [contrib_as, guest_as])\\n roles = self.project.get_roles(min_rank=40)\\n self.assertEqual(roles, [guest_as])\\n roles = self.project.get_roles(min_rank=30, max_rank=30)\\n self.assertEqual(roles, [contrib_as])\",\n \"def rank(args: argparse.Namespace, cookie: str) -> None:\\n cookies = {\\n \\\"spip_session\\\": cookie,\\n }\\n\\n max_len_user = max((len(x) for x in args.user))\\n\\n for username in args.user:\\n\\n users: Dict[str, Searchresult] = search_user(username, cookies)\\n\\n userids = []\\n # Parse API search results in order to find the correct users\\n for user in users:\\n if users[user][\\\"nom\\\"] == username:\\n userids.append(users[user][\\\"id_auteur\\\"])\\n\\n if not userids:\\n raise ValueError(\\n (\\n \\\"User %s not found in 50 first search terms.\\\"\\n \\\" Did you mispell it? \\\"\\n ),\\n username,\\n )\\n\\n for userid in userids:\\n userinfo = get_user_info(userid, cookies)\\n if userinfo is None:\\n continue\\n\\n print(\\n f\\\"[{Fore.GREEN}+{Style.RESET_ALL}] \\\"\\n f\\\"{userinfo['nom']:<{max_len_user + 2}} \\\"\\n f\\\"Score: {userinfo['score']:<6} \\\"\\n f\\\"Position: {userinfo['position']}\\\"\\n )\",\n \"def diversified_ranking(self):\\n self.error_throw('rank')\\n instance = Instance(self.table_name)\\n instance.addTable(Table(instance,False,'','')) # 'False'->transformed '',''->no describe yet\\n if self.import_method == 'mysql': instance = self.mysql_handle(instance)\\n elif self.import_method == 'csv': instance = self.csv_handle(instance)\\n\\n self.rank_partial(instance)\\n\\n self.rank_method = methods_of_ranking[3] # = 'diversified_ranking'\",\n \"def get_num_hit_rank(boxes_truth, boxes_pred, rank):\\n\\n def is_hit(box_truth, box_pred):\\n return is_label_match_rank(box_truth, box_pred, rank)\\n\\n return get_num_hit(boxes_truth, boxes_pred, is_hit)\",\n \"def calc_rank(id=13197473):\\r\\n player_url = urllib.parse.urlparse(\\\"http://osu.ppy.sh/pages/include/profile-general.php?u=player_id&m=0\\\".replace('player_id', str(id)))\\r\\n page = urlopen(player_url.geturl())\\r\\n soup = BeautifulSoup(page, features=\\\"html.parser\\\")\\r\\n table_divs = soup.findAll('div', attrs={'class': 'profileStatLine'})\\r\\n\\r\\n import re\\r\\n pattern = '\\\\(#\\\\d*,*\\\\d+\\\\)'\\r\\n for div in table_divs:\\r\\n for childdiv in div.find_all('b'):\\r\\n result = re.search(pattern, str(childdiv.text))\\r\\n my_ranking = int(result.group(0).replace(',', '').replace(\\\"(#\\\", '').replace(\\\")\\\", ''))\\r\\n break\\r\\n break\\r\\n return my_ranking\",\n \"def test(\\n self, render=True, opponent=None, wormzero_player=None, num_tests=1, num_gpus=0\\n ):\\n opponent = opponent if opponent else self.config.opponent\\n wormzero_player = wormzero_player if wormzero_player else self.config.wormzero_player\\n self_play_worker = self_play.SelfPlay.options(\\n num_cpus=0, num_gpus=num_gpus,\\n ).remote(self.checkpoint, self.Game, self.config, np.random.randint(10000))\\n results = []\\n for i in range(num_tests):\\n print(f\\\"Testing {i+1}/{num_tests}\\\")\\n results.append(\\n ray.get(\\n self_play_worker.play_game.remote(\\n 0, 0, render, opponent, wormzero_player,\\n )\\n )\\n )\\n self_play_worker.close_game.remote()\\n\\n if len(self.config.players) == 1:\\n result = np.mean([sum(history.reward_history) for history in results])\\n else:\\n result = np.mean(\\n [\\n sum(\\n reward\\n for i, reward in enumerate(history.reward_history)\\n if history.to_play_history[i - 1] == wormzero_player\\n )\\n for history in results\\n ]\\n )\\n return result\",\n \"def __rank__(self) -> int:\",\n \"def get_ranks(d): \\n raise NotImplementedError(\\\"Problem 3 Incomplete\\\")\",\n \"def get_rank(self) -> int:\\r\\n return self.rank\",\n \"def test_verify_non_existing_rank(self):\\n data = {'rank_id': 5}\\n res = self.post(url='/verify/4', data=data, role='admin')\\n self.assertEqual(res.status_code, 401)\\n self.assertException(res, exc.EntryNotFound)\",\n \"def get_rank(self):\\n details = self._tab.find(\\\"table\\\", class_=\\\"details\\\")\\n rank, = self.rank_re.match(details.find(\\\"td\\\", class_=\\\"value\\\").get_text()).groups()\\n return rank\",\n \"def test_profile_page_returns_correct_html(self):\\n self.add_testuser()\\n response = self.client.get('/profile/testuser/')\\n self.assertContains(response, 'Album Count')\",\n \"def test_03_leaderboard(self):\\r\\n # As Anonymou user\\r\\n url = \\\"/leaderboard\\\"\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should not be shown to anonymous users\\\"\\r\\n assert dom.find(id='enforce_privacy') is not None, err_msg\\r\\n # As Authenticated user but NOT ADMIN\\r\\n self.signin()\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should not be shown to authenticated users\\\"\\r\\n assert dom.find(id='enforce_privacy') is not None, err_msg\\r\\n self.signout\\r\\n # As Authenticated user but ADMIN\\r\\n self.signin(email=self.root_addr, password=self.root_password)\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should be shown to admin users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout()\",\n \"def test_figure1(self):\\n\\n searcher = SimpleSearcher.from_prebuilt_index('msmarco-passage')\\n hits = searcher.search('what is a lobster roll?', 10)\\n\\n self.assertAlmostEqual(hits[0].score, 11.00830, delta=0.0001)\\n self.assertEqual(hits[0].docid, '7157707')\\n\\n self.assertAlmostEqual(hits[9].score, 9.92200, delta=0.0001)\\n self.assertEqual(hits[9].docid, '6234461')\\n\\n self.assertEqual(len(hits), 10)\",\n \"def test_03_leaderboard(self):\\r\\n # As Anonymou user\\r\\n url = \\\"/leaderboard\\\"\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should be shown to anonymous users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n # As Authenticated user but NOT ADMIN\\r\\n self.signin()\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should be shown to authenticated users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout\\r\\n # As Authenticated user but ADMIN\\r\\n self.signin(email=self.root_addr, password=self.root_password)\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Leaderboard page should be shown to admin users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout()\",\n \"def get_rank(self):\\r\\n return self.rank\",\n \"def test_pages_with_same_name_ranked_by_number(w_driver):\\n w_driver.get('localhost:8000/add_form')\\n #1.) Add two websites with same name, but different number into db\\n #First site, with number=5\\n name_element=w_driver.find_element_by_name('name')\\n name_element.send_keys('test site')\\n url_element=w_driver.find_element_by_name('url')\\n url_element.send_keys('www.testsitewithnumber5.com')\\n number_element=w_driver.find_element_by_name('number')\\n number_element.send_keys(5)\\n name_element.send_keys(Keys.RETURN)\\n\\n #Second site, with number=6\\n name_element=w_driver.find_element_by_name('name')\\n name_element.send_keys('test site')\\n url_element=w_driver.find_element_by_name('url')\\n url_element.send_keys('www.testsitewithnumber6.com')\\n number_element=w_driver.find_element_by_name('number')\\n number_element.send_keys(6)\\n name_element.send_keys(Keys.RETURN)\\n\\n #2.) Do a Kasner search and verify that the websites are sorted in \\n # descending order in the search results.\\n w_driver.get('localhost:8000')\\n element = w_driver.find_element_by_name('query')\\n element.send_keys('test site' + Keys.RETURN)\\n\\n results=w_driver.page_source\\n #results should list testsitewithnumber6.com before testsitewithnumber5.com\\n text6='www.testsitewithnumber6.com' \\n #find the character that begins the reference to testsitewithnumber6.com\\n matches6 = [matches.start() \\n for matches in re.finditer(r'{}'.format(re.escape(text6)), results)]\\n #there are 2 references for testsitewithnumber6.com; we get the first.\\n matches6=matches6[0]\\n\\n #find the character that begins the reference to testsitewithnumber5.com\\n text5='www.testsitewithnumber5.com' \\n matches5 = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text5)), results)]\\n #there are 2 references for testsitewithnumber5.com. we get the first.\\n matches5=matches5[0]\\n\\n #matches6 should occur first, so its reference character should be smaller\\n assert (matches6<matches5)\",\n \"def article_rank():\\n db, c = connect(DBNAME)\\n c.execute(\\\"select title, count(title) as views from \\\\\\\"pathslug\\\\\\\" \\\"\\n \\\"group by title order by views desc limit 3\\\")\\n article_table = c.fetchall()\\n db.close()\\n print \\\"\\\\nThree Most Popular Articles All Time:\\\"\\n for article in article_table:\\n print str(article[0]) + \\\" - \\\" + str(article[1]) + \\\" views\\\"\",\n \"def test_calculate_number_page(self):\\n self.requestapi = RequestApi('snack')\\n result = self.requestapi.calculate_number_page()\\n self.assertEqual(result, 1)\",\n \"def report(results, n_top=1):\\n for i in range(1, n_top + 1):\\n candidates = np.flatnonzero(results['rank_test_score'] == i)\\n for candidate in candidates:\\n print(f\\\"Model with rank: {i}\\\")\\n print(f\\\"Mean validation score: {results['mean_test_score'][candidate]} (std: {results['std_test_score'][candidate]}\\\")\\n print(f\\\"Parameters: {results['params'][candidate]}\\\")\",\n \"def print_top_pageranks(topn):\\n print('Top page ranks:\\\\n%s' % ('\\\\n'.join('%s\\\\t%.5f' % (u, v) for u, v in topn)))\",\n \"def test_scrape(self):\\n self.assertEqual(self.scraped.title, 'Heading!')\\n self.assertEqual(self.scraped.link_text, 'Go to Google')\\n self.assertEqual(self.scraped.link_url, 'http://Google.com')\",\n \"def rank():\\n return int(os.environ['RANK'])\",\n \"def testHighscore(self):\\n \\n game = Game.objects.get(title='game0')\\n gameplayeds = game.gameplayed_set\\n ply_group = Group.objects.get(name='Player')\\n for i in range(4):\\n user = ply_group.user_set.get(username='ply{}'.format(i))\\n gameplayed = gameplayeds.get(user=user)\\n gameplayed.gameScore = i\\n gameplayed.save()\\n \\n response = self.client.get(\\n reverse('api:game-buyers', args=['v1', 'game0']),\\n {'order_by': 'gameScore'},\\n format='json'\\n )\\n self.assertEquals(response.status_code, 200)\\n content = self.parser.parse(BytesIO(response.content))\\n for i in range(4):\\n self.assertEquals(content['results'][i]['user'], 'ply{}'.format(i))\\n \\n response = self.client.get(\\n reverse('api:game-buyers', args=['v1', 'game0']),\\n {'order_by': '-gameScore'},\\n format='json'\\n )\\n self.assertEquals(response.status_code, 200)\\n content = self.parser.parse(BytesIO(response.content))\\n for i in range(4):\\n self.assertEquals(content['results'][i]['user'], 'ply{}'.format(3 - i))\",\n \"def test_page(self):\\n result = self.test_client.page\\n\\n assert result == 1\",\n \"def test_rank():\\n keys = list(range(100))\\n d = OrderedTreeDict((key, None) for key in keys)\\n assert all(list(\\n keys[d.rank(k)] == k for k in keys\\n ))\",\n \"def test_summary_with_results(self):\\n add_constituency_result_line('X, 10, C')\\n r = self.client.get('/summary')\\n self.assertEqual(r.status_code, 200)\\n soup = BeautifulSoup(r.data, 'html.parser')\\n self.assertIs(soup.find(id='no-results'), None)\\n self.assertIsNot(soup.find(id='results-table'), None)\",\n \"def test_get_score(self):\\r\\n score_dict = self.combinedoe.get_score()\\r\\n self.assertEqual(score_dict['score'], 0)\\r\\n self.assertEqual(score_dict['total'], 1)\",\n \"def evaluate_ranks(articles, rank_tuples):\\n rank_tp=defaultdict(int)\\n rank_fn=defaultdict(int)\\n rank_fp=defaultdict(int)\\n \\n for article in articles:\\n for mention in article.entity_mentions:\\n form=mention.mention\\n meaning=mention.gold_link\\n sys_meaning=mention.sys_link\\n t_gold=(form, meaning)\\n t_sys=(form, sys_meaning)\\n for rank, r_tuples in rank_tuples.items():\\n if t_gold in r_tuples and t_sys in r_tuples:\\n rank_tp[rank]+=1\\n break\\n elif t_gold in r_tuples:\\n rank_fn[rank]+=1\\n elif t_sys in r_tuples:\\n rank_fp[rank]+=1\\n print('tp', rank_tp)\\n print('fp', rank_fp)\\n print('fn', rank_fn)\\n \\n rank_prec={}\\n rank_recall={}\\n rank_f1={}\\n \\n for rank in range(1,13):\\n if rank_tp[rank]+rank_fp[rank]>0:\\n rank_prec[rank]=rank_tp[rank]/(rank_tp[rank]+rank_fp[rank])\\n else:\\n rank_prec[rank]=0.0\\n if rank_tp[rank]+rank_fn[rank]>0:\\n rank_recall[rank]=rank_tp[rank]/(rank_tp[rank]+rank_fn[rank])\\n else:\\n rank_recall[rank]=0.0\\n if rank_prec[rank]+rank_recall[rank]>0:\\n rank_f1[rank]=2*rank_prec[rank]*rank_recall[rank]/(rank_prec[rank]+rank_recall[rank])\\n else:\\n rank_f1[rank]=0.0\\n print('precision', rank_prec)\\n print()\\n print('recall', rank_recall)\\n print()\\n print('f1', rank_f1)\\n print()\\n return rank_prec, rank_recall, rank_f1\",\n \"def alexa_rank(url):\\n return ALEXA_MAP.index(url)\",\n \"def test_index_page(self):\\n response = self.client.get(\\\"\\\")\\n self.assertEqual(response.status_code, 200)\\n self.assertTemplateUsed(response, \\\"search/base.html\\\")\\n self.assertTemplateUsed(response, \\\"search/search_form.html\\\")\\n self.assertTemplateUsed(response, \\\"search/index.html\\\")\",\n \"def test_scores(self):\\n pig = game.pig.Pig('PlayerA', 'PlayerB', 'PlayerC')\\n self.assertEqual(\\n pig.get_score(),\\n {\\n 'PlayerA': 0,\\n 'PlayerB': 0,\\n 'PlayerC': 0\\n }\\n )\",\n \"def score(self, test_data):\\n\\n\\t\\tpass\",\n \"def testGetsPresentRichness(self):\\n self.assertEqual(37, self.coal.get_species_richness(1))\\n self.assertEqual(53, self.coal.get_species_richness(3))\",\n \"def rerank_candidates(s, pred2sub_rank, all_predictions, rerank_top=20):\\n predicted_smiles = []\\n model_input = []\\n for (predict_smi, label), _ in Counter(all_predictions).most_common(rerank_top):\\n if predict_smi == s:\\n continue\\n features = get_all_features(\\n get_all_ranking_info(pred2sub_rank[predict_smi]))\\n predicted_smiles.append((predict_smi, label))\\n model_input.append(features)\\n\\n model = RankingModel()\\n model.load_state_dict(torch.load('./models/ranker/rank_model.pt', map_location='cpu'))\\n model.eval()\\n\\n test_loader = DataLoader(RankingTestDataset(\\n model_input), batch_size=1000, shuffle=False, num_workers=2)\\n ranking_scores = []\\n for data in test_loader:\\n outputs = model(data)[0]\\n ranking_scores.extend(outputs.detach().cpu().numpy())\\n\\n assert len(predicted_smiles) == len(ranking_scores)\\n pred_smi2score = {k: v[1]\\n for k, v in zip(predicted_smiles, ranking_scores)}\\n return pred_smi2score\",\n \"def rank(self):\\n return self.lib.calculate_rank()\",\n \"def test_default_page_size(self):\\n response = self.client.get('/')\\n self.assertEqual(10, len(response.data['results']))\",\n \"def test_wikipedia_matches_ranked_higher_than_keyword_matches(w_driver):\\n keyword='test site'\\n w_driver.get('localhost:8000/add_form')\\n \\n #1.) Add 2 websites into db with keyword=test site\\n name_element1=w_driver.find_element_by_name('name')\\n name_element1.send_keys('first one')\\n url_element1=w_driver.find_element_by_name('url')\\n url_element1.send_keys('www.firstsite5.com')\\n number_element1=w_driver.find_element_by_name('number')\\n number_element1.send_keys(5)\\n words_element1=w_driver.find_element_by_name('words')\\n words_element1.send_keys(keyword)\\n words_element1.send_keys(Keys.RETURN)\\n\\n name_element2=w_driver.find_element_by_name('name')\\n name_element2.send_keys('first one')\\n url_element2=w_driver.find_element_by_name('url')\\n url_element2.send_keys('www.secondsite5.com')\\n number_element2=w_driver.find_element_by_name('number')\\n number_element2.send_keys(5)\\n words_element2=w_driver.find_element_by_name('words')\\n words_element2.send_keys(keyword)\\n words_element2.send_keys(Keys.RETURN)\\n\\n #2.) Do a Kasner search and verify that the websites with keyword=tests is\\n #ranked higher than the wikipedia site.\\n w_driver.get('localhost:8000')\\n element = w_driver.find_element_by_name('query')\\n element.send_keys(keyword + Keys.RETURN)\\n\\n results=w_driver.page_source\\n #results should list wikipedia site before keyword match results\\n text_wikip='wikipedia.org/wiki/'+keyword\\n #find the character that begins the reference to wikipedia site \\n matches_wikip = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_wikip)), results)]\\n #there are 2 references for wikipedia site; we get the first.\\n matches_wikip=matches_wikip[0]\\n\\n text_kw_1='www.firstsite5.com'\\n #find the character that begins the reference to www.firstsite5.com\\n matches_kw_1 = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_kw_1)),results)]\\n #there are 2 references for the url; we get the first.\\n matches_kw_1=matches_kw_1[0]\\n\\n text_kw_2='www.secondsite5.com'\\n #find the character that begins the reference to www.firstsite5.com\\n matches_kw_2 = [matches.start()\\n for matches in re.finditer(r'{}'.format(re.escape(text_kw_2)),results)]\\n #there are 2 references for the url; we get the first.\\n matches_kw_2=matches_kw_2[0]\\n \\n assert (matches_wikip<matches_kw_1)\\n assert (matches_wikip<matches_kw_2)\",\n \"def index(self):\\n\\t\\t# Return a rendered template\\n\\t\\t# return render('/some/template.mako')\\n\\t\\t# or, Return a response\\n\\t\\tScoreRankHistogramType = model.Stock_250kDB.ScoreRankHistogramType\\n\\t\\trows = ScoreRankHistogramType.query.order_by(ScoreRankHistogramType.results_type).\\\\\\n\\t\\t\\torder_by(ScoreRankHistogramType.null_distribution_type_id).\\\\\\n\\t\\t\\torder_by(ScoreRankHistogramType.min_distance).order_by(ScoreRankHistogramType.get_closest).\\\\\\n\\t\\t\\torder_by(ScoreRankHistogramType.min_MAF).order_by(ScoreRankHistogramType.allow_two_sample_overlapping).all()\\n\\t\\tc.score_rank_hist_type_ls = rows\\n\\t\\treturn render('/score_rank_histgram.html')\",\n \"def test_quiz_statistics_view(self):\\n target_url = url_for('quiz.quiz_statistics')\\n redirect_url = url_for('users.login', next=target_url)\\n response = self.client.get(target_url)\\n self.assertEqual(response.status_code, 302)\\n self.assertRedirects(response, redirect_url)\",\n \"def test_github_explore(self):\\n python_trending_url = \\\"https://github.com/trending/python\\\"\\n driver = self.driver\\n driver.get(self.base_url)\\n search_box = driver.find_element_by_link_text(\\\"Explore\\\")\\n search_box.send_keys(Keys.RETURN)\\n time.sleep(2)\\n trending_elem = driver.find_element_by_partial_link_text(\\\"Trending\\\")\\n trending_elem.send_keys(Keys.ENTER)\\n time.sleep(2)\\n python_lang = driver.find_element_by_link_text(\\\"Python\\\")\\n python_lang.send_keys(Keys.ENTER)\\n time.sleep(2)\\n assert python_trending_url == driver.current_url\",\n \"def test_homepage(self):\\n\\n response = self.client.get(\\\"/\\\")\\n self.assertIn(\\\"Books</title>\\\", response.data)\\n self.assertIn(\\\"Goodreads ID\\\", response.data)\",\n \"def get_ranking(self,\\n query=\\\"\\\",\\n number=10,\\n use_SVD=True #if false: VSM based similarity used\\n ):\\n if self.is_valid:\\n print(\\\"Ranking - Query: \\\", query)\\n\\n # Preprocess string: Convert into bag of words\\n bow = fp.string_transformation(query)\\n # bow = ['underground','underwater','wireless','communications']\\n # Doc sci.electronics/52434\\n print(\\\"Ranking - BoW: \\\", bow)\\n\\n # Convert into term vector\\n term_vec = self.tfidf_handler.convertBoWToTermVector(bow)\\n print(\\\"Ranking - Term Vector: Non-Zero positions:\\\", term_vec.nonzero())\\n print(\\\"Ranking - Term Vector: Non-Zero values:\\\", term_vec[term_vec.nonzero()])\\n print(\\\"Ranking - Term Vector: Size:\\\", term_vec.size)\\n\\n if(use_SVD):\\n # Calc. LSI ranking (SVD)\\n doc_ind, similarities = self.svd_handler.get_ranking(vec=term_vec, number=number)\\n else:\\n # Calc. VSM ranking\\n train_dtm = self.tfidf_handler.getTDM().transpose().tocsr()\\n norms = norm(train_dtm, axis=1) * np.linalg.norm(term_vec) # Normalize\\n sims = np.array(train_dtm.dot(term_vec) / norms)\\n\\n # Derive top-n\\n topn = np.argsort(sims)[::-1][:number]\\n similarities = sims[topn]\\n doc_ind = topn\\n\\n return self.tfidf_handler.getDocuments(doc_ind), similarities\\n else:\\n print(\\\"Ranking: Error - Instantiation of LSI failed\\\")\\n return [\\\"\\\"], [0]\",\n \"def test_index(self):\\n result = self.client.get('/')\\n self.assertEqual(result.status, '200 OK')\\n self.assertIn(b'Game', result.data)\",\n \"def rank(ctx, path, metric, revision, limit, desc, threshold, wrap):\\n config = ctx.obj[\\\"CONFIG\\\"]\\n\\n if not exists(config):\\n handle_no_cache(ctx)\\n\\n from wily.commands.rank import rank\\n\\n logger.debug(f\\\"Running rank on {path} for metric {metric} and revision {revision}\\\")\\n rank(\\n config=config,\\n path=path,\\n metric=metric,\\n revision_index=revision,\\n limit=limit,\\n threshold=threshold,\\n descending=desc,\\n wrap=wrap,\\n )\",\n \"def test_04_global_stats_index(self):\\r\\n # As Anonymou user\\r\\n url = \\\"/stats\\\"\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should not be shown to anonymous users\\\"\\r\\n assert dom.find(id='enforce_privacy') is not None, err_msg\\r\\n # As Authenticated user but NOT ADMIN\\r\\n self.signin()\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should not be shown to authenticated users\\\"\\r\\n assert dom.find(id='enforce_privacy') is not None, err_msg\\r\\n self.signout\\r\\n # As Authenticated user but ADMIN\\r\\n self.signin(email=self.root_addr, password=self.root_password)\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should be shown to admin users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout()\",\n \"def test_get_score(self):\\r\\n score_dict = self.combinedoe.get_score()\\r\\n self.assertEqual(score_dict['score'], 15.0)\\r\\n self.assertEqual(score_dict['total'], 5.0)\",\n \"def assignRanks(self):\\r\\n\\t\\trank = 0\\r\\n\\t\\tscores = list(self._playerScores)\\r\\n\\t\\tscores.reverse()\\r\\n\\t\\tfor playerScore in scores:\\r\\n\\t\\t\\tif not playerScore.has(NOT_MET) or not playerScore.value(NOT_MET):\\r\\n\\t\\t\\t\\trank += 1\\r\\n\\t\\t\\t\\tplayerScore.set(RANK, smallText(BugUtil.colorText(u\\\"%d\\\" % rank, ScoreOpt.getRankColor())))\\r\\n\\t\\tif rank > 0:\\r\\n\\t\\t\\tself._anyHas[RANK] = True\",\n \"def disp_score():\",\n \"def getPageRank(elistPath, alpha, maxiter, tolerance):\\n\\n adjGraph = AdjGraph(elistPath, separator=\\\" \\\")\\n graph = adjGraph.SNAPGraph\\n\\n preference_vector = []\\n for node in graph.Nodes():\\n id = node.GetId()\\n if (id % 4) == 0:\\n preference_vector.append(id)\\n\\n pageRank, convIter, time = biasedPageRank(\\n adjGraph, preference_vector=preference_vector, alpha=alpha,\\n max_iterations=maxiter, tolerance=tolerance)\\n\\n writeCentrality(\\\"pagerank.txt\\\", pageRank)\\n return pageRank, convIter, time\",\n \"def test_04_global_stats_index(self):\\r\\n # As Anonymou user\\r\\n url = \\\"/stats\\\"\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should be shown to anonymous users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n # As Authenticated user but NOT ADMIN\\r\\n self.signin()\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should be shown to authenticated users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout\\r\\n # As Authenticated user but ADMIN\\r\\n self.signin(email=self.root_addr, password=self.root_password)\\r\\n res = self.app.get(url, follow_redirects=True)\\r\\n dom = BeautifulSoup(res.data)\\r\\n err_msg = \\\"Stats page should be shown to admin users\\\"\\r\\n assert dom.find(id='enforce_privacy') is None, err_msg\\r\\n self.signout()\",\n \"async def handle_rank_ups(self, user: discord.User, brawler: str):\\n\\n brawler_data = await self.get_player_stat(\\n user, 'brawlers', is_iter=True, substat=brawler)\\n\\n pb = brawler_data['pb']\\n rank = brawler_data['rank']\\n\\n rank_as_per_pb = self.get_rank(pb)\\n\\n if rank_as_per_pb <= rank:\\n return False\\n\\n await self.update_player_stat(\\n user, 'brawlers', rank_as_per_pb, brawler, 'rank')\\n\\n rank_up_tokens = self.RANKS[str(rank)][\\\"PrimaryLvlUpRewardCount\\\"]\\n\\n token_doubler = await self.get_player_stat(user, 'token_doubler')\\n\\n upd_td = token_doubler - rank_up_tokens\\n if upd_td < 0:\\n upd_td = 0\\n\\n if token_doubler > rank_up_tokens:\\n rank_up_tokens *= 2\\n else:\\n rank_up_tokens += token_doubler\\n\\n rank_up_starpoints = self.RANKS[str(rank)][\\\"SecondaryLvlUpRewardCount\\\"]\\n\\n await self.update_player_stat(\\n user, 'tokens', rank_up_tokens, add_self=True)\\n await self.update_player_stat(\\n user, 'starpoints', rank_up_starpoints, add_self=True)\\n await self.update_player_stat(\\n user, 'token_doubler', upd_td)\\n\\n embed = discord.Embed(\\n color=EMBED_COLOR,\\n title=f\\\"Brawler Rank Up! {rank} → {rank_as_per_pb}\\\"\\n )\\n embed.set_author(name=user.name, icon_url=user.avatar_url)\\n embed.add_field(\\n name=\\\"Brawler\\\", value=f\\\"{brawler_emojis[brawler]} {brawler}\\\")\\n embed.add_field(\\n name=\\\"Tokens\\\", value=f\\\"{emojis['token']} {rank_up_tokens}\\\")\\n if rank_up_starpoints:\\n embed.add_field(\\n name=\\\"Star Points\\\",\\n value=f\\\"{emojis['starpoints']} {rank_up_starpoints}\\\"\\n )\\n if token_doubler > 0:\\n embed.add_field(\\n name=\\\"Token Doubler\\\",\\n value=f\\\"{emojis['tokendoubler']} x{upd_td} remaining!\\\",\\n inline=False\\n )\\n return embed\",\n \"def test_happy_display_points_board_welcome(self):\\n\\n rv = self.login(\\\"john@simplylift.co\\\")\\n\\n assert rv.status_code in [200]\\n assert b\\\"Logout\\\" in rv.data\\n\\n assert b\\\"Points Board\\\" in rv.data\\n for club in self.clubs:\\n assert str.encode(club[\\\"name\\\"]) in rv.data\\n assert str.encode(f\\\"Current Points: {club['points']}\\\") in rv.data\",\n \"def page_rank_score(self, rows):\\n pageranks = dict([(row[0], self.con.execute('select score from pagerank where urlid=%d'\\n % row[0]).fetchone()[0]) for row in rows])\\n maxrank = max(pageranks.values())\\n normalizedscores = dict([(u, float(l) / maxrank) for (u, l) in pageranks.items()])\\n return normalizedscores\",\n \"def test_recommendation_evaluation_17(model, interactions_ds):\\n fun = lambda x: 1\\n try:\\n recommendation_evaluation(model, interactions_ds[1], n_test_users=None, k=5, n_pos_interactions=None,\\n novelty=False, metrics=[fun], verbose=False)\\n assert False\\n except Exception as e:\\n assert str(e) == f'Expected metric {fun} to be an instance of type RankingMetricABC.'\",\n \"def learning_to_rank(self):\\n self.error_throw('rank')\\n\\n instance = Instance(self.table_name)\\n instance.addTable(Table_LTR(instance,False,'',''))\\n if self.import_method == 'mysql': instance = self.mysql_handle(instance)\\n elif self.import_method == 'csv': instance = self.csv_handle(instance)\\n\\n self.rank_learning(instance)\\n\\n self.rank_method = methods_of_ranking[1] # = 'learn_to_rank'\",\n \"def getRank(self):\\r\\n return self.rank\",\n \"def test_get_grid_rank(flopy_dis_mf6):\\n flopy_dis, mf6 = flopy_dis_mf6\\n mf6.initialize()\\n\\n if flopy_dis.nlay == 1:\\n prescribed_grid_rank = 2\\n else:\\n prescribed_grid_rank = 3\\n\\n # Getting the grid id from the model, requires specifying one variable\\n k11_tag = mf6.get_var_address(\\\"K11\\\", flopy_dis.model_name, \\\"NPF\\\")\\n grid_id = mf6.get_var_grid(k11_tag)\\n\\n assert prescribed_grid_rank == mf6.get_grid_rank(grid_id)\",\n \"def test_posts_page(self):\\n\\n with self.client:\\n result = self.client.get('/posts')\\n self.assertEqual(result.status_code, 200)\\n self.assertIn(b'<h5 class=\\\"card-title title-preview\\\">test 1 title</h5>', result.data)\",\n \"def stats_page():\\n import alltheitems.stats\\n return alltheitems.stats.index()\",\n \"def isranked(self):\\n rank = self.console.getCvar('sv_ranked')\\n return rank and rank.getInt() == 2\",\n \"def test(self,name,curr,expected):\\n rank = mpicom.rank()\\n res = \\\"passed\\\"\\n if curr != expected:\\n res = \\\"failed\\\"\\n self.failures += 1\\n rtxt = \\\"%d:%s=%s:%s\\\\n\\\" % (rank,name,repr(str(curr)),res)\\n self.results += rtxt\",\n \"def test_default_score_post(self):\\n post = Post.objects.get(body=\\\"123ABC Body\\\")\\n self.assertEqual(post.score, DEFAULT_SCORE)\",\n \"def result_page():\\n results = session.get('results')\\n points = 0\\n if results:\\n for k,p in results.items():\\n points += p['points']\\n return render_template('results.html', results=results, points=points),\",\n \"def top_ten(subreddit):\\n url = \\\"https://www.reddit.com/r/\\\" + subreddit + \\\"/top/.json\\\"\\n r = requests.get(url,\\n headers={'User-agent': 'norman'},\\n params={\\\"limit\\\": 10},\\n allow_redirects=False)\\n if r.status_code == 200:\\n for dic in r.json().get('data').get('children'):\\n print(dic.get('data').get('title'))\\n else:\\n print(\\\"None\\\")\",\n \"def get_rank() -> int:\\n return collective.get_rank()\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.71264744","0.6681979","0.6591885","0.6474593","0.6324985","0.6264448","0.6260697","0.6223638","0.61334574","0.61309284","0.6111487","0.61088467","0.60755545","0.6048969","0.6043513","0.5932428","0.5879655","0.58348274","0.58152354","0.5808058","0.56979764","0.56732076","0.5644784","0.5640055","0.5635797","0.56289405","0.5628108","0.5608872","0.56008327","0.5573004","0.55646855","0.55630946","0.55615556","0.55521214","0.55427086","0.55398345","0.55301064","0.5527893","0.55180913","0.5489926","0.54554385","0.5451728","0.5425614","0.5417755","0.5417164","0.54150367","0.53946376","0.53910685","0.5387045","0.5384161","0.5381905","0.53721255","0.5363917","0.536197","0.5353352","0.5352088","0.5351602","0.5346537","0.53399444","0.53370047","0.53276","0.53219384","0.53165233","0.5313955","0.5310923","0.5304398","0.53039545","0.5293326","0.52894026","0.5288893","0.5286916","0.5285984","0.52835995","0.5274037","0.527331","0.5265363","0.52509785","0.52493525","0.5223505","0.5221242","0.5219821","0.5211456","0.5211308","0.5208754","0.5206316","0.5205448","0.5202285","0.5191759","0.5189071","0.5186799","0.5185746","0.51844436","0.5183619","0.5181014","0.51745176","0.5172385","0.5172032","0.516866","0.5167372","0.5166256"],"string":"[\n \"0.71264744\",\n \"0.6681979\",\n \"0.6591885\",\n \"0.6474593\",\n \"0.6324985\",\n \"0.6264448\",\n \"0.6260697\",\n \"0.6223638\",\n \"0.61334574\",\n \"0.61309284\",\n \"0.6111487\",\n \"0.61088467\",\n \"0.60755545\",\n \"0.6048969\",\n \"0.6043513\",\n \"0.5932428\",\n \"0.5879655\",\n \"0.58348274\",\n \"0.58152354\",\n \"0.5808058\",\n \"0.56979764\",\n \"0.56732076\",\n \"0.5644784\",\n \"0.5640055\",\n \"0.5635797\",\n \"0.56289405\",\n \"0.5628108\",\n \"0.5608872\",\n \"0.56008327\",\n \"0.5573004\",\n \"0.55646855\",\n \"0.55630946\",\n \"0.55615556\",\n \"0.55521214\",\n \"0.55427086\",\n \"0.55398345\",\n \"0.55301064\",\n \"0.5527893\",\n \"0.55180913\",\n \"0.5489926\",\n \"0.54554385\",\n \"0.5451728\",\n \"0.5425614\",\n \"0.5417755\",\n \"0.5417164\",\n \"0.54150367\",\n \"0.53946376\",\n \"0.53910685\",\n \"0.5387045\",\n \"0.5384161\",\n \"0.5381905\",\n \"0.53721255\",\n \"0.5363917\",\n \"0.536197\",\n \"0.5353352\",\n \"0.5352088\",\n \"0.5351602\",\n \"0.5346537\",\n \"0.53399444\",\n \"0.53370047\",\n \"0.53276\",\n \"0.53219384\",\n \"0.53165233\",\n \"0.5313955\",\n \"0.5310923\",\n \"0.5304398\",\n \"0.53039545\",\n \"0.5293326\",\n \"0.52894026\",\n \"0.5288893\",\n \"0.5286916\",\n \"0.5285984\",\n \"0.52835995\",\n \"0.5274037\",\n \"0.527331\",\n \"0.5265363\",\n \"0.52509785\",\n \"0.52493525\",\n \"0.5223505\",\n \"0.5221242\",\n \"0.5219821\",\n \"0.5211456\",\n \"0.5211308\",\n \"0.5208754\",\n \"0.5206316\",\n \"0.5205448\",\n \"0.5202285\",\n \"0.5191759\",\n \"0.5189071\",\n \"0.5186799\",\n \"0.5185746\",\n \"0.51844436\",\n \"0.5183619\",\n \"0.5181014\",\n \"0.51745176\",\n \"0.5172385\",\n \"0.5172032\",\n \"0.516866\",\n \"0.5167372\",\n \"0.5166256\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.5600284"},"document_rank":{"kind":"string","value":"29"}}},{"rowIdx":90886,"cells":{"query":{"kind":"string","value":"Test translations are displayed correctly in hindi, as well as bottom rankings rendered correctly."},"document":{"kind":"string","value":"def test_energy_consumption_bottom_ranking_hi(self):\n self.driver.get(\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FIND&unit=kg&hl=hi&bottom='\n )\n\n self.assertEqual(\n self.driver.find_elements_by_class_name('navbar-brand')[0].text,\n 'Data Commons प्लेस रैंकिंग')\n\n subtitle_present = EC.text_to_be_present_in_element(\n (By.TAG_NAME, 'h3'), 'the World से नीचे से 100 देश')\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(subtitle_present)\n self.assertEqual(\n self.driver.find_element_by_tag_name('h1').text,\n 'प्रति व्यक्ति ऊर्जा की खपत के हिसाब से रैंकिंग')\n self.assertEqual(\n self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/h3/a').get_attribute('href'),\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FIND&unit=kg&hl=hi'\n )\n\n table = self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/table')\n headers = table.find_elements_by_xpath('.//thead/tr/th')\n self.assertEqual(headers[0].text, 'रैंक')\n self.assertEqual(headers[1].text, 'देश')\n self.assertEqual(headers[2].text, 'किलोग्राम')\n row = table.find_elements_by_xpath('.//tbody/tr[1]/td')\n self.assertEqual(row[0].text, '172')\n self.assertEqual(row[1].text, 'लेसोथो')\n self.assertEqual(\n row[1].find_element_by_tag_name('a').get_attribute('href'),\n self.url_ + '/place/country/LSO?hl=hi')\n\n chart = self.driver.find_element_by_id('ranking-chart')\n y_text = chart.find_elements_by_class_name(\n 'y')[0].find_elements_by_tag_name('text')\n self.assertEqual(y_text[0].text, '0 कि॰ग्रा॰')\n self.assertEqual(y_text[-1].text, '1.5 हज़ार कि॰ग्रा॰')\n\n x_text = chart.find_elements_by_class_name(\n 'x')[0].find_elements_by_tag_name('text')\n self.assertEqual(x_text[0].text, 'साइप्रस')\n self.assertEqual(x_text[-1].text, 'लेसोथो')"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def test_i18n10(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n10', {'bool': True})\n self.assertEqual(output, 'Ja')","def test_i18n27(self):\n with translation.override(\"ru\"):\n output = self.engine.render_to_string(\"i18n27\", {\"number\": 1})\n self.assertEqual(\n output, \"1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442\"\n )","def test_content(google_translator):\n # from bs4 import BeautifulSoup\n # assert 'GitHub' in BeautifulSoup(response.content).title.string\n assert google_translator.translate(text='좋은') == \"good\"","def test_i18n27(self):\n with translation.override('ru'):\n output = self.engine.render_to_string('i18n27', {'number': 1})\n self.assertEqual(output, '1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442')","def test_i18n27(self):\n with translation.override('ru'):\n output = self.engine.render_to_string('i18n27', {'number': 1})\n self.assertEqual(output, '1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442')","def test_i18n09(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n09')\n self.assertEqual(output, 'Page not found')","def setUp(self):\n super().setUp()\n translation.activate(\"en-us\")","def test_i18n23(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n23')\n self.assertEqual(output, 'nicht gefunden')","def test_i18n16(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n16')\n self.assertEqual(output, '<')","def test_i18n(self):\n # If you want to modify this code, please get acquainted with\n # Python's locale module. In particular:\n # http://docs.python.org/library/locale.html#locale.getdefaultlocale\n\n # Set the standard C locale.\n os.environ['LANG'] = 'C'\n os.environ['LC_ALL'] = 'C.UTF-8'\n\n #must be after above\n out1 = tr('Hello!')\n expected1 = 'Hello!'\n msg = 'Expected %s, got %s' % (expected1, out1)\n assert out1 == expected1, msg\n\n # Set the Indonesian locale to test translations.\n os.environ['LANG'] = 'id'\n os.environ['LC_ALL'] = 'id_ID.UTF-8'\n\n #must be after above\n #indoout1 = tr('Hello!') # translate as 'Hi'\n #indoexpected1 = 'Hi!'\n #msg = 'Expected %s, got %s' % (indoexpected1, indoout1)\n #assert indoout1 == indoexpected1, msg","def use_en(self):\n pass","def test_i18n06(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n06')\n self.assertEqual(output, 'Seite nicht gefunden')","def test_issue3625():\n nlp = Hindi()\n doc = nlp(u\"hi. how हुए. होटल, होटल\")\n assert [token.text for token in doc] == ['hi', '.', 'how', 'हुए', '.', 'होटल', ',', 'होटल']","def test_default_translations(self):\n\t\t\n\t\tself.assertTrue(data.get_default_translation('Catholicism', 3) == 'DRA')\n\t\tself.assertTrue(data.get_default_translation('Christianity', 3) == 'ESV')","def test_i18n28(self):\n output = self.engine.render_to_string(\"i18n28\", {\"anton\": \"α\", \"berta\": \"β\"})\n self.assertEqual(output, \"α + β\")","def use_zh(self):\n pass","def test_hero_all_langs(self):\n\n self.driver.get(self.url_ + '/?hl=de')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons ist ein Open Knowledge Repository'))\n\n self.driver.get(self.url_ + '/?hl=es')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons es un repositorio abierto'))\n\n self.driver.get(self.url_ + '/?hl=fr')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons est un dépôt de connaissances'))\n\n self.driver.get(self.url_ + '/?hl=hi')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons, एक ऐसा प्रोजेक्ट है जिसमें डेटा'))\n\n self.driver.get(self.url_ + '/?hl=ja')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(hero_msg.text.startswith('データコモンズは、マッピングされた共通'))\n\n self.driver.get(self.url_ + '/?hl=ko')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons는 매핑된 공통 항목을 사용해'))\n\n self.driver.get(self.url_ + '/?hl=ru')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons – это открытая база данных'))","def test_i18n28(self):\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\n self.assertEqual(output, 'α + β')","def test_i18n28(self):\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\n self.assertEqual(output, 'α + β')","def get_translation(self):","def test_language_sensitivity(self): \n \n for lang in self.LANGUAGES:\n activate(lang)\n \n self.assertEqual(get_wording_text('test_1'), lang)","def test_i18n11(self):\n output = self.engine.render_to_string('i18n11', {'bool': True})\n self.assertEqual(output, 'ja')","def question_new_translate():","def test_translate_unique_langs(self):\n\n trans_msgs_dict = MessageController.translate_unique_langs({'2': 'es', '4': 'fr'}, \n 'hi', 'en', False, False)\n\n self.assertEqual(trans_msgs_dict, {'es': u'{hola}', 'fr': u'salut'})","def test_international(self):\n\n test_cases = (\n makeSimpleTestCase('the beatles'),\n makeSimpleTestCase('david guetta'),\n makeSimpleTestCase('susan boyle'),\n makeSimpleTestCase('rammstein'),\n makeSimpleTestCase('deadmau5'),\n makeTestCase('deadmau 5', Matcher(title=Equals('deadmau5'))),\n makeSimpleTestCase('2face idibia'),\n makeSimpleTestCase('angelique kidjo'),\n makeSimpleTestCase('gotye'),\n makeSimpleTestCase('michel telo'),\n makeSimpleTestCase('stereo total'),\n makeSimpleTestCase('yelle'),\n makeSimpleTestCase('stefanie heinzmann'),\n makeSimpleTestCase('olly murs')\n )\n\n self._run_tests('i18n_artist', test_cases)","def test_i18n01(self):\n output = self.engine.render_to_string('i18n01')\n self.assertEqual(output, 'xxxyyyxxx')","def test_word_translation(self):\n self.assertEqual(translator.translate_word(\"hour\"), \"ourhay\")\n self.assertEqual(translator.translate_word(\"\"), \"\")\n self.assertEqual(translator.translate_word(\"aaa\"), \"aaayay\")","def translate():\n pass","def test_i18n03(self):\n output = self.engine.render_to_string(\"i18n03\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"Å\")","def test_ukrainian_symbols(self):\n string = \"Минає ніч від’їзду\"\n expected = \"Minaye nich vid’yizdu\"\n self.assertEqual(transliterate(string), expected)","def test_fallback_language(self):\n x = SimpleModel()\n x.set_current_language(self.conf_fallback)\n x.tr_title = \"TITLE_FALLBACK\"\n\n x.set_current_language(self.other_lang1)\n x.tr_title = \"TITLE_XX\"\n x.save()\n\n with translation.override(self.other_lang2):\n x = SimpleModel.objects.get(pk=x.pk)\n self.assertEqual(x.tr_title, \"TITLE_FALLBACK\")","def test_fallback_values_2(self):\n title1_de = \"title de\"\n text1_de = \"text in german\"\n n = TestModelWithFallback2()\n n.title = title1_de\n n.text = text1_de\n n.save()\n del n\n n = TestModelWithFallback2.objects.get(title=title1_de)\n trans_real.activate(\"en\")\n self.failUnlessEqual(n.title, title1_de)\n self.failUnlessEqual(n.text,\\\n TestTranslationOptionsWithFallback2.fallback_values['text'])","def translate(self, language=None):","def test_i18n13(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n13')\n self.assertEqual(output, 'Passwort')","def test_i18n04(self):\n output = self.engine.render_to_string(\"i18n04\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"å\")","def test_langid_benchmark(basic_multilingual):\n examples = [\n {\"text\": \"contingentiam in naturalibus causis.\", \"label\": \"la\"},\n {\"text\": \"I jak opowiadał nieżyjący już pan Czesław\", \"label\": \"pl\"},\n {\"text\": \"Sonera gilt seit längerem als Übernahmekandidat\", \"label\": \"de\"},\n {\"text\": \"与银类似，汞也可以与空气中的硫化氢反应。\", \"label\": \"zh-hans\"},\n {\"text\": \"contradictionem implicat.\", \"label\": \"la\"},\n {\"text\": \"Bis zu Prozent gingen die Offerten etwa im\", \"label\": \"de\"},\n {\"text\": \"inneren Sicherheit vorgeschlagene Ausweitung der\", \"label\": \"de\"},\n {\"text\": \"Multimedia-PDA mit Mini-Tastatur\", \"label\": \"de\"},\n {\"text\": \"Ponášalo sa to na rovnicu o dvoch neznámych.\", \"label\": \"sk\"},\n {\"text\": \"이처럼 앞으로 심판의 그 날에 다시 올 메시아가 예수 그리스도이며 , 그는 모든 인류의\", \"label\": \"ko\"},\n {\"text\": \"Die Arbeitsgruppe bedauert , dass der weit über\", \"label\": \"de\"},\n {\"text\": \"И только раз довелось поговорить с ним не вполне\", \"label\": \"ru\"},\n {\"text\": \"de a-l lovi cu piciorul și conștiința că era\", \"label\": \"ro\"},\n {\"text\": \"relación coas pretensións do demandante e que, nos\", \"label\": \"gl\"},\n {\"text\": \"med petdeset in sedemdeset\", \"label\": \"sl\"},\n {\"text\": \"Catalunya; el Consell Comarcal del Vallès Oriental\", \"label\": \"ca\"},\n {\"text\": \"kunnen worden.\", \"label\": \"nl\"},\n {\"text\": \"Witkin je ve většině ohledů zcela jiný.\", \"label\": \"cs\"},\n {\"text\": \"lernen, so zu agieren, dass sie positive oder auch\", \"label\": \"de\"},\n {\"text\": \"olurmuş...\", \"label\": \"tr\"},\n {\"text\": \"sarcasmo de Altman, desde as «peruas» que discutem\", \"label\": \"pt\"},\n {\"text\": \"خلاف فوجداری مقدمہ درج کرے۔\", \"label\": \"ur\"},\n {\"text\": \"Norddal kommune :\", \"label\": \"no\"},\n {\"text\": \"dem Windows-.-Zeitalter , soll in diesem Jahr\", \"label\": \"de\"},\n {\"text\": \"przeklętych ucieleśniają mit poety-cygana,\", \"label\": \"pl\"},\n {\"text\": \"We do not believe the suspect has ties to this\", \"label\": \"en\"},\n {\"text\": \"groziņu pīšanu.\", \"label\": \"lv\"},\n {\"text\": \"Senior Vice-President David M. Thomas möchte\", \"label\": \"de\"},\n {\"text\": \"neomylně vybral nějakou knihu a začetl se.\", \"label\": \"cs\"},\n {\"text\": \"Statt dessen darf beispielsweise der Browser des\", \"label\": \"de\"},\n {\"text\": \"outubro, alcançando R $ bilhões em .\", \"label\": \"pt\"},\n {\"text\": \"(Porte, ), as it does other disciplines\", \"label\": \"en\"},\n {\"text\": \"uskupení se mylně domnívaly, že podporu\", \"label\": \"cs\"},\n {\"text\": \"Übernahme von Next Ende an dem System herum , das\", \"label\": \"de\"},\n {\"text\": \"No podemos decir a la Hacienda que los alemanes\", \"label\": \"es\"},\n {\"text\": \"и рѣста еи братья\", \"label\": \"orv\"},\n {\"text\": \"الذي اتخذ قرارا بتجميد اعلان الدولة الفلسطينية\", \"label\": \"ar\"},\n {\"text\": \"uurides Rootsi sõjaarhiivist toodud . sajandi\", \"label\": \"et\"},\n {\"text\": \"selskapets penger til å pusse opp sin enebolig på\", \"label\": \"no\"},\n {\"text\": \"средней полосе и севернее в Ярославской,\", \"label\": \"ru\"},\n {\"text\": \"il-massa żejda fil-ġemgħat u superġemgħat ta'\", \"label\": \"mt\"},\n {\"text\": \"The Global Beauties on internetilehekülg, mida\", \"label\": \"et\"},\n {\"text\": \"이스라엘 인들은 하나님이 그 큰 팔을 펴 이집트 인들을 치는 것을 보고 하나님을 두려워하며\", \"label\": \"ko\"},\n {\"text\": \"Snad ještě dodejme jeden ekonomický argument.\", \"label\": \"cs\"},\n {\"text\": \"Spalio d. vykusiame pirmajame rinkimų ture\", \"label\": \"lt\"},\n {\"text\": \"und schlechter Journalismus ein gutes Geschäft .\", \"label\": \"de\"},\n {\"text\": \"Du sodiečiai sėdi ant potvynio apsemtų namų stogo.\", \"label\": \"lt\"},\n {\"text\": \"цей є автентичним.\", \"label\": \"uk\"},\n {\"text\": \"Și îndegrabă fu cu îngerul mulțime de șireaguri\", \"label\": \"ro\"},\n {\"text\": \"sobra personal cualificado.\", \"label\": \"es\"},\n {\"text\": \"Tako se u Njemačkoj dvije trećine liječnika služe\", \"label\": \"hr\"},\n {\"text\": \"Dual-Athlon-Chipsatz noch in diesem Jahr\", \"label\": \"de\"},\n {\"text\": \"यहां तक कि चीन के चीफ ऑफ जनरल स्टाफ भी भारत का\", \"label\": \"hi\"},\n {\"text\": \"Li forestier du mont avale\", \"label\": \"fro\"},\n {\"text\": \"Netzwerken für Privatanwender zu bewundern .\", \"label\": \"de\"},\n {\"text\": \"만해는 승적을 가진 중이 결혼할 수 없다는 불교의 계율을 시대에 맞지 않는 것으로 보았다\", \"label\": \"ko\"},\n {\"text\": \"balance and weight distribution but not really for\", \"label\": \"en\"},\n {\"text\": \"og så e # tente vi opp den om morgonen å sfyrte\", \"label\": \"nn\"},\n {\"text\": \"변화는 의심의 여지가 없는 것이지만 반면에 진화는 논쟁의 씨앗이다 .\", \"label\": \"ko\"},\n {\"text\": \"puteare fac aceastea.\", \"label\": \"ro\"},\n {\"text\": \"Waitt seine Führungsmannschaft nicht dem\", \"label\": \"de\"},\n {\"text\": \"juhtimisega, tulid sealt.\", \"label\": \"et\"},\n {\"text\": \"Veränderungen .\", \"label\": \"de\"},\n {\"text\": \"banda en el Bayer Leverkusen de la Bundesliga de\", \"label\": \"es\"},\n {\"text\": \"В туже зиму посла всеволодъ сн҃а своѥго ст҃ослава\", \"label\": \"orv\"},\n {\"text\": \"пославъ приведе я мастеры ѿ грекъ\", \"label\": \"orv\"},\n {\"text\": \"En un nou escenari difícil d'imaginar fa poques\", \"label\": \"ca\"},\n {\"text\": \"καὶ γὰρ τινὲς αὐτοὺς εὐεργεσίαι εἶχον ἐκ Κροίσου\", \"label\": \"grc\"},\n {\"text\": \"직접적인 관련이 있다 .\", \"label\": \"ko\"},\n {\"text\": \"가까운 듯하면서도 멀다 .\", \"label\": \"ko\"},\n {\"text\": \"Er bietet ein ähnliches Leistungsniveau und\", \"label\": \"de\"},\n {\"text\": \"民都洛水牛是獨居的，並不會以群族聚居。\", \"label\": \"zh-hant\"},\n {\"text\": \"την τρομοκρατία.\", \"label\": \"el\"},\n {\"text\": \"hurbiltzen diren neurrian.\", \"label\": \"eu\"},\n {\"text\": \"Ah dimenticavo, ma tutta sta caciara per fare un\", \"label\": \"it\"},\n {\"text\": \"На первом этапе (-) прошла так называемая\", \"label\": \"ru\"},\n {\"text\": \"of games are on the market.\", \"label\": \"en\"},\n {\"text\": \"находится Мост дружбы, соединяющий узбекский и\", \"label\": \"ru\"},\n {\"text\": \"lessié je voldroie que li saint fussent aporté\", \"label\": \"fro\"},\n {\"text\": \"Дошла очередь и до Гималаев.\", \"label\": \"ru\"},\n {\"text\": \"vzácným suknem táhly pouští, si jednou chtěl do\", \"label\": \"cs\"},\n {\"text\": \"E no terceiro tipo sitúa a familias (%), nos que a\", \"label\": \"gl\"},\n {\"text\": \"وجابت دوريات امريكية وعراقية شوارع المدينة، فيما\", \"label\": \"ar\"},\n {\"text\": \"Jeg har bodd her i år .\", \"label\": \"no\"},\n {\"text\": \"Pohrozil, že odbory zostří postoj, pokud se\", \"label\": \"cs\"},\n {\"text\": \"tinham conseguido.\", \"label\": \"pt\"},\n {\"text\": \"Nicht-Erkrankten einen Anfangsverdacht für einen\", \"label\": \"de\"},\n {\"text\": \"permanece em aberto.\", \"label\": \"pt\"},\n {\"text\": \"questi possono promettere rendimenti fino a un\", \"label\": \"it\"},\n {\"text\": \"Tema juurutatud kahevedurisüsteemita oleksid\", \"label\": \"et\"},\n {\"text\": \"Поведение внешне простой игрушки оказалось\", \"label\": \"ru\"},\n {\"text\": \"Bundesländern war vom Börsenverein des Deutschen\", \"label\": \"de\"},\n {\"text\": \"acció, 'a mesura que avanci l'estiu, amb l'augment\", \"label\": \"ca\"},\n {\"text\": \"Dove trovare queste risorse? Jay Naidoo, ministro\", \"label\": \"it\"},\n {\"text\": \"essas gordurinhas.\", \"label\": \"pt\"},\n {\"text\": \"Im zweiten Schritt sollen im übernächsten Jahr\", \"label\": \"de\"},\n {\"text\": \"allveelaeva pole enam vaja, kuna külm sõda on läbi\", \"label\": \"et\"},\n {\"text\": \"उपद्रवी दुकानों को लूटने के साथ ही उनमें आग लगा\", \"label\": \"hi\"},\n {\"text\": \"@user nella sfortuna sei fortunata ..\", \"label\": \"it\"},\n {\"text\": \"математических школ в виде грозовых туч.\", \"label\": \"ru\"},\n {\"text\": \"No cambiaremos nunca nuestra forma de jugar por un\", \"label\": \"es\"},\n {\"text\": \"dla tej klasy ani wymogów minimalnych, z wyjątkiem\", \"label\": \"pl\"},\n {\"text\": \"en todo el mundo, mientras que en España consiguió\", \"label\": \"es\"},\n {\"text\": \"политики считать надежное обеспечение военной\", \"label\": \"ru\"},\n {\"text\": \"gogoratzen du, genio alemana delakoaren\", \"label\": \"eu\"},\n {\"text\": \"Бычий глаз.\", \"label\": \"ru\"},\n {\"text\": \"Opeření se v pravidelných obdobích obnovuje\", \"label\": \"cs\"},\n {\"text\": \"I no és només la seva, es tracta d'una resposta\", \"label\": \"ca\"},\n {\"text\": \"오경을 가르쳤다 .\", \"label\": \"ko\"},\n {\"text\": \"Nach der so genannten Start-up-Periode vergibt die\", \"label\": \"de\"},\n {\"text\": \"Saulista huomasi jo lapsena , että hänellä on\", \"label\": \"fi\"},\n {\"text\": \"Министерство культуры сочло нецелесообразным, и\", \"label\": \"ru\"},\n {\"text\": \"znepřátelené tábory v Tádžikistánu předseda\", \"label\": \"cs\"},\n {\"text\": \"καὶ ἦν ὁ λαὸς προσδοκῶν τὸν Ζαχαρίαν καὶ ἐθαύμαζον\", \"label\": \"grc\"},\n {\"text\": \"Вечером, в продукте, этот же человек говорил о\", \"label\": \"ru\"},\n {\"text\": \"lugar á formación de xuizos máis complexos.\", \"label\": \"gl\"},\n {\"text\": \"cheaper, in the end?\", \"label\": \"en\"},\n {\"text\": \"الوزارة في شأن صفقات بيع الشركات العامة التي تم\", \"label\": \"ar\"},\n {\"text\": \"tärkeintä elämässäni .\", \"label\": \"fi\"},\n {\"text\": \"Виконання Мінських угод було заблоковано Росією та\", \"label\": \"uk\"},\n {\"text\": \"Aby szybko rozpoznać żołnierzy desantu, należy\", \"label\": \"pl\"},\n {\"text\": \"Bankengeschäfte liegen vorn , sagte Strothmann .\", \"label\": \"de\"},\n {\"text\": \"продолжение работы.\", \"label\": \"ru\"},\n {\"text\": \"Metro AG plant Online-Offensive\", \"label\": \"de\"},\n {\"text\": \"nu vor veni, și să vor osîndi, aceia nu pot porni\", \"label\": \"ro\"},\n {\"text\": \"Ich denke , es geht in Wirklichkeit darum , NT bei\", \"label\": \"de\"},\n {\"text\": \"de turism care încasează contravaloarea\", \"label\": \"ro\"},\n {\"text\": \"Aurkaria itotzea da helburua, baloia lapurtu eta\", \"label\": \"eu\"},\n {\"text\": \"com a centre de formació en Tecnologies de la\", \"label\": \"ca\"},\n {\"text\": \"oportet igitur quod omne agens in agendo intendat\", \"label\": \"la\"},\n {\"text\": \"Jerzego Andrzejewskiego, oparty na chińskich\", \"label\": \"pl\"},\n {\"text\": \"sau một vài câu chuyện xã giao không dính dáng tới\", \"label\": \"vi\"},\n {\"text\": \"что экономическому прорыву жесткий авторитарный\", \"label\": \"ru\"},\n {\"text\": \"DRAM-Preisen scheinen DSPs ein\", \"label\": \"de\"},\n {\"text\": \"Jos dajan nubbái: Mana!\", \"label\": \"sme\"},\n {\"text\": \"toți carii ascultară de el să răsipiră.\", \"label\": \"ro\"},\n {\"text\": \"odpowiedzialności, które w systemie własności\", \"label\": \"pl\"},\n {\"text\": \"Dvomesečno potovanje do Mollenda v Peruju je\", \"label\": \"sl\"},\n {\"text\": \"d'entre les agències internacionals.\", \"label\": \"ca\"},\n {\"text\": \"Fahrzeugzugangssysteme gefertigt und an viele\", \"label\": \"de\"},\n {\"text\": \"in an answer to the sharers' petition in Cuthbert\", \"label\": \"en\"},\n {\"text\": \"Europa-Domain per Verordnung zu regeln .\", \"label\": \"de\"},\n {\"text\": \"#Balotelli. Su ebay prezzi stracciati per Silvio\", \"label\": \"it\"},\n {\"text\": \"Ne na košickém trávníku, ale už včera v letadle se\", \"label\": \"cs\"},\n {\"text\": \"zaměstnanosti a investičních strategií.\", \"label\": \"cs\"},\n {\"text\": \"Tatínku, udělej den\", \"label\": \"cs\"},\n {\"text\": \"frecuencia con Mary.\", \"label\": \"es\"},\n {\"text\": \"Свеаборге.\", \"label\": \"ru\"},\n {\"text\": \"opatření slovenské strany o certifikaci nejvíce\", \"label\": \"cs\"},\n {\"text\": \"En todas me decían: 'Espera que hagamos un estudio\", \"label\": \"es\"},\n {\"text\": \"Die Demonstration sollte nach Darstellung der\", \"label\": \"de\"},\n {\"text\": \"Ci vorrà un assoluto rigore se dietro i disavanzi\", \"label\": \"it\"},\n {\"text\": \"Tatínku, víš, že Honzovi odešla maminka?\", \"label\": \"cs\"},\n {\"text\": \"Die Anzahl der Rechner wuchs um % auf und die\", \"label\": \"de\"},\n {\"text\": \"האמריקאית על אדמת סעודיה עלולה לסבך את ישראל, אין\", \"label\": \"he\"},\n {\"text\": \"Volán Egyesülés, a Közlekedési Főfelügyelet is.\", \"label\": \"hu\"},\n {\"text\": \"Schejbala, který stejnou hru s velkým úspěchem\", \"label\": \"cs\"},\n {\"text\": \"depends on the data type of the field.\", \"label\": \"en\"},\n {\"text\": \"Umsatzwarnung zu Wochenbeginn zeitweise auf ein\", \"label\": \"de\"},\n {\"text\": \"niin heti nukun .\", \"label\": \"fi\"},\n {\"text\": \"Mobilfunkunternehmen gegen die Anwendung der so\", \"label\": \"de\"},\n {\"text\": \"sapessi le intenzioni del governo Monti e dell'UE\", \"label\": \"it\"},\n {\"text\": \"Di chi è figlia Martine Aubry?\", \"label\": \"it\"},\n {\"text\": \"avec le reste du monde.\", \"label\": \"fr\"},\n {\"text\": \"Այդ մաքոքը ինքնին նոր չէ, աշխարհը արդեն մի քանի\", \"label\": \"hy\"},\n {\"text\": \"și în cazul destrămării cenaclului.\", \"label\": \"ro\"},\n {\"text\": \"befriedigen kann , und ohne die auftretenden\", \"label\": \"de\"},\n {\"text\": \"Κύκνον τ̓ ἐξεναρεῖν καὶ ἀπὸ κλυτὰ τεύχεα δῦσαι.\", \"label\": \"grc\"},\n {\"text\": \"færdiguddannede.\", \"label\": \"da\"},\n {\"text\": \"Schmidt war Sohn eines Rittergutsbesitzers.\", \"label\": \"de\"},\n {\"text\": \"и вдаша попадь ѡпрати\", \"label\": \"orv\"},\n {\"text\": \"cine nu știe învățătură”.\", \"label\": \"ro\"},\n {\"text\": \"détacha et cette dernière tenta de tuer le jeune\", \"label\": \"fr\"},\n {\"text\": \"Der har saka også ei lengre forhistorie.\", \"label\": \"nn\"},\n {\"text\": \"Pieprz roztłuc w moździerzu, dodać do pasty,\", \"label\": \"pl\"},\n {\"text\": \"Лежа за гребнем оврага, как за бруствером, Ушаков\", \"label\": \"ru\"},\n {\"text\": \"gesucht habe, vielen Dank nochmals!\", \"label\": \"de\"},\n {\"text\": \"инструментальных сталей, повышения\", \"label\": \"ru\"},\n {\"text\": \"im Halbfinale Patrick Smith und im Finale dann\", \"label\": \"de\"},\n {\"text\": \"البنوك التريث في منح تسهيلات جديدة لمنتجي حديد\", \"label\": \"ar\"},\n {\"text\": \"una bolsa ventral, la cual se encuentra debajo de\", \"label\": \"es\"},\n {\"text\": \"za SETimes.\", \"label\": \"sr\"},\n {\"text\": \"de Irak, a un piloto italiano que había violado el\", \"label\": \"es\"},\n {\"text\": \"Er könne sich nicht erklären , wie die Zeitung auf\", \"label\": \"de\"},\n {\"text\": \"Прохорова.\", \"label\": \"ru\"},\n {\"text\": \"la democrazia perde sulla tecnocrazia? #\", \"label\": \"it\"},\n {\"text\": \"entre ambas instituciones, confirmó al medio que\", \"label\": \"es\"},\n {\"text\": \"Austlandet, vart det funne om lag førti\", \"label\": \"nn\"},\n {\"text\": \"уровнями власти.\", \"label\": \"ru\"},\n {\"text\": \"Dá tedy primáři úplatek, a často ne malý.\", \"label\": \"cs\"},\n {\"text\": \"brillantes del acto, al llevar a cabo en el\", \"label\": \"es\"},\n {\"text\": \"eee druga zadeva je majhen priročen gre kamorkoli\", \"label\": \"sl\"},\n {\"text\": \"Das ATX-Board paßt in herkömmliche PC-ATX-Gehäuse\", \"label\": \"de\"},\n {\"text\": \"Za vodné bylo v prvním pololetí zaplaceno v ČR\", \"label\": \"cs\"},\n {\"text\": \"Даже на полсантиметра.\", \"label\": \"ru\"},\n {\"text\": \"com la del primer tinent d'alcalde en funcions,\", \"label\": \"ca\"},\n {\"text\": \"кількох оповідань в цілості — щось на зразок того\", \"label\": \"uk\"},\n {\"text\": \"sed ad divitias congregandas, vel superfluum\", \"label\": \"la\"},\n {\"text\": \"Norma Talmadge, spela mot Valentino i en version\", \"label\": \"sv\"},\n {\"text\": \"Dlatego chciał się jej oświadczyć w niezwykłym\", \"label\": \"pl\"},\n {\"text\": \"будут выступать на одинаковых снарядах.\", \"label\": \"ru\"},\n {\"text\": \"Orang-orang terbunuh di sana.\", \"label\": \"id\"},\n {\"text\": \"لدى رايت شقيق اسمه أوسكار, وهو يعمل كرسام للكتب\", \"label\": \"ar\"},\n {\"text\": \"Wirklichkeit verlagerten und kaum noch\", \"label\": \"de\"},\n {\"text\": \"как перемешивают костяшки перед игрой в домино, и\", \"label\": \"ru\"},\n {\"text\": \"В средине дня, когда солнце светило в нашу\", \"label\": \"ru\"},\n {\"text\": \"d'aventure aux rôles de jeune romantique avec une\", \"label\": \"fr\"},\n {\"text\": \"My teď hledáme organizace, jež by s námi chtěly\", \"label\": \"cs\"},\n {\"text\": \"Urteilsfähigkeit einbüßen , wenn ich eigene\", \"label\": \"de\"},\n {\"text\": \"sua appartenenza anche a voci diverse da quella in\", \"label\": \"it\"},\n {\"text\": \"Aufträge dieses Jahr verdoppeln werden .\", \"label\": \"de\"},\n {\"text\": \"M.E.: Miała szanse mnie odnaleźć, gdyby naprawdę\", \"label\": \"pl\"},\n {\"text\": \"secundum contactum virtutis, cum careat dimensiva\", \"label\": \"la\"},\n {\"text\": \"ezinbestekoa dela esan zuen.\", \"label\": \"eu\"},\n {\"text\": \"Anek hurbiltzeko eskatzen zion besaulkitik, eta\", \"label\": \"eu\"},\n {\"text\": \"perfectius alio videat, quamvis uterque videat\", \"label\": \"la\"},\n {\"text\": \"Die Strecke war anspruchsvoll und führte unter\", \"label\": \"de\"},\n {\"text\": \"саморазоблачительным уроком, западные СМИ не\", \"label\": \"ru\"},\n {\"text\": \"han representerer radikal islamisme .\", \"label\": \"no\"},\n {\"text\": \"Què s'hi respira pel que fa a la reforma del\", \"label\": \"ca\"},\n {\"text\": \"previsto para também ser desconstruido.\", \"label\": \"pt\"},\n {\"text\": \"Ὠκεανοῦ βαθυκόλποις ἄνθεά τ̓ αἰνυμένην, ῥόδα καὶ\", \"label\": \"grc\"},\n {\"text\": \"para jovens de a anos nos Cieps.\", \"label\": \"pt\"},\n {\"text\": \"संघर्ष को अंजाम तक पहुंचाने का ऐलान किया है ।\", \"label\": \"hi\"},\n {\"text\": \"objeví i u nás.\", \"label\": \"cs\"},\n {\"text\": \"kvitteringer.\", \"label\": \"da\"},\n {\"text\": \"This report is no exception.\", \"label\": \"en\"},\n {\"text\": \"Разлепват доносниците до избирателните списъци\", \"label\": \"bg\"},\n {\"text\": \"anderem ihre Bewegungsfreiheit in den USA\", \"label\": \"de\"},\n {\"text\": \"Ñu tegoon ca kaw gor ña ay njotti bopp yu kenn\", \"label\": \"wo\"},\n {\"text\": \"Struktur kann beispielsweise der Schwerpunkt mehr\", \"label\": \"de\"},\n {\"text\": \"% la velocidad permitida, la sanción es muy grave.\", \"label\": \"es\"},\n {\"text\": \"Teles-Einstieg in ADSL-Markt\", \"label\": \"de\"},\n {\"text\": \"ettekäändeks liiga suure osamaksu.\", \"label\": \"et\"},\n {\"text\": \"als Indiz für die geänderte Marktpolitik des\", \"label\": \"de\"},\n {\"text\": \"quod quidem aperte consequitur ponentes\", \"label\": \"la\"},\n {\"text\": \"de negociación para el próximo de junio.\", \"label\": \"es\"},\n {\"text\": \"Tyto důmyslné dekorace doznaly v poslední době\", \"label\": \"cs\"},\n {\"text\": \"največjega uspeha doslej.\", \"label\": \"sl\"},\n {\"text\": \"Paul Allen je jedan od suosnivača Interval\", \"label\": \"hr\"},\n {\"text\": \"Federal (Seac / DF) eo Sindicato das Empresas de\", \"label\": \"pt\"},\n {\"text\": \"Quartal mit . Mark gegenüber dem gleichen Quartal\", \"label\": \"de\"},\n {\"text\": \"otros clubes y del Barça B saldrán varios\", \"label\": \"es\"},\n {\"text\": \"Jaskula (Pol.) -\", \"label\": \"cs\"},\n {\"text\": \"umožnily říci, že je možné přejít k mnohem\", \"label\": \"cs\"},\n {\"text\": \"اعلن الجنرال تومي فرانكس قائد القوات الامريكية\", \"label\": \"ar\"},\n {\"text\": \"Telekom-Chef Ron Sommer und der Vorstandssprecher\", \"label\": \"de\"},\n {\"text\": \"My, jako průmyslový a finanční holding, můžeme\", \"label\": \"cs\"},\n {\"text\": \"voorlichting onder andere betrekking kan hebben:\", \"label\": \"nl\"},\n {\"text\": \"Hinrichtung geistig Behinderter applaudiert oder\", \"label\": \"de\"},\n {\"text\": \"wie beispielsweise Anzahl erzielte Klicks ,\", \"label\": \"de\"},\n {\"text\": \"Intel-PC-SDRAM-Spezifikation in der Version . (\", \"label\": \"de\"},\n {\"text\": \"plângere în termen de zile de la comunicarea\", \"label\": \"ro\"},\n {\"text\": \"и Испания ще изгубят втория си комисар в ЕК.\", \"label\": \"bg\"},\n {\"text\": \"इसके चलते इस आदिवासी जनजाति का क्षरण हो रहा है ।\", \"label\": \"hi\"},\n {\"text\": \"aunque se mostró contrario a establecer un\", \"label\": \"es\"},\n {\"text\": \"des letzten Jahres von auf Millionen Euro .\", \"label\": \"de\"},\n {\"text\": \"Ankara se također poziva da u cijelosti ratificira\", \"label\": \"hr\"},\n {\"text\": \"herunterlädt .\", \"label\": \"de\"},\n {\"text\": \"стрессовую ситуацию для организма, каковой\", \"label\": \"ru\"},\n {\"text\": \"Státního shromáždění (parlamentu).\", \"label\": \"cs\"},\n {\"text\": \"diskutieren , ob und wie dieser Dienst weiterhin\", \"label\": \"de\"},\n {\"text\": \"Verbindungen zu FPÖ-nahen Polizisten gepflegt und\", \"label\": \"de\"},\n {\"text\": \"Pražského volebního lídra ovšem nevybírá Miloš\", \"label\": \"cs\"},\n {\"text\": \"Nach einem Bericht der Washington Post bleibt das\", \"label\": \"de\"},\n {\"text\": \"للوضع آنذاك، لكني في قرارة نفسي كنت سعيداً لما\", \"label\": \"ar\"},\n {\"text\": \"не желаят запазването на статуквото.\", \"label\": \"bg\"},\n {\"text\": \"Offenburg gewesen .\", \"label\": \"de\"},\n {\"text\": \"ἐὰν ὑμῖν εἴπω οὐ μὴ πιστεύσητε\", \"label\": \"grc\"},\n {\"text\": \"all'odiato compagno di squadra Prost, il quale\", \"label\": \"it\"},\n {\"text\": \"historischen Gänselieselbrunnens.\", \"label\": \"de\"},\n {\"text\": \"למידע מלווייני הריגול האמריקאיים העוקבים אחר\", \"label\": \"he\"},\n {\"text\": \"οὐδὲν ἄρα διαφέρεις Ἀμάσιος τοῦ Ἠλείου, ὃν\", \"label\": \"grc\"},\n {\"text\": \"movementos migratorios.\", \"label\": \"gl\"},\n {\"text\": \"Handy und ein Spracherkennungsprogramm sämtliche\", \"label\": \"de\"},\n {\"text\": \"Kümne aasta jooksul on Eestisse ohjeldamatult\", \"label\": \"et\"},\n {\"text\": \"H.G. Bücknera.\", \"label\": \"pl\"},\n {\"text\": \"protiv krijumčarenja, ili pak traženju ukidanja\", \"label\": \"hr\"},\n {\"text\": \"Topware-Anteile mehrere Millionen Mark gefordert\", \"label\": \"de\"},\n {\"text\": \"Maar de mensen die nu over Van Dijk bij FC Twente\", \"label\": \"nl\"},\n {\"text\": \"poidan experimentar as percepcións do interesado,\", \"label\": \"gl\"},\n {\"text\": \"Miał przecież w kieszeni nóż.\", \"label\": \"pl\"},\n {\"text\": \"Avšak žádná z nich nepronikla za hranice přímé\", \"label\": \"cs\"},\n {\"text\": \"esim. helpottamalla luottoja muiden\", \"label\": \"fi\"},\n {\"text\": \"Podle předběžných výsledků zvítězila v\", \"label\": \"cs\"},\n {\"text\": \"Nicht nur das Web-Frontend , auch die\", \"label\": \"de\"},\n {\"text\": \"Regierungsinstitutionen oder Universitäten bei\", \"label\": \"de\"},\n {\"text\": \"Խուլեն Լոպետեգիին, պատճառաբանելով, որ վերջինս\", \"label\": \"hy\"},\n {\"text\": \"Афганистана, где в последние дни идут ожесточенные\", \"label\": \"ru\"},\n {\"text\": \"лѧхове же не идоша\", \"label\": \"orv\"},\n {\"text\": \"Mit Hilfe von IBMs Chip-Management-Systemen sollen\", \"label\": \"de\"},\n {\"text\": \", als Manager zu Telefonica zu wechseln .\", \"label\": \"de\"},\n {\"text\": \"którym zajmuje się człowiek, zmienia go i pozwala\", \"label\": \"pl\"},\n {\"text\": \"činí kyperských liber, to je asi USD.\", \"label\": \"cs\"},\n {\"text\": \"Studienplätze getauscht werden .\", \"label\": \"de\"},\n {\"text\": \"учёных, орнитологов признают вид.\", \"label\": \"ru\"},\n {\"text\": \"acordare a concediilor prevăzute de legislațiile\", \"label\": \"ro\"},\n {\"text\": \"at større innsats for fornybar, berekraftig energi\", \"label\": \"nn\"},\n {\"text\": \"Politiet veit ikkje kor mange personar som deltok\", \"label\": \"nn\"},\n {\"text\": \"offentligheten av unge , sinte menn som har\", \"label\": \"no\"},\n {\"text\": \"însuși în jurul lapunei, care încet DISPARE în\", \"label\": \"ro\"},\n {\"text\": \"O motivo da decisão é evitar uma sobrecarga ainda\", \"label\": \"pt\"},\n {\"text\": \"El Apostolado de la prensa contribuye en modo\", \"label\": \"es\"},\n {\"text\": \"Teltow ( Kreis Teltow-Fläming ) ist Schmitt einer\", \"label\": \"de\"},\n {\"text\": \"grozījumus un iesniegt tos Apvienoto Nāciju\", \"label\": \"lv\"},\n {\"text\": \"Gestalt einer deutschen Nationalmannschaft als\", \"label\": \"de\"},\n {\"text\": \"D überholt zu haben , konterte am heutigen Montag\", \"label\": \"de\"},\n {\"text\": \"Softwarehersteller Oracle hat im dritten Quartal\", \"label\": \"de\"},\n {\"text\": \"Během nich se ekonomické podmínky mohou radikálně\", \"label\": \"cs\"},\n {\"text\": \"Dziki kot w górach zeskakuje z kamienia.\", \"label\": \"pl\"},\n {\"text\": \"Ačkoliv ligový nováček prohrál, opět potvrdil, že\", \"label\": \"cs\"},\n {\"text\": \"des Tages , Portraits internationaler Stars sowie\", \"label\": \"de\"},\n {\"text\": \"Communicator bekannt wurde .\", \"label\": \"de\"},\n {\"text\": \"τῷ δ’ ἄρα καὶ αὐτῷ ἡ γυνή ἐπίτεξ ἐοῦσα πᾶσαν\", \"label\": \"grc\"},\n {\"text\": \"Triadú tenia, mentre redactava 'Dies de memòria',\", \"label\": \"ca\"},\n {\"text\": \"دستهجمعی در درخشندگی ماه سیمگون زمزمه ستاینده و\", \"label\": \"fa\"},\n {\"text\": \"Книгу, наполненную мелочной заботой об одежде,\", \"label\": \"ru\"},\n {\"text\": \"putares canem leporem persequi.\", \"label\": \"la\"},\n {\"text\": \"В дальнейшем эта яркость слегка померкла, но в\", \"label\": \"ru\"},\n {\"text\": \"offizielles Verfahren gegen die Telekom\", \"label\": \"de\"},\n {\"text\": \"podrían haber sido habitantes de la Península\", \"label\": \"es\"},\n {\"text\": \"Grundlage für dieses Verfahren sind spezielle\", \"label\": \"de\"},\n {\"text\": \"Rechtsausschuß vorgelegten Entwurf der Richtlinie\", \"label\": \"de\"},\n {\"text\": \"Im so genannten Portalgeschäft sei das Unternehmen\", \"label\": \"de\"},\n {\"text\": \"ⲏ ⲉⲓϣⲁⲛϥⲓ ⲛⲉⲓⲇⲱⲗⲟⲛ ⲉⲧϩⲙⲡⲉⲕⲏⲓ ⲙⲏ ⲉⲓⲛⲁϣϩⲱⲡ ⲟⲛ ⲙⲡⲣⲏ\", \"label\": \"cop\"},\n {\"text\": \"juego podían matar a cualquier herbívoro, pero\", \"label\": \"es\"},\n {\"text\": \"Nach Angaben von Axent nutzen Unternehmen aus der\", \"label\": \"de\"},\n {\"text\": \"hrdiny Havlovy Zahradní slavnosti (premiéra ) se\", \"label\": \"cs\"},\n {\"text\": \"Een zin van heb ik jou daar\", \"label\": \"nl\"},\n {\"text\": \"hat sein Hirn an der CeBIT-Kasse vergessen .\", \"label\": \"de\"},\n {\"text\": \"καὶ τοὺς ἐκπλαγέντας οὐκ ἔχειν ἔτι ἐλεγχομένους\", \"label\": \"grc\"},\n {\"text\": \"nachgewiesenen langfristigen Kosten , sowie den im\", \"label\": \"de\"},\n {\"text\": \"jučer nakon četiri dana putovanja u Helsinki.\", \"label\": \"hr\"},\n {\"text\": \"pašto paslaugos teikėjas gali susitarti su\", \"label\": \"lt\"},\n {\"text\": \"В результате, эти золотые кадры переходят из одной\", \"label\": \"ru\"},\n {\"text\": \"द फाइव-ईयर एंगेजमेंट में अभिनय किया जिसमें जैसन\", \"label\": \"hi\"},\n {\"text\": \"výpis o počtu akcií.\", \"label\": \"cs\"},\n {\"text\": \"Enfin, elles arrivent à un pavillon chinois\", \"label\": \"fr\"},\n {\"text\": \"Tentu saja, tren yang berhubungandengan\", \"label\": \"id\"},\n {\"text\": \"Arbeidarpartiet og SV har sikra seg fleirtal mot\", \"label\": \"nn\"},\n {\"text\": \"eles: 'Tudo isso está errado' , disse um\", \"label\": \"pt\"},\n {\"text\": \"The islands are in their own time zone, minutes\", \"label\": \"en\"},\n {\"text\": \"Auswahl debütierte er am .\", \"label\": \"de\"},\n {\"text\": \"Bu komisyonlar, arazilerini satın almak için\", \"label\": \"tr\"},\n {\"text\": \"Geschütze gegen Redmond aufgefahren .\", \"label\": \"de\"},\n {\"text\": \"Time scything the hours, but at the top, over the\", \"label\": \"en\"},\n {\"text\": \"Di musim semi , berharap mengadaptasi Tintin untuk\", \"label\": \"id\"},\n {\"text\": \"крупнейшей геополитической катастрофой XX века.\", \"label\": \"ru\"},\n {\"text\": \"Rajojen avaaminen ei suju ongelmitta .\", \"label\": \"fi\"},\n {\"text\": \"непроницаемым, как для СССР.\", \"label\": \"ru\"},\n {\"text\": \"Ma non mancano le polemiche.\", \"label\": \"it\"},\n {\"text\": \"Internet als Ort politischer Diskussion und auch\", \"label\": \"de\"},\n {\"text\": \"incomplets.\", \"label\": \"ca\"},\n {\"text\": \"Su padre luchó al lado de Luis Moya, primer Jefe\", \"label\": \"es\"},\n {\"text\": \"informazione.\", \"label\": \"it\"},\n {\"text\": \"Primacom bietet für Telekom-Kabelnetz\", \"label\": \"de\"},\n {\"text\": \"Oświadczenie prezydencji w imieniu Unii\", \"label\": \"pl\"},\n {\"text\": \"foran rattet i familiens gamle Baleno hvis døra på\", \"label\": \"no\"},\n {\"text\": \"[speaker:laughter]\", \"label\": \"sl\"},\n {\"text\": \"Dog med langt mindre utstyr med seg.\", \"label\": \"nn\"},\n {\"text\": \"dass es nicht schon mit der anfänglichen\", \"label\": \"de\"},\n {\"text\": \"इस पर दोनों पक्षों में नोकझोंक शुरू हो गई ।\", \"label\": \"hi\"},\n {\"text\": \"کے ترجمان منیش تیواری اور دگ وجئے سنگھ نے بھی یہ\", \"label\": \"ur\"},\n {\"text\": \"dell'Assemblea Costituente che posseggono i\", \"label\": \"it\"},\n {\"text\": \"и аште вьси съблазнѧтъ сѧ нъ не азъ\", \"label\": \"cu\"},\n {\"text\": \"In Irvine hat auch das Logistikunternehmen Atlas\", \"label\": \"de\"},\n {\"text\": \"законодательных норм, принимаемых существующей\", \"label\": \"ru\"},\n {\"text\": \"Κροίσῳ προτείνων τὰς χεῖρας ἐπικατασφάξαι μιν\", \"label\": \"grc\"},\n {\"text\": \"МИНУСЫ: ИНФЛЯЦИЯ И КРИЗИС В ЖИВОТНОВОДСТВЕ.\", \"label\": \"ru\"},\n {\"text\": \"unterschiedlicher Meinung .\", \"label\": \"de\"},\n {\"text\": \"Jospa joku ystävällinen sielu auttaisi kassieni\", \"label\": \"fi\"},\n {\"text\": \"Añadió que, en el futuro se harán otros\", \"label\": \"es\"},\n {\"text\": \"Sessiz tonlama hem Fince, hem de Kuzey Sami\", \"label\": \"tr\"},\n {\"text\": \"nicht ihnen gehört und sie nicht alles , was sie\", \"label\": \"de\"},\n {\"text\": \"Etelästä Kuivajärveen laskee Tammelan Liesjärvestä\", \"label\": \"fi\"},\n {\"text\": \"ICANNs Vorsitzender Vint Cerf warb mit dem Hinweis\", \"label\": \"de\"},\n {\"text\": \"Norsk politikk frå til kan dermed, i\", \"label\": \"nn\"},\n {\"text\": \"Głosowało posłów.\", \"label\": \"pl\"},\n {\"text\": \"Danny Jones -- smithjones@ev.net\", \"label\": \"en\"},\n {\"text\": \"sebeuvědomění moderní civilizace sehrála lučavka\", \"label\": \"cs\"},\n {\"text\": \"относительно спокойный сон: тому гарантия\", \"label\": \"ru\"},\n {\"text\": \"A halte voiz prist li pedra a crïer\", \"label\": \"fro\"},\n {\"text\": \"آنها امیدوارند این واکسن بهزودی در دسترس بیماران\", \"label\": \"fa\"},\n {\"text\": \"vlastní důstojnou vousatou tváří.\", \"label\": \"cs\"},\n {\"text\": \"ora aprire la strada a nuove cause e alimentare il\", \"label\": \"it\"},\n {\"text\": \"Die Zahl der Vielleser nahm von auf Prozent zu ,\", \"label\": \"de\"},\n {\"text\": \"Finanzvorstand von Hotline-Dienstleister InfoGenie\", \"label\": \"de\"},\n {\"text\": \"entwickeln .\", \"label\": \"de\"},\n {\"text\": \"incolumità pubblica.\", \"label\": \"it\"},\n {\"text\": \"lehtija televisiomainonta\", \"label\": \"fi\"},\n {\"text\": \"joistakin kohdista eri mieltä.\", \"label\": \"fi\"},\n {\"text\": \"Hlavně anglická nezávislá scéna, Dead Can Dance,\", \"label\": \"cs\"},\n {\"text\": \"pásmech od do bodů bodové stupnice.\", \"label\": \"cs\"},\n {\"text\": \"Zu Beginn des Ersten Weltkrieges zählte das\", \"label\": \"de\"},\n {\"text\": \"Així van sorgir, damunt els antics cementiris,\", \"label\": \"ca\"},\n {\"text\": \"In manchem Gedicht der spätern Alten, wie zum\", \"label\": \"de\"},\n {\"text\": \"gaweihaida jah insandida in þana fairƕu jus qiþiþ\", \"label\": \"got\"},\n {\"text\": \"Beides sollte gelöscht werden!\", \"label\": \"de\"},\n {\"text\": \"modifiqués la seva petició inicial de anys de\", \"label\": \"ca\"},\n {\"text\": \"В день открытия симпозиума состоялась закладка\", \"label\": \"ru\"},\n {\"text\": \"tõestatud.\", \"label\": \"et\"},\n {\"text\": \"ἵππῳ πίπτει αὐτοῦ ταύτῃ\", \"label\": \"grc\"},\n {\"text\": \"bisher nie enttäuscht!\", \"label\": \"de\"},\n {\"text\": \"De bohte ollu tuollárat ja suttolaččat ja\", \"label\": \"sme\"},\n {\"text\": \"Klarsignal från röstlängdsläsaren, tre tryck i\", \"label\": \"sv\"},\n {\"text\": \"Tvůrcem nového termínu je Joseph Fisher.\", \"label\": \"cs\"},\n {\"text\": \"Nie miałem czasu na reakcję twierdzi Norbert,\", \"label\": \"pl\"},\n {\"text\": \"potentia Schöpfer.\", \"label\": \"de\"},\n {\"text\": \"Un poquito caro, pero vale mucho la pena;\", \"label\": \"es\"},\n {\"text\": \"οὔ τε γὰρ ἴφθιμοι Λύκιοι Δαναῶν ἐδύναντο τεῖχος\", \"label\": \"grc\"},\n {\"text\": \"vajec, sladového výtažku a některých vitamínových\", \"label\": \"cs\"},\n {\"text\": \"Настоящие герои, те, чьи истории потом\", \"label\": \"ru\"},\n {\"text\": \"praesumptio:\", \"label\": \"la\"},\n {\"text\": \"Olin justkui nende vastutusel.\", \"label\": \"et\"},\n {\"text\": \"Jokainen keinahdus tuo lähemmäksi hetkeä jolloin\", \"label\": \"fi\"},\n {\"text\": \"ekonomicky výhodných způsobů odvodnění těžkých,\", \"label\": \"cs\"},\n {\"text\": \"Poprvé ve své historii dokázala v kvalifikaci pro\", \"label\": \"cs\"},\n {\"text\": \"zpracovatelského a spotřebního průmyslu bude nutné\", \"label\": \"cs\"},\n {\"text\": \"Windows CE zu integrieren .\", \"label\": \"de\"},\n {\"text\": \"Armangué, a través d'un decret, ordenés l'aturada\", \"label\": \"ca\"},\n {\"text\": \"to, co nás Evropany spojuje, než to, co nás od\", \"label\": \"cs\"},\n {\"text\": \"ergänzt durch einen gesetzlich verankertes\", \"label\": \"de\"},\n {\"text\": \"Насчитал, что с начала года всего три дня были\", \"label\": \"ru\"},\n {\"text\": \"Borisovu tražeći od njega da prihvati njenu\", \"label\": \"sr\"},\n {\"text\": \"la presenza di ben veleni diversi: . chili di\", \"label\": \"it\"},\n {\"text\": \"καὶ τῶν ἐκλεκτῶν ἀγγέλων ἵνα ταῦτα φυλάξῃς χωρὶς\", \"label\": \"grc\"},\n {\"text\": \"pretraživale obližnju bolnicu i stambene zgrade u\", \"label\": \"hr\"},\n {\"text\": \"An rund Katzen habe Wolf seine Spiele getestet ,\", \"label\": \"de\"},\n {\"text\": \"investigating since March.\", \"label\": \"en\"},\n {\"text\": \"Tonböden (Mullböden).\", \"label\": \"de\"},\n {\"text\": \"Stálý dopisovatel LN v SRN Bedřich Utitz\", \"label\": \"cs\"},\n {\"text\": \"červnu předložené smlouvy.\", \"label\": \"cs\"},\n {\"text\": \"πνεύματι ᾧ ἐλάλει\", \"label\": \"grc\"},\n {\"text\": \".%의 신장세를 보였다.\", \"label\": \"ko\"},\n {\"text\": \"Foae verde, foi de nuc, Prin pădure, prin colnic,\", \"label\": \"ro\"},\n {\"text\": \"διαπέμψας ἄλλους ἄλλῃ τοὺς μὲν ἐς Δελφοὺς ἰέναι\", \"label\": \"grc\"},\n {\"text\": \"المسلمين أو أي تيار سياسي طالما عمل ذلك التيار في\", \"label\": \"ar\"},\n {\"text\": \"As informações são da Dow Jones.\", \"label\": \"pt\"},\n {\"text\": \"Milliarde DM ausgestattet sein .\", \"label\": \"de\"},\n {\"text\": \"De utgår fortfarande från att kvinnans jämlikhet\", \"label\": \"sv\"},\n {\"text\": \"Sneeuw maakte in Davos bij de voorbereiding een\", \"label\": \"nl\"},\n {\"text\": \"De ahí que en este mercado puedan negociarse\", \"label\": \"es\"},\n {\"text\": \"intenzívnějšímu sbírání a studiu.\", \"label\": \"cs\"},\n {\"text\": \"और औसकर ४.० पैकेज का प्रयोग किया गया है ।\", \"label\": \"hi\"},\n {\"text\": \"Adipati Kuningan karena Kuningan menjadi bagian\", \"label\": \"id\"},\n {\"text\": \"Svako je bar jednom poželeo da mašine prosto umeju\", \"label\": \"sr\"},\n {\"text\": \"Im vergangenen Jahr haben die Regierungen einen\", \"label\": \"de\"},\n {\"text\": \"durat motus, aliquid fit et non est;\", \"label\": \"la\"},\n {\"text\": \"Dominować będą piosenki do tekstów Edwarda\", \"label\": \"pl\"},\n {\"text\": \"beantwortet .\", \"label\": \"de\"},\n {\"text\": \"О гуманитариях было кому рассказывать, а вот за\", \"label\": \"ru\"},\n {\"text\": \"Helsingin kaupunki riitautti vuokrasopimuksen\", \"label\": \"fi\"},\n {\"text\": \"chợt tan biến.\", \"label\": \"vi\"},\n {\"text\": \"avtomobil ločuje od drugih.\", \"label\": \"sl\"},\n {\"text\": \"Congress has proven itself ineffective as a body.\", \"label\": \"en\"},\n {\"text\": \"मैक्सिको ने इस तरह का शो इस समय आयोजित करने का\", \"label\": \"hi\"},\n {\"text\": \"No minimum order amount.\", \"label\": \"en\"},\n {\"text\": \"Convertassa .\", \"label\": \"fi\"},\n {\"text\": \"Как это можно сделать?\", \"label\": \"ru\"},\n {\"text\": \"tha mi creidsinn gu robh iad ceart cho saor shuas\", \"label\": \"gd\"},\n {\"text\": \"실제 일제는 이런 만해의 논리를 묵살하고 한반도를 침략한 다음 , 이어 만주를 침략하고\", \"label\": \"ko\"},\n {\"text\": \"Da un semplice richiamo all'ordine fino a grandi\", \"label\": \"it\"},\n {\"text\": \"pozoruhodný nejen po umělecké stránce, jež\", \"label\": \"cs\"},\n {\"text\": \"La comida y el servicio aprueban.\", \"label\": \"es\"},\n {\"text\": \"again, connected not with each other but to the\", \"label\": \"en\"},\n {\"text\": \"Protokol výslovně stanoví, že nikdo nemůže být\", \"label\": \"cs\"},\n {\"text\": \"ఒక విషయం అడగాలని ఉంది .\", \"label\": \"te\"},\n {\"text\": \"Безгранично почитая дирекцию, ловя на лету каждое\", \"label\": \"ru\"},\n {\"text\": \"rovnoběžných růstových vrstev, zůstávají krychlové\", \"label\": \"cs\"},\n {\"text\": \"प्रवेश और पूर्व प्रधानमंत्री लाल बहादुर शास्त्री\", \"label\": \"hi\"},\n {\"text\": \"Bronzen medaille in de Europese marathon.\", \"label\": \"nl\"},\n {\"text\": \"- gadu vecumā viņi to nesaprot.\", \"label\": \"lv\"},\n {\"text\": \"Realizó sus estudios primarios en la Escuela Julia\", \"label\": \"es\"},\n {\"text\": \"cuartos de final, su clasificación para la final a\", \"label\": \"es\"},\n {\"text\": \"Sem si pro něho přiletí americký raketoplán, na\", \"label\": \"cs\"},\n {\"text\": \"Way to go!\", \"label\": \"en\"},\n {\"text\": \"gehört der neuen SPD-Führung unter Parteichef\", \"label\": \"de\"},\n {\"text\": \"Somit simuliert der Player mit einer GByte-Platte\", \"label\": \"de\"},\n {\"text\": \"Berufung auf kommissionsnahe Kreise , die bereits\", \"label\": \"de\"},\n {\"text\": \"Dist Clarïen\", \"label\": \"fro\"},\n {\"text\": \"Schon nach den Gerüchten , die Telekom wolle den\", \"label\": \"de\"},\n {\"text\": \"Software von NetObjects ist nach Angaben des\", \"label\": \"de\"},\n {\"text\": \"si enim per legem iustitia ergo Christus gratis\", \"label\": \"la\"},\n {\"text\": \"ducerent in ipsam magis quam in corpus christi,\", \"label\": \"la\"},\n {\"text\": \"Neustar-Melbourne-IT-Partnerschaft NeuLevel .\", \"label\": \"de\"},\n {\"text\": \"forderte dagegen seine drastische Verschärfung.\", \"label\": \"de\"},\n {\"text\": \"pemmican på hundrede forskellige måder.\", \"label\": \"da\"},\n {\"text\": \"Lehån, själv matematiklärare, visar hur den nya\", \"label\": \"sv\"},\n {\"text\": \"I highly recommend his shop.\", \"label\": \"en\"},\n {\"text\": \"verità, giovani fedeli prostratevi #amen\", \"label\": \"it\"},\n {\"text\": \"उत्तर प्रदेश के अध्यक्ष पद से हटाए गए विनय कटियार\", \"label\": \"hi\"},\n {\"text\": \"() روزی مےں کشادگی ہوتی ہے۔\", \"label\": \"ur\"},\n {\"text\": \"Prozessorgeschäft profitieren kann , stellen\", \"label\": \"de\"},\n {\"text\": \"školy začalo počítat pytle s moukou a zjistilo, že\", \"label\": \"cs\"},\n {\"text\": \"प्रभावशाली पर गैर सरकारी लोगों के घरों में भी\", \"label\": \"hi\"},\n {\"text\": \"geschichtslos , oder eine Farce , wie sich\", \"label\": \"de\"},\n {\"text\": \"Ústrednými mocnosťami v marci však spôsobilo, že\", \"label\": \"sk\"},\n {\"text\": \"التسليح بدون مبرر، واستمرار الأضرار الناجمة عن فرض\", \"label\": \"ar\"},\n {\"text\": \"Například Pedagogická fakulta Univerzity Karlovy\", \"label\": \"cs\"},\n {\"text\": \"nostris ut eriperet nos de praesenti saeculo\", \"label\": \"la\"}]\n \n docs = [Document([], text=example[\"text\"]) for example in examples]\n gold_labels = [example[\"label\"] for example in examples]\n basic_multilingual(docs)\n accuracy = sum([(doc.lang == label) for doc,label in zip(docs,gold_labels)])/len(docs)\n assert accuracy >= 0.98","def test_translation_smoke():\n english_to_morse = get_translator(\"english\", \"morse\")\n morse_to_english = get_translator(\"morse\", \"english\")\n morse = english_to_morse.translate(\"hello world\")\n english = morse_to_english.translate(morse)\n assert english == \"HELLO WORLD\"","def showTranslatedWithoutJoin(cls):\n print (\"ALL WORDS WITH TRANSLATIONS STORED IN DATABASE:\")\n for word1 in EnglishHelper.query(\"SELECT english_word FROM EnglishWords\", fetchAll=True):\n try:\n print word1[0],\" - \", (EnglishHelper.query(\"select polish_word from PolishWords where \"\n \" id_pl=(select id_pl from translations where \"\n \"id_eng = (select id_eng from EnglishWords \"\n \"where english_word = '%s'))\"%word1))[0].encode('utf-8')\n except:\n print \"There is no translation, sorry :(\"","def test_single_locale_activation(self):\n with translation.override(\"fr\"):\n self.assertEqual(\n self.get_template(\n \"{% load i18n %}{% blocktranslate %}Yes{% endblocktranslate %}\"\n ).render(Context({})),\n \"Oui\",\n )","def test_i18n02(self):\n output = self.engine.render_to_string('i18n02')\n self.assertEqual(output, 'xxxyyyxxx')","def test_i18n08(self):\n output = self.engine.render_to_string('i18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_i18n08(self):\n output = self.engine.render_to_string('i18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_i18n08(self):\n output = self.engine.render_to_string(\"i18n08\", {\"number\": 2})\n self.assertEqual(output, \"2 plural\")","def test_i18n14(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n14')\n self.assertEqual(output, 'foo Passwort Passwort')","def test_legacyi18n08(self):\n output = self.engine.render_to_string(\"legacyi18n08\", {\"number\": 2})\n self.assertEqual(output, \"2 plural\")","def test_legacyi18n04(self):\n output = self.engine.render_to_string(\"legacyi18n04\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"å\")","def test_i18n12(self):\n output = self.engine.render_to_string('i18n12')\n self.assertEqual(output, 'de')","def test_legacyi18n08(self):\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_legacyi18n08(self):\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def stats_text(test):\n\n stats_text_en(test) \n \n stats_text_cn(test)","def test_fallback_language_no_current(self):\n x = SimpleModel()\n x.set_current_language(self.conf_fallback)\n x.tr_title = \"TITLE_FALLBACK\"\n\n self.assertEqual(\n x.safe_translation_getter(\"tr_title\", language_code=self.other_lang1), \"TITLE_FALLBACK\"\n )","def test_language_fix(self):\n #TODO\n \n for lang in self.LANGUAGES:\n activate(lang)\n \n self.assertEqual(lang, get_language())","def test_get_languages_multiple(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n translation = StoryTranslation(story=story, title=\"Spanish Title\",\n summary=\"Spanish Summary\", language=\"es\")\n translation.save()\n self.assertEqual([settings.LANGUAGE_CODE, 'es'], story.get_languages())","def translate_wrapper(atext):\n print(\"translating:\",atext)\n res=\"\"\n res=translate(atext,\"pl\",\"fr\")\n time.sleep(0.5)\n print(\"translation:\",res)\n return res","def test_english_gb_no_translation(self):\n resp = ResponseFactory(\n locale=u'en-GB',\n description=u'hello',\n translated_description=u''\n )\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_english_with_dennis(self):\n resp = ResponseFactory(\n locale=u'en-US',\n product=u'firefox',\n description=u'hello',\n translated_description=u''\n )\n\n # Set the product up for translation *after* creating the response\n # so that it doesn't get auto-translated because Response is set up\n # for auto-translation.\n prod = Product.objects.get(db_name='firefox')\n prod.translation_system = u'dennis'\n prod.save()\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_get_translation_resources(self):\n pass","def test_i18n05(self):\n output = self.engine.render_to_string(\"i18n05\", {\"anton\": \"yyy\"})\n self.assertEqual(output, \"xxxyyyxxx\")","def test_get_languages_multiple(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n translation = StoryTranslation(story=story, title=\"Spanish Title\",\n summary=\"Spanish Summary\", language=\"es\")\n translation.save()\n story_languages = story.get_languages()\n self.assertEqual(len(story_languages), 2)\n for code in (settings.LANGUAGE_CODE, 'es'):\n self.assertIn(code, story_languages)","def test_single_locale_activation(self):\n with translation.override('fr'):\n self.assertEqual(\n Template(\"{% load i18n %}{% blocktrans %}Yes{% endblocktrans %}\").render(Context({})),\n 'Oui'\n )","def test_00_i18n_anonymous(self):\r\n # First default 'en' locale\r\n with self.app as c:\r\n err_msg = \"The page should be in English\"\r\n res = c.get('/', headers=[('Accept-Language', 'en')])\r\n assert \"Community\" in res.data, err_msg\r\n # Second with 'es' locale\r\n with self.app as c:\r\n err_msg = \"The page should be in Spanish\"\r\n res = c.get('/', headers=[('Accept-Language', 'es')])\r\n assert \"Comunidad\" in res.data, err_msg","def test_functionallity(self):\n\n pp = Lexpp(external_dict=pkg_resources.resource_filename(\"lexpp\", \"tests/test.dict\"))\n\n test_word = \"キャプテン\"\n entries = list(pp.lookup(test_word))\n\n self.assertEqual(len(entries), 4)\n\n for e in entries:\n self.assertEqual(type(e), Entry)\n rep = pp.get_representative_form(e)\n self.assertEqual(rep, test_word)","def test_get_languages(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())","def test_get_languages(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())","def test_i18n05(self):\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\n self.assertEqual(output, 'xxxyyyxxx')","def test_i18n05(self):\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\n self.assertEqual(output, 'xxxyyyxxx')","def test_fallback_variant(self):\n x = SimpleModel()\n\n x.set_current_language(\"de\")\n x.tr_title = \"Hallo-de\"\n\n x.set_current_language(\"en\")\n x.tr_title = \"Hello-en\"\n\n x.save()\n\n with translation.override(\"de-ch\"):\n x = SimpleModel.objects.get(pk=x.pk)\n self.assertEqual(x.tr_title, \"Hallo-de\")","def translate(self):\n pass","def _load_transliterated_regional(self):\n# global approved, conflicts, suggestions, unknown, cldr, current\n start = self.lblFallback['text'].find('=>') + 2\n if self.lblFallback['text'][start:]:\n self.preferred.set(\\\n self._transliterate_text(self.lblFallback['text'][start:]))\n pass","def test_i18n15(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n15', {'absent': ''})\n self.assertEqual(output, 'Passwort')","def test_i18n04(self):\n output = self.engine.render_to_string('i18n04', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'å')","def test_i18n04(self):\n output = self.engine.render_to_string('i18n04', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'å')","def test_legacyi18n04(self):\n output = self.engine.render_to_string('legacyi18n04', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'å')","def test_legacyi18n04(self):\n output = self.engine.render_to_string('legacyi18n04', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'å')","def test_list_unicode_locale(self):\n response = self.client.get(self.url, {'lang': 'ru'})\n assert response.data\n\n assert response.data['count'] == len(discopane_items)\n assert response.data['next'] is None\n assert response.data['previous'] is None\n assert response.data['results']","def test_fallback_values_1(self):\n title1_de = \"title de\"\n n = TestModelWithFallback()\n n.title = title1_de\n n.save()\n del n\n n = TestModelWithFallback.objects.get(title=title1_de)\n self.failUnlessEqual(n.title, title1_de)\n trans_real.activate(\"en\")\n self.failUnlessEqual(n.title, \"\")","def goToTranslate():\n\n try:\n global driver\n\n options = Options()\n\n # Code to disable notifications pop up of Chrome Browser\n options.add_argument(\"--disable-notifications\")\n options.add_argument(\"--disable-infobars\")\n options.add_argument(\"--mute-audio\")\n # options.add_argument(\"headless\")\n\n try:\n platform_ = platform.system().lower()\n if platform_ in ['linux', 'darwin']:\n driver = webdriver.Chrome(executable_path=\"./chromedriver\", options=options)\n else:\n driver = webdriver.Chrome(executable_path=\"./chromedriver.exe\", options=options)\n except:\n print(\"Kindly replace the Chrome Web Driver with the latest one from\"\n \"http://chromedriver.chromium.org/downloads\"\n \"\\nYour OS: {}\".format(platform_)\n )\n exit()\n\n driver.get(\"https://translate.google.com\") # go to facebook home page\n driver.maximize_window()\n\n time.sleep(30)\n\n try:\n\n f=open('C:\\\\Users\\\\Alvi Rahman\\\\Desktop\\\\Farhana Miss\\\\Data\\\\banglaTweets.txt',mode=\"r\", encoding=\"utf-8\")\n tweets=f.readlines()\n f.close()\n\n tweets=[s.replace('\\n','') for s in tweets]\n\n output_file=open(\"translated_tweet.txt\", mode=\"w\", encoding=\"utf-16\")\n\n inputBox=driver.find_element_by_id('source')\n\n tweet_count = 0\n for i in range (8001,12000):\n inputBox.send_keys(tweets[i])\n translatedTweet = \"xxxxxxxxxxxx\"\n time.sleep(10)\n\n try:\n\n\t TranslationBox=driver.find_element_by_xpath(\"/html/body/div[2]/div[1]/div[2]/div[1]/div[1]/div[2]/div[3]/div[1]/div[2]/div/span[1]/span\")\n\t translatedTweet = TranslationBox.text\n\n\t translatedTweet=translatedTweet.replace('\\b',' ')\n\t translatedTweet=translatedTweet.replace('\\n',' ')\n\t translatedTweet=translatedTweet.replace('\\t',' ')\n\t translatedTweet=translatedTweet.replace('\\r',' ')\n\t translatedTweet=translatedTweet.replace('\\f',' ')\n\t translatedTweet=reg.sub(' +',' ',translatedTweet)\n\n except:\n pass\n\n\n output_file.writelines(translatedTweet+\"\\n\")\n inputBox.clear()\n tweet_count = tweet_count + 1\n\n if tweet_count%100 ==0:\n time.sleep(270)\n\n elif tweet_count%50 ==0:\n time.sleep(150)\n\n elif tweet_count%10 ==0:\n time.sleep(60)\n\n output_file.close()\n driver.close()\n\n except:\n exc_type = sys.exc_info()[0]\n exc_tb = sys.exc_info()[2]\n print(exc_type, exc_tb.tb_lineno)\n pass\n\n # filling the form\n # driver.find_element_by_name('email').send_keys(email)\n # driver.find_element_by_name('pass').send_keys(password)\n\n # # clicking on login button\n # driver.find_element_by_id('loginbutton').click()\n\n print(\"Successfully logged in\")\n\n except Exception as e:\n print(\"There's some error in log in.\")\n print(sys.exc_info()[0])\n exit()","def test_english_no_translation(self):\n resp = ResponseFactory(\n locale=u'en-US',\n description=u'hello',\n translated_description=u''\n )\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_home_view_cyrillic(self):\n UserData.objects.create(\n first_name=\"Андрей\",\n last_name=\"Герасько\",\n date_of_birth='1998-02-23',\n bio=\"Био\",\n email=\"aaa@gmail.com\",\n jabber=\"aaaaa\",\n skype=\"aaaaa\",\n other_contacts=\"Контакты\"\n )\n response = self.client.get(self.url)\n self.assertEqual(response.status_code, 200)\n self.assertIn('42 Coffee Cups Test Assignment', response.content)\n self.assertIn('Андрей', response.content)\n self.assertIn('Герасько', response.content)\n self.assertIn('Date of birth', response.content)\n self.assertIn('Био', response.content)\n self.assertIn('Контакты', response.content)","def test_translate(self):\n body = Question()\n response = self.client.open('/api/rtx/v1/translate',\n method='POST',\n data=json.dumps(body),\n content_type='application/json')\n self.assert200(response, \"Response body is : \" + response.data.decode('utf-8'))","def test_get_supported_locales_for_voice_datasets(self):\n pass","def test_energy_consumption_top_ranking_ko(self):\n self.driver.get(\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko'\n )\n\n self.assertEqual(\n self.driver.find_elements_by_class_name('navbar-brand')[0].text,\n 'Data Commons 장소 순위')\n\n subtitle_present = EC.text_to_be_present_in_element(\n (By.TAG_NAME, 'h3'), 'the World 상위 국가 100개')\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(subtitle_present)\n self.assertEqual(\n self.driver.find_element_by_tag_name('h1').text, '1인당 에너지 소비량 순위')\n self.assertEqual(\n self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/h3/a').get_attribute('href'),\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko&bottom='\n )\n\n table = self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/table')\n headers = table.find_elements_by_xpath('.//thead/tr/th')\n self.assertEqual(headers[0].text, '순위')\n self.assertEqual(headers[1].text, '국가')\n self.assertEqual(headers[2].text, '킬로그램')\n row = table.find_elements_by_xpath('.//tbody/tr[1]/td')\n self.assertEqual(row[0].text, '1')\n self.assertEqual(row[1].text, '카타르')\n self.assertEqual(\n row[1].find_element_by_tag_name('a').get_attribute('href'),\n self.url_ + '/place/country/QAT?hl=ko')\n\n chart = self.driver.find_element_by_id('ranking-chart')\n y_text = chart.find_elements_by_class_name(\n 'y')[0].find_elements_by_tag_name('text')\n self.assertEqual(y_text[0].text, '0kg')\n self.assertEqual(y_text[-1].text, '1.5만kg')\n\n x_text = chart.find_elements_by_class_name(\n 'x')[0].find_elements_by_tag_name('text')\n self.assertEqual(x_text[0].text, '카타르')\n self.assertEqual(x_text[-1].text, '에콰도르')","def test_i18n07(self):\n output = self.engine.render_to_string('i18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_i18n07(self):\n output = self.engine.render_to_string('i18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_i18n07(self):\n output = self.engine.render_to_string(\"i18n07\", {\"number\": 1})\n self.assertEqual(output, \"singular\")","def test_shipped_locales(self):\n url = reverse('shipping.views.status.shipped_locales')\n url += '?av=fx1.0'\n response = self.client.get(url)\n eq_(response.status_code, 200)\n eq_(response.content, \"\"\"da\nde\nen-US\n\"\"\")","def test_legacyi18n07(self):\n output = self.engine.render_to_string(\"legacyi18n07\", {\"number\": 1})\n self.assertEqual(output, \"singular\")","def test_legacyi18n07(self):\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_legacyi18n07(self):\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_i18n03(self):\n output = self.engine.render_to_string('i18n03', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'Å')","def test_i18n03(self):\n output = self.engine.render_to_string('i18n03', {'anton': b'\\xc3\\x85'})\n self.assertEqual(output, 'Å')","def test_get_unique_langs(self):\n\n contacts = Contact.query.filter_by(user_id=6).all()\n\n unique_lang_dict = MessageController.get_unique_langs(contacts)\n self.assertEqual(unique_lang_dict, {4: u'fr'})","def _test (self) :\n print (_T (ckw.title or \"Baz\"))\n print (_T (\"Foo\"))\n foo = _(\"Markup %d\")\n print (_T(foo) % 42)\n print (_Tn (\"Singular\", \"Plural\", 4))","def test_i18n26(self):\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\n self.assertEqual(output, 'singular test')","def test_i18n26(self):\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\n self.assertEqual(output, 'singular test')","def test_spanish_with_dennis_and_existing_translations(self):\n existing_resp = ResponseFactory(\n locale=u'es',\n product=u'firefox',\n description=u'hola',\n translated_description=u'DUDE!'\n )\n\n resp = ResponseFactory(\n locale=u'es',\n product=u'firefox',\n description=u'hola',\n translated_description=u''\n )\n\n # Set the product up for translation *after* creating the response\n # so that it doesn't get auto-translated because Response is set up\n # for auto-translation.\n prod = Product.objects.get(db_name='firefox')\n prod.translation_system = u'dennis'\n prod.save()\n\n # No jobs should be translated\n eq_(len(resp.generate_translation_jobs()), 0)\n eq_(resp.translated_description, existing_resp.translated_description)","def test_spanish_no_translation(self):\n resp = ResponseFactory(\n locale=u'es',\n product=u'firefox',\n description=u'hola',\n translated_description=u''\n )\n\n # No jobs should be translated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Nothing should be translated\n eq_(resp.translated_description, u'')","def test_homepage_en(self):\n\n self.driver.get(self.url_ + '/')\n\n title_present = EC.text_to_be_present_in_element(\n (By.XPATH, '//*[@id=\"main-nav\"]/div/div[1]/a'), 'Data Commons')\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(title_present)\n\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons is an open knowledge repository'))\n\n explore_callout_msg = self.driver.find_element_by_xpath(\n '//*[@id=\"homepage\"]/section[3]/p')\n self.assertTrue(\n explore_callout_msg.text.startswith(\n 'We cleaned and processed the data so you don\\'t have to'))\n\n nyc_health = self.driver.find_element_by_xpath(\n '//*[@id=\"homepage\"]/section[3]/ul/li[1]/ul/li[2]/a')\n self.assertEqual(nyc_health.text, 'New York City, NY Health')\n self.assertEqual(nyc_health.get_attribute('href'),\n self.url_ + '/place/geoId/3651000?topic=Health')\n\n schema_org = self.driver.find_element_by_xpath(\n '//*[@id=\"homepage\"]/section[5]/ul/li[2]')\n self.assertEqual(schema_org.text,\n 'Open sourced, built using Schema.org.')\n\n more_msg = self.driver.find_element_by_xpath(\n '//*[@id=\"homepage\"]/section[3]/ul/li[3]/ul/li[4]/a')\n self.assertEqual(more_msg.text, 'more ...')","def try_translation(text):\n try:\n trad = Translator().translate(text = text, dest = 'fr')\n except:\n sleep(random() * 60)\n return False\n \n return trad.text","def translation():\n class TranslationTable(tables.Table):\n normal = tables.Column(verbose_name=ugettext(\"Normal\"))\n lazy = tables.Column(verbose_name=ugettext(\"Lazy\"))\n\n table = TranslationTable([])\n Assert(\"Normal\") == table.columns[\"normal\"].header\n Assert(\"Lazy\") == table.columns[\"lazy\"].header"],"string":"[\n \"def test_i18n10(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n10', {'bool': True})\\n self.assertEqual(output, 'Ja')\",\n \"def test_i18n27(self):\\n with translation.override(\\\"ru\\\"):\\n output = self.engine.render_to_string(\\\"i18n27\\\", {\\\"number\\\": 1})\\n self.assertEqual(\\n output, \\\"1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442\\\"\\n )\",\n \"def test_content(google_translator):\\n # from bs4 import BeautifulSoup\\n # assert 'GitHub' in BeautifulSoup(response.content).title.string\\n assert google_translator.translate(text='좋은') == \\\"good\\\"\",\n \"def test_i18n27(self):\\n with translation.override('ru'):\\n output = self.engine.render_to_string('i18n27', {'number': 1})\\n self.assertEqual(output, '1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442')\",\n \"def test_i18n27(self):\\n with translation.override('ru'):\\n output = self.engine.render_to_string('i18n27', {'number': 1})\\n self.assertEqual(output, '1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442')\",\n \"def test_i18n09(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n09')\\n self.assertEqual(output, 'Page not found')\",\n \"def setUp(self):\\n super().setUp()\\n translation.activate(\\\"en-us\\\")\",\n \"def test_i18n23(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n23')\\n self.assertEqual(output, 'nicht gefunden')\",\n \"def test_i18n16(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n16')\\n self.assertEqual(output, '<')\",\n \"def test_i18n(self):\\n # If you want to modify this code, please get acquainted with\\n # Python's locale module. In particular:\\n # http://docs.python.org/library/locale.html#locale.getdefaultlocale\\n\\n # Set the standard C locale.\\n os.environ['LANG'] = 'C'\\n os.environ['LC_ALL'] = 'C.UTF-8'\\n\\n #must be after above\\n out1 = tr('Hello!')\\n expected1 = 'Hello!'\\n msg = 'Expected %s, got %s' % (expected1, out1)\\n assert out1 == expected1, msg\\n\\n # Set the Indonesian locale to test translations.\\n os.environ['LANG'] = 'id'\\n os.environ['LC_ALL'] = 'id_ID.UTF-8'\\n\\n #must be after above\\n #indoout1 = tr('Hello!') # translate as 'Hi'\\n #indoexpected1 = 'Hi!'\\n #msg = 'Expected %s, got %s' % (indoexpected1, indoout1)\\n #assert indoout1 == indoexpected1, msg\",\n \"def use_en(self):\\n pass\",\n \"def test_i18n06(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n06')\\n self.assertEqual(output, 'Seite nicht gefunden')\",\n \"def test_issue3625():\\n nlp = Hindi()\\n doc = nlp(u\\\"hi. how हुए. होटल, होटल\\\")\\n assert [token.text for token in doc] == ['hi', '.', 'how', 'हुए', '.', 'होटल', ',', 'होटल']\",\n \"def test_default_translations(self):\\n\\t\\t\\n\\t\\tself.assertTrue(data.get_default_translation('Catholicism', 3) == 'DRA')\\n\\t\\tself.assertTrue(data.get_default_translation('Christianity', 3) == 'ESV')\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string(\\\"i18n28\\\", {\\\"anton\\\": \\\"α\\\", \\\"berta\\\": \\\"β\\\"})\\n self.assertEqual(output, \\\"α + β\\\")\",\n \"def use_zh(self):\\n pass\",\n \"def test_hero_all_langs(self):\\n\\n self.driver.get(self.url_ + '/?hl=de')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons ist ein Open Knowledge Repository'))\\n\\n self.driver.get(self.url_ + '/?hl=es')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons es un repositorio abierto'))\\n\\n self.driver.get(self.url_ + '/?hl=fr')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons est un dépôt de connaissances'))\\n\\n self.driver.get(self.url_ + '/?hl=hi')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons, एक ऐसा प्रोजेक्ट है जिसमें डेटा'))\\n\\n self.driver.get(self.url_ + '/?hl=ja')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(hero_msg.text.startswith('データコモンズは、マッピングされた共通'))\\n\\n self.driver.get(self.url_ + '/?hl=ko')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons는 매핑된 공통 항목을 사용해'))\\n\\n self.driver.get(self.url_ + '/?hl=ru')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons – это открытая база данных'))\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\\n self.assertEqual(output, 'α + β')\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\\n self.assertEqual(output, 'α + β')\",\n \"def get_translation(self):\",\n \"def test_language_sensitivity(self): \\n \\n for lang in self.LANGUAGES:\\n activate(lang)\\n \\n self.assertEqual(get_wording_text('test_1'), lang)\",\n \"def test_i18n11(self):\\n output = self.engine.render_to_string('i18n11', {'bool': True})\\n self.assertEqual(output, 'ja')\",\n \"def question_new_translate():\",\n \"def test_translate_unique_langs(self):\\n\\n trans_msgs_dict = MessageController.translate_unique_langs({'2': 'es', '4': 'fr'}, \\n 'hi', 'en', False, False)\\n\\n self.assertEqual(trans_msgs_dict, {'es': u'{hola}', 'fr': u'salut'})\",\n \"def test_international(self):\\n\\n test_cases = (\\n makeSimpleTestCase('the beatles'),\\n makeSimpleTestCase('david guetta'),\\n makeSimpleTestCase('susan boyle'),\\n makeSimpleTestCase('rammstein'),\\n makeSimpleTestCase('deadmau5'),\\n makeTestCase('deadmau 5', Matcher(title=Equals('deadmau5'))),\\n makeSimpleTestCase('2face idibia'),\\n makeSimpleTestCase('angelique kidjo'),\\n makeSimpleTestCase('gotye'),\\n makeSimpleTestCase('michel telo'),\\n makeSimpleTestCase('stereo total'),\\n makeSimpleTestCase('yelle'),\\n makeSimpleTestCase('stefanie heinzmann'),\\n makeSimpleTestCase('olly murs')\\n )\\n\\n self._run_tests('i18n_artist', test_cases)\",\n \"def test_i18n01(self):\\n output = self.engine.render_to_string('i18n01')\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_word_translation(self):\\n self.assertEqual(translator.translate_word(\\\"hour\\\"), \\\"ourhay\\\")\\n self.assertEqual(translator.translate_word(\\\"\\\"), \\\"\\\")\\n self.assertEqual(translator.translate_word(\\\"aaa\\\"), \\\"aaayay\\\")\",\n \"def translate():\\n pass\",\n \"def test_i18n03(self):\\n output = self.engine.render_to_string(\\\"i18n03\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"Å\\\")\",\n \"def test_ukrainian_symbols(self):\\n string = \\\"Минає ніч від’їзду\\\"\\n expected = \\\"Minaye nich vid’yizdu\\\"\\n self.assertEqual(transliterate(string), expected)\",\n \"def test_fallback_language(self):\\n x = SimpleModel()\\n x.set_current_language(self.conf_fallback)\\n x.tr_title = \\\"TITLE_FALLBACK\\\"\\n\\n x.set_current_language(self.other_lang1)\\n x.tr_title = \\\"TITLE_XX\\\"\\n x.save()\\n\\n with translation.override(self.other_lang2):\\n x = SimpleModel.objects.get(pk=x.pk)\\n self.assertEqual(x.tr_title, \\\"TITLE_FALLBACK\\\")\",\n \"def test_fallback_values_2(self):\\n title1_de = \\\"title de\\\"\\n text1_de = \\\"text in german\\\"\\n n = TestModelWithFallback2()\\n n.title = title1_de\\n n.text = text1_de\\n n.save()\\n del n\\n n = TestModelWithFallback2.objects.get(title=title1_de)\\n trans_real.activate(\\\"en\\\")\\n self.failUnlessEqual(n.title, title1_de)\\n self.failUnlessEqual(n.text,\\\\\\n TestTranslationOptionsWithFallback2.fallback_values['text'])\",\n \"def translate(self, language=None):\",\n \"def test_i18n13(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n13')\\n self.assertEqual(output, 'Passwort')\",\n \"def test_i18n04(self):\\n output = self.engine.render_to_string(\\\"i18n04\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"å\\\")\",\n \"def test_langid_benchmark(basic_multilingual):\\n examples = [\\n {\\\"text\\\": \\\"contingentiam in naturalibus causis.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"I jak opowiadał nieżyjący już pan Czesław\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Sonera gilt seit längerem als Übernahmekandidat\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"与银类似，汞也可以与空气中的硫化氢反应。\\\", \\\"label\\\": \\\"zh-hans\\\"},\\n {\\\"text\\\": \\\"contradictionem implicat.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Bis zu Prozent gingen die Offerten etwa im\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"inneren Sicherheit vorgeschlagene Ausweitung der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Multimedia-PDA mit Mini-Tastatur\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ponášalo sa to na rovnicu o dvoch neznámych.\\\", \\\"label\\\": \\\"sk\\\"},\\n {\\\"text\\\": \\\"이처럼 앞으로 심판의 그 날에 다시 올 메시아가 예수 그리스도이며 , 그는 모든 인류의\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Die Arbeitsgruppe bedauert , dass der weit über\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"И только раз довелось поговорить с ним не вполне\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"de a-l lovi cu piciorul și conștiința că era\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"relación coas pretensións do demandante e que, nos\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"med petdeset in sedemdeset\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Catalunya; el Consell Comarcal del Vallès Oriental\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"kunnen worden.\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"Witkin je ve většině ohledů zcela jiný.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"lernen, so zu agieren, dass sie positive oder auch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"olurmuş...\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"sarcasmo de Altman, desde as «peruas» que discutem\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"خلاف فوجداری مقدمہ درج کرے۔\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"Norddal kommune :\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"dem Windows-.-Zeitalter , soll in diesem Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"przeklętych ucieleśniają mit poety-cygana,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"We do not believe the suspect has ties to this\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"groziņu pīšanu.\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Senior Vice-President David M. Thomas möchte\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"neomylně vybral nějakou knihu a začetl se.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Statt dessen darf beispielsweise der Browser des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"outubro, alcançando R $ bilhões em .\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"(Porte, ), as it does other disciplines\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"uskupení se mylně domnívaly, že podporu\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Übernahme von Next Ende an dem System herum , das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"No podemos decir a la Hacienda que los alemanes\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"и рѣста еи братья\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"الذي اتخذ قرارا بتجميد اعلان الدولة الفلسطينية\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"uurides Rootsi sõjaarhiivist toodud . sajandi\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"selskapets penger til å pusse opp sin enebolig på\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"средней полосе и севернее в Ярославской,\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"il-massa żejda fil-ġemgħat u superġemgħat ta'\\\", \\\"label\\\": \\\"mt\\\"},\\n {\\\"text\\\": \\\"The Global Beauties on internetilehekülg, mida\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"이스라엘 인들은 하나님이 그 큰 팔을 펴 이집트 인들을 치는 것을 보고 하나님을 두려워하며\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Snad ještě dodejme jeden ekonomický argument.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Spalio d. vykusiame pirmajame rinkimų ture\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"und schlechter Journalismus ein gutes Geschäft .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Du sodiečiai sėdi ant potvynio apsemtų namų stogo.\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"цей є автентичним.\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"Și îndegrabă fu cu îngerul mulțime de șireaguri\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"sobra personal cualificado.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Tako se u Njemačkoj dvije trećine liječnika služe\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Dual-Athlon-Chipsatz noch in diesem Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"यहां तक कि चीन के चीफ ऑफ जनरल स्टाफ भी भारत का\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Li forestier du mont avale\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Netzwerken für Privatanwender zu bewundern .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"만해는 승적을 가진 중이 결혼할 수 없다는 불교의 계율을 시대에 맞지 않는 것으로 보았다\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"balance and weight distribution but not really for\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"og så e # tente vi opp den om morgonen å sfyrte\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"변화는 의심의 여지가 없는 것이지만 반면에 진화는 논쟁의 씨앗이다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"puteare fac aceastea.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Waitt seine Führungsmannschaft nicht dem\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"juhtimisega, tulid sealt.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Veränderungen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"banda en el Bayer Leverkusen de la Bundesliga de\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"В туже зиму посла всеволодъ сн҃а своѥго ст҃ослава\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"пославъ приведе я мастеры ѿ грекъ\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"En un nou escenari difícil d'imaginar fa poques\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"καὶ γὰρ τινὲς αὐτοὺς εὐεργεσίαι εἶχον ἐκ Κροίσου\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"직접적인 관련이 있다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"가까운 듯하면서도 멀다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Er bietet ein ähnliches Leistungsniveau und\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"民都洛水牛是獨居的，並不會以群族聚居。\\\", \\\"label\\\": \\\"zh-hant\\\"},\\n {\\\"text\\\": \\\"την τρομοκρατία.\\\", \\\"label\\\": \\\"el\\\"},\\n {\\\"text\\\": \\\"hurbiltzen diren neurrian.\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Ah dimenticavo, ma tutta sta caciara per fare un\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"На первом этапе (-) прошла так называемая\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"of games are on the market.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"находится Мост дружбы, соединяющий узбекский и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"lessié je voldroie que li saint fussent aporté\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Дошла очередь и до Гималаев.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"vzácným suknem táhly pouští, si jednou chtěl do\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"E no terceiro tipo sitúa a familias (%), nos que a\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"وجابت دوريات امريكية وعراقية شوارع المدينة، فيما\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Jeg har bodd her i år .\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"Pohrozil, že odbory zostří postoj, pokud se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"tinham conseguido.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Nicht-Erkrankten einen Anfangsverdacht für einen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"permanece em aberto.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"questi possono promettere rendimenti fino a un\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Tema juurutatud kahevedurisüsteemita oleksid\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Поведение внешне простой игрушки оказалось\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Bundesländern war vom Börsenverein des Deutschen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"acció, 'a mesura que avanci l'estiu, amb l'augment\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Dove trovare queste risorse? Jay Naidoo, ministro\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"essas gordurinhas.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Im zweiten Schritt sollen im übernächsten Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"allveelaeva pole enam vaja, kuna külm sõda on läbi\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"उपद्रवी दुकानों को लूटने के साथ ही उनमें आग लगा\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"@user nella sfortuna sei fortunata ..\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"математических школ в виде грозовых туч.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"No cambiaremos nunca nuestra forma de jugar por un\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"dla tej klasy ani wymogów minimalnych, z wyjątkiem\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"en todo el mundo, mientras que en España consiguió\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"политики считать надежное обеспечение военной\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"gogoratzen du, genio alemana delakoaren\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Бычий глаз.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Opeření se v pravidelných obdobích obnovuje\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"I no és només la seva, es tracta d'una resposta\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"오경을 가르쳤다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Nach der so genannten Start-up-Periode vergibt die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Saulista huomasi jo lapsena , että hänellä on\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Министерство культуры сочло нецелесообразным, и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"znepřátelené tábory v Tádžikistánu předseda\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"καὶ ἦν ὁ λαὸς προσδοκῶν τὸν Ζαχαρίαν καὶ ἐθαύμαζον\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"Вечером, в продукте, этот же человек говорил о\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"lugar á formación de xuizos máis complexos.\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"cheaper, in the end?\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"الوزارة في شأن صفقات بيع الشركات العامة التي تم\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"tärkeintä elämässäni .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Виконання Мінських угод було заблоковано Росією та\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"Aby szybko rozpoznać żołnierzy desantu, należy\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Bankengeschäfte liegen vorn , sagte Strothmann .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"продолжение работы.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Metro AG plant Online-Offensive\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"nu vor veni, și să vor osîndi, aceia nu pot porni\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Ich denke , es geht in Wirklichkeit darum , NT bei\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"de turism care încasează contravaloarea\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Aurkaria itotzea da helburua, baloia lapurtu eta\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"com a centre de formació en Tecnologies de la\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"oportet igitur quod omne agens in agendo intendat\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Jerzego Andrzejewskiego, oparty na chińskich\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"sau một vài câu chuyện xã giao không dính dáng tới\\\", \\\"label\\\": \\\"vi\\\"},\\n {\\\"text\\\": \\\"что экономическому прорыву жесткий авторитарный\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"DRAM-Preisen scheinen DSPs ein\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Jos dajan nubbái: Mana!\\\", \\\"label\\\": \\\"sme\\\"},\\n {\\\"text\\\": \\\"toți carii ascultară de el să răsipiră.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"odpowiedzialności, które w systemie własności\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Dvomesečno potovanje do Mollenda v Peruju je\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"d'entre les agències internacionals.\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Fahrzeugzugangssysteme gefertigt und an viele\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"in an answer to the sharers' petition in Cuthbert\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Europa-Domain per Verordnung zu regeln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"#Balotelli. Su ebay prezzi stracciati per Silvio\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Ne na košickém trávníku, ale už včera v letadle se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"zaměstnanosti a investičních strategií.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Tatínku, udělej den\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"frecuencia con Mary.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Свеаборге.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"opatření slovenské strany o certifikaci nejvíce\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"En todas me decían: 'Espera que hagamos un estudio\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Die Demonstration sollte nach Darstellung der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ci vorrà un assoluto rigore se dietro i disavanzi\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Tatínku, víš, že Honzovi odešla maminka?\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Die Anzahl der Rechner wuchs um % auf und die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"האמריקאית על אדמת סעודיה עלולה לסבך את ישראל, אין\\\", \\\"label\\\": \\\"he\\\"},\\n {\\\"text\\\": \\\"Volán Egyesülés, a Közlekedési Főfelügyelet is.\\\", \\\"label\\\": \\\"hu\\\"},\\n {\\\"text\\\": \\\"Schejbala, který stejnou hru s velkým úspěchem\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"depends on the data type of the field.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Umsatzwarnung zu Wochenbeginn zeitweise auf ein\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"niin heti nukun .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Mobilfunkunternehmen gegen die Anwendung der so\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"sapessi le intenzioni del governo Monti e dell'UE\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Di chi è figlia Martine Aubry?\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"avec le reste du monde.\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Այդ մաքոքը ինքնին նոր չէ, աշխարհը արդեն մի քանի\\\", \\\"label\\\": \\\"hy\\\"},\\n {\\\"text\\\": \\\"și în cazul destrămării cenaclului.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"befriedigen kann , und ohne die auftretenden\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Κύκνον τ̓ ἐξεναρεῖν καὶ ἀπὸ κλυτὰ τεύχεα δῦσαι.\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"færdiguddannede.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"Schmidt war Sohn eines Rittergutsbesitzers.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"и вдаша попадь ѡпрати\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"cine nu știe învățătură”.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"détacha et cette dernière tenta de tuer le jeune\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Der har saka også ei lengre forhistorie.\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Pieprz roztłuc w moździerzu, dodać do pasty,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Лежа за гребнем оврага, как за бруствером, Ушаков\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"gesucht habe, vielen Dank nochmals!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"инструментальных сталей, повышения\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"im Halbfinale Patrick Smith und im Finale dann\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"البنوك التريث في منح تسهيلات جديدة لمنتجي حديد\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"una bolsa ventral, la cual se encuentra debajo de\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"za SETimes.\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"de Irak, a un piloto italiano que había violado el\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Er könne sich nicht erklären , wie die Zeitung auf\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Прохорова.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"la democrazia perde sulla tecnocrazia? #\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"entre ambas instituciones, confirmó al medio que\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Austlandet, vart det funne om lag førti\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"уровнями власти.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Dá tedy primáři úplatek, a často ne malý.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"brillantes del acto, al llevar a cabo en el\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"eee druga zadeva je majhen priročen gre kamorkoli\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Das ATX-Board paßt in herkömmliche PC-ATX-Gehäuse\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Za vodné bylo v prvním pololetí zaplaceno v ČR\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Даже на полсантиметра.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"com la del primer tinent d'alcalde en funcions,\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"кількох оповідань в цілості — щось на зразок того\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"sed ad divitias congregandas, vel superfluum\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Norma Talmadge, spela mot Valentino i en version\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Dlatego chciał się jej oświadczyć w niezwykłym\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"будут выступать на одинаковых снарядах.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Orang-orang terbunuh di sana.\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"لدى رايت شقيق اسمه أوسكار, وهو يعمل كرسام للكتب\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Wirklichkeit verlagerten und kaum noch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"как перемешивают костяшки перед игрой в домино, и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"В средине дня, когда солнце светило в нашу\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"d'aventure aux rôles de jeune romantique avec une\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"My teď hledáme organizace, jež by s námi chtěly\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Urteilsfähigkeit einbüßen , wenn ich eigene\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"sua appartenenza anche a voci diverse da quella in\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Aufträge dieses Jahr verdoppeln werden .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"M.E.: Miała szanse mnie odnaleźć, gdyby naprawdę\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"secundum contactum virtutis, cum careat dimensiva\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"ezinbestekoa dela esan zuen.\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Anek hurbiltzeko eskatzen zion besaulkitik, eta\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"perfectius alio videat, quamvis uterque videat\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Die Strecke war anspruchsvoll und führte unter\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"саморазоблачительным уроком, западные СМИ не\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"han representerer radikal islamisme .\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"Què s'hi respira pel que fa a la reforma del\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"previsto para também ser desconstruido.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Ὠκεανοῦ βαθυκόλποις ἄνθεά τ̓ αἰνυμένην, ῥόδα καὶ\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"para jovens de a anos nos Cieps.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"संघर्ष को अंजाम तक पहुंचाने का ऐलान किया है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"objeví i u nás.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"kvitteringer.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"This report is no exception.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Разлепват доносниците до избирателните списъци\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"anderem ihre Bewegungsfreiheit in den USA\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ñu tegoon ca kaw gor ña ay njotti bopp yu kenn\\\", \\\"label\\\": \\\"wo\\\"},\\n {\\\"text\\\": \\\"Struktur kann beispielsweise der Schwerpunkt mehr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"% la velocidad permitida, la sanción es muy grave.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Teles-Einstieg in ADSL-Markt\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ettekäändeks liiga suure osamaksu.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"als Indiz für die geänderte Marktpolitik des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"quod quidem aperte consequitur ponentes\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"de negociación para el próximo de junio.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Tyto důmyslné dekorace doznaly v poslední době\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"največjega uspeha doslej.\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Paul Allen je jedan od suosnivača Interval\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Federal (Seac / DF) eo Sindicato das Empresas de\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Quartal mit . Mark gegenüber dem gleichen Quartal\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"otros clubes y del Barça B saldrán varios\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Jaskula (Pol.) -\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"umožnily říci, že je možné přejít k mnohem\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"اعلن الجنرال تومي فرانكس قائد القوات الامريكية\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Telekom-Chef Ron Sommer und der Vorstandssprecher\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"My, jako průmyslový a finanční holding, můžeme\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"voorlichting onder andere betrekking kan hebben:\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"Hinrichtung geistig Behinderter applaudiert oder\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"wie beispielsweise Anzahl erzielte Klicks ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Intel-PC-SDRAM-Spezifikation in der Version . (\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"plângere în termen de zile de la comunicarea\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"и Испания ще изгубят втория си комисар в ЕК.\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"इसके चलते इस आदिवासी जनजाति का क्षरण हो रहा है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"aunque se mostró contrario a establecer un\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"des letzten Jahres von auf Millionen Euro .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ankara se također poziva da u cijelosti ratificira\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"herunterlädt .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"стрессовую ситуацию для организма, каковой\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Státního shromáždění (parlamentu).\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"diskutieren , ob und wie dieser Dienst weiterhin\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Verbindungen zu FPÖ-nahen Polizisten gepflegt und\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Pražského volebního lídra ovšem nevybírá Miloš\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nach einem Bericht der Washington Post bleibt das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"للوضع آنذاك، لكني في قرارة نفسي كنت سعيداً لما\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"не желаят запазването на статуквото.\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"Offenburg gewesen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ἐὰν ὑμῖν εἴπω οὐ μὴ πιστεύσητε\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"all'odiato compagno di squadra Prost, il quale\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"historischen Gänselieselbrunnens.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"למידע מלווייני הריגול האמריקאיים העוקבים אחר\\\", \\\"label\\\": \\\"he\\\"},\\n {\\\"text\\\": \\\"οὐδὲν ἄρα διαφέρεις Ἀμάσιος τοῦ Ἠλείου, ὃν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"movementos migratorios.\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"Handy und ein Spracherkennungsprogramm sämtliche\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Kümne aasta jooksul on Eestisse ohjeldamatult\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"H.G. Bücknera.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"protiv krijumčarenja, ili pak traženju ukidanja\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Topware-Anteile mehrere Millionen Mark gefordert\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Maar de mensen die nu over Van Dijk bij FC Twente\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"poidan experimentar as percepcións do interesado,\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"Miał przecież w kieszeni nóż.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Avšak žádná z nich nepronikla za hranice přímé\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"esim. helpottamalla luottoja muiden\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Podle předběžných výsledků zvítězila v\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nicht nur das Web-Frontend , auch die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Regierungsinstitutionen oder Universitäten bei\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Խուլեն Լոպետեգիին, պատճառաբանելով, որ վերջինս\\\", \\\"label\\\": \\\"hy\\\"},\\n {\\\"text\\\": \\\"Афганистана, где в последние дни идут ожесточенные\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"лѧхове же не идоша\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"Mit Hilfe von IBMs Chip-Management-Systemen sollen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\", als Manager zu Telefonica zu wechseln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"którym zajmuje się człowiek, zmienia go i pozwala\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"činí kyperských liber, to je asi USD.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Studienplätze getauscht werden .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"учёных, орнитологов признают вид.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"acordare a concediilor prevăzute de legislațiile\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"at større innsats for fornybar, berekraftig energi\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Politiet veit ikkje kor mange personar som deltok\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"offentligheten av unge , sinte menn som har\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"însuși în jurul lapunei, care încet DISPARE în\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"O motivo da decisão é evitar uma sobrecarga ainda\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"El Apostolado de la prensa contribuye en modo\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Teltow ( Kreis Teltow-Fläming ) ist Schmitt einer\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"grozījumus un iesniegt tos Apvienoto Nāciju\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Gestalt einer deutschen Nationalmannschaft als\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"D überholt zu haben , konterte am heutigen Montag\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Softwarehersteller Oracle hat im dritten Quartal\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Během nich se ekonomické podmínky mohou radikálně\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Dziki kot w górach zeskakuje z kamienia.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Ačkoliv ligový nováček prohrál, opět potvrdil, že\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"des Tages , Portraits internationaler Stars sowie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Communicator bekannt wurde .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"τῷ δ’ ἄρα καὶ αὐτῷ ἡ γυνή ἐπίτεξ ἐοῦσα πᾶσαν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"Triadú tenia, mentre redactava 'Dies de memòria',\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"دستهجمعی در درخشندگی ماه سیمگون زمزمه ستاینده و\\\", \\\"label\\\": \\\"fa\\\"},\\n {\\\"text\\\": \\\"Книгу, наполненную мелочной заботой об одежде,\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"putares canem leporem persequi.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"В дальнейшем эта яркость слегка померкла, но в\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"offizielles Verfahren gegen die Telekom\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"podrían haber sido habitantes de la Península\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Grundlage für dieses Verfahren sind spezielle\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Rechtsausschuß vorgelegten Entwurf der Richtlinie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Im so genannten Portalgeschäft sei das Unternehmen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ⲏ ⲉⲓϣⲁⲛϥⲓ ⲛⲉⲓⲇⲱⲗⲟⲛ ⲉⲧϩⲙⲡⲉⲕⲏⲓ ⲙⲏ ⲉⲓⲛⲁϣϩⲱⲡ ⲟⲛ ⲙⲡⲣⲏ\\\", \\\"label\\\": \\\"cop\\\"},\\n {\\\"text\\\": \\\"juego podían matar a cualquier herbívoro, pero\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Nach Angaben von Axent nutzen Unternehmen aus der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"hrdiny Havlovy Zahradní slavnosti (premiéra ) se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Een zin van heb ik jou daar\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"hat sein Hirn an der CeBIT-Kasse vergessen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"καὶ τοὺς ἐκπλαγέντας οὐκ ἔχειν ἔτι ἐλεγχομένους\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"nachgewiesenen langfristigen Kosten , sowie den im\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"jučer nakon četiri dana putovanja u Helsinki.\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"pašto paslaugos teikėjas gali susitarti su\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"В результате, эти золотые кадры переходят из одной\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"द फाइव-ईयर एंगेजमेंट में अभिनय किया जिसमें जैसन\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"výpis o počtu akcií.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Enfin, elles arrivent à un pavillon chinois\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Tentu saja, tren yang berhubungandengan\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"Arbeidarpartiet og SV har sikra seg fleirtal mot\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"eles: 'Tudo isso está errado' , disse um\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"The islands are in their own time zone, minutes\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Auswahl debütierte er am .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Bu komisyonlar, arazilerini satın almak için\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"Geschütze gegen Redmond aufgefahren .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Time scything the hours, but at the top, over the\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Di musim semi , berharap mengadaptasi Tintin untuk\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"крупнейшей геополитической катастрофой XX века.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Rajojen avaaminen ei suju ongelmitta .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"непроницаемым, как для СССР.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Ma non mancano le polemiche.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Internet als Ort politischer Diskussion und auch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"incomplets.\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Su padre luchó al lado de Luis Moya, primer Jefe\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"informazione.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Primacom bietet für Telekom-Kabelnetz\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Oświadczenie prezydencji w imieniu Unii\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"foran rattet i familiens gamle Baleno hvis døra på\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"[speaker:laughter]\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Dog med langt mindre utstyr med seg.\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"dass es nicht schon mit der anfänglichen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"इस पर दोनों पक्षों में नोकझोंक शुरू हो गई ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"کے ترجمان منیش تیواری اور دگ وجئے سنگھ نے بھی یہ\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"dell'Assemblea Costituente che posseggono i\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"и аште вьси съблазнѧтъ сѧ нъ не азъ\\\", \\\"label\\\": \\\"cu\\\"},\\n {\\\"text\\\": \\\"In Irvine hat auch das Logistikunternehmen Atlas\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"законодательных норм, принимаемых существующей\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Κροίσῳ προτείνων τὰς χεῖρας ἐπικατασφάξαι μιν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"МИНУСЫ: ИНФЛЯЦИЯ И КРИЗИС В ЖИВОТНОВОДСТВЕ.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"unterschiedlicher Meinung .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Jospa joku ystävällinen sielu auttaisi kassieni\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Añadió que, en el futuro se harán otros\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Sessiz tonlama hem Fince, hem de Kuzey Sami\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"nicht ihnen gehört und sie nicht alles , was sie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Etelästä Kuivajärveen laskee Tammelan Liesjärvestä\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"ICANNs Vorsitzender Vint Cerf warb mit dem Hinweis\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Norsk politikk frå til kan dermed, i\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Głosowało posłów.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Danny Jones -- smithjones@ev.net\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"sebeuvědomění moderní civilizace sehrála lučavka\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"относительно спокойный сон: тому гарантия\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"A halte voiz prist li pedra a crïer\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"آنها امیدوارند این واکسن بهزودی در دسترس بیماران\\\", \\\"label\\\": \\\"fa\\\"},\\n {\\\"text\\\": \\\"vlastní důstojnou vousatou tváří.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ora aprire la strada a nuove cause e alimentare il\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Die Zahl der Vielleser nahm von auf Prozent zu ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Finanzvorstand von Hotline-Dienstleister InfoGenie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"entwickeln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"incolumità pubblica.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"lehtija televisiomainonta\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"joistakin kohdista eri mieltä.\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Hlavně anglická nezávislá scéna, Dead Can Dance,\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"pásmech od do bodů bodové stupnice.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Zu Beginn des Ersten Weltkrieges zählte das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Així van sorgir, damunt els antics cementiris,\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"In manchem Gedicht der spätern Alten, wie zum\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"gaweihaida jah insandida in þana fairƕu jus qiþiþ\\\", \\\"label\\\": \\\"got\\\"},\\n {\\\"text\\\": \\\"Beides sollte gelöscht werden!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"modifiqués la seva petició inicial de anys de\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"В день открытия симпозиума состоялась закладка\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"tõestatud.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"ἵππῳ πίπτει αὐτοῦ ταύτῃ\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"bisher nie enttäuscht!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"De bohte ollu tuollárat ja suttolaččat ja\\\", \\\"label\\\": \\\"sme\\\"},\\n {\\\"text\\\": \\\"Klarsignal från röstlängdsläsaren, tre tryck i\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Tvůrcem nového termínu je Joseph Fisher.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nie miałem czasu na reakcję twierdzi Norbert,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"potentia Schöpfer.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Un poquito caro, pero vale mucho la pena;\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"οὔ τε γὰρ ἴφθιμοι Λύκιοι Δαναῶν ἐδύναντο τεῖχος\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"vajec, sladového výtažku a některých vitamínových\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Настоящие герои, те, чьи истории потом\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"praesumptio:\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Olin justkui nende vastutusel.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Jokainen keinahdus tuo lähemmäksi hetkeä jolloin\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"ekonomicky výhodných způsobů odvodnění těžkých,\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Poprvé ve své historii dokázala v kvalifikaci pro\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"zpracovatelského a spotřebního průmyslu bude nutné\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Windows CE zu integrieren .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Armangué, a través d'un decret, ordenés l'aturada\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"to, co nás Evropany spojuje, než to, co nás od\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ergänzt durch einen gesetzlich verankertes\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Насчитал, что с начала года всего три дня были\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Borisovu tražeći od njega da prihvati njenu\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"la presenza di ben veleni diversi: . chili di\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"καὶ τῶν ἐκλεκτῶν ἀγγέλων ἵνα ταῦτα φυλάξῃς χωρὶς\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"pretraživale obližnju bolnicu i stambene zgrade u\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"An rund Katzen habe Wolf seine Spiele getestet ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"investigating since March.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Tonböden (Mullböden).\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Stálý dopisovatel LN v SRN Bedřich Utitz\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"červnu předložené smlouvy.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"πνεύματι ᾧ ἐλάλει\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\".%의 신장세를 보였다.\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Foae verde, foi de nuc, Prin pădure, prin colnic,\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"διαπέμψας ἄλλους ἄλλῃ τοὺς μὲν ἐς Δελφοὺς ἰέναι\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"المسلمين أو أي تيار سياسي طالما عمل ذلك التيار في\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"As informações são da Dow Jones.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Milliarde DM ausgestattet sein .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"De utgår fortfarande från att kvinnans jämlikhet\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Sneeuw maakte in Davos bij de voorbereiding een\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"De ahí que en este mercado puedan negociarse\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"intenzívnějšímu sbírání a studiu.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"और औसकर ४.० पैकेज का प्रयोग किया गया है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Adipati Kuningan karena Kuningan menjadi bagian\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"Svako je bar jednom poželeo da mašine prosto umeju\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"Im vergangenen Jahr haben die Regierungen einen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"durat motus, aliquid fit et non est;\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Dominować będą piosenki do tekstów Edwarda\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"beantwortet .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"О гуманитариях было кому рассказывать, а вот за\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Helsingin kaupunki riitautti vuokrasopimuksen\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"chợt tan biến.\\\", \\\"label\\\": \\\"vi\\\"},\\n {\\\"text\\\": \\\"avtomobil ločuje od drugih.\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Congress has proven itself ineffective as a body.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"मैक्सिको ने इस तरह का शो इस समय आयोजित करने का\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"No minimum order amount.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Convertassa .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Как это можно сделать?\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"tha mi creidsinn gu robh iad ceart cho saor shuas\\\", \\\"label\\\": \\\"gd\\\"},\\n {\\\"text\\\": \\\"실제 일제는 이런 만해의 논리를 묵살하고 한반도를 침략한 다음 , 이어 만주를 침략하고\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Da un semplice richiamo all'ordine fino a grandi\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"pozoruhodný nejen po umělecké stránce, jež\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"La comida y el servicio aprueban.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"again, connected not with each other but to the\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Protokol výslovně stanoví, že nikdo nemůže být\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ఒక విషయం అడగాలని ఉంది .\\\", \\\"label\\\": \\\"te\\\"},\\n {\\\"text\\\": \\\"Безгранично почитая дирекцию, ловя на лету каждое\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"rovnoběžných růstových vrstev, zůstávají krychlové\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"प्रवेश और पूर्व प्रधानमंत्री लाल बहादुर शास्त्री\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Bronzen medaille in de Europese marathon.\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"- gadu vecumā viņi to nesaprot.\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Realizó sus estudios primarios en la Escuela Julia\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"cuartos de final, su clasificación para la final a\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Sem si pro něho přiletí americký raketoplán, na\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Way to go!\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"gehört der neuen SPD-Führung unter Parteichef\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Somit simuliert der Player mit einer GByte-Platte\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Berufung auf kommissionsnahe Kreise , die bereits\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Dist Clarïen\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Schon nach den Gerüchten , die Telekom wolle den\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Software von NetObjects ist nach Angaben des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"si enim per legem iustitia ergo Christus gratis\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"ducerent in ipsam magis quam in corpus christi,\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Neustar-Melbourne-IT-Partnerschaft NeuLevel .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"forderte dagegen seine drastische Verschärfung.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"pemmican på hundrede forskellige måder.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"Lehån, själv matematiklärare, visar hur den nya\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"I highly recommend his shop.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"verità, giovani fedeli prostratevi #amen\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"उत्तर प्रदेश के अध्यक्ष पद से हटाए गए विनय कटियार\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"() روزی مےں کشادگی ہوتی ہے۔\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"Prozessorgeschäft profitieren kann , stellen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"školy začalo počítat pytle s moukou a zjistilo, že\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"प्रभावशाली पर गैर सरकारी लोगों के घरों में भी\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"geschichtslos , oder eine Farce , wie sich\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ústrednými mocnosťami v marci však spôsobilo, že\\\", \\\"label\\\": \\\"sk\\\"},\\n {\\\"text\\\": \\\"التسليح بدون مبرر، واستمرار الأضرار الناجمة عن فرض\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Například Pedagogická fakulta Univerzity Karlovy\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"nostris ut eriperet nos de praesenti saeculo\\\", \\\"label\\\": \\\"la\\\"}]\\n \\n docs = [Document([], text=example[\\\"text\\\"]) for example in examples]\\n gold_labels = [example[\\\"label\\\"] for example in examples]\\n basic_multilingual(docs)\\n accuracy = sum([(doc.lang == label) for doc,label in zip(docs,gold_labels)])/len(docs)\\n assert accuracy >= 0.98\",\n \"def test_translation_smoke():\\n english_to_morse = get_translator(\\\"english\\\", \\\"morse\\\")\\n morse_to_english = get_translator(\\\"morse\\\", \\\"english\\\")\\n morse = english_to_morse.translate(\\\"hello world\\\")\\n english = morse_to_english.translate(morse)\\n assert english == \\\"HELLO WORLD\\\"\",\n \"def showTranslatedWithoutJoin(cls):\\n print (\\\"ALL WORDS WITH TRANSLATIONS STORED IN DATABASE:\\\")\\n for word1 in EnglishHelper.query(\\\"SELECT english_word FROM EnglishWords\\\", fetchAll=True):\\n try:\\n print word1[0],\\\" - \\\", (EnglishHelper.query(\\\"select polish_word from PolishWords where \\\"\\n \\\" id_pl=(select id_pl from translations where \\\"\\n \\\"id_eng = (select id_eng from EnglishWords \\\"\\n \\\"where english_word = '%s'))\\\"%word1))[0].encode('utf-8')\\n except:\\n print \\\"There is no translation, sorry :(\\\"\",\n \"def test_single_locale_activation(self):\\n with translation.override(\\\"fr\\\"):\\n self.assertEqual(\\n self.get_template(\\n \\\"{% load i18n %}{% blocktranslate %}Yes{% endblocktranslate %}\\\"\\n ).render(Context({})),\\n \\\"Oui\\\",\\n )\",\n \"def test_i18n02(self):\\n output = self.engine.render_to_string('i18n02')\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string('i18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string('i18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string(\\\"i18n08\\\", {\\\"number\\\": 2})\\n self.assertEqual(output, \\\"2 plural\\\")\",\n \"def test_i18n14(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n14')\\n self.assertEqual(output, 'foo Passwort Passwort')\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string(\\\"legacyi18n08\\\", {\\\"number\\\": 2})\\n self.assertEqual(output, \\\"2 plural\\\")\",\n \"def test_legacyi18n04(self):\\n output = self.engine.render_to_string(\\\"legacyi18n04\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"å\\\")\",\n \"def test_i18n12(self):\\n output = self.engine.render_to_string('i18n12')\\n self.assertEqual(output, 'de')\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def stats_text(test):\\n\\n stats_text_en(test) \\n \\n stats_text_cn(test)\",\n \"def test_fallback_language_no_current(self):\\n x = SimpleModel()\\n x.set_current_language(self.conf_fallback)\\n x.tr_title = \\\"TITLE_FALLBACK\\\"\\n\\n self.assertEqual(\\n x.safe_translation_getter(\\\"tr_title\\\", language_code=self.other_lang1), \\\"TITLE_FALLBACK\\\"\\n )\",\n \"def test_language_fix(self):\\n #TODO\\n \\n for lang in self.LANGUAGES:\\n activate(lang)\\n \\n self.assertEqual(lang, get_language())\",\n \"def test_get_languages_multiple(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n translation = StoryTranslation(story=story, title=\\\"Spanish Title\\\",\\n summary=\\\"Spanish Summary\\\", language=\\\"es\\\")\\n translation.save()\\n self.assertEqual([settings.LANGUAGE_CODE, 'es'], story.get_languages())\",\n \"def translate_wrapper(atext):\\n print(\\\"translating:\\\",atext)\\n res=\\\"\\\"\\n res=translate(atext,\\\"pl\\\",\\\"fr\\\")\\n time.sleep(0.5)\\n print(\\\"translation:\\\",res)\\n return res\",\n \"def test_english_gb_no_translation(self):\\n resp = ResponseFactory(\\n locale=u'en-GB',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_english_with_dennis(self):\\n resp = ResponseFactory(\\n locale=u'en-US',\\n product=u'firefox',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # Set the product up for translation *after* creating the response\\n # so that it doesn't get auto-translated because Response is set up\\n # for auto-translation.\\n prod = Product.objects.get(db_name='firefox')\\n prod.translation_system = u'dennis'\\n prod.save()\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_get_translation_resources(self):\\n pass\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string(\\\"i18n05\\\", {\\\"anton\\\": \\\"yyy\\\"})\\n self.assertEqual(output, \\\"xxxyyyxxx\\\")\",\n \"def test_get_languages_multiple(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n translation = StoryTranslation(story=story, title=\\\"Spanish Title\\\",\\n summary=\\\"Spanish Summary\\\", language=\\\"es\\\")\\n translation.save()\\n story_languages = story.get_languages()\\n self.assertEqual(len(story_languages), 2)\\n for code in (settings.LANGUAGE_CODE, 'es'):\\n self.assertIn(code, story_languages)\",\n \"def test_single_locale_activation(self):\\n with translation.override('fr'):\\n self.assertEqual(\\n Template(\\\"{% load i18n %}{% blocktrans %}Yes{% endblocktrans %}\\\").render(Context({})),\\n 'Oui'\\n )\",\n \"def test_00_i18n_anonymous(self):\\r\\n # First default 'en' locale\\r\\n with self.app as c:\\r\\n err_msg = \\\"The page should be in English\\\"\\r\\n res = c.get('/', headers=[('Accept-Language', 'en')])\\r\\n assert \\\"Community\\\" in res.data, err_msg\\r\\n # Second with 'es' locale\\r\\n with self.app as c:\\r\\n err_msg = \\\"The page should be in Spanish\\\"\\r\\n res = c.get('/', headers=[('Accept-Language', 'es')])\\r\\n assert \\\"Comunidad\\\" in res.data, err_msg\",\n \"def test_functionallity(self):\\n\\n pp = Lexpp(external_dict=pkg_resources.resource_filename(\\\"lexpp\\\", \\\"tests/test.dict\\\"))\\n\\n test_word = \\\"キャプテン\\\"\\n entries = list(pp.lookup(test_word))\\n\\n self.assertEqual(len(entries), 4)\\n\\n for e in entries:\\n self.assertEqual(type(e), Entry)\\n rep = pp.get_representative_form(e)\\n self.assertEqual(rep, test_word)\",\n \"def test_get_languages(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())\",\n \"def test_get_languages(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_fallback_variant(self):\\n x = SimpleModel()\\n\\n x.set_current_language(\\\"de\\\")\\n x.tr_title = \\\"Hallo-de\\\"\\n\\n x.set_current_language(\\\"en\\\")\\n x.tr_title = \\\"Hello-en\\\"\\n\\n x.save()\\n\\n with translation.override(\\\"de-ch\\\"):\\n x = SimpleModel.objects.get(pk=x.pk)\\n self.assertEqual(x.tr_title, \\\"Hallo-de\\\")\",\n \"def translate(self):\\n pass\",\n \"def _load_transliterated_regional(self):\\n# global approved, conflicts, suggestions, unknown, cldr, current\\n start = self.lblFallback['text'].find('=>') + 2\\n if self.lblFallback['text'][start:]:\\n self.preferred.set(\\\\\\n self._transliterate_text(self.lblFallback['text'][start:]))\\n pass\",\n \"def test_i18n15(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n15', {'absent': ''})\\n self.assertEqual(output, 'Passwort')\",\n \"def test_i18n04(self):\\n output = self.engine.render_to_string('i18n04', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'å')\",\n \"def test_i18n04(self):\\n output = self.engine.render_to_string('i18n04', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'å')\",\n \"def test_legacyi18n04(self):\\n output = self.engine.render_to_string('legacyi18n04', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'å')\",\n \"def test_legacyi18n04(self):\\n output = self.engine.render_to_string('legacyi18n04', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'å')\",\n \"def test_list_unicode_locale(self):\\n response = self.client.get(self.url, {'lang': 'ru'})\\n assert response.data\\n\\n assert response.data['count'] == len(discopane_items)\\n assert response.data['next'] is None\\n assert response.data['previous'] is None\\n assert response.data['results']\",\n \"def test_fallback_values_1(self):\\n title1_de = \\\"title de\\\"\\n n = TestModelWithFallback()\\n n.title = title1_de\\n n.save()\\n del n\\n n = TestModelWithFallback.objects.get(title=title1_de)\\n self.failUnlessEqual(n.title, title1_de)\\n trans_real.activate(\\\"en\\\")\\n self.failUnlessEqual(n.title, \\\"\\\")\",\n \"def goToTranslate():\\n\\n try:\\n global driver\\n\\n options = Options()\\n\\n # Code to disable notifications pop up of Chrome Browser\\n options.add_argument(\\\"--disable-notifications\\\")\\n options.add_argument(\\\"--disable-infobars\\\")\\n options.add_argument(\\\"--mute-audio\\\")\\n # options.add_argument(\\\"headless\\\")\\n\\n try:\\n platform_ = platform.system().lower()\\n if platform_ in ['linux', 'darwin']:\\n driver = webdriver.Chrome(executable_path=\\\"./chromedriver\\\", options=options)\\n else:\\n driver = webdriver.Chrome(executable_path=\\\"./chromedriver.exe\\\", options=options)\\n except:\\n print(\\\"Kindly replace the Chrome Web Driver with the latest one from\\\"\\n \\\"http://chromedriver.chromium.org/downloads\\\"\\n \\\"\\\\nYour OS: {}\\\".format(platform_)\\n )\\n exit()\\n\\n driver.get(\\\"https://translate.google.com\\\") # go to facebook home page\\n driver.maximize_window()\\n\\n time.sleep(30)\\n\\n try:\\n\\n f=open('C:\\\\\\\\Users\\\\\\\\Alvi Rahman\\\\\\\\Desktop\\\\\\\\Farhana Miss\\\\\\\\Data\\\\\\\\banglaTweets.txt',mode=\\\"r\\\", encoding=\\\"utf-8\\\")\\n tweets=f.readlines()\\n f.close()\\n\\n tweets=[s.replace('\\\\n','') for s in tweets]\\n\\n output_file=open(\\\"translated_tweet.txt\\\", mode=\\\"w\\\", encoding=\\\"utf-16\\\")\\n\\n inputBox=driver.find_element_by_id('source')\\n\\n tweet_count = 0\\n for i in range (8001,12000):\\n inputBox.send_keys(tweets[i])\\n translatedTweet = \\\"xxxxxxxxxxxx\\\"\\n time.sleep(10)\\n\\n try:\\n\\n\\t TranslationBox=driver.find_element_by_xpath(\\\"/html/body/div[2]/div[1]/div[2]/div[1]/div[1]/div[2]/div[3]/div[1]/div[2]/div/span[1]/span\\\")\\n\\t translatedTweet = TranslationBox.text\\n\\n\\t translatedTweet=translatedTweet.replace('\\\\b',' ')\\n\\t translatedTweet=translatedTweet.replace('\\\\n',' ')\\n\\t translatedTweet=translatedTweet.replace('\\\\t',' ')\\n\\t translatedTweet=translatedTweet.replace('\\\\r',' ')\\n\\t translatedTweet=translatedTweet.replace('\\\\f',' ')\\n\\t translatedTweet=reg.sub(' +',' ',translatedTweet)\\n\\n except:\\n pass\\n\\n\\n output_file.writelines(translatedTweet+\\\"\\\\n\\\")\\n inputBox.clear()\\n tweet_count = tweet_count + 1\\n\\n if tweet_count%100 ==0:\\n time.sleep(270)\\n\\n elif tweet_count%50 ==0:\\n time.sleep(150)\\n\\n elif tweet_count%10 ==0:\\n time.sleep(60)\\n\\n output_file.close()\\n driver.close()\\n\\n except:\\n exc_type = sys.exc_info()[0]\\n exc_tb = sys.exc_info()[2]\\n print(exc_type, exc_tb.tb_lineno)\\n pass\\n\\n # filling the form\\n # driver.find_element_by_name('email').send_keys(email)\\n # driver.find_element_by_name('pass').send_keys(password)\\n\\n # # clicking on login button\\n # driver.find_element_by_id('loginbutton').click()\\n\\n print(\\\"Successfully logged in\\\")\\n\\n except Exception as e:\\n print(\\\"There's some error in log in.\\\")\\n print(sys.exc_info()[0])\\n exit()\",\n \"def test_english_no_translation(self):\\n resp = ResponseFactory(\\n locale=u'en-US',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_home_view_cyrillic(self):\\n UserData.objects.create(\\n first_name=\\\"Андрей\\\",\\n last_name=\\\"Герасько\\\",\\n date_of_birth='1998-02-23',\\n bio=\\\"Био\\\",\\n email=\\\"aaa@gmail.com\\\",\\n jabber=\\\"aaaaa\\\",\\n skype=\\\"aaaaa\\\",\\n other_contacts=\\\"Контакты\\\"\\n )\\n response = self.client.get(self.url)\\n self.assertEqual(response.status_code, 200)\\n self.assertIn('42 Coffee Cups Test Assignment', response.content)\\n self.assertIn('Андрей', response.content)\\n self.assertIn('Герасько', response.content)\\n self.assertIn('Date of birth', response.content)\\n self.assertIn('Био', response.content)\\n self.assertIn('Контакты', response.content)\",\n \"def test_translate(self):\\n body = Question()\\n response = self.client.open('/api/rtx/v1/translate',\\n method='POST',\\n data=json.dumps(body),\\n content_type='application/json')\\n self.assert200(response, \\\"Response body is : \\\" + response.data.decode('utf-8'))\",\n \"def test_get_supported_locales_for_voice_datasets(self):\\n pass\",\n \"def test_energy_consumption_top_ranking_ko(self):\\n self.driver.get(\\n self.url_ +\\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko'\\n )\\n\\n self.assertEqual(\\n self.driver.find_elements_by_class_name('navbar-brand')[0].text,\\n 'Data Commons 장소 순위')\\n\\n subtitle_present = EC.text_to_be_present_in_element(\\n (By.TAG_NAME, 'h3'), 'the World 상위 국가 100개')\\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(subtitle_present)\\n self.assertEqual(\\n self.driver.find_element_by_tag_name('h1').text, '1인당 에너지 소비량 순위')\\n self.assertEqual(\\n self.driver.find_element_by_xpath(\\n '//*[@id=\\\"main-pane\\\"]/div/h3/a').get_attribute('href'),\\n self.url_ +\\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko&bottom='\\n )\\n\\n table = self.driver.find_element_by_xpath(\\n '//*[@id=\\\"main-pane\\\"]/div/table')\\n headers = table.find_elements_by_xpath('.//thead/tr/th')\\n self.assertEqual(headers[0].text, '순위')\\n self.assertEqual(headers[1].text, '국가')\\n self.assertEqual(headers[2].text, '킬로그램')\\n row = table.find_elements_by_xpath('.//tbody/tr[1]/td')\\n self.assertEqual(row[0].text, '1')\\n self.assertEqual(row[1].text, '카타르')\\n self.assertEqual(\\n row[1].find_element_by_tag_name('a').get_attribute('href'),\\n self.url_ + '/place/country/QAT?hl=ko')\\n\\n chart = self.driver.find_element_by_id('ranking-chart')\\n y_text = chart.find_elements_by_class_name(\\n 'y')[0].find_elements_by_tag_name('text')\\n self.assertEqual(y_text[0].text, '0kg')\\n self.assertEqual(y_text[-1].text, '1.5만kg')\\n\\n x_text = chart.find_elements_by_class_name(\\n 'x')[0].find_elements_by_tag_name('text')\\n self.assertEqual(x_text[0].text, '카타르')\\n self.assertEqual(x_text[-1].text, '에콰도르')\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string('i18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string('i18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string(\\\"i18n07\\\", {\\\"number\\\": 1})\\n self.assertEqual(output, \\\"singular\\\")\",\n \"def test_shipped_locales(self):\\n url = reverse('shipping.views.status.shipped_locales')\\n url += '?av=fx1.0'\\n response = self.client.get(url)\\n eq_(response.status_code, 200)\\n eq_(response.content, \\\"\\\"\\\"da\\nde\\nen-US\\n\\\"\\\"\\\")\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string(\\\"legacyi18n07\\\", {\\\"number\\\": 1})\\n self.assertEqual(output, \\\"singular\\\")\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_i18n03(self):\\n output = self.engine.render_to_string('i18n03', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'Å')\",\n \"def test_i18n03(self):\\n output = self.engine.render_to_string('i18n03', {'anton': b'\\\\xc3\\\\x85'})\\n self.assertEqual(output, 'Å')\",\n \"def test_get_unique_langs(self):\\n\\n contacts = Contact.query.filter_by(user_id=6).all()\\n\\n unique_lang_dict = MessageController.get_unique_langs(contacts)\\n self.assertEqual(unique_lang_dict, {4: u'fr'})\",\n \"def _test (self) :\\n print (_T (ckw.title or \\\"Baz\\\"))\\n print (_T (\\\"Foo\\\"))\\n foo = _(\\\"Markup %d\\\")\\n print (_T(foo) % 42)\\n print (_Tn (\\\"Singular\\\", \\\"Plural\\\", 4))\",\n \"def test_i18n26(self):\\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\\n self.assertEqual(output, 'singular test')\",\n \"def test_i18n26(self):\\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\\n self.assertEqual(output, 'singular test')\",\n \"def test_spanish_with_dennis_and_existing_translations(self):\\n existing_resp = ResponseFactory(\\n locale=u'es',\\n product=u'firefox',\\n description=u'hola',\\n translated_description=u'DUDE!'\\n )\\n\\n resp = ResponseFactory(\\n locale=u'es',\\n product=u'firefox',\\n description=u'hola',\\n translated_description=u''\\n )\\n\\n # Set the product up for translation *after* creating the response\\n # so that it doesn't get auto-translated because Response is set up\\n # for auto-translation.\\n prod = Product.objects.get(db_name='firefox')\\n prod.translation_system = u'dennis'\\n prod.save()\\n\\n # No jobs should be translated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n eq_(resp.translated_description, existing_resp.translated_description)\",\n \"def test_spanish_no_translation(self):\\n resp = ResponseFactory(\\n locale=u'es',\\n product=u'firefox',\\n description=u'hola',\\n translated_description=u''\\n )\\n\\n # No jobs should be translated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Nothing should be translated\\n eq_(resp.translated_description, u'')\",\n \"def test_homepage_en(self):\\n\\n self.driver.get(self.url_ + '/')\\n\\n title_present = EC.text_to_be_present_in_element(\\n (By.XPATH, '//*[@id=\\\"main-nav\\\"]/div/div[1]/a'), 'Data Commons')\\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(title_present)\\n\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons is an open knowledge repository'))\\n\\n explore_callout_msg = self.driver.find_element_by_xpath(\\n '//*[@id=\\\"homepage\\\"]/section[3]/p')\\n self.assertTrue(\\n explore_callout_msg.text.startswith(\\n 'We cleaned and processed the data so you don\\\\'t have to'))\\n\\n nyc_health = self.driver.find_element_by_xpath(\\n '//*[@id=\\\"homepage\\\"]/section[3]/ul/li[1]/ul/li[2]/a')\\n self.assertEqual(nyc_health.text, 'New York City, NY Health')\\n self.assertEqual(nyc_health.get_attribute('href'),\\n self.url_ + '/place/geoId/3651000?topic=Health')\\n\\n schema_org = self.driver.find_element_by_xpath(\\n '//*[@id=\\\"homepage\\\"]/section[5]/ul/li[2]')\\n self.assertEqual(schema_org.text,\\n 'Open sourced, built using Schema.org.')\\n\\n more_msg = self.driver.find_element_by_xpath(\\n '//*[@id=\\\"homepage\\\"]/section[3]/ul/li[3]/ul/li[4]/a')\\n self.assertEqual(more_msg.text, 'more ...')\",\n \"def try_translation(text):\\n try:\\n trad = Translator().translate(text = text, dest = 'fr')\\n except:\\n sleep(random() * 60)\\n return False\\n \\n return trad.text\",\n \"def translation():\\n class TranslationTable(tables.Table):\\n normal = tables.Column(verbose_name=ugettext(\\\"Normal\\\"))\\n lazy = tables.Column(verbose_name=ugettext(\\\"Lazy\\\"))\\n\\n table = TranslationTable([])\\n Assert(\\\"Normal\\\") == table.columns[\\\"normal\\\"].header\\n Assert(\\\"Lazy\\\") == table.columns[\\\"lazy\\\"].header\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.66185194","0.65907204","0.65775216","0.65519655","0.65519655","0.65463674","0.6530294","0.6447057","0.64026135","0.63868695","0.6383359","0.63773143","0.63244116","0.6323046","0.6316519","0.6313638","0.63092667","0.62892264","0.62892264","0.6254798","0.6225774","0.6220153","0.62187886","0.62006176","0.61804575","0.618008","0.61532485","0.61377525","0.6134953","0.6071378","0.6069911","0.60664296","0.606236","0.60539615","0.60299593","0.60128576","0.60080713","0.6006107","0.59998196","0.5996066","0.5978001","0.5978001","0.59746027","0.5970351","0.59553397","0.59494466","0.59461707","0.5938071","0.5938071","0.59178","0.5908871","0.5888681","0.5887712","0.58819085","0.586043","0.58383405","0.58335257","0.58272487","0.58240336","0.57941276","0.57697546","0.5762837","0.57608736","0.57608736","0.57555526","0.57555526","0.57528436","0.57357776","0.5699352","0.5698747","0.56887764","0.56887764","0.5683346","0.5683346","0.56821704","0.5658707","0.56555897","0.5637075","0.5612181","0.56022596","0.5601521","0.5593238","0.5590103","0.5590103","0.55840844","0.55820894","0.5570769","0.5558774","0.5558774","0.5558715","0.5558715","0.5556942","0.5549655","0.55484575","0.55484575","0.55365914","0.5528148","0.55098677","0.55030894","0.5499714"],"string":"[\n \"0.66185194\",\n \"0.65907204\",\n \"0.65775216\",\n \"0.65519655\",\n \"0.65519655\",\n \"0.65463674\",\n \"0.6530294\",\n \"0.6447057\",\n \"0.64026135\",\n \"0.63868695\",\n \"0.6383359\",\n \"0.63773143\",\n \"0.63244116\",\n \"0.6323046\",\n \"0.6316519\",\n \"0.6313638\",\n \"0.63092667\",\n \"0.62892264\",\n \"0.62892264\",\n \"0.6254798\",\n \"0.6225774\",\n \"0.6220153\",\n \"0.62187886\",\n \"0.62006176\",\n \"0.61804575\",\n \"0.618008\",\n \"0.61532485\",\n \"0.61377525\",\n \"0.6134953\",\n \"0.6071378\",\n \"0.6069911\",\n \"0.60664296\",\n \"0.606236\",\n \"0.60539615\",\n \"0.60299593\",\n \"0.60128576\",\n \"0.60080713\",\n \"0.6006107\",\n \"0.59998196\",\n \"0.5996066\",\n \"0.5978001\",\n \"0.5978001\",\n \"0.59746027\",\n \"0.5970351\",\n \"0.59553397\",\n \"0.59494466\",\n \"0.59461707\",\n \"0.5938071\",\n \"0.5938071\",\n \"0.59178\",\n \"0.5908871\",\n \"0.5888681\",\n \"0.5887712\",\n \"0.58819085\",\n \"0.586043\",\n \"0.58383405\",\n \"0.58335257\",\n \"0.58272487\",\n \"0.58240336\",\n \"0.57941276\",\n \"0.57697546\",\n \"0.5762837\",\n \"0.57608736\",\n \"0.57608736\",\n \"0.57555526\",\n \"0.57555526\",\n \"0.57528436\",\n \"0.57357776\",\n \"0.5699352\",\n \"0.5698747\",\n \"0.56887764\",\n \"0.56887764\",\n \"0.5683346\",\n \"0.5683346\",\n \"0.56821704\",\n \"0.5658707\",\n \"0.56555897\",\n \"0.5637075\",\n \"0.5612181\",\n \"0.56022596\",\n \"0.5601521\",\n \"0.5593238\",\n \"0.5590103\",\n \"0.5590103\",\n \"0.55840844\",\n \"0.55820894\",\n \"0.5570769\",\n \"0.5558774\",\n \"0.5558774\",\n \"0.5558715\",\n \"0.5558715\",\n \"0.5556942\",\n \"0.5549655\",\n \"0.55484575\",\n \"0.55484575\",\n \"0.55365914\",\n \"0.5528148\",\n \"0.55098677\",\n \"0.55030894\",\n \"0.5499714\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.58505183"},"document_rank":{"kind":"string","value":"55"}}},{"rowIdx":90887,"cells":{"query":{"kind":"string","value":"Test translations are displayed correctly in korean, as well as top rankings rendered correctly."},"document":{"kind":"string","value":"def test_energy_consumption_top_ranking_ko(self):\n self.driver.get(\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko'\n )\n\n self.assertEqual(\n self.driver.find_elements_by_class_name('navbar-brand')[0].text,\n 'Data Commons 장소 순위')\n\n subtitle_present = EC.text_to_be_present_in_element(\n (By.TAG_NAME, 'h3'), 'the World 상위 국가 100개')\n WebDriverWait(self.driver, self.TIMEOUT_SEC).until(subtitle_present)\n self.assertEqual(\n self.driver.find_element_by_tag_name('h1').text, '1인당 에너지 소비량 순위')\n self.assertEqual(\n self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/h3/a').get_attribute('href'),\n self.url_ +\n '/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko&bottom='\n )\n\n table = self.driver.find_element_by_xpath(\n '//*[@id=\"main-pane\"]/div/table')\n headers = table.find_elements_by_xpath('.//thead/tr/th')\n self.assertEqual(headers[0].text, '순위')\n self.assertEqual(headers[1].text, '국가')\n self.assertEqual(headers[2].text, '킬로그램')\n row = table.find_elements_by_xpath('.//tbody/tr[1]/td')\n self.assertEqual(row[0].text, '1')\n self.assertEqual(row[1].text, '카타르')\n self.assertEqual(\n row[1].find_element_by_tag_name('a').get_attribute('href'),\n self.url_ + '/place/country/QAT?hl=ko')\n\n chart = self.driver.find_element_by_id('ranking-chart')\n y_text = chart.find_elements_by_class_name(\n 'y')[0].find_elements_by_tag_name('text')\n self.assertEqual(y_text[0].text, '0kg')\n self.assertEqual(y_text[-1].text, '1.5만kg')\n\n x_text = chart.find_elements_by_class_name(\n 'x')[0].find_elements_by_tag_name('text')\n self.assertEqual(x_text[0].text, '카타르')\n self.assertEqual(x_text[-1].text, '에콰도르')"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def test_i18n10(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n10', {'bool': True})\n self.assertEqual(output, 'Ja')","def test_i18n27(self):\n with translation.override(\"ru\"):\n output = self.engine.render_to_string(\"i18n27\", {\"number\": 1})\n self.assertEqual(\n output, \"1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442\"\n )","def test_i18n09(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n09')\n self.assertEqual(output, 'Page not found')","def test_i18n27(self):\n with translation.override('ru'):\n output = self.engine.render_to_string('i18n27', {'number': 1})\n self.assertEqual(output, '1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442')","def test_i18n27(self):\n with translation.override('ru'):\n output = self.engine.render_to_string('i18n27', {'number': 1})\n self.assertEqual(output, '1 \\u0440\\u0435\\u0437\\u0443\\u043b\\u044c\\u0442\\u0430\\u0442')","def test_i18n23(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n23')\n self.assertEqual(output, 'nicht gefunden')","def test_i18n(self):\n # If you want to modify this code, please get acquainted with\n # Python's locale module. In particular:\n # http://docs.python.org/library/locale.html#locale.getdefaultlocale\n\n # Set the standard C locale.\n os.environ['LANG'] = 'C'\n os.environ['LC_ALL'] = 'C.UTF-8'\n\n #must be after above\n out1 = tr('Hello!')\n expected1 = 'Hello!'\n msg = 'Expected %s, got %s' % (expected1, out1)\n assert out1 == expected1, msg\n\n # Set the Indonesian locale to test translations.\n os.environ['LANG'] = 'id'\n os.environ['LC_ALL'] = 'id_ID.UTF-8'\n\n #must be after above\n #indoout1 = tr('Hello!') # translate as 'Hi'\n #indoexpected1 = 'Hi!'\n #msg = 'Expected %s, got %s' % (indoexpected1, indoout1)\n #assert indoout1 == indoexpected1, msg","def test_i18n11(self):\n output = self.engine.render_to_string('i18n11', {'bool': True})\n self.assertEqual(output, 'ja')","def test_default_translations(self):\n\t\t\n\t\tself.assertTrue(data.get_default_translation('Catholicism', 3) == 'DRA')\n\t\tself.assertTrue(data.get_default_translation('Christianity', 3) == 'ESV')","def test_i18n06(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n06')\n self.assertEqual(output, 'Seite nicht gefunden')","def setUp(self):\n super().setUp()\n translation.activate(\"en-us\")","def test_i18n28(self):\n output = self.engine.render_to_string(\"i18n28\", {\"anton\": \"α\", \"berta\": \"β\"})\n self.assertEqual(output, \"α + β\")","def test_i18n16(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n16')\n self.assertEqual(output, '<')","def test_i18n28(self):\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\n self.assertEqual(output, 'α + β')","def test_i18n28(self):\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\n self.assertEqual(output, 'α + β')","def test_language_sensitivity(self): \n \n for lang in self.LANGUAGES:\n activate(lang)\n \n self.assertEqual(get_wording_text('test_1'), lang)","def use_en(self):\n pass","def test_i18n13(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n13')\n self.assertEqual(output, 'Passwort')","def test_international(self):\n\n test_cases = (\n makeSimpleTestCase('the beatles'),\n makeSimpleTestCase('david guetta'),\n makeSimpleTestCase('susan boyle'),\n makeSimpleTestCase('rammstein'),\n makeSimpleTestCase('deadmau5'),\n makeTestCase('deadmau 5', Matcher(title=Equals('deadmau5'))),\n makeSimpleTestCase('2face idibia'),\n makeSimpleTestCase('angelique kidjo'),\n makeSimpleTestCase('gotye'),\n makeSimpleTestCase('michel telo'),\n makeSimpleTestCase('stereo total'),\n makeSimpleTestCase('yelle'),\n makeSimpleTestCase('stefanie heinzmann'),\n makeSimpleTestCase('olly murs')\n )\n\n self._run_tests('i18n_artist', test_cases)","def test_i18n12(self):\n output = self.engine.render_to_string('i18n12')\n self.assertEqual(output, 'de')","def test_translate_unique_langs(self):\n\n trans_msgs_dict = MessageController.translate_unique_langs({'2': 'es', '4': 'fr'}, \n 'hi', 'en', False, False)\n\n self.assertEqual(trans_msgs_dict, {'es': u'{hola}', 'fr': u'salut'})","def test_i18n01(self):\n output = self.engine.render_to_string('i18n01')\n self.assertEqual(output, 'xxxyyyxxx')","def test_legacyi18n08(self):\n output = self.engine.render_to_string(\"legacyi18n08\", {\"number\": 2})\n self.assertEqual(output, \"2 plural\")","def test_fallback_language_no_current(self):\n x = SimpleModel()\n x.set_current_language(self.conf_fallback)\n x.tr_title = \"TITLE_FALLBACK\"\n\n self.assertEqual(\n x.safe_translation_getter(\"tr_title\", language_code=self.other_lang1), \"TITLE_FALLBACK\"\n )","def test_i18n08(self):\n output = self.engine.render_to_string(\"i18n08\", {\"number\": 2})\n self.assertEqual(output, \"2 plural\")","def test_i18n14(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n14')\n self.assertEqual(output, 'foo Passwort Passwort')","def test_i18n08(self):\n output = self.engine.render_to_string('i18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_i18n08(self):\n output = self.engine.render_to_string('i18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_switching_to_korean(web_browser):\n\n page = MainPage(web_browser)\n page.welcoming_popup_close_button.click()\n\n # Changing language to korean\n page.regional_settings_button.click()\n page.current_language_button.click()\n page.korean_language_button.click()\n page.finish_login_button.click()\n page.wait_page_loaded()\n\n assert \"더 똑똑한 쇼핑, 더 나은 생활!\" == page.logo_slogan_title.get_text(), \"Switching to korean lang failed!\"","def test_fallback_language(self):\n x = SimpleModel()\n x.set_current_language(self.conf_fallback)\n x.tr_title = \"TITLE_FALLBACK\"\n\n x.set_current_language(self.other_lang1)\n x.tr_title = \"TITLE_XX\"\n x.save()\n\n with translation.override(self.other_lang2):\n x = SimpleModel.objects.get(pk=x.pk)\n self.assertEqual(x.tr_title, \"TITLE_FALLBACK\")","def test_legacyi18n08(self):\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_legacyi18n08(self):\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\n self.assertEqual(output, '2 plural')","def test_language_fix(self):\n #TODO\n \n for lang in self.LANGUAGES:\n activate(lang)\n \n self.assertEqual(lang, get_language())","def test_content(google_translator):\n # from bs4 import BeautifulSoup\n # assert 'GitHub' in BeautifulSoup(response.content).title.string\n assert google_translator.translate(text='좋은') == \"good\"","def test_langid_benchmark(basic_multilingual):\n examples = [\n {\"text\": \"contingentiam in naturalibus causis.\", \"label\": \"la\"},\n {\"text\": \"I jak opowiadał nieżyjący już pan Czesław\", \"label\": \"pl\"},\n {\"text\": \"Sonera gilt seit längerem als Übernahmekandidat\", \"label\": \"de\"},\n {\"text\": \"与银类似，汞也可以与空气中的硫化氢反应。\", \"label\": \"zh-hans\"},\n {\"text\": \"contradictionem implicat.\", \"label\": \"la\"},\n {\"text\": \"Bis zu Prozent gingen die Offerten etwa im\", \"label\": \"de\"},\n {\"text\": \"inneren Sicherheit vorgeschlagene Ausweitung der\", \"label\": \"de\"},\n {\"text\": \"Multimedia-PDA mit Mini-Tastatur\", \"label\": \"de\"},\n {\"text\": \"Ponášalo sa to na rovnicu o dvoch neznámych.\", \"label\": \"sk\"},\n {\"text\": \"이처럼 앞으로 심판의 그 날에 다시 올 메시아가 예수 그리스도이며 , 그는 모든 인류의\", \"label\": \"ko\"},\n {\"text\": \"Die Arbeitsgruppe bedauert , dass der weit über\", \"label\": \"de\"},\n {\"text\": \"И только раз довелось поговорить с ним не вполне\", \"label\": \"ru\"},\n {\"text\": \"de a-l lovi cu piciorul și conștiința că era\", \"label\": \"ro\"},\n {\"text\": \"relación coas pretensións do demandante e que, nos\", \"label\": \"gl\"},\n {\"text\": \"med petdeset in sedemdeset\", \"label\": \"sl\"},\n {\"text\": \"Catalunya; el Consell Comarcal del Vallès Oriental\", \"label\": \"ca\"},\n {\"text\": \"kunnen worden.\", \"label\": \"nl\"},\n {\"text\": \"Witkin je ve většině ohledů zcela jiný.\", \"label\": \"cs\"},\n {\"text\": \"lernen, so zu agieren, dass sie positive oder auch\", \"label\": \"de\"},\n {\"text\": \"olurmuş...\", \"label\": \"tr\"},\n {\"text\": \"sarcasmo de Altman, desde as «peruas» que discutem\", \"label\": \"pt\"},\n {\"text\": \"خلاف فوجداری مقدمہ درج کرے۔\", \"label\": \"ur\"},\n {\"text\": \"Norddal kommune :\", \"label\": \"no\"},\n {\"text\": \"dem Windows-.-Zeitalter , soll in diesem Jahr\", \"label\": \"de\"},\n {\"text\": \"przeklętych ucieleśniają mit poety-cygana,\", \"label\": \"pl\"},\n {\"text\": \"We do not believe the suspect has ties to this\", \"label\": \"en\"},\n {\"text\": \"groziņu pīšanu.\", \"label\": \"lv\"},\n {\"text\": \"Senior Vice-President David M. Thomas möchte\", \"label\": \"de\"},\n {\"text\": \"neomylně vybral nějakou knihu a začetl se.\", \"label\": \"cs\"},\n {\"text\": \"Statt dessen darf beispielsweise der Browser des\", \"label\": \"de\"},\n {\"text\": \"outubro, alcançando R $ bilhões em .\", \"label\": \"pt\"},\n {\"text\": \"(Porte, ), as it does other disciplines\", \"label\": \"en\"},\n {\"text\": \"uskupení se mylně domnívaly, že podporu\", \"label\": \"cs\"},\n {\"text\": \"Übernahme von Next Ende an dem System herum , das\", \"label\": \"de\"},\n {\"text\": \"No podemos decir a la Hacienda que los alemanes\", \"label\": \"es\"},\n {\"text\": \"и рѣста еи братья\", \"label\": \"orv\"},\n {\"text\": \"الذي اتخذ قرارا بتجميد اعلان الدولة الفلسطينية\", \"label\": \"ar\"},\n {\"text\": \"uurides Rootsi sõjaarhiivist toodud . sajandi\", \"label\": \"et\"},\n {\"text\": \"selskapets penger til å pusse opp sin enebolig på\", \"label\": \"no\"},\n {\"text\": \"средней полосе и севернее в Ярославской,\", \"label\": \"ru\"},\n {\"text\": \"il-massa żejda fil-ġemgħat u superġemgħat ta'\", \"label\": \"mt\"},\n {\"text\": \"The Global Beauties on internetilehekülg, mida\", \"label\": \"et\"},\n {\"text\": \"이스라엘 인들은 하나님이 그 큰 팔을 펴 이집트 인들을 치는 것을 보고 하나님을 두려워하며\", \"label\": \"ko\"},\n {\"text\": \"Snad ještě dodejme jeden ekonomický argument.\", \"label\": \"cs\"},\n {\"text\": \"Spalio d. vykusiame pirmajame rinkimų ture\", \"label\": \"lt\"},\n {\"text\": \"und schlechter Journalismus ein gutes Geschäft .\", \"label\": \"de\"},\n {\"text\": \"Du sodiečiai sėdi ant potvynio apsemtų namų stogo.\", \"label\": \"lt\"},\n {\"text\": \"цей є автентичним.\", \"label\": \"uk\"},\n {\"text\": \"Și îndegrabă fu cu îngerul mulțime de șireaguri\", \"label\": \"ro\"},\n {\"text\": \"sobra personal cualificado.\", \"label\": \"es\"},\n {\"text\": \"Tako se u Njemačkoj dvije trećine liječnika služe\", \"label\": \"hr\"},\n {\"text\": \"Dual-Athlon-Chipsatz noch in diesem Jahr\", \"label\": \"de\"},\n {\"text\": \"यहां तक कि चीन के चीफ ऑफ जनरल स्टाफ भी भारत का\", \"label\": \"hi\"},\n {\"text\": \"Li forestier du mont avale\", \"label\": \"fro\"},\n {\"text\": \"Netzwerken für Privatanwender zu bewundern .\", \"label\": \"de\"},\n {\"text\": \"만해는 승적을 가진 중이 결혼할 수 없다는 불교의 계율을 시대에 맞지 않는 것으로 보았다\", \"label\": \"ko\"},\n {\"text\": \"balance and weight distribution but not really for\", \"label\": \"en\"},\n {\"text\": \"og så e # tente vi opp den om morgonen å sfyrte\", \"label\": \"nn\"},\n {\"text\": \"변화는 의심의 여지가 없는 것이지만 반면에 진화는 논쟁의 씨앗이다 .\", \"label\": \"ko\"},\n {\"text\": \"puteare fac aceastea.\", \"label\": \"ro\"},\n {\"text\": \"Waitt seine Führungsmannschaft nicht dem\", \"label\": \"de\"},\n {\"text\": \"juhtimisega, tulid sealt.\", \"label\": \"et\"},\n {\"text\": \"Veränderungen .\", \"label\": \"de\"},\n {\"text\": \"banda en el Bayer Leverkusen de la Bundesliga de\", \"label\": \"es\"},\n {\"text\": \"В туже зиму посла всеволодъ сн҃а своѥго ст҃ослава\", \"label\": \"orv\"},\n {\"text\": \"пославъ приведе я мастеры ѿ грекъ\", \"label\": \"orv\"},\n {\"text\": \"En un nou escenari difícil d'imaginar fa poques\", \"label\": \"ca\"},\n {\"text\": \"καὶ γὰρ τινὲς αὐτοὺς εὐεργεσίαι εἶχον ἐκ Κροίσου\", \"label\": \"grc\"},\n {\"text\": \"직접적인 관련이 있다 .\", \"label\": \"ko\"},\n {\"text\": \"가까운 듯하면서도 멀다 .\", \"label\": \"ko\"},\n {\"text\": \"Er bietet ein ähnliches Leistungsniveau und\", \"label\": \"de\"},\n {\"text\": \"民都洛水牛是獨居的，並不會以群族聚居。\", \"label\": \"zh-hant\"},\n {\"text\": \"την τρομοκρατία.\", \"label\": \"el\"},\n {\"text\": \"hurbiltzen diren neurrian.\", \"label\": \"eu\"},\n {\"text\": \"Ah dimenticavo, ma tutta sta caciara per fare un\", \"label\": \"it\"},\n {\"text\": \"На первом этапе (-) прошла так называемая\", \"label\": \"ru\"},\n {\"text\": \"of games are on the market.\", \"label\": \"en\"},\n {\"text\": \"находится Мост дружбы, соединяющий узбекский и\", \"label\": \"ru\"},\n {\"text\": \"lessié je voldroie que li saint fussent aporté\", \"label\": \"fro\"},\n {\"text\": \"Дошла очередь и до Гималаев.\", \"label\": \"ru\"},\n {\"text\": \"vzácným suknem táhly pouští, si jednou chtěl do\", \"label\": \"cs\"},\n {\"text\": \"E no terceiro tipo sitúa a familias (%), nos que a\", \"label\": \"gl\"},\n {\"text\": \"وجابت دوريات امريكية وعراقية شوارع المدينة، فيما\", \"label\": \"ar\"},\n {\"text\": \"Jeg har bodd her i år .\", \"label\": \"no\"},\n {\"text\": \"Pohrozil, že odbory zostří postoj, pokud se\", \"label\": \"cs\"},\n {\"text\": \"tinham conseguido.\", \"label\": \"pt\"},\n {\"text\": \"Nicht-Erkrankten einen Anfangsverdacht für einen\", \"label\": \"de\"},\n {\"text\": \"permanece em aberto.\", \"label\": \"pt\"},\n {\"text\": \"questi possono promettere rendimenti fino a un\", \"label\": \"it\"},\n {\"text\": \"Tema juurutatud kahevedurisüsteemita oleksid\", \"label\": \"et\"},\n {\"text\": \"Поведение внешне простой игрушки оказалось\", \"label\": \"ru\"},\n {\"text\": \"Bundesländern war vom Börsenverein des Deutschen\", \"label\": \"de\"},\n {\"text\": \"acció, 'a mesura que avanci l'estiu, amb l'augment\", \"label\": \"ca\"},\n {\"text\": \"Dove trovare queste risorse? Jay Naidoo, ministro\", \"label\": \"it\"},\n {\"text\": \"essas gordurinhas.\", \"label\": \"pt\"},\n {\"text\": \"Im zweiten Schritt sollen im übernächsten Jahr\", \"label\": \"de\"},\n {\"text\": \"allveelaeva pole enam vaja, kuna külm sõda on läbi\", \"label\": \"et\"},\n {\"text\": \"उपद्रवी दुकानों को लूटने के साथ ही उनमें आग लगा\", \"label\": \"hi\"},\n {\"text\": \"@user nella sfortuna sei fortunata ..\", \"label\": \"it\"},\n {\"text\": \"математических школ в виде грозовых туч.\", \"label\": \"ru\"},\n {\"text\": \"No cambiaremos nunca nuestra forma de jugar por un\", \"label\": \"es\"},\n {\"text\": \"dla tej klasy ani wymogów minimalnych, z wyjątkiem\", \"label\": \"pl\"},\n {\"text\": \"en todo el mundo, mientras que en España consiguió\", \"label\": \"es\"},\n {\"text\": \"политики считать надежное обеспечение военной\", \"label\": \"ru\"},\n {\"text\": \"gogoratzen du, genio alemana delakoaren\", \"label\": \"eu\"},\n {\"text\": \"Бычий глаз.\", \"label\": \"ru\"},\n {\"text\": \"Opeření se v pravidelných obdobích obnovuje\", \"label\": \"cs\"},\n {\"text\": \"I no és només la seva, es tracta d'una resposta\", \"label\": \"ca\"},\n {\"text\": \"오경을 가르쳤다 .\", \"label\": \"ko\"},\n {\"text\": \"Nach der so genannten Start-up-Periode vergibt die\", \"label\": \"de\"},\n {\"text\": \"Saulista huomasi jo lapsena , että hänellä on\", \"label\": \"fi\"},\n {\"text\": \"Министерство культуры сочло нецелесообразным, и\", \"label\": \"ru\"},\n {\"text\": \"znepřátelené tábory v Tádžikistánu předseda\", \"label\": \"cs\"},\n {\"text\": \"καὶ ἦν ὁ λαὸς προσδοκῶν τὸν Ζαχαρίαν καὶ ἐθαύμαζον\", \"label\": \"grc\"},\n {\"text\": \"Вечером, в продукте, этот же человек говорил о\", \"label\": \"ru\"},\n {\"text\": \"lugar á formación de xuizos máis complexos.\", \"label\": \"gl\"},\n {\"text\": \"cheaper, in the end?\", \"label\": \"en\"},\n {\"text\": \"الوزارة في شأن صفقات بيع الشركات العامة التي تم\", \"label\": \"ar\"},\n {\"text\": \"tärkeintä elämässäni .\", \"label\": \"fi\"},\n {\"text\": \"Виконання Мінських угод було заблоковано Росією та\", \"label\": \"uk\"},\n {\"text\": \"Aby szybko rozpoznać żołnierzy desantu, należy\", \"label\": \"pl\"},\n {\"text\": \"Bankengeschäfte liegen vorn , sagte Strothmann .\", \"label\": \"de\"},\n {\"text\": \"продолжение работы.\", \"label\": \"ru\"},\n {\"text\": \"Metro AG plant Online-Offensive\", \"label\": \"de\"},\n {\"text\": \"nu vor veni, și să vor osîndi, aceia nu pot porni\", \"label\": \"ro\"},\n {\"text\": \"Ich denke , es geht in Wirklichkeit darum , NT bei\", \"label\": \"de\"},\n {\"text\": \"de turism care încasează contravaloarea\", \"label\": \"ro\"},\n {\"text\": \"Aurkaria itotzea da helburua, baloia lapurtu eta\", \"label\": \"eu\"},\n {\"text\": \"com a centre de formació en Tecnologies de la\", \"label\": \"ca\"},\n {\"text\": \"oportet igitur quod omne agens in agendo intendat\", \"label\": \"la\"},\n {\"text\": \"Jerzego Andrzejewskiego, oparty na chińskich\", \"label\": \"pl\"},\n {\"text\": \"sau một vài câu chuyện xã giao không dính dáng tới\", \"label\": \"vi\"},\n {\"text\": \"что экономическому прорыву жесткий авторитарный\", \"label\": \"ru\"},\n {\"text\": \"DRAM-Preisen scheinen DSPs ein\", \"label\": \"de\"},\n {\"text\": \"Jos dajan nubbái: Mana!\", \"label\": \"sme\"},\n {\"text\": \"toți carii ascultară de el să răsipiră.\", \"label\": \"ro\"},\n {\"text\": \"odpowiedzialności, które w systemie własności\", \"label\": \"pl\"},\n {\"text\": \"Dvomesečno potovanje do Mollenda v Peruju je\", \"label\": \"sl\"},\n {\"text\": \"d'entre les agències internacionals.\", \"label\": \"ca\"},\n {\"text\": \"Fahrzeugzugangssysteme gefertigt und an viele\", \"label\": \"de\"},\n {\"text\": \"in an answer to the sharers' petition in Cuthbert\", \"label\": \"en\"},\n {\"text\": \"Europa-Domain per Verordnung zu regeln .\", \"label\": \"de\"},\n {\"text\": \"#Balotelli. Su ebay prezzi stracciati per Silvio\", \"label\": \"it\"},\n {\"text\": \"Ne na košickém trávníku, ale už včera v letadle se\", \"label\": \"cs\"},\n {\"text\": \"zaměstnanosti a investičních strategií.\", \"label\": \"cs\"},\n {\"text\": \"Tatínku, udělej den\", \"label\": \"cs\"},\n {\"text\": \"frecuencia con Mary.\", \"label\": \"es\"},\n {\"text\": \"Свеаборге.\", \"label\": \"ru\"},\n {\"text\": \"opatření slovenské strany o certifikaci nejvíce\", \"label\": \"cs\"},\n {\"text\": \"En todas me decían: 'Espera que hagamos un estudio\", \"label\": \"es\"},\n {\"text\": \"Die Demonstration sollte nach Darstellung der\", \"label\": \"de\"},\n {\"text\": \"Ci vorrà un assoluto rigore se dietro i disavanzi\", \"label\": \"it\"},\n {\"text\": \"Tatínku, víš, že Honzovi odešla maminka?\", \"label\": \"cs\"},\n {\"text\": \"Die Anzahl der Rechner wuchs um % auf und die\", \"label\": \"de\"},\n {\"text\": \"האמריקאית על אדמת סעודיה עלולה לסבך את ישראל, אין\", \"label\": \"he\"},\n {\"text\": \"Volán Egyesülés, a Közlekedési Főfelügyelet is.\", \"label\": \"hu\"},\n {\"text\": \"Schejbala, který stejnou hru s velkým úspěchem\", \"label\": \"cs\"},\n {\"text\": \"depends on the data type of the field.\", \"label\": \"en\"},\n {\"text\": \"Umsatzwarnung zu Wochenbeginn zeitweise auf ein\", \"label\": \"de\"},\n {\"text\": \"niin heti nukun .\", \"label\": \"fi\"},\n {\"text\": \"Mobilfunkunternehmen gegen die Anwendung der so\", \"label\": \"de\"},\n {\"text\": \"sapessi le intenzioni del governo Monti e dell'UE\", \"label\": \"it\"},\n {\"text\": \"Di chi è figlia Martine Aubry?\", \"label\": \"it\"},\n {\"text\": \"avec le reste du monde.\", \"label\": \"fr\"},\n {\"text\": \"Այդ մաքոքը ինքնին նոր չէ, աշխարհը արդեն մի քանի\", \"label\": \"hy\"},\n {\"text\": \"și în cazul destrămării cenaclului.\", \"label\": \"ro\"},\n {\"text\": \"befriedigen kann , und ohne die auftretenden\", \"label\": \"de\"},\n {\"text\": \"Κύκνον τ̓ ἐξεναρεῖν καὶ ἀπὸ κλυτὰ τεύχεα δῦσαι.\", \"label\": \"grc\"},\n {\"text\": \"færdiguddannede.\", \"label\": \"da\"},\n {\"text\": \"Schmidt war Sohn eines Rittergutsbesitzers.\", \"label\": \"de\"},\n {\"text\": \"и вдаша попадь ѡпрати\", \"label\": \"orv\"},\n {\"text\": \"cine nu știe învățătură”.\", \"label\": \"ro\"},\n {\"text\": \"détacha et cette dernière tenta de tuer le jeune\", \"label\": \"fr\"},\n {\"text\": \"Der har saka også ei lengre forhistorie.\", \"label\": \"nn\"},\n {\"text\": \"Pieprz roztłuc w moździerzu, dodać do pasty,\", \"label\": \"pl\"},\n {\"text\": \"Лежа за гребнем оврага, как за бруствером, Ушаков\", \"label\": \"ru\"},\n {\"text\": \"gesucht habe, vielen Dank nochmals!\", \"label\": \"de\"},\n {\"text\": \"инструментальных сталей, повышения\", \"label\": \"ru\"},\n {\"text\": \"im Halbfinale Patrick Smith und im Finale dann\", \"label\": \"de\"},\n {\"text\": \"البنوك التريث في منح تسهيلات جديدة لمنتجي حديد\", \"label\": \"ar\"},\n {\"text\": \"una bolsa ventral, la cual se encuentra debajo de\", \"label\": \"es\"},\n {\"text\": \"za SETimes.\", \"label\": \"sr\"},\n {\"text\": \"de Irak, a un piloto italiano que había violado el\", \"label\": \"es\"},\n {\"text\": \"Er könne sich nicht erklären , wie die Zeitung auf\", \"label\": \"de\"},\n {\"text\": \"Прохорова.\", \"label\": \"ru\"},\n {\"text\": \"la democrazia perde sulla tecnocrazia? #\", \"label\": \"it\"},\n {\"text\": \"entre ambas instituciones, confirmó al medio que\", \"label\": \"es\"},\n {\"text\": \"Austlandet, vart det funne om lag førti\", \"label\": \"nn\"},\n {\"text\": \"уровнями власти.\", \"label\": \"ru\"},\n {\"text\": \"Dá tedy primáři úplatek, a často ne malý.\", \"label\": \"cs\"},\n {\"text\": \"brillantes del acto, al llevar a cabo en el\", \"label\": \"es\"},\n {\"text\": \"eee druga zadeva je majhen priročen gre kamorkoli\", \"label\": \"sl\"},\n {\"text\": \"Das ATX-Board paßt in herkömmliche PC-ATX-Gehäuse\", \"label\": \"de\"},\n {\"text\": \"Za vodné bylo v prvním pololetí zaplaceno v ČR\", \"label\": \"cs\"},\n {\"text\": \"Даже на полсантиметра.\", \"label\": \"ru\"},\n {\"text\": \"com la del primer tinent d'alcalde en funcions,\", \"label\": \"ca\"},\n {\"text\": \"кількох оповідань в цілості — щось на зразок того\", \"label\": \"uk\"},\n {\"text\": \"sed ad divitias congregandas, vel superfluum\", \"label\": \"la\"},\n {\"text\": \"Norma Talmadge, spela mot Valentino i en version\", \"label\": \"sv\"},\n {\"text\": \"Dlatego chciał się jej oświadczyć w niezwykłym\", \"label\": \"pl\"},\n {\"text\": \"будут выступать на одинаковых снарядах.\", \"label\": \"ru\"},\n {\"text\": \"Orang-orang terbunuh di sana.\", \"label\": \"id\"},\n {\"text\": \"لدى رايت شقيق اسمه أوسكار, وهو يعمل كرسام للكتب\", \"label\": \"ar\"},\n {\"text\": \"Wirklichkeit verlagerten und kaum noch\", \"label\": \"de\"},\n {\"text\": \"как перемешивают костяшки перед игрой в домино, и\", \"label\": \"ru\"},\n {\"text\": \"В средине дня, когда солнце светило в нашу\", \"label\": \"ru\"},\n {\"text\": \"d'aventure aux rôles de jeune romantique avec une\", \"label\": \"fr\"},\n {\"text\": \"My teď hledáme organizace, jež by s námi chtěly\", \"label\": \"cs\"},\n {\"text\": \"Urteilsfähigkeit einbüßen , wenn ich eigene\", \"label\": \"de\"},\n {\"text\": \"sua appartenenza anche a voci diverse da quella in\", \"label\": \"it\"},\n {\"text\": \"Aufträge dieses Jahr verdoppeln werden .\", \"label\": \"de\"},\n {\"text\": \"M.E.: Miała szanse mnie odnaleźć, gdyby naprawdę\", \"label\": \"pl\"},\n {\"text\": \"secundum contactum virtutis, cum careat dimensiva\", \"label\": \"la\"},\n {\"text\": \"ezinbestekoa dela esan zuen.\", \"label\": \"eu\"},\n {\"text\": \"Anek hurbiltzeko eskatzen zion besaulkitik, eta\", \"label\": \"eu\"},\n {\"text\": \"perfectius alio videat, quamvis uterque videat\", \"label\": \"la\"},\n {\"text\": \"Die Strecke war anspruchsvoll und führte unter\", \"label\": \"de\"},\n {\"text\": \"саморазоблачительным уроком, западные СМИ не\", \"label\": \"ru\"},\n {\"text\": \"han representerer radikal islamisme .\", \"label\": \"no\"},\n {\"text\": \"Què s'hi respira pel que fa a la reforma del\", \"label\": \"ca\"},\n {\"text\": \"previsto para também ser desconstruido.\", \"label\": \"pt\"},\n {\"text\": \"Ὠκεανοῦ βαθυκόλποις ἄνθεά τ̓ αἰνυμένην, ῥόδα καὶ\", \"label\": \"grc\"},\n {\"text\": \"para jovens de a anos nos Cieps.\", \"label\": \"pt\"},\n {\"text\": \"संघर्ष को अंजाम तक पहुंचाने का ऐलान किया है ।\", \"label\": \"hi\"},\n {\"text\": \"objeví i u nás.\", \"label\": \"cs\"},\n {\"text\": \"kvitteringer.\", \"label\": \"da\"},\n {\"text\": \"This report is no exception.\", \"label\": \"en\"},\n {\"text\": \"Разлепват доносниците до избирателните списъци\", \"label\": \"bg\"},\n {\"text\": \"anderem ihre Bewegungsfreiheit in den USA\", \"label\": \"de\"},\n {\"text\": \"Ñu tegoon ca kaw gor ña ay njotti bopp yu kenn\", \"label\": \"wo\"},\n {\"text\": \"Struktur kann beispielsweise der Schwerpunkt mehr\", \"label\": \"de\"},\n {\"text\": \"% la velocidad permitida, la sanción es muy grave.\", \"label\": \"es\"},\n {\"text\": \"Teles-Einstieg in ADSL-Markt\", \"label\": \"de\"},\n {\"text\": \"ettekäändeks liiga suure osamaksu.\", \"label\": \"et\"},\n {\"text\": \"als Indiz für die geänderte Marktpolitik des\", \"label\": \"de\"},\n {\"text\": \"quod quidem aperte consequitur ponentes\", \"label\": \"la\"},\n {\"text\": \"de negociación para el próximo de junio.\", \"label\": \"es\"},\n {\"text\": \"Tyto důmyslné dekorace doznaly v poslední době\", \"label\": \"cs\"},\n {\"text\": \"največjega uspeha doslej.\", \"label\": \"sl\"},\n {\"text\": \"Paul Allen je jedan od suosnivača Interval\", \"label\": \"hr\"},\n {\"text\": \"Federal (Seac / DF) eo Sindicato das Empresas de\", \"label\": \"pt\"},\n {\"text\": \"Quartal mit . Mark gegenüber dem gleichen Quartal\", \"label\": \"de\"},\n {\"text\": \"otros clubes y del Barça B saldrán varios\", \"label\": \"es\"},\n {\"text\": \"Jaskula (Pol.) -\", \"label\": \"cs\"},\n {\"text\": \"umožnily říci, že je možné přejít k mnohem\", \"label\": \"cs\"},\n {\"text\": \"اعلن الجنرال تومي فرانكس قائد القوات الامريكية\", \"label\": \"ar\"},\n {\"text\": \"Telekom-Chef Ron Sommer und der Vorstandssprecher\", \"label\": \"de\"},\n {\"text\": \"My, jako průmyslový a finanční holding, můžeme\", \"label\": \"cs\"},\n {\"text\": \"voorlichting onder andere betrekking kan hebben:\", \"label\": \"nl\"},\n {\"text\": \"Hinrichtung geistig Behinderter applaudiert oder\", \"label\": \"de\"},\n {\"text\": \"wie beispielsweise Anzahl erzielte Klicks ,\", \"label\": \"de\"},\n {\"text\": \"Intel-PC-SDRAM-Spezifikation in der Version . (\", \"label\": \"de\"},\n {\"text\": \"plângere în termen de zile de la comunicarea\", \"label\": \"ro\"},\n {\"text\": \"и Испания ще изгубят втория си комисар в ЕК.\", \"label\": \"bg\"},\n {\"text\": \"इसके चलते इस आदिवासी जनजाति का क्षरण हो रहा है ।\", \"label\": \"hi\"},\n {\"text\": \"aunque se mostró contrario a establecer un\", \"label\": \"es\"},\n {\"text\": \"des letzten Jahres von auf Millionen Euro .\", \"label\": \"de\"},\n {\"text\": \"Ankara se također poziva da u cijelosti ratificira\", \"label\": \"hr\"},\n {\"text\": \"herunterlädt .\", \"label\": \"de\"},\n {\"text\": \"стрессовую ситуацию для организма, каковой\", \"label\": \"ru\"},\n {\"text\": \"Státního shromáždění (parlamentu).\", \"label\": \"cs\"},\n {\"text\": \"diskutieren , ob und wie dieser Dienst weiterhin\", \"label\": \"de\"},\n {\"text\": \"Verbindungen zu FPÖ-nahen Polizisten gepflegt und\", \"label\": \"de\"},\n {\"text\": \"Pražského volebního lídra ovšem nevybírá Miloš\", \"label\": \"cs\"},\n {\"text\": \"Nach einem Bericht der Washington Post bleibt das\", \"label\": \"de\"},\n {\"text\": \"للوضع آنذاك، لكني في قرارة نفسي كنت سعيداً لما\", \"label\": \"ar\"},\n {\"text\": \"не желаят запазването на статуквото.\", \"label\": \"bg\"},\n {\"text\": \"Offenburg gewesen .\", \"label\": \"de\"},\n {\"text\": \"ἐὰν ὑμῖν εἴπω οὐ μὴ πιστεύσητε\", \"label\": \"grc\"},\n {\"text\": \"all'odiato compagno di squadra Prost, il quale\", \"label\": \"it\"},\n {\"text\": \"historischen Gänselieselbrunnens.\", \"label\": \"de\"},\n {\"text\": \"למידע מלווייני הריגול האמריקאיים העוקבים אחר\", \"label\": \"he\"},\n {\"text\": \"οὐδὲν ἄρα διαφέρεις Ἀμάσιος τοῦ Ἠλείου, ὃν\", \"label\": \"grc\"},\n {\"text\": \"movementos migratorios.\", \"label\": \"gl\"},\n {\"text\": \"Handy und ein Spracherkennungsprogramm sämtliche\", \"label\": \"de\"},\n {\"text\": \"Kümne aasta jooksul on Eestisse ohjeldamatult\", \"label\": \"et\"},\n {\"text\": \"H.G. Bücknera.\", \"label\": \"pl\"},\n {\"text\": \"protiv krijumčarenja, ili pak traženju ukidanja\", \"label\": \"hr\"},\n {\"text\": \"Topware-Anteile mehrere Millionen Mark gefordert\", \"label\": \"de\"},\n {\"text\": \"Maar de mensen die nu over Van Dijk bij FC Twente\", \"label\": \"nl\"},\n {\"text\": \"poidan experimentar as percepcións do interesado,\", \"label\": \"gl\"},\n {\"text\": \"Miał przecież w kieszeni nóż.\", \"label\": \"pl\"},\n {\"text\": \"Avšak žádná z nich nepronikla za hranice přímé\", \"label\": \"cs\"},\n {\"text\": \"esim. helpottamalla luottoja muiden\", \"label\": \"fi\"},\n {\"text\": \"Podle předběžných výsledků zvítězila v\", \"label\": \"cs\"},\n {\"text\": \"Nicht nur das Web-Frontend , auch die\", \"label\": \"de\"},\n {\"text\": \"Regierungsinstitutionen oder Universitäten bei\", \"label\": \"de\"},\n {\"text\": \"Խուլեն Լոպետեգիին, պատճառաբանելով, որ վերջինս\", \"label\": \"hy\"},\n {\"text\": \"Афганистана, где в последние дни идут ожесточенные\", \"label\": \"ru\"},\n {\"text\": \"лѧхове же не идоша\", \"label\": \"orv\"},\n {\"text\": \"Mit Hilfe von IBMs Chip-Management-Systemen sollen\", \"label\": \"de\"},\n {\"text\": \", als Manager zu Telefonica zu wechseln .\", \"label\": \"de\"},\n {\"text\": \"którym zajmuje się człowiek, zmienia go i pozwala\", \"label\": \"pl\"},\n {\"text\": \"činí kyperských liber, to je asi USD.\", \"label\": \"cs\"},\n {\"text\": \"Studienplätze getauscht werden .\", \"label\": \"de\"},\n {\"text\": \"учёных, орнитологов признают вид.\", \"label\": \"ru\"},\n {\"text\": \"acordare a concediilor prevăzute de legislațiile\", \"label\": \"ro\"},\n {\"text\": \"at større innsats for fornybar, berekraftig energi\", \"label\": \"nn\"},\n {\"text\": \"Politiet veit ikkje kor mange personar som deltok\", \"label\": \"nn\"},\n {\"text\": \"offentligheten av unge , sinte menn som har\", \"label\": \"no\"},\n {\"text\": \"însuși în jurul lapunei, care încet DISPARE în\", \"label\": \"ro\"},\n {\"text\": \"O motivo da decisão é evitar uma sobrecarga ainda\", \"label\": \"pt\"},\n {\"text\": \"El Apostolado de la prensa contribuye en modo\", \"label\": \"es\"},\n {\"text\": \"Teltow ( Kreis Teltow-Fläming ) ist Schmitt einer\", \"label\": \"de\"},\n {\"text\": \"grozījumus un iesniegt tos Apvienoto Nāciju\", \"label\": \"lv\"},\n {\"text\": \"Gestalt einer deutschen Nationalmannschaft als\", \"label\": \"de\"},\n {\"text\": \"D überholt zu haben , konterte am heutigen Montag\", \"label\": \"de\"},\n {\"text\": \"Softwarehersteller Oracle hat im dritten Quartal\", \"label\": \"de\"},\n {\"text\": \"Během nich se ekonomické podmínky mohou radikálně\", \"label\": \"cs\"},\n {\"text\": \"Dziki kot w górach zeskakuje z kamienia.\", \"label\": \"pl\"},\n {\"text\": \"Ačkoliv ligový nováček prohrál, opět potvrdil, že\", \"label\": \"cs\"},\n {\"text\": \"des Tages , Portraits internationaler Stars sowie\", \"label\": \"de\"},\n {\"text\": \"Communicator bekannt wurde .\", \"label\": \"de\"},\n {\"text\": \"τῷ δ’ ἄρα καὶ αὐτῷ ἡ γυνή ἐπίτεξ ἐοῦσα πᾶσαν\", \"label\": \"grc\"},\n {\"text\": \"Triadú tenia, mentre redactava 'Dies de memòria',\", \"label\": \"ca\"},\n {\"text\": \"دستهجمعی در درخشندگی ماه سیمگون زمزمه ستاینده و\", \"label\": \"fa\"},\n {\"text\": \"Книгу, наполненную мелочной заботой об одежде,\", \"label\": \"ru\"},\n {\"text\": \"putares canem leporem persequi.\", \"label\": \"la\"},\n {\"text\": \"В дальнейшем эта яркость слегка померкла, но в\", \"label\": \"ru\"},\n {\"text\": \"offizielles Verfahren gegen die Telekom\", \"label\": \"de\"},\n {\"text\": \"podrían haber sido habitantes de la Península\", \"label\": \"es\"},\n {\"text\": \"Grundlage für dieses Verfahren sind spezielle\", \"label\": \"de\"},\n {\"text\": \"Rechtsausschuß vorgelegten Entwurf der Richtlinie\", \"label\": \"de\"},\n {\"text\": \"Im so genannten Portalgeschäft sei das Unternehmen\", \"label\": \"de\"},\n {\"text\": \"ⲏ ⲉⲓϣⲁⲛϥⲓ ⲛⲉⲓⲇⲱⲗⲟⲛ ⲉⲧϩⲙⲡⲉⲕⲏⲓ ⲙⲏ ⲉⲓⲛⲁϣϩⲱⲡ ⲟⲛ ⲙⲡⲣⲏ\", \"label\": \"cop\"},\n {\"text\": \"juego podían matar a cualquier herbívoro, pero\", \"label\": \"es\"},\n {\"text\": \"Nach Angaben von Axent nutzen Unternehmen aus der\", \"label\": \"de\"},\n {\"text\": \"hrdiny Havlovy Zahradní slavnosti (premiéra ) se\", \"label\": \"cs\"},\n {\"text\": \"Een zin van heb ik jou daar\", \"label\": \"nl\"},\n {\"text\": \"hat sein Hirn an der CeBIT-Kasse vergessen .\", \"label\": \"de\"},\n {\"text\": \"καὶ τοὺς ἐκπλαγέντας οὐκ ἔχειν ἔτι ἐλεγχομένους\", \"label\": \"grc\"},\n {\"text\": \"nachgewiesenen langfristigen Kosten , sowie den im\", \"label\": \"de\"},\n {\"text\": \"jučer nakon četiri dana putovanja u Helsinki.\", \"label\": \"hr\"},\n {\"text\": \"pašto paslaugos teikėjas gali susitarti su\", \"label\": \"lt\"},\n {\"text\": \"В результате, эти золотые кадры переходят из одной\", \"label\": \"ru\"},\n {\"text\": \"द फाइव-ईयर एंगेजमेंट में अभिनय किया जिसमें जैसन\", \"label\": \"hi\"},\n {\"text\": \"výpis o počtu akcií.\", \"label\": \"cs\"},\n {\"text\": \"Enfin, elles arrivent à un pavillon chinois\", \"label\": \"fr\"},\n {\"text\": \"Tentu saja, tren yang berhubungandengan\", \"label\": \"id\"},\n {\"text\": \"Arbeidarpartiet og SV har sikra seg fleirtal mot\", \"label\": \"nn\"},\n {\"text\": \"eles: 'Tudo isso está errado' , disse um\", \"label\": \"pt\"},\n {\"text\": \"The islands are in their own time zone, minutes\", \"label\": \"en\"},\n {\"text\": \"Auswahl debütierte er am .\", \"label\": \"de\"},\n {\"text\": \"Bu komisyonlar, arazilerini satın almak için\", \"label\": \"tr\"},\n {\"text\": \"Geschütze gegen Redmond aufgefahren .\", \"label\": \"de\"},\n {\"text\": \"Time scything the hours, but at the top, over the\", \"label\": \"en\"},\n {\"text\": \"Di musim semi , berharap mengadaptasi Tintin untuk\", \"label\": \"id\"},\n {\"text\": \"крупнейшей геополитической катастрофой XX века.\", \"label\": \"ru\"},\n {\"text\": \"Rajojen avaaminen ei suju ongelmitta .\", \"label\": \"fi\"},\n {\"text\": \"непроницаемым, как для СССР.\", \"label\": \"ru\"},\n {\"text\": \"Ma non mancano le polemiche.\", \"label\": \"it\"},\n {\"text\": \"Internet als Ort politischer Diskussion und auch\", \"label\": \"de\"},\n {\"text\": \"incomplets.\", \"label\": \"ca\"},\n {\"text\": \"Su padre luchó al lado de Luis Moya, primer Jefe\", \"label\": \"es\"},\n {\"text\": \"informazione.\", \"label\": \"it\"},\n {\"text\": \"Primacom bietet für Telekom-Kabelnetz\", \"label\": \"de\"},\n {\"text\": \"Oświadczenie prezydencji w imieniu Unii\", \"label\": \"pl\"},\n {\"text\": \"foran rattet i familiens gamle Baleno hvis døra på\", \"label\": \"no\"},\n {\"text\": \"[speaker:laughter]\", \"label\": \"sl\"},\n {\"text\": \"Dog med langt mindre utstyr med seg.\", \"label\": \"nn\"},\n {\"text\": \"dass es nicht schon mit der anfänglichen\", \"label\": \"de\"},\n {\"text\": \"इस पर दोनों पक्षों में नोकझोंक शुरू हो गई ।\", \"label\": \"hi\"},\n {\"text\": \"کے ترجمان منیش تیواری اور دگ وجئے سنگھ نے بھی یہ\", \"label\": \"ur\"},\n {\"text\": \"dell'Assemblea Costituente che posseggono i\", \"label\": \"it\"},\n {\"text\": \"и аште вьси съблазнѧтъ сѧ нъ не азъ\", \"label\": \"cu\"},\n {\"text\": \"In Irvine hat auch das Logistikunternehmen Atlas\", \"label\": \"de\"},\n {\"text\": \"законодательных норм, принимаемых существующей\", \"label\": \"ru\"},\n {\"text\": \"Κροίσῳ προτείνων τὰς χεῖρας ἐπικατασφάξαι μιν\", \"label\": \"grc\"},\n {\"text\": \"МИНУСЫ: ИНФЛЯЦИЯ И КРИЗИС В ЖИВОТНОВОДСТВЕ.\", \"label\": \"ru\"},\n {\"text\": \"unterschiedlicher Meinung .\", \"label\": \"de\"},\n {\"text\": \"Jospa joku ystävällinen sielu auttaisi kassieni\", \"label\": \"fi\"},\n {\"text\": \"Añadió que, en el futuro se harán otros\", \"label\": \"es\"},\n {\"text\": \"Sessiz tonlama hem Fince, hem de Kuzey Sami\", \"label\": \"tr\"},\n {\"text\": \"nicht ihnen gehört und sie nicht alles , was sie\", \"label\": \"de\"},\n {\"text\": \"Etelästä Kuivajärveen laskee Tammelan Liesjärvestä\", \"label\": \"fi\"},\n {\"text\": \"ICANNs Vorsitzender Vint Cerf warb mit dem Hinweis\", \"label\": \"de\"},\n {\"text\": \"Norsk politikk frå til kan dermed, i\", \"label\": \"nn\"},\n {\"text\": \"Głosowało posłów.\", \"label\": \"pl\"},\n {\"text\": \"Danny Jones -- smithjones@ev.net\", \"label\": \"en\"},\n {\"text\": \"sebeuvědomění moderní civilizace sehrála lučavka\", \"label\": \"cs\"},\n {\"text\": \"относительно спокойный сон: тому гарантия\", \"label\": \"ru\"},\n {\"text\": \"A halte voiz prist li pedra a crïer\", \"label\": \"fro\"},\n {\"text\": \"آنها امیدوارند این واکسن بهزودی در دسترس بیماران\", \"label\": \"fa\"},\n {\"text\": \"vlastní důstojnou vousatou tváří.\", \"label\": \"cs\"},\n {\"text\": \"ora aprire la strada a nuove cause e alimentare il\", \"label\": \"it\"},\n {\"text\": \"Die Zahl der Vielleser nahm von auf Prozent zu ,\", \"label\": \"de\"},\n {\"text\": \"Finanzvorstand von Hotline-Dienstleister InfoGenie\", \"label\": \"de\"},\n {\"text\": \"entwickeln .\", \"label\": \"de\"},\n {\"text\": \"incolumità pubblica.\", \"label\": \"it\"},\n {\"text\": \"lehtija televisiomainonta\", \"label\": \"fi\"},\n {\"text\": \"joistakin kohdista eri mieltä.\", \"label\": \"fi\"},\n {\"text\": \"Hlavně anglická nezávislá scéna, Dead Can Dance,\", \"label\": \"cs\"},\n {\"text\": \"pásmech od do bodů bodové stupnice.\", \"label\": \"cs\"},\n {\"text\": \"Zu Beginn des Ersten Weltkrieges zählte das\", \"label\": \"de\"},\n {\"text\": \"Així van sorgir, damunt els antics cementiris,\", \"label\": \"ca\"},\n {\"text\": \"In manchem Gedicht der spätern Alten, wie zum\", \"label\": \"de\"},\n {\"text\": \"gaweihaida jah insandida in þana fairƕu jus qiþiþ\", \"label\": \"got\"},\n {\"text\": \"Beides sollte gelöscht werden!\", \"label\": \"de\"},\n {\"text\": \"modifiqués la seva petició inicial de anys de\", \"label\": \"ca\"},\n {\"text\": \"В день открытия симпозиума состоялась закладка\", \"label\": \"ru\"},\n {\"text\": \"tõestatud.\", \"label\": \"et\"},\n {\"text\": \"ἵππῳ πίπτει αὐτοῦ ταύτῃ\", \"label\": \"grc\"},\n {\"text\": \"bisher nie enttäuscht!\", \"label\": \"de\"},\n {\"text\": \"De bohte ollu tuollárat ja suttolaččat ja\", \"label\": \"sme\"},\n {\"text\": \"Klarsignal från röstlängdsläsaren, tre tryck i\", \"label\": \"sv\"},\n {\"text\": \"Tvůrcem nového termínu je Joseph Fisher.\", \"label\": \"cs\"},\n {\"text\": \"Nie miałem czasu na reakcję twierdzi Norbert,\", \"label\": \"pl\"},\n {\"text\": \"potentia Schöpfer.\", \"label\": \"de\"},\n {\"text\": \"Un poquito caro, pero vale mucho la pena;\", \"label\": \"es\"},\n {\"text\": \"οὔ τε γὰρ ἴφθιμοι Λύκιοι Δαναῶν ἐδύναντο τεῖχος\", \"label\": \"grc\"},\n {\"text\": \"vajec, sladového výtažku a některých vitamínových\", \"label\": \"cs\"},\n {\"text\": \"Настоящие герои, те, чьи истории потом\", \"label\": \"ru\"},\n {\"text\": \"praesumptio:\", \"label\": \"la\"},\n {\"text\": \"Olin justkui nende vastutusel.\", \"label\": \"et\"},\n {\"text\": \"Jokainen keinahdus tuo lähemmäksi hetkeä jolloin\", \"label\": \"fi\"},\n {\"text\": \"ekonomicky výhodných způsobů odvodnění těžkých,\", \"label\": \"cs\"},\n {\"text\": \"Poprvé ve své historii dokázala v kvalifikaci pro\", \"label\": \"cs\"},\n {\"text\": \"zpracovatelského a spotřebního průmyslu bude nutné\", \"label\": \"cs\"},\n {\"text\": \"Windows CE zu integrieren .\", \"label\": \"de\"},\n {\"text\": \"Armangué, a través d'un decret, ordenés l'aturada\", \"label\": \"ca\"},\n {\"text\": \"to, co nás Evropany spojuje, než to, co nás od\", \"label\": \"cs\"},\n {\"text\": \"ergänzt durch einen gesetzlich verankertes\", \"label\": \"de\"},\n {\"text\": \"Насчитал, что с начала года всего три дня были\", \"label\": \"ru\"},\n {\"text\": \"Borisovu tražeći od njega da prihvati njenu\", \"label\": \"sr\"},\n {\"text\": \"la presenza di ben veleni diversi: . chili di\", \"label\": \"it\"},\n {\"text\": \"καὶ τῶν ἐκλεκτῶν ἀγγέλων ἵνα ταῦτα φυλάξῃς χωρὶς\", \"label\": \"grc\"},\n {\"text\": \"pretraživale obližnju bolnicu i stambene zgrade u\", \"label\": \"hr\"},\n {\"text\": \"An rund Katzen habe Wolf seine Spiele getestet ,\", \"label\": \"de\"},\n {\"text\": \"investigating since March.\", \"label\": \"en\"},\n {\"text\": \"Tonböden (Mullböden).\", \"label\": \"de\"},\n {\"text\": \"Stálý dopisovatel LN v SRN Bedřich Utitz\", \"label\": \"cs\"},\n {\"text\": \"červnu předložené smlouvy.\", \"label\": \"cs\"},\n {\"text\": \"πνεύματι ᾧ ἐλάλει\", \"label\": \"grc\"},\n {\"text\": \".%의 신장세를 보였다.\", \"label\": \"ko\"},\n {\"text\": \"Foae verde, foi de nuc, Prin pădure, prin colnic,\", \"label\": \"ro\"},\n {\"text\": \"διαπέμψας ἄλλους ἄλλῃ τοὺς μὲν ἐς Δελφοὺς ἰέναι\", \"label\": \"grc\"},\n {\"text\": \"المسلمين أو أي تيار سياسي طالما عمل ذلك التيار في\", \"label\": \"ar\"},\n {\"text\": \"As informações são da Dow Jones.\", \"label\": \"pt\"},\n {\"text\": \"Milliarde DM ausgestattet sein .\", \"label\": \"de\"},\n {\"text\": \"De utgår fortfarande från att kvinnans jämlikhet\", \"label\": \"sv\"},\n {\"text\": \"Sneeuw maakte in Davos bij de voorbereiding een\", \"label\": \"nl\"},\n {\"text\": \"De ahí que en este mercado puedan negociarse\", \"label\": \"es\"},\n {\"text\": \"intenzívnějšímu sbírání a studiu.\", \"label\": \"cs\"},\n {\"text\": \"और औसकर ४.० पैकेज का प्रयोग किया गया है ।\", \"label\": \"hi\"},\n {\"text\": \"Adipati Kuningan karena Kuningan menjadi bagian\", \"label\": \"id\"},\n {\"text\": \"Svako je bar jednom poželeo da mašine prosto umeju\", \"label\": \"sr\"},\n {\"text\": \"Im vergangenen Jahr haben die Regierungen einen\", \"label\": \"de\"},\n {\"text\": \"durat motus, aliquid fit et non est;\", \"label\": \"la\"},\n {\"text\": \"Dominować będą piosenki do tekstów Edwarda\", \"label\": \"pl\"},\n {\"text\": \"beantwortet .\", \"label\": \"de\"},\n {\"text\": \"О гуманитариях было кому рассказывать, а вот за\", \"label\": \"ru\"},\n {\"text\": \"Helsingin kaupunki riitautti vuokrasopimuksen\", \"label\": \"fi\"},\n {\"text\": \"chợt tan biến.\", \"label\": \"vi\"},\n {\"text\": \"avtomobil ločuje od drugih.\", \"label\": \"sl\"},\n {\"text\": \"Congress has proven itself ineffective as a body.\", \"label\": \"en\"},\n {\"text\": \"मैक्सिको ने इस तरह का शो इस समय आयोजित करने का\", \"label\": \"hi\"},\n {\"text\": \"No minimum order amount.\", \"label\": \"en\"},\n {\"text\": \"Convertassa .\", \"label\": \"fi\"},\n {\"text\": \"Как это можно сделать?\", \"label\": \"ru\"},\n {\"text\": \"tha mi creidsinn gu robh iad ceart cho saor shuas\", \"label\": \"gd\"},\n {\"text\": \"실제 일제는 이런 만해의 논리를 묵살하고 한반도를 침략한 다음 , 이어 만주를 침략하고\", \"label\": \"ko\"},\n {\"text\": \"Da un semplice richiamo all'ordine fino a grandi\", \"label\": \"it\"},\n {\"text\": \"pozoruhodný nejen po umělecké stránce, jež\", \"label\": \"cs\"},\n {\"text\": \"La comida y el servicio aprueban.\", \"label\": \"es\"},\n {\"text\": \"again, connected not with each other but to the\", \"label\": \"en\"},\n {\"text\": \"Protokol výslovně stanoví, že nikdo nemůže být\", \"label\": \"cs\"},\n {\"text\": \"ఒక విషయం అడగాలని ఉంది .\", \"label\": \"te\"},\n {\"text\": \"Безгранично почитая дирекцию, ловя на лету каждое\", \"label\": \"ru\"},\n {\"text\": \"rovnoběžných růstových vrstev, zůstávají krychlové\", \"label\": \"cs\"},\n {\"text\": \"प्रवेश और पूर्व प्रधानमंत्री लाल बहादुर शास्त्री\", \"label\": \"hi\"},\n {\"text\": \"Bronzen medaille in de Europese marathon.\", \"label\": \"nl\"},\n {\"text\": \"- gadu vecumā viņi to nesaprot.\", \"label\": \"lv\"},\n {\"text\": \"Realizó sus estudios primarios en la Escuela Julia\", \"label\": \"es\"},\n {\"text\": \"cuartos de final, su clasificación para la final a\", \"label\": \"es\"},\n {\"text\": \"Sem si pro něho přiletí americký raketoplán, na\", \"label\": \"cs\"},\n {\"text\": \"Way to go!\", \"label\": \"en\"},\n {\"text\": \"gehört der neuen SPD-Führung unter Parteichef\", \"label\": \"de\"},\n {\"text\": \"Somit simuliert der Player mit einer GByte-Platte\", \"label\": \"de\"},\n {\"text\": \"Berufung auf kommissionsnahe Kreise , die bereits\", \"label\": \"de\"},\n {\"text\": \"Dist Clarïen\", \"label\": \"fro\"},\n {\"text\": \"Schon nach den Gerüchten , die Telekom wolle den\", \"label\": \"de\"},\n {\"text\": \"Software von NetObjects ist nach Angaben des\", \"label\": \"de\"},\n {\"text\": \"si enim per legem iustitia ergo Christus gratis\", \"label\": \"la\"},\n {\"text\": \"ducerent in ipsam magis quam in corpus christi,\", \"label\": \"la\"},\n {\"text\": \"Neustar-Melbourne-IT-Partnerschaft NeuLevel .\", \"label\": \"de\"},\n {\"text\": \"forderte dagegen seine drastische Verschärfung.\", \"label\": \"de\"},\n {\"text\": \"pemmican på hundrede forskellige måder.\", \"label\": \"da\"},\n {\"text\": \"Lehån, själv matematiklärare, visar hur den nya\", \"label\": \"sv\"},\n {\"text\": \"I highly recommend his shop.\", \"label\": \"en\"},\n {\"text\": \"verità, giovani fedeli prostratevi #amen\", \"label\": \"it\"},\n {\"text\": \"उत्तर प्रदेश के अध्यक्ष पद से हटाए गए विनय कटियार\", \"label\": \"hi\"},\n {\"text\": \"() روزی مےں کشادگی ہوتی ہے۔\", \"label\": \"ur\"},\n {\"text\": \"Prozessorgeschäft profitieren kann , stellen\", \"label\": \"de\"},\n {\"text\": \"školy začalo počítat pytle s moukou a zjistilo, že\", \"label\": \"cs\"},\n {\"text\": \"प्रभावशाली पर गैर सरकारी लोगों के घरों में भी\", \"label\": \"hi\"},\n {\"text\": \"geschichtslos , oder eine Farce , wie sich\", \"label\": \"de\"},\n {\"text\": \"Ústrednými mocnosťami v marci však spôsobilo, že\", \"label\": \"sk\"},\n {\"text\": \"التسليح بدون مبرر، واستمرار الأضرار الناجمة عن فرض\", \"label\": \"ar\"},\n {\"text\": \"Například Pedagogická fakulta Univerzity Karlovy\", \"label\": \"cs\"},\n {\"text\": \"nostris ut eriperet nos de praesenti saeculo\", \"label\": \"la\"}]\n \n docs = [Document([], text=example[\"text\"]) for example in examples]\n gold_labels = [example[\"label\"] for example in examples]\n basic_multilingual(docs)\n accuracy = sum([(doc.lang == label) for doc,label in zip(docs,gold_labels)])/len(docs)\n assert accuracy >= 0.98","def test_i18n03(self):\n output = self.engine.render_to_string(\"i18n03\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"Å\")","def test_00_i18n_anonymous(self):\r\n # First default 'en' locale\r\n with self.app as c:\r\n err_msg = \"The page should be in English\"\r\n res = c.get('/', headers=[('Accept-Language', 'en')])\r\n assert \"Community\" in res.data, err_msg\r\n # Second with 'es' locale\r\n with self.app as c:\r\n err_msg = \"The page should be in Spanish\"\r\n res = c.get('/', headers=[('Accept-Language', 'es')])\r\n assert \"Comunidad\" in res.data, err_msg","def test_i18n02(self):\n output = self.engine.render_to_string('i18n02')\n self.assertEqual(output, 'xxxyyyxxx')","def question_new_translate():","def test_ukrainian_symbols(self):\n string = \"Минає ніч від’їзду\"\n expected = \"Minaye nich vid’yizdu\"\n self.assertEqual(transliterate(string), expected)","def test_single_locale_activation(self):\n with translation.override(\"fr\"):\n self.assertEqual(\n self.get_template(\n \"{% load i18n %}{% blocktranslate %}Yes{% endblocktranslate %}\"\n ).render(Context({})),\n \"Oui\",\n )","def use_zh(self):\n pass","def test_list_unicode_locale(self):\n response = self.client.get(self.url, {'lang': 'ru'})\n assert response.data\n\n assert response.data['count'] == len(discopane_items)\n assert response.data['next'] is None\n assert response.data['previous'] is None\n assert response.data['results']","def showTranslatedWithoutJoin(cls):\n print (\"ALL WORDS WITH TRANSLATIONS STORED IN DATABASE:\")\n for word1 in EnglishHelper.query(\"SELECT english_word FROM EnglishWords\", fetchAll=True):\n try:\n print word1[0],\" - \", (EnglishHelper.query(\"select polish_word from PolishWords where \"\n \" id_pl=(select id_pl from translations where \"\n \"id_eng = (select id_eng from EnglishWords \"\n \"where english_word = '%s'))\"%word1))[0].encode('utf-8')\n except:\n print \"There is no translation, sorry :(\"","def _test (self) :\n print (_T (ckw.title or \"Baz\"))\n print (_T (\"Foo\"))\n foo = _(\"Markup %d\")\n print (_T(foo) % 42)\n print (_Tn (\"Singular\", \"Plural\", 4))","def test_fallback_values_2(self):\n title1_de = \"title de\"\n text1_de = \"text in german\"\n n = TestModelWithFallback2()\n n.title = title1_de\n n.text = text1_de\n n.save()\n del n\n n = TestModelWithFallback2.objects.get(title=title1_de)\n trans_real.activate(\"en\")\n self.failUnlessEqual(n.title, title1_de)\n self.failUnlessEqual(n.text,\\\n TestTranslationOptionsWithFallback2.fallback_values['text'])","def test_fallback_variant(self):\n x = SimpleModel()\n\n x.set_current_language(\"de\")\n x.tr_title = \"Hallo-de\"\n\n x.set_current_language(\"en\")\n x.tr_title = \"Hello-en\"\n\n x.save()\n\n with translation.override(\"de-ch\"):\n x = SimpleModel.objects.get(pk=x.pk)\n self.assertEqual(x.tr_title, \"Hallo-de\")","def test_hero_all_langs(self):\n\n self.driver.get(self.url_ + '/?hl=de')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons ist ein Open Knowledge Repository'))\n\n self.driver.get(self.url_ + '/?hl=es')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons es un repositorio abierto'))\n\n self.driver.get(self.url_ + '/?hl=fr')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons est un dépôt de connaissances'))\n\n self.driver.get(self.url_ + '/?hl=hi')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith(\n 'Data Commons, एक ऐसा प्रोजेक्ट है जिसमें डेटा'))\n\n self.driver.get(self.url_ + '/?hl=ja')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(hero_msg.text.startswith('データコモンズは、マッピングされた共通'))\n\n self.driver.get(self.url_ + '/?hl=ko')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons는 매핑된 공통 항목을 사용해'))\n\n self.driver.get(self.url_ + '/?hl=ru')\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\n self.assertTrue(\n hero_msg.text.startswith('Data Commons – это открытая база данных'))","def get_translation(self):","def test_i18n05(self):\n output = self.engine.render_to_string(\"i18n05\", {\"anton\": \"yyy\"})\n self.assertEqual(output, \"xxxyyyxxx\")","def test_i18n04(self):\n output = self.engine.render_to_string(\"i18n04\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"å\")","def translate(self, language=None):","def test_get_translation_resources(self):\n pass","def test_single_locale_activation(self):\n with translation.override('fr'):\n self.assertEqual(\n Template(\"{% load i18n %}{% blocktrans %}Yes{% endblocktrans %}\").render(Context({})),\n 'Oui'\n )","def test_i18n05(self):\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\n self.assertEqual(output, 'xxxyyyxxx')","def test_i18n05(self):\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\n self.assertEqual(output, 'xxxyyyxxx')","def test_i18n15(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n15', {'absent': ''})\n self.assertEqual(output, 'Passwort')","def test_word_translation(self):\n self.assertEqual(translator.translate_word(\"hour\"), \"ourhay\")\n self.assertEqual(translator.translate_word(\"\"), \"\")\n self.assertEqual(translator.translate_word(\"aaa\"), \"aaayay\")","def init_translations():\n if \"@lang\" in input.load_input():\n lang = input.get_lang()\n try:\n trad = gettext.GNUTranslations(open(\"../course/common_student/$i18n/\" + lang + \".mo\", \"rb\"))\n except FileNotFoundError:\n trad = gettext.NullTranslations()\n trad.install()\n return lang\n trad = gettext.NullTranslations()\n trad.install()\n return \"en\"","def test_legacyi18n04(self):\n output = self.engine.render_to_string(\"legacyi18n04\", {\"anton\": \"Å\"})\n self.assertEqual(output, \"å\")","def translate():\n pass","def test_i18n26(self):\n output = self.engine.render_to_string(\n \"i18n26\", {\"myextra_field\": \"test\", \"number\": 1}\n )\n self.assertEqual(output, \"singular test\")","def test_get_unique_langs(self):\n\n contacts = Contact.query.filter_by(user_id=6).all()\n\n unique_lang_dict = MessageController.get_unique_langs(contacts)\n self.assertEqual(unique_lang_dict, {4: u'fr'})","def test_i18n26(self):\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\n self.assertEqual(output, 'singular test')","def test_i18n26(self):\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\n self.assertEqual(output, 'singular test')","def test_english_with_dennis(self):\n resp = ResponseFactory(\n locale=u'en-US',\n product=u'firefox',\n description=u'hello',\n translated_description=u''\n )\n\n # Set the product up for translation *after* creating the response\n # so that it doesn't get auto-translated because Response is set up\n # for auto-translation.\n prod = Product.objects.get(db_name='firefox')\n prod.translation_system = u'dennis'\n prod.save()\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_translation_smoke():\n english_to_morse = get_translator(\"english\", \"morse\")\n morse_to_english = get_translator(\"morse\", \"english\")\n morse = english_to_morse.translate(\"hello world\")\n english = morse_to_english.translate(morse)\n assert english == \"HELLO WORLD\"","def test_get_languages(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())","def test_get_languages(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())","def test_shipped_locales(self):\n url = reverse('shipping.views.status.shipped_locales')\n url += '?av=fx1.0'\n response = self.client.get(url)\n eq_(response.status_code, 200)\n eq_(response.content, \"\"\"da\nde\nen-US\n\"\"\")","def stats_text(test):\n\n stats_text_en(test) \n \n stats_text_cn(test)","def test_get_languages_multiple(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n translation = StoryTranslation(story=story, title=\"Spanish Title\",\n summary=\"Spanish Summary\", language=\"es\")\n translation.save()\n self.assertEqual([settings.LANGUAGE_CODE, 'es'], story.get_languages())","def test_langs(self):\n self._api.Init(lang=\"eng\")\n lang = self._api.GetInitLanguagesAsString()\n self.assertEqual(lang, \"eng\")\n langs = self._api.GetLoadedLanguages()\n self.assertEqual(langs, [\"eng\"])\n self.assertIn(\"eng\", self._api.GetAvailableLanguages())","def test_home_view_cyrillic(self):\n UserData.objects.create(\n first_name=\"Андрей\",\n last_name=\"Герасько\",\n date_of_birth='1998-02-23',\n bio=\"Био\",\n email=\"aaa@gmail.com\",\n jabber=\"aaaaa\",\n skype=\"aaaaa\",\n other_contacts=\"Контакты\"\n )\n response = self.client.get(self.url)\n self.assertEqual(response.status_code, 200)\n self.assertIn('42 Coffee Cups Test Assignment', response.content)\n self.assertIn('Андрей', response.content)\n self.assertIn('Герасько', response.content)\n self.assertIn('Date of birth', response.content)\n self.assertIn('Био', response.content)\n self.assertIn('Контакты', response.content)","def test_get_languages_multiple(self):\n title = ('Transportation Challenges Limit Education Choices for '\n 'Denver Parents')\n summary = \"\"\"\n Many families in the Denver metro area use public\n transportation instead of a school bus because for them, a\n quality education is worth hours of daily commuting. Colorado's\n school choice program is meant to foster educational equity,\n but the families who benefit most are those who have time and\n money to travel. Low-income families are often left in a lurch.\n \"\"\"\n byline = \"Mile High Connects\"\n story = create_story(title=title, summary=summary, byline=byline)\n translation = StoryTranslation(story=story, title=\"Spanish Title\",\n summary=\"Spanish Summary\", language=\"es\")\n translation.save()\n story_languages = story.get_languages()\n self.assertEqual(len(story_languages), 2)\n for code in (settings.LANGUAGE_CODE, 'es'):\n self.assertIn(code, story_languages)","def test_set_language(self):\n # Test for default languages\n self.assertEqual(self.scraper.language_original, 'jpn')\n self.assertEqual(self.scraper.language_translated, 'eng')\n\n # Test after setting supported languages\n self.scraper.set_languages('jpn', 'eng')\n self.assertEqual(self.scraper.language_translated, 'jpn')\n self.assertEqual(self.scraper.language_original, 'eng')\n\n # Test after setting non-supported languages\n self.scraper.set_languages('eng', 'lol')\n self.assertEqual(self.scraper.language_translated, 'jpn')\n self.assertEqual(self.scraper.language_original, 'eng')","def test_english_no_translation(self):\n resp = ResponseFactory(\n locale=u'en-US',\n description=u'hello',\n translated_description=u''\n )\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_language_conversion(self):\n\n self.assertEqual(self.converter.language, LANGUAGE_CODE,\n 'Check the language conversion.')","def test_english_gb_no_translation(self):\n resp = ResponseFactory(\n locale=u'en-GB',\n description=u'hello',\n translated_description=u''\n )\n\n # No new jobs should be generated\n eq_(len(resp.generate_translation_jobs()), 0)\n\n # Re-fetch from the db and make sure the description was copied over\n resp = Response.objects.get(id=resp.id)\n eq_(resp.description, resp.translated_description)","def test_functionallity(self):\n\n pp = Lexpp(external_dict=pkg_resources.resource_filename(\"lexpp\", \"tests/test.dict\"))\n\n test_word = \"キャプテン\"\n entries = list(pp.lookup(test_word))\n\n self.assertEqual(len(entries), 4)\n\n for e in entries:\n self.assertEqual(type(e), Entry)\n rep = pp.get_representative_form(e)\n self.assertEqual(rep, test_word)","def test_settings(self):\n \n self.assertTrue(settings.USE_I18N, msg=\"setting USE_I18N must be True to have languages working\")","def test_translate(self):\n body = Question()\n response = self.client.open('/api/rtx/v1/translate',\n method='POST',\n data=json.dumps(body),\n content_type='application/json')\n self.assert200(response, \"Response body is : \" + response.data.decode('utf-8'))","def test_i18n07(self):\n output = self.engine.render_to_string(\"i18n07\", {\"number\": 1})\n self.assertEqual(output, \"singular\")","def test_languages(self):\n for i, item in enumerate(self._letters_proto.item):\n for code in item.language:\n # The language code should be in ISO 639 format and consists of\n # two letters for ISO 639-1 languages and three letters otherwise.\n self.assertLess(1, len(code))\n self.assertGreater(4, len(code))\n self.assertTrue(code.islower(), f'Line {i}: Language code should be '\n 'lower-case')\n if len(code) == 3:\n lang = pycountry.languages.get(alpha_3=code)\n self.assertTrue(lang, f'Failed to find language for code {code}')\n if hasattr(lang, 'alpha_2'):\n self.fail(f'Letter {i}: Please use two-letter code `{lang.alpha_2}`'\n f' instead of `{lang.alpha_3}` for {lang.name}')\n else:\n lang = pycountry.languages.get(alpha_2=code)\n self.assertTrue(lang, f'Failed to find language for code {code}')","def test_i18n07(self):\n output = self.engine.render_to_string('i18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_i18n07(self):\n output = self.engine.render_to_string('i18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_legacyi18n07(self):\n output = self.engine.render_to_string(\"legacyi18n07\", {\"number\": 1})\n self.assertEqual(output, \"singular\")","def test_simple_translation_using_get(self):\n pass","def test_get_language(self):\n with translation.override(\"fr\"):\n # Despite being\n # Initialize form in other language.\n x = SimpleModel(shared=\"SHARED\", tr_title=\"TRANS\", _current_language=\"nl\")\n self.assertEqual(x.get_current_language(), \"nl\")\n x.save()\n\n x2 = SimpleModel.objects.language(\"nl\").get(pk=x.pk)\n self.assertEqual(x2.get_current_language(), \"nl\")\n self.assertEqual(x2.shared, \"SHARED\")\n self.assertEqual(x2.tr_title, \"TRANS\")","def test_legacyi18n07(self):\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def test_legacyi18n07(self):\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\n self.assertEqual(output, 'singular')","def make_translated_text():\n return {\n code: ''\n for code, name\n in settings.LANGUAGES\n }","def test_course_explicit_english(self):\r\n self.client = AjaxEnabledTestClient()\r\n self.client.login(username=self.uname, password=self.password)\r\n\r\n resp = self.client.get_html('/course/',\r\n {},\r\n HTTP_ACCEPT_LANGUAGE='en'\r\n )\r\n\r\n self.assertContains(resp,\r\n '<h1 class=\"page-header\">My Courses</h1>',\r\n status_code=200,\r\n html=True)","def test_fallback_values_1(self):\n title1_de = \"title de\"\n n = TestModelWithFallback()\n n.title = title1_de\n n.save()\n del n\n n = TestModelWithFallback.objects.get(title=title1_de)\n self.failUnlessEqual(n.title, title1_de)\n trans_real.activate(\"en\")\n self.failUnlessEqual(n.title, \"\")","def test_issue3625():\n nlp = Hindi()\n doc = nlp(u\"hi. how हुए. होटल, होटल\")\n assert [token.text for token in doc] == ['hi', '.', 'how', 'हुए', '.', 'होटल', ',', 'होटल']","def test_get_supported_locales_for_voice_datasets(self):\n pass","def nintl(self):","def test_01_i18n_authenticated(self):\r\n with self.app as c:\r\n # First default 'en' locale\r\n err_msg = \"The page should be in English\"\r\n res = c.get('/', follow_redirects=True)\r\n assert \"Community\" in res.data, err_msg\r\n self.register()\r\n self.signin()\r\n # After signing in it should be in English\r\n err_msg = \"The page should be in English\"\r\n res = c.get('/', follow_redirects=True)\r\n assert \"Community\" in res.data, err_msg\r\n\r\n # Change it to Spanish\r\n user = db.session.query(User).filter_by(name='johndoe').first()\r\n user.locale = 'es'\r\n db.session.add(user)\r\n db.session.commit()\r\n\r\n res = c.get('/', follow_redirects=True)\r\n err_msg = \"The page should be in Spanish\"\r\n assert \"Comunidad\" in res.data, err_msg\r\n # Sign out should revert it to English\r\n self.signout()\r\n err_msg = \"The page should be in English\"\r\n res = c.get('/', follow_redirects=True)\r\n assert \"Community\" in res.data, err_msg","def test_no_cyrillic_string(self):\n string = \"test\"\n expected = \"test\"\n self.assertEqual(transliterate(string), expected)","def t(eng, chinese):\n return chinese if 'zh' in get_info().user_language else eng"],"string":"[\n \"def test_i18n10(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n10', {'bool': True})\\n self.assertEqual(output, 'Ja')\",\n \"def test_i18n27(self):\\n with translation.override(\\\"ru\\\"):\\n output = self.engine.render_to_string(\\\"i18n27\\\", {\\\"number\\\": 1})\\n self.assertEqual(\\n output, \\\"1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442\\\"\\n )\",\n \"def test_i18n09(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n09')\\n self.assertEqual(output, 'Page not found')\",\n \"def test_i18n27(self):\\n with translation.override('ru'):\\n output = self.engine.render_to_string('i18n27', {'number': 1})\\n self.assertEqual(output, '1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442')\",\n \"def test_i18n27(self):\\n with translation.override('ru'):\\n output = self.engine.render_to_string('i18n27', {'number': 1})\\n self.assertEqual(output, '1 \\\\u0440\\\\u0435\\\\u0437\\\\u0443\\\\u043b\\\\u044c\\\\u0442\\\\u0430\\\\u0442')\",\n \"def test_i18n23(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n23')\\n self.assertEqual(output, 'nicht gefunden')\",\n \"def test_i18n(self):\\n # If you want to modify this code, please get acquainted with\\n # Python's locale module. In particular:\\n # http://docs.python.org/library/locale.html#locale.getdefaultlocale\\n\\n # Set the standard C locale.\\n os.environ['LANG'] = 'C'\\n os.environ['LC_ALL'] = 'C.UTF-8'\\n\\n #must be after above\\n out1 = tr('Hello!')\\n expected1 = 'Hello!'\\n msg = 'Expected %s, got %s' % (expected1, out1)\\n assert out1 == expected1, msg\\n\\n # Set the Indonesian locale to test translations.\\n os.environ['LANG'] = 'id'\\n os.environ['LC_ALL'] = 'id_ID.UTF-8'\\n\\n #must be after above\\n #indoout1 = tr('Hello!') # translate as 'Hi'\\n #indoexpected1 = 'Hi!'\\n #msg = 'Expected %s, got %s' % (indoexpected1, indoout1)\\n #assert indoout1 == indoexpected1, msg\",\n \"def test_i18n11(self):\\n output = self.engine.render_to_string('i18n11', {'bool': True})\\n self.assertEqual(output, 'ja')\",\n \"def test_default_translations(self):\\n\\t\\t\\n\\t\\tself.assertTrue(data.get_default_translation('Catholicism', 3) == 'DRA')\\n\\t\\tself.assertTrue(data.get_default_translation('Christianity', 3) == 'ESV')\",\n \"def test_i18n06(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n06')\\n self.assertEqual(output, 'Seite nicht gefunden')\",\n \"def setUp(self):\\n super().setUp()\\n translation.activate(\\\"en-us\\\")\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string(\\\"i18n28\\\", {\\\"anton\\\": \\\"α\\\", \\\"berta\\\": \\\"β\\\"})\\n self.assertEqual(output, \\\"α + β\\\")\",\n \"def test_i18n16(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n16')\\n self.assertEqual(output, '<')\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\\n self.assertEqual(output, 'α + β')\",\n \"def test_i18n28(self):\\n output = self.engine.render_to_string('i18n28', {'anton': 'α', 'berta': 'β'})\\n self.assertEqual(output, 'α + β')\",\n \"def test_language_sensitivity(self): \\n \\n for lang in self.LANGUAGES:\\n activate(lang)\\n \\n self.assertEqual(get_wording_text('test_1'), lang)\",\n \"def use_en(self):\\n pass\",\n \"def test_i18n13(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n13')\\n self.assertEqual(output, 'Passwort')\",\n \"def test_international(self):\\n\\n test_cases = (\\n makeSimpleTestCase('the beatles'),\\n makeSimpleTestCase('david guetta'),\\n makeSimpleTestCase('susan boyle'),\\n makeSimpleTestCase('rammstein'),\\n makeSimpleTestCase('deadmau5'),\\n makeTestCase('deadmau 5', Matcher(title=Equals('deadmau5'))),\\n makeSimpleTestCase('2face idibia'),\\n makeSimpleTestCase('angelique kidjo'),\\n makeSimpleTestCase('gotye'),\\n makeSimpleTestCase('michel telo'),\\n makeSimpleTestCase('stereo total'),\\n makeSimpleTestCase('yelle'),\\n makeSimpleTestCase('stefanie heinzmann'),\\n makeSimpleTestCase('olly murs')\\n )\\n\\n self._run_tests('i18n_artist', test_cases)\",\n \"def test_i18n12(self):\\n output = self.engine.render_to_string('i18n12')\\n self.assertEqual(output, 'de')\",\n \"def test_translate_unique_langs(self):\\n\\n trans_msgs_dict = MessageController.translate_unique_langs({'2': 'es', '4': 'fr'}, \\n 'hi', 'en', False, False)\\n\\n self.assertEqual(trans_msgs_dict, {'es': u'{hola}', 'fr': u'salut'})\",\n \"def test_i18n01(self):\\n output = self.engine.render_to_string('i18n01')\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string(\\\"legacyi18n08\\\", {\\\"number\\\": 2})\\n self.assertEqual(output, \\\"2 plural\\\")\",\n \"def test_fallback_language_no_current(self):\\n x = SimpleModel()\\n x.set_current_language(self.conf_fallback)\\n x.tr_title = \\\"TITLE_FALLBACK\\\"\\n\\n self.assertEqual(\\n x.safe_translation_getter(\\\"tr_title\\\", language_code=self.other_lang1), \\\"TITLE_FALLBACK\\\"\\n )\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string(\\\"i18n08\\\", {\\\"number\\\": 2})\\n self.assertEqual(output, \\\"2 plural\\\")\",\n \"def test_i18n14(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n14')\\n self.assertEqual(output, 'foo Passwort Passwort')\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string('i18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_i18n08(self):\\n output = self.engine.render_to_string('i18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_switching_to_korean(web_browser):\\n\\n page = MainPage(web_browser)\\n page.welcoming_popup_close_button.click()\\n\\n # Changing language to korean\\n page.regional_settings_button.click()\\n page.current_language_button.click()\\n page.korean_language_button.click()\\n page.finish_login_button.click()\\n page.wait_page_loaded()\\n\\n assert \\\"더 똑똑한 쇼핑, 더 나은 생활!\\\" == page.logo_slogan_title.get_text(), \\\"Switching to korean lang failed!\\\"\",\n \"def test_fallback_language(self):\\n x = SimpleModel()\\n x.set_current_language(self.conf_fallback)\\n x.tr_title = \\\"TITLE_FALLBACK\\\"\\n\\n x.set_current_language(self.other_lang1)\\n x.tr_title = \\\"TITLE_XX\\\"\\n x.save()\\n\\n with translation.override(self.other_lang2):\\n x = SimpleModel.objects.get(pk=x.pk)\\n self.assertEqual(x.tr_title, \\\"TITLE_FALLBACK\\\")\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_legacyi18n08(self):\\n output = self.engine.render_to_string('legacyi18n08', {'number': 2})\\n self.assertEqual(output, '2 plural')\",\n \"def test_language_fix(self):\\n #TODO\\n \\n for lang in self.LANGUAGES:\\n activate(lang)\\n \\n self.assertEqual(lang, get_language())\",\n \"def test_content(google_translator):\\n # from bs4 import BeautifulSoup\\n # assert 'GitHub' in BeautifulSoup(response.content).title.string\\n assert google_translator.translate(text='좋은') == \\\"good\\\"\",\n \"def test_langid_benchmark(basic_multilingual):\\n examples = [\\n {\\\"text\\\": \\\"contingentiam in naturalibus causis.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"I jak opowiadał nieżyjący już pan Czesław\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Sonera gilt seit längerem als Übernahmekandidat\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"与银类似，汞也可以与空气中的硫化氢反应。\\\", \\\"label\\\": \\\"zh-hans\\\"},\\n {\\\"text\\\": \\\"contradictionem implicat.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Bis zu Prozent gingen die Offerten etwa im\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"inneren Sicherheit vorgeschlagene Ausweitung der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Multimedia-PDA mit Mini-Tastatur\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ponášalo sa to na rovnicu o dvoch neznámych.\\\", \\\"label\\\": \\\"sk\\\"},\\n {\\\"text\\\": \\\"이처럼 앞으로 심판의 그 날에 다시 올 메시아가 예수 그리스도이며 , 그는 모든 인류의\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Die Arbeitsgruppe bedauert , dass der weit über\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"И только раз довелось поговорить с ним не вполне\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"de a-l lovi cu piciorul și conștiința că era\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"relación coas pretensións do demandante e que, nos\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"med petdeset in sedemdeset\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Catalunya; el Consell Comarcal del Vallès Oriental\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"kunnen worden.\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"Witkin je ve většině ohledů zcela jiný.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"lernen, so zu agieren, dass sie positive oder auch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"olurmuş...\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"sarcasmo de Altman, desde as «peruas» que discutem\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"خلاف فوجداری مقدمہ درج کرے۔\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"Norddal kommune :\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"dem Windows-.-Zeitalter , soll in diesem Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"przeklętych ucieleśniają mit poety-cygana,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"We do not believe the suspect has ties to this\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"groziņu pīšanu.\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Senior Vice-President David M. Thomas möchte\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"neomylně vybral nějakou knihu a začetl se.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Statt dessen darf beispielsweise der Browser des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"outubro, alcançando R $ bilhões em .\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"(Porte, ), as it does other disciplines\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"uskupení se mylně domnívaly, že podporu\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Übernahme von Next Ende an dem System herum , das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"No podemos decir a la Hacienda que los alemanes\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"и рѣста еи братья\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"الذي اتخذ قرارا بتجميد اعلان الدولة الفلسطينية\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"uurides Rootsi sõjaarhiivist toodud . sajandi\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"selskapets penger til å pusse opp sin enebolig på\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"средней полосе и севернее в Ярославской,\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"il-massa żejda fil-ġemgħat u superġemgħat ta'\\\", \\\"label\\\": \\\"mt\\\"},\\n {\\\"text\\\": \\\"The Global Beauties on internetilehekülg, mida\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"이스라엘 인들은 하나님이 그 큰 팔을 펴 이집트 인들을 치는 것을 보고 하나님을 두려워하며\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Snad ještě dodejme jeden ekonomický argument.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Spalio d. vykusiame pirmajame rinkimų ture\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"und schlechter Journalismus ein gutes Geschäft .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Du sodiečiai sėdi ant potvynio apsemtų namų stogo.\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"цей є автентичним.\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"Și îndegrabă fu cu îngerul mulțime de șireaguri\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"sobra personal cualificado.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Tako se u Njemačkoj dvije trećine liječnika služe\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Dual-Athlon-Chipsatz noch in diesem Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"यहां तक कि चीन के चीफ ऑफ जनरल स्टाफ भी भारत का\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Li forestier du mont avale\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Netzwerken für Privatanwender zu bewundern .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"만해는 승적을 가진 중이 결혼할 수 없다는 불교의 계율을 시대에 맞지 않는 것으로 보았다\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"balance and weight distribution but not really for\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"og så e # tente vi opp den om morgonen å sfyrte\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"변화는 의심의 여지가 없는 것이지만 반면에 진화는 논쟁의 씨앗이다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"puteare fac aceastea.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Waitt seine Führungsmannschaft nicht dem\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"juhtimisega, tulid sealt.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Veränderungen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"banda en el Bayer Leverkusen de la Bundesliga de\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"В туже зиму посла всеволодъ сн҃а своѥго ст҃ослава\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"пославъ приведе я мастеры ѿ грекъ\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"En un nou escenari difícil d'imaginar fa poques\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"καὶ γὰρ τινὲς αὐτοὺς εὐεργεσίαι εἶχον ἐκ Κροίσου\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"직접적인 관련이 있다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"가까운 듯하면서도 멀다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Er bietet ein ähnliches Leistungsniveau und\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"民都洛水牛是獨居的，並不會以群族聚居。\\\", \\\"label\\\": \\\"zh-hant\\\"},\\n {\\\"text\\\": \\\"την τρομοκρατία.\\\", \\\"label\\\": \\\"el\\\"},\\n {\\\"text\\\": \\\"hurbiltzen diren neurrian.\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Ah dimenticavo, ma tutta sta caciara per fare un\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"На первом этапе (-) прошла так называемая\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"of games are on the market.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"находится Мост дружбы, соединяющий узбекский и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"lessié je voldroie que li saint fussent aporté\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Дошла очередь и до Гималаев.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"vzácným suknem táhly pouští, si jednou chtěl do\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"E no terceiro tipo sitúa a familias (%), nos que a\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"وجابت دوريات امريكية وعراقية شوارع المدينة، فيما\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Jeg har bodd her i år .\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"Pohrozil, že odbory zostří postoj, pokud se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"tinham conseguido.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Nicht-Erkrankten einen Anfangsverdacht für einen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"permanece em aberto.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"questi possono promettere rendimenti fino a un\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Tema juurutatud kahevedurisüsteemita oleksid\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Поведение внешне простой игрушки оказалось\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Bundesländern war vom Börsenverein des Deutschen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"acció, 'a mesura que avanci l'estiu, amb l'augment\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Dove trovare queste risorse? Jay Naidoo, ministro\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"essas gordurinhas.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Im zweiten Schritt sollen im übernächsten Jahr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"allveelaeva pole enam vaja, kuna külm sõda on läbi\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"उपद्रवी दुकानों को लूटने के साथ ही उनमें आग लगा\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"@user nella sfortuna sei fortunata ..\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"математических школ в виде грозовых туч.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"No cambiaremos nunca nuestra forma de jugar por un\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"dla tej klasy ani wymogów minimalnych, z wyjątkiem\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"en todo el mundo, mientras que en España consiguió\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"политики считать надежное обеспечение военной\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"gogoratzen du, genio alemana delakoaren\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Бычий глаз.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Opeření se v pravidelných obdobích obnovuje\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"I no és només la seva, es tracta d'una resposta\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"오경을 가르쳤다 .\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Nach der so genannten Start-up-Periode vergibt die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Saulista huomasi jo lapsena , että hänellä on\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Министерство культуры сочло нецелесообразным, и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"znepřátelené tábory v Tádžikistánu předseda\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"καὶ ἦν ὁ λαὸς προσδοκῶν τὸν Ζαχαρίαν καὶ ἐθαύμαζον\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"Вечером, в продукте, этот же человек говорил о\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"lugar á formación de xuizos máis complexos.\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"cheaper, in the end?\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"الوزارة في شأن صفقات بيع الشركات العامة التي تم\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"tärkeintä elämässäni .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Виконання Мінських угод було заблоковано Росією та\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"Aby szybko rozpoznać żołnierzy desantu, należy\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Bankengeschäfte liegen vorn , sagte Strothmann .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"продолжение работы.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Metro AG plant Online-Offensive\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"nu vor veni, și să vor osîndi, aceia nu pot porni\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Ich denke , es geht in Wirklichkeit darum , NT bei\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"de turism care încasează contravaloarea\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"Aurkaria itotzea da helburua, baloia lapurtu eta\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"com a centre de formació en Tecnologies de la\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"oportet igitur quod omne agens in agendo intendat\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Jerzego Andrzejewskiego, oparty na chińskich\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"sau một vài câu chuyện xã giao không dính dáng tới\\\", \\\"label\\\": \\\"vi\\\"},\\n {\\\"text\\\": \\\"что экономическому прорыву жесткий авторитарный\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"DRAM-Preisen scheinen DSPs ein\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Jos dajan nubbái: Mana!\\\", \\\"label\\\": \\\"sme\\\"},\\n {\\\"text\\\": \\\"toți carii ascultară de el să răsipiră.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"odpowiedzialności, które w systemie własności\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Dvomesečno potovanje do Mollenda v Peruju je\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"d'entre les agències internacionals.\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Fahrzeugzugangssysteme gefertigt und an viele\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"in an answer to the sharers' petition in Cuthbert\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Europa-Domain per Verordnung zu regeln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"#Balotelli. Su ebay prezzi stracciati per Silvio\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Ne na košickém trávníku, ale už včera v letadle se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"zaměstnanosti a investičních strategií.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Tatínku, udělej den\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"frecuencia con Mary.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Свеаборге.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"opatření slovenské strany o certifikaci nejvíce\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"En todas me decían: 'Espera que hagamos un estudio\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Die Demonstration sollte nach Darstellung der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ci vorrà un assoluto rigore se dietro i disavanzi\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Tatínku, víš, že Honzovi odešla maminka?\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Die Anzahl der Rechner wuchs um % auf und die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"האמריקאית על אדמת סעודיה עלולה לסבך את ישראל, אין\\\", \\\"label\\\": \\\"he\\\"},\\n {\\\"text\\\": \\\"Volán Egyesülés, a Közlekedési Főfelügyelet is.\\\", \\\"label\\\": \\\"hu\\\"},\\n {\\\"text\\\": \\\"Schejbala, který stejnou hru s velkým úspěchem\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"depends on the data type of the field.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Umsatzwarnung zu Wochenbeginn zeitweise auf ein\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"niin heti nukun .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Mobilfunkunternehmen gegen die Anwendung der so\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"sapessi le intenzioni del governo Monti e dell'UE\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Di chi è figlia Martine Aubry?\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"avec le reste du monde.\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Այդ մաքոքը ինքնին նոր չէ, աշխարհը արդեն մի քանի\\\", \\\"label\\\": \\\"hy\\\"},\\n {\\\"text\\\": \\\"și în cazul destrămării cenaclului.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"befriedigen kann , und ohne die auftretenden\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Κύκνον τ̓ ἐξεναρεῖν καὶ ἀπὸ κλυτὰ τεύχεα δῦσαι.\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"færdiguddannede.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"Schmidt war Sohn eines Rittergutsbesitzers.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"и вдаша попадь ѡпрати\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"cine nu știe învățătură”.\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"détacha et cette dernière tenta de tuer le jeune\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Der har saka også ei lengre forhistorie.\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Pieprz roztłuc w moździerzu, dodać do pasty,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Лежа за гребнем оврага, как за бруствером, Ушаков\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"gesucht habe, vielen Dank nochmals!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"инструментальных сталей, повышения\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"im Halbfinale Patrick Smith und im Finale dann\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"البنوك التريث في منح تسهيلات جديدة لمنتجي حديد\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"una bolsa ventral, la cual se encuentra debajo de\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"za SETimes.\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"de Irak, a un piloto italiano que había violado el\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Er könne sich nicht erklären , wie die Zeitung auf\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Прохорова.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"la democrazia perde sulla tecnocrazia? #\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"entre ambas instituciones, confirmó al medio que\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Austlandet, vart det funne om lag førti\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"уровнями власти.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Dá tedy primáři úplatek, a často ne malý.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"brillantes del acto, al llevar a cabo en el\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"eee druga zadeva je majhen priročen gre kamorkoli\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Das ATX-Board paßt in herkömmliche PC-ATX-Gehäuse\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Za vodné bylo v prvním pololetí zaplaceno v ČR\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Даже на полсантиметра.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"com la del primer tinent d'alcalde en funcions,\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"кількох оповідань в цілості — щось на зразок того\\\", \\\"label\\\": \\\"uk\\\"},\\n {\\\"text\\\": \\\"sed ad divitias congregandas, vel superfluum\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Norma Talmadge, spela mot Valentino i en version\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Dlatego chciał się jej oświadczyć w niezwykłym\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"будут выступать на одинаковых снарядах.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Orang-orang terbunuh di sana.\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"لدى رايت شقيق اسمه أوسكار, وهو يعمل كرسام للكتب\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Wirklichkeit verlagerten und kaum noch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"как перемешивают костяшки перед игрой в домино, и\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"В средине дня, когда солнце светило в нашу\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"d'aventure aux rôles de jeune romantique avec une\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"My teď hledáme organizace, jež by s námi chtěly\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Urteilsfähigkeit einbüßen , wenn ich eigene\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"sua appartenenza anche a voci diverse da quella in\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Aufträge dieses Jahr verdoppeln werden .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"M.E.: Miała szanse mnie odnaleźć, gdyby naprawdę\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"secundum contactum virtutis, cum careat dimensiva\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"ezinbestekoa dela esan zuen.\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"Anek hurbiltzeko eskatzen zion besaulkitik, eta\\\", \\\"label\\\": \\\"eu\\\"},\\n {\\\"text\\\": \\\"perfectius alio videat, quamvis uterque videat\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Die Strecke war anspruchsvoll und führte unter\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"саморазоблачительным уроком, западные СМИ не\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"han representerer radikal islamisme .\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"Què s'hi respira pel que fa a la reforma del\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"previsto para também ser desconstruido.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Ὠκεανοῦ βαθυκόλποις ἄνθεά τ̓ αἰνυμένην, ῥόδα καὶ\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"para jovens de a anos nos Cieps.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"संघर्ष को अंजाम तक पहुंचाने का ऐलान किया है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"objeví i u nás.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"kvitteringer.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"This report is no exception.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Разлепват доносниците до избирателните списъци\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"anderem ihre Bewegungsfreiheit in den USA\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ñu tegoon ca kaw gor ña ay njotti bopp yu kenn\\\", \\\"label\\\": \\\"wo\\\"},\\n {\\\"text\\\": \\\"Struktur kann beispielsweise der Schwerpunkt mehr\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"% la velocidad permitida, la sanción es muy grave.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Teles-Einstieg in ADSL-Markt\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ettekäändeks liiga suure osamaksu.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"als Indiz für die geänderte Marktpolitik des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"quod quidem aperte consequitur ponentes\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"de negociación para el próximo de junio.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Tyto důmyslné dekorace doznaly v poslední době\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"največjega uspeha doslej.\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Paul Allen je jedan od suosnivača Interval\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Federal (Seac / DF) eo Sindicato das Empresas de\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Quartal mit . Mark gegenüber dem gleichen Quartal\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"otros clubes y del Barça B saldrán varios\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Jaskula (Pol.) -\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"umožnily říci, že je možné přejít k mnohem\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"اعلن الجنرال تومي فرانكس قائد القوات الامريكية\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Telekom-Chef Ron Sommer und der Vorstandssprecher\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"My, jako průmyslový a finanční holding, můžeme\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"voorlichting onder andere betrekking kan hebben:\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"Hinrichtung geistig Behinderter applaudiert oder\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"wie beispielsweise Anzahl erzielte Klicks ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Intel-PC-SDRAM-Spezifikation in der Version . (\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"plângere în termen de zile de la comunicarea\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"и Испания ще изгубят втория си комисар в ЕК.\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"इसके चलते इस आदिवासी जनजाति का क्षरण हो रहा है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"aunque se mostró contrario a establecer un\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"des letzten Jahres von auf Millionen Euro .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ankara se također poziva da u cijelosti ratificira\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"herunterlädt .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"стрессовую ситуацию для организма, каковой\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Státního shromáždění (parlamentu).\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"diskutieren , ob und wie dieser Dienst weiterhin\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Verbindungen zu FPÖ-nahen Polizisten gepflegt und\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Pražského volebního lídra ovšem nevybírá Miloš\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nach einem Bericht der Washington Post bleibt das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"للوضع آنذاك، لكني في قرارة نفسي كنت سعيداً لما\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"не желаят запазването на статуквото.\\\", \\\"label\\\": \\\"bg\\\"},\\n {\\\"text\\\": \\\"Offenburg gewesen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ἐὰν ὑμῖν εἴπω οὐ μὴ πιστεύσητε\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"all'odiato compagno di squadra Prost, il quale\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"historischen Gänselieselbrunnens.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"למידע מלווייני הריגול האמריקאיים העוקבים אחר\\\", \\\"label\\\": \\\"he\\\"},\\n {\\\"text\\\": \\\"οὐδὲν ἄρα διαφέρεις Ἀμάσιος τοῦ Ἠλείου, ὃν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"movementos migratorios.\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"Handy und ein Spracherkennungsprogramm sämtliche\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Kümne aasta jooksul on Eestisse ohjeldamatult\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"H.G. Bücknera.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"protiv krijumčarenja, ili pak traženju ukidanja\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"Topware-Anteile mehrere Millionen Mark gefordert\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Maar de mensen die nu over Van Dijk bij FC Twente\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"poidan experimentar as percepcións do interesado,\\\", \\\"label\\\": \\\"gl\\\"},\\n {\\\"text\\\": \\\"Miał przecież w kieszeni nóż.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Avšak žádná z nich nepronikla za hranice přímé\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"esim. helpottamalla luottoja muiden\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Podle předběžných výsledků zvítězila v\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nicht nur das Web-Frontend , auch die\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Regierungsinstitutionen oder Universitäten bei\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Խուլեն Լոպետեգիին, պատճառաբանելով, որ վերջինս\\\", \\\"label\\\": \\\"hy\\\"},\\n {\\\"text\\\": \\\"Афганистана, где в последние дни идут ожесточенные\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"лѧхове же не идоша\\\", \\\"label\\\": \\\"orv\\\"},\\n {\\\"text\\\": \\\"Mit Hilfe von IBMs Chip-Management-Systemen sollen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\", als Manager zu Telefonica zu wechseln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"którym zajmuje się człowiek, zmienia go i pozwala\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"činí kyperských liber, to je asi USD.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Studienplätze getauscht werden .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"учёных, орнитологов признают вид.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"acordare a concediilor prevăzute de legislațiile\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"at større innsats for fornybar, berekraftig energi\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Politiet veit ikkje kor mange personar som deltok\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"offentligheten av unge , sinte menn som har\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"însuși în jurul lapunei, care încet DISPARE în\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"O motivo da decisão é evitar uma sobrecarga ainda\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"El Apostolado de la prensa contribuye en modo\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Teltow ( Kreis Teltow-Fläming ) ist Schmitt einer\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"grozījumus un iesniegt tos Apvienoto Nāciju\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Gestalt einer deutschen Nationalmannschaft als\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"D überholt zu haben , konterte am heutigen Montag\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Softwarehersteller Oracle hat im dritten Quartal\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Během nich se ekonomické podmínky mohou radikálně\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Dziki kot w górach zeskakuje z kamienia.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Ačkoliv ligový nováček prohrál, opět potvrdil, že\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"des Tages , Portraits internationaler Stars sowie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Communicator bekannt wurde .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"τῷ δ’ ἄρα καὶ αὐτῷ ἡ γυνή ἐπίτεξ ἐοῦσα πᾶσαν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"Triadú tenia, mentre redactava 'Dies de memòria',\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"دستهجمعی در درخشندگی ماه سیمگون زمزمه ستاینده و\\\", \\\"label\\\": \\\"fa\\\"},\\n {\\\"text\\\": \\\"Книгу, наполненную мелочной заботой об одежде,\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"putares canem leporem persequi.\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"В дальнейшем эта яркость слегка померкла, но в\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"offizielles Verfahren gegen die Telekom\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"podrían haber sido habitantes de la Península\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Grundlage für dieses Verfahren sind spezielle\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Rechtsausschuß vorgelegten Entwurf der Richtlinie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Im so genannten Portalgeschäft sei das Unternehmen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"ⲏ ⲉⲓϣⲁⲛϥⲓ ⲛⲉⲓⲇⲱⲗⲟⲛ ⲉⲧϩⲙⲡⲉⲕⲏⲓ ⲙⲏ ⲉⲓⲛⲁϣϩⲱⲡ ⲟⲛ ⲙⲡⲣⲏ\\\", \\\"label\\\": \\\"cop\\\"},\\n {\\\"text\\\": \\\"juego podían matar a cualquier herbívoro, pero\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Nach Angaben von Axent nutzen Unternehmen aus der\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"hrdiny Havlovy Zahradní slavnosti (premiéra ) se\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Een zin van heb ik jou daar\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"hat sein Hirn an der CeBIT-Kasse vergessen .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"καὶ τοὺς ἐκπλαγέντας οὐκ ἔχειν ἔτι ἐλεγχομένους\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"nachgewiesenen langfristigen Kosten , sowie den im\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"jučer nakon četiri dana putovanja u Helsinki.\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"pašto paslaugos teikėjas gali susitarti su\\\", \\\"label\\\": \\\"lt\\\"},\\n {\\\"text\\\": \\\"В результате, эти золотые кадры переходят из одной\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"द फाइव-ईयर एंगेजमेंट में अभिनय किया जिसमें जैसन\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"výpis o počtu akcií.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Enfin, elles arrivent à un pavillon chinois\\\", \\\"label\\\": \\\"fr\\\"},\\n {\\\"text\\\": \\\"Tentu saja, tren yang berhubungandengan\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"Arbeidarpartiet og SV har sikra seg fleirtal mot\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"eles: 'Tudo isso está errado' , disse um\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"The islands are in their own time zone, minutes\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Auswahl debütierte er am .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Bu komisyonlar, arazilerini satın almak için\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"Geschütze gegen Redmond aufgefahren .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Time scything the hours, but at the top, over the\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Di musim semi , berharap mengadaptasi Tintin untuk\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"крупнейшей геополитической катастрофой XX века.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Rajojen avaaminen ei suju ongelmitta .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"непроницаемым, как для СССР.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Ma non mancano le polemiche.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Internet als Ort politischer Diskussion und auch\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"incomplets.\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"Su padre luchó al lado de Luis Moya, primer Jefe\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"informazione.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Primacom bietet für Telekom-Kabelnetz\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Oświadczenie prezydencji w imieniu Unii\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"foran rattet i familiens gamle Baleno hvis døra på\\\", \\\"label\\\": \\\"no\\\"},\\n {\\\"text\\\": \\\"[speaker:laughter]\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Dog med langt mindre utstyr med seg.\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"dass es nicht schon mit der anfänglichen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"इस पर दोनों पक्षों में नोकझोंक शुरू हो गई ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"کے ترجمان منیش تیواری اور دگ وجئے سنگھ نے بھی یہ\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"dell'Assemblea Costituente che posseggono i\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"и аште вьси съблазнѧтъ сѧ нъ не азъ\\\", \\\"label\\\": \\\"cu\\\"},\\n {\\\"text\\\": \\\"In Irvine hat auch das Logistikunternehmen Atlas\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"законодательных норм, принимаемых существующей\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Κροίσῳ προτείνων τὰς χεῖρας ἐπικατασφάξαι μιν\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"МИНУСЫ: ИНФЛЯЦИЯ И КРИЗИС В ЖИВОТНОВОДСТВЕ.\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"unterschiedlicher Meinung .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Jospa joku ystävällinen sielu auttaisi kassieni\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Añadió que, en el futuro se harán otros\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Sessiz tonlama hem Fince, hem de Kuzey Sami\\\", \\\"label\\\": \\\"tr\\\"},\\n {\\\"text\\\": \\\"nicht ihnen gehört und sie nicht alles , was sie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Etelästä Kuivajärveen laskee Tammelan Liesjärvestä\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"ICANNs Vorsitzender Vint Cerf warb mit dem Hinweis\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Norsk politikk frå til kan dermed, i\\\", \\\"label\\\": \\\"nn\\\"},\\n {\\\"text\\\": \\\"Głosowało posłów.\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"Danny Jones -- smithjones@ev.net\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"sebeuvědomění moderní civilizace sehrála lučavka\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"относительно спокойный сон: тому гарантия\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"A halte voiz prist li pedra a crïer\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"آنها امیدوارند این واکسن بهزودی در دسترس بیماران\\\", \\\"label\\\": \\\"fa\\\"},\\n {\\\"text\\\": \\\"vlastní důstojnou vousatou tváří.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ora aprire la strada a nuove cause e alimentare il\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"Die Zahl der Vielleser nahm von auf Prozent zu ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Finanzvorstand von Hotline-Dienstleister InfoGenie\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"entwickeln .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"incolumità pubblica.\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"lehtija televisiomainonta\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"joistakin kohdista eri mieltä.\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Hlavně anglická nezávislá scéna, Dead Can Dance,\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"pásmech od do bodů bodové stupnice.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Zu Beginn des Ersten Weltkrieges zählte das\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Així van sorgir, damunt els antics cementiris,\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"In manchem Gedicht der spätern Alten, wie zum\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"gaweihaida jah insandida in þana fairƕu jus qiþiþ\\\", \\\"label\\\": \\\"got\\\"},\\n {\\\"text\\\": \\\"Beides sollte gelöscht werden!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"modifiqués la seva petició inicial de anys de\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"В день открытия симпозиума состоялась закладка\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"tõestatud.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"ἵππῳ πίπτει αὐτοῦ ταύτῃ\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"bisher nie enttäuscht!\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"De bohte ollu tuollárat ja suttolaččat ja\\\", \\\"label\\\": \\\"sme\\\"},\\n {\\\"text\\\": \\\"Klarsignal från röstlängdsläsaren, tre tryck i\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Tvůrcem nového termínu je Joseph Fisher.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Nie miałem czasu na reakcję twierdzi Norbert,\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"potentia Schöpfer.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Un poquito caro, pero vale mucho la pena;\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"οὔ τε γὰρ ἴφθιμοι Λύκιοι Δαναῶν ἐδύναντο τεῖχος\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"vajec, sladového výtažku a některých vitamínových\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Настоящие герои, те, чьи истории потом\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"praesumptio:\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Olin justkui nende vastutusel.\\\", \\\"label\\\": \\\"et\\\"},\\n {\\\"text\\\": \\\"Jokainen keinahdus tuo lähemmäksi hetkeä jolloin\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"ekonomicky výhodných způsobů odvodnění těžkých,\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Poprvé ve své historii dokázala v kvalifikaci pro\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"zpracovatelského a spotřebního průmyslu bude nutné\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Windows CE zu integrieren .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Armangué, a través d'un decret, ordenés l'aturada\\\", \\\"label\\\": \\\"ca\\\"},\\n {\\\"text\\\": \\\"to, co nás Evropany spojuje, než to, co nás od\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ergänzt durch einen gesetzlich verankertes\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Насчитал, что с начала года всего три дня были\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Borisovu tražeći od njega da prihvati njenu\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"la presenza di ben veleni diversi: . chili di\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"καὶ τῶν ἐκλεκτῶν ἀγγέλων ἵνα ταῦτα φυλάξῃς χωρὶς\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"pretraživale obližnju bolnicu i stambene zgrade u\\\", \\\"label\\\": \\\"hr\\\"},\\n {\\\"text\\\": \\\"An rund Katzen habe Wolf seine Spiele getestet ,\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"investigating since March.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Tonböden (Mullböden).\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Stálý dopisovatel LN v SRN Bedřich Utitz\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"červnu předložené smlouvy.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"πνεύματι ᾧ ἐλάλει\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\".%의 신장세를 보였다.\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Foae verde, foi de nuc, Prin pădure, prin colnic,\\\", \\\"label\\\": \\\"ro\\\"},\\n {\\\"text\\\": \\\"διαπέμψας ἄλλους ἄλλῃ τοὺς μὲν ἐς Δελφοὺς ἰέναι\\\", \\\"label\\\": \\\"grc\\\"},\\n {\\\"text\\\": \\\"المسلمين أو أي تيار سياسي طالما عمل ذلك التيار في\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"As informações são da Dow Jones.\\\", \\\"label\\\": \\\"pt\\\"},\\n {\\\"text\\\": \\\"Milliarde DM ausgestattet sein .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"De utgår fortfarande från att kvinnans jämlikhet\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"Sneeuw maakte in Davos bij de voorbereiding een\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"De ahí que en este mercado puedan negociarse\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"intenzívnějšímu sbírání a studiu.\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"और औसकर ४.० पैकेज का प्रयोग किया गया है ।\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Adipati Kuningan karena Kuningan menjadi bagian\\\", \\\"label\\\": \\\"id\\\"},\\n {\\\"text\\\": \\\"Svako je bar jednom poželeo da mašine prosto umeju\\\", \\\"label\\\": \\\"sr\\\"},\\n {\\\"text\\\": \\\"Im vergangenen Jahr haben die Regierungen einen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"durat motus, aliquid fit et non est;\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Dominować będą piosenki do tekstów Edwarda\\\", \\\"label\\\": \\\"pl\\\"},\\n {\\\"text\\\": \\\"beantwortet .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"О гуманитариях было кому рассказывать, а вот за\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"Helsingin kaupunki riitautti vuokrasopimuksen\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"chợt tan biến.\\\", \\\"label\\\": \\\"vi\\\"},\\n {\\\"text\\\": \\\"avtomobil ločuje od drugih.\\\", \\\"label\\\": \\\"sl\\\"},\\n {\\\"text\\\": \\\"Congress has proven itself ineffective as a body.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"मैक्सिको ने इस तरह का शो इस समय आयोजित करने का\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"No minimum order amount.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Convertassa .\\\", \\\"label\\\": \\\"fi\\\"},\\n {\\\"text\\\": \\\"Как это можно сделать?\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"tha mi creidsinn gu robh iad ceart cho saor shuas\\\", \\\"label\\\": \\\"gd\\\"},\\n {\\\"text\\\": \\\"실제 일제는 이런 만해의 논리를 묵살하고 한반도를 침략한 다음 , 이어 만주를 침략하고\\\", \\\"label\\\": \\\"ko\\\"},\\n {\\\"text\\\": \\\"Da un semplice richiamo all'ordine fino a grandi\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"pozoruhodný nejen po umělecké stránce, jež\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"La comida y el servicio aprueban.\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"again, connected not with each other but to the\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"Protokol výslovně stanoví, že nikdo nemůže být\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"ఒక విషయం అడగాలని ఉంది .\\\", \\\"label\\\": \\\"te\\\"},\\n {\\\"text\\\": \\\"Безгранично почитая дирекцию, ловя на лету каждое\\\", \\\"label\\\": \\\"ru\\\"},\\n {\\\"text\\\": \\\"rovnoběžných růstových vrstev, zůstávají krychlové\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"प्रवेश और पूर्व प्रधानमंत्री लाल बहादुर शास्त्री\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"Bronzen medaille in de Europese marathon.\\\", \\\"label\\\": \\\"nl\\\"},\\n {\\\"text\\\": \\\"- gadu vecumā viņi to nesaprot.\\\", \\\"label\\\": \\\"lv\\\"},\\n {\\\"text\\\": \\\"Realizó sus estudios primarios en la Escuela Julia\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"cuartos de final, su clasificación para la final a\\\", \\\"label\\\": \\\"es\\\"},\\n {\\\"text\\\": \\\"Sem si pro něho přiletí americký raketoplán, na\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"Way to go!\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"gehört der neuen SPD-Führung unter Parteichef\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Somit simuliert der Player mit einer GByte-Platte\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Berufung auf kommissionsnahe Kreise , die bereits\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Dist Clarïen\\\", \\\"label\\\": \\\"fro\\\"},\\n {\\\"text\\\": \\\"Schon nach den Gerüchten , die Telekom wolle den\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Software von NetObjects ist nach Angaben des\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"si enim per legem iustitia ergo Christus gratis\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"ducerent in ipsam magis quam in corpus christi,\\\", \\\"label\\\": \\\"la\\\"},\\n {\\\"text\\\": \\\"Neustar-Melbourne-IT-Partnerschaft NeuLevel .\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"forderte dagegen seine drastische Verschärfung.\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"pemmican på hundrede forskellige måder.\\\", \\\"label\\\": \\\"da\\\"},\\n {\\\"text\\\": \\\"Lehån, själv matematiklärare, visar hur den nya\\\", \\\"label\\\": \\\"sv\\\"},\\n {\\\"text\\\": \\\"I highly recommend his shop.\\\", \\\"label\\\": \\\"en\\\"},\\n {\\\"text\\\": \\\"verità, giovani fedeli prostratevi #amen\\\", \\\"label\\\": \\\"it\\\"},\\n {\\\"text\\\": \\\"उत्तर प्रदेश के अध्यक्ष पद से हटाए गए विनय कटियार\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"() روزی مےں کشادگی ہوتی ہے۔\\\", \\\"label\\\": \\\"ur\\\"},\\n {\\\"text\\\": \\\"Prozessorgeschäft profitieren kann , stellen\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"školy začalo počítat pytle s moukou a zjistilo, že\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"प्रभावशाली पर गैर सरकारी लोगों के घरों में भी\\\", \\\"label\\\": \\\"hi\\\"},\\n {\\\"text\\\": \\\"geschichtslos , oder eine Farce , wie sich\\\", \\\"label\\\": \\\"de\\\"},\\n {\\\"text\\\": \\\"Ústrednými mocnosťami v marci však spôsobilo, že\\\", \\\"label\\\": \\\"sk\\\"},\\n {\\\"text\\\": \\\"التسليح بدون مبرر، واستمرار الأضرار الناجمة عن فرض\\\", \\\"label\\\": \\\"ar\\\"},\\n {\\\"text\\\": \\\"Například Pedagogická fakulta Univerzity Karlovy\\\", \\\"label\\\": \\\"cs\\\"},\\n {\\\"text\\\": \\\"nostris ut eriperet nos de praesenti saeculo\\\", \\\"label\\\": \\\"la\\\"}]\\n \\n docs = [Document([], text=example[\\\"text\\\"]) for example in examples]\\n gold_labels = [example[\\\"label\\\"] for example in examples]\\n basic_multilingual(docs)\\n accuracy = sum([(doc.lang == label) for doc,label in zip(docs,gold_labels)])/len(docs)\\n assert accuracy >= 0.98\",\n \"def test_i18n03(self):\\n output = self.engine.render_to_string(\\\"i18n03\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"Å\\\")\",\n \"def test_00_i18n_anonymous(self):\\r\\n # First default 'en' locale\\r\\n with self.app as c:\\r\\n err_msg = \\\"The page should be in English\\\"\\r\\n res = c.get('/', headers=[('Accept-Language', 'en')])\\r\\n assert \\\"Community\\\" in res.data, err_msg\\r\\n # Second with 'es' locale\\r\\n with self.app as c:\\r\\n err_msg = \\\"The page should be in Spanish\\\"\\r\\n res = c.get('/', headers=[('Accept-Language', 'es')])\\r\\n assert \\\"Comunidad\\\" in res.data, err_msg\",\n \"def test_i18n02(self):\\n output = self.engine.render_to_string('i18n02')\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def question_new_translate():\",\n \"def test_ukrainian_symbols(self):\\n string = \\\"Минає ніч від’їзду\\\"\\n expected = \\\"Minaye nich vid’yizdu\\\"\\n self.assertEqual(transliterate(string), expected)\",\n \"def test_single_locale_activation(self):\\n with translation.override(\\\"fr\\\"):\\n self.assertEqual(\\n self.get_template(\\n \\\"{% load i18n %}{% blocktranslate %}Yes{% endblocktranslate %}\\\"\\n ).render(Context({})),\\n \\\"Oui\\\",\\n )\",\n \"def use_zh(self):\\n pass\",\n \"def test_list_unicode_locale(self):\\n response = self.client.get(self.url, {'lang': 'ru'})\\n assert response.data\\n\\n assert response.data['count'] == len(discopane_items)\\n assert response.data['next'] is None\\n assert response.data['previous'] is None\\n assert response.data['results']\",\n \"def showTranslatedWithoutJoin(cls):\\n print (\\\"ALL WORDS WITH TRANSLATIONS STORED IN DATABASE:\\\")\\n for word1 in EnglishHelper.query(\\\"SELECT english_word FROM EnglishWords\\\", fetchAll=True):\\n try:\\n print word1[0],\\\" - \\\", (EnglishHelper.query(\\\"select polish_word from PolishWords where \\\"\\n \\\" id_pl=(select id_pl from translations where \\\"\\n \\\"id_eng = (select id_eng from EnglishWords \\\"\\n \\\"where english_word = '%s'))\\\"%word1))[0].encode('utf-8')\\n except:\\n print \\\"There is no translation, sorry :(\\\"\",\n \"def _test (self) :\\n print (_T (ckw.title or \\\"Baz\\\"))\\n print (_T (\\\"Foo\\\"))\\n foo = _(\\\"Markup %d\\\")\\n print (_T(foo) % 42)\\n print (_Tn (\\\"Singular\\\", \\\"Plural\\\", 4))\",\n \"def test_fallback_values_2(self):\\n title1_de = \\\"title de\\\"\\n text1_de = \\\"text in german\\\"\\n n = TestModelWithFallback2()\\n n.title = title1_de\\n n.text = text1_de\\n n.save()\\n del n\\n n = TestModelWithFallback2.objects.get(title=title1_de)\\n trans_real.activate(\\\"en\\\")\\n self.failUnlessEqual(n.title, title1_de)\\n self.failUnlessEqual(n.text,\\\\\\n TestTranslationOptionsWithFallback2.fallback_values['text'])\",\n \"def test_fallback_variant(self):\\n x = SimpleModel()\\n\\n x.set_current_language(\\\"de\\\")\\n x.tr_title = \\\"Hallo-de\\\"\\n\\n x.set_current_language(\\\"en\\\")\\n x.tr_title = \\\"Hello-en\\\"\\n\\n x.save()\\n\\n with translation.override(\\\"de-ch\\\"):\\n x = SimpleModel.objects.get(pk=x.pk)\\n self.assertEqual(x.tr_title, \\\"Hallo-de\\\")\",\n \"def test_hero_all_langs(self):\\n\\n self.driver.get(self.url_ + '/?hl=de')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons ist ein Open Knowledge Repository'))\\n\\n self.driver.get(self.url_ + '/?hl=es')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons es un repositorio abierto'))\\n\\n self.driver.get(self.url_ + '/?hl=fr')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons est un dépôt de connaissances'))\\n\\n self.driver.get(self.url_ + '/?hl=hi')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith(\\n 'Data Commons, एक ऐसा प्रोजेक्ट है जिसमें डेटा'))\\n\\n self.driver.get(self.url_ + '/?hl=ja')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(hero_msg.text.startswith('データコモンズは、マッピングされた共通'))\\n\\n self.driver.get(self.url_ + '/?hl=ko')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons는 매핑된 공통 항목을 사용해'))\\n\\n self.driver.get(self.url_ + '/?hl=ru')\\n hero_msg = self.driver.find_elements_by_class_name('lead')[0]\\n self.assertTrue(\\n hero_msg.text.startswith('Data Commons – это открытая база данных'))\",\n \"def get_translation(self):\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string(\\\"i18n05\\\", {\\\"anton\\\": \\\"yyy\\\"})\\n self.assertEqual(output, \\\"xxxyyyxxx\\\")\",\n \"def test_i18n04(self):\\n output = self.engine.render_to_string(\\\"i18n04\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"å\\\")\",\n \"def translate(self, language=None):\",\n \"def test_get_translation_resources(self):\\n pass\",\n \"def test_single_locale_activation(self):\\n with translation.override('fr'):\\n self.assertEqual(\\n Template(\\\"{% load i18n %}{% blocktrans %}Yes{% endblocktrans %}\\\").render(Context({})),\\n 'Oui'\\n )\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_i18n05(self):\\n output = self.engine.render_to_string('i18n05', {'anton': 'yyy'})\\n self.assertEqual(output, 'xxxyyyxxx')\",\n \"def test_i18n15(self):\\n with translation.override('de'):\\n output = self.engine.render_to_string('i18n15', {'absent': ''})\\n self.assertEqual(output, 'Passwort')\",\n \"def test_word_translation(self):\\n self.assertEqual(translator.translate_word(\\\"hour\\\"), \\\"ourhay\\\")\\n self.assertEqual(translator.translate_word(\\\"\\\"), \\\"\\\")\\n self.assertEqual(translator.translate_word(\\\"aaa\\\"), \\\"aaayay\\\")\",\n \"def init_translations():\\n if \\\"@lang\\\" in input.load_input():\\n lang = input.get_lang()\\n try:\\n trad = gettext.GNUTranslations(open(\\\"../course/common_student/$i18n/\\\" + lang + \\\".mo\\\", \\\"rb\\\"))\\n except FileNotFoundError:\\n trad = gettext.NullTranslations()\\n trad.install()\\n return lang\\n trad = gettext.NullTranslations()\\n trad.install()\\n return \\\"en\\\"\",\n \"def test_legacyi18n04(self):\\n output = self.engine.render_to_string(\\\"legacyi18n04\\\", {\\\"anton\\\": \\\"Å\\\"})\\n self.assertEqual(output, \\\"å\\\")\",\n \"def translate():\\n pass\",\n \"def test_i18n26(self):\\n output = self.engine.render_to_string(\\n \\\"i18n26\\\", {\\\"myextra_field\\\": \\\"test\\\", \\\"number\\\": 1}\\n )\\n self.assertEqual(output, \\\"singular test\\\")\",\n \"def test_get_unique_langs(self):\\n\\n contacts = Contact.query.filter_by(user_id=6).all()\\n\\n unique_lang_dict = MessageController.get_unique_langs(contacts)\\n self.assertEqual(unique_lang_dict, {4: u'fr'})\",\n \"def test_i18n26(self):\\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\\n self.assertEqual(output, 'singular test')\",\n \"def test_i18n26(self):\\n output = self.engine.render_to_string('i18n26', {'myextra_field': 'test', 'number': 1})\\n self.assertEqual(output, 'singular test')\",\n \"def test_english_with_dennis(self):\\n resp = ResponseFactory(\\n locale=u'en-US',\\n product=u'firefox',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # Set the product up for translation *after* creating the response\\n # so that it doesn't get auto-translated because Response is set up\\n # for auto-translation.\\n prod = Product.objects.get(db_name='firefox')\\n prod.translation_system = u'dennis'\\n prod.save()\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_translation_smoke():\\n english_to_morse = get_translator(\\\"english\\\", \\\"morse\\\")\\n morse_to_english = get_translator(\\\"morse\\\", \\\"english\\\")\\n morse = english_to_morse.translate(\\\"hello world\\\")\\n english = morse_to_english.translate(morse)\\n assert english == \\\"HELLO WORLD\\\"\",\n \"def test_get_languages(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())\",\n \"def test_get_languages(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n self.assertEqual([settings.LANGUAGE_CODE], story.get_languages())\",\n \"def test_shipped_locales(self):\\n url = reverse('shipping.views.status.shipped_locales')\\n url += '?av=fx1.0'\\n response = self.client.get(url)\\n eq_(response.status_code, 200)\\n eq_(response.content, \\\"\\\"\\\"da\\nde\\nen-US\\n\\\"\\\"\\\")\",\n \"def stats_text(test):\\n\\n stats_text_en(test) \\n \\n stats_text_cn(test)\",\n \"def test_get_languages_multiple(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n translation = StoryTranslation(story=story, title=\\\"Spanish Title\\\",\\n summary=\\\"Spanish Summary\\\", language=\\\"es\\\")\\n translation.save()\\n self.assertEqual([settings.LANGUAGE_CODE, 'es'], story.get_languages())\",\n \"def test_langs(self):\\n self._api.Init(lang=\\\"eng\\\")\\n lang = self._api.GetInitLanguagesAsString()\\n self.assertEqual(lang, \\\"eng\\\")\\n langs = self._api.GetLoadedLanguages()\\n self.assertEqual(langs, [\\\"eng\\\"])\\n self.assertIn(\\\"eng\\\", self._api.GetAvailableLanguages())\",\n \"def test_home_view_cyrillic(self):\\n UserData.objects.create(\\n first_name=\\\"Андрей\\\",\\n last_name=\\\"Герасько\\\",\\n date_of_birth='1998-02-23',\\n bio=\\\"Био\\\",\\n email=\\\"aaa@gmail.com\\\",\\n jabber=\\\"aaaaa\\\",\\n skype=\\\"aaaaa\\\",\\n other_contacts=\\\"Контакты\\\"\\n )\\n response = self.client.get(self.url)\\n self.assertEqual(response.status_code, 200)\\n self.assertIn('42 Coffee Cups Test Assignment', response.content)\\n self.assertIn('Андрей', response.content)\\n self.assertIn('Герасько', response.content)\\n self.assertIn('Date of birth', response.content)\\n self.assertIn('Био', response.content)\\n self.assertIn('Контакты', response.content)\",\n \"def test_get_languages_multiple(self):\\n title = ('Transportation Challenges Limit Education Choices for '\\n 'Denver Parents')\\n summary = \\\"\\\"\\\"\\n Many families in the Denver metro area use public\\n transportation instead of a school bus because for them, a\\n quality education is worth hours of daily commuting. Colorado's\\n school choice program is meant to foster educational equity,\\n but the families who benefit most are those who have time and\\n money to travel. Low-income families are often left in a lurch.\\n \\\"\\\"\\\"\\n byline = \\\"Mile High Connects\\\"\\n story = create_story(title=title, summary=summary, byline=byline)\\n translation = StoryTranslation(story=story, title=\\\"Spanish Title\\\",\\n summary=\\\"Spanish Summary\\\", language=\\\"es\\\")\\n translation.save()\\n story_languages = story.get_languages()\\n self.assertEqual(len(story_languages), 2)\\n for code in (settings.LANGUAGE_CODE, 'es'):\\n self.assertIn(code, story_languages)\",\n \"def test_set_language(self):\\n # Test for default languages\\n self.assertEqual(self.scraper.language_original, 'jpn')\\n self.assertEqual(self.scraper.language_translated, 'eng')\\n\\n # Test after setting supported languages\\n self.scraper.set_languages('jpn', 'eng')\\n self.assertEqual(self.scraper.language_translated, 'jpn')\\n self.assertEqual(self.scraper.language_original, 'eng')\\n\\n # Test after setting non-supported languages\\n self.scraper.set_languages('eng', 'lol')\\n self.assertEqual(self.scraper.language_translated, 'jpn')\\n self.assertEqual(self.scraper.language_original, 'eng')\",\n \"def test_english_no_translation(self):\\n resp = ResponseFactory(\\n locale=u'en-US',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_language_conversion(self):\\n\\n self.assertEqual(self.converter.language, LANGUAGE_CODE,\\n 'Check the language conversion.')\",\n \"def test_english_gb_no_translation(self):\\n resp = ResponseFactory(\\n locale=u'en-GB',\\n description=u'hello',\\n translated_description=u''\\n )\\n\\n # No new jobs should be generated\\n eq_(len(resp.generate_translation_jobs()), 0)\\n\\n # Re-fetch from the db and make sure the description was copied over\\n resp = Response.objects.get(id=resp.id)\\n eq_(resp.description, resp.translated_description)\",\n \"def test_functionallity(self):\\n\\n pp = Lexpp(external_dict=pkg_resources.resource_filename(\\\"lexpp\\\", \\\"tests/test.dict\\\"))\\n\\n test_word = \\\"キャプテン\\\"\\n entries = list(pp.lookup(test_word))\\n\\n self.assertEqual(len(entries), 4)\\n\\n for e in entries:\\n self.assertEqual(type(e), Entry)\\n rep = pp.get_representative_form(e)\\n self.assertEqual(rep, test_word)\",\n \"def test_settings(self):\\n \\n self.assertTrue(settings.USE_I18N, msg=\\\"setting USE_I18N must be True to have languages working\\\")\",\n \"def test_translate(self):\\n body = Question()\\n response = self.client.open('/api/rtx/v1/translate',\\n method='POST',\\n data=json.dumps(body),\\n content_type='application/json')\\n self.assert200(response, \\\"Response body is : \\\" + response.data.decode('utf-8'))\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string(\\\"i18n07\\\", {\\\"number\\\": 1})\\n self.assertEqual(output, \\\"singular\\\")\",\n \"def test_languages(self):\\n for i, item in enumerate(self._letters_proto.item):\\n for code in item.language:\\n # The language code should be in ISO 639 format and consists of\\n # two letters for ISO 639-1 languages and three letters otherwise.\\n self.assertLess(1, len(code))\\n self.assertGreater(4, len(code))\\n self.assertTrue(code.islower(), f'Line {i}: Language code should be '\\n 'lower-case')\\n if len(code) == 3:\\n lang = pycountry.languages.get(alpha_3=code)\\n self.assertTrue(lang, f'Failed to find language for code {code}')\\n if hasattr(lang, 'alpha_2'):\\n self.fail(f'Letter {i}: Please use two-letter code `{lang.alpha_2}`'\\n f' instead of `{lang.alpha_3}` for {lang.name}')\\n else:\\n lang = pycountry.languages.get(alpha_2=code)\\n self.assertTrue(lang, f'Failed to find language for code {code}')\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string('i18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_i18n07(self):\\n output = self.engine.render_to_string('i18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string(\\\"legacyi18n07\\\", {\\\"number\\\": 1})\\n self.assertEqual(output, \\\"singular\\\")\",\n \"def test_simple_translation_using_get(self):\\n pass\",\n \"def test_get_language(self):\\n with translation.override(\\\"fr\\\"):\\n # Despite being\\n # Initialize form in other language.\\n x = SimpleModel(shared=\\\"SHARED\\\", tr_title=\\\"TRANS\\\", _current_language=\\\"nl\\\")\\n self.assertEqual(x.get_current_language(), \\\"nl\\\")\\n x.save()\\n\\n x2 = SimpleModel.objects.language(\\\"nl\\\").get(pk=x.pk)\\n self.assertEqual(x2.get_current_language(), \\\"nl\\\")\\n self.assertEqual(x2.shared, \\\"SHARED\\\")\\n self.assertEqual(x2.tr_title, \\\"TRANS\\\")\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def test_legacyi18n07(self):\\n output = self.engine.render_to_string('legacyi18n07', {'number': 1})\\n self.assertEqual(output, 'singular')\",\n \"def make_translated_text():\\n return {\\n code: ''\\n for code, name\\n in settings.LANGUAGES\\n }\",\n \"def test_course_explicit_english(self):\\r\\n self.client = AjaxEnabledTestClient()\\r\\n self.client.login(username=self.uname, password=self.password)\\r\\n\\r\\n resp = self.client.get_html('/course/',\\r\\n {},\\r\\n HTTP_ACCEPT_LANGUAGE='en'\\r\\n )\\r\\n\\r\\n self.assertContains(resp,\\r\\n '<h1 class=\\\"page-header\\\">My Courses</h1>',\\r\\n status_code=200,\\r\\n html=True)\",\n \"def test_fallback_values_1(self):\\n title1_de = \\\"title de\\\"\\n n = TestModelWithFallback()\\n n.title = title1_de\\n n.save()\\n del n\\n n = TestModelWithFallback.objects.get(title=title1_de)\\n self.failUnlessEqual(n.title, title1_de)\\n trans_real.activate(\\\"en\\\")\\n self.failUnlessEqual(n.title, \\\"\\\")\",\n \"def test_issue3625():\\n nlp = Hindi()\\n doc = nlp(u\\\"hi. how हुए. होटल, होटल\\\")\\n assert [token.text for token in doc] == ['hi', '.', 'how', 'हुए', '.', 'होटल', ',', 'होटल']\",\n \"def test_get_supported_locales_for_voice_datasets(self):\\n pass\",\n \"def nintl(self):\",\n \"def test_01_i18n_authenticated(self):\\r\\n with self.app as c:\\r\\n # First default 'en' locale\\r\\n err_msg = \\\"The page should be in English\\\"\\r\\n res = c.get('/', follow_redirects=True)\\r\\n assert \\\"Community\\\" in res.data, err_msg\\r\\n self.register()\\r\\n self.signin()\\r\\n # After signing in it should be in English\\r\\n err_msg = \\\"The page should be in English\\\"\\r\\n res = c.get('/', follow_redirects=True)\\r\\n assert \\\"Community\\\" in res.data, err_msg\\r\\n\\r\\n # Change it to Spanish\\r\\n user = db.session.query(User).filter_by(name='johndoe').first()\\r\\n user.locale = 'es'\\r\\n db.session.add(user)\\r\\n db.session.commit()\\r\\n\\r\\n res = c.get('/', follow_redirects=True)\\r\\n err_msg = \\\"The page should be in Spanish\\\"\\r\\n assert \\\"Comunidad\\\" in res.data, err_msg\\r\\n # Sign out should revert it to English\\r\\n self.signout()\\r\\n err_msg = \\\"The page should be in English\\\"\\r\\n res = c.get('/', follow_redirects=True)\\r\\n assert \\\"Community\\\" in res.data, err_msg\",\n \"def test_no_cyrillic_string(self):\\n string = \\\"test\\\"\\n expected = \\\"test\\\"\\n self.assertEqual(transliterate(string), expected)\",\n \"def t(eng, chinese):\\n return chinese if 'zh' in get_info().user_language else eng\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.70812863","0.6640255","0.6615368","0.6592845","0.6592845","0.6576762","0.65381277","0.64886284","0.64697886","0.6454866","0.642305","0.63667893","0.636136","0.632902","0.632902","0.6252181","0.61971015","0.6178152","0.61683005","0.6137992","0.6137528","0.61306524","0.6122047","0.6120748","0.6118362","0.61151403","0.6104679","0.6104679","0.6098567","0.60962194","0.6090324","0.6090324","0.6063938","0.60591656","0.60494673","0.60418123","0.60024965","0.5997934","0.59861493","0.5951402","0.59497595","0.5924202","0.58975905","0.5894607","0.58796555","0.5864898","0.5862795","0.5862482","0.5860084","0.5856459","0.58301514","0.58119893","0.58022463","0.57895446","0.57796913","0.57796913","0.5767002","0.57658666","0.5760826","0.5756322","0.5734087","0.5730883","0.5725482","0.5724834","0.5724834","0.5718158","0.57089293","0.5698896","0.5698896","0.569512","0.56925064","0.5652747","0.56489486","0.5638387","0.5624883","0.5617755","0.5612267","0.56090194","0.56053805","0.56037974","0.5593602","0.5593167","0.5592861","0.55916727","0.55888504","0.55888504","0.558748","0.55862176","0.55763716","0.55689883","0.55689883","0.5562557","0.5548645","0.55310535","0.55203056","0.55097693","0.5493467","0.5447561","0.54467326","0.54416347"],"string":"[\n \"0.70812863\",\n \"0.6640255\",\n \"0.6615368\",\n \"0.6592845\",\n \"0.6592845\",\n \"0.6576762\",\n \"0.65381277\",\n \"0.64886284\",\n \"0.64697886\",\n \"0.6454866\",\n \"0.642305\",\n \"0.63667893\",\n \"0.636136\",\n \"0.632902\",\n \"0.632902\",\n \"0.6252181\",\n \"0.61971015\",\n \"0.6178152\",\n \"0.61683005\",\n \"0.6137992\",\n \"0.6137528\",\n \"0.61306524\",\n \"0.6122047\",\n \"0.6120748\",\n \"0.6118362\",\n \"0.61151403\",\n \"0.6104679\",\n \"0.6104679\",\n \"0.6098567\",\n \"0.60962194\",\n \"0.6090324\",\n \"0.6090324\",\n \"0.6063938\",\n \"0.60591656\",\n \"0.60494673\",\n \"0.60418123\",\n \"0.60024965\",\n \"0.5997934\",\n \"0.59861493\",\n \"0.5951402\",\n \"0.59497595\",\n \"0.5924202\",\n \"0.58975905\",\n \"0.5894607\",\n \"0.58796555\",\n \"0.5864898\",\n \"0.5862795\",\n \"0.5862482\",\n \"0.5860084\",\n \"0.5856459\",\n \"0.58301514\",\n \"0.58119893\",\n \"0.58022463\",\n \"0.57895446\",\n \"0.57796913\",\n \"0.57796913\",\n \"0.5767002\",\n \"0.57658666\",\n \"0.5760826\",\n \"0.5756322\",\n \"0.5734087\",\n \"0.5730883\",\n \"0.5725482\",\n \"0.5724834\",\n \"0.5724834\",\n \"0.5718158\",\n \"0.57089293\",\n \"0.5698896\",\n \"0.5698896\",\n \"0.569512\",\n \"0.56925064\",\n \"0.5652747\",\n \"0.56489486\",\n \"0.5638387\",\n \"0.5624883\",\n \"0.5617755\",\n \"0.5612267\",\n \"0.56090194\",\n \"0.56053805\",\n \"0.56037974\",\n \"0.5593602\",\n \"0.5593167\",\n \"0.5592861\",\n \"0.55916727\",\n \"0.55888504\",\n \"0.55888504\",\n \"0.558748\",\n \"0.55862176\",\n \"0.55763716\",\n \"0.55689883\",\n \"0.55689883\",\n \"0.5562557\",\n \"0.5548645\",\n \"0.55310535\",\n \"0.55203056\",\n \"0.55097693\",\n \"0.5493467\",\n \"0.5447561\",\n \"0.54467326\",\n \"0.54416347\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.57219356"},"document_rank":{"kind":"string","value":"65"}}},{"rowIdx":90888,"cells":{"query":{"kind":"string","value":"Create an instance of the FriCAS interpreter."},"document":{"kind":"string","value":"def __init__(self, name='fricas', command='fricas -nox -noclef',\n script_subdirectory=None, logfile=None,\n server=None, server_tmpdir=None,\n init_code=[')lisp (si::readline-off)']):\n PanAxiom.__init__(self, name, command,\n script_subdirectory, logfile,\n server, server_tmpdir,\n init_code)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def __create_execution_engine(self):\n target = self.binding.Target.from_default_triple()\n target_machine = target.create_target_machine()\n backing_mod = binding.parse_assembly(\"\")\n engine = binding.create_mcjit_compiler(backing_mod, target_machine)\n self.engine = engine","def Program(self):\n funcs = list()\n while self.currtok[1].name in {\"INT\", \"FLOAT\", \"BOOLEAN\"}:\n func = self.FunctionDef()\n funcs.append(func)\n return Program(funcs)","def factory(args, cxn, log):\n name = args['assembler'].lower()\n assembler = ASSEMBLERS[name]\n return assembler(args, cxn, log)","def __init__ ( self , syms , defn ):\n\n self.logic = cognitiveDefiner.convertDefinition(syms,defn)","def __init__(self, *args):\n this = _ida_hexrays.new_casm_t(*args)\n try: self.this.append(this)\n except: self.this = this","def __init__(self, fen): \n self.d_engine = Stockfish()\n self.setFen(fen)","def __init__(self):\n self._lib_vscf_ecc = VscfEcc()\n self._c_impl = None\n self._ctx = None\n self.ctx = self._lib_vscf_ecc.vscf_ecc_new()","def create_engine(self, input):\n return","def create_execution_engine():\n # Create a target machine representing the host\n target = llvm.Target.from_default_triple()\n target_machine = target.create_target_machine()\n # And an execution engine with an empty backing module\n backing_mod = llvm.parse_assembly(\"\")\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\n return engine","def create_execution_engine():\n # Create a target machine representing the host\n target = llvm.Target.from_default_triple()\n target_machine = target.create_target_machine()\n # And an execution engine with an empty backing module\n backing_mod = llvm.parse_assembly(\"\")\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\n return engine","def create_execution_engine():\n # Create a target machine representing the host\n target = llvm.Target.from_default_triple()\n target_machine = target.create_target_machine()\n # And an execution engine with an empty backing module\n backing_mod = llvm.parse_assembly(\"\")\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\n return engine","def __init__(self, engine: str = \"sfdp\"):\n self.engine = engine","def __init__(self, regex, fixed_slice):\n self._regex = regex\n self._fixed_slice = fixed_slice\n self._dfa = fte.dfa.from_regex(self._regex, self._fixed_slice)\n self._encrypter = fte.encrypter.Encrypter()","def _new_instance(self):\n return self.__class__(self._fmodule)","def _new_instance(self):\n return self.__class__(self._fmodule)","def instantiate_algorithm(args):\n if args.algorithm == 'xor_encoding':\n return XorEncoding(block_size=args.block_size, intensity=args.intensity)\n\n raise RuntimeError('Algorithm type not detected')","def parse_main(self):\n self.decompiler = self.createDecompiler(self.compilerEnv)\n self.nf = IRFactory(self)\n self.currentScriptOrFn = self.nf.createScript()\n sourceStartOffset = self.decompiler.getCurrentOffset();\n self.encodedSource = None;\n self.decompiler.addToken(Token.SCRIPT);\n \n self.currentFlaggedToken = Token.EOF;\n self.syntaxErrorCount = 0;\n\n baseLineno = self.ts.getLineno(); #// line number where source starts\n\n #/* so we have something to add nodes to until\n # * we've collected all the source */\n pn = self.nf.createLeaf(Token.BLOCK);\n try:\n while(True):\n tt = self.peekToken()\n if (tt <= Token.EOF):\n break\n n = None\n \n if tt == Token.FUNCTION:\n self.consumeToken()\n try: \n n = self.function( \\\n (FunctionNode.FUNCTION_EXPRESSION,\\\n FunctionNode.FUNCTION_STATEMENT)[self.calledByCompileFunction])\n except ParserException:\n raise NotImplementedError()# should this have to raise?\n break;\n #raise NotImplementedError <- I think this is now implemented - TW\n \n else:\n n = self.statement()\n \n self.nf.addChildToBack(pn, n)\n \n \n except RuntimeError:\n # Was StackOverflowError\n raise\n # TODO: exception handling\n \n if (self.syntaxErrorCount != 0) :\n msg = str(self.syntaxErrorCount)\n #msg = ScriptRuntime.getMessage1(\"msg.got.syntax.errors\", msg);\n #throw errorReporter.runtimeError(msg, sourceURI, baseLineno,\n # null, 0);\n\n self.currentScriptOrFn.setSourceName(self.sourceURI);\n self.currentScriptOrFn.setBaseLineno(baseLineno);\n self.currentScriptOrFn.setEndLineno(self.ts.getLineno());\n\n sourceEndOffset = self.decompiler.getCurrentOffset();\n self.currentScriptOrFn.setEncodedSourceBounds(sourceStartOffset,\n sourceEndOffset);\n\n self.nf.initScript(self.currentScriptOrFn, pn);\n\n if (self.compilerEnv.isGeneratingSource()):\n self.encodedSource = self.decompiler.getEncodedSource();\n \n del self.decompiler# comment was //\"It helps GC\" \n # - can't do any harm on CPython either\n\n return self.currentScriptOrFn;","def __init__(self, foretext=\"::>>> \"):\n self.keywords = []\n\n # self.functions is a *dict* with\n # key: keyword, value: (function, args, kwargs)\n self.functions = {}\n\n self.foretext = foretext\n self.input_asked = False\n self.retrieved_input = None","def New(*args, **kargs):\n obj = itkIsoDataThresholdCalculatorHFSS.__New_orig__()\n import itkTemplate\n itkTemplate.New(obj, *args, **kargs)\n return obj","def run(_):\n fscad = types.ModuleType(\"fscad.fscad\")\n fscad.__path__ = [os.path.dirname(os.path.realpath(__file__))]\n sys.modules['fscad.fscad'] = fscad\n\n for key in __all__:\n fscad.__setattr__(key, globals()[key])","def main():\n create_fc(\"layers/WAPR.lyr\",\n \"TravelerInfo.gdb/HighwayAlerts\")","def __init__(self, *args):\n _BRepAlgo.BRepAlgo_Fuse_swiginit(self,_BRepAlgo.new_BRepAlgo_Fuse(*args))","def build(self, name='impala'):\n program = lp.Program(name=name)\n\n with program.group('replay'):\n queue = program.add_node(lp.ReverbNode(self.queue))\n\n with program.group('counter'):\n counter = program.add_node(lp.CourierNode(self.counter))\n\n with program.group('learner'):\n learner = program.add_node(\n lp.CourierNode(self.learner, queue, counter))\n\n with program.group('evaluator'):\n program.add_node(lp.CourierNode(self.evaluator, learner, counter))\n\n with program.group('cacher'):\n cacher = program.add_node(\n lp.CacherNode(learner, refresh_interval_ms=2000, stale_after_ms=4000))\n\n with program.group('actor'):\n for _ in range(self._num_actors):\n program.add_node(lp.CourierNode(self.actor, queue, cacher, counter))\n\n return program","def New(*args, **kargs):\n obj = itkIsoDataThresholdCalculatorHFUS.__New_orig__()\n import itkTemplate\n itkTemplate.New(obj, *args, **kargs)\n return obj","def __init__(self, pele_exec, pele_src, pele_license,\n ploprottemp_src, schrodinger_src,\n charge_method='am1bcc', solvent='OBC',\n opls_nonbonding=False, opls_bonds_angles=False,\n n_proc=1, forcefield_name=None, forcefield=None):\n self.pele_exec = pele_exec\n self.pele_src = pele_src\n self.pele_license = pele_license\n self.charge_method = charge_method\n self.solvent = solvent\n self.opls_nonbonding = opls_nonbonding\n self.opls_bonds_angles = opls_bonds_angles\n self.ploprottemp_src = ploprottemp_src\n self.schrodinger_src = schrodinger_src\n self._n_proc = n_proc\n self.forcefield_name = forcefield_name\n self.forcefield = forcefield\n self._results = dict()\n\n # Deactivate peleffy output\n from peleffy.utils import Logger\n logger = Logger()\n logger.set_level('WARNING')\n\n # Supress OpenForceField toolkit warnings\n import logging\n logging.getLogger().setLevel(logging.ERROR)","def init(config=config, version='crystal'):\n setup_logging()\n configure_for_version(version, config)\n get_constants(config=config)\n return main(config=config)","def disassembler_factory():\n \n ir_id, ir_cls, dis_cls = find_current_arch()\n \n class disassembler(dis_cls, ir_cls): # disassembler (host) class must be left-most.\n def __init__(self, ir_id):\n self.ir_id = ir_id\n dis_cls.__init__(self)\n ir_cls.__init__(self)\n return\n \n dis = disassembler(ir_id)\n \n return dis","def __init__(self, caldb_name, irf_name, verbose=False):\n\n self.caldb_path = os.environ['CALDB']\n self.caldb_name = caldb_name\n self.irf = irf_name\n self.verbose = verbose\n\n self.am_ok = True\n\n self._aeff = dict()\n self._psf = dict()\n self._edips = dict()\n\n self._check_available_irfs()\n\n self.input_irf_file_name = '{path:s}/data/cta/{caldb:s}/bcf/{irf:s}/irf_file.fits'.format(path=self.caldb_path,\n caldb=self.caldb_name,\n irf=irf_name)","def __init__(self):\n self.reg = [0] * 8\n self.ram = [0] * 256\n self.pc = 0\n self.sp = 7\n\n # state (running)\n self.running = True\n\n # REPL (FETCH, DECODE, EXECUTE)\n self.op_pc = False\n\n self.branch_table = {}\n self.branch_table[PRN] = self.prn\n self.branch_table[LDI] = self.ldi\n self.branch_table[CMP] = self.cmp\n self.branch_table[JMP] = self.jmp\n self.branch_table[JEQ] = self.jeq\n self.branch_table[JNE] = self.jne","def initialize_implementation(self):\n # call overridable methods to get number of inputs/outputs\n n_inputs = self.get_number_of_input_connections()\n n_outputs = self.get_number_of_output_ports()\n\n # call overrides to get implementation for teca execution\n # phase implementations\n import teca_py\n self.impl = teca_py.teca_programmable_algorithm.New()\n self.impl.set_number_of_input_connections(n_inputs)\n self.impl.set_number_of_output_ports(n_outputs)\n self.impl.set_name(self.__class__.__name__)\n self.impl.set_report_callback(self.get_report_callback())\n self.impl.set_request_callback(self.get_request_callback())\n self.impl.set_execute_callback(self.get_execute_callback())","def main() -> None:\n version_path = './src/fqe/_version.py'\n\n __version__ = version_number(version_path)\n\n if __version__ is None:\n raise ValueError('Version information not found in ' + version_path)\n\n long_description = ('OpenFermion-FQE\\n' +\n '===============\\n')\n stream = io.open('README.md', encoding='utf-8')\n stream.readline()\n long_description += stream.read()\n\n requirements_buffer = open('requirements.txt').readlines()\n requirements = [r.strip() for r in requirements_buffer]\n\n # C code extension\n config_vars = get_config_vars()\n config_vars[\"EXT_SUFFIX\"] = '.' + config_vars[\"EXT_SUFFIX\"].split('.')[-1]\n libdir = os.path.join(\"src\", \"fqe\", \"lib\")\n cfiles = [\n \"macros.c\",\n \"mylapack.c\",\n \"fci_graph.c\",\n \"fqe_data.c\",\n \"cirq_utils.c\",\n \"wick.c\",\n \"bitstring.c\",\n \"binom.c\",\n ]\n srcs = [os.path.join(libdir, cf) for cf in cfiles]\n libraries = []\n extensions = [\n Extension(\"fqe.lib.libfqe\",\n srcs,\n include_dirs=[libdir],\n library_dirs=[libdir],\n libraries=libraries,\n language='c')\n ]\n\n cythonfiles = [\"_fqe_data.pyx\"]\n srcs = [os.path.join(libdir, cf) for cf in cythonfiles]\n extensions.append(Extension(\"fqe.lib.fqe_data\", srcs, language='c'))\n\n setup(name='fqe',\n version=__version__,\n author='The OpenFermion FQE Developers',\n author_email='help@openfermion.org',\n url='http://www.openfermion.org',\n description='OpenFermion Fermionic Quantum Emulator',\n ext_modules=cythonize(extensions,\n compiler_directives={'language_level': \"3\"}),\n long_description=long_description,\n long_description_content_type=\"text/markdown\",\n install_requires=requirements,\n license='Apache 2',\n packages=find_packages(where='src'),\n package_dir={'': 'src'},\n cmdclass={'build_ext': CustomBuildOptions})","def create(cls, name, defn, state=None):\n prog = cls(name, False)\n prog.raw_data = defn\n prog.state = state or {}\n prog.mk_data()\n\n return prog","def init(self):\n self._frame_idx = 0\n if self.initialization is not None:\n del self.initialization\n self.initialization = None\n\n if self.config.initialization is not None:\n self.initialization = INITIALIZATION.load(self.config.initialization, **self.__kwargs)\n self.initialization.init()\n\n if self.preprocessing is not None:\n del self.preprocessing\n self.preprocessing = None\n\n if self.config.preprocessing is not None:\n self.preprocessing = Preprocessing(self.config.preprocessing, **self.__kwargs)\n\n if self.odometry is None:\n assert self.config.odometry is not None\n self.odometry = ODOMETRY.load(self.config.odometry, **self.__kwargs)\n\n assert self.odometry is not None\n self.odometry.init()\n if self.loop_closure is None and self.config.loop_closure is not None:\n self.loop_closure = LOOP_CLOSURE.load(self.config.loop_closure, **self.__kwargs)\n if self.loop_closure is not None:\n self.loop_closure.init()\n if self.config.backend is not None:\n self.backend = BACKEND.load(self.config.backend, **self.__kwargs)\n if self.backend is not None:\n self.backend.init()\n else:\n logging.warning(\"[SLAMAlgorithm]Defined a Loop Closure Algorithm Without a Backend\")","def make(self, *args, **kwargs):\r\n self.make_module_instance(args,kwargs)\r\n\r\n mode = kwargs.pop('mode', theano.compile.mode.get_default_mode())\r\n rval = self.make_no_init(mode)\r\n if hasattr(rval, 'initialize'):\r\n rval.initialize(*args, **kwargs)\r\n return rval","def __init__(self):\r\n\t\tself.introducer()\r\n\t\tif self.code_mode == \"1\":\r\n\t\t\tif self.input_mode == \"1\":\r\n\t\t\t\tself.encrypt_message()\r\n\t\t\telse:\r\n\t\t\t\tself.encrypt_text_file()\r\n\t\t\t\t#print(\"work in progress\")\r\n\t\telif self.code_mode == \"2\":\r\n\t\t\tif self.input_mode == \"1\":\r\n\t\t\t\tself.decrypt_message()\r\n\t\t\telse:\r\n\t\t\t\tself.decrypt_text_file()\r\n\t\telse:\r\n\t\t\tif self.input_mode == \"1\":\r\n\t\t\t\tself.hack_message()\r\n\t\t\telse:\r\n\t\t\t\tself.hack_text_file()","def main():\n get_engine(onnx_file_path, engine_file_path)","def main() -> None:\n # The first thing to do is get the lines of the PyFlex file we are given.\n parser = Parser(filename=sys.argv[1])\n parsed_data = parser.ParseFile()\n\n # Upon retrieving the Parsed Data, assign the parsed data to the\n # Symbol Table.\n SymbolTable.RULESET = parsed_data['ruleset']\n SymbolTable.INSTRUCTIONS = parsed_data['instructions']\n SymbolTable.CODE = parsed_data['code']\n # SymbolTable.PrintTable()\n\n # Using the Generator backend, we can build the generated script\n generator = Generator()\n generator.GenerateNewScript()\n\n autopep8.fix_file(filename=generator.file_main)\n\n print(\"Generated Script can be found in {}\".format(generator.file_main))","def __init__(self, *args):\n _hypre.HypreTriSolve_swiginit(self, _hypre.new_HypreTriSolve(*args))","def __init__(self, vplprog):\n\n self.validated = False\n self.consts = dict()\n self.functions = []\n funcnames = set()\n\n for child in vplprog.children:\n func = FunctionNode(child, self.consts)\n\n if func.name in funcnames:\n raise VPLSyntaxError(\"Duplicate function name '{0}'\".format(func.name))\n\n funcnames.add(func.name)\n self.functions.append(func)","def __init__(self):\n self.label = \"TOC Fuzzification\"\n self.description = \"This fuzzification method utilized the symbolization of the input raster that has been applied in the map document table of contects (TOC). The symbolization in the TOC defines the number of classes and this tool rescales those classes (1...N) to the range [0,1] by (C - 1)/(N-1) where C is the class value and N is the number of classes.\"\n self.canRunInBackground = False\n self.category = \"Fuzzy Logic\\\\Fuzzy Membership\"","def __init__(self, *args):\n _snap.TAscFlt_swiginit(self, _snap.new_TAscFlt(*args))","def __init__(self):\n self.libpath = os.sep.join(os.path.abspath(__file__).split(os.sep)[:-1])\n sys.path.append(self.libpath)\n libpath2 = os.sep.join(self.libpath.split(os.sep)[:-1])\n sys.path.append(libpath2)\n # Initialize TCMetaSchema with correct libpath\n TCMetaSchema(self.libpath)\n self.args, self.unknown = IceteaManager._parse_arguments()\n # If called with --clean, clean up logs.\n if self.args.clean:\n _cleanlogs(silent=self.args.silent, log_location=self.args.log)\n\n LogManager.init_base_logging(self.args.log, verbose=self.args.verbose,\n silent=self.args.silent, color=self.args.color,\n no_file=(self.args.list or self.args.listsuites),\n truncate=not self.args.disable_log_truncate)\n\n self.logger = LogManager.get_logger(\"icetea\")\n self.pluginmanager = None\n self.resourceprovider = ResourceProvider(self.args)\n self._init_pluginmanager()\n self.resourceprovider.set_pluginmanager(self.pluginmanager)","def base_calc(fresh_aiida_env, vasp_code):\n from aiida_vasp.calcs.base import VaspCalcBase\n manager = get_manager()\n runner = manager.get_runner()\n inputs = AttributeDict()\n\n metadata = AttributeDict({'options': {'resources': {'num_machines': 1, 'num_mpiprocs_per_machine': 1}}})\n\n inputs.code = vasp_code\n inputs.metadata = metadata\n\n return instantiate_process(runner, VaspCalcBase, **inputs)","def main():\n openff = Openff()\n openff.console.welcome()","def main() -> None:\n parser = argparse.ArgumentParser(description=\"Flower\")\n parser.add_argument(\n \"--server_address\",\n type=str,\n default=DEFAULT_SERVER_ADDRESS,\n help=f\"gRPC server address (default: {DEFAULT_SERVER_ADDRESS})\",\n )\n parser.add_argument(\n \"--cid\", type=str, required=True, help=\"Client CID (no default)\"\n )\n parser.add_argument(\n \"--log_host\", type=str, help=\"Logserver address (no default)\",\n )\n parser.add_argument(\n \"--nb_clients\", type=int, default=10, help=\"Total number of clients\",\n )\n args = parser.parse_args()\n\n # Configure logger\n fl.common.logger.configure(f\"client_{args.cid}\", host=args.log_host)\n\n # Load model and data\n #model = cifar.load_model()\n #model.to(DEVICE)\n #trainset, testset = cifar.load_data()\n\n # Start client\n #client = CifarClient(args.cid, model, trainset, testset, args.nb_clients)\n #client = CifarClient(args.cid, trainset, testset, args.nb_clients)\n client = CifarClient(args.cid,args.nb_clients)\n fl.client.start_client(args.server_address, client)","def __init__(self) -> None:\n flags.reset()\n self.flags = flags\n self.super_memory = SuperMemory()\n self.memory = self.super_memory.memory\n self.super_memory.PC(\"0x0800\")\n self._registers_list = {\n \"A\": self.super_memory.A,\n \"PSW\": self.flags.PSW,\n \"B\": self.super_memory.BC,\n \"C\": self.super_memory.BC,\n \"D\": self.super_memory.DE,\n \"E\": self.super_memory.DE,\n \"H\": self.super_memory.HL,\n \"L\": self.super_memory.HL,\n \"SP\": self.super_memory.SP,\n \"PC\": self.super_memory.PC,\n \"M\": self.super_memory.M,\n }\n self._lookup_opcodes_dir = opcodes_lookup\n\n self._keywords = []\n self._generate_keywords()\n self._assembler = {}\n pass","def __init__(self):\n self.RRTFamilySolver = RRTFamilyPathPlanner()\n self.PRMSolver = PRMPathPlanner()","def __init__(self, *args):\n if self.__class__ == codegen_t:\n _self = None\n else:\n _self = self\n this = _ida_hexrays.new_codegen_t(_self, *args)\n try: self.this.append(this)\n except: self.this = this","def main():\n init_latex()","def __init__(self, *args):\n this = _libsbml.new_FunctionDefinition(*args)\n try: self.this.append(this)\n except: self.this = this","def __init__(self, osi, fy, fu, es, esh, eps_sh, eps_ult, cf, alpha_2, cd):\n self.osi = osi\n self.fy = float(fy)\n self.fu = float(fu)\n self.es = float(es)\n self.esh = float(esh)\n self.eps_sh = float(eps_sh)\n self.eps_ult = float(eps_ult)\n self.cf = float(cf)\n self.alpha_2 = alpha_2\n self.cd = float(cd)\n if osi is not None:\n osi.n_mat += 1\n self._tag = osi.n_mat\n self._parameters = [self.op_type, self._tag, self.fy, self.fu, self.es, self.esh, self.eps_sh, self.eps_ult, '-CMFatigue', self.cf, self.alpha_2, self.cd]\n if osi is None:\n self.built = 0\n if osi is not None:\n self.to_process(osi)","def create_instance(c_instance):\n return MonoPedal(c_instance)","def __init__(self, comm):\n _hypre.HypreAME_swiginit(self, _hypre.new_HypreAME(comm))","def __init__(self): \n\t\n\t # get the environment\n\t\tself.env = env()","def Build(self, mode):\n return _gmat_py.Interpreter_Build(self, mode)","def __init__(self):\n\n # Initialize the fields that all formulae have\n self.nvars = 0\n self.nclauses = 0\n self.base_clauses = []\n self.aux_clauses = []\n\n # Create variable meaning table\n #\n # Variable meaning is a table which maps boolean varibles to their\n # semantics. The semantics of a variable are reprsented a 3-tuple.\n # The first element of the tuple says whether the variable is an\n # assignment, clustering, activation, or commander variable. If the\n # variable is an assignment variable, the second element is the\n # relevant signature and the third is the relevenant methyl. If the\n # variable is a clustering variable, the second element is the\n # relevant NOE and the third is the relevant signature. If it is an\n # activation variable, then the 2nd a 3rd elements are the relevant\n # NOEs (in arbitrary order).\n\n self.variable_meaning = {}","def __init__(self, *args):\n this = _ida_hexrays.new_cinsn_t(*args)\n try: self.this.append(this)\n except: self.this = this","def __init__(self, *args):\n this = _ida_hexrays.new_cif_t(*args)\n try: self.this.append(this)\n except: self.this = this","def New(*args, **kargs):\n obj = itkIsoDataThresholdCalculatorHFF.__New_orig__()\n import itkTemplate\n itkTemplate.New(obj, *args, **kargs)\n return obj","def __init__(self, hostname = '', port = 0):\n ScriptForge.hostname = hostname\n ScriptForge.port = port\n # Determine main pyuno entry points\n ScriptForge.componentcontext = self.ConnectToLOProcess(hostname, port) # com.sun.star.uno.XComponentContext\n ScriptForge.scriptprovider = self.ScriptProvider(self.componentcontext) # ...script.provider.XScriptProvider\n #\n # Establish a list of the available services as a dictionary (servicename, serviceclass)\n ScriptForge.serviceslist = dict((cls.servicename, cls) for cls in SFServices.__subclasses__())\n ScriptForge.servicesdispatcher = None\n #\n # All properties and methods of the ScriptForge API are ProperCased\n # Compute their synonyms as lowercased and camelCased names\n ScriptForge.SetAttributeSynonyms()\n #\n ScriptForge.SCRIPTFORGEINITDONE = True","def __init__(self, *args):\n this = _ida_hexrays.new_cloop_t(*args)\n try: self.this.append(this)\n except: self.this = this","def __init__(self, restart=None, ignore_bad_restart_file=False,\n label='dftb', atoms=None, kpts=None,\n run_manyDftb_steps=False, slako_dir='./',\n **kwargs):\n\n # skf tables\n #if 'DFTB_PREFIX' in os.environ:\n # self.slako_dir = os.environ['DFTB_PREFIX'].rstrip('/') + '/'\n #else:\n # self.slako_dir = './'\n #slako_dir = os.path.abspath(slako_dir) + '/'\n self.slako_dir = slako_dir.rstrip('/') + '/'\n\n # file names\n self.geo_fname = 'geo_in.gen'\n\n if run_manyDftb_steps:\n # minimisation of molecular dynamics is run by native DFTB+\n self.default_parameters = dict(\n Hamiltonian_='DFTB',\n Hamiltonian_SlaterKosterFiles_='Type2FileNames',\n Hamiltonian_SlaterKosterFiles_Prefix=self.slako_dir,\n Hamiltonian_SlaterKosterFiles_Separator='\"-\"',\n Hamiltonian_SlaterKosterFiles_Suffix='\".skf\"',\n Hamiltonian_MaxAngularMomentum_='')\n else:\n # using ase to get forces and energy only\n # (single point calculation)\n self.default_parameters = dict(\n Hamiltonian_='DFTB',\n Hamiltonian_SlaterKosterFiles_='Type2FileNames',\n Hamiltonian_SlaterKosterFiles_Prefix=self.slako_dir,\n Hamiltonian_SlaterKosterFiles_Separator='\"-\"',\n Hamiltonian_SlaterKosterFiles_Suffix='\".skf\"',\n Hamiltonian_MaxAngularMomentum_='')\n\n self.pcpot = None\n self.lines = None # results.tag\n\n self.atoms = None\n self.atoms_input = None\n\n self.do_forces = False\n self.outfilename = 'dftb.out'\n\n FileIOCalculator.__init__(self, restart, ignore_bad_restart_file,\n label, atoms,\n **kwargs)\n\n # Determine number of spin channels\n try:\n entry = kwargs['Hamiltonian_SpinPolarisation']\n spinpol = 'colinear' in entry.lower()\n except KeyError:\n spinpol = False\n self.nspin = 2 if spinpol else 1\n\n # kpoint stuff by ase\n self.kpts = kpts\n self.kpts_coord = None\n\n if self.kpts is not None:\n initkey = 'Hamiltonian_KPointsAndWeights'\n mp_mesh = None\n offsets = None\n \n if isinstance(self.kpts, dict):\n if 'path' in self.kpts:\n # kpts is path in Brillouin zone\n self.parameters[initkey + '_'] = 'Klines '\n self.kpts_coord = kpts2ndarray(self.kpts, atoms=atoms)\n else:\n # kpts is (implicit) definition of\n # Monkhorst-Pack grid\n self.parameters[initkey + '_'] = 'SupercellFolding '\n mp_mesh, offsets = kpts2sizeandoffsets(atoms=atoms,\n **self.kpts)\n elif np.array(self.kpts).ndim == 1:\n # kpts is Monkhorst-Pack grid\n self.parameters[initkey + '_'] = 'SupercellFolding '\n mp_mesh = self.kpts\n offsets = [0.] * 3\n elif np.array(self.kpts).ndim == 2:\n # kpts is (N x 3) list/array of k-point coordinates\n # each will be given equal weight\n self.parameters[initkey + '_'] = ''\n self.kpts_coord = np.array(self.kpts)\n else:\n raise ValueError('Illegal kpts definition:' + str(self.kpts))\n\n if mp_mesh is not None:\n eps = 1e-10\n for i in range(3):\n key = initkey + '_empty%03d' % i\n val = [mp_mesh[i] if j == i else 0 for j in range(3)]\n self.parameters[key] = ' '.join(map(str, val))\n offsets[i] *= mp_mesh[i]\n assert abs(offsets[i]) < eps or abs(offsets[i] - 0.5) < eps\n # DFTB+ uses a different offset convention, where\n # the k-point mesh is already Gamma-centered prior\n # to the addition of any offsets\n #if mp_mesh[i] % 2 == 0:\n # offsets[i] += 0.5 # use gamma-centred\n key = initkey + '_empty%03d' % 3\n self.parameters[key] = ' '.join(map(str, offsets))\n\n elif self.kpts_coord is not None:\n for i, c in enumerate(self.kpts_coord):\n key = initkey + '_empty%09d' % i\n c_str = ' '.join(map(str, c))\n if 'Klines' in self.parameters[initkey + '_']:\n c_str = '1 ' + c_str\n else:\n c_str += ' 1.0'\n self.parameters[key] = c_str","def solidity_compiler():\n compiler = SolidityCompiler(test_contract_dir=TEST_CONTRACTS_DIR)\n yield compiler","def __init__(self, config, fe_space, scheme_type=SchemeType.EXPLICIT_ORDERONE, init_cond_type=InitialConditionType.NONE):\n self.config = config\n self.fe_space = fe_space\n self.init_cond_type = init_cond_type\n self.scheme_type = scheme_type\n\n self.u0 = None\n self.u1 = None\n self.u2 = None\n self.ustar = None\n self.operator1 = None\n self.operator2 = None\n self.rhs_operator = None\n self.inv_operator = None\n self.timestep = None\n self.time = 0.0","def main():\n run_nutanix_vm_creation_module()","def __init__(self, static_lib):\n self.tool = self.find_tool()\n if self.tool is None:\n # ERROR no NM found\n print(\"ERROR: Could not find readelf\")\n sys.exit(1)\n # TODO: Support readelf for reading symbols from archives\n assert not static_lib and \"RealElf does not yet support static libs\"\n self.flags = ['--wide', '--symbols']","def __init__(self):\n self.read_input()\n self.update_binaries()","def initialise(self, args, environ):","def test_init(paper):\n atSi = Atoms(\"Si8\",positions=[[0,0,0],[0.25,0.25,0.25],[0.5,0.5,0],[0.75,0.75,0.25],\n [0.5,0,0.5],[0.75,0.25,0.75],[0,0.5,0.5],[0.25,0.75,0.75]],\n cell=[5.43,5.43,5.43])\n\n entry = paper[0]\n kwargs = {}\n args = []\n\n calc = Aflow(atSi, '.', '.', 0, entry=entry, *args, **kwargs)\n\n assert calc.entry_file == \"./entry.pkl\"\n assert calc.can_execute()\n assert calc.can_extract()\n assert not calc.is_executing()\n calc.extract()","def __init__(self, *args):\n _hypre.HyprePCG_swiginit(self, _hypre.new_HyprePCG(*args))","def __init__(self, name, rootdir=None, create=True):\n # Guess the dtype based on the class name\n dtype = FreezableAPI.guess_type(self)\n self.m80 = FreezableAPI(dtype, name, rootdir)","def createACTopol(self):\n #sleap = False\n if self.engine == 'sleap':\n if self.execSleap():\n self.printError(\"Sleap failed\")\n self.printMess(\"... trying Tleap\")\n if self.execTleap():\n self.printError(\"Tleap failed\")\n if self.engine == 'tleap':\n if self.execTleap():\n self.printError(\"Tleap failed\")\n if self.extOld == '.pdb':\n self.printMess(\"... trying Sleap\")\n self.ext = self.extOld\n self.inputFile = self.baseName+self.ext\n if self.execSleap():\n self.printError(\"Sleap failed\")\n if not self.debug:\n self.delOutputFiles()","def __init__(self, evaluation_only=False, seed=0):\n os = platform.system()\n if os == 'Darwin':\n file_name = 'Soccer.app'\n elif os == 'Linux':\n file_name = 'Soccer_Linux/Soccer.x86_64'\n self.env = UnityEnvironment(file_name='unity_envs/' + file_name, seed=seed)\n self.brain_names = self.env.brain_names\n self.evaluation_only = evaluation_only","def __init__(self, filename, default_seq=None, key_function=None, as_raw=False, strict_bounds=False):\n self.filename = filename\n self.faidx = Faidx(filename, key_function=key_function, as_raw=as_raw,\n default_seq=default_seq, strict_bounds=strict_bounds)","def New(*args, **kargs):\n obj = itkIsoDataThresholdCalculatorHFUC.__New_orig__()\n import itkTemplate\n itkTemplate.New(obj, *args, **kargs)\n return obj","def __init__(self, *args):\n _snap.TAscFltVP_swiginit(self, _snap.new_TAscFltVP(*args))","def __init__(self, *args):\n _snap.TAscFltPr_swiginit(self, _snap.new_TAscFltPr(*args))","def create_constructor(self, *args, **kws):\n node = FunctionNode(\"<constructor>\", *args, **kws)\n node.class_member = True\n self.add_object(node)\n self.constructors.append(node)\n return node","def __init__(self):\n\n # Memory (RAM)\n self.ram = [0] * 256\n # Registers\n self.reg = [0] * 8\n # Program Counter\n self.pc = 0\n # Stack pointer\n self.reg[7] = 244\n # CPU status\n self.running = False\n # Flags register\n self.fl = 0\n # Branch table to handle instruction codes\n self.branchtable = {}\n self.branchtable[HLT] = self.halt\n self.branchtable[LDI] = self.ldi\n self.branchtable[PRN] = self.prn\n self.branchtable[PUS] = self.push\n self.branchtable[POP] = self.pop\n \n # Jump table that sets program counter\n self.jumptable = {}\n self.jumptable[CAL] = self.call\n self.jumptable[RET] = self.ret\n self.jumptable[JMP] = self.jmp\n self.jumptable[JEQ] = self.jeq\n self.jumptable[JNE] = self.jne","def test_new():\n from qiskit import BasicAer\n from qiskit.aqua.algorithms import Grover\n from qiskit.aqua.components.oracles import LogicalExpressionOracle\n\n expr = \"your logical expression goes here\"\n algorithm = Grover(LogicalExpressionOracle(expr))\n backend = BasicAer.get_backend('qasm_simulator')\n result = algorithm.run(backend, seed=101110)\n print(result)","def init_analysis(session, args):\r\n path = os.path.join(session.abs_path,'{}_{}_{}_{}'.format(args[12],args[13],args[14],args[15]))\r\n session.case = Case(path, session.method)\r\n case = session.case\r\n if args[0].split('#')[0]=='R':\r\n args[7]= -args[7]\r\n\r\n case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[6], args[7]))\r\n\r\n case.file_U.set_field('boundaryField',\r\n {'inlet': {'type': 'freestream',\r\n 'freestreamValue': 'uniform ({} {} 0)'.format(args[6], args[7])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'fixedValue', 'value':'uniform (0 0 0)'},\r\n 'extrados': {'type': 'fixedValue', 'value':'uniform (0 0 0)'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\r\n case.file_p.set_field('boundaryField',\r\n {'inlet': {'type': 'freestreamPressure'},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\r\n session.case.file_T.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[11])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'slip'},\r\n 'extrados': {'type': 'slip'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n \"\"\"\r\n\r\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[6], args[7]))\r\n session.case.file_U.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue',\r\n 'value': 'uniform ({} {} 0)'.format(args[6], args[7])},\r\n 'outlet': {'type': 'inletOutlet','inletValue':'uniform ({} {} 0)'.format(args[6], args[7]),\r\n 'value':'uniform ({} {} 0)'.format(args[6], args[7])},\r\n 'intrados': {'type': 'noSlip'},\r\n 'extrados': {'type': 'noSlip'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\r\n session.case.file_p.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\r\n session.case.file_T.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue','value':'uniform {}'.format(args[11])},\r\n 'outlet': {'type': 'inletOutlet','inletValue':'uniform {}'.format(args[11]),'value':'uniform {}'.format(args[11])},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n \"\"\"\r\n \"\"\"\r\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[7], args[6]))\r\n session.case.file_U.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue',\r\n 'value': 'uniform ({} {} 0)'.format(args[7], args[6])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'noSlip'},\r\n 'extrados': {'type': 'noSlip'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\r\n session.case.file_p.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\r\n session.case.file_T.set_field('boundaryField',\r\n {'inlet': {'type': 'zeroGradient'},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n \"\"\"\r\n \"\"\"\r\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[5], args[6]))\r\n session.case.file_U.set_field('boundaryField',\r\n {'inlet': {'type': 'fixedValue', 'value': 'uniform ({} {} 0)'.format(args[5], args[6])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'noSlip'},\r\n 'extrados': {'type': 'noSlip'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_p.set_field('boundaryField', {'inlet': {'type': 'fixedValue','value':'uniform {}'.format(args[9])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[9]))\r\n\r\n session.case.file_T.set_field('boundaryField', {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\r\n 'outlet': {'type': 'zeroGradient'},\r\n 'intrados': {'type': 'zeroGradient'},\r\n 'extrados': {'type': 'zeroGradient'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n session.case.file_T.set_field('internalField','uniform 300')\r\n\r\n session.case.file_nut.set_field('boundaryField', {'inlet':{'type':'calculated', 'value':'uniform 0'},\r\n 'outlet':{'type':'calculated', 'value':'uniform 0'},\r\n 'intrados': {'type': 'nutkWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 0'},\r\n 'extrados': {'type': 'nutkWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 0'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_k.set_field('internalField', 'uniform 1')\r\n session.case.file_k.set_field('boundaryField', {\r\n 'inlet': {'type': 'turbulentIntensityKineticEnergyInlet', 'intensity': '0.05', 'value': 'uniform 1'},\r\n 'outlet': {'type': 'inletOutlet', 'inletValue': 'uniform 1', 'value': 'uniform 1'},\r\n 'intrados': {'type': 'kqRWallFunction','value':'uniform 1'},\r\n 'extrados': {'type': 'kqRWallFunction','value':'uniform 1'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n\r\n session.case.file_epsilon.set_field('boundaryField', {'inlet': {'type': 'turbulentMixingLengthDissipationRateInlet', 'mixingLength': '0.005', 'value': 'uniform 200'},\r\n 'outlet': {'type': 'inletOutlet', 'inletValue': 'uniform 200', 'value': 'uniform 200'},\r\n 'intrados': {'type': 'epsilonWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 200'},\r\n 'extrados': {'type': 'epsilonWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 200'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n session.case.file_epsilon.set_field('internalField', 'uniform 200')\r\n\r\n session.case.file_alphat.set_field('boundaryField', {'inlet':{'type':'calculated', 'value':'uniform 0'},\r\n 'outlet':{'type':'calculated', 'value':'uniform 0'},\r\n 'intrados': {'type': 'compressible::alphatWallFunction', 'Prt':'0.85', 'value':'uniform 0'},\r\n 'extrados': {'type': 'compressible::alphatWallFunction', 'Prt':'0.85', 'value':'uniform 0'},\r\n 'top_down': {'type': 'empty'},\r\n 'cyclic_in_1': {'type': 'cyclic'},\r\n 'cyclic_in_2': {'type': 'cyclic'},\r\n 'cyclic_out_1': {'type': 'cyclic'},\r\n 'cyclic_out_2': {'type': 'cyclic'}})\r\n \"\"\"\r\n session.case.file_controlDict.set_field('endTime', '10000')\r\n session.case.file_controlDict.set_field('startFrom', 'latestTime')\r\n session.case.file_controlDict.set_field('functions', {\"#includeFunc\":\"MachNo\"})\r\n session.case.file_turbulenceProperties.set_field('simulationType', 'laminar')\r\n session.case.interacting(100)\r\n sim = session.case.simulation(\"open40\") # Build files\r\n sim.limit_write = 50\r\n sim.block_mesh(string=write_block_mesh(args[1], args[2], args[3], args[4], args[5], session.mesh))\r\n sim.check_mesh()\r\n\r\n result_dict={\"T\": 0, \"p\":0, \"Theta\":0, \"z\":0, \"profile\":args[14]}\r\n\r\n def _function(container, args):\r\n current_time = container['current_time']\r\n if float(current_time)>=0.000015:\r\n print('Parsing results')\r\n sim.foamToVTK()\r\n results = sim.get_last_results('outlet')\r\n result_U = results.GetCellData('U')\r\n result_p = results.GetCellData('p')\r\n result_T = results.GetCellData('T')\r\n theta = 0.0\r\n z = 0.0\r\n p=0.0\r\n t=0.0\r\n U_length = len(result_U)\r\n p_length = len(result_p)\r\n t_length = len(result_T)\r\n for i,j,k in zip(result_p, result_T, result_U):\r\n p+= float(i[0])/p_length\r\n t+= float(j[0])/t_length\r\n theta += float(k[1])/U_length\r\n z += float(k[0])/U_length\r\n\r\n args[\"T\"] = t\r\n args[\"p\"] = p\r\n args[\"Theta\"] = theta\r\n args[\"z\"] = z\r\n return True\r\n return False\r\n \r\n #sim.run(_function, result_dict)\r\n #result_dict = {'T': 195.38959999999997, 'z': 429.3120571428572, 'p': 74001.90285714286, 'Theta': -207.19442857142855, 'profile': 0}\r\n\r\n print('Sending results')\r\n if args[0].split('#')[0]=='R':\r\n result_dict['Theta']= -result_dict['Theta']\r\n\r\n session.socket_design.send({'new_data':result_dict})","def create_and_run():\n\n app = App()\n app.run()","def init():\n safe_call(backend.get().af_init())","def __init__(self, program):\r\n self._program = program","def __init__(self, *tocs, **kwargs):\n\n from PyInstaller.config import CONF\n Target.__init__(self)\n name = kwargs.get('name', None)\n cipher = kwargs.get('cipher', None)\n self.toc = TOC()\n # If available, use code objects directly from ModuleGraph to speed up PyInstaller.\n self.code_dict = {}\n for t in tocs:\n self.toc.extend(t)\n self.code_dict.update(getattr(t, '_code_cache', {}))\n\n self.name = name\n if name is None:\n self.name = os.path.splitext(self.tocfilename)[0] + '.pyz'\n # PyInstaller bootstrapping modules.\n self.dependencies = get_bootstrap_modules()\n # Bundle the crypto key.\n self.cipher = cipher\n if cipher:\n key_file = ('pyimod00_crypto_key', os.path.join(CONF['workpath'], 'pyimod00_crypto_key.pyc'), 'PYMODULE')\n # Insert the key as the first module in the list. The key module contains just variables and does not depend\n # on other modules.\n self.dependencies.insert(0, key_file)\n # Compile the top-level modules so that they end up in the CArchive and can be imported by the bootstrap script.\n self.dependencies = misc.compile_py_files(self.dependencies, CONF['workpath'])\n self.__postinit__()","def __init__(self):\n import arcgisscripting as ARC\n self.GP = ARC.create(9.3)\n self.params = self.GP.getparameterinfo()","def __init__(self, *args):\n _snap.TAscFltV_swiginit(self, _snap.new_TAscFltV(*args))","def __init__(self, fips_codes, horiz_pos, vert_pos, population, risk):\n self._fips_codes = fips_codes\n self._horiz_center = horiz_pos\n self._vert_center = vert_pos\n self._total_population = population\n self._averaged_risk = risk","def __init__(self, fips_codes, horiz_pos, vert_pos, population, risk):\n self._fips_codes = fips_codes\n self._horiz_center = horiz_pos\n self._vert_center = vert_pos\n self._total_population = population\n self._averaged_risk = risk","def __init__(self):\n _hypre.HypreIdentity_swiginit(self, _hypre.new_HypreIdentity())","def main():\n model = Calculator()","def init():\n try:\n compile_contract(\n \"fishcake\", f\"Fishcake(sp.address('{pub_key_hash}'),{default_supply})\")\n fishcake_addr = deploy(\"fishcake\")\n print(\"\\n\")\n compile_contract(\n \"fishcakeBox\", f\"FishcakeBox({default_redeem_amt}, sp.address('{fishcake_addr}'))\")\n fishcake_box_addr = deploy(\"fishcakeBox\")\n setup(fishcake_addr, fishcake_box_addr)\n print(\"\\n\\n[!] Details :\\n\")\n print(f\"-- Fishcake Token Address : {fishcake_addr}\")\n print(f\"-- Fishcake Box Address : {fishcake_box_addr}\")\n except Exception as e:\n print(\"Failed to originate Contracts : \", e)","def create_factory(cls, *args):\n raise NotImplementedError","def __init__(self, *args):\n _snap.TSFlt_swiginit(self, _snap.new_TSFlt(*args))","def __init__(self, source_code, bytecode=None, address=None, abi=SOPHIA, client=None):\n if client is None:\n client = epoch.EpochClient()\n self.client = client\n self.abi = abi\n self.source_code = source_code\n self.bytecode = bytecode\n self.address = address\n if self.bytecode is None:\n self.bytecode = self.compile(self.source_code)","def __init__(self):\n BasicEngine.__init__(self)\n self._active_qubits = 0\n self._num_rotations = 0\n self._rotations = []","def my_main() -> None: # pragma: no cover\n universe = AutomataUniverse(SQUARE_GRID_NEIGHBORS, [2,3], [3])\n instance = AutomataTransforms(universe)\n assert isinstance(instance, AutomataTransforms)\n # # _is_rot_mat_test(instance)\n # # _rotations_check(instance)\n # # _prime_cells_check(instance)\n # _check_transform_test(instance)\n # # _hashable_transform_test(instance)\n # _duplicate_test(instance)\n # _collision_test(instance)\n # _end_cycle_test(instance)\n # _add_transform_test(instance)\n # instance.generate_combination_transforms()\n\n # # _matrix_rotate_test(instance)\n # # _duplicate_test(instance) # test again after transform(s) added\n # # _collision_test(instance) # test again after transform(s) added «also refactoring»\n # instance.dbg_report_instance() # DEBUG","def make(self, *args, **kwargs):\r\n mode = kwargs.pop('mode', theano.compile.mode.get_default_mode())\r\n rval = self.make_no_init(mode)\r\n if hasattr(rval, 'initialize'):\r\n rval.initialize(*args, **kwargs)\r\n return rval","def __init__(self):\n\n super().__init__(BabelCLIHandlersEnum.COMPILE)","def __init__(self, base=None, mvtype=None, fct=False, blade_rep=False):\n\n def make_scalar(self, base): # make a scalar (grade 0)\n if isinstance(base, str):\n if self.fct:\n self.obj = Function(base)(*MV.coords) * MV.ONE\n else:\n self.obj = make_coef(self, base) * MV.ONE\n else:\n self.obj = base * MV.ONE\n self.igrade = 0\n self.blade_rep = True\n return self\n\n def make_vector(self, base): # make a vector (grade 1)\n if isinstance(base, str):\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=1, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[1]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=1, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=1, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[1]))))\n else:\n result = S.Zero\n for (coef, base) in zip(base, MV.blades[1]):\n result += coef * base\n self.obj = result\n self.igrade = 1\n self.blade_rep = True\n return self\n\n def make_basisvector(self, base):\n raise NotImplementedError(\"Don't know how to compute basis vectors of class %\" % self.__class__)\n\n def make_basisbivector(self, base):\n raise NotImplementedError(\"Don't know how to compute basis bivectors of class %\" % self.__class__)\n\n def make_grade(self, base): # if base is 'A,n' then make a grade n multivector\n if isinstance(base, str):\n base_lst = base.split(',')\n base = base_lst[0]\n n = int(base_lst[1])\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=n, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[n]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=n, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=n, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[n]))))\n else:\n raise TypeError('Cannot make_grade for base = %s' % base)\n self.igrade = n\n self.blade_rep = True\n return self\n\n def make_grade2(self, base): # grade 2 multivector\n if isinstance(base, str):\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=2, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[2]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=2, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=2, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[2]))))\n else:\n raise TypeError('!!!!Cannot make_grade2 for base = ' + str(base) + '!!!!\\n')\n self.igrade = 2\n self.blade_rep = True\n return self\n\n def make_pseudo(self, base): # multivector of grade MV.dim\n if isinstance(base, str):\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=MV.dim, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[MV.dim]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=MV.dim, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=MV.dim, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[MV.dim]))))\n else:\n raise TypeError('!!!!Cannot make_pseudo for base = ' + str(base) + '!!!!\\n')\n self.igrade = MV.dim\n self.blade_rep = True\n return self\n\n def make_spinor(self, base): # multivector with all even grades\n if isinstance(base, str):\n if self.fct:\n self.obj = Function(base)(*MV.coords) * MV.ONE\n else:\n self.obj = Symbol(base) * MV.ONE\n for rank in range(2, MV.dim1, 2):\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=rank, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\n else:\n raise TypeError('Cannot make_mv for base = %s' % base)\n self.igrade = -1\n self.blade_rep = True\n return self\n\n def make_mv(self, base):\n if isinstance(base, str):\n if self.fct:\n self.obj = Function(base)(*MV.coords) * MV.ONE\n else:\n self.obj = Symbol(base) * MV.ONE\n for rank in range(1, MV.dim1):\n if self.fct:\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\n fct_lst = fct_sym_array(base_lst, MV.coords)\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\n else:\n if MV.coords is not None:\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\n else:\n base_lst = str_combinations(base, MV.subscripts, rank=rank, mode='__')\n fct_lst = fct_sym_array(base_lst, None)\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\n else:\n raise TypeError('!!!!Cannot make_mv for base = ' + str(base) + '!!!!\\n')\n self.igrade = -1\n self.blade_rep = True\n return self\n\n MVtypes = {'scalar': make_scalar,\n 'vector': make_vector,\n 'basisvector': make_basisvector,\n 'basisbivector': make_basisbivector,\n 'grade': make_grade,\n 'grade2': make_grade2,\n 'bivector': make_grade2,\n 'pseudo': make_pseudo,\n 'spinor': make_spinor,\n 'mv': make_mv}\n\n self.fct = fct\n self.is_base = False\n self.is_grad = False\n self.print_blades = MV.print_blades\n self.fmt = 1\n\n if mvtype is None:\n if base in (None, S.Zero): # Default is zero multivector\n self.blade_rep = True\n self.obj = S.Zero\n self.igrade = 0\n elif isinstance(base, str): # Base or blade basis multivector\n self.is_base = True\n if '*' in base:\n self.blade_rep = False\n self.igrade = -1\n else:\n if '^' in base:\n self.blade_rep = True\n self.igrade = base.count('^') + 1\n else:\n self.blade_rep = blade_rep\n self.igrade = 1\n self.obj = Symbol(base, commutative=False)\n elif isinstance(base, MV): # Copy constructor\n self.blade_rep = base.blade_rep\n self.obj = base.obj\n self.igrade = base.igrade\n self.fct = base.fct\n self.is_base = base.is_base\n self.is_grad = base.is_grad\n elif isinstance(base, (Expr, Symbol)): # Gets properties of multivector from Expr\n if base.is_commutative:\n self.obj = base * MV.ONE\n self.blade_rep = True\n self.igrade = 0\n else:\n if isinstance(base, (Add, Mul)): # Complex expression\n MV.characterize_expression(self, base)\n elif isinstance(base, Symbol):\n if not base.is_commutative:\n if base == MV.ONE:\n self.obj = base\n self.blade_rep = True\n self.igrade = 0\n elif base in MV.blades_flat: # basis blade\n self.obj = base\n self.blade_rep = True\n self.igrade = MV.blade_grades[base]\n elif base in MV.bases_flat: # basis base\n self.obj = base\n self.blade_rep = False\n self.igrade = -1\n else:\n raise ValueError('MV(' + str(base) + ') is not allowed in constructor\\n' +\n 'non-commutative argument is not a base\\n')\n else: # scalar sympy symbol\n self.obj = base * MV.ONE\n self.igrade = 0\n self.blade_rep = True\n elif isinstance(base, Number):\n self.obj = base * MV.ONE\n self.igrade = 0\n self.blade_rep = True\n else: # Preconfigured multivector types\n MVtypes[mvtype](self, base)"],"string":"[\n \"def __create_execution_engine(self):\\n target = self.binding.Target.from_default_triple()\\n target_machine = target.create_target_machine()\\n backing_mod = binding.parse_assembly(\\\"\\\")\\n engine = binding.create_mcjit_compiler(backing_mod, target_machine)\\n self.engine = engine\",\n \"def Program(self):\\n funcs = list()\\n while self.currtok[1].name in {\\\"INT\\\", \\\"FLOAT\\\", \\\"BOOLEAN\\\"}:\\n func = self.FunctionDef()\\n funcs.append(func)\\n return Program(funcs)\",\n \"def factory(args, cxn, log):\\n name = args['assembler'].lower()\\n assembler = ASSEMBLERS[name]\\n return assembler(args, cxn, log)\",\n \"def __init__ ( self , syms , defn ):\\n\\n self.logic = cognitiveDefiner.convertDefinition(syms,defn)\",\n \"def __init__(self, *args):\\n this = _ida_hexrays.new_casm_t(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def __init__(self, fen): \\n self.d_engine = Stockfish()\\n self.setFen(fen)\",\n \"def __init__(self):\\n self._lib_vscf_ecc = VscfEcc()\\n self._c_impl = None\\n self._ctx = None\\n self.ctx = self._lib_vscf_ecc.vscf_ecc_new()\",\n \"def create_engine(self, input):\\n return\",\n \"def create_execution_engine():\\n # Create a target machine representing the host\\n target = llvm.Target.from_default_triple()\\n target_machine = target.create_target_machine()\\n # And an execution engine with an empty backing module\\n backing_mod = llvm.parse_assembly(\\\"\\\")\\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\\n return engine\",\n \"def create_execution_engine():\\n # Create a target machine representing the host\\n target = llvm.Target.from_default_triple()\\n target_machine = target.create_target_machine()\\n # And an execution engine with an empty backing module\\n backing_mod = llvm.parse_assembly(\\\"\\\")\\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\\n return engine\",\n \"def create_execution_engine():\\n # Create a target machine representing the host\\n target = llvm.Target.from_default_triple()\\n target_machine = target.create_target_machine()\\n # And an execution engine with an empty backing module\\n backing_mod = llvm.parse_assembly(\\\"\\\")\\n engine = llvm.create_mcjit_compiler(backing_mod, target_machine)\\n return engine\",\n \"def __init__(self, engine: str = \\\"sfdp\\\"):\\n self.engine = engine\",\n \"def __init__(self, regex, fixed_slice):\\n self._regex = regex\\n self._fixed_slice = fixed_slice\\n self._dfa = fte.dfa.from_regex(self._regex, self._fixed_slice)\\n self._encrypter = fte.encrypter.Encrypter()\",\n \"def _new_instance(self):\\n return self.__class__(self._fmodule)\",\n \"def _new_instance(self):\\n return self.__class__(self._fmodule)\",\n \"def instantiate_algorithm(args):\\n if args.algorithm == 'xor_encoding':\\n return XorEncoding(block_size=args.block_size, intensity=args.intensity)\\n\\n raise RuntimeError('Algorithm type not detected')\",\n \"def parse_main(self):\\n self.decompiler = self.createDecompiler(self.compilerEnv)\\n self.nf = IRFactory(self)\\n self.currentScriptOrFn = self.nf.createScript()\\n sourceStartOffset = self.decompiler.getCurrentOffset();\\n self.encodedSource = None;\\n self.decompiler.addToken(Token.SCRIPT);\\n \\n self.currentFlaggedToken = Token.EOF;\\n self.syntaxErrorCount = 0;\\n\\n baseLineno = self.ts.getLineno(); #// line number where source starts\\n\\n #/* so we have something to add nodes to until\\n # * we've collected all the source */\\n pn = self.nf.createLeaf(Token.BLOCK);\\n try:\\n while(True):\\n tt = self.peekToken()\\n if (tt <= Token.EOF):\\n break\\n n = None\\n \\n if tt == Token.FUNCTION:\\n self.consumeToken()\\n try: \\n n = self.function( \\\\\\n (FunctionNode.FUNCTION_EXPRESSION,\\\\\\n FunctionNode.FUNCTION_STATEMENT)[self.calledByCompileFunction])\\n except ParserException:\\n raise NotImplementedError()# should this have to raise?\\n break;\\n #raise NotImplementedError <- I think this is now implemented - TW\\n \\n else:\\n n = self.statement()\\n \\n self.nf.addChildToBack(pn, n)\\n \\n \\n except RuntimeError:\\n # Was StackOverflowError\\n raise\\n # TODO: exception handling\\n \\n if (self.syntaxErrorCount != 0) :\\n msg = str(self.syntaxErrorCount)\\n #msg = ScriptRuntime.getMessage1(\\\"msg.got.syntax.errors\\\", msg);\\n #throw errorReporter.runtimeError(msg, sourceURI, baseLineno,\\n # null, 0);\\n\\n self.currentScriptOrFn.setSourceName(self.sourceURI);\\n self.currentScriptOrFn.setBaseLineno(baseLineno);\\n self.currentScriptOrFn.setEndLineno(self.ts.getLineno());\\n\\n sourceEndOffset = self.decompiler.getCurrentOffset();\\n self.currentScriptOrFn.setEncodedSourceBounds(sourceStartOffset,\\n sourceEndOffset);\\n\\n self.nf.initScript(self.currentScriptOrFn, pn);\\n\\n if (self.compilerEnv.isGeneratingSource()):\\n self.encodedSource = self.decompiler.getEncodedSource();\\n \\n del self.decompiler# comment was //\\\"It helps GC\\\" \\n # - can't do any harm on CPython either\\n\\n return self.currentScriptOrFn;\",\n \"def __init__(self, foretext=\\\"::>>> \\\"):\\n self.keywords = []\\n\\n # self.functions is a *dict* with\\n # key: keyword, value: (function, args, kwargs)\\n self.functions = {}\\n\\n self.foretext = foretext\\n self.input_asked = False\\n self.retrieved_input = None\",\n \"def New(*args, **kargs):\\n obj = itkIsoDataThresholdCalculatorHFSS.__New_orig__()\\n import itkTemplate\\n itkTemplate.New(obj, *args, **kargs)\\n return obj\",\n \"def run(_):\\n fscad = types.ModuleType(\\\"fscad.fscad\\\")\\n fscad.__path__ = [os.path.dirname(os.path.realpath(__file__))]\\n sys.modules['fscad.fscad'] = fscad\\n\\n for key in __all__:\\n fscad.__setattr__(key, globals()[key])\",\n \"def main():\\n create_fc(\\\"layers/WAPR.lyr\\\",\\n \\\"TravelerInfo.gdb/HighwayAlerts\\\")\",\n \"def __init__(self, *args):\\n _BRepAlgo.BRepAlgo_Fuse_swiginit(self,_BRepAlgo.new_BRepAlgo_Fuse(*args))\",\n \"def build(self, name='impala'):\\n program = lp.Program(name=name)\\n\\n with program.group('replay'):\\n queue = program.add_node(lp.ReverbNode(self.queue))\\n\\n with program.group('counter'):\\n counter = program.add_node(lp.CourierNode(self.counter))\\n\\n with program.group('learner'):\\n learner = program.add_node(\\n lp.CourierNode(self.learner, queue, counter))\\n\\n with program.group('evaluator'):\\n program.add_node(lp.CourierNode(self.evaluator, learner, counter))\\n\\n with program.group('cacher'):\\n cacher = program.add_node(\\n lp.CacherNode(learner, refresh_interval_ms=2000, stale_after_ms=4000))\\n\\n with program.group('actor'):\\n for _ in range(self._num_actors):\\n program.add_node(lp.CourierNode(self.actor, queue, cacher, counter))\\n\\n return program\",\n \"def New(*args, **kargs):\\n obj = itkIsoDataThresholdCalculatorHFUS.__New_orig__()\\n import itkTemplate\\n itkTemplate.New(obj, *args, **kargs)\\n return obj\",\n \"def __init__(self, pele_exec, pele_src, pele_license,\\n ploprottemp_src, schrodinger_src,\\n charge_method='am1bcc', solvent='OBC',\\n opls_nonbonding=False, opls_bonds_angles=False,\\n n_proc=1, forcefield_name=None, forcefield=None):\\n self.pele_exec = pele_exec\\n self.pele_src = pele_src\\n self.pele_license = pele_license\\n self.charge_method = charge_method\\n self.solvent = solvent\\n self.opls_nonbonding = opls_nonbonding\\n self.opls_bonds_angles = opls_bonds_angles\\n self.ploprottemp_src = ploprottemp_src\\n self.schrodinger_src = schrodinger_src\\n self._n_proc = n_proc\\n self.forcefield_name = forcefield_name\\n self.forcefield = forcefield\\n self._results = dict()\\n\\n # Deactivate peleffy output\\n from peleffy.utils import Logger\\n logger = Logger()\\n logger.set_level('WARNING')\\n\\n # Supress OpenForceField toolkit warnings\\n import logging\\n logging.getLogger().setLevel(logging.ERROR)\",\n \"def init(config=config, version='crystal'):\\n setup_logging()\\n configure_for_version(version, config)\\n get_constants(config=config)\\n return main(config=config)\",\n \"def disassembler_factory():\\n \\n ir_id, ir_cls, dis_cls = find_current_arch()\\n \\n class disassembler(dis_cls, ir_cls): # disassembler (host) class must be left-most.\\n def __init__(self, ir_id):\\n self.ir_id = ir_id\\n dis_cls.__init__(self)\\n ir_cls.__init__(self)\\n return\\n \\n dis = disassembler(ir_id)\\n \\n return dis\",\n \"def __init__(self, caldb_name, irf_name, verbose=False):\\n\\n self.caldb_path = os.environ['CALDB']\\n self.caldb_name = caldb_name\\n self.irf = irf_name\\n self.verbose = verbose\\n\\n self.am_ok = True\\n\\n self._aeff = dict()\\n self._psf = dict()\\n self._edips = dict()\\n\\n self._check_available_irfs()\\n\\n self.input_irf_file_name = '{path:s}/data/cta/{caldb:s}/bcf/{irf:s}/irf_file.fits'.format(path=self.caldb_path,\\n caldb=self.caldb_name,\\n irf=irf_name)\",\n \"def __init__(self):\\n self.reg = [0] * 8\\n self.ram = [0] * 256\\n self.pc = 0\\n self.sp = 7\\n\\n # state (running)\\n self.running = True\\n\\n # REPL (FETCH, DECODE, EXECUTE)\\n self.op_pc = False\\n\\n self.branch_table = {}\\n self.branch_table[PRN] = self.prn\\n self.branch_table[LDI] = self.ldi\\n self.branch_table[CMP] = self.cmp\\n self.branch_table[JMP] = self.jmp\\n self.branch_table[JEQ] = self.jeq\\n self.branch_table[JNE] = self.jne\",\n \"def initialize_implementation(self):\\n # call overridable methods to get number of inputs/outputs\\n n_inputs = self.get_number_of_input_connections()\\n n_outputs = self.get_number_of_output_ports()\\n\\n # call overrides to get implementation for teca execution\\n # phase implementations\\n import teca_py\\n self.impl = teca_py.teca_programmable_algorithm.New()\\n self.impl.set_number_of_input_connections(n_inputs)\\n self.impl.set_number_of_output_ports(n_outputs)\\n self.impl.set_name(self.__class__.__name__)\\n self.impl.set_report_callback(self.get_report_callback())\\n self.impl.set_request_callback(self.get_request_callback())\\n self.impl.set_execute_callback(self.get_execute_callback())\",\n \"def main() -> None:\\n version_path = './src/fqe/_version.py'\\n\\n __version__ = version_number(version_path)\\n\\n if __version__ is None:\\n raise ValueError('Version information not found in ' + version_path)\\n\\n long_description = ('OpenFermion-FQE\\\\n' +\\n '===============\\\\n')\\n stream = io.open('README.md', encoding='utf-8')\\n stream.readline()\\n long_description += stream.read()\\n\\n requirements_buffer = open('requirements.txt').readlines()\\n requirements = [r.strip() for r in requirements_buffer]\\n\\n # C code extension\\n config_vars = get_config_vars()\\n config_vars[\\\"EXT_SUFFIX\\\"] = '.' + config_vars[\\\"EXT_SUFFIX\\\"].split('.')[-1]\\n libdir = os.path.join(\\\"src\\\", \\\"fqe\\\", \\\"lib\\\")\\n cfiles = [\\n \\\"macros.c\\\",\\n \\\"mylapack.c\\\",\\n \\\"fci_graph.c\\\",\\n \\\"fqe_data.c\\\",\\n \\\"cirq_utils.c\\\",\\n \\\"wick.c\\\",\\n \\\"bitstring.c\\\",\\n \\\"binom.c\\\",\\n ]\\n srcs = [os.path.join(libdir, cf) for cf in cfiles]\\n libraries = []\\n extensions = [\\n Extension(\\\"fqe.lib.libfqe\\\",\\n srcs,\\n include_dirs=[libdir],\\n library_dirs=[libdir],\\n libraries=libraries,\\n language='c')\\n ]\\n\\n cythonfiles = [\\\"_fqe_data.pyx\\\"]\\n srcs = [os.path.join(libdir, cf) for cf in cythonfiles]\\n extensions.append(Extension(\\\"fqe.lib.fqe_data\\\", srcs, language='c'))\\n\\n setup(name='fqe',\\n version=__version__,\\n author='The OpenFermion FQE Developers',\\n author_email='help@openfermion.org',\\n url='http://www.openfermion.org',\\n description='OpenFermion Fermionic Quantum Emulator',\\n ext_modules=cythonize(extensions,\\n compiler_directives={'language_level': \\\"3\\\"}),\\n long_description=long_description,\\n long_description_content_type=\\\"text/markdown\\\",\\n install_requires=requirements,\\n license='Apache 2',\\n packages=find_packages(where='src'),\\n package_dir={'': 'src'},\\n cmdclass={'build_ext': CustomBuildOptions})\",\n \"def create(cls, name, defn, state=None):\\n prog = cls(name, False)\\n prog.raw_data = defn\\n prog.state = state or {}\\n prog.mk_data()\\n\\n return prog\",\n \"def init(self):\\n self._frame_idx = 0\\n if self.initialization is not None:\\n del self.initialization\\n self.initialization = None\\n\\n if self.config.initialization is not None:\\n self.initialization = INITIALIZATION.load(self.config.initialization, **self.__kwargs)\\n self.initialization.init()\\n\\n if self.preprocessing is not None:\\n del self.preprocessing\\n self.preprocessing = None\\n\\n if self.config.preprocessing is not None:\\n self.preprocessing = Preprocessing(self.config.preprocessing, **self.__kwargs)\\n\\n if self.odometry is None:\\n assert self.config.odometry is not None\\n self.odometry = ODOMETRY.load(self.config.odometry, **self.__kwargs)\\n\\n assert self.odometry is not None\\n self.odometry.init()\\n if self.loop_closure is None and self.config.loop_closure is not None:\\n self.loop_closure = LOOP_CLOSURE.load(self.config.loop_closure, **self.__kwargs)\\n if self.loop_closure is not None:\\n self.loop_closure.init()\\n if self.config.backend is not None:\\n self.backend = BACKEND.load(self.config.backend, **self.__kwargs)\\n if self.backend is not None:\\n self.backend.init()\\n else:\\n logging.warning(\\\"[SLAMAlgorithm]Defined a Loop Closure Algorithm Without a Backend\\\")\",\n \"def make(self, *args, **kwargs):\\r\\n self.make_module_instance(args,kwargs)\\r\\n\\r\\n mode = kwargs.pop('mode', theano.compile.mode.get_default_mode())\\r\\n rval = self.make_no_init(mode)\\r\\n if hasattr(rval, 'initialize'):\\r\\n rval.initialize(*args, **kwargs)\\r\\n return rval\",\n \"def __init__(self):\\r\\n\\t\\tself.introducer()\\r\\n\\t\\tif self.code_mode == \\\"1\\\":\\r\\n\\t\\t\\tif self.input_mode == \\\"1\\\":\\r\\n\\t\\t\\t\\tself.encrypt_message()\\r\\n\\t\\t\\telse:\\r\\n\\t\\t\\t\\tself.encrypt_text_file()\\r\\n\\t\\t\\t\\t#print(\\\"work in progress\\\")\\r\\n\\t\\telif self.code_mode == \\\"2\\\":\\r\\n\\t\\t\\tif self.input_mode == \\\"1\\\":\\r\\n\\t\\t\\t\\tself.decrypt_message()\\r\\n\\t\\t\\telse:\\r\\n\\t\\t\\t\\tself.decrypt_text_file()\\r\\n\\t\\telse:\\r\\n\\t\\t\\tif self.input_mode == \\\"1\\\":\\r\\n\\t\\t\\t\\tself.hack_message()\\r\\n\\t\\t\\telse:\\r\\n\\t\\t\\t\\tself.hack_text_file()\",\n \"def main():\\n get_engine(onnx_file_path, engine_file_path)\",\n \"def main() -> None:\\n # The first thing to do is get the lines of the PyFlex file we are given.\\n parser = Parser(filename=sys.argv[1])\\n parsed_data = parser.ParseFile()\\n\\n # Upon retrieving the Parsed Data, assign the parsed data to the\\n # Symbol Table.\\n SymbolTable.RULESET = parsed_data['ruleset']\\n SymbolTable.INSTRUCTIONS = parsed_data['instructions']\\n SymbolTable.CODE = parsed_data['code']\\n # SymbolTable.PrintTable()\\n\\n # Using the Generator backend, we can build the generated script\\n generator = Generator()\\n generator.GenerateNewScript()\\n\\n autopep8.fix_file(filename=generator.file_main)\\n\\n print(\\\"Generated Script can be found in {}\\\".format(generator.file_main))\",\n \"def __init__(self, *args):\\n _hypre.HypreTriSolve_swiginit(self, _hypre.new_HypreTriSolve(*args))\",\n \"def __init__(self, vplprog):\\n\\n self.validated = False\\n self.consts = dict()\\n self.functions = []\\n funcnames = set()\\n\\n for child in vplprog.children:\\n func = FunctionNode(child, self.consts)\\n\\n if func.name in funcnames:\\n raise VPLSyntaxError(\\\"Duplicate function name '{0}'\\\".format(func.name))\\n\\n funcnames.add(func.name)\\n self.functions.append(func)\",\n \"def __init__(self):\\n self.label = \\\"TOC Fuzzification\\\"\\n self.description = \\\"This fuzzification method utilized the symbolization of the input raster that has been applied in the map document table of contects (TOC). The symbolization in the TOC defines the number of classes and this tool rescales those classes (1...N) to the range [0,1] by (C - 1)/(N-1) where C is the class value and N is the number of classes.\\\"\\n self.canRunInBackground = False\\n self.category = \\\"Fuzzy Logic\\\\\\\\Fuzzy Membership\\\"\",\n \"def __init__(self, *args):\\n _snap.TAscFlt_swiginit(self, _snap.new_TAscFlt(*args))\",\n \"def __init__(self):\\n self.libpath = os.sep.join(os.path.abspath(__file__).split(os.sep)[:-1])\\n sys.path.append(self.libpath)\\n libpath2 = os.sep.join(self.libpath.split(os.sep)[:-1])\\n sys.path.append(libpath2)\\n # Initialize TCMetaSchema with correct libpath\\n TCMetaSchema(self.libpath)\\n self.args, self.unknown = IceteaManager._parse_arguments()\\n # If called with --clean, clean up logs.\\n if self.args.clean:\\n _cleanlogs(silent=self.args.silent, log_location=self.args.log)\\n\\n LogManager.init_base_logging(self.args.log, verbose=self.args.verbose,\\n silent=self.args.silent, color=self.args.color,\\n no_file=(self.args.list or self.args.listsuites),\\n truncate=not self.args.disable_log_truncate)\\n\\n self.logger = LogManager.get_logger(\\\"icetea\\\")\\n self.pluginmanager = None\\n self.resourceprovider = ResourceProvider(self.args)\\n self._init_pluginmanager()\\n self.resourceprovider.set_pluginmanager(self.pluginmanager)\",\n \"def base_calc(fresh_aiida_env, vasp_code):\\n from aiida_vasp.calcs.base import VaspCalcBase\\n manager = get_manager()\\n runner = manager.get_runner()\\n inputs = AttributeDict()\\n\\n metadata = AttributeDict({'options': {'resources': {'num_machines': 1, 'num_mpiprocs_per_machine': 1}}})\\n\\n inputs.code = vasp_code\\n inputs.metadata = metadata\\n\\n return instantiate_process(runner, VaspCalcBase, **inputs)\",\n \"def main():\\n openff = Openff()\\n openff.console.welcome()\",\n \"def main() -> None:\\n parser = argparse.ArgumentParser(description=\\\"Flower\\\")\\n parser.add_argument(\\n \\\"--server_address\\\",\\n type=str,\\n default=DEFAULT_SERVER_ADDRESS,\\n help=f\\\"gRPC server address (default: {DEFAULT_SERVER_ADDRESS})\\\",\\n )\\n parser.add_argument(\\n \\\"--cid\\\", type=str, required=True, help=\\\"Client CID (no default)\\\"\\n )\\n parser.add_argument(\\n \\\"--log_host\\\", type=str, help=\\\"Logserver address (no default)\\\",\\n )\\n parser.add_argument(\\n \\\"--nb_clients\\\", type=int, default=10, help=\\\"Total number of clients\\\",\\n )\\n args = parser.parse_args()\\n\\n # Configure logger\\n fl.common.logger.configure(f\\\"client_{args.cid}\\\", host=args.log_host)\\n\\n # Load model and data\\n #model = cifar.load_model()\\n #model.to(DEVICE)\\n #trainset, testset = cifar.load_data()\\n\\n # Start client\\n #client = CifarClient(args.cid, model, trainset, testset, args.nb_clients)\\n #client = CifarClient(args.cid, trainset, testset, args.nb_clients)\\n client = CifarClient(args.cid,args.nb_clients)\\n fl.client.start_client(args.server_address, client)\",\n \"def __init__(self) -> None:\\n flags.reset()\\n self.flags = flags\\n self.super_memory = SuperMemory()\\n self.memory = self.super_memory.memory\\n self.super_memory.PC(\\\"0x0800\\\")\\n self._registers_list = {\\n \\\"A\\\": self.super_memory.A,\\n \\\"PSW\\\": self.flags.PSW,\\n \\\"B\\\": self.super_memory.BC,\\n \\\"C\\\": self.super_memory.BC,\\n \\\"D\\\": self.super_memory.DE,\\n \\\"E\\\": self.super_memory.DE,\\n \\\"H\\\": self.super_memory.HL,\\n \\\"L\\\": self.super_memory.HL,\\n \\\"SP\\\": self.super_memory.SP,\\n \\\"PC\\\": self.super_memory.PC,\\n \\\"M\\\": self.super_memory.M,\\n }\\n self._lookup_opcodes_dir = opcodes_lookup\\n\\n self._keywords = []\\n self._generate_keywords()\\n self._assembler = {}\\n pass\",\n \"def __init__(self):\\n self.RRTFamilySolver = RRTFamilyPathPlanner()\\n self.PRMSolver = PRMPathPlanner()\",\n \"def __init__(self, *args):\\n if self.__class__ == codegen_t:\\n _self = None\\n else:\\n _self = self\\n this = _ida_hexrays.new_codegen_t(_self, *args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def main():\\n init_latex()\",\n \"def __init__(self, *args):\\n this = _libsbml.new_FunctionDefinition(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def __init__(self, osi, fy, fu, es, esh, eps_sh, eps_ult, cf, alpha_2, cd):\\n self.osi = osi\\n self.fy = float(fy)\\n self.fu = float(fu)\\n self.es = float(es)\\n self.esh = float(esh)\\n self.eps_sh = float(eps_sh)\\n self.eps_ult = float(eps_ult)\\n self.cf = float(cf)\\n self.alpha_2 = alpha_2\\n self.cd = float(cd)\\n if osi is not None:\\n osi.n_mat += 1\\n self._tag = osi.n_mat\\n self._parameters = [self.op_type, self._tag, self.fy, self.fu, self.es, self.esh, self.eps_sh, self.eps_ult, '-CMFatigue', self.cf, self.alpha_2, self.cd]\\n if osi is None:\\n self.built = 0\\n if osi is not None:\\n self.to_process(osi)\",\n \"def create_instance(c_instance):\\n return MonoPedal(c_instance)\",\n \"def __init__(self, comm):\\n _hypre.HypreAME_swiginit(self, _hypre.new_HypreAME(comm))\",\n \"def __init__(self): \\n\\t\\n\\t # get the environment\\n\\t\\tself.env = env()\",\n \"def Build(self, mode):\\n return _gmat_py.Interpreter_Build(self, mode)\",\n \"def __init__(self):\\n\\n # Initialize the fields that all formulae have\\n self.nvars = 0\\n self.nclauses = 0\\n self.base_clauses = []\\n self.aux_clauses = []\\n\\n # Create variable meaning table\\n #\\n # Variable meaning is a table which maps boolean varibles to their\\n # semantics. The semantics of a variable are reprsented a 3-tuple.\\n # The first element of the tuple says whether the variable is an\\n # assignment, clustering, activation, or commander variable. If the\\n # variable is an assignment variable, the second element is the\\n # relevant signature and the third is the relevenant methyl. If the\\n # variable is a clustering variable, the second element is the\\n # relevant NOE and the third is the relevant signature. If it is an\\n # activation variable, then the 2nd a 3rd elements are the relevant\\n # NOEs (in arbitrary order).\\n\\n self.variable_meaning = {}\",\n \"def __init__(self, *args):\\n this = _ida_hexrays.new_cinsn_t(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def __init__(self, *args):\\n this = _ida_hexrays.new_cif_t(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def New(*args, **kargs):\\n obj = itkIsoDataThresholdCalculatorHFF.__New_orig__()\\n import itkTemplate\\n itkTemplate.New(obj, *args, **kargs)\\n return obj\",\n \"def __init__(self, hostname = '', port = 0):\\n ScriptForge.hostname = hostname\\n ScriptForge.port = port\\n # Determine main pyuno entry points\\n ScriptForge.componentcontext = self.ConnectToLOProcess(hostname, port) # com.sun.star.uno.XComponentContext\\n ScriptForge.scriptprovider = self.ScriptProvider(self.componentcontext) # ...script.provider.XScriptProvider\\n #\\n # Establish a list of the available services as a dictionary (servicename, serviceclass)\\n ScriptForge.serviceslist = dict((cls.servicename, cls) for cls in SFServices.__subclasses__())\\n ScriptForge.servicesdispatcher = None\\n #\\n # All properties and methods of the ScriptForge API are ProperCased\\n # Compute their synonyms as lowercased and camelCased names\\n ScriptForge.SetAttributeSynonyms()\\n #\\n ScriptForge.SCRIPTFORGEINITDONE = True\",\n \"def __init__(self, *args):\\n this = _ida_hexrays.new_cloop_t(*args)\\n try: self.this.append(this)\\n except: self.this = this\",\n \"def __init__(self, restart=None, ignore_bad_restart_file=False,\\n label='dftb', atoms=None, kpts=None,\\n run_manyDftb_steps=False, slako_dir='./',\\n **kwargs):\\n\\n # skf tables\\n #if 'DFTB_PREFIX' in os.environ:\\n # self.slako_dir = os.environ['DFTB_PREFIX'].rstrip('/') + '/'\\n #else:\\n # self.slako_dir = './'\\n #slako_dir = os.path.abspath(slako_dir) + '/'\\n self.slako_dir = slako_dir.rstrip('/') + '/'\\n\\n # file names\\n self.geo_fname = 'geo_in.gen'\\n\\n if run_manyDftb_steps:\\n # minimisation of molecular dynamics is run by native DFTB+\\n self.default_parameters = dict(\\n Hamiltonian_='DFTB',\\n Hamiltonian_SlaterKosterFiles_='Type2FileNames',\\n Hamiltonian_SlaterKosterFiles_Prefix=self.slako_dir,\\n Hamiltonian_SlaterKosterFiles_Separator='\\\"-\\\"',\\n Hamiltonian_SlaterKosterFiles_Suffix='\\\".skf\\\"',\\n Hamiltonian_MaxAngularMomentum_='')\\n else:\\n # using ase to get forces and energy only\\n # (single point calculation)\\n self.default_parameters = dict(\\n Hamiltonian_='DFTB',\\n Hamiltonian_SlaterKosterFiles_='Type2FileNames',\\n Hamiltonian_SlaterKosterFiles_Prefix=self.slako_dir,\\n Hamiltonian_SlaterKosterFiles_Separator='\\\"-\\\"',\\n Hamiltonian_SlaterKosterFiles_Suffix='\\\".skf\\\"',\\n Hamiltonian_MaxAngularMomentum_='')\\n\\n self.pcpot = None\\n self.lines = None # results.tag\\n\\n self.atoms = None\\n self.atoms_input = None\\n\\n self.do_forces = False\\n self.outfilename = 'dftb.out'\\n\\n FileIOCalculator.__init__(self, restart, ignore_bad_restart_file,\\n label, atoms,\\n **kwargs)\\n\\n # Determine number of spin channels\\n try:\\n entry = kwargs['Hamiltonian_SpinPolarisation']\\n spinpol = 'colinear' in entry.lower()\\n except KeyError:\\n spinpol = False\\n self.nspin = 2 if spinpol else 1\\n\\n # kpoint stuff by ase\\n self.kpts = kpts\\n self.kpts_coord = None\\n\\n if self.kpts is not None:\\n initkey = 'Hamiltonian_KPointsAndWeights'\\n mp_mesh = None\\n offsets = None\\n \\n if isinstance(self.kpts, dict):\\n if 'path' in self.kpts:\\n # kpts is path in Brillouin zone\\n self.parameters[initkey + '_'] = 'Klines '\\n self.kpts_coord = kpts2ndarray(self.kpts, atoms=atoms)\\n else:\\n # kpts is (implicit) definition of\\n # Monkhorst-Pack grid\\n self.parameters[initkey + '_'] = 'SupercellFolding '\\n mp_mesh, offsets = kpts2sizeandoffsets(atoms=atoms,\\n **self.kpts)\\n elif np.array(self.kpts).ndim == 1:\\n # kpts is Monkhorst-Pack grid\\n self.parameters[initkey + '_'] = 'SupercellFolding '\\n mp_mesh = self.kpts\\n offsets = [0.] * 3\\n elif np.array(self.kpts).ndim == 2:\\n # kpts is (N x 3) list/array of k-point coordinates\\n # each will be given equal weight\\n self.parameters[initkey + '_'] = ''\\n self.kpts_coord = np.array(self.kpts)\\n else:\\n raise ValueError('Illegal kpts definition:' + str(self.kpts))\\n\\n if mp_mesh is not None:\\n eps = 1e-10\\n for i in range(3):\\n key = initkey + '_empty%03d' % i\\n val = [mp_mesh[i] if j == i else 0 for j in range(3)]\\n self.parameters[key] = ' '.join(map(str, val))\\n offsets[i] *= mp_mesh[i]\\n assert abs(offsets[i]) < eps or abs(offsets[i] - 0.5) < eps\\n # DFTB+ uses a different offset convention, where\\n # the k-point mesh is already Gamma-centered prior\\n # to the addition of any offsets\\n #if mp_mesh[i] % 2 == 0:\\n # offsets[i] += 0.5 # use gamma-centred\\n key = initkey + '_empty%03d' % 3\\n self.parameters[key] = ' '.join(map(str, offsets))\\n\\n elif self.kpts_coord is not None:\\n for i, c in enumerate(self.kpts_coord):\\n key = initkey + '_empty%09d' % i\\n c_str = ' '.join(map(str, c))\\n if 'Klines' in self.parameters[initkey + '_']:\\n c_str = '1 ' + c_str\\n else:\\n c_str += ' 1.0'\\n self.parameters[key] = c_str\",\n \"def solidity_compiler():\\n compiler = SolidityCompiler(test_contract_dir=TEST_CONTRACTS_DIR)\\n yield compiler\",\n \"def __init__(self, config, fe_space, scheme_type=SchemeType.EXPLICIT_ORDERONE, init_cond_type=InitialConditionType.NONE):\\n self.config = config\\n self.fe_space = fe_space\\n self.init_cond_type = init_cond_type\\n self.scheme_type = scheme_type\\n\\n self.u0 = None\\n self.u1 = None\\n self.u2 = None\\n self.ustar = None\\n self.operator1 = None\\n self.operator2 = None\\n self.rhs_operator = None\\n self.inv_operator = None\\n self.timestep = None\\n self.time = 0.0\",\n \"def main():\\n run_nutanix_vm_creation_module()\",\n \"def __init__(self, static_lib):\\n self.tool = self.find_tool()\\n if self.tool is None:\\n # ERROR no NM found\\n print(\\\"ERROR: Could not find readelf\\\")\\n sys.exit(1)\\n # TODO: Support readelf for reading symbols from archives\\n assert not static_lib and \\\"RealElf does not yet support static libs\\\"\\n self.flags = ['--wide', '--symbols']\",\n \"def __init__(self):\\n self.read_input()\\n self.update_binaries()\",\n \"def initialise(self, args, environ):\",\n \"def test_init(paper):\\n atSi = Atoms(\\\"Si8\\\",positions=[[0,0,0],[0.25,0.25,0.25],[0.5,0.5,0],[0.75,0.75,0.25],\\n [0.5,0,0.5],[0.75,0.25,0.75],[0,0.5,0.5],[0.25,0.75,0.75]],\\n cell=[5.43,5.43,5.43])\\n\\n entry = paper[0]\\n kwargs = {}\\n args = []\\n\\n calc = Aflow(atSi, '.', '.', 0, entry=entry, *args, **kwargs)\\n\\n assert calc.entry_file == \\\"./entry.pkl\\\"\\n assert calc.can_execute()\\n assert calc.can_extract()\\n assert not calc.is_executing()\\n calc.extract()\",\n \"def __init__(self, *args):\\n _hypre.HyprePCG_swiginit(self, _hypre.new_HyprePCG(*args))\",\n \"def __init__(self, name, rootdir=None, create=True):\\n # Guess the dtype based on the class name\\n dtype = FreezableAPI.guess_type(self)\\n self.m80 = FreezableAPI(dtype, name, rootdir)\",\n \"def createACTopol(self):\\n #sleap = False\\n if self.engine == 'sleap':\\n if self.execSleap():\\n self.printError(\\\"Sleap failed\\\")\\n self.printMess(\\\"... trying Tleap\\\")\\n if self.execTleap():\\n self.printError(\\\"Tleap failed\\\")\\n if self.engine == 'tleap':\\n if self.execTleap():\\n self.printError(\\\"Tleap failed\\\")\\n if self.extOld == '.pdb':\\n self.printMess(\\\"... trying Sleap\\\")\\n self.ext = self.extOld\\n self.inputFile = self.baseName+self.ext\\n if self.execSleap():\\n self.printError(\\\"Sleap failed\\\")\\n if not self.debug:\\n self.delOutputFiles()\",\n \"def __init__(self, evaluation_only=False, seed=0):\\n os = platform.system()\\n if os == 'Darwin':\\n file_name = 'Soccer.app'\\n elif os == 'Linux':\\n file_name = 'Soccer_Linux/Soccer.x86_64'\\n self.env = UnityEnvironment(file_name='unity_envs/' + file_name, seed=seed)\\n self.brain_names = self.env.brain_names\\n self.evaluation_only = evaluation_only\",\n \"def __init__(self, filename, default_seq=None, key_function=None, as_raw=False, strict_bounds=False):\\n self.filename = filename\\n self.faidx = Faidx(filename, key_function=key_function, as_raw=as_raw,\\n default_seq=default_seq, strict_bounds=strict_bounds)\",\n \"def New(*args, **kargs):\\n obj = itkIsoDataThresholdCalculatorHFUC.__New_orig__()\\n import itkTemplate\\n itkTemplate.New(obj, *args, **kargs)\\n return obj\",\n \"def __init__(self, *args):\\n _snap.TAscFltVP_swiginit(self, _snap.new_TAscFltVP(*args))\",\n \"def __init__(self, *args):\\n _snap.TAscFltPr_swiginit(self, _snap.new_TAscFltPr(*args))\",\n \"def create_constructor(self, *args, **kws):\\n node = FunctionNode(\\\"<constructor>\\\", *args, **kws)\\n node.class_member = True\\n self.add_object(node)\\n self.constructors.append(node)\\n return node\",\n \"def __init__(self):\\n\\n # Memory (RAM)\\n self.ram = [0] * 256\\n # Registers\\n self.reg = [0] * 8\\n # Program Counter\\n self.pc = 0\\n # Stack pointer\\n self.reg[7] = 244\\n # CPU status\\n self.running = False\\n # Flags register\\n self.fl = 0\\n # Branch table to handle instruction codes\\n self.branchtable = {}\\n self.branchtable[HLT] = self.halt\\n self.branchtable[LDI] = self.ldi\\n self.branchtable[PRN] = self.prn\\n self.branchtable[PUS] = self.push\\n self.branchtable[POP] = self.pop\\n \\n # Jump table that sets program counter\\n self.jumptable = {}\\n self.jumptable[CAL] = self.call\\n self.jumptable[RET] = self.ret\\n self.jumptable[JMP] = self.jmp\\n self.jumptable[JEQ] = self.jeq\\n self.jumptable[JNE] = self.jne\",\n \"def test_new():\\n from qiskit import BasicAer\\n from qiskit.aqua.algorithms import Grover\\n from qiskit.aqua.components.oracles import LogicalExpressionOracle\\n\\n expr = \\\"your logical expression goes here\\\"\\n algorithm = Grover(LogicalExpressionOracle(expr))\\n backend = BasicAer.get_backend('qasm_simulator')\\n result = algorithm.run(backend, seed=101110)\\n print(result)\",\n \"def init_analysis(session, args):\\r\\n path = os.path.join(session.abs_path,'{}_{}_{}_{}'.format(args[12],args[13],args[14],args[15]))\\r\\n session.case = Case(path, session.method)\\r\\n case = session.case\\r\\n if args[0].split('#')[0]=='R':\\r\\n args[7]= -args[7]\\r\\n\\r\\n case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[6], args[7]))\\r\\n\\r\\n case.file_U.set_field('boundaryField',\\r\\n {'inlet': {'type': 'freestream',\\r\\n 'freestreamValue': 'uniform ({} {} 0)'.format(args[6], args[7])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'fixedValue', 'value':'uniform (0 0 0)'},\\r\\n 'extrados': {'type': 'fixedValue', 'value':'uniform (0 0 0)'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\\r\\n case.file_p.set_field('boundaryField',\\r\\n {'inlet': {'type': 'freestreamPressure'},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\\r\\n session.case.file_T.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[11])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'slip'},\\r\\n 'extrados': {'type': 'slip'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n \\\"\\\"\\\"\\r\\n\\r\\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[6], args[7]))\\r\\n session.case.file_U.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue',\\r\\n 'value': 'uniform ({} {} 0)'.format(args[6], args[7])},\\r\\n 'outlet': {'type': 'inletOutlet','inletValue':'uniform ({} {} 0)'.format(args[6], args[7]),\\r\\n 'value':'uniform ({} {} 0)'.format(args[6], args[7])},\\r\\n 'intrados': {'type': 'noSlip'},\\r\\n 'extrados': {'type': 'noSlip'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\\r\\n session.case.file_p.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\\r\\n session.case.file_T.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue','value':'uniform {}'.format(args[11])},\\r\\n 'outlet': {'type': 'inletOutlet','inletValue':'uniform {}'.format(args[11]),'value':'uniform {}'.format(args[11])},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n \\\"\\\"\\\"\\r\\n \\\"\\\"\\\"\\r\\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[7], args[6]))\\r\\n session.case.file_U.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue',\\r\\n 'value': 'uniform ({} {} 0)'.format(args[7], args[6])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'noSlip'},\\r\\n 'extrados': {'type': 'noSlip'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[10]))\\r\\n session.case.file_p.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_T.set_field('internalField', 'uniform {}'.format(args[11]))\\r\\n session.case.file_T.set_field('boundaryField',\\r\\n {'inlet': {'type': 'zeroGradient'},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n \\\"\\\"\\\"\\r\\n \\\"\\\"\\\"\\r\\n session.case.file_U.set_field('internalField', 'uniform ({} {} 0)'.format(args[5], args[6]))\\r\\n session.case.file_U.set_field('boundaryField',\\r\\n {'inlet': {'type': 'fixedValue', 'value': 'uniform ({} {} 0)'.format(args[5], args[6])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'noSlip'},\\r\\n 'extrados': {'type': 'noSlip'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_p.set_field('boundaryField', {'inlet': {'type': 'fixedValue','value':'uniform {}'.format(args[9])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n session.case.file_p.set_field('internalField', 'uniform {}'.format(args[9]))\\r\\n\\r\\n session.case.file_T.set_field('boundaryField', {'inlet': {'type': 'fixedValue', 'value': 'uniform {}'.format(args[10])},\\r\\n 'outlet': {'type': 'zeroGradient'},\\r\\n 'intrados': {'type': 'zeroGradient'},\\r\\n 'extrados': {'type': 'zeroGradient'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n session.case.file_T.set_field('internalField','uniform 300')\\r\\n\\r\\n session.case.file_nut.set_field('boundaryField', {'inlet':{'type':'calculated', 'value':'uniform 0'},\\r\\n 'outlet':{'type':'calculated', 'value':'uniform 0'},\\r\\n 'intrados': {'type': 'nutkWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 0'},\\r\\n 'extrados': {'type': 'nutkWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 0'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_k.set_field('internalField', 'uniform 1')\\r\\n session.case.file_k.set_field('boundaryField', {\\r\\n 'inlet': {'type': 'turbulentIntensityKineticEnergyInlet', 'intensity': '0.05', 'value': 'uniform 1'},\\r\\n 'outlet': {'type': 'inletOutlet', 'inletValue': 'uniform 1', 'value': 'uniform 1'},\\r\\n 'intrados': {'type': 'kqRWallFunction','value':'uniform 1'},\\r\\n 'extrados': {'type': 'kqRWallFunction','value':'uniform 1'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n\\r\\n session.case.file_epsilon.set_field('boundaryField', {'inlet': {'type': 'turbulentMixingLengthDissipationRateInlet', 'mixingLength': '0.005', 'value': 'uniform 200'},\\r\\n 'outlet': {'type': 'inletOutlet', 'inletValue': 'uniform 200', 'value': 'uniform 200'},\\r\\n 'intrados': {'type': 'epsilonWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 200'},\\r\\n 'extrados': {'type': 'epsilonWallFunction', 'Cmu':'0.09', 'kappa':'0.41', 'E':'9.8', 'value':'uniform 200'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n session.case.file_epsilon.set_field('internalField', 'uniform 200')\\r\\n\\r\\n session.case.file_alphat.set_field('boundaryField', {'inlet':{'type':'calculated', 'value':'uniform 0'},\\r\\n 'outlet':{'type':'calculated', 'value':'uniform 0'},\\r\\n 'intrados': {'type': 'compressible::alphatWallFunction', 'Prt':'0.85', 'value':'uniform 0'},\\r\\n 'extrados': {'type': 'compressible::alphatWallFunction', 'Prt':'0.85', 'value':'uniform 0'},\\r\\n 'top_down': {'type': 'empty'},\\r\\n 'cyclic_in_1': {'type': 'cyclic'},\\r\\n 'cyclic_in_2': {'type': 'cyclic'},\\r\\n 'cyclic_out_1': {'type': 'cyclic'},\\r\\n 'cyclic_out_2': {'type': 'cyclic'}})\\r\\n \\\"\\\"\\\"\\r\\n session.case.file_controlDict.set_field('endTime', '10000')\\r\\n session.case.file_controlDict.set_field('startFrom', 'latestTime')\\r\\n session.case.file_controlDict.set_field('functions', {\\\"#includeFunc\\\":\\\"MachNo\\\"})\\r\\n session.case.file_turbulenceProperties.set_field('simulationType', 'laminar')\\r\\n session.case.interacting(100)\\r\\n sim = session.case.simulation(\\\"open40\\\") # Build files\\r\\n sim.limit_write = 50\\r\\n sim.block_mesh(string=write_block_mesh(args[1], args[2], args[3], args[4], args[5], session.mesh))\\r\\n sim.check_mesh()\\r\\n\\r\\n result_dict={\\\"T\\\": 0, \\\"p\\\":0, \\\"Theta\\\":0, \\\"z\\\":0, \\\"profile\\\":args[14]}\\r\\n\\r\\n def _function(container, args):\\r\\n current_time = container['current_time']\\r\\n if float(current_time)>=0.000015:\\r\\n print('Parsing results')\\r\\n sim.foamToVTK()\\r\\n results = sim.get_last_results('outlet')\\r\\n result_U = results.GetCellData('U')\\r\\n result_p = results.GetCellData('p')\\r\\n result_T = results.GetCellData('T')\\r\\n theta = 0.0\\r\\n z = 0.0\\r\\n p=0.0\\r\\n t=0.0\\r\\n U_length = len(result_U)\\r\\n p_length = len(result_p)\\r\\n t_length = len(result_T)\\r\\n for i,j,k in zip(result_p, result_T, result_U):\\r\\n p+= float(i[0])/p_length\\r\\n t+= float(j[0])/t_length\\r\\n theta += float(k[1])/U_length\\r\\n z += float(k[0])/U_length\\r\\n\\r\\n args[\\\"T\\\"] = t\\r\\n args[\\\"p\\\"] = p\\r\\n args[\\\"Theta\\\"] = theta\\r\\n args[\\\"z\\\"] = z\\r\\n return True\\r\\n return False\\r\\n \\r\\n #sim.run(_function, result_dict)\\r\\n #result_dict = {'T': 195.38959999999997, 'z': 429.3120571428572, 'p': 74001.90285714286, 'Theta': -207.19442857142855, 'profile': 0}\\r\\n\\r\\n print('Sending results')\\r\\n if args[0].split('#')[0]=='R':\\r\\n result_dict['Theta']= -result_dict['Theta']\\r\\n\\r\\n session.socket_design.send({'new_data':result_dict})\",\n \"def create_and_run():\\n\\n app = App()\\n app.run()\",\n \"def init():\\n safe_call(backend.get().af_init())\",\n \"def __init__(self, program):\\r\\n self._program = program\",\n \"def __init__(self, *tocs, **kwargs):\\n\\n from PyInstaller.config import CONF\\n Target.__init__(self)\\n name = kwargs.get('name', None)\\n cipher = kwargs.get('cipher', None)\\n self.toc = TOC()\\n # If available, use code objects directly from ModuleGraph to speed up PyInstaller.\\n self.code_dict = {}\\n for t in tocs:\\n self.toc.extend(t)\\n self.code_dict.update(getattr(t, '_code_cache', {}))\\n\\n self.name = name\\n if name is None:\\n self.name = os.path.splitext(self.tocfilename)[0] + '.pyz'\\n # PyInstaller bootstrapping modules.\\n self.dependencies = get_bootstrap_modules()\\n # Bundle the crypto key.\\n self.cipher = cipher\\n if cipher:\\n key_file = ('pyimod00_crypto_key', os.path.join(CONF['workpath'], 'pyimod00_crypto_key.pyc'), 'PYMODULE')\\n # Insert the key as the first module in the list. The key module contains just variables and does not depend\\n # on other modules.\\n self.dependencies.insert(0, key_file)\\n # Compile the top-level modules so that they end up in the CArchive and can be imported by the bootstrap script.\\n self.dependencies = misc.compile_py_files(self.dependencies, CONF['workpath'])\\n self.__postinit__()\",\n \"def __init__(self):\\n import arcgisscripting as ARC\\n self.GP = ARC.create(9.3)\\n self.params = self.GP.getparameterinfo()\",\n \"def __init__(self, *args):\\n _snap.TAscFltV_swiginit(self, _snap.new_TAscFltV(*args))\",\n \"def __init__(self, fips_codes, horiz_pos, vert_pos, population, risk):\\n self._fips_codes = fips_codes\\n self._horiz_center = horiz_pos\\n self._vert_center = vert_pos\\n self._total_population = population\\n self._averaged_risk = risk\",\n \"def __init__(self, fips_codes, horiz_pos, vert_pos, population, risk):\\n self._fips_codes = fips_codes\\n self._horiz_center = horiz_pos\\n self._vert_center = vert_pos\\n self._total_population = population\\n self._averaged_risk = risk\",\n \"def __init__(self):\\n _hypre.HypreIdentity_swiginit(self, _hypre.new_HypreIdentity())\",\n \"def main():\\n model = Calculator()\",\n \"def init():\\n try:\\n compile_contract(\\n \\\"fishcake\\\", f\\\"Fishcake(sp.address('{pub_key_hash}'),{default_supply})\\\")\\n fishcake_addr = deploy(\\\"fishcake\\\")\\n print(\\\"\\\\n\\\")\\n compile_contract(\\n \\\"fishcakeBox\\\", f\\\"FishcakeBox({default_redeem_amt}, sp.address('{fishcake_addr}'))\\\")\\n fishcake_box_addr = deploy(\\\"fishcakeBox\\\")\\n setup(fishcake_addr, fishcake_box_addr)\\n print(\\\"\\\\n\\\\n[!] Details :\\\\n\\\")\\n print(f\\\"-- Fishcake Token Address : {fishcake_addr}\\\")\\n print(f\\\"-- Fishcake Box Address : {fishcake_box_addr}\\\")\\n except Exception as e:\\n print(\\\"Failed to originate Contracts : \\\", e)\",\n \"def create_factory(cls, *args):\\n raise NotImplementedError\",\n \"def __init__(self, *args):\\n _snap.TSFlt_swiginit(self, _snap.new_TSFlt(*args))\",\n \"def __init__(self, source_code, bytecode=None, address=None, abi=SOPHIA, client=None):\\n if client is None:\\n client = epoch.EpochClient()\\n self.client = client\\n self.abi = abi\\n self.source_code = source_code\\n self.bytecode = bytecode\\n self.address = address\\n if self.bytecode is None:\\n self.bytecode = self.compile(self.source_code)\",\n \"def __init__(self):\\n BasicEngine.__init__(self)\\n self._active_qubits = 0\\n self._num_rotations = 0\\n self._rotations = []\",\n \"def my_main() -> None: # pragma: no cover\\n universe = AutomataUniverse(SQUARE_GRID_NEIGHBORS, [2,3], [3])\\n instance = AutomataTransforms(universe)\\n assert isinstance(instance, AutomataTransforms)\\n # # _is_rot_mat_test(instance)\\n # # _rotations_check(instance)\\n # # _prime_cells_check(instance)\\n # _check_transform_test(instance)\\n # # _hashable_transform_test(instance)\\n # _duplicate_test(instance)\\n # _collision_test(instance)\\n # _end_cycle_test(instance)\\n # _add_transform_test(instance)\\n # instance.generate_combination_transforms()\\n\\n # # _matrix_rotate_test(instance)\\n # # _duplicate_test(instance) # test again after transform(s) added\\n # # _collision_test(instance) # test again after transform(s) added «also refactoring»\\n # instance.dbg_report_instance() # DEBUG\",\n \"def make(self, *args, **kwargs):\\r\\n mode = kwargs.pop('mode', theano.compile.mode.get_default_mode())\\r\\n rval = self.make_no_init(mode)\\r\\n if hasattr(rval, 'initialize'):\\r\\n rval.initialize(*args, **kwargs)\\r\\n return rval\",\n \"def __init__(self):\\n\\n super().__init__(BabelCLIHandlersEnum.COMPILE)\",\n \"def __init__(self, base=None, mvtype=None, fct=False, blade_rep=False):\\n\\n def make_scalar(self, base): # make a scalar (grade 0)\\n if isinstance(base, str):\\n if self.fct:\\n self.obj = Function(base)(*MV.coords) * MV.ONE\\n else:\\n self.obj = make_coef(self, base) * MV.ONE\\n else:\\n self.obj = base * MV.ONE\\n self.igrade = 0\\n self.blade_rep = True\\n return self\\n\\n def make_vector(self, base): # make a vector (grade 1)\\n if isinstance(base, str):\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=1, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[1]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=1, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=1, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[1]))))\\n else:\\n result = S.Zero\\n for (coef, base) in zip(base, MV.blades[1]):\\n result += coef * base\\n self.obj = result\\n self.igrade = 1\\n self.blade_rep = True\\n return self\\n\\n def make_basisvector(self, base):\\n raise NotImplementedError(\\\"Don't know how to compute basis vectors of class %\\\" % self.__class__)\\n\\n def make_basisbivector(self, base):\\n raise NotImplementedError(\\\"Don't know how to compute basis bivectors of class %\\\" % self.__class__)\\n\\n def make_grade(self, base): # if base is 'A,n' then make a grade n multivector\\n if isinstance(base, str):\\n base_lst = base.split(',')\\n base = base_lst[0]\\n n = int(base_lst[1])\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=n, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[n]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=n, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=n, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[n]))))\\n else:\\n raise TypeError('Cannot make_grade for base = %s' % base)\\n self.igrade = n\\n self.blade_rep = True\\n return self\\n\\n def make_grade2(self, base): # grade 2 multivector\\n if isinstance(base, str):\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=2, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[2]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=2, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=2, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[2]))))\\n else:\\n raise TypeError('!!!!Cannot make_grade2 for base = ' + str(base) + '!!!!\\\\n')\\n self.igrade = 2\\n self.blade_rep = True\\n return self\\n\\n def make_pseudo(self, base): # multivector of grade MV.dim\\n if isinstance(base, str):\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=MV.dim, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[MV.dim]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=MV.dim, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=MV.dim, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj = reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[MV.dim]))))\\n else:\\n raise TypeError('!!!!Cannot make_pseudo for base = ' + str(base) + '!!!!\\\\n')\\n self.igrade = MV.dim\\n self.blade_rep = True\\n return self\\n\\n def make_spinor(self, base): # multivector with all even grades\\n if isinstance(base, str):\\n if self.fct:\\n self.obj = Function(base)(*MV.coords) * MV.ONE\\n else:\\n self.obj = Symbol(base) * MV.ONE\\n for rank in range(2, MV.dim1, 2):\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=rank, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\\n else:\\n raise TypeError('Cannot make_mv for base = %s' % base)\\n self.igrade = -1\\n self.blade_rep = True\\n return self\\n\\n def make_mv(self, base):\\n if isinstance(base, str):\\n if self.fct:\\n self.obj = Function(base)(*MV.coords) * MV.ONE\\n else:\\n self.obj = Symbol(base) * MV.ONE\\n for rank in range(1, MV.dim1):\\n if self.fct:\\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\\n fct_lst = fct_sym_array(base_lst, MV.coords)\\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\\n else:\\n if MV.coords is not None:\\n base_lst = str_combinations(base, MV.coords, rank=rank, mode='__')\\n else:\\n base_lst = str_combinations(base, MV.subscripts, rank=rank, mode='__')\\n fct_lst = fct_sym_array(base_lst, None)\\n self.obj += reduce(operator.add, tuple(map(lambda x: x[0] * x[1], zip(fct_lst, MV.blades[rank]))))\\n else:\\n raise TypeError('!!!!Cannot make_mv for base = ' + str(base) + '!!!!\\\\n')\\n self.igrade = -1\\n self.blade_rep = True\\n return self\\n\\n MVtypes = {'scalar': make_scalar,\\n 'vector': make_vector,\\n 'basisvector': make_basisvector,\\n 'basisbivector': make_basisbivector,\\n 'grade': make_grade,\\n 'grade2': make_grade2,\\n 'bivector': make_grade2,\\n 'pseudo': make_pseudo,\\n 'spinor': make_spinor,\\n 'mv': make_mv}\\n\\n self.fct = fct\\n self.is_base = False\\n self.is_grad = False\\n self.print_blades = MV.print_blades\\n self.fmt = 1\\n\\n if mvtype is None:\\n if base in (None, S.Zero): # Default is zero multivector\\n self.blade_rep = True\\n self.obj = S.Zero\\n self.igrade = 0\\n elif isinstance(base, str): # Base or blade basis multivector\\n self.is_base = True\\n if '*' in base:\\n self.blade_rep = False\\n self.igrade = -1\\n else:\\n if '^' in base:\\n self.blade_rep = True\\n self.igrade = base.count('^') + 1\\n else:\\n self.blade_rep = blade_rep\\n self.igrade = 1\\n self.obj = Symbol(base, commutative=False)\\n elif isinstance(base, MV): # Copy constructor\\n self.blade_rep = base.blade_rep\\n self.obj = base.obj\\n self.igrade = base.igrade\\n self.fct = base.fct\\n self.is_base = base.is_base\\n self.is_grad = base.is_grad\\n elif isinstance(base, (Expr, Symbol)): # Gets properties of multivector from Expr\\n if base.is_commutative:\\n self.obj = base * MV.ONE\\n self.blade_rep = True\\n self.igrade = 0\\n else:\\n if isinstance(base, (Add, Mul)): # Complex expression\\n MV.characterize_expression(self, base)\\n elif isinstance(base, Symbol):\\n if not base.is_commutative:\\n if base == MV.ONE:\\n self.obj = base\\n self.blade_rep = True\\n self.igrade = 0\\n elif base in MV.blades_flat: # basis blade\\n self.obj = base\\n self.blade_rep = True\\n self.igrade = MV.blade_grades[base]\\n elif base in MV.bases_flat: # basis base\\n self.obj = base\\n self.blade_rep = False\\n self.igrade = -1\\n else:\\n raise ValueError('MV(' + str(base) + ') is not allowed in constructor\\\\n' +\\n 'non-commutative argument is not a base\\\\n')\\n else: # scalar sympy symbol\\n self.obj = base * MV.ONE\\n self.igrade = 0\\n self.blade_rep = True\\n elif isinstance(base, Number):\\n self.obj = base * MV.ONE\\n self.igrade = 0\\n self.blade_rep = True\\n else: # Preconfigured multivector types\\n MVtypes[mvtype](self, base)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.57146007","0.56284124","0.56038344","0.55532616","0.5481942","0.54455805","0.5445103","0.54039806","0.5394971","0.5394971","0.5394971","0.5316189","0.5302524","0.5259621","0.5259621","0.52579147","0.52550256","0.52283067","0.519386","0.5180122","0.51798546","0.51767415","0.5131241","0.50992894","0.509765","0.5092811","0.5083334","0.5060924","0.5060009","0.50365174","0.50316226","0.50307584","0.5018654","0.50123847","0.50099134","0.49937305","0.49637154","0.49601915","0.4951824","0.4947635","0.4944229","0.49431902","0.49306953","0.49304506","0.4929569","0.49276817","0.4927248","0.4925283","0.49241108","0.49174756","0.48943472","0.48870382","0.48827353","0.4877591","0.4871639","0.48694456","0.48687828","0.48663265","0.48612276","0.48567513","0.48529905","0.48527285","0.48515365","0.48483697","0.4836743","0.4827254","0.48216957","0.48203668","0.4817109","0.48161522","0.48145285","0.48110816","0.47956392","0.47937933","0.47909793","0.47902173","0.47856823","0.47840533","0.47792667","0.47765198","0.47635046","0.47625947","0.47612873","0.47541025","0.4751699","0.47499073","0.47497252","0.47485897","0.47485897","0.47460067","0.47395894","0.47385582","0.47382113","0.47360194","0.4733779","0.47323477","0.47317678","0.47303218","0.4727161","0.47224417"],"string":"[\n \"0.57146007\",\n \"0.56284124\",\n \"0.56038344\",\n \"0.55532616\",\n \"0.5481942\",\n \"0.54455805\",\n \"0.5445103\",\n \"0.54039806\",\n \"0.5394971\",\n \"0.5394971\",\n \"0.5394971\",\n \"0.5316189\",\n \"0.5302524\",\n \"0.5259621\",\n \"0.5259621\",\n \"0.52579147\",\n \"0.52550256\",\n \"0.52283067\",\n \"0.519386\",\n \"0.5180122\",\n \"0.51798546\",\n \"0.51767415\",\n \"0.5131241\",\n \"0.50992894\",\n \"0.509765\",\n \"0.5092811\",\n \"0.5083334\",\n \"0.5060924\",\n \"0.5060009\",\n \"0.50365174\",\n \"0.50316226\",\n \"0.50307584\",\n \"0.5018654\",\n \"0.50123847\",\n \"0.50099134\",\n \"0.49937305\",\n \"0.49637154\",\n \"0.49601915\",\n \"0.4951824\",\n \"0.4947635\",\n \"0.4944229\",\n \"0.49431902\",\n \"0.49306953\",\n \"0.49304506\",\n \"0.4929569\",\n \"0.49276817\",\n \"0.4927248\",\n \"0.4925283\",\n \"0.49241108\",\n \"0.49174756\",\n \"0.48943472\",\n \"0.48870382\",\n \"0.48827353\",\n \"0.4877591\",\n \"0.4871639\",\n \"0.48694456\",\n \"0.48687828\",\n \"0.48663265\",\n \"0.48612276\",\n \"0.48567513\",\n \"0.48529905\",\n \"0.48527285\",\n \"0.48515365\",\n \"0.48483697\",\n \"0.4836743\",\n \"0.4827254\",\n \"0.48216957\",\n \"0.48203668\",\n \"0.4817109\",\n \"0.48161522\",\n \"0.48145285\",\n \"0.48110816\",\n \"0.47956392\",\n \"0.47937933\",\n \"0.47909793\",\n \"0.47902173\",\n \"0.47856823\",\n \"0.47840533\",\n \"0.47792667\",\n \"0.47765198\",\n \"0.47635046\",\n \"0.47625947\",\n \"0.47612873\",\n \"0.47541025\",\n \"0.4751699\",\n \"0.47499073\",\n \"0.47497252\",\n \"0.47485897\",\n \"0.47485897\",\n \"0.47460067\",\n \"0.47395894\",\n \"0.47385582\",\n \"0.47382113\",\n \"0.47360194\",\n \"0.4733779\",\n \"0.47323477\",\n \"0.47317678\",\n \"0.47303218\",\n \"0.4727161\",\n \"0.47224417\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6181505"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90889,"cells":{"query":{"kind":"string","value":"Return the FriCASExpectFunction class."},"document":{"kind":"string","value":"def _function_class(self):\n return FriCASExpectFunction"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def _function_element_class(self):\n return FriCASFunctionElement","def func(self):\n return self.__class__","def choose_botorch_acqf_class() -> Type[AcquisitionFunction]:\n # NOTE: In the future, this dispatch function could leverage any\n # of the attributes of `BoTorchModel` or kwargs passed to\n # `BoTorchModel.gen` to intelligently select acquisition function.\n return qNoisyExpectedImprovement","def fortran_c_wrapper(self) -> str:\n if self.fc_override is not None:\n return self.fc_override.replace('$CLASSNAME$', self.class_name).replace(\n \"$C_PREFIX$\", self.c_prefix).replace(\"$F_PREFIX$\", self.f_prefix)\n\n result = ''\n\n # declaration\n in_parameters = self._fc_in_parameters()\n return_type, out_parameters = self._fc_out_parameters()\n if self.may_throw:\n out_parameters.append('int * err_code')\n out_parameters.append('char ** err_msg')\n out_parameters.append('std::size_t * err_msg_len')\n\n func_name = '{}_{}_{}_'.format(\n self.c_prefix, self.class_name, self.name)\n\n par_str = ', '.join(in_parameters + out_parameters)\n result += '{} {}({}) {{\\n'.format(return_type, func_name, par_str)\n\n # convert input\n for par in self.params:\n result += '{}'.format(par.fc_convert_input())\n\n # call C++ function and return result\n if self.may_throw:\n result += ' try {\\n'\n result += ' *err_code = 0;\\n'\n result += indent(self._fc_cpp_call(), 4*' ')\n result += indent(self._fc_return(), 4*' ')\n result += ' }\\n'\n for exception, code in error_codes.items():\n if code != 0:\n catch = ''\n catch += 'catch (std::{} const & e) {{\\n'.format(exception)\n catch += ' *err_code = {};\\n'.format(code)\n catch += ' static std::string msg;\\n'\n catch += ' msg = e.what();\\n'\n catch += ' *err_msg = const_cast<char*>(msg.data());\\n'\n catch += ' *err_msg_len = msg.size();\\n'\n catch += '}\\n'\n result += indent(catch, 4*' ')\n result += self._fc_return_default()\n else:\n result += self._fc_cpp_call()\n result += self._fc_return()\n result += '}\\n\\n'\n return result","def _func(self):\n return self._get_flint_func(self.domain)","def getFunctionClass(functionID):\n d = { 1: Linear,\n 2: LinearDrag,\n 11: Gaussian,\n 12: GaussianDrag,\n 21: Lorentzian,\n 22: LorentzianDrag }\n return d[functionID]","def getFunction(self) -> ghidra.program.model.listing.Function:\n ...","def _make_cas_function():\n # Generate IR\n mod = lc.Module.new('generate-cas')\n llint = lc.Type.int()\n llintp = lc.Type.pointer(llint)\n fnty = lc.Type.function(llint, [llintp, llint, llint])\n fn = mod.add_function(fnty, name='.numba.parallel.ufunc.cas')\n ptr, old, repl = fn.args\n bb = fn.append_basic_block('')\n builder = lc.Builder.new(bb)\n outpack = builder.cmpxchg(ptr, old, repl, ordering='monotonic')\n out = builder.extract_value(outpack, 0)\n failed = builder.extract_value(outpack, 1)\n builder.ret(builder.select(failed, old, out))\n\n # Build & Link\n llmod = ll.parse_assembly(str(mod))\n\n target = ll.Target.from_triple(ll.get_process_triple())\n tm = target.create_target_machine()\n engine = ll.create_mcjit_compiler(llmod, tm)\n ptr = engine.get_function_address(fn.name)\n return engine, ptr","def arg_to_CFI(self, node, ordered_functions):\n options = node.options\n fmt_func = node.fmtdict\n\n if options.wrap_fortran is False:\n # The buffer function is intended to be called by Fortran.\n # No Fortran, no need for buffer function.\n return\n\n ast = node.ast\n declarator = ast.declarator\n result_typemap = ast.typemap\n # shadow classes have not been added yet.\n # Only care about string, vector here.\n result_is_ptr = declarator.is_indirect()\n if (\n result_typemap\n and result_typemap.base in [\"string\", \"vector\"]\n and result_typemap.name != \"char\"\n and not result_is_ptr\n ):\n node.wrap.c = False\n # node.wrap.fortran = False\n self.config.log.write(\n \"Skipping {}, unable to create C wrapper \"\n \"for function returning {} instance\"\n \" (must return a pointer or reference).\"\n \" Bufferify version will still be created.\\n\".format(\n result_typemap.cxx_type, declarator.user_name\n )\n )\n \n cfi_args = {}\n for arg in ast.declarator.params:\n declarator = arg.declarator\n name = declarator.user_name\n attrs = declarator.attrs\n meta = declarator.metaattrs\n cfi_args[name] = False\n arg_typemap = arg.typemap\n if meta[\"api\"]:\n # API explicitly set by user.\n continue\n elif meta[\"assumed-rank\"]:\n cfi_args[name] = True\n elif attrs[\"rank\"]:\n cfi_args[name] = True\n elif arg_typemap.sgroup == \"string\":\n cfi_args[name] = True\n elif arg_typemap.sgroup == \"char\":\n if declarator.is_indirect():\n cfi_args[name] = True\n elif meta[\"deref\"] in [\"allocatable\", \"pointer\"]:\n cfi_args[name] = True\n has_cfi_arg = any(cfi_args.values())\n\n # Function result.\n need_buf_result = None\n\n result_as_arg = \"\" # Only applies to string functions\n # when the result is added as an argument to the Fortran api.\n\n # Check if result needs to be an argument.\n declarator = ast.declarator\n attrs = declarator.attrs\n meta = declarator.metaattrs\n if meta[\"deref\"] == \"raw\":\n # No bufferify required for raw pointer result.\n pass\n elif result_typemap.sgroup == \"string\":\n need_buf_result = \"cfi\"\n result_as_arg = fmt_func.F_string_result_as_arg\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\n elif result_typemap.sgroup == \"char\" and result_is_ptr:\n need_buf_result = \"cfi\"\n result_as_arg = fmt_func.F_string_result_as_arg\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\n elif meta[\"deref\"] in [\"allocatable\", \"pointer\"]:\n need_buf_result = \"cfi\"\n\n if not (need_buf_result or\n has_cfi_arg):\n return False\n\n options.wrap_fortran = False\n\n # Create a new C function and change arguments\n # and add attributes.\n C_new = node.clone()\n ordered_functions.append(C_new)\n self.append_function_index(C_new)\n\n generated_suffix = \"cfi\"\n C_new._generated = \"arg_to_cfi\"\n C_new.splicer_group = \"cfi\"\n if need_buf_result:\n C_new.ast.declarator.metaattrs[\"api\"] = need_buf_result\n fmt_func = C_new.fmtdict\n fmt_func.function_suffix = fmt_func.function_suffix + fmt_func.C_cfi_suffix\n\n C_new.wrap.assign(c=True)#, fortran=True)\n C_new._PTR_C_CXX_index = node._function_index\n\n for arg in C_new.ast.declarator.params:\n name = arg.declarator.user_name\n if cfi_args[name]:\n arg.declarator.metaattrs[\"api\"] = generated_suffix\n\n ast = C_new.ast\n if True: # preserve to avoid changing indention for now.\n f_attrs = node.ast.declarator.attrs # Fortran function attributes\n f_meta = node.ast.declarator.metaattrs # Fortran function attributes\n if result_as_arg:\n # decl: const char * getCharPtr2() +len(30)\n # +len implies copying into users buffer.\n result_as_string = ast.result_as_arg(result_name)\n result_as_string.const = False # must be writeable\n attrs = result_as_string.declarator.attrs\n # Special case for wrapf.py to override \"allocatable\"\n f_meta[\"deref\"] = None\n result_as_string.declarator.metaattrs[\"api\"] = \"cfi\"\n result_as_string.declarator.metaattrs[\"deref\"] = \"result\"\n result_as_string.declarator.metaattrs[\"is_result\"] = True\n C_new.ast.declarator.metaattrs[\"api\"] = None\n C_new.ast.declarator.metaattrs[\"intent\"] = \"subroutine\"\n C_new.ast.declarator.metaattrs[\"deref\"] = None\n\n if result_as_arg:\n F_new = self.result_as_arg(node, C_new)\n ordered_functions.append(F_new)\n self.append_function_index(F_new)\n else:\n if node._generated in [\"result_to_arg\", \"fortran_generic\", \"getter/setter\"]:\n node.wrap.c = False\n # Fortran function may call C subroutine if string/vector result\n # Fortran function calls bufferify function.\n node._PTR_F_C_index = C_new._function_index\n return True","def _fc_function_definitions(self) -> str:\n result = 'extern \"C\" {\\n\\n'\n for namespace in self.namespaces:\n for member in namespace.members:\n result += member.fortran_c_wrapper()\n\n result += '}\\n\\n'\n return result","def func(cls):\n return cls.get_wrapper()(cls.callable)","def get_function(self):\n return SSAFunction(self.get_graph())","def createFunctionTerm(self):\n return _libsbml.Transition_createFunctionTerm(self)","def get_function(self):\n raise NotImplementedError()","def get_tested_func(self):\n t_ast = [x for x in self.AST.ext if type(x) is c_ast.FuncDef and x.decl.name == self.Fname]\n if len(t_ast) < 1:\n raise Exception(\"Func is not in ast\")\n else:\n return t_ast[0]","def test_cos():\n c=14\n def myfunc(x):\n f1=EF.cos(x)\n return f1\n\n f_obj=ADiff(myfunc)\n res=f_obj.Jac(c)\n\n expectAns={'diff': -math.sin(c), 'value': math.cos(c)}\n\n assert res==expectAns","def __init__(self, function='cogscore/'):\n self.function = function","def get_func_type(self, *args):\n return _ida_hexrays.cfunc_t_get_func_type(self, *args)","def fortran_function(self) -> str:\n if self.f_override is not None:\n return indent(\n self.f_override.replace('$CLASSNAME$', self.class_name).replace(\n \"$C_PREFIX$\", self.c_prefix).replace(\n \"$F_PREFIX$\", self.f_prefix),\n 4*' ')\n\n result = ''\n\n # declaration\n func_name = '{}_{}_{}'.format(\n self.f_prefix, self.class_name, self.name)\n in_parameters = self._f_in_parameters()\n return_type, out_parameters = self._f_out_parameters()\n if self.may_throw:\n out_parameters.append(('integer, optional', 'err_code'))\n out_parameters.append(('character(:), allocatable, optional',\n 'err_msg'))\n\n all_parameters = in_parameters + out_parameters\n arg_list = ', &\\n'.join([par_name for _, par_name in all_parameters])\n arg_ilist = indent(arg_list, 8*' ')\n if return_type != '':\n result += 'function {}( &\\n{})\\n'.format(func_name, arg_ilist)\n else:\n result += 'subroutine {}( &\\n{})\\n'.format(func_name, arg_ilist)\n\n # parameter declarations\n result += ' implicit none\\n'\n for par_type, par_name in in_parameters:\n result += ' {}, intent(in) :: {}\\n'.format(\n par_type, par_name)\n for par_type, par_name in out_parameters:\n result += ' {}, intent(out) :: {}\\n'.format(par_type, par_name)\n if return_type != '':\n result += ' {} :: {}\\n'.format(return_type, func_name)\n result += '\\n'\n\n # variable declarations\n c_return_type, fi_out_parameters = self._fi_out_parameters()\n if c_return_type:\n result += ' {} :: ret_val\\n'.format(c_return_type)\n for par_type, par_name in fi_out_parameters:\n result += ' {} :: {}\\n'.format(par_type, par_name)\n for par_type, par_name in self.ret_type.f_aux_variables():\n result += ' {} :: {}\\n'.format(par_type, par_name)\n if self.may_throw:\n result += ' integer (c_int) :: err_code_v\\n'\n result += ' type (c_ptr) :: err_msg_v\\n'\n result += ' integer (c_size_t) :: err_msg_len_v\\n'\n result += ' character (c_char), dimension(:), pointer :: err_msg_f\\n'\n result += ' character(:), allocatable :: err_msg_p\\n'\n result += ' integer (c_size_t) :: err_msg_i\\n'\n if c_return_type or fi_out_parameters or self.may_throw:\n result += '\\n'\n\n # convert input\n args = [param.f_chain_arg() for param in self.params]\n args += [par_name for _, par_name in fi_out_parameters]\n if self.may_throw:\n args += ['err_code_v', 'err_msg_v', 'err_msg_len_v']\n arg_str = ', &\\n'.join([8*' ' + arg for arg in args])\n\n # call C function\n fc_func_name = '{}_{}_{}_'.format(\n self.c_prefix, self.class_name, self.name)\n chain_call = self.fc_chain_call(\n ns_prefix=self.c_prefix, class_name=self.class_name,\n fc_func_name=fc_func_name, fc_args=arg_str)\n result_name = ''\n if return_type != '':\n result_name = func_name\n elif out_parameters:\n result_name = out_parameters[0][1]\n result += self.ret_type.f_call_c('ret_val', chain_call)\n\n # handle errors if necessary\n if self.may_throw:\n # Note: I tried to factor this out into a function, but Fortran\n # makes that near-impossible. Since we're generating anyway, it's\n # not really duplication, so leave it as is.\n result += indent(dedent(f\"\"\"\\\n if (err_code_v .ne. 0) then\n if (present(err_code)) then\n err_code = err_code_v\n if (present(err_msg)) then\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\n allocate (character(err_msg_len_v) :: err_msg)\n do err_msg_i = 1, err_msg_len_v\n err_msg(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\n end do\n end if\n {dedent(\n self.ret_type.f_return_dummy_result(result_name))}\n return\n else\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\n allocate (character(err_msg_len_v) :: err_msg_p)\n do err_msg_i = 1, err_msg_len_v\n err_msg_p(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\n end do\n print *, err_msg_p\n stop\n end if\n else\n if (present(err_code)) then\n err_code = 0\n end if\n end if\n\n \"\"\"), 4*' ')\n\n # convert and return result\n result += self.ret_type.f_return_result(result_name, 'ret_val')\n\n # end\n if return_type != '':\n result += 'end function {}\\n\\n'.format(func_name)\n else:\n result += 'end subroutine {}\\n\\n'.format(func_name)\n return indent(result, 4*' ')","def __init__(self, func_name, spin):\n self.xc_func = None\n self._xc_func_init = False\n\n # Handle func_name\n if isinstance(func_name, str):\n func_id = util.xc_functional_get_number(func_name)\n if func_id == -1:\n raise KeyError(\"LibXCFunctional: name '%s' not found.\" % func_name)\n elif isinstance(func_name, (int, np.integer)):\n func_id = func_name\n if util.xc_functional_get_name(func_name) is None:\n raise KeyError(\"LibXCFunctional: ID '%d' not found.\" % func_name)\n else:\n raise TypeError(\"LibXCFunctional: func_name must either be a string or int. Got {}\".format(func_name))\n\n self._xc_func_name = util.xc_functional_get_name(func_id)\n\n # Handle spin\n if isinstance(spin, str):\n spin = spin.lower()\n if spin == \"polarized\":\n self._spin = 2\n elif spin == \"unpolarized\":\n self._spin = 1\n else:\n raise KeyError(\"LibXCFunctional: spin must either be 'polarized' or 'unpolarized' if represented by a string. Got {}\".format(spin))\n else:\n self._spin = spin\n\n if self._spin not in [1, 2]:\n raise KeyError(\"LibXCFunctional: spin must either be 1 or 2 if represented by a integer. Got {}\".format(self._spin))\n\n # Build the LibXC functional\n self.xc_func = core.xc_func_alloc()\n self.xc_func_size_names = [x for x in dir(self.xc_func.contents.dim) if not \"_\" in x]\n\n # Set all int attributes to zero (not all set to zero in libxc)\n for attr in self.xc_func_size_names:\n setattr(self.xc_func.contents, attr, 0)\n\n ret = core.xc_func_init(self.xc_func, func_id, self._spin)\n if ret != 0:\n raise ValueError(\"LibXC Functional construction did not complete. Error code %d\" % ret)\n self._xc_func_init = True\n\n # Pull out all sizes after init\n self.xc_func_sizes = {}\n for attr in self.xc_func_size_names:\n self.xc_func_sizes[attr] = getattr(self.xc_func.contents.dim, attr)\n\n # Unpack functional info\n self.xc_func_info = core.xc_func_get_info(self.xc_func)\n self._number = core.xc_func_info_get_number(self.xc_func_info)\n self._kind = core.xc_func_info_get_kind(self.xc_func_info)\n self._name = core.xc_func_info_get_name(self.xc_func_info).decode(\"UTF-8\")\n self._family = core.xc_func_info_get_family(self.xc_func_info)\n self._flags = core.xc_func_info_get_flags(self.xc_func_info)\n\n # Set needed flags\n self._needs_laplacian = self._flags & flags.XC_FLAGS_NEEDS_LAPLACIAN\n\n # Set derivatives\n self._have_exc = self._flags & flags.XC_FLAGS_HAVE_EXC\n self._have_vxc = self._flags & flags.XC_FLAGS_HAVE_VXC\n self._have_fxc = self._flags & flags.XC_FLAGS_HAVE_FXC\n self._have_kxc = self._flags & flags.XC_FLAGS_HAVE_KXC\n self._have_lxc = self._flags & flags.XC_FLAGS_HAVE_LXC\n\n # Set omega\n self._have_cam = self._flags & flags.XC_FLAGS_HYB_CAM\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_CAMY\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LC\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LCY\n self._cam_omega = self._cam_alpha = self._cam_beta = False\n if self._have_cam:\n self._cam_omega = self.xc_func.contents.cam_omega\n self._cam_alpha = self.xc_func.contents.cam_alpha\n self._cam_beta = self.xc_func.contents.cam_beta\n\n elif self._family in [flags.XC_FAMILY_HYB_LDA, flags.XC_FAMILY_HYB_GGA, flags.XC_FAMILY_HYB_MGGA]:\n self._cam_alpha = self.xc_func.contents.cam_alpha\n\n # VV10\n self._have_vv10 = self._flags & flags.XC_FLAGS_VV10\n self._nlc_b = self._nlc_C = False\n if self._have_vv10:\n self._nlc_b = self.xc_func.contents.nlc_b\n self._nlc_C = self.xc_func.contents.nlc_C\n\n # Stable\n self._stable = self._flags & flags.XC_FLAGS_STABLE\n self._dev = self._flags & flags.XC_FLAGS_DEVELOPMENT\n\n # Pull out references\n self._refs = []\n self._bibtexs = []\n self._dois = []\n\n for pos in range(flags.XC_MAX_REFERENCES):\n ref = core.xc_func_info_get_references(self.xc_func_info, pos)\n if not ref: break\n\n self._refs.append(ref.contents.ref.decode(\"UTF-8\"))\n self._bibtexs.append(ref.contents.bibtex.decode(\"UTF-8\"))\n self._dois.append(ref.contents.doi.decode(\"UTF-8\"))","def get_unc_f(self):\n return self.uncf","def __init__(self, expected, test_func):\n self._f = test_func\n self._exp = expected","def cfunc_type(self):\n tif = ida_typeinf.tinfo_t()\n result = self.get_func_type(tif)\n if not result:\n return\n return tif","def __init__(self,\n function: Callable):\n\n self._function = function","def _get_class():\n return ASParameters","def test_arccos():\n c=0.5\n def myfunc(x):\n f1=EF.arccos(x)\n return f1\n\n f_obj=ADiff(myfunc)\n res=f_obj.Jac(c)\n expectAns={'diff':-1/math.sqrt(1-c**2), 'value':math.acos(c)}\n assert res==expectAns","def fortran_function(self) -> str:\n return ''.join([i.fortran_function() for i in self.instances])","def getColorTransferFunction(self):\n\t\treturn self.ctf","def _constructor(self):\n return self.__class__","def getFunction(self):\n return gpio.gpio_function(self.bcm_id)","def get_decorated_function(self):","def getCallable():","def __init__(self, func, args, kwargs, res): # make test\n\n # Check if func is a callable function\n if callable(func):\n self.func = func\n else:\n self.func = lambda *args, **kwargs: func\n\n # Store parameters in class\n self.args = args or [] # If args == None, set self.args to an empty list\n self.kwargs = kwargs or {} # If kwargs == None, set self.kwargs to an empty dict\n self.res = res\n\n\n self._tolerance = 1.0e-10 # Variable used to check for equality in case of float","def get_ea(func_ea):\n if isinstance(func_ea, idaapi.func_t):\n return func_ea.startEA\n return func_ea","def __init__(self, function: ast3.FunctionDef, module: \"Module\") -> None:\n\n # easy data\n self._function = function\n self.name = function.name\n self.line = function.lineno\n self.column = function.col_offset\n self.body = function.body\n self.module = module\n self.decorators = function.decorator_list\n\n # time to parse arguments\n self._args = function.args.args\n self.args = [Arg(x) for x in self._args]\n self.functions = [\n Function(x, self.module)\n for x in self.body\n if isinstance(x, ast3.FunctionDef)\n ]\n self.classes = [\n Class(x, self.module) for x in self.body if isinstance(x, ast3.ClassDef)\n ]\n self.untyped = [\n x for x in self.args if not x.typed and x not in DEFAULT_ARG_IGNORE\n ]\n self.doc = None\n self.returns = None\n self.return_typed = False\n self.missing_args: Set[str] = set()\n self.unexpected_args: Set[str] = set()\n arg_names = set(x.name for x in self.args if x.name not in DEFAULT_ARG_IGNORE)\n self.missing_args = arg_names\n if isinstance(self.body[0], ast3.Expr):\n # this is most likely a doc string\n self.doc = Doc(self.body[0], Doc.Type.FUNCTION)\n doc_arg_names = set(x for x, y in self.doc.args.items())\n self.missing_args = arg_names - doc_arg_names\n self.unexpected_args = doc_arg_names - arg_names\n if function.returns:\n self.return_typed = True\n self.returns = parse_elt(function.returns) # type: ignore\n\n # complexity checks\n self._radon = cc_visit_ast(self._function)[0]\n self.complexity = self._radon.complexity\n self.is_method = self._radon.is_method\n self._halstead = h_visit_ast(self._function)","def __call__(self, f_or_klass):\n\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\n # 0.3.0\n # -- implement \"kill switch\", NO_DECO\n # -- handle decorating both functions and classes\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\n\n # 0.3.0b16: if it isn't callable, scram'\n if not callable(f_or_klass):\n return f_or_klass\n\n # Special-case handling for ``NO_DECO``: remove from settings of ``self``\n if self._effective_settings.get('NO_DECO'):\n return f_or_klass\n # else, delete that item wherever it might be\n if 'NO_DECO' in self._effective_settings:\n del self._effective_settings['NO_DECO']\n if 'NO_DECO' in self._changed_settings:\n del self._changed_settings['NO_DECO']\n\n f = f_or_klass if inspect.isfunction(f_or_klass) else None\n klass = f_or_klass if inspect.isclass(f_or_klass) else None\n\n self.f = f\n self.cls = klass\n\n if klass:\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a class\" -- namely, klass\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n\n self._class__call__(klass) # modifies klass (methods & inner classes) (if not builtin)\n self._add_class_attrs(klass) # v0.3.0v20 traps TypeError for builtins\n return klass\n\n elif not f:\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a callable but not a function\"\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # functools.partial objects are callable, have no __name__ much less __qualname__,\n # and trying to deco __call__ gets messy.\n # Callable builtins e.g. len are not functions in the isfunction sense,\n # can't deco anyway. Just give up (quietly):\n return f_or_klass\n\n else: # not a class, f nonempty is a function of f_or_klass callable\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a function\" -- namely, f\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n\n #----------------------------------------------------------------\n # Don't double-decorate -- don't wanna, & it doesn't work anyway!\n #----------------------------------------------------------------\n # Note: As with methods of classes,\n # . if f is deco'd, its existing EXPLICITLY GIVEN settings take precedence.\n\n # # From _class__call__, props & methods cases, w/a few name changes\n deco_obj = getattr(f, self._sentinels['DECO_OF'], None) # type: _deco_base\n\n # get a fresh copy for each attr\n new_settings = self._changed_settings.copy() # updated below\n\n # __init__ fixup, a nicety:\n # By default, don't log retval for __init__.\n # If user insists on it with 'log_retval=True' in __init__ deco,\n # that will override this.\n if f.__name__ == '__init__':\n self.fixup_for_init(new_settings)\n\n if deco_obj: # f is deco'd by this decorator\n # Yes. Figure out settings for f,\n ### 0.3.0b18 -- Use self._override\n self._update_settings(new=new_settings,\n old=deco_obj._changed_settings,\n override_existing=self._override)\n # update func's settings (_force_mutable=True to handle `max_history` properly)\n deco_obj._settings_mapping.update(new_settings, _force_mutable=True)\n return f\n\n #----------------------------------------------------------------\n # f is a function & is NOT already deco'd\n #----------------------------------------------------------------\n\n # 0.3.0.x -- f may not have a .__qualname__\n try:\n self._classname_of_f = '.'.join( f.__qualname__.split('.')[:-1] )\n except AttributeError as e:\n self._classname_of_f = ''\n\n # Special-case '__repr__' handling, if deco subclass doesn't allow it.\n if f.__name__ == '__repr__' and self._classname_of_f and not self.allow_repr():\n # v0.3.0b23 -- Instead of refusing to deco, use recursive_repr\n # return f\n return recursive_repr(fillvalue=\"...\")(f)\n\n # 0.3.0\n # Use __qualname__ ALL the time, unless user provides `name=display_name_str`\n # where `display_name_str` is either the name to be used for the fn in logged output,\n # or is an oldstyle format str into which f.__name__ will be substituted\n # to obtain the display name.\n # We require Py3.3+, so __qualname__ is available.\n\n # setup f_display_name\n if self._name_param:\n try:\n self.f_display_name = (self._name_param % f.__name__)\n except TypeError:\n self.f_display_name = self._name_param\n else:\n self.f_display_name = f.__qualname__\n\n # TODO TRY THIS -- anything break?\n # 0.3.1 Inspired by fractions.Fraction.__sub__ et al:\n # __name__ may be very different from __qualname__;\n # if so, show both\n if f.__name__ not in f.__qualname__:\n self.f_display_name += \" (\" + f.__name__ + \")\"\n\n #================================================================\n # 0.3.0 -- Init things (migrated from __init__)\n #----------------------------------------------------------------\n # set up pseudo-dict (DecoSettingsMapping),\n # using settings given by self._effective_settings.\n #\n # *** DecoSettingsMapping \"API\" --\n # (2) construct DecoSettingsMapping object\n # that will provide mapping & attribute access to settings, & more\n #----------------------------------------------------------------\n self._settings_mapping = DecoSettingsMapping(\n deco_class=self.__class__,\n # DecoSettingsMapping calls the rest ** values_dict\n ** self._effective_settings # 0.3.0 set by __init__\n )\n\n #----------------------------------------------------------------\n # Init more stuff\n #----------------------------------------------------------------\n self._stats = ClassInstanceAttrProxy(\n class_instance=self,\n data_descriptor_names=self.__class__._data_descriptor_names,\n method_descriptor_names=self.__class__._method_descriptor_names)\n # Accessed by descriptors on the stats obj\n self._num_calls_total = 0\n self._num_calls_logged = 0\n # max_history > 0 --> size of self._call_history; <= 0 --> unbounded\n # Set before calling _make_call_history\n\n # 0.3.0 self._other_values_dict set by __init__\n self.max_history = self._other_values_dict.get('max_history', 0) # <-- Nota bene\n self._call_history = self._make_call_history()\n\n # Accumulate this (for logged calls only)\n # even when record_history is false:\n self._elapsed_secs_logged = 0.0\n self._process_secs_logged = 0.0\n\n # 0.2.2.post1\n # stack(s), pushed & popped wrapper of deco'd function\n # by _logging_state_push, _logging_state_pop\n # 0.3.0 convert to pushing/popping single namedtuples\n self.logging_state_stack = [] # 0.3.0 stack of LoggingState namedtuples\n self._enabled_stack = [] # 0.3.0 - um, stack, of 'enabled's\n\n #----------------------------------------------------------------\n # end of Init passage\n #================================================================\n\n # Save signature and parameters of f\n self.f_signature = inspect.signature(f) # Py >= 3.3\n self.f_params = self.f_signature.parameters\n\n # 0.3.0 We assume Py3.3 so we use perf_counter, process_time all the time\n wall_time_fn = time.perf_counter\n process_time_fn = time.process_time\n\n #############################\n # The wrapper of a callable\n #############################\n\n @wraps(f)\n def _deco_base_f_wrapper_(*args, **kwargs):\n \"\"\"Wrapper around the wrapped function f.\n When this runs, f has been called, so we can now resolve\n any indirect values for the settings/keyword-params\n of log_calls, using info in kwargs and self.f_params.\"\"\"\n # *** Part of the DecoSettingsMapping \"API\" --\n # (4) using self._settings_mapping.get_final_value in wrapper\n # [[[ This/these is/are 4th chronologically ]]]\n\n # inner/local fn -- save a few cycles and characters -\n # we call this a lot (<= 9x).\n def _get_final_value(setting_name):\n \"Use outer scope's kwargs and self.f_params\"\n return self._settings_mapping.get_final_value(\n setting_name, kwargs, fparams=self.f_params)\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # if nothing to do, hurry up & don't do it.\n # NOTE: call_chain_to_next_log_calls_fn looks in stack frames\n # to find (0.2.4) STACKFRAME_HACK_DICT_NAME (really!)\n # It and its values (the following _XXX variables)\n # must be set before calling f.\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n _enabled = _get_final_value('enabled')\n # 0.3.0 in case f calls log_message (no output if f disabled)\n self._enabled_state_push(_enabled)\n\n # 0.2.4.post5 \"true bypass\": if 'enabled' < 0 then scram\n if _enabled < 0:\n ret = f(*args, **kwargs)\n self._enabled_state_pop()\n return ret\n\n # Bump call counters, before calling fn.\n # Note: elapsed_secs, process_secs not reflected yet of course\n self._add_call(logged=_enabled)\n\n _log_call_numbers = _get_final_value('log_call_numbers')\n # counters just got bumped\n _active_call_number = (self._stats.num_calls_logged\n if _log_call_numbers else\n 0)\n # Get list of callers up to & including first log_call's-deco'd fn\n # (or just caller, if no such fn)\n call_list, prev_indent_level = self.call_chain_to_next_log_calls_fn()\n\n # Bump _extra_indent_level if last fn on call_list is deco'd AND enabled,\n # o/w it's the _extra_indent_level which that fn 'inherited'.\n # _extra_indent_level: prev_indent_level, or prev_indent_level + 1\n do_indent = _get_final_value('indent')\n _extra_indent_level = (prev_indent_level +\n int(not not do_indent and not not _enabled))\n # 0.3.0\n ########## prefixed_fname = _get_final_value('prefix') + f.__name__\n prefixed_fname = _get_final_value('prefix') + self.f_display_name\n\n # Stackframe hack:\n assert '_deco_base__active_call_items__' == STACKFRAME_HACK_DICT_NAME\n _deco_base__active_call_items__ = {\n '_enabled': _enabled,\n '_log_call_numbers': _log_call_numbers,\n '_prefixed_fname': prefixed_fname, # Hack alert (Pt 1)\n '_active_call_number': _active_call_number,\n '_extra_indent_level': _extra_indent_level,\n # 0.3.0 for _get_own_deco_wrapper\n '_wrapper_deco': self\n }\n\n # Get logging function IF ANY.\n # For the benefit of callees further down the call chain,\n # if this f is not enabled (_enabled <= 0).\n # Subclass can return None to suppress printed/logged output.\n logging_fn = self.get_logging_fn(_get_final_value)\n\n # Only do global indentation for print, not for loggers\n global_indent_len = max(_extra_indent_level, 0) * self.INDENT\n\n # 0.2.2.post1 - save output_fname for log_message use\n call_number_str = ((' [%d]' % _active_call_number)\n if _log_call_numbers else '')\n output_fname = prefixed_fname + call_number_str\n\n # 0.3.0\n # Note: DON'T combine with global_mute(),\n # cuz this value will be pushed,\n # and when popped any realtime changes to global mute\n # made during call to f would be ignored.\n mute = _get_final_value('mute')\n\n # 0.2.2 -- self._log_message() will use\n # the logging_fn, indent_len and output_fname at top of these stacks;\n # thus, verbose functions should use log_calls.print (~ log_message)\n # to write their blather.\n # There's a stack of logging-state ,\n # used by self._log_message(), maintained in this wrapper.\n self._logging_state_push(logging_fn, global_indent_len, output_fname, mute)\n\n # (_xxx variables set, ok to call f)\n if not _enabled:\n ret = f(*args, **kwargs)\n self._logging_state_pop(enabled_too=True)\n return ret\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Set up context, for pre-call handlers\n # (after calling f, add to it for post-call handlers)\n # THIS is the time sink - 23x slower than other 'blocks'\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Key/values of \"context\" whose values we know so far:\n context = {\n 'decorator': self,\n 'settings': self._settings_mapping,\n 'stats': self._stats,\n 'prefixed_fname': prefixed_fname,\n 'fparams': self.f_params,\n 'call_list': call_list,\n 'args': args,\n 'kwargs': kwargs,\n 'indent': \" \" * self.INDENT, # our unit of indentation\n 'output_fname': output_fname,\n }\n\n # Gather all the things we need (for log output, & for history)\n # Use inspect module's Signature.bind method.\n # bound_args.arguments -- contains only explicitly bound arguments\n # 0.2.4.post5 - using\n # inspect.signature(f).bind(*args, **kwargs)\n # took 45% of execution time of entire wrapper; this takes 23%:\n # 0.3.1 TODO BUG No args is a problem?!\n bound_args = self.f_signature.bind(*args, **kwargs)\n \"\"\"\n File \"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/tests/_temp.py\", line 12, in <module>\n g(f())\n File \"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/log_calls.py\", line 1935, in _deco_base_f_wrapper_\n bound_args = self.f_signature.bind(*args, **kwargs)\n File \"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\", line 2646, in bind\n return args[0]._bind(args[1:], kwargs)\n File \"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\", line 2571, in _bind\n raise TypeError('too many positional arguments') from None\n TypeError: too many positional arguments\n \"\"\"\n\n varargs_pos = get_args_pos(self.f_params) # -1 if no *args in signature\n argcount = varargs_pos if varargs_pos >= 0 else len(args)\n context['argcount'] = argcount\n # The first argcount-many things in bound_args\n context['argnames'] = list(bound_args.arguments)[:argcount]\n context['argvals'] = args[:argcount]\n\n context['varargs'] = args[argcount:]\n (context['varargs_name'],\n context['kwargs_name']) = get_args_kwargs_param_names(self.f_params)\n\n # These 3 statements = 31% of execution time of wrapper\n context['defaulted_kwargs'] = get_defaulted_kwargs_OD(self.f_params, bound_args)\n context['explicit_kwargs'] = get_explicit_kwargs_OD(self.f_params, bound_args, kwargs)\n # context['implicit_kwargs'] = {\n # k: kwargs[k] for k in kwargs if k not in context['explicit_kwargs']\n # }\n # At least 2x as fast:\n context['implicit_kwargs'] = \\\n difference_update(kwargs.copy(), context['explicit_kwargs'])\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call pre-call handlers, collect nonempty return values\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # only consult global mute in r/t\n if not (mute or self.global_mute()): # 0.3.0\n pre_msgs = []\n for setting_name in self._settings_mapping._pre_call_handlers: # keys\n if _get_final_value(setting_name):\n info = self._settings_mapping._get_DecoSetting(setting_name)\n msg = info.pre_call_handler(context)\n if msg:\n pre_msgs.append(msg)\n\n # Write pre-call messages\n if logging_fn:\n for msg in pre_msgs:\n self._log_message(msg, extra_indent_level=0)\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call f(*args, **kwargs) and get its retval; time it.\n # Add timestamp, elapsed time(s) and retval to context.\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # No dictionary overhead between timer(s) start & stop.\n t0 = time.time() # for timestamp\n t0_wall = wall_time_fn()\n t0_process = process_time_fn()\n retval = f(*args, **kwargs)\n t_end_wall = wall_time_fn()\n t_end_process = process_time_fn()\n context['elapsed_secs'] = (t_end_wall - t0_wall)\n context['process_secs'] = (t_end_process - t0_process)\n context['timestamp'] = t0\n context['retval'] = retval\n\n self._add_to_elapsed(context['elapsed_secs'], context['process_secs'])\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call post-call handlers, collect nonempty return values\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # only consult global mute in r/t\n if not (mute or self.global_mute()): # 0.3.0\n post_msgs = []\n for setting_name in self._settings_mapping._post_call_handlers: # keys\n if _get_final_value(setting_name):\n info = self._settings_mapping._get_DecoSetting(setting_name)\n msg = info.post_call_handler(context)\n if msg:\n post_msgs.append(msg)\n\n # Write post-call messages\n if logging_fn:\n for msg in post_msgs:\n self._log_message(msg, extra_indent_level=0)\n # v0.3.0b22 -- if recording history, add record of call even if we're muted(!)\n elif _get_final_value('record_history'):\n info = self._settings_mapping._get_DecoSetting('record_history')\n _ = info.post_call_handler(context)\n\n self._logging_state_pop(enabled_too=True)\n\n return retval\n\n self._add_function_attrs(f, _deco_base_f_wrapper_)\n return _deco_base_f_wrapper_\n\n #-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+\n # end else (case \"f_or_klass is a function\",\n # subcase \"f is a function & is NOT already deco'd\")\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*","def as_ret(self):\n if self.type in OCClass.classes_indexed_by_name:\n oc_class = OCClass.classes_indexed_by_name[self.type]\n if oc_class.class_meths:\n for f in oc_class.class_meths:\n #TODO MachOTask.currentTask.analyze_function(start_addr=f)\n print 'T'","def get_function(self):\n return subs(self.f.get_function(), self.sub_pre, self.sub_post)","def __init__(self, function):\n self.function = function","def get_func_type(self, *args):\n return _ida_hexrays.cfuncptr_t_get_func_type(self, *args)","def functional(self):\n return self.__functional","def get_real_function( this, fn):\n\t\treturn this._get_native_function(fn)","def createFunctionDefinition(self):\n return _libsbml.Model_createFunctionDefinition(self)","def f(self):\n return self._f","def get_ctf(ima):\n\tfrom EMAN2 import EMAN2Ctf\n\tctf_params = ima.get_attr(\"ctf\")\t\n\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang","def convert(self):\n return _libsbml.SBMLFunctionDefinitionConverter_convert(self)","def for_function(self, func):\n return self.for_pysig(utils.pysignature(func))","def getcfix(self):\n cfix_ = ctypes.c_double()\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\n if res != 0:\n _,msg = self.__getlasterror(res)\n raise Error(rescode(res),msg)\n cfix_ = cfix_.value\n _cfix_return_value = cfix_\n return (_cfix_return_value)","def func ( self ) :\n return self.__func","def getcfix(self): # 3\n res,resargs = self.__obj.getcfix()\n if res != 0:\n result,msg = self.__getlasterror(res)\n raise Error(rescode(res),msg)\n _cfix_return_value = resargs\n return _cfix_return_value","def classFactory(iface): # pylint: disable=invalid-name\n #\n from .CADDigitize import CADDigitize\n return CADDigitize(iface)","def __init__(self, function=None):\n self._function = function","def getTestClass(self):\n\t\treturn AbstraccionPrueba","def instantiate_callable_class(builder: IRBuilder, fn_info: FuncInfo) -> Value:\n fitem = fn_info.fitem\n func_reg = builder.add(Call(fn_info.callable_class.ir.ctor, [], fitem.line))\n\n # Set the environment attribute of the callable class to point at\n # the environment class defined in the callable class' immediate\n # outer scope. Note that there are three possible environment\n # class registers we may use. This depends on what the encapsulating\n # (parent) function is:\n #\n # - A nested function: the callable class is instantiated\n # from the current callable class' '__call__' function, and hence\n # the callable class' environment register is used.\n # - A generator function: the callable class is instantiated\n # from the '__next__' method of the generator class, and hence the\n # environment of the generator class is used.\n # - Regular function: we use the environment of the original function.\n curr_env_reg = None\n if builder.fn_info.is_generator:\n curr_env_reg = builder.fn_info.generator_class.curr_env_reg\n elif builder.fn_info.is_nested:\n curr_env_reg = builder.fn_info.callable_class.curr_env_reg\n elif builder.fn_info.contains_nested:\n curr_env_reg = builder.fn_info.curr_env_reg\n if curr_env_reg:\n builder.add(SetAttr(func_reg, ENV_ATTR_NAME, curr_env_reg, fitem.line))\n return func_reg","def ICRFFile_Instance():\n return _gmat_py.ICRFFile_Instance()","def func ( self ) :\n return self.__func","def fortran_interface(self) -> str:\n result = ''\n if self.fc_override == '':\n return result\n\n func_name = '{}_{}_{}_'.format(\n self.c_prefix, self.class_name, self.name)\n\n # declaration\n in_parameters = self._fi_in_parameters()\n return_type, out_parameters = self._fi_out_parameters()\n if self.may_throw:\n out_parameters.append(('integer (c_int)', 'err_code'))\n out_parameters.append(('type (c_ptr)', 'err_msg'))\n out_parameters.append(('integer (c_size_t)', 'err_msg_len'))\n\n arg_list = [par_name for _, par_name in in_parameters + out_parameters]\n if len(arg_list) > 1:\n arg_vlist = ' &\\n' + indent(', &\\n'.join(arg_list), 8*' ')\n else:\n arg_vlist = ', '.join(arg_list)\n\n if return_type != '':\n result += '{} function {}({}) &\\n'.format(\n return_type, func_name, arg_vlist)\n else:\n result += 'subroutine {}({}) &\\n'.format(func_name, arg_vlist)\n result += ' bind(C, name=\"{}\")\\n'.format(func_name)\n result += '\\n'\n result += ' use iso_c_binding\\n'\n\n # parameter declarations\n for par_type, par_name in in_parameters:\n result += ' {}, intent(in) :: {}\\n'.format(\n par_type, par_name)\n for par_type, par_name in out_parameters:\n result += ' {}, intent(out) :: {}\\n'.format(par_type, par_name)\n\n # end\n if return_type != '':\n result += 'end function {}\\n\\n'.format(func_name)\n else:\n result += 'end subroutine {}\\n\\n'.format(func_name)\n return indent(result, 8*' ')","def _wrap_FunctionDefResult(self, expr):\n var = expr.var\n name = var.name\n scope = self.scope\n # Make name available for later\n scope.insert_symbol(name)\n local_var = var.clone(scope.get_expected_name(name))\n\n if local_var.rank:\n # Allocatable is not returned so it must appear in local scope\n scope.insert_variable(local_var, name)\n\n # Create the C-compatible data pointer\n bind_var = Variable(dtype=BindCPointer(),\n name=scope.get_new_name('bound_'+name),\n is_const=False, memory_handling='alias')\n scope.insert_variable(bind_var)\n\n result = BindCFunctionDefResult(bind_var, var, scope)\n\n # Save the shapes of the array\n self._additional_exprs.extend([Assign(result.shape[i], var.shape[i]) for i in range(var.rank)])\n\n # Create an array variable which can be passed to CLocFunc\n ptr_var = var.clone(scope.get_new_name(name+'_ptr'),\n memory_handling='alias')\n scope.insert_variable(ptr_var)\n\n # Define the additional steps necessary to define and fill ptr_var\n alloc = Allocate(ptr_var, shape=result.shape,\n order=var.order, status='unallocated')\n copy = Assign(ptr_var, var)\n c_loc = CLocFunc(ptr_var, bind_var)\n self._additional_exprs.extend([alloc, copy, c_loc])\n\n return result\n else:\n return BindCFunctionDefResult(local_var, var, scope)","def class_from_invokedynamic(ins, cf):\n info = InvokeDynamicInfo.create(ins, cf)\n assert info.created_type != \"void\"\n return info.created_type","def FunctionDef(self):\n type = self.Type()\n key = self.currtok[0]\n self.functions[self.currtok[0]] = \"first_call\"\n id = self.primary()\n self.functions[key] = type\n if self.currtok[1].name == \"LPAREN\":\n self.currtok = next(self.tg)\n params = self.Params()\n if self.currtok[1].name == \"RPAREN\":\n self.currtok = next(self.tg)\n if self.currtok[1].name == \"LCURLY\":\n self.currtok = next(self.tg)\n decs = self.Declarations()\n states = self.Statements()\n if self.currtok[1].name == \"RCURLY\":\n self.currtok = next(self.tg)\n return FunctionDef(type, id, params, decs, states)\n raise SLUCSyntaxError(\"ERROR: Missing Right Curly Brace on line {0}\"\n .format(str(self.currtok[2] - 1 - 1)))\n raise SLUCSyntaxError(\"ERROR: Missing Left Curly Brace on line {0}\".format(str(self.currtok[2] - 1)))\n raise SLUCSyntaxError(\"ERROR: Missing Right Paren on line {0}\".format(str(self.currtok[2] - 1)))\n raise SLUCSyntaxError(\"ERROR: Missing Left Paren on line {0}\".format(str(self.currtok[2] - 1)))","def test_cot():\n c=0.5\n\n def myfunc(x):\n f1=EF.cot(x)\n return f1\n\n f_obj=ADiff(myfunc)\n res=f_obj.Jac(c)\n expectAns={'diff':2/(math.cos(c*2)-1), 'value':math.cos(c)/math.sin(c)}\n assert res==expectAns","def setup_callable_class(builder: IRBuilder) -> None:\n # Check to see that the name has not already been taken. If so,\n # rename the class. We allow multiple uses of the same function\n # name because this is valid in if-else blocks. Example:\n #\n # if True:\n # def foo(): ----> foo_obj()\n # return True\n # else:\n # def foo(): ----> foo_obj_0()\n # return False\n name = base_name = \"{}_obj\".format(builder.fn_info.namespaced_name())\n count = 0\n while name in builder.callable_class_names:\n name = base_name + \"_\" + str(count)\n count += 1\n builder.callable_class_names.add(name)\n\n # Define the actual callable class ClassIR, and set its\n # environment to point at the previously defined environment\n # class.\n callable_class_ir = ClassIR(name, builder.module_name, is_generated=True)\n\n # The functools @wraps decorator attempts to call setattr on\n # nested functions, so we create a dict for these nested\n # functions.\n # https://github.com/python/cpython/blob/3.7/Lib/functools.py#L58\n if builder.fn_info.is_nested:\n callable_class_ir.has_dict = True\n\n # If the enclosing class doesn't contain nested (which will happen if\n # this is a toplevel lambda), don't set up an environment.\n if builder.fn_infos[-2].contains_nested:\n callable_class_ir.attributes[ENV_ATTR_NAME] = RInstance(\n builder.fn_infos[-2].env_class\n )\n callable_class_ir.mro = [callable_class_ir]\n builder.fn_info.callable_class = ImplicitClass(callable_class_ir)\n builder.classes.append(callable_class_ir)\n\n # Add a 'self' variable to the environment of the callable class,\n # and store that variable in a register to be accessed later.\n self_target = add_self_to_env(builder.environment, callable_class_ir)\n builder.fn_info.callable_class.self_reg = builder.read(\n self_target, builder.fn_info.fitem.line\n )","def init():\n return _libsbml.SBMLFunctionDefinitionConverter_init()","def C(classname):\n return objc.objc_getClass(_utf8(classname))","def _get_ciphering_function(algorithm: Algorithm) -> Callable:\n ciphering_function = None\n if algorithm == Algorithm.caesar:\n ciphering_function = cifra.cipher.caesar.cipher\n elif algorithm == Algorithm.affine:\n ciphering_function = cifra.cipher.affine.cipher\n elif algorithm == Algorithm.substitution:\n ciphering_function = cifra.cipher.substitution.cipher\n elif algorithm == Algorithm.transposition:\n ciphering_function = cifra.cipher.transposition.cipher\n elif algorithm == Algorithm.vigenere:\n ciphering_function = cifra.cipher.vigenere.cipher\n return ciphering_function","def construction(self):\n from sage.categories.pushout import AlgebraicClosureFunctor\n return (AlgebraicClosureFunctor(), self._real_field())","def make_fee_class():\r\n fee = 1\r\n return make_class(locals())","def functions(self):\n return functions(self.startEA, self.endEA)","def python_function(self):\n return self._python_function # pylint: disable=protected-access","def get_class(self):\n\t\treturn self.CLASS","def getCall(self):\n if hasattr(self,'_filterfunctype'):\n return self._filterfunctype\n else:\n return None","def __init__(self, fn: callable):\n self.fn = fn","def functionURI(self):\n ret = libxml2mod.xmlXPathGetFunctionURI(self._o)\n return ret","def from_function(cls, py_func, py_file):\n raise NotImplementedError","def from_function(cls, py_func, py_file):\n raise NotImplementedError","def _get_flint_func(cls, domain):\n if domain == ZZ:\n return flint.fmpz_mat\n elif domain == QQ:\n return flint.fmpq_mat\n else:\n raise NotImplementedError(\"Only ZZ and QQ are supported by DFM\")","def get_func(func_ea):\n if isinstance(func_ea, idaapi.func_t):\n return func_ea\n func = idaapi.get_func(func_ea)\n if func is None:\n raise exceptions.SarkNoFunction(\"No function at 0x{:08X}\".format(func_ea))\n\n return func","def get_fns(self):\n return self.key_to_fn","def __init__(self, f, callback=None, errback=None):\n self.f = f\n self.exception = None\n self.result = None\n self.callback = callback\n self.errback = errback","def __init__(self, func, type):\n self.func = func\n self.type = type","def getInstance(self):\n return FSDecorator.instance","def function(self):\n return self.devicefuncs[self._funcname]","def _make_callable(func):\n try:\n return func.evaluator()\n except AttributeError:\n return func","def fn(self):\n return self._fn","def fortran_functions(self) -> str:\n fun_name = f'{self.f_prefix}_{self.name}_to_int_'\n result = f'integer function {fun_name}(flags)\\n'\n result += ' implicit none\\n'\n result += '\\n'\n result += f' class({self.f_prefix}_{self.name}), intent(in) :: flags\\n'\n result += ' integer :: ret_val\\n'\n result += '\\n'\n result += ' ret_val = 0\\n'\n for i, value in enumerate(self.values):\n result += f\" if (flags%{value}) ret_val = ret_val + {1 << i}\\n\"\n result += f' {fun_name} = ret_val\\n'\n result += f'end function {fun_name}\\n'\n return indent(result, 4*' ')","def get_func_lookup():\n return {\n \"randomstr\": randomstr,\n \"random\": random,\n \"sha256\": sha256,\n \"ed25519\": ed25519_private_key,\n \"rsa\": rsa_private_key,\n \"rsapublic\": rsa_public_key,\n \"publickey\": public_key,\n \"reveal\": reveal,\n \"loweralphanum\": loweralphanum,\n \"basicauth\": basicauth,\n }","def func_in_class():\n return \"just a function hanging out in a class\"","def get_function(self):\n return self.element.get_basis_functions()[self.n]","def test_sin():\n c=14\n def myfunc(x):\n f1=EF.sin(x)\n return f1\n\n f_obj=ADiff(myfunc)\n res=f_obj.Jac(c)\n\n expectAns={'diff': math.cos(c), 'value': math.sin(c)}\n assert res==expectAns","def __call__(self, function: FuncSpeechArg):\n self._add_attr(function)\n return function","def get_const_func(const):\n # pylint: disable=unused-argument\n def env(t):\n return const\n\n return env","def ifc_definition(cls):\n definitions[cls.__name__.upper()] = cls\n return cls","def __init__(self, function, function_representation):\n self.function = function\n self.function_representation = function_representation","def get_code(self):\n if self.code is None:\n try:\n import playhouse.sqlite_ext as sqlext\n except:\n raise KGTKException(f\"you need to 'pip install peewee' to use table-valued functions\")\n # multiple-inheritance breaks things, so we inherit the relevant information by hand:\n self.code = type(self.get_name(), (sqlext.TableFunction,), {\n 'name': self.get_name(),\n 'params': self.params,\n 'columns': self.columns,\n 'initialize': getattr(type(self), 'initialize'),\n 'iterate': getattr(type(self), 'iterate'),\n 'compute_result_rows': getattr(type(self), 'compute_result_rows'),\n })\n return self.code","def get_class():\n return CuBoltServerScript","def eurologic_backplane_function(self):\n return self._eurologic_backplane_function","def get_corrmat(self, f):\n return self._get_corrmat(f)","def res_cls(self):\n return self._res_cls","def fortran_c_wrapper(self) -> str:\n result = banner('//')\n result += self._fc_includes()\n result += self._fc_using_statements()\n result += self._fc_function_definitions()\n return result","def result_as_arg(self, node, C_new):\n F_new = C_new.clone()\n\n # Fortran function should wrap the new C function\n F_new._PTR_F_C_index = C_new._function_index\n F_new.wrap.assign(fortran=True)\n # Do not add '_bufferify'\n F_new.fmtdict.function_suffix = node.fmtdict.function_suffix\n\n # Do not wrap original function (does not have result argument)\n node.wrap.fortran = False\n return F_new"],"string":"[\n \"def _function_element_class(self):\\n return FriCASFunctionElement\",\n \"def func(self):\\n return self.__class__\",\n \"def choose_botorch_acqf_class() -> Type[AcquisitionFunction]:\\n # NOTE: In the future, this dispatch function could leverage any\\n # of the attributes of `BoTorchModel` or kwargs passed to\\n # `BoTorchModel.gen` to intelligently select acquisition function.\\n return qNoisyExpectedImprovement\",\n \"def fortran_c_wrapper(self) -> str:\\n if self.fc_override is not None:\\n return self.fc_override.replace('$CLASSNAME$', self.class_name).replace(\\n \\\"$C_PREFIX$\\\", self.c_prefix).replace(\\\"$F_PREFIX$\\\", self.f_prefix)\\n\\n result = ''\\n\\n # declaration\\n in_parameters = self._fc_in_parameters()\\n return_type, out_parameters = self._fc_out_parameters()\\n if self.may_throw:\\n out_parameters.append('int * err_code')\\n out_parameters.append('char ** err_msg')\\n out_parameters.append('std::size_t * err_msg_len')\\n\\n func_name = '{}_{}_{}_'.format(\\n self.c_prefix, self.class_name, self.name)\\n\\n par_str = ', '.join(in_parameters + out_parameters)\\n result += '{} {}({}) {{\\\\n'.format(return_type, func_name, par_str)\\n\\n # convert input\\n for par in self.params:\\n result += '{}'.format(par.fc_convert_input())\\n\\n # call C++ function and return result\\n if self.may_throw:\\n result += ' try {\\\\n'\\n result += ' *err_code = 0;\\\\n'\\n result += indent(self._fc_cpp_call(), 4*' ')\\n result += indent(self._fc_return(), 4*' ')\\n result += ' }\\\\n'\\n for exception, code in error_codes.items():\\n if code != 0:\\n catch = ''\\n catch += 'catch (std::{} const & e) {{\\\\n'.format(exception)\\n catch += ' *err_code = {};\\\\n'.format(code)\\n catch += ' static std::string msg;\\\\n'\\n catch += ' msg = e.what();\\\\n'\\n catch += ' *err_msg = const_cast<char*>(msg.data());\\\\n'\\n catch += ' *err_msg_len = msg.size();\\\\n'\\n catch += '}\\\\n'\\n result += indent(catch, 4*' ')\\n result += self._fc_return_default()\\n else:\\n result += self._fc_cpp_call()\\n result += self._fc_return()\\n result += '}\\\\n\\\\n'\\n return result\",\n \"def _func(self):\\n return self._get_flint_func(self.domain)\",\n \"def getFunctionClass(functionID):\\n d = { 1: Linear,\\n 2: LinearDrag,\\n 11: Gaussian,\\n 12: GaussianDrag,\\n 21: Lorentzian,\\n 22: LorentzianDrag }\\n return d[functionID]\",\n \"def getFunction(self) -> ghidra.program.model.listing.Function:\\n ...\",\n \"def _make_cas_function():\\n # Generate IR\\n mod = lc.Module.new('generate-cas')\\n llint = lc.Type.int()\\n llintp = lc.Type.pointer(llint)\\n fnty = lc.Type.function(llint, [llintp, llint, llint])\\n fn = mod.add_function(fnty, name='.numba.parallel.ufunc.cas')\\n ptr, old, repl = fn.args\\n bb = fn.append_basic_block('')\\n builder = lc.Builder.new(bb)\\n outpack = builder.cmpxchg(ptr, old, repl, ordering='monotonic')\\n out = builder.extract_value(outpack, 0)\\n failed = builder.extract_value(outpack, 1)\\n builder.ret(builder.select(failed, old, out))\\n\\n # Build & Link\\n llmod = ll.parse_assembly(str(mod))\\n\\n target = ll.Target.from_triple(ll.get_process_triple())\\n tm = target.create_target_machine()\\n engine = ll.create_mcjit_compiler(llmod, tm)\\n ptr = engine.get_function_address(fn.name)\\n return engine, ptr\",\n \"def arg_to_CFI(self, node, ordered_functions):\\n options = node.options\\n fmt_func = node.fmtdict\\n\\n if options.wrap_fortran is False:\\n # The buffer function is intended to be called by Fortran.\\n # No Fortran, no need for buffer function.\\n return\\n\\n ast = node.ast\\n declarator = ast.declarator\\n result_typemap = ast.typemap\\n # shadow classes have not been added yet.\\n # Only care about string, vector here.\\n result_is_ptr = declarator.is_indirect()\\n if (\\n result_typemap\\n and result_typemap.base in [\\\"string\\\", \\\"vector\\\"]\\n and result_typemap.name != \\\"char\\\"\\n and not result_is_ptr\\n ):\\n node.wrap.c = False\\n # node.wrap.fortran = False\\n self.config.log.write(\\n \\\"Skipping {}, unable to create C wrapper \\\"\\n \\\"for function returning {} instance\\\"\\n \\\" (must return a pointer or reference).\\\"\\n \\\" Bufferify version will still be created.\\\\n\\\".format(\\n result_typemap.cxx_type, declarator.user_name\\n )\\n )\\n \\n cfi_args = {}\\n for arg in ast.declarator.params:\\n declarator = arg.declarator\\n name = declarator.user_name\\n attrs = declarator.attrs\\n meta = declarator.metaattrs\\n cfi_args[name] = False\\n arg_typemap = arg.typemap\\n if meta[\\\"api\\\"]:\\n # API explicitly set by user.\\n continue\\n elif meta[\\\"assumed-rank\\\"]:\\n cfi_args[name] = True\\n elif attrs[\\\"rank\\\"]:\\n cfi_args[name] = True\\n elif arg_typemap.sgroup == \\\"string\\\":\\n cfi_args[name] = True\\n elif arg_typemap.sgroup == \\\"char\\\":\\n if declarator.is_indirect():\\n cfi_args[name] = True\\n elif meta[\\\"deref\\\"] in [\\\"allocatable\\\", \\\"pointer\\\"]:\\n cfi_args[name] = True\\n has_cfi_arg = any(cfi_args.values())\\n\\n # Function result.\\n need_buf_result = None\\n\\n result_as_arg = \\\"\\\" # Only applies to string functions\\n # when the result is added as an argument to the Fortran api.\\n\\n # Check if result needs to be an argument.\\n declarator = ast.declarator\\n attrs = declarator.attrs\\n meta = declarator.metaattrs\\n if meta[\\\"deref\\\"] == \\\"raw\\\":\\n # No bufferify required for raw pointer result.\\n pass\\n elif result_typemap.sgroup == \\\"string\\\":\\n need_buf_result = \\\"cfi\\\"\\n result_as_arg = fmt_func.F_string_result_as_arg\\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\\n elif result_typemap.sgroup == \\\"char\\\" and result_is_ptr:\\n need_buf_result = \\\"cfi\\\"\\n result_as_arg = fmt_func.F_string_result_as_arg\\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\\n elif meta[\\\"deref\\\"] in [\\\"allocatable\\\", \\\"pointer\\\"]:\\n need_buf_result = \\\"cfi\\\"\\n\\n if not (need_buf_result or\\n has_cfi_arg):\\n return False\\n\\n options.wrap_fortran = False\\n\\n # Create a new C function and change arguments\\n # and add attributes.\\n C_new = node.clone()\\n ordered_functions.append(C_new)\\n self.append_function_index(C_new)\\n\\n generated_suffix = \\\"cfi\\\"\\n C_new._generated = \\\"arg_to_cfi\\\"\\n C_new.splicer_group = \\\"cfi\\\"\\n if need_buf_result:\\n C_new.ast.declarator.metaattrs[\\\"api\\\"] = need_buf_result\\n fmt_func = C_new.fmtdict\\n fmt_func.function_suffix = fmt_func.function_suffix + fmt_func.C_cfi_suffix\\n\\n C_new.wrap.assign(c=True)#, fortran=True)\\n C_new._PTR_C_CXX_index = node._function_index\\n\\n for arg in C_new.ast.declarator.params:\\n name = arg.declarator.user_name\\n if cfi_args[name]:\\n arg.declarator.metaattrs[\\\"api\\\"] = generated_suffix\\n\\n ast = C_new.ast\\n if True: # preserve to avoid changing indention for now.\\n f_attrs = node.ast.declarator.attrs # Fortran function attributes\\n f_meta = node.ast.declarator.metaattrs # Fortran function attributes\\n if result_as_arg:\\n # decl: const char * getCharPtr2() +len(30)\\n # +len implies copying into users buffer.\\n result_as_string = ast.result_as_arg(result_name)\\n result_as_string.const = False # must be writeable\\n attrs = result_as_string.declarator.attrs\\n # Special case for wrapf.py to override \\\"allocatable\\\"\\n f_meta[\\\"deref\\\"] = None\\n result_as_string.declarator.metaattrs[\\\"api\\\"] = \\\"cfi\\\"\\n result_as_string.declarator.metaattrs[\\\"deref\\\"] = \\\"result\\\"\\n result_as_string.declarator.metaattrs[\\\"is_result\\\"] = True\\n C_new.ast.declarator.metaattrs[\\\"api\\\"] = None\\n C_new.ast.declarator.metaattrs[\\\"intent\\\"] = \\\"subroutine\\\"\\n C_new.ast.declarator.metaattrs[\\\"deref\\\"] = None\\n\\n if result_as_arg:\\n F_new = self.result_as_arg(node, C_new)\\n ordered_functions.append(F_new)\\n self.append_function_index(F_new)\\n else:\\n if node._generated in [\\\"result_to_arg\\\", \\\"fortran_generic\\\", \\\"getter/setter\\\"]:\\n node.wrap.c = False\\n # Fortran function may call C subroutine if string/vector result\\n # Fortran function calls bufferify function.\\n node._PTR_F_C_index = C_new._function_index\\n return True\",\n \"def _fc_function_definitions(self) -> str:\\n result = 'extern \\\"C\\\" {\\\\n\\\\n'\\n for namespace in self.namespaces:\\n for member in namespace.members:\\n result += member.fortran_c_wrapper()\\n\\n result += '}\\\\n\\\\n'\\n return result\",\n \"def func(cls):\\n return cls.get_wrapper()(cls.callable)\",\n \"def get_function(self):\\n return SSAFunction(self.get_graph())\",\n \"def createFunctionTerm(self):\\n return _libsbml.Transition_createFunctionTerm(self)\",\n \"def get_function(self):\\n raise NotImplementedError()\",\n \"def get_tested_func(self):\\n t_ast = [x for x in self.AST.ext if type(x) is c_ast.FuncDef and x.decl.name == self.Fname]\\n if len(t_ast) < 1:\\n raise Exception(\\\"Func is not in ast\\\")\\n else:\\n return t_ast[0]\",\n \"def test_cos():\\n c=14\\n def myfunc(x):\\n f1=EF.cos(x)\\n return f1\\n\\n f_obj=ADiff(myfunc)\\n res=f_obj.Jac(c)\\n\\n expectAns={'diff': -math.sin(c), 'value': math.cos(c)}\\n\\n assert res==expectAns\",\n \"def __init__(self, function='cogscore/'):\\n self.function = function\",\n \"def get_func_type(self, *args):\\n return _ida_hexrays.cfunc_t_get_func_type(self, *args)\",\n \"def fortran_function(self) -> str:\\n if self.f_override is not None:\\n return indent(\\n self.f_override.replace('$CLASSNAME$', self.class_name).replace(\\n \\\"$C_PREFIX$\\\", self.c_prefix).replace(\\n \\\"$F_PREFIX$\\\", self.f_prefix),\\n 4*' ')\\n\\n result = ''\\n\\n # declaration\\n func_name = '{}_{}_{}'.format(\\n self.f_prefix, self.class_name, self.name)\\n in_parameters = self._f_in_parameters()\\n return_type, out_parameters = self._f_out_parameters()\\n if self.may_throw:\\n out_parameters.append(('integer, optional', 'err_code'))\\n out_parameters.append(('character(:), allocatable, optional',\\n 'err_msg'))\\n\\n all_parameters = in_parameters + out_parameters\\n arg_list = ', &\\\\n'.join([par_name for _, par_name in all_parameters])\\n arg_ilist = indent(arg_list, 8*' ')\\n if return_type != '':\\n result += 'function {}( &\\\\n{})\\\\n'.format(func_name, arg_ilist)\\n else:\\n result += 'subroutine {}( &\\\\n{})\\\\n'.format(func_name, arg_ilist)\\n\\n # parameter declarations\\n result += ' implicit none\\\\n'\\n for par_type, par_name in in_parameters:\\n result += ' {}, intent(in) :: {}\\\\n'.format(\\n par_type, par_name)\\n for par_type, par_name in out_parameters:\\n result += ' {}, intent(out) :: {}\\\\n'.format(par_type, par_name)\\n if return_type != '':\\n result += ' {} :: {}\\\\n'.format(return_type, func_name)\\n result += '\\\\n'\\n\\n # variable declarations\\n c_return_type, fi_out_parameters = self._fi_out_parameters()\\n if c_return_type:\\n result += ' {} :: ret_val\\\\n'.format(c_return_type)\\n for par_type, par_name in fi_out_parameters:\\n result += ' {} :: {}\\\\n'.format(par_type, par_name)\\n for par_type, par_name in self.ret_type.f_aux_variables():\\n result += ' {} :: {}\\\\n'.format(par_type, par_name)\\n if self.may_throw:\\n result += ' integer (c_int) :: err_code_v\\\\n'\\n result += ' type (c_ptr) :: err_msg_v\\\\n'\\n result += ' integer (c_size_t) :: err_msg_len_v\\\\n'\\n result += ' character (c_char), dimension(:), pointer :: err_msg_f\\\\n'\\n result += ' character(:), allocatable :: err_msg_p\\\\n'\\n result += ' integer (c_size_t) :: err_msg_i\\\\n'\\n if c_return_type or fi_out_parameters or self.may_throw:\\n result += '\\\\n'\\n\\n # convert input\\n args = [param.f_chain_arg() for param in self.params]\\n args += [par_name for _, par_name in fi_out_parameters]\\n if self.may_throw:\\n args += ['err_code_v', 'err_msg_v', 'err_msg_len_v']\\n arg_str = ', &\\\\n'.join([8*' ' + arg for arg in args])\\n\\n # call C function\\n fc_func_name = '{}_{}_{}_'.format(\\n self.c_prefix, self.class_name, self.name)\\n chain_call = self.fc_chain_call(\\n ns_prefix=self.c_prefix, class_name=self.class_name,\\n fc_func_name=fc_func_name, fc_args=arg_str)\\n result_name = ''\\n if return_type != '':\\n result_name = func_name\\n elif out_parameters:\\n result_name = out_parameters[0][1]\\n result += self.ret_type.f_call_c('ret_val', chain_call)\\n\\n # handle errors if necessary\\n if self.may_throw:\\n # Note: I tried to factor this out into a function, but Fortran\\n # makes that near-impossible. Since we're generating anyway, it's\\n # not really duplication, so leave it as is.\\n result += indent(dedent(f\\\"\\\"\\\"\\\\\\n if (err_code_v .ne. 0) then\\n if (present(err_code)) then\\n err_code = err_code_v\\n if (present(err_msg)) then\\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\\n allocate (character(err_msg_len_v) :: err_msg)\\n do err_msg_i = 1, err_msg_len_v\\n err_msg(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\\n end do\\n end if\\n {dedent(\\n self.ret_type.f_return_dummy_result(result_name))}\\n return\\n else\\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\\n allocate (character(err_msg_len_v) :: err_msg_p)\\n do err_msg_i = 1, err_msg_len_v\\n err_msg_p(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\\n end do\\n print *, err_msg_p\\n stop\\n end if\\n else\\n if (present(err_code)) then\\n err_code = 0\\n end if\\n end if\\n\\n \\\"\\\"\\\"), 4*' ')\\n\\n # convert and return result\\n result += self.ret_type.f_return_result(result_name, 'ret_val')\\n\\n # end\\n if return_type != '':\\n result += 'end function {}\\\\n\\\\n'.format(func_name)\\n else:\\n result += 'end subroutine {}\\\\n\\\\n'.format(func_name)\\n return indent(result, 4*' ')\",\n \"def __init__(self, func_name, spin):\\n self.xc_func = None\\n self._xc_func_init = False\\n\\n # Handle func_name\\n if isinstance(func_name, str):\\n func_id = util.xc_functional_get_number(func_name)\\n if func_id == -1:\\n raise KeyError(\\\"LibXCFunctional: name '%s' not found.\\\" % func_name)\\n elif isinstance(func_name, (int, np.integer)):\\n func_id = func_name\\n if util.xc_functional_get_name(func_name) is None:\\n raise KeyError(\\\"LibXCFunctional: ID '%d' not found.\\\" % func_name)\\n else:\\n raise TypeError(\\\"LibXCFunctional: func_name must either be a string or int. Got {}\\\".format(func_name))\\n\\n self._xc_func_name = util.xc_functional_get_name(func_id)\\n\\n # Handle spin\\n if isinstance(spin, str):\\n spin = spin.lower()\\n if spin == \\\"polarized\\\":\\n self._spin = 2\\n elif spin == \\\"unpolarized\\\":\\n self._spin = 1\\n else:\\n raise KeyError(\\\"LibXCFunctional: spin must either be 'polarized' or 'unpolarized' if represented by a string. Got {}\\\".format(spin))\\n else:\\n self._spin = spin\\n\\n if self._spin not in [1, 2]:\\n raise KeyError(\\\"LibXCFunctional: spin must either be 1 or 2 if represented by a integer. Got {}\\\".format(self._spin))\\n\\n # Build the LibXC functional\\n self.xc_func = core.xc_func_alloc()\\n self.xc_func_size_names = [x for x in dir(self.xc_func.contents.dim) if not \\\"_\\\" in x]\\n\\n # Set all int attributes to zero (not all set to zero in libxc)\\n for attr in self.xc_func_size_names:\\n setattr(self.xc_func.contents, attr, 0)\\n\\n ret = core.xc_func_init(self.xc_func, func_id, self._spin)\\n if ret != 0:\\n raise ValueError(\\\"LibXC Functional construction did not complete. Error code %d\\\" % ret)\\n self._xc_func_init = True\\n\\n # Pull out all sizes after init\\n self.xc_func_sizes = {}\\n for attr in self.xc_func_size_names:\\n self.xc_func_sizes[attr] = getattr(self.xc_func.contents.dim, attr)\\n\\n # Unpack functional info\\n self.xc_func_info = core.xc_func_get_info(self.xc_func)\\n self._number = core.xc_func_info_get_number(self.xc_func_info)\\n self._kind = core.xc_func_info_get_kind(self.xc_func_info)\\n self._name = core.xc_func_info_get_name(self.xc_func_info).decode(\\\"UTF-8\\\")\\n self._family = core.xc_func_info_get_family(self.xc_func_info)\\n self._flags = core.xc_func_info_get_flags(self.xc_func_info)\\n\\n # Set needed flags\\n self._needs_laplacian = self._flags & flags.XC_FLAGS_NEEDS_LAPLACIAN\\n\\n # Set derivatives\\n self._have_exc = self._flags & flags.XC_FLAGS_HAVE_EXC\\n self._have_vxc = self._flags & flags.XC_FLAGS_HAVE_VXC\\n self._have_fxc = self._flags & flags.XC_FLAGS_HAVE_FXC\\n self._have_kxc = self._flags & flags.XC_FLAGS_HAVE_KXC\\n self._have_lxc = self._flags & flags.XC_FLAGS_HAVE_LXC\\n\\n # Set omega\\n self._have_cam = self._flags & flags.XC_FLAGS_HYB_CAM\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_CAMY\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LC\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LCY\\n self._cam_omega = self._cam_alpha = self._cam_beta = False\\n if self._have_cam:\\n self._cam_omega = self.xc_func.contents.cam_omega\\n self._cam_alpha = self.xc_func.contents.cam_alpha\\n self._cam_beta = self.xc_func.contents.cam_beta\\n\\n elif self._family in [flags.XC_FAMILY_HYB_LDA, flags.XC_FAMILY_HYB_GGA, flags.XC_FAMILY_HYB_MGGA]:\\n self._cam_alpha = self.xc_func.contents.cam_alpha\\n\\n # VV10\\n self._have_vv10 = self._flags & flags.XC_FLAGS_VV10\\n self._nlc_b = self._nlc_C = False\\n if self._have_vv10:\\n self._nlc_b = self.xc_func.contents.nlc_b\\n self._nlc_C = self.xc_func.contents.nlc_C\\n\\n # Stable\\n self._stable = self._flags & flags.XC_FLAGS_STABLE\\n self._dev = self._flags & flags.XC_FLAGS_DEVELOPMENT\\n\\n # Pull out references\\n self._refs = []\\n self._bibtexs = []\\n self._dois = []\\n\\n for pos in range(flags.XC_MAX_REFERENCES):\\n ref = core.xc_func_info_get_references(self.xc_func_info, pos)\\n if not ref: break\\n\\n self._refs.append(ref.contents.ref.decode(\\\"UTF-8\\\"))\\n self._bibtexs.append(ref.contents.bibtex.decode(\\\"UTF-8\\\"))\\n self._dois.append(ref.contents.doi.decode(\\\"UTF-8\\\"))\",\n \"def get_unc_f(self):\\n return self.uncf\",\n \"def __init__(self, expected, test_func):\\n self._f = test_func\\n self._exp = expected\",\n \"def cfunc_type(self):\\n tif = ida_typeinf.tinfo_t()\\n result = self.get_func_type(tif)\\n if not result:\\n return\\n return tif\",\n \"def __init__(self,\\n function: Callable):\\n\\n self._function = function\",\n \"def _get_class():\\n return ASParameters\",\n \"def test_arccos():\\n c=0.5\\n def myfunc(x):\\n f1=EF.arccos(x)\\n return f1\\n\\n f_obj=ADiff(myfunc)\\n res=f_obj.Jac(c)\\n expectAns={'diff':-1/math.sqrt(1-c**2), 'value':math.acos(c)}\\n assert res==expectAns\",\n \"def fortran_function(self) -> str:\\n return ''.join([i.fortran_function() for i in self.instances])\",\n \"def getColorTransferFunction(self):\\n\\t\\treturn self.ctf\",\n \"def _constructor(self):\\n return self.__class__\",\n \"def getFunction(self):\\n return gpio.gpio_function(self.bcm_id)\",\n \"def get_decorated_function(self):\",\n \"def getCallable():\",\n \"def __init__(self, func, args, kwargs, res): # make test\\n\\n # Check if func is a callable function\\n if callable(func):\\n self.func = func\\n else:\\n self.func = lambda *args, **kwargs: func\\n\\n # Store parameters in class\\n self.args = args or [] # If args == None, set self.args to an empty list\\n self.kwargs = kwargs or {} # If kwargs == None, set self.kwargs to an empty dict\\n self.res = res\\n\\n\\n self._tolerance = 1.0e-10 # Variable used to check for equality in case of float\",\n \"def get_ea(func_ea):\\n if isinstance(func_ea, idaapi.func_t):\\n return func_ea.startEA\\n return func_ea\",\n \"def __init__(self, function: ast3.FunctionDef, module: \\\"Module\\\") -> None:\\n\\n # easy data\\n self._function = function\\n self.name = function.name\\n self.line = function.lineno\\n self.column = function.col_offset\\n self.body = function.body\\n self.module = module\\n self.decorators = function.decorator_list\\n\\n # time to parse arguments\\n self._args = function.args.args\\n self.args = [Arg(x) for x in self._args]\\n self.functions = [\\n Function(x, self.module)\\n for x in self.body\\n if isinstance(x, ast3.FunctionDef)\\n ]\\n self.classes = [\\n Class(x, self.module) for x in self.body if isinstance(x, ast3.ClassDef)\\n ]\\n self.untyped = [\\n x for x in self.args if not x.typed and x not in DEFAULT_ARG_IGNORE\\n ]\\n self.doc = None\\n self.returns = None\\n self.return_typed = False\\n self.missing_args: Set[str] = set()\\n self.unexpected_args: Set[str] = set()\\n arg_names = set(x.name for x in self.args if x.name not in DEFAULT_ARG_IGNORE)\\n self.missing_args = arg_names\\n if isinstance(self.body[0], ast3.Expr):\\n # this is most likely a doc string\\n self.doc = Doc(self.body[0], Doc.Type.FUNCTION)\\n doc_arg_names = set(x for x, y in self.doc.args.items())\\n self.missing_args = arg_names - doc_arg_names\\n self.unexpected_args = doc_arg_names - arg_names\\n if function.returns:\\n self.return_typed = True\\n self.returns = parse_elt(function.returns) # type: ignore\\n\\n # complexity checks\\n self._radon = cc_visit_ast(self._function)[0]\\n self.complexity = self._radon.complexity\\n self.is_method = self._radon.is_method\\n self._halstead = h_visit_ast(self._function)\",\n \"def __call__(self, f_or_klass):\\n\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\\n # 0.3.0\\n # -- implement \\\"kill switch\\\", NO_DECO\\n # -- handle decorating both functions and classes\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\\n\\n # 0.3.0b16: if it isn't callable, scram'\\n if not callable(f_or_klass):\\n return f_or_klass\\n\\n # Special-case handling for ``NO_DECO``: remove from settings of ``self``\\n if self._effective_settings.get('NO_DECO'):\\n return f_or_klass\\n # else, delete that item wherever it might be\\n if 'NO_DECO' in self._effective_settings:\\n del self._effective_settings['NO_DECO']\\n if 'NO_DECO' in self._changed_settings:\\n del self._changed_settings['NO_DECO']\\n\\n f = f_or_klass if inspect.isfunction(f_or_klass) else None\\n klass = f_or_klass if inspect.isclass(f_or_klass) else None\\n\\n self.f = f\\n self.cls = klass\\n\\n if klass:\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a class\\\" -- namely, klass\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n\\n self._class__call__(klass) # modifies klass (methods & inner classes) (if not builtin)\\n self._add_class_attrs(klass) # v0.3.0v20 traps TypeError for builtins\\n return klass\\n\\n elif not f:\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a callable but not a function\\\"\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # functools.partial objects are callable, have no __name__ much less __qualname__,\\n # and trying to deco __call__ gets messy.\\n # Callable builtins e.g. len are not functions in the isfunction sense,\\n # can't deco anyway. Just give up (quietly):\\n return f_or_klass\\n\\n else: # not a class, f nonempty is a function of f_or_klass callable\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a function\\\" -- namely, f\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n\\n #----------------------------------------------------------------\\n # Don't double-decorate -- don't wanna, & it doesn't work anyway!\\n #----------------------------------------------------------------\\n # Note: As with methods of classes,\\n # . if f is deco'd, its existing EXPLICITLY GIVEN settings take precedence.\\n\\n # # From _class__call__, props & methods cases, w/a few name changes\\n deco_obj = getattr(f, self._sentinels['DECO_OF'], None) # type: _deco_base\\n\\n # get a fresh copy for each attr\\n new_settings = self._changed_settings.copy() # updated below\\n\\n # __init__ fixup, a nicety:\\n # By default, don't log retval for __init__.\\n # If user insists on it with 'log_retval=True' in __init__ deco,\\n # that will override this.\\n if f.__name__ == '__init__':\\n self.fixup_for_init(new_settings)\\n\\n if deco_obj: # f is deco'd by this decorator\\n # Yes. Figure out settings for f,\\n ### 0.3.0b18 -- Use self._override\\n self._update_settings(new=new_settings,\\n old=deco_obj._changed_settings,\\n override_existing=self._override)\\n # update func's settings (_force_mutable=True to handle `max_history` properly)\\n deco_obj._settings_mapping.update(new_settings, _force_mutable=True)\\n return f\\n\\n #----------------------------------------------------------------\\n # f is a function & is NOT already deco'd\\n #----------------------------------------------------------------\\n\\n # 0.3.0.x -- f may not have a .__qualname__\\n try:\\n self._classname_of_f = '.'.join( f.__qualname__.split('.')[:-1] )\\n except AttributeError as e:\\n self._classname_of_f = ''\\n\\n # Special-case '__repr__' handling, if deco subclass doesn't allow it.\\n if f.__name__ == '__repr__' and self._classname_of_f and not self.allow_repr():\\n # v0.3.0b23 -- Instead of refusing to deco, use recursive_repr\\n # return f\\n return recursive_repr(fillvalue=\\\"...\\\")(f)\\n\\n # 0.3.0\\n # Use __qualname__ ALL the time, unless user provides `name=display_name_str`\\n # where `display_name_str` is either the name to be used for the fn in logged output,\\n # or is an oldstyle format str into which f.__name__ will be substituted\\n # to obtain the display name.\\n # We require Py3.3+, so __qualname__ is available.\\n\\n # setup f_display_name\\n if self._name_param:\\n try:\\n self.f_display_name = (self._name_param % f.__name__)\\n except TypeError:\\n self.f_display_name = self._name_param\\n else:\\n self.f_display_name = f.__qualname__\\n\\n # TODO TRY THIS -- anything break?\\n # 0.3.1 Inspired by fractions.Fraction.__sub__ et al:\\n # __name__ may be very different from __qualname__;\\n # if so, show both\\n if f.__name__ not in f.__qualname__:\\n self.f_display_name += \\\" (\\\" + f.__name__ + \\\")\\\"\\n\\n #================================================================\\n # 0.3.0 -- Init things (migrated from __init__)\\n #----------------------------------------------------------------\\n # set up pseudo-dict (DecoSettingsMapping),\\n # using settings given by self._effective_settings.\\n #\\n # *** DecoSettingsMapping \\\"API\\\" --\\n # (2) construct DecoSettingsMapping object\\n # that will provide mapping & attribute access to settings, & more\\n #----------------------------------------------------------------\\n self._settings_mapping = DecoSettingsMapping(\\n deco_class=self.__class__,\\n # DecoSettingsMapping calls the rest ** values_dict\\n ** self._effective_settings # 0.3.0 set by __init__\\n )\\n\\n #----------------------------------------------------------------\\n # Init more stuff\\n #----------------------------------------------------------------\\n self._stats = ClassInstanceAttrProxy(\\n class_instance=self,\\n data_descriptor_names=self.__class__._data_descriptor_names,\\n method_descriptor_names=self.__class__._method_descriptor_names)\\n # Accessed by descriptors on the stats obj\\n self._num_calls_total = 0\\n self._num_calls_logged = 0\\n # max_history > 0 --> size of self._call_history; <= 0 --> unbounded\\n # Set before calling _make_call_history\\n\\n # 0.3.0 self._other_values_dict set by __init__\\n self.max_history = self._other_values_dict.get('max_history', 0) # <-- Nota bene\\n self._call_history = self._make_call_history()\\n\\n # Accumulate this (for logged calls only)\\n # even when record_history is false:\\n self._elapsed_secs_logged = 0.0\\n self._process_secs_logged = 0.0\\n\\n # 0.2.2.post1\\n # stack(s), pushed & popped wrapper of deco'd function\\n # by _logging_state_push, _logging_state_pop\\n # 0.3.0 convert to pushing/popping single namedtuples\\n self.logging_state_stack = [] # 0.3.0 stack of LoggingState namedtuples\\n self._enabled_stack = [] # 0.3.0 - um, stack, of 'enabled's\\n\\n #----------------------------------------------------------------\\n # end of Init passage\\n #================================================================\\n\\n # Save signature and parameters of f\\n self.f_signature = inspect.signature(f) # Py >= 3.3\\n self.f_params = self.f_signature.parameters\\n\\n # 0.3.0 We assume Py3.3 so we use perf_counter, process_time all the time\\n wall_time_fn = time.perf_counter\\n process_time_fn = time.process_time\\n\\n #############################\\n # The wrapper of a callable\\n #############################\\n\\n @wraps(f)\\n def _deco_base_f_wrapper_(*args, **kwargs):\\n \\\"\\\"\\\"Wrapper around the wrapped function f.\\n When this runs, f has been called, so we can now resolve\\n any indirect values for the settings/keyword-params\\n of log_calls, using info in kwargs and self.f_params.\\\"\\\"\\\"\\n # *** Part of the DecoSettingsMapping \\\"API\\\" --\\n # (4) using self._settings_mapping.get_final_value in wrapper\\n # [[[ This/these is/are 4th chronologically ]]]\\n\\n # inner/local fn -- save a few cycles and characters -\\n # we call this a lot (<= 9x).\\n def _get_final_value(setting_name):\\n \\\"Use outer scope's kwargs and self.f_params\\\"\\n return self._settings_mapping.get_final_value(\\n setting_name, kwargs, fparams=self.f_params)\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # if nothing to do, hurry up & don't do it.\\n # NOTE: call_chain_to_next_log_calls_fn looks in stack frames\\n # to find (0.2.4) STACKFRAME_HACK_DICT_NAME (really!)\\n # It and its values (the following _XXX variables)\\n # must be set before calling f.\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n _enabled = _get_final_value('enabled')\\n # 0.3.0 in case f calls log_message (no output if f disabled)\\n self._enabled_state_push(_enabled)\\n\\n # 0.2.4.post5 \\\"true bypass\\\": if 'enabled' < 0 then scram\\n if _enabled < 0:\\n ret = f(*args, **kwargs)\\n self._enabled_state_pop()\\n return ret\\n\\n # Bump call counters, before calling fn.\\n # Note: elapsed_secs, process_secs not reflected yet of course\\n self._add_call(logged=_enabled)\\n\\n _log_call_numbers = _get_final_value('log_call_numbers')\\n # counters just got bumped\\n _active_call_number = (self._stats.num_calls_logged\\n if _log_call_numbers else\\n 0)\\n # Get list of callers up to & including first log_call's-deco'd fn\\n # (or just caller, if no such fn)\\n call_list, prev_indent_level = self.call_chain_to_next_log_calls_fn()\\n\\n # Bump _extra_indent_level if last fn on call_list is deco'd AND enabled,\\n # o/w it's the _extra_indent_level which that fn 'inherited'.\\n # _extra_indent_level: prev_indent_level, or prev_indent_level + 1\\n do_indent = _get_final_value('indent')\\n _extra_indent_level = (prev_indent_level +\\n int(not not do_indent and not not _enabled))\\n # 0.3.0\\n ########## prefixed_fname = _get_final_value('prefix') + f.__name__\\n prefixed_fname = _get_final_value('prefix') + self.f_display_name\\n\\n # Stackframe hack:\\n assert '_deco_base__active_call_items__' == STACKFRAME_HACK_DICT_NAME\\n _deco_base__active_call_items__ = {\\n '_enabled': _enabled,\\n '_log_call_numbers': _log_call_numbers,\\n '_prefixed_fname': prefixed_fname, # Hack alert (Pt 1)\\n '_active_call_number': _active_call_number,\\n '_extra_indent_level': _extra_indent_level,\\n # 0.3.0 for _get_own_deco_wrapper\\n '_wrapper_deco': self\\n }\\n\\n # Get logging function IF ANY.\\n # For the benefit of callees further down the call chain,\\n # if this f is not enabled (_enabled <= 0).\\n # Subclass can return None to suppress printed/logged output.\\n logging_fn = self.get_logging_fn(_get_final_value)\\n\\n # Only do global indentation for print, not for loggers\\n global_indent_len = max(_extra_indent_level, 0) * self.INDENT\\n\\n # 0.2.2.post1 - save output_fname for log_message use\\n call_number_str = ((' [%d]' % _active_call_number)\\n if _log_call_numbers else '')\\n output_fname = prefixed_fname + call_number_str\\n\\n # 0.3.0\\n # Note: DON'T combine with global_mute(),\\n # cuz this value will be pushed,\\n # and when popped any realtime changes to global mute\\n # made during call to f would be ignored.\\n mute = _get_final_value('mute')\\n\\n # 0.2.2 -- self._log_message() will use\\n # the logging_fn, indent_len and output_fname at top of these stacks;\\n # thus, verbose functions should use log_calls.print (~ log_message)\\n # to write their blather.\\n # There's a stack of logging-state ,\\n # used by self._log_message(), maintained in this wrapper.\\n self._logging_state_push(logging_fn, global_indent_len, output_fname, mute)\\n\\n # (_xxx variables set, ok to call f)\\n if not _enabled:\\n ret = f(*args, **kwargs)\\n self._logging_state_pop(enabled_too=True)\\n return ret\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Set up context, for pre-call handlers\\n # (after calling f, add to it for post-call handlers)\\n # THIS is the time sink - 23x slower than other 'blocks'\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Key/values of \\\"context\\\" whose values we know so far:\\n context = {\\n 'decorator': self,\\n 'settings': self._settings_mapping,\\n 'stats': self._stats,\\n 'prefixed_fname': prefixed_fname,\\n 'fparams': self.f_params,\\n 'call_list': call_list,\\n 'args': args,\\n 'kwargs': kwargs,\\n 'indent': \\\" \\\" * self.INDENT, # our unit of indentation\\n 'output_fname': output_fname,\\n }\\n\\n # Gather all the things we need (for log output, & for history)\\n # Use inspect module's Signature.bind method.\\n # bound_args.arguments -- contains only explicitly bound arguments\\n # 0.2.4.post5 - using\\n # inspect.signature(f).bind(*args, **kwargs)\\n # took 45% of execution time of entire wrapper; this takes 23%:\\n # 0.3.1 TODO BUG No args is a problem?!\\n bound_args = self.f_signature.bind(*args, **kwargs)\\n \\\"\\\"\\\"\\n File \\\"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/tests/_temp.py\\\", line 12, in <module>\\n g(f())\\n File \\\"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/log_calls.py\\\", line 1935, in _deco_base_f_wrapper_\\n bound_args = self.f_signature.bind(*args, **kwargs)\\n File \\\"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\\\", line 2646, in bind\\n return args[0]._bind(args[1:], kwargs)\\n File \\\"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\\\", line 2571, in _bind\\n raise TypeError('too many positional arguments') from None\\n TypeError: too many positional arguments\\n \\\"\\\"\\\"\\n\\n varargs_pos = get_args_pos(self.f_params) # -1 if no *args in signature\\n argcount = varargs_pos if varargs_pos >= 0 else len(args)\\n context['argcount'] = argcount\\n # The first argcount-many things in bound_args\\n context['argnames'] = list(bound_args.arguments)[:argcount]\\n context['argvals'] = args[:argcount]\\n\\n context['varargs'] = args[argcount:]\\n (context['varargs_name'],\\n context['kwargs_name']) = get_args_kwargs_param_names(self.f_params)\\n\\n # These 3 statements = 31% of execution time of wrapper\\n context['defaulted_kwargs'] = get_defaulted_kwargs_OD(self.f_params, bound_args)\\n context['explicit_kwargs'] = get_explicit_kwargs_OD(self.f_params, bound_args, kwargs)\\n # context['implicit_kwargs'] = {\\n # k: kwargs[k] for k in kwargs if k not in context['explicit_kwargs']\\n # }\\n # At least 2x as fast:\\n context['implicit_kwargs'] = \\\\\\n difference_update(kwargs.copy(), context['explicit_kwargs'])\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call pre-call handlers, collect nonempty return values\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # only consult global mute in r/t\\n if not (mute or self.global_mute()): # 0.3.0\\n pre_msgs = []\\n for setting_name in self._settings_mapping._pre_call_handlers: # keys\\n if _get_final_value(setting_name):\\n info = self._settings_mapping._get_DecoSetting(setting_name)\\n msg = info.pre_call_handler(context)\\n if msg:\\n pre_msgs.append(msg)\\n\\n # Write pre-call messages\\n if logging_fn:\\n for msg in pre_msgs:\\n self._log_message(msg, extra_indent_level=0)\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call f(*args, **kwargs) and get its retval; time it.\\n # Add timestamp, elapsed time(s) and retval to context.\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # No dictionary overhead between timer(s) start & stop.\\n t0 = time.time() # for timestamp\\n t0_wall = wall_time_fn()\\n t0_process = process_time_fn()\\n retval = f(*args, **kwargs)\\n t_end_wall = wall_time_fn()\\n t_end_process = process_time_fn()\\n context['elapsed_secs'] = (t_end_wall - t0_wall)\\n context['process_secs'] = (t_end_process - t0_process)\\n context['timestamp'] = t0\\n context['retval'] = retval\\n\\n self._add_to_elapsed(context['elapsed_secs'], context['process_secs'])\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call post-call handlers, collect nonempty return values\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # only consult global mute in r/t\\n if not (mute or self.global_mute()): # 0.3.0\\n post_msgs = []\\n for setting_name in self._settings_mapping._post_call_handlers: # keys\\n if _get_final_value(setting_name):\\n info = self._settings_mapping._get_DecoSetting(setting_name)\\n msg = info.post_call_handler(context)\\n if msg:\\n post_msgs.append(msg)\\n\\n # Write post-call messages\\n if logging_fn:\\n for msg in post_msgs:\\n self._log_message(msg, extra_indent_level=0)\\n # v0.3.0b22 -- if recording history, add record of call even if we're muted(!)\\n elif _get_final_value('record_history'):\\n info = self._settings_mapping._get_DecoSetting('record_history')\\n _ = info.post_call_handler(context)\\n\\n self._logging_state_pop(enabled_too=True)\\n\\n return retval\\n\\n self._add_function_attrs(f, _deco_base_f_wrapper_)\\n return _deco_base_f_wrapper_\\n\\n #-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+\\n # end else (case \\\"f_or_klass is a function\\\",\\n # subcase \\\"f is a function & is NOT already deco'd\\\")\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\",\n \"def as_ret(self):\\n if self.type in OCClass.classes_indexed_by_name:\\n oc_class = OCClass.classes_indexed_by_name[self.type]\\n if oc_class.class_meths:\\n for f in oc_class.class_meths:\\n #TODO MachOTask.currentTask.analyze_function(start_addr=f)\\n print 'T'\",\n \"def get_function(self):\\n return subs(self.f.get_function(), self.sub_pre, self.sub_post)\",\n \"def __init__(self, function):\\n self.function = function\",\n \"def get_func_type(self, *args):\\n return _ida_hexrays.cfuncptr_t_get_func_type(self, *args)\",\n \"def functional(self):\\n return self.__functional\",\n \"def get_real_function( this, fn):\\n\\t\\treturn this._get_native_function(fn)\",\n \"def createFunctionDefinition(self):\\n return _libsbml.Model_createFunctionDefinition(self)\",\n \"def f(self):\\n return self._f\",\n \"def get_ctf(ima):\\n\\tfrom EMAN2 import EMAN2Ctf\\n\\tctf_params = ima.get_attr(\\\"ctf\\\")\\t\\n\\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang\",\n \"def convert(self):\\n return _libsbml.SBMLFunctionDefinitionConverter_convert(self)\",\n \"def for_function(self, func):\\n return self.for_pysig(utils.pysignature(func))\",\n \"def getcfix(self):\\n cfix_ = ctypes.c_double()\\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\\n if res != 0:\\n _,msg = self.__getlasterror(res)\\n raise Error(rescode(res),msg)\\n cfix_ = cfix_.value\\n _cfix_return_value = cfix_\\n return (_cfix_return_value)\",\n \"def func ( self ) :\\n return self.__func\",\n \"def getcfix(self): # 3\\n res,resargs = self.__obj.getcfix()\\n if res != 0:\\n result,msg = self.__getlasterror(res)\\n raise Error(rescode(res),msg)\\n _cfix_return_value = resargs\\n return _cfix_return_value\",\n \"def classFactory(iface): # pylint: disable=invalid-name\\n #\\n from .CADDigitize import CADDigitize\\n return CADDigitize(iface)\",\n \"def __init__(self, function=None):\\n self._function = function\",\n \"def getTestClass(self):\\n\\t\\treturn AbstraccionPrueba\",\n \"def instantiate_callable_class(builder: IRBuilder, fn_info: FuncInfo) -> Value:\\n fitem = fn_info.fitem\\n func_reg = builder.add(Call(fn_info.callable_class.ir.ctor, [], fitem.line))\\n\\n # Set the environment attribute of the callable class to point at\\n # the environment class defined in the callable class' immediate\\n # outer scope. Note that there are three possible environment\\n # class registers we may use. This depends on what the encapsulating\\n # (parent) function is:\\n #\\n # - A nested function: the callable class is instantiated\\n # from the current callable class' '__call__' function, and hence\\n # the callable class' environment register is used.\\n # - A generator function: the callable class is instantiated\\n # from the '__next__' method of the generator class, and hence the\\n # environment of the generator class is used.\\n # - Regular function: we use the environment of the original function.\\n curr_env_reg = None\\n if builder.fn_info.is_generator:\\n curr_env_reg = builder.fn_info.generator_class.curr_env_reg\\n elif builder.fn_info.is_nested:\\n curr_env_reg = builder.fn_info.callable_class.curr_env_reg\\n elif builder.fn_info.contains_nested:\\n curr_env_reg = builder.fn_info.curr_env_reg\\n if curr_env_reg:\\n builder.add(SetAttr(func_reg, ENV_ATTR_NAME, curr_env_reg, fitem.line))\\n return func_reg\",\n \"def ICRFFile_Instance():\\n return _gmat_py.ICRFFile_Instance()\",\n \"def func ( self ) :\\n return self.__func\",\n \"def fortran_interface(self) -> str:\\n result = ''\\n if self.fc_override == '':\\n return result\\n\\n func_name = '{}_{}_{}_'.format(\\n self.c_prefix, self.class_name, self.name)\\n\\n # declaration\\n in_parameters = self._fi_in_parameters()\\n return_type, out_parameters = self._fi_out_parameters()\\n if self.may_throw:\\n out_parameters.append(('integer (c_int)', 'err_code'))\\n out_parameters.append(('type (c_ptr)', 'err_msg'))\\n out_parameters.append(('integer (c_size_t)', 'err_msg_len'))\\n\\n arg_list = [par_name for _, par_name in in_parameters + out_parameters]\\n if len(arg_list) > 1:\\n arg_vlist = ' &\\\\n' + indent(', &\\\\n'.join(arg_list), 8*' ')\\n else:\\n arg_vlist = ', '.join(arg_list)\\n\\n if return_type != '':\\n result += '{} function {}({}) &\\\\n'.format(\\n return_type, func_name, arg_vlist)\\n else:\\n result += 'subroutine {}({}) &\\\\n'.format(func_name, arg_vlist)\\n result += ' bind(C, name=\\\"{}\\\")\\\\n'.format(func_name)\\n result += '\\\\n'\\n result += ' use iso_c_binding\\\\n'\\n\\n # parameter declarations\\n for par_type, par_name in in_parameters:\\n result += ' {}, intent(in) :: {}\\\\n'.format(\\n par_type, par_name)\\n for par_type, par_name in out_parameters:\\n result += ' {}, intent(out) :: {}\\\\n'.format(par_type, par_name)\\n\\n # end\\n if return_type != '':\\n result += 'end function {}\\\\n\\\\n'.format(func_name)\\n else:\\n result += 'end subroutine {}\\\\n\\\\n'.format(func_name)\\n return indent(result, 8*' ')\",\n \"def _wrap_FunctionDefResult(self, expr):\\n var = expr.var\\n name = var.name\\n scope = self.scope\\n # Make name available for later\\n scope.insert_symbol(name)\\n local_var = var.clone(scope.get_expected_name(name))\\n\\n if local_var.rank:\\n # Allocatable is not returned so it must appear in local scope\\n scope.insert_variable(local_var, name)\\n\\n # Create the C-compatible data pointer\\n bind_var = Variable(dtype=BindCPointer(),\\n name=scope.get_new_name('bound_'+name),\\n is_const=False, memory_handling='alias')\\n scope.insert_variable(bind_var)\\n\\n result = BindCFunctionDefResult(bind_var, var, scope)\\n\\n # Save the shapes of the array\\n self._additional_exprs.extend([Assign(result.shape[i], var.shape[i]) for i in range(var.rank)])\\n\\n # Create an array variable which can be passed to CLocFunc\\n ptr_var = var.clone(scope.get_new_name(name+'_ptr'),\\n memory_handling='alias')\\n scope.insert_variable(ptr_var)\\n\\n # Define the additional steps necessary to define and fill ptr_var\\n alloc = Allocate(ptr_var, shape=result.shape,\\n order=var.order, status='unallocated')\\n copy = Assign(ptr_var, var)\\n c_loc = CLocFunc(ptr_var, bind_var)\\n self._additional_exprs.extend([alloc, copy, c_loc])\\n\\n return result\\n else:\\n return BindCFunctionDefResult(local_var, var, scope)\",\n \"def class_from_invokedynamic(ins, cf):\\n info = InvokeDynamicInfo.create(ins, cf)\\n assert info.created_type != \\\"void\\\"\\n return info.created_type\",\n \"def FunctionDef(self):\\n type = self.Type()\\n key = self.currtok[0]\\n self.functions[self.currtok[0]] = \\\"first_call\\\"\\n id = self.primary()\\n self.functions[key] = type\\n if self.currtok[1].name == \\\"LPAREN\\\":\\n self.currtok = next(self.tg)\\n params = self.Params()\\n if self.currtok[1].name == \\\"RPAREN\\\":\\n self.currtok = next(self.tg)\\n if self.currtok[1].name == \\\"LCURLY\\\":\\n self.currtok = next(self.tg)\\n decs = self.Declarations()\\n states = self.Statements()\\n if self.currtok[1].name == \\\"RCURLY\\\":\\n self.currtok = next(self.tg)\\n return FunctionDef(type, id, params, decs, states)\\n raise SLUCSyntaxError(\\\"ERROR: Missing Right Curly Brace on line {0}\\\"\\n .format(str(self.currtok[2] - 1 - 1)))\\n raise SLUCSyntaxError(\\\"ERROR: Missing Left Curly Brace on line {0}\\\".format(str(self.currtok[2] - 1)))\\n raise SLUCSyntaxError(\\\"ERROR: Missing Right Paren on line {0}\\\".format(str(self.currtok[2] - 1)))\\n raise SLUCSyntaxError(\\\"ERROR: Missing Left Paren on line {0}\\\".format(str(self.currtok[2] - 1)))\",\n \"def test_cot():\\n c=0.5\\n\\n def myfunc(x):\\n f1=EF.cot(x)\\n return f1\\n\\n f_obj=ADiff(myfunc)\\n res=f_obj.Jac(c)\\n expectAns={'diff':2/(math.cos(c*2)-1), 'value':math.cos(c)/math.sin(c)}\\n assert res==expectAns\",\n \"def setup_callable_class(builder: IRBuilder) -> None:\\n # Check to see that the name has not already been taken. If so,\\n # rename the class. We allow multiple uses of the same function\\n # name because this is valid in if-else blocks. Example:\\n #\\n # if True:\\n # def foo(): ----> foo_obj()\\n # return True\\n # else:\\n # def foo(): ----> foo_obj_0()\\n # return False\\n name = base_name = \\\"{}_obj\\\".format(builder.fn_info.namespaced_name())\\n count = 0\\n while name in builder.callable_class_names:\\n name = base_name + \\\"_\\\" + str(count)\\n count += 1\\n builder.callable_class_names.add(name)\\n\\n # Define the actual callable class ClassIR, and set its\\n # environment to point at the previously defined environment\\n # class.\\n callable_class_ir = ClassIR(name, builder.module_name, is_generated=True)\\n\\n # The functools @wraps decorator attempts to call setattr on\\n # nested functions, so we create a dict for these nested\\n # functions.\\n # https://github.com/python/cpython/blob/3.7/Lib/functools.py#L58\\n if builder.fn_info.is_nested:\\n callable_class_ir.has_dict = True\\n\\n # If the enclosing class doesn't contain nested (which will happen if\\n # this is a toplevel lambda), don't set up an environment.\\n if builder.fn_infos[-2].contains_nested:\\n callable_class_ir.attributes[ENV_ATTR_NAME] = RInstance(\\n builder.fn_infos[-2].env_class\\n )\\n callable_class_ir.mro = [callable_class_ir]\\n builder.fn_info.callable_class = ImplicitClass(callable_class_ir)\\n builder.classes.append(callable_class_ir)\\n\\n # Add a 'self' variable to the environment of the callable class,\\n # and store that variable in a register to be accessed later.\\n self_target = add_self_to_env(builder.environment, callable_class_ir)\\n builder.fn_info.callable_class.self_reg = builder.read(\\n self_target, builder.fn_info.fitem.line\\n )\",\n \"def init():\\n return _libsbml.SBMLFunctionDefinitionConverter_init()\",\n \"def C(classname):\\n return objc.objc_getClass(_utf8(classname))\",\n \"def _get_ciphering_function(algorithm: Algorithm) -> Callable:\\n ciphering_function = None\\n if algorithm == Algorithm.caesar:\\n ciphering_function = cifra.cipher.caesar.cipher\\n elif algorithm == Algorithm.affine:\\n ciphering_function = cifra.cipher.affine.cipher\\n elif algorithm == Algorithm.substitution:\\n ciphering_function = cifra.cipher.substitution.cipher\\n elif algorithm == Algorithm.transposition:\\n ciphering_function = cifra.cipher.transposition.cipher\\n elif algorithm == Algorithm.vigenere:\\n ciphering_function = cifra.cipher.vigenere.cipher\\n return ciphering_function\",\n \"def construction(self):\\n from sage.categories.pushout import AlgebraicClosureFunctor\\n return (AlgebraicClosureFunctor(), self._real_field())\",\n \"def make_fee_class():\\r\\n fee = 1\\r\\n return make_class(locals())\",\n \"def functions(self):\\n return functions(self.startEA, self.endEA)\",\n \"def python_function(self):\\n return self._python_function # pylint: disable=protected-access\",\n \"def get_class(self):\\n\\t\\treturn self.CLASS\",\n \"def getCall(self):\\n if hasattr(self,'_filterfunctype'):\\n return self._filterfunctype\\n else:\\n return None\",\n \"def __init__(self, fn: callable):\\n self.fn = fn\",\n \"def functionURI(self):\\n ret = libxml2mod.xmlXPathGetFunctionURI(self._o)\\n return ret\",\n \"def from_function(cls, py_func, py_file):\\n raise NotImplementedError\",\n \"def from_function(cls, py_func, py_file):\\n raise NotImplementedError\",\n \"def _get_flint_func(cls, domain):\\n if domain == ZZ:\\n return flint.fmpz_mat\\n elif domain == QQ:\\n return flint.fmpq_mat\\n else:\\n raise NotImplementedError(\\\"Only ZZ and QQ are supported by DFM\\\")\",\n \"def get_func(func_ea):\\n if isinstance(func_ea, idaapi.func_t):\\n return func_ea\\n func = idaapi.get_func(func_ea)\\n if func is None:\\n raise exceptions.SarkNoFunction(\\\"No function at 0x{:08X}\\\".format(func_ea))\\n\\n return func\",\n \"def get_fns(self):\\n return self.key_to_fn\",\n \"def __init__(self, f, callback=None, errback=None):\\n self.f = f\\n self.exception = None\\n self.result = None\\n self.callback = callback\\n self.errback = errback\",\n \"def __init__(self, func, type):\\n self.func = func\\n self.type = type\",\n \"def getInstance(self):\\n return FSDecorator.instance\",\n \"def function(self):\\n return self.devicefuncs[self._funcname]\",\n \"def _make_callable(func):\\n try:\\n return func.evaluator()\\n except AttributeError:\\n return func\",\n \"def fn(self):\\n return self._fn\",\n \"def fortran_functions(self) -> str:\\n fun_name = f'{self.f_prefix}_{self.name}_to_int_'\\n result = f'integer function {fun_name}(flags)\\\\n'\\n result += ' implicit none\\\\n'\\n result += '\\\\n'\\n result += f' class({self.f_prefix}_{self.name}), intent(in) :: flags\\\\n'\\n result += ' integer :: ret_val\\\\n'\\n result += '\\\\n'\\n result += ' ret_val = 0\\\\n'\\n for i, value in enumerate(self.values):\\n result += f\\\" if (flags%{value}) ret_val = ret_val + {1 << i}\\\\n\\\"\\n result += f' {fun_name} = ret_val\\\\n'\\n result += f'end function {fun_name}\\\\n'\\n return indent(result, 4*' ')\",\n \"def get_func_lookup():\\n return {\\n \\\"randomstr\\\": randomstr,\\n \\\"random\\\": random,\\n \\\"sha256\\\": sha256,\\n \\\"ed25519\\\": ed25519_private_key,\\n \\\"rsa\\\": rsa_private_key,\\n \\\"rsapublic\\\": rsa_public_key,\\n \\\"publickey\\\": public_key,\\n \\\"reveal\\\": reveal,\\n \\\"loweralphanum\\\": loweralphanum,\\n \\\"basicauth\\\": basicauth,\\n }\",\n \"def func_in_class():\\n return \\\"just a function hanging out in a class\\\"\",\n \"def get_function(self):\\n return self.element.get_basis_functions()[self.n]\",\n \"def test_sin():\\n c=14\\n def myfunc(x):\\n f1=EF.sin(x)\\n return f1\\n\\n f_obj=ADiff(myfunc)\\n res=f_obj.Jac(c)\\n\\n expectAns={'diff': math.cos(c), 'value': math.sin(c)}\\n assert res==expectAns\",\n \"def __call__(self, function: FuncSpeechArg):\\n self._add_attr(function)\\n return function\",\n \"def get_const_func(const):\\n # pylint: disable=unused-argument\\n def env(t):\\n return const\\n\\n return env\",\n \"def ifc_definition(cls):\\n definitions[cls.__name__.upper()] = cls\\n return cls\",\n \"def __init__(self, function, function_representation):\\n self.function = function\\n self.function_representation = function_representation\",\n \"def get_code(self):\\n if self.code is None:\\n try:\\n import playhouse.sqlite_ext as sqlext\\n except:\\n raise KGTKException(f\\\"you need to 'pip install peewee' to use table-valued functions\\\")\\n # multiple-inheritance breaks things, so we inherit the relevant information by hand:\\n self.code = type(self.get_name(), (sqlext.TableFunction,), {\\n 'name': self.get_name(),\\n 'params': self.params,\\n 'columns': self.columns,\\n 'initialize': getattr(type(self), 'initialize'),\\n 'iterate': getattr(type(self), 'iterate'),\\n 'compute_result_rows': getattr(type(self), 'compute_result_rows'),\\n })\\n return self.code\",\n \"def get_class():\\n return CuBoltServerScript\",\n \"def eurologic_backplane_function(self):\\n return self._eurologic_backplane_function\",\n \"def get_corrmat(self, f):\\n return self._get_corrmat(f)\",\n \"def res_cls(self):\\n return self._res_cls\",\n \"def fortran_c_wrapper(self) -> str:\\n result = banner('//')\\n result += self._fc_includes()\\n result += self._fc_using_statements()\\n result += self._fc_function_definitions()\\n return result\",\n \"def result_as_arg(self, node, C_new):\\n F_new = C_new.clone()\\n\\n # Fortran function should wrap the new C function\\n F_new._PTR_F_C_index = C_new._function_index\\n F_new.wrap.assign(fortran=True)\\n # Do not add '_bufferify'\\n F_new.fmtdict.function_suffix = node.fmtdict.function_suffix\\n\\n # Do not wrap original function (does not have result argument)\\n node.wrap.fortran = False\\n return F_new\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.7460688","0.55869156","0.5528139","0.5359988","0.52939034","0.52370685","0.51732045","0.5147477","0.5134911","0.51322496","0.51305723","0.5104154","0.5016477","0.5000046","0.49879676","0.4959492","0.49496135","0.49427474","0.49407926","0.49324685","0.4931695","0.4903135","0.4884697","0.48673907","0.4860595","0.485764","0.48565164","0.48551172","0.48517746","0.4808122","0.48000586","0.47932723","0.4779233","0.4774073","0.4774024","0.4760248","0.47590503","0.47532454","0.47476813","0.47444686","0.4717364","0.47101438","0.47097495","0.46998802","0.4696885","0.4690215","0.46895716","0.4688081","0.46865004","0.46850654","0.46700558","0.4667724","0.4658604","0.4656677","0.4654761","0.46490473","0.46479905","0.46299607","0.4623355","0.46197706","0.46130505","0.46114606","0.45843998","0.45821577","0.45731878","0.45638368","0.45552063","0.455494","0.45392412","0.45356834","0.45352584","0.45282423","0.45249227","0.452088","0.452088","0.45168507","0.45165256","0.4515878","0.45132595","0.45124772","0.4510274","0.45071718","0.45046917","0.44991553","0.4494359","0.44936818","0.44936076","0.44833612","0.44829768","0.4480108","0.44696665","0.44634718","0.44634616","0.445322","0.4452098","0.44509986","0.44507334","0.44492254","0.44490668","0.44447434"],"string":"[\n \"0.7460688\",\n \"0.55869156\",\n \"0.5528139\",\n \"0.5359988\",\n \"0.52939034\",\n \"0.52370685\",\n \"0.51732045\",\n \"0.5147477\",\n \"0.5134911\",\n \"0.51322496\",\n \"0.51305723\",\n \"0.5104154\",\n \"0.5016477\",\n \"0.5000046\",\n \"0.49879676\",\n \"0.4959492\",\n \"0.49496135\",\n \"0.49427474\",\n \"0.49407926\",\n \"0.49324685\",\n \"0.4931695\",\n \"0.4903135\",\n \"0.4884697\",\n \"0.48673907\",\n \"0.4860595\",\n \"0.485764\",\n \"0.48565164\",\n \"0.48551172\",\n \"0.48517746\",\n \"0.4808122\",\n \"0.48000586\",\n \"0.47932723\",\n \"0.4779233\",\n \"0.4774073\",\n \"0.4774024\",\n \"0.4760248\",\n \"0.47590503\",\n \"0.47532454\",\n \"0.47476813\",\n \"0.47444686\",\n \"0.4717364\",\n \"0.47101438\",\n \"0.47097495\",\n \"0.46998802\",\n \"0.4696885\",\n \"0.4690215\",\n \"0.46895716\",\n \"0.4688081\",\n \"0.46865004\",\n \"0.46850654\",\n \"0.46700558\",\n \"0.4667724\",\n \"0.4658604\",\n \"0.4656677\",\n \"0.4654761\",\n \"0.46490473\",\n \"0.46479905\",\n \"0.46299607\",\n \"0.4623355\",\n \"0.46197706\",\n \"0.46130505\",\n \"0.46114606\",\n \"0.45843998\",\n \"0.45821577\",\n \"0.45731878\",\n \"0.45638368\",\n \"0.45552063\",\n \"0.455494\",\n \"0.45392412\",\n \"0.45356834\",\n \"0.45352584\",\n \"0.45282423\",\n \"0.45249227\",\n \"0.452088\",\n \"0.452088\",\n \"0.45168507\",\n \"0.45165256\",\n \"0.4515878\",\n \"0.45132595\",\n \"0.45124772\",\n \"0.4510274\",\n \"0.45071718\",\n \"0.45046917\",\n \"0.44991553\",\n \"0.4494359\",\n \"0.44936818\",\n \"0.44936076\",\n \"0.44833612\",\n \"0.44829768\",\n \"0.4480108\",\n \"0.44696665\",\n \"0.44634718\",\n \"0.44634616\",\n \"0.445322\",\n \"0.4452098\",\n \"0.44509986\",\n \"0.44507334\",\n \"0.44492254\",\n \"0.44490668\",\n \"0.44447434\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.8935946"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90890,"cells":{"query":{"kind":"string","value":"Returns the FriCAS function element class."},"document":{"kind":"string","value":"def _function_element_class(self):\n return FriCASFunctionElement"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def _function_class(self):\n return FriCASExpectFunction","def getFunctionClass(functionID):\n d = { 1: Linear,\n 2: LinearDrag,\n 11: Gaussian,\n 12: GaussianDrag,\n 21: Lorentzian,\n 22: LorentzianDrag }\n return d[functionID]","def cfunc_type(self):\n tif = ida_typeinf.tinfo_t()\n result = self.get_func_type(tif)\n if not result:\n return\n return tif","def get_func_type(self, *args):\n return _ida_hexrays.cfunc_t_get_func_type(self, *args)","def get_class_for(self, elem):\r\n\t\treturn self.__tag_to_cls.get(elem.tag, self.__default_cls)","def get_function(self):\n return self.element.get_basis_functions()[self.n]","def func(self):\n return self.__class__","def get_class_functional_name(name):\n name = _strip_class_name(name)\n return name","def get_ctf(ima):\n\tfrom EMAN2 import EMAN2Ctf\n\tctf_params = ima.get_attr(\"ctf\")\t\n\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang","def element_type(self) -> global___Type:","def w_class_a(self):\n return self._w_class_a","def getFunction(self) -> ghidra.program.model.listing.Function:\n ...","def arg_to_CFI(self, node, ordered_functions):\n options = node.options\n fmt_func = node.fmtdict\n\n if options.wrap_fortran is False:\n # The buffer function is intended to be called by Fortran.\n # No Fortran, no need for buffer function.\n return\n\n ast = node.ast\n declarator = ast.declarator\n result_typemap = ast.typemap\n # shadow classes have not been added yet.\n # Only care about string, vector here.\n result_is_ptr = declarator.is_indirect()\n if (\n result_typemap\n and result_typemap.base in [\"string\", \"vector\"]\n and result_typemap.name != \"char\"\n and not result_is_ptr\n ):\n node.wrap.c = False\n # node.wrap.fortran = False\n self.config.log.write(\n \"Skipping {}, unable to create C wrapper \"\n \"for function returning {} instance\"\n \" (must return a pointer or reference).\"\n \" Bufferify version will still be created.\\n\".format(\n result_typemap.cxx_type, declarator.user_name\n )\n )\n \n cfi_args = {}\n for arg in ast.declarator.params:\n declarator = arg.declarator\n name = declarator.user_name\n attrs = declarator.attrs\n meta = declarator.metaattrs\n cfi_args[name] = False\n arg_typemap = arg.typemap\n if meta[\"api\"]:\n # API explicitly set by user.\n continue\n elif meta[\"assumed-rank\"]:\n cfi_args[name] = True\n elif attrs[\"rank\"]:\n cfi_args[name] = True\n elif arg_typemap.sgroup == \"string\":\n cfi_args[name] = True\n elif arg_typemap.sgroup == \"char\":\n if declarator.is_indirect():\n cfi_args[name] = True\n elif meta[\"deref\"] in [\"allocatable\", \"pointer\"]:\n cfi_args[name] = True\n has_cfi_arg = any(cfi_args.values())\n\n # Function result.\n need_buf_result = None\n\n result_as_arg = \"\" # Only applies to string functions\n # when the result is added as an argument to the Fortran api.\n\n # Check if result needs to be an argument.\n declarator = ast.declarator\n attrs = declarator.attrs\n meta = declarator.metaattrs\n if meta[\"deref\"] == \"raw\":\n # No bufferify required for raw pointer result.\n pass\n elif result_typemap.sgroup == \"string\":\n need_buf_result = \"cfi\"\n result_as_arg = fmt_func.F_string_result_as_arg\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\n elif result_typemap.sgroup == \"char\" and result_is_ptr:\n need_buf_result = \"cfi\"\n result_as_arg = fmt_func.F_string_result_as_arg\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\n elif meta[\"deref\"] in [\"allocatable\", \"pointer\"]:\n need_buf_result = \"cfi\"\n\n if not (need_buf_result or\n has_cfi_arg):\n return False\n\n options.wrap_fortran = False\n\n # Create a new C function and change arguments\n # and add attributes.\n C_new = node.clone()\n ordered_functions.append(C_new)\n self.append_function_index(C_new)\n\n generated_suffix = \"cfi\"\n C_new._generated = \"arg_to_cfi\"\n C_new.splicer_group = \"cfi\"\n if need_buf_result:\n C_new.ast.declarator.metaattrs[\"api\"] = need_buf_result\n fmt_func = C_new.fmtdict\n fmt_func.function_suffix = fmt_func.function_suffix + fmt_func.C_cfi_suffix\n\n C_new.wrap.assign(c=True)#, fortran=True)\n C_new._PTR_C_CXX_index = node._function_index\n\n for arg in C_new.ast.declarator.params:\n name = arg.declarator.user_name\n if cfi_args[name]:\n arg.declarator.metaattrs[\"api\"] = generated_suffix\n\n ast = C_new.ast\n if True: # preserve to avoid changing indention for now.\n f_attrs = node.ast.declarator.attrs # Fortran function attributes\n f_meta = node.ast.declarator.metaattrs # Fortran function attributes\n if result_as_arg:\n # decl: const char * getCharPtr2() +len(30)\n # +len implies copying into users buffer.\n result_as_string = ast.result_as_arg(result_name)\n result_as_string.const = False # must be writeable\n attrs = result_as_string.declarator.attrs\n # Special case for wrapf.py to override \"allocatable\"\n f_meta[\"deref\"] = None\n result_as_string.declarator.metaattrs[\"api\"] = \"cfi\"\n result_as_string.declarator.metaattrs[\"deref\"] = \"result\"\n result_as_string.declarator.metaattrs[\"is_result\"] = True\n C_new.ast.declarator.metaattrs[\"api\"] = None\n C_new.ast.declarator.metaattrs[\"intent\"] = \"subroutine\"\n C_new.ast.declarator.metaattrs[\"deref\"] = None\n\n if result_as_arg:\n F_new = self.result_as_arg(node, C_new)\n ordered_functions.append(F_new)\n self.append_function_index(F_new)\n else:\n if node._generated in [\"result_to_arg\", \"fortran_generic\", \"getter/setter\"]:\n node.wrap.c = False\n # Fortran function may call C subroutine if string/vector result\n # Fortran function calls bufferify function.\n node._PTR_F_C_index = C_new._function_index\n return True","def get_class(self):\n\t\treturn self.CLASS","def _fc_function_definitions(self) -> str:\n result = 'extern \"C\" {\\n\\n'\n for namespace in self.namespaces:\n for member in namespace.members:\n result += member.fortran_c_wrapper()\n\n result += '}\\n\\n'\n return result","def GetClassConstructor(filename):\n ode_name = GetOdeSystemName(filename)\n\n class_constr_str = (ode_name + \"::~\" + ode_name + \"()\\n\"\n \"{\\n\" \n \"}\\n\\n\")\n\n return class_constr_str","def getColorTransferFunction(self):\n\t\treturn self.ctf","def C(classname):\n return objc.objc_getClass(_utf8(classname))","def fortran_c_wrapper(self) -> str:\n if self.fc_override is not None:\n return self.fc_override.replace('$CLASSNAME$', self.class_name).replace(\n \"$C_PREFIX$\", self.c_prefix).replace(\"$F_PREFIX$\", self.f_prefix)\n\n result = ''\n\n # declaration\n in_parameters = self._fc_in_parameters()\n return_type, out_parameters = self._fc_out_parameters()\n if self.may_throw:\n out_parameters.append('int * err_code')\n out_parameters.append('char ** err_msg')\n out_parameters.append('std::size_t * err_msg_len')\n\n func_name = '{}_{}_{}_'.format(\n self.c_prefix, self.class_name, self.name)\n\n par_str = ', '.join(in_parameters + out_parameters)\n result += '{} {}({}) {{\\n'.format(return_type, func_name, par_str)\n\n # convert input\n for par in self.params:\n result += '{}'.format(par.fc_convert_input())\n\n # call C++ function and return result\n if self.may_throw:\n result += ' try {\\n'\n result += ' *err_code = 0;\\n'\n result += indent(self._fc_cpp_call(), 4*' ')\n result += indent(self._fc_return(), 4*' ')\n result += ' }\\n'\n for exception, code in error_codes.items():\n if code != 0:\n catch = ''\n catch += 'catch (std::{} const & e) {{\\n'.format(exception)\n catch += ' *err_code = {};\\n'.format(code)\n catch += ' static std::string msg;\\n'\n catch += ' msg = e.what();\\n'\n catch += ' *err_msg = const_cast<char*>(msg.data());\\n'\n catch += ' *err_msg_len = msg.size();\\n'\n catch += '}\\n'\n result += indent(catch, 4*' ')\n result += self._fc_return_default()\n else:\n result += self._fc_cpp_call()\n result += self._fc_return()\n result += '}\\n\\n'\n return result","def getClass(self):\n return _libsbml.ASTNode_getClass(self)","def CFL(self):\n return self.__CFL","def _getTypeClass(self, namespaceURI, localName):\r\n bti = BaseTypeInterpreter()\r\n simpleTypeClass = bti.get_typeclass(localName, namespaceURI)\r\n return simpleTypeClass","def get_func_type(self, *args):\n return _ida_hexrays.cfuncptr_t_get_func_type(self, *args)","def fortran_function(self) -> str:\n return ''.join([i.fortran_function() for i in self.instances])","def get_css_class(self):\n\n return None","def get_unc_f(self):\n return self.uncf","def tctfdfc(x):\n if isinstance(x,Fdf) :\n pass\n else : \n x = Fdf.constant(x)\n return x","def _func(self):\n return self._get_flint_func(self.domain)","def _get_class():\n return ASParameters","def get_asset_class_feature_value(self):\n ac_feature = AssetFeature.Standard.ASSET_CLASS.get_object()\n return AssetFeatureValue.objects.get_or_create(name=self.asset_class.display_name, feature=ac_feature)[0]","def get_surface_class(self) -> SurfaceClassStr:\n return SURFACE_CLASS.inverse[self.surfaceClass()]","def choose_botorch_acqf_class() -> Type[AcquisitionFunction]:\n # NOTE: In the future, this dispatch function could leverage any\n # of the attributes of `BoTorchModel` or kwargs passed to\n # `BoTorchModel.gen` to intelligently select acquisition function.\n return qNoisyExpectedImprovement","def get_class_name(self):\n\n if \"class\" in self._root.attrib:\n return self._root.attrib['class']\n else:\n return self._root.tag","def ICRFFile_Instance():\n return _gmat_py.ICRFFile_Instance()","def getTypeCode(self):\n return _libsbml.FunctionDefinition_getTypeCode(self)","def ffb_c (cls):\r\n b=(bin(c0+jmax)[-(cls)])\r\n return (2*int(b)-1) # or -(2*int(b)-1), depending on the final direction of the hierarchy\r","def getClass():\n\treturn AnimatorMode","def make_fee_class():\r\n fee = 1\r\n return make_class(locals())","def get_func_type(header):\n func_type = header.functionType\n if func_type == SSE.SCALAR:\n return FunctionType.Scalar\n elif func_type == SSE.AGGREGATION:\n return FunctionType.Aggregation\n elif func_type == SSE.TENSOR:\n return FunctionType.Tensor","def get_elf_class(afile):\n if afile in g_elf_class_db:\n return g_elf_class_db[afile]\n cmd = 'readelf -h ' + cmd_quote(afile) + ' | grep \"Class:\" || true'\n output = get_shell_cmd_output(cmd)\n tokens = output.split(\":\")\n elf_class = tokens[1].strip()\n verbose(afile + \" ELF Class is \" + elf_class, LEVEL_2)\n g_elf_class_db[afile] = elf_class\n return elf_class","def get_classless_functions(root_node):\n\n file_ast = ast.parse(root_node)\n\n visitor = ClasslessFunctionVisitor()\n\n visitor.visit(file_ast)\n\n return visitor.function_list","def get_element_type(cls):\r\n return cls._type_name(cls.element_type)","def fortran_function(self) -> str:\n if self.f_override is not None:\n return indent(\n self.f_override.replace('$CLASSNAME$', self.class_name).replace(\n \"$C_PREFIX$\", self.c_prefix).replace(\n \"$F_PREFIX$\", self.f_prefix),\n 4*' ')\n\n result = ''\n\n # declaration\n func_name = '{}_{}_{}'.format(\n self.f_prefix, self.class_name, self.name)\n in_parameters = self._f_in_parameters()\n return_type, out_parameters = self._f_out_parameters()\n if self.may_throw:\n out_parameters.append(('integer, optional', 'err_code'))\n out_parameters.append(('character(:), allocatable, optional',\n 'err_msg'))\n\n all_parameters = in_parameters + out_parameters\n arg_list = ', &\\n'.join([par_name for _, par_name in all_parameters])\n arg_ilist = indent(arg_list, 8*' ')\n if return_type != '':\n result += 'function {}( &\\n{})\\n'.format(func_name, arg_ilist)\n else:\n result += 'subroutine {}( &\\n{})\\n'.format(func_name, arg_ilist)\n\n # parameter declarations\n result += ' implicit none\\n'\n for par_type, par_name in in_parameters:\n result += ' {}, intent(in) :: {}\\n'.format(\n par_type, par_name)\n for par_type, par_name in out_parameters:\n result += ' {}, intent(out) :: {}\\n'.format(par_type, par_name)\n if return_type != '':\n result += ' {} :: {}\\n'.format(return_type, func_name)\n result += '\\n'\n\n # variable declarations\n c_return_type, fi_out_parameters = self._fi_out_parameters()\n if c_return_type:\n result += ' {} :: ret_val\\n'.format(c_return_type)\n for par_type, par_name in fi_out_parameters:\n result += ' {} :: {}\\n'.format(par_type, par_name)\n for par_type, par_name in self.ret_type.f_aux_variables():\n result += ' {} :: {}\\n'.format(par_type, par_name)\n if self.may_throw:\n result += ' integer (c_int) :: err_code_v\\n'\n result += ' type (c_ptr) :: err_msg_v\\n'\n result += ' integer (c_size_t) :: err_msg_len_v\\n'\n result += ' character (c_char), dimension(:), pointer :: err_msg_f\\n'\n result += ' character(:), allocatable :: err_msg_p\\n'\n result += ' integer (c_size_t) :: err_msg_i\\n'\n if c_return_type or fi_out_parameters or self.may_throw:\n result += '\\n'\n\n # convert input\n args = [param.f_chain_arg() for param in self.params]\n args += [par_name for _, par_name in fi_out_parameters]\n if self.may_throw:\n args += ['err_code_v', 'err_msg_v', 'err_msg_len_v']\n arg_str = ', &\\n'.join([8*' ' + arg for arg in args])\n\n # call C function\n fc_func_name = '{}_{}_{}_'.format(\n self.c_prefix, self.class_name, self.name)\n chain_call = self.fc_chain_call(\n ns_prefix=self.c_prefix, class_name=self.class_name,\n fc_func_name=fc_func_name, fc_args=arg_str)\n result_name = ''\n if return_type != '':\n result_name = func_name\n elif out_parameters:\n result_name = out_parameters[0][1]\n result += self.ret_type.f_call_c('ret_val', chain_call)\n\n # handle errors if necessary\n if self.may_throw:\n # Note: I tried to factor this out into a function, but Fortran\n # makes that near-impossible. Since we're generating anyway, it's\n # not really duplication, so leave it as is.\n result += indent(dedent(f\"\"\"\\\n if (err_code_v .ne. 0) then\n if (present(err_code)) then\n err_code = err_code_v\n if (present(err_msg)) then\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\n allocate (character(err_msg_len_v) :: err_msg)\n do err_msg_i = 1, err_msg_len_v\n err_msg(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\n end do\n end if\n {dedent(\n self.ret_type.f_return_dummy_result(result_name))}\n return\n else\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\n allocate (character(err_msg_len_v) :: err_msg_p)\n do err_msg_i = 1, err_msg_len_v\n err_msg_p(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\n end do\n print *, err_msg_p\n stop\n end if\n else\n if (present(err_code)) then\n err_code = 0\n end if\n end if\n\n \"\"\"), 4*' ')\n\n # convert and return result\n result += self.ret_type.f_return_result(result_name, 'ret_val')\n\n # end\n if return_type != '':\n result += 'end function {}\\n\\n'.format(func_name)\n else:\n result += 'end subroutine {}\\n\\n'.format(func_name)\n return indent(result, 4*' ')","def Point_to_FeatureClass(self, fc):\n\n\n feature_class = []\n for index, traectory in enumerate(self.__traectory_list):\n point_row = arcpy.Point(X=traectory[0], Y=traectory[1], Z=traectory[2], ID=index)\n feature_class.append(arcpy.PointGeometry(point_row, arcpy.SpatialReference(2436)))\n arcpy.CopyFeatures_management(feature_class, (self.workspace + '\\\\' + fc))\n print 'Complete Creating a Point Feature Class'\n\n return None","def create_class_instance(element, element_id, doc_id):\n xsi_type = get_xsi_type(element)\n element_class = XSI_TYPE_CLASSES[xsi_type]\n return element_class.from_etree(element)","def eurologic_backplane_function(self):\n return self._eurologic_backplane_function","def class_from_invokedynamic(ins, cf):\n info = InvokeDynamicInfo.create(ins, cf)\n assert info.created_type != \"void\"\n return info.created_type","def visit_class(self, flags, scope, token, parent):\r\n\r\n # define the class name in the current scope\r\n # see visit_block\r\n #scope.define(SC_FUNCTION, token.children[0])\r\n scope.defer(token)","def _get_cos_traffic_class(self):\n return self.__cos_traffic_class","def get_fc(self, layer):\n return getattr(self, 'attention_{}'.format(layer))","def get_function(self):\n return SSAFunction(self.get_graph())","def function(self):\n ret = libxml2mod.xmlXPathGetFunction(self._o)\n return ret","def element_type(self):\r\n result = conf.lib.clang_getElementType(self)\r\n if result.kind == TypeKind.INVALID:\r\n raise Exception('Element type not available on this type.')\r\n\r\n return result","def fce(B):\n return wce(B)/(2.*np.pi)","def convert(self):\n return _libsbml.SBMLFunctionDefinitionConverter_convert(self)","def getFunction(self):\n return gpio.gpio_function(self.bcm_id)","def fortran_functions(self) -> str:\n fun_name = f'{self.f_prefix}_{self.name}_to_int_'\n result = f'integer function {fun_name}(flags)\\n'\n result += ' implicit none\\n'\n result += '\\n'\n result += f' class({self.f_prefix}_{self.name}), intent(in) :: flags\\n'\n result += ' integer :: ret_val\\n'\n result += '\\n'\n result += ' ret_val = 0\\n'\n for i, value in enumerate(self.values):\n result += f\" if (flags%{value}) ret_val = ret_val + {1 << i}\\n\"\n result += f' {fun_name} = ret_val\\n'\n result += f'end function {fun_name}\\n'\n return indent(result, 4*' ')","def define_function_spaces(self):\n\n cell = self.mesh.ufl_cell()\n vec_degree = int(self.config['formulation']['element'][0][-1])\n if vec_degree == 0:\n vec_family = \"DG\"\n else:\n vec_family = \"CG\"\n vec_element = dlf.VectorElement(vec_family, cell, vec_degree)\n self.vectorSpace = dlf.FunctionSpace(self.mesh, vec_element)\n\n if self.config['material']['incompressible']:\n scalar_degree = int(self.config['formulation']['element'][1][-1])\n if scalar_degree == 0:\n scalar_family = \"DG\"\n else:\n scalar_family = \"CG\"\n scalar_element = dlf.FiniteElement(scalar_family, cell, scalar_degree)\n self.scalarSpace = dlf.FunctionSpace(self.mesh, scalar_element)\n else:\n self.scalarSpace = None\n\n return None","def get_attribute_class(self):\n return self._attr_cls","def getFunctionStyle(self,cpArray,X0):\n cpArray=tf.cast(tf.convert_to_tensor(cpArray),dtype=tf.float64)\n X0 = tf.cast(tf.convert_to_tensor(X0),dtype=tf.float64)\n func = self.lamda_computeCPdiff(X0)\n gatheredCps = tf.map_fn(func,cpArray,parallel_iterations=32,back_prop=False)\n return gatheredCps","def getElementName(self):\n return _libsbml.FunctionDefinition_getElementName(self)","def cython_ctype(self):\n if self._cython_ctype is None:\n self._cython_ctype = self.ts.cython_ctype(self.t)\n return self._cython_ctype","def feature_class():\n if deployment_settings.get_security_map() and not s3_has_role(\"MapAdmin\"):\n unauthorised()\n\n tablename = \"%s_%s\" % (module, resourcename)\n table = db[tablename]\n\n # Model options\n table.gps_marker.comment = DIV( _class=\"tooltip\",\n _title=\"%s|%s\" % (T(\"GPS Marker\"),\n T(\"Defines the icon used for display of features on handheld GPS.\")))\n\n # CRUD Strings\n LIST_FEATURE_CLASS = T(\"List Feature Classes\")\n s3.crud_strings[tablename] = Storage(\n title_create = ADD_FEATURE_CLASS,\n title_display = T(\"Feature Class Details\"),\n title_list = T(\"Feature Classes\"),\n title_update = T(\"Edit Feature Class\"),\n title_search = T(\"Search Feature Class\"),\n subtitle_create = T(\"Add New Feature Class\"),\n subtitle_list = LIST_FEATURE_CLASS,\n label_list_button = LIST_FEATURE_CLASS,\n label_create_button = ADD_FEATURE_CLASS,\n label_delete_button = T(\"Delete Feature Class\"),\n msg_record_created = T(\"Feature Class added\"),\n msg_record_modified = T(\"Feature Class updated\"),\n msg_record_deleted = T(\"Feature Class deleted\"),\n msg_list_empty = T(\"No Feature Classes currently defined\"))\n\n output = s3_rest_controller(module, resourcename)\n\n if not \"gis\" in response.view and response.view != \"popup.html\":\n response.view = \"gis/\" + response.view\n\n return output","def func_in_class():\n return \"just a function hanging out in a class\"","def __init__(self, function='cogscore/'):\n self.function = function","def frtype(self):\n return self._frtype","def record_cls(self):\n return self._ELE_CLS","def class_ref(self):\n return self._class_ref","def _get_ciphering_function(algorithm: Algorithm) -> Callable:\n ciphering_function = None\n if algorithm == Algorithm.caesar:\n ciphering_function = cifra.cipher.caesar.cipher\n elif algorithm == Algorithm.affine:\n ciphering_function = cifra.cipher.affine.cipher\n elif algorithm == Algorithm.substitution:\n ciphering_function = cifra.cipher.substitution.cipher\n elif algorithm == Algorithm.transposition:\n ciphering_function = cifra.cipher.transposition.cipher\n elif algorithm == Algorithm.vigenere:\n ciphering_function = cifra.cipher.vigenere.cipher\n return ciphering_function","def get_attribute_class(self, attr_name):\n return self.attrs.get_attribute_class(attr_name)","def get_SrcClass(args):\n return Reactome(args)","def getFunction(self, name: unicode) -> ghidra.program.model.listing.Function:\n ...","def functionURI(self):\n ret = libxml2mod.xmlXPathGetFunctionURI(self._o)\n return ret","def createFunctionTerm(self):\n return _libsbml.Transition_createFunctionTerm(self)","def _class(self):\n return self.__class","def get_function(self):\n return Gumtree.gumtree.getFunction()","def get_float_type(*args):\n return _ida_hexrays.get_float_type(*args)","def getTypeCode(self):\n return _libsbml.FbcAnd_getTypeCode(self)","def return_regressor_class(self):\n pass","def __init__(self, func_name, spin):\n self.xc_func = None\n self._xc_func_init = False\n\n # Handle func_name\n if isinstance(func_name, str):\n func_id = util.xc_functional_get_number(func_name)\n if func_id == -1:\n raise KeyError(\"LibXCFunctional: name '%s' not found.\" % func_name)\n elif isinstance(func_name, (int, np.integer)):\n func_id = func_name\n if util.xc_functional_get_name(func_name) is None:\n raise KeyError(\"LibXCFunctional: ID '%d' not found.\" % func_name)\n else:\n raise TypeError(\"LibXCFunctional: func_name must either be a string or int. Got {}\".format(func_name))\n\n self._xc_func_name = util.xc_functional_get_name(func_id)\n\n # Handle spin\n if isinstance(spin, str):\n spin = spin.lower()\n if spin == \"polarized\":\n self._spin = 2\n elif spin == \"unpolarized\":\n self._spin = 1\n else:\n raise KeyError(\"LibXCFunctional: spin must either be 'polarized' or 'unpolarized' if represented by a string. Got {}\".format(spin))\n else:\n self._spin = spin\n\n if self._spin not in [1, 2]:\n raise KeyError(\"LibXCFunctional: spin must either be 1 or 2 if represented by a integer. Got {}\".format(self._spin))\n\n # Build the LibXC functional\n self.xc_func = core.xc_func_alloc()\n self.xc_func_size_names = [x for x in dir(self.xc_func.contents.dim) if not \"_\" in x]\n\n # Set all int attributes to zero (not all set to zero in libxc)\n for attr in self.xc_func_size_names:\n setattr(self.xc_func.contents, attr, 0)\n\n ret = core.xc_func_init(self.xc_func, func_id, self._spin)\n if ret != 0:\n raise ValueError(\"LibXC Functional construction did not complete. Error code %d\" % ret)\n self._xc_func_init = True\n\n # Pull out all sizes after init\n self.xc_func_sizes = {}\n for attr in self.xc_func_size_names:\n self.xc_func_sizes[attr] = getattr(self.xc_func.contents.dim, attr)\n\n # Unpack functional info\n self.xc_func_info = core.xc_func_get_info(self.xc_func)\n self._number = core.xc_func_info_get_number(self.xc_func_info)\n self._kind = core.xc_func_info_get_kind(self.xc_func_info)\n self._name = core.xc_func_info_get_name(self.xc_func_info).decode(\"UTF-8\")\n self._family = core.xc_func_info_get_family(self.xc_func_info)\n self._flags = core.xc_func_info_get_flags(self.xc_func_info)\n\n # Set needed flags\n self._needs_laplacian = self._flags & flags.XC_FLAGS_NEEDS_LAPLACIAN\n\n # Set derivatives\n self._have_exc = self._flags & flags.XC_FLAGS_HAVE_EXC\n self._have_vxc = self._flags & flags.XC_FLAGS_HAVE_VXC\n self._have_fxc = self._flags & flags.XC_FLAGS_HAVE_FXC\n self._have_kxc = self._flags & flags.XC_FLAGS_HAVE_KXC\n self._have_lxc = self._flags & flags.XC_FLAGS_HAVE_LXC\n\n # Set omega\n self._have_cam = self._flags & flags.XC_FLAGS_HYB_CAM\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_CAMY\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LC\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LCY\n self._cam_omega = self._cam_alpha = self._cam_beta = False\n if self._have_cam:\n self._cam_omega = self.xc_func.contents.cam_omega\n self._cam_alpha = self.xc_func.contents.cam_alpha\n self._cam_beta = self.xc_func.contents.cam_beta\n\n elif self._family in [flags.XC_FAMILY_HYB_LDA, flags.XC_FAMILY_HYB_GGA, flags.XC_FAMILY_HYB_MGGA]:\n self._cam_alpha = self.xc_func.contents.cam_alpha\n\n # VV10\n self._have_vv10 = self._flags & flags.XC_FLAGS_VV10\n self._nlc_b = self._nlc_C = False\n if self._have_vv10:\n self._nlc_b = self.xc_func.contents.nlc_b\n self._nlc_C = self.xc_func.contents.nlc_C\n\n # Stable\n self._stable = self._flags & flags.XC_FLAGS_STABLE\n self._dev = self._flags & flags.XC_FLAGS_DEVELOPMENT\n\n # Pull out references\n self._refs = []\n self._bibtexs = []\n self._dois = []\n\n for pos in range(flags.XC_MAX_REFERENCES):\n ref = core.xc_func_info_get_references(self.xc_func_info, pos)\n if not ref: break\n\n self._refs.append(ref.contents.ref.decode(\"UTF-8\"))\n self._bibtexs.append(ref.contents.bibtex.decode(\"UTF-8\"))\n self._dois.append(ref.contents.doi.decode(\"UTF-8\"))","def get_class_from_frame(fr):\n args, _, _, value_dict = inspect.getargvalues(fr)\n # Check the first parameter of the frame function is named `self` - if it is,\n # `self` will be referenced in `value_dict`.\n if len(args) != 0 and args[0] == 'self':\n instance = value_dict.get('self', None)\n if instance:\n return getattr(instance, '__class__', None)\n return None","def get_attribute_class(self, name):\n self.validate_attribute_name(name)\n return self.schema[name].get_attribute_class()","def as_ret(self):\n if self.type in OCClass.classes_indexed_by_name:\n oc_class = OCClass.classes_indexed_by_name[self.type]\n if oc_class.class_meths:\n for f in oc_class.class_meths:\n #TODO MachOTask.currentTask.analyze_function(start_addr=f)\n print 'T'","def __call__(self, f_or_klass):\n\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\n # 0.3.0\n # -- implement \"kill switch\", NO_DECO\n # -- handle decorating both functions and classes\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\n\n # 0.3.0b16: if it isn't callable, scram'\n if not callable(f_or_klass):\n return f_or_klass\n\n # Special-case handling for ``NO_DECO``: remove from settings of ``self``\n if self._effective_settings.get('NO_DECO'):\n return f_or_klass\n # else, delete that item wherever it might be\n if 'NO_DECO' in self._effective_settings:\n del self._effective_settings['NO_DECO']\n if 'NO_DECO' in self._changed_settings:\n del self._changed_settings['NO_DECO']\n\n f = f_or_klass if inspect.isfunction(f_or_klass) else None\n klass = f_or_klass if inspect.isclass(f_or_klass) else None\n\n self.f = f\n self.cls = klass\n\n if klass:\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a class\" -- namely, klass\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n\n self._class__call__(klass) # modifies klass (methods & inner classes) (if not builtin)\n self._add_class_attrs(klass) # v0.3.0v20 traps TypeError for builtins\n return klass\n\n elif not f:\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a callable but not a function\"\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # functools.partial objects are callable, have no __name__ much less __qualname__,\n # and trying to deco __call__ gets messy.\n # Callable builtins e.g. len are not functions in the isfunction sense,\n # can't deco anyway. Just give up (quietly):\n return f_or_klass\n\n else: # not a class, f nonempty is a function of f_or_klass callable\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n # 0.3.0 -- case \"f_or_klass is a function\" -- namely, f\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\n\n #----------------------------------------------------------------\n # Don't double-decorate -- don't wanna, & it doesn't work anyway!\n #----------------------------------------------------------------\n # Note: As with methods of classes,\n # . if f is deco'd, its existing EXPLICITLY GIVEN settings take precedence.\n\n # # From _class__call__, props & methods cases, w/a few name changes\n deco_obj = getattr(f, self._sentinels['DECO_OF'], None) # type: _deco_base\n\n # get a fresh copy for each attr\n new_settings = self._changed_settings.copy() # updated below\n\n # __init__ fixup, a nicety:\n # By default, don't log retval for __init__.\n # If user insists on it with 'log_retval=True' in __init__ deco,\n # that will override this.\n if f.__name__ == '__init__':\n self.fixup_for_init(new_settings)\n\n if deco_obj: # f is deco'd by this decorator\n # Yes. Figure out settings for f,\n ### 0.3.0b18 -- Use self._override\n self._update_settings(new=new_settings,\n old=deco_obj._changed_settings,\n override_existing=self._override)\n # update func's settings (_force_mutable=True to handle `max_history` properly)\n deco_obj._settings_mapping.update(new_settings, _force_mutable=True)\n return f\n\n #----------------------------------------------------------------\n # f is a function & is NOT already deco'd\n #----------------------------------------------------------------\n\n # 0.3.0.x -- f may not have a .__qualname__\n try:\n self._classname_of_f = '.'.join( f.__qualname__.split('.')[:-1] )\n except AttributeError as e:\n self._classname_of_f = ''\n\n # Special-case '__repr__' handling, if deco subclass doesn't allow it.\n if f.__name__ == '__repr__' and self._classname_of_f and not self.allow_repr():\n # v0.3.0b23 -- Instead of refusing to deco, use recursive_repr\n # return f\n return recursive_repr(fillvalue=\"...\")(f)\n\n # 0.3.0\n # Use __qualname__ ALL the time, unless user provides `name=display_name_str`\n # where `display_name_str` is either the name to be used for the fn in logged output,\n # or is an oldstyle format str into which f.__name__ will be substituted\n # to obtain the display name.\n # We require Py3.3+, so __qualname__ is available.\n\n # setup f_display_name\n if self._name_param:\n try:\n self.f_display_name = (self._name_param % f.__name__)\n except TypeError:\n self.f_display_name = self._name_param\n else:\n self.f_display_name = f.__qualname__\n\n # TODO TRY THIS -- anything break?\n # 0.3.1 Inspired by fractions.Fraction.__sub__ et al:\n # __name__ may be very different from __qualname__;\n # if so, show both\n if f.__name__ not in f.__qualname__:\n self.f_display_name += \" (\" + f.__name__ + \")\"\n\n #================================================================\n # 0.3.0 -- Init things (migrated from __init__)\n #----------------------------------------------------------------\n # set up pseudo-dict (DecoSettingsMapping),\n # using settings given by self._effective_settings.\n #\n # *** DecoSettingsMapping \"API\" --\n # (2) construct DecoSettingsMapping object\n # that will provide mapping & attribute access to settings, & more\n #----------------------------------------------------------------\n self._settings_mapping = DecoSettingsMapping(\n deco_class=self.__class__,\n # DecoSettingsMapping calls the rest ** values_dict\n ** self._effective_settings # 0.3.0 set by __init__\n )\n\n #----------------------------------------------------------------\n # Init more stuff\n #----------------------------------------------------------------\n self._stats = ClassInstanceAttrProxy(\n class_instance=self,\n data_descriptor_names=self.__class__._data_descriptor_names,\n method_descriptor_names=self.__class__._method_descriptor_names)\n # Accessed by descriptors on the stats obj\n self._num_calls_total = 0\n self._num_calls_logged = 0\n # max_history > 0 --> size of self._call_history; <= 0 --> unbounded\n # Set before calling _make_call_history\n\n # 0.3.0 self._other_values_dict set by __init__\n self.max_history = self._other_values_dict.get('max_history', 0) # <-- Nota bene\n self._call_history = self._make_call_history()\n\n # Accumulate this (for logged calls only)\n # even when record_history is false:\n self._elapsed_secs_logged = 0.0\n self._process_secs_logged = 0.0\n\n # 0.2.2.post1\n # stack(s), pushed & popped wrapper of deco'd function\n # by _logging_state_push, _logging_state_pop\n # 0.3.0 convert to pushing/popping single namedtuples\n self.logging_state_stack = [] # 0.3.0 stack of LoggingState namedtuples\n self._enabled_stack = [] # 0.3.0 - um, stack, of 'enabled's\n\n #----------------------------------------------------------------\n # end of Init passage\n #================================================================\n\n # Save signature and parameters of f\n self.f_signature = inspect.signature(f) # Py >= 3.3\n self.f_params = self.f_signature.parameters\n\n # 0.3.0 We assume Py3.3 so we use perf_counter, process_time all the time\n wall_time_fn = time.perf_counter\n process_time_fn = time.process_time\n\n #############################\n # The wrapper of a callable\n #############################\n\n @wraps(f)\n def _deco_base_f_wrapper_(*args, **kwargs):\n \"\"\"Wrapper around the wrapped function f.\n When this runs, f has been called, so we can now resolve\n any indirect values for the settings/keyword-params\n of log_calls, using info in kwargs and self.f_params.\"\"\"\n # *** Part of the DecoSettingsMapping \"API\" --\n # (4) using self._settings_mapping.get_final_value in wrapper\n # [[[ This/these is/are 4th chronologically ]]]\n\n # inner/local fn -- save a few cycles and characters -\n # we call this a lot (<= 9x).\n def _get_final_value(setting_name):\n \"Use outer scope's kwargs and self.f_params\"\n return self._settings_mapping.get_final_value(\n setting_name, kwargs, fparams=self.f_params)\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # if nothing to do, hurry up & don't do it.\n # NOTE: call_chain_to_next_log_calls_fn looks in stack frames\n # to find (0.2.4) STACKFRAME_HACK_DICT_NAME (really!)\n # It and its values (the following _XXX variables)\n # must be set before calling f.\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n _enabled = _get_final_value('enabled')\n # 0.3.0 in case f calls log_message (no output if f disabled)\n self._enabled_state_push(_enabled)\n\n # 0.2.4.post5 \"true bypass\": if 'enabled' < 0 then scram\n if _enabled < 0:\n ret = f(*args, **kwargs)\n self._enabled_state_pop()\n return ret\n\n # Bump call counters, before calling fn.\n # Note: elapsed_secs, process_secs not reflected yet of course\n self._add_call(logged=_enabled)\n\n _log_call_numbers = _get_final_value('log_call_numbers')\n # counters just got bumped\n _active_call_number = (self._stats.num_calls_logged\n if _log_call_numbers else\n 0)\n # Get list of callers up to & including first log_call's-deco'd fn\n # (or just caller, if no such fn)\n call_list, prev_indent_level = self.call_chain_to_next_log_calls_fn()\n\n # Bump _extra_indent_level if last fn on call_list is deco'd AND enabled,\n # o/w it's the _extra_indent_level which that fn 'inherited'.\n # _extra_indent_level: prev_indent_level, or prev_indent_level + 1\n do_indent = _get_final_value('indent')\n _extra_indent_level = (prev_indent_level +\n int(not not do_indent and not not _enabled))\n # 0.3.0\n ########## prefixed_fname = _get_final_value('prefix') + f.__name__\n prefixed_fname = _get_final_value('prefix') + self.f_display_name\n\n # Stackframe hack:\n assert '_deco_base__active_call_items__' == STACKFRAME_HACK_DICT_NAME\n _deco_base__active_call_items__ = {\n '_enabled': _enabled,\n '_log_call_numbers': _log_call_numbers,\n '_prefixed_fname': prefixed_fname, # Hack alert (Pt 1)\n '_active_call_number': _active_call_number,\n '_extra_indent_level': _extra_indent_level,\n # 0.3.0 for _get_own_deco_wrapper\n '_wrapper_deco': self\n }\n\n # Get logging function IF ANY.\n # For the benefit of callees further down the call chain,\n # if this f is not enabled (_enabled <= 0).\n # Subclass can return None to suppress printed/logged output.\n logging_fn = self.get_logging_fn(_get_final_value)\n\n # Only do global indentation for print, not for loggers\n global_indent_len = max(_extra_indent_level, 0) * self.INDENT\n\n # 0.2.2.post1 - save output_fname for log_message use\n call_number_str = ((' [%d]' % _active_call_number)\n if _log_call_numbers else '')\n output_fname = prefixed_fname + call_number_str\n\n # 0.3.0\n # Note: DON'T combine with global_mute(),\n # cuz this value will be pushed,\n # and when popped any realtime changes to global mute\n # made during call to f would be ignored.\n mute = _get_final_value('mute')\n\n # 0.2.2 -- self._log_message() will use\n # the logging_fn, indent_len and output_fname at top of these stacks;\n # thus, verbose functions should use log_calls.print (~ log_message)\n # to write their blather.\n # There's a stack of logging-state ,\n # used by self._log_message(), maintained in this wrapper.\n self._logging_state_push(logging_fn, global_indent_len, output_fname, mute)\n\n # (_xxx variables set, ok to call f)\n if not _enabled:\n ret = f(*args, **kwargs)\n self._logging_state_pop(enabled_too=True)\n return ret\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Set up context, for pre-call handlers\n # (after calling f, add to it for post-call handlers)\n # THIS is the time sink - 23x slower than other 'blocks'\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Key/values of \"context\" whose values we know so far:\n context = {\n 'decorator': self,\n 'settings': self._settings_mapping,\n 'stats': self._stats,\n 'prefixed_fname': prefixed_fname,\n 'fparams': self.f_params,\n 'call_list': call_list,\n 'args': args,\n 'kwargs': kwargs,\n 'indent': \" \" * self.INDENT, # our unit of indentation\n 'output_fname': output_fname,\n }\n\n # Gather all the things we need (for log output, & for history)\n # Use inspect module's Signature.bind method.\n # bound_args.arguments -- contains only explicitly bound arguments\n # 0.2.4.post5 - using\n # inspect.signature(f).bind(*args, **kwargs)\n # took 45% of execution time of entire wrapper; this takes 23%:\n # 0.3.1 TODO BUG No args is a problem?!\n bound_args = self.f_signature.bind(*args, **kwargs)\n \"\"\"\n File \"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/tests/_temp.py\", line 12, in <module>\n g(f())\n File \"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/log_calls.py\", line 1935, in _deco_base_f_wrapper_\n bound_args = self.f_signature.bind(*args, **kwargs)\n File \"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\", line 2646, in bind\n return args[0]._bind(args[1:], kwargs)\n File \"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\", line 2571, in _bind\n raise TypeError('too many positional arguments') from None\n TypeError: too many positional arguments\n \"\"\"\n\n varargs_pos = get_args_pos(self.f_params) # -1 if no *args in signature\n argcount = varargs_pos if varargs_pos >= 0 else len(args)\n context['argcount'] = argcount\n # The first argcount-many things in bound_args\n context['argnames'] = list(bound_args.arguments)[:argcount]\n context['argvals'] = args[:argcount]\n\n context['varargs'] = args[argcount:]\n (context['varargs_name'],\n context['kwargs_name']) = get_args_kwargs_param_names(self.f_params)\n\n # These 3 statements = 31% of execution time of wrapper\n context['defaulted_kwargs'] = get_defaulted_kwargs_OD(self.f_params, bound_args)\n context['explicit_kwargs'] = get_explicit_kwargs_OD(self.f_params, bound_args, kwargs)\n # context['implicit_kwargs'] = {\n # k: kwargs[k] for k in kwargs if k not in context['explicit_kwargs']\n # }\n # At least 2x as fast:\n context['implicit_kwargs'] = \\\n difference_update(kwargs.copy(), context['explicit_kwargs'])\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call pre-call handlers, collect nonempty return values\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # only consult global mute in r/t\n if not (mute or self.global_mute()): # 0.3.0\n pre_msgs = []\n for setting_name in self._settings_mapping._pre_call_handlers: # keys\n if _get_final_value(setting_name):\n info = self._settings_mapping._get_DecoSetting(setting_name)\n msg = info.pre_call_handler(context)\n if msg:\n pre_msgs.append(msg)\n\n # Write pre-call messages\n if logging_fn:\n for msg in pre_msgs:\n self._log_message(msg, extra_indent_level=0)\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call f(*args, **kwargs) and get its retval; time it.\n # Add timestamp, elapsed time(s) and retval to context.\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # No dictionary overhead between timer(s) start & stop.\n t0 = time.time() # for timestamp\n t0_wall = wall_time_fn()\n t0_process = process_time_fn()\n retval = f(*args, **kwargs)\n t_end_wall = wall_time_fn()\n t_end_process = process_time_fn()\n context['elapsed_secs'] = (t_end_wall - t0_wall)\n context['process_secs'] = (t_end_process - t0_process)\n context['timestamp'] = t0\n context['retval'] = retval\n\n self._add_to_elapsed(context['elapsed_secs'], context['process_secs'])\n\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # Call post-call handlers, collect nonempty return values\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n # only consult global mute in r/t\n if not (mute or self.global_mute()): # 0.3.0\n post_msgs = []\n for setting_name in self._settings_mapping._post_call_handlers: # keys\n if _get_final_value(setting_name):\n info = self._settings_mapping._get_DecoSetting(setting_name)\n msg = info.post_call_handler(context)\n if msg:\n post_msgs.append(msg)\n\n # Write post-call messages\n if logging_fn:\n for msg in post_msgs:\n self._log_message(msg, extra_indent_level=0)\n # v0.3.0b22 -- if recording history, add record of call even if we're muted(!)\n elif _get_final_value('record_history'):\n info = self._settings_mapping._get_DecoSetting('record_history')\n _ = info.post_call_handler(context)\n\n self._logging_state_pop(enabled_too=True)\n\n return retval\n\n self._add_function_attrs(f, _deco_base_f_wrapper_)\n return _deco_base_f_wrapper_\n\n #-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+\n # end else (case \"f_or_klass is a function\",\n # subcase \"f is a function & is NOT already deco'd\")\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*","def f0(self):\n return self._f0","def getcfix(self):\n cfix_ = ctypes.c_double()\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\n if res != 0:\n _,msg = self.__getlasterror(res)\n raise Error(rescode(res),msg)\n cfix_ = cfix_.value\n _cfix_return_value = cfix_\n return (_cfix_return_value)","def AceType(self):\n raw_type = super(AceHeader, self).AceType\n return ACE_CLASS_TYPE_MAPPER[raw_type]","def f_get_class_name(self):\n return self.__class__.__name__","def getElementName(self):\n return _libsbml.ListOfFunctionDefinitions_getElementName(self)","def XPGetWidgetClassFunc(inWidgetClass):\n pass","def get_function(self,attr):\n func_name = self._user_funcs[attr] \n if hasattr(func_name,'__call__'):\n func_name = func_name(self)\n return func_name","def cpf(self):\n return self._cpf","def get_score_function(self, idattr):\n return self.get_node(\n '//ScoreFunctions/ScoreFunction[@id=\"%s\"]' % idattr)","def f1(self, option='average'):\n\t\tif option == 'classwise':\treturn self.class_f1_\n\t\telif option == 'average':\treturn self.avg_f1_","def get_cfft(self):\n return self.get_rfft().get_cfft()","def getTypeCode(self):\n return _libsbml.CubicBezier_getTypeCode(self)","def functional(self):\n return self.__functional","def functional(self):\n return self.vasprun.run_type","def firedrake_fspace(self, shape=None):\n if shape is None:\n from firedrake import FunctionSpace\n return FunctionSpace(self._mesh_geometry,\n self._ufl_element.family(),\n degree=self._ufl_element.degree())\n elif isinstance(shape, int):\n from firedrake import VectorFunctionSpace\n return VectorFunctionSpace(self._mesh_geometry,\n self._ufl_element.family(),\n degree=self._ufl_element.degree(),\n dim=shape)\n elif isinstance(shape, tuple):\n from firedrake import TensorFunctionSpace\n return TensorFunctionSpace(self._mesh_geometry,\n self._ufl_element.family(),\n degree=self._ufl_element.degree(),\n shape=shape)\n else:\n raise TypeError(\"'shape' must be *None*, an integer, \"\n \" or a tuple of integers, not of type '%s'.\"\n % type(shape))","def class_impl(self):\n # timing is stored by node, we compute timing by class on demand\n rval = {}\n for fgraph, node in self.apply_callcount:\n typ = type(node.op)\n if self.apply_cimpl[node]:\n impl = \"C \"\n else:\n impl = \"Py\"\n rval.setdefault(typ, impl)\n if rval[typ] != impl and len(rval[typ]) == 2:\n rval[typ] += impl\n return rval"],"string":"[\n \"def _function_class(self):\\n return FriCASExpectFunction\",\n \"def getFunctionClass(functionID):\\n d = { 1: Linear,\\n 2: LinearDrag,\\n 11: Gaussian,\\n 12: GaussianDrag,\\n 21: Lorentzian,\\n 22: LorentzianDrag }\\n return d[functionID]\",\n \"def cfunc_type(self):\\n tif = ida_typeinf.tinfo_t()\\n result = self.get_func_type(tif)\\n if not result:\\n return\\n return tif\",\n \"def get_func_type(self, *args):\\n return _ida_hexrays.cfunc_t_get_func_type(self, *args)\",\n \"def get_class_for(self, elem):\\r\\n\\t\\treturn self.__tag_to_cls.get(elem.tag, self.__default_cls)\",\n \"def get_function(self):\\n return self.element.get_basis_functions()[self.n]\",\n \"def func(self):\\n return self.__class__\",\n \"def get_class_functional_name(name):\\n name = _strip_class_name(name)\\n return name\",\n \"def get_ctf(ima):\\n\\tfrom EMAN2 import EMAN2Ctf\\n\\tctf_params = ima.get_attr(\\\"ctf\\\")\\t\\n\\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang\",\n \"def element_type(self) -> global___Type:\",\n \"def w_class_a(self):\\n return self._w_class_a\",\n \"def getFunction(self) -> ghidra.program.model.listing.Function:\\n ...\",\n \"def arg_to_CFI(self, node, ordered_functions):\\n options = node.options\\n fmt_func = node.fmtdict\\n\\n if options.wrap_fortran is False:\\n # The buffer function is intended to be called by Fortran.\\n # No Fortran, no need for buffer function.\\n return\\n\\n ast = node.ast\\n declarator = ast.declarator\\n result_typemap = ast.typemap\\n # shadow classes have not been added yet.\\n # Only care about string, vector here.\\n result_is_ptr = declarator.is_indirect()\\n if (\\n result_typemap\\n and result_typemap.base in [\\\"string\\\", \\\"vector\\\"]\\n and result_typemap.name != \\\"char\\\"\\n and not result_is_ptr\\n ):\\n node.wrap.c = False\\n # node.wrap.fortran = False\\n self.config.log.write(\\n \\\"Skipping {}, unable to create C wrapper \\\"\\n \\\"for function returning {} instance\\\"\\n \\\" (must return a pointer or reference).\\\"\\n \\\" Bufferify version will still be created.\\\\n\\\".format(\\n result_typemap.cxx_type, declarator.user_name\\n )\\n )\\n \\n cfi_args = {}\\n for arg in ast.declarator.params:\\n declarator = arg.declarator\\n name = declarator.user_name\\n attrs = declarator.attrs\\n meta = declarator.metaattrs\\n cfi_args[name] = False\\n arg_typemap = arg.typemap\\n if meta[\\\"api\\\"]:\\n # API explicitly set by user.\\n continue\\n elif meta[\\\"assumed-rank\\\"]:\\n cfi_args[name] = True\\n elif attrs[\\\"rank\\\"]:\\n cfi_args[name] = True\\n elif arg_typemap.sgroup == \\\"string\\\":\\n cfi_args[name] = True\\n elif arg_typemap.sgroup == \\\"char\\\":\\n if declarator.is_indirect():\\n cfi_args[name] = True\\n elif meta[\\\"deref\\\"] in [\\\"allocatable\\\", \\\"pointer\\\"]:\\n cfi_args[name] = True\\n has_cfi_arg = any(cfi_args.values())\\n\\n # Function result.\\n need_buf_result = None\\n\\n result_as_arg = \\\"\\\" # Only applies to string functions\\n # when the result is added as an argument to the Fortran api.\\n\\n # Check if result needs to be an argument.\\n declarator = ast.declarator\\n attrs = declarator.attrs\\n meta = declarator.metaattrs\\n if meta[\\\"deref\\\"] == \\\"raw\\\":\\n # No bufferify required for raw pointer result.\\n pass\\n elif result_typemap.sgroup == \\\"string\\\":\\n need_buf_result = \\\"cfi\\\"\\n result_as_arg = fmt_func.F_string_result_as_arg\\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\\n elif result_typemap.sgroup == \\\"char\\\" and result_is_ptr:\\n need_buf_result = \\\"cfi\\\"\\n result_as_arg = fmt_func.F_string_result_as_arg\\n result_name = result_as_arg or fmt_func.C_string_result_as_arg\\n elif meta[\\\"deref\\\"] in [\\\"allocatable\\\", \\\"pointer\\\"]:\\n need_buf_result = \\\"cfi\\\"\\n\\n if not (need_buf_result or\\n has_cfi_arg):\\n return False\\n\\n options.wrap_fortran = False\\n\\n # Create a new C function and change arguments\\n # and add attributes.\\n C_new = node.clone()\\n ordered_functions.append(C_new)\\n self.append_function_index(C_new)\\n\\n generated_suffix = \\\"cfi\\\"\\n C_new._generated = \\\"arg_to_cfi\\\"\\n C_new.splicer_group = \\\"cfi\\\"\\n if need_buf_result:\\n C_new.ast.declarator.metaattrs[\\\"api\\\"] = need_buf_result\\n fmt_func = C_new.fmtdict\\n fmt_func.function_suffix = fmt_func.function_suffix + fmt_func.C_cfi_suffix\\n\\n C_new.wrap.assign(c=True)#, fortran=True)\\n C_new._PTR_C_CXX_index = node._function_index\\n\\n for arg in C_new.ast.declarator.params:\\n name = arg.declarator.user_name\\n if cfi_args[name]:\\n arg.declarator.metaattrs[\\\"api\\\"] = generated_suffix\\n\\n ast = C_new.ast\\n if True: # preserve to avoid changing indention for now.\\n f_attrs = node.ast.declarator.attrs # Fortran function attributes\\n f_meta = node.ast.declarator.metaattrs # Fortran function attributes\\n if result_as_arg:\\n # decl: const char * getCharPtr2() +len(30)\\n # +len implies copying into users buffer.\\n result_as_string = ast.result_as_arg(result_name)\\n result_as_string.const = False # must be writeable\\n attrs = result_as_string.declarator.attrs\\n # Special case for wrapf.py to override \\\"allocatable\\\"\\n f_meta[\\\"deref\\\"] = None\\n result_as_string.declarator.metaattrs[\\\"api\\\"] = \\\"cfi\\\"\\n result_as_string.declarator.metaattrs[\\\"deref\\\"] = \\\"result\\\"\\n result_as_string.declarator.metaattrs[\\\"is_result\\\"] = True\\n C_new.ast.declarator.metaattrs[\\\"api\\\"] = None\\n C_new.ast.declarator.metaattrs[\\\"intent\\\"] = \\\"subroutine\\\"\\n C_new.ast.declarator.metaattrs[\\\"deref\\\"] = None\\n\\n if result_as_arg:\\n F_new = self.result_as_arg(node, C_new)\\n ordered_functions.append(F_new)\\n self.append_function_index(F_new)\\n else:\\n if node._generated in [\\\"result_to_arg\\\", \\\"fortran_generic\\\", \\\"getter/setter\\\"]:\\n node.wrap.c = False\\n # Fortran function may call C subroutine if string/vector result\\n # Fortran function calls bufferify function.\\n node._PTR_F_C_index = C_new._function_index\\n return True\",\n \"def get_class(self):\\n\\t\\treturn self.CLASS\",\n \"def _fc_function_definitions(self) -> str:\\n result = 'extern \\\"C\\\" {\\\\n\\\\n'\\n for namespace in self.namespaces:\\n for member in namespace.members:\\n result += member.fortran_c_wrapper()\\n\\n result += '}\\\\n\\\\n'\\n return result\",\n \"def GetClassConstructor(filename):\\n ode_name = GetOdeSystemName(filename)\\n\\n class_constr_str = (ode_name + \\\"::~\\\" + ode_name + \\\"()\\\\n\\\"\\n \\\"{\\\\n\\\" \\n \\\"}\\\\n\\\\n\\\")\\n\\n return class_constr_str\",\n \"def getColorTransferFunction(self):\\n\\t\\treturn self.ctf\",\n \"def C(classname):\\n return objc.objc_getClass(_utf8(classname))\",\n \"def fortran_c_wrapper(self) -> str:\\n if self.fc_override is not None:\\n return self.fc_override.replace('$CLASSNAME$', self.class_name).replace(\\n \\\"$C_PREFIX$\\\", self.c_prefix).replace(\\\"$F_PREFIX$\\\", self.f_prefix)\\n\\n result = ''\\n\\n # declaration\\n in_parameters = self._fc_in_parameters()\\n return_type, out_parameters = self._fc_out_parameters()\\n if self.may_throw:\\n out_parameters.append('int * err_code')\\n out_parameters.append('char ** err_msg')\\n out_parameters.append('std::size_t * err_msg_len')\\n\\n func_name = '{}_{}_{}_'.format(\\n self.c_prefix, self.class_name, self.name)\\n\\n par_str = ', '.join(in_parameters + out_parameters)\\n result += '{} {}({}) {{\\\\n'.format(return_type, func_name, par_str)\\n\\n # convert input\\n for par in self.params:\\n result += '{}'.format(par.fc_convert_input())\\n\\n # call C++ function and return result\\n if self.may_throw:\\n result += ' try {\\\\n'\\n result += ' *err_code = 0;\\\\n'\\n result += indent(self._fc_cpp_call(), 4*' ')\\n result += indent(self._fc_return(), 4*' ')\\n result += ' }\\\\n'\\n for exception, code in error_codes.items():\\n if code != 0:\\n catch = ''\\n catch += 'catch (std::{} const & e) {{\\\\n'.format(exception)\\n catch += ' *err_code = {};\\\\n'.format(code)\\n catch += ' static std::string msg;\\\\n'\\n catch += ' msg = e.what();\\\\n'\\n catch += ' *err_msg = const_cast<char*>(msg.data());\\\\n'\\n catch += ' *err_msg_len = msg.size();\\\\n'\\n catch += '}\\\\n'\\n result += indent(catch, 4*' ')\\n result += self._fc_return_default()\\n else:\\n result += self._fc_cpp_call()\\n result += self._fc_return()\\n result += '}\\\\n\\\\n'\\n return result\",\n \"def getClass(self):\\n return _libsbml.ASTNode_getClass(self)\",\n \"def CFL(self):\\n return self.__CFL\",\n \"def _getTypeClass(self, namespaceURI, localName):\\r\\n bti = BaseTypeInterpreter()\\r\\n simpleTypeClass = bti.get_typeclass(localName, namespaceURI)\\r\\n return simpleTypeClass\",\n \"def get_func_type(self, *args):\\n return _ida_hexrays.cfuncptr_t_get_func_type(self, *args)\",\n \"def fortran_function(self) -> str:\\n return ''.join([i.fortran_function() for i in self.instances])\",\n \"def get_css_class(self):\\n\\n return None\",\n \"def get_unc_f(self):\\n return self.uncf\",\n \"def tctfdfc(x):\\n if isinstance(x,Fdf) :\\n pass\\n else : \\n x = Fdf.constant(x)\\n return x\",\n \"def _func(self):\\n return self._get_flint_func(self.domain)\",\n \"def _get_class():\\n return ASParameters\",\n \"def get_asset_class_feature_value(self):\\n ac_feature = AssetFeature.Standard.ASSET_CLASS.get_object()\\n return AssetFeatureValue.objects.get_or_create(name=self.asset_class.display_name, feature=ac_feature)[0]\",\n \"def get_surface_class(self) -> SurfaceClassStr:\\n return SURFACE_CLASS.inverse[self.surfaceClass()]\",\n \"def choose_botorch_acqf_class() -> Type[AcquisitionFunction]:\\n # NOTE: In the future, this dispatch function could leverage any\\n # of the attributes of `BoTorchModel` or kwargs passed to\\n # `BoTorchModel.gen` to intelligently select acquisition function.\\n return qNoisyExpectedImprovement\",\n \"def get_class_name(self):\\n\\n if \\\"class\\\" in self._root.attrib:\\n return self._root.attrib['class']\\n else:\\n return self._root.tag\",\n \"def ICRFFile_Instance():\\n return _gmat_py.ICRFFile_Instance()\",\n \"def getTypeCode(self):\\n return _libsbml.FunctionDefinition_getTypeCode(self)\",\n \"def ffb_c (cls):\\r\\n b=(bin(c0+jmax)[-(cls)])\\r\\n return (2*int(b)-1) # or -(2*int(b)-1), depending on the final direction of the hierarchy\\r\",\n \"def getClass():\\n\\treturn AnimatorMode\",\n \"def make_fee_class():\\r\\n fee = 1\\r\\n return make_class(locals())\",\n \"def get_func_type(header):\\n func_type = header.functionType\\n if func_type == SSE.SCALAR:\\n return FunctionType.Scalar\\n elif func_type == SSE.AGGREGATION:\\n return FunctionType.Aggregation\\n elif func_type == SSE.TENSOR:\\n return FunctionType.Tensor\",\n \"def get_elf_class(afile):\\n if afile in g_elf_class_db:\\n return g_elf_class_db[afile]\\n cmd = 'readelf -h ' + cmd_quote(afile) + ' | grep \\\"Class:\\\" || true'\\n output = get_shell_cmd_output(cmd)\\n tokens = output.split(\\\":\\\")\\n elf_class = tokens[1].strip()\\n verbose(afile + \\\" ELF Class is \\\" + elf_class, LEVEL_2)\\n g_elf_class_db[afile] = elf_class\\n return elf_class\",\n \"def get_classless_functions(root_node):\\n\\n file_ast = ast.parse(root_node)\\n\\n visitor = ClasslessFunctionVisitor()\\n\\n visitor.visit(file_ast)\\n\\n return visitor.function_list\",\n \"def get_element_type(cls):\\r\\n return cls._type_name(cls.element_type)\",\n \"def fortran_function(self) -> str:\\n if self.f_override is not None:\\n return indent(\\n self.f_override.replace('$CLASSNAME$', self.class_name).replace(\\n \\\"$C_PREFIX$\\\", self.c_prefix).replace(\\n \\\"$F_PREFIX$\\\", self.f_prefix),\\n 4*' ')\\n\\n result = ''\\n\\n # declaration\\n func_name = '{}_{}_{}'.format(\\n self.f_prefix, self.class_name, self.name)\\n in_parameters = self._f_in_parameters()\\n return_type, out_parameters = self._f_out_parameters()\\n if self.may_throw:\\n out_parameters.append(('integer, optional', 'err_code'))\\n out_parameters.append(('character(:), allocatable, optional',\\n 'err_msg'))\\n\\n all_parameters = in_parameters + out_parameters\\n arg_list = ', &\\\\n'.join([par_name for _, par_name in all_parameters])\\n arg_ilist = indent(arg_list, 8*' ')\\n if return_type != '':\\n result += 'function {}( &\\\\n{})\\\\n'.format(func_name, arg_ilist)\\n else:\\n result += 'subroutine {}( &\\\\n{})\\\\n'.format(func_name, arg_ilist)\\n\\n # parameter declarations\\n result += ' implicit none\\\\n'\\n for par_type, par_name in in_parameters:\\n result += ' {}, intent(in) :: {}\\\\n'.format(\\n par_type, par_name)\\n for par_type, par_name in out_parameters:\\n result += ' {}, intent(out) :: {}\\\\n'.format(par_type, par_name)\\n if return_type != '':\\n result += ' {} :: {}\\\\n'.format(return_type, func_name)\\n result += '\\\\n'\\n\\n # variable declarations\\n c_return_type, fi_out_parameters = self._fi_out_parameters()\\n if c_return_type:\\n result += ' {} :: ret_val\\\\n'.format(c_return_type)\\n for par_type, par_name in fi_out_parameters:\\n result += ' {} :: {}\\\\n'.format(par_type, par_name)\\n for par_type, par_name in self.ret_type.f_aux_variables():\\n result += ' {} :: {}\\\\n'.format(par_type, par_name)\\n if self.may_throw:\\n result += ' integer (c_int) :: err_code_v\\\\n'\\n result += ' type (c_ptr) :: err_msg_v\\\\n'\\n result += ' integer (c_size_t) :: err_msg_len_v\\\\n'\\n result += ' character (c_char), dimension(:), pointer :: err_msg_f\\\\n'\\n result += ' character(:), allocatable :: err_msg_p\\\\n'\\n result += ' integer (c_size_t) :: err_msg_i\\\\n'\\n if c_return_type or fi_out_parameters or self.may_throw:\\n result += '\\\\n'\\n\\n # convert input\\n args = [param.f_chain_arg() for param in self.params]\\n args += [par_name for _, par_name in fi_out_parameters]\\n if self.may_throw:\\n args += ['err_code_v', 'err_msg_v', 'err_msg_len_v']\\n arg_str = ', &\\\\n'.join([8*' ' + arg for arg in args])\\n\\n # call C function\\n fc_func_name = '{}_{}_{}_'.format(\\n self.c_prefix, self.class_name, self.name)\\n chain_call = self.fc_chain_call(\\n ns_prefix=self.c_prefix, class_name=self.class_name,\\n fc_func_name=fc_func_name, fc_args=arg_str)\\n result_name = ''\\n if return_type != '':\\n result_name = func_name\\n elif out_parameters:\\n result_name = out_parameters[0][1]\\n result += self.ret_type.f_call_c('ret_val', chain_call)\\n\\n # handle errors if necessary\\n if self.may_throw:\\n # Note: I tried to factor this out into a function, but Fortran\\n # makes that near-impossible. Since we're generating anyway, it's\\n # not really duplication, so leave it as is.\\n result += indent(dedent(f\\\"\\\"\\\"\\\\\\n if (err_code_v .ne. 0) then\\n if (present(err_code)) then\\n err_code = err_code_v\\n if (present(err_msg)) then\\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\\n allocate (character(err_msg_len_v) :: err_msg)\\n do err_msg_i = 1, err_msg_len_v\\n err_msg(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\\n end do\\n end if\\n {dedent(\\n self.ret_type.f_return_dummy_result(result_name))}\\n return\\n else\\n call c_f_pointer(err_msg_v, err_msg_f, (/err_msg_len_v/))\\n allocate (character(err_msg_len_v) :: err_msg_p)\\n do err_msg_i = 1, err_msg_len_v\\n err_msg_p(err_msg_i:err_msg_i) = err_msg_f(err_msg_i)\\n end do\\n print *, err_msg_p\\n stop\\n end if\\n else\\n if (present(err_code)) then\\n err_code = 0\\n end if\\n end if\\n\\n \\\"\\\"\\\"), 4*' ')\\n\\n # convert and return result\\n result += self.ret_type.f_return_result(result_name, 'ret_val')\\n\\n # end\\n if return_type != '':\\n result += 'end function {}\\\\n\\\\n'.format(func_name)\\n else:\\n result += 'end subroutine {}\\\\n\\\\n'.format(func_name)\\n return indent(result, 4*' ')\",\n \"def Point_to_FeatureClass(self, fc):\\n\\n\\n feature_class = []\\n for index, traectory in enumerate(self.__traectory_list):\\n point_row = arcpy.Point(X=traectory[0], Y=traectory[1], Z=traectory[2], ID=index)\\n feature_class.append(arcpy.PointGeometry(point_row, arcpy.SpatialReference(2436)))\\n arcpy.CopyFeatures_management(feature_class, (self.workspace + '\\\\\\\\' + fc))\\n print 'Complete Creating a Point Feature Class'\\n\\n return None\",\n \"def create_class_instance(element, element_id, doc_id):\\n xsi_type = get_xsi_type(element)\\n element_class = XSI_TYPE_CLASSES[xsi_type]\\n return element_class.from_etree(element)\",\n \"def eurologic_backplane_function(self):\\n return self._eurologic_backplane_function\",\n \"def class_from_invokedynamic(ins, cf):\\n info = InvokeDynamicInfo.create(ins, cf)\\n assert info.created_type != \\\"void\\\"\\n return info.created_type\",\n \"def visit_class(self, flags, scope, token, parent):\\r\\n\\r\\n # define the class name in the current scope\\r\\n # see visit_block\\r\\n #scope.define(SC_FUNCTION, token.children[0])\\r\\n scope.defer(token)\",\n \"def _get_cos_traffic_class(self):\\n return self.__cos_traffic_class\",\n \"def get_fc(self, layer):\\n return getattr(self, 'attention_{}'.format(layer))\",\n \"def get_function(self):\\n return SSAFunction(self.get_graph())\",\n \"def function(self):\\n ret = libxml2mod.xmlXPathGetFunction(self._o)\\n return ret\",\n \"def element_type(self):\\r\\n result = conf.lib.clang_getElementType(self)\\r\\n if result.kind == TypeKind.INVALID:\\r\\n raise Exception('Element type not available on this type.')\\r\\n\\r\\n return result\",\n \"def fce(B):\\n return wce(B)/(2.*np.pi)\",\n \"def convert(self):\\n return _libsbml.SBMLFunctionDefinitionConverter_convert(self)\",\n \"def getFunction(self):\\n return gpio.gpio_function(self.bcm_id)\",\n \"def fortran_functions(self) -> str:\\n fun_name = f'{self.f_prefix}_{self.name}_to_int_'\\n result = f'integer function {fun_name}(flags)\\\\n'\\n result += ' implicit none\\\\n'\\n result += '\\\\n'\\n result += f' class({self.f_prefix}_{self.name}), intent(in) :: flags\\\\n'\\n result += ' integer :: ret_val\\\\n'\\n result += '\\\\n'\\n result += ' ret_val = 0\\\\n'\\n for i, value in enumerate(self.values):\\n result += f\\\" if (flags%{value}) ret_val = ret_val + {1 << i}\\\\n\\\"\\n result += f' {fun_name} = ret_val\\\\n'\\n result += f'end function {fun_name}\\\\n'\\n return indent(result, 4*' ')\",\n \"def define_function_spaces(self):\\n\\n cell = self.mesh.ufl_cell()\\n vec_degree = int(self.config['formulation']['element'][0][-1])\\n if vec_degree == 0:\\n vec_family = \\\"DG\\\"\\n else:\\n vec_family = \\\"CG\\\"\\n vec_element = dlf.VectorElement(vec_family, cell, vec_degree)\\n self.vectorSpace = dlf.FunctionSpace(self.mesh, vec_element)\\n\\n if self.config['material']['incompressible']:\\n scalar_degree = int(self.config['formulation']['element'][1][-1])\\n if scalar_degree == 0:\\n scalar_family = \\\"DG\\\"\\n else:\\n scalar_family = \\\"CG\\\"\\n scalar_element = dlf.FiniteElement(scalar_family, cell, scalar_degree)\\n self.scalarSpace = dlf.FunctionSpace(self.mesh, scalar_element)\\n else:\\n self.scalarSpace = None\\n\\n return None\",\n \"def get_attribute_class(self):\\n return self._attr_cls\",\n \"def getFunctionStyle(self,cpArray,X0):\\n cpArray=tf.cast(tf.convert_to_tensor(cpArray),dtype=tf.float64)\\n X0 = tf.cast(tf.convert_to_tensor(X0),dtype=tf.float64)\\n func = self.lamda_computeCPdiff(X0)\\n gatheredCps = tf.map_fn(func,cpArray,parallel_iterations=32,back_prop=False)\\n return gatheredCps\",\n \"def getElementName(self):\\n return _libsbml.FunctionDefinition_getElementName(self)\",\n \"def cython_ctype(self):\\n if self._cython_ctype is None:\\n self._cython_ctype = self.ts.cython_ctype(self.t)\\n return self._cython_ctype\",\n \"def feature_class():\\n if deployment_settings.get_security_map() and not s3_has_role(\\\"MapAdmin\\\"):\\n unauthorised()\\n\\n tablename = \\\"%s_%s\\\" % (module, resourcename)\\n table = db[tablename]\\n\\n # Model options\\n table.gps_marker.comment = DIV( _class=\\\"tooltip\\\",\\n _title=\\\"%s|%s\\\" % (T(\\\"GPS Marker\\\"),\\n T(\\\"Defines the icon used for display of features on handheld GPS.\\\")))\\n\\n # CRUD Strings\\n LIST_FEATURE_CLASS = T(\\\"List Feature Classes\\\")\\n s3.crud_strings[tablename] = Storage(\\n title_create = ADD_FEATURE_CLASS,\\n title_display = T(\\\"Feature Class Details\\\"),\\n title_list = T(\\\"Feature Classes\\\"),\\n title_update = T(\\\"Edit Feature Class\\\"),\\n title_search = T(\\\"Search Feature Class\\\"),\\n subtitle_create = T(\\\"Add New Feature Class\\\"),\\n subtitle_list = LIST_FEATURE_CLASS,\\n label_list_button = LIST_FEATURE_CLASS,\\n label_create_button = ADD_FEATURE_CLASS,\\n label_delete_button = T(\\\"Delete Feature Class\\\"),\\n msg_record_created = T(\\\"Feature Class added\\\"),\\n msg_record_modified = T(\\\"Feature Class updated\\\"),\\n msg_record_deleted = T(\\\"Feature Class deleted\\\"),\\n msg_list_empty = T(\\\"No Feature Classes currently defined\\\"))\\n\\n output = s3_rest_controller(module, resourcename)\\n\\n if not \\\"gis\\\" in response.view and response.view != \\\"popup.html\\\":\\n response.view = \\\"gis/\\\" + response.view\\n\\n return output\",\n \"def func_in_class():\\n return \\\"just a function hanging out in a class\\\"\",\n \"def __init__(self, function='cogscore/'):\\n self.function = function\",\n \"def frtype(self):\\n return self._frtype\",\n \"def record_cls(self):\\n return self._ELE_CLS\",\n \"def class_ref(self):\\n return self._class_ref\",\n \"def _get_ciphering_function(algorithm: Algorithm) -> Callable:\\n ciphering_function = None\\n if algorithm == Algorithm.caesar:\\n ciphering_function = cifra.cipher.caesar.cipher\\n elif algorithm == Algorithm.affine:\\n ciphering_function = cifra.cipher.affine.cipher\\n elif algorithm == Algorithm.substitution:\\n ciphering_function = cifra.cipher.substitution.cipher\\n elif algorithm == Algorithm.transposition:\\n ciphering_function = cifra.cipher.transposition.cipher\\n elif algorithm == Algorithm.vigenere:\\n ciphering_function = cifra.cipher.vigenere.cipher\\n return ciphering_function\",\n \"def get_attribute_class(self, attr_name):\\n return self.attrs.get_attribute_class(attr_name)\",\n \"def get_SrcClass(args):\\n return Reactome(args)\",\n \"def getFunction(self, name: unicode) -> ghidra.program.model.listing.Function:\\n ...\",\n \"def functionURI(self):\\n ret = libxml2mod.xmlXPathGetFunctionURI(self._o)\\n return ret\",\n \"def createFunctionTerm(self):\\n return _libsbml.Transition_createFunctionTerm(self)\",\n \"def _class(self):\\n return self.__class\",\n \"def get_function(self):\\n return Gumtree.gumtree.getFunction()\",\n \"def get_float_type(*args):\\n return _ida_hexrays.get_float_type(*args)\",\n \"def getTypeCode(self):\\n return _libsbml.FbcAnd_getTypeCode(self)\",\n \"def return_regressor_class(self):\\n pass\",\n \"def __init__(self, func_name, spin):\\n self.xc_func = None\\n self._xc_func_init = False\\n\\n # Handle func_name\\n if isinstance(func_name, str):\\n func_id = util.xc_functional_get_number(func_name)\\n if func_id == -1:\\n raise KeyError(\\\"LibXCFunctional: name '%s' not found.\\\" % func_name)\\n elif isinstance(func_name, (int, np.integer)):\\n func_id = func_name\\n if util.xc_functional_get_name(func_name) is None:\\n raise KeyError(\\\"LibXCFunctional: ID '%d' not found.\\\" % func_name)\\n else:\\n raise TypeError(\\\"LibXCFunctional: func_name must either be a string or int. Got {}\\\".format(func_name))\\n\\n self._xc_func_name = util.xc_functional_get_name(func_id)\\n\\n # Handle spin\\n if isinstance(spin, str):\\n spin = spin.lower()\\n if spin == \\\"polarized\\\":\\n self._spin = 2\\n elif spin == \\\"unpolarized\\\":\\n self._spin = 1\\n else:\\n raise KeyError(\\\"LibXCFunctional: spin must either be 'polarized' or 'unpolarized' if represented by a string. Got {}\\\".format(spin))\\n else:\\n self._spin = spin\\n\\n if self._spin not in [1, 2]:\\n raise KeyError(\\\"LibXCFunctional: spin must either be 1 or 2 if represented by a integer. Got {}\\\".format(self._spin))\\n\\n # Build the LibXC functional\\n self.xc_func = core.xc_func_alloc()\\n self.xc_func_size_names = [x for x in dir(self.xc_func.contents.dim) if not \\\"_\\\" in x]\\n\\n # Set all int attributes to zero (not all set to zero in libxc)\\n for attr in self.xc_func_size_names:\\n setattr(self.xc_func.contents, attr, 0)\\n\\n ret = core.xc_func_init(self.xc_func, func_id, self._spin)\\n if ret != 0:\\n raise ValueError(\\\"LibXC Functional construction did not complete. Error code %d\\\" % ret)\\n self._xc_func_init = True\\n\\n # Pull out all sizes after init\\n self.xc_func_sizes = {}\\n for attr in self.xc_func_size_names:\\n self.xc_func_sizes[attr] = getattr(self.xc_func.contents.dim, attr)\\n\\n # Unpack functional info\\n self.xc_func_info = core.xc_func_get_info(self.xc_func)\\n self._number = core.xc_func_info_get_number(self.xc_func_info)\\n self._kind = core.xc_func_info_get_kind(self.xc_func_info)\\n self._name = core.xc_func_info_get_name(self.xc_func_info).decode(\\\"UTF-8\\\")\\n self._family = core.xc_func_info_get_family(self.xc_func_info)\\n self._flags = core.xc_func_info_get_flags(self.xc_func_info)\\n\\n # Set needed flags\\n self._needs_laplacian = self._flags & flags.XC_FLAGS_NEEDS_LAPLACIAN\\n\\n # Set derivatives\\n self._have_exc = self._flags & flags.XC_FLAGS_HAVE_EXC\\n self._have_vxc = self._flags & flags.XC_FLAGS_HAVE_VXC\\n self._have_fxc = self._flags & flags.XC_FLAGS_HAVE_FXC\\n self._have_kxc = self._flags & flags.XC_FLAGS_HAVE_KXC\\n self._have_lxc = self._flags & flags.XC_FLAGS_HAVE_LXC\\n\\n # Set omega\\n self._have_cam = self._flags & flags.XC_FLAGS_HYB_CAM\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_CAMY\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LC\\n self._have_cam |= self._flags & flags.XC_FLAGS_HYB_LCY\\n self._cam_omega = self._cam_alpha = self._cam_beta = False\\n if self._have_cam:\\n self._cam_omega = self.xc_func.contents.cam_omega\\n self._cam_alpha = self.xc_func.contents.cam_alpha\\n self._cam_beta = self.xc_func.contents.cam_beta\\n\\n elif self._family in [flags.XC_FAMILY_HYB_LDA, flags.XC_FAMILY_HYB_GGA, flags.XC_FAMILY_HYB_MGGA]:\\n self._cam_alpha = self.xc_func.contents.cam_alpha\\n\\n # VV10\\n self._have_vv10 = self._flags & flags.XC_FLAGS_VV10\\n self._nlc_b = self._nlc_C = False\\n if self._have_vv10:\\n self._nlc_b = self.xc_func.contents.nlc_b\\n self._nlc_C = self.xc_func.contents.nlc_C\\n\\n # Stable\\n self._stable = self._flags & flags.XC_FLAGS_STABLE\\n self._dev = self._flags & flags.XC_FLAGS_DEVELOPMENT\\n\\n # Pull out references\\n self._refs = []\\n self._bibtexs = []\\n self._dois = []\\n\\n for pos in range(flags.XC_MAX_REFERENCES):\\n ref = core.xc_func_info_get_references(self.xc_func_info, pos)\\n if not ref: break\\n\\n self._refs.append(ref.contents.ref.decode(\\\"UTF-8\\\"))\\n self._bibtexs.append(ref.contents.bibtex.decode(\\\"UTF-8\\\"))\\n self._dois.append(ref.contents.doi.decode(\\\"UTF-8\\\"))\",\n \"def get_class_from_frame(fr):\\n args, _, _, value_dict = inspect.getargvalues(fr)\\n # Check the first parameter of the frame function is named `self` - if it is,\\n # `self` will be referenced in `value_dict`.\\n if len(args) != 0 and args[0] == 'self':\\n instance = value_dict.get('self', None)\\n if instance:\\n return getattr(instance, '__class__', None)\\n return None\",\n \"def get_attribute_class(self, name):\\n self.validate_attribute_name(name)\\n return self.schema[name].get_attribute_class()\",\n \"def as_ret(self):\\n if self.type in OCClass.classes_indexed_by_name:\\n oc_class = OCClass.classes_indexed_by_name[self.type]\\n if oc_class.class_meths:\\n for f in oc_class.class_meths:\\n #TODO MachOTask.currentTask.analyze_function(start_addr=f)\\n print 'T'\",\n \"def __call__(self, f_or_klass):\\n\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\\n # 0.3.0\\n # -- implement \\\"kill switch\\\", NO_DECO\\n # -- handle decorating both functions and classes\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*****\\n\\n # 0.3.0b16: if it isn't callable, scram'\\n if not callable(f_or_klass):\\n return f_or_klass\\n\\n # Special-case handling for ``NO_DECO``: remove from settings of ``self``\\n if self._effective_settings.get('NO_DECO'):\\n return f_or_klass\\n # else, delete that item wherever it might be\\n if 'NO_DECO' in self._effective_settings:\\n del self._effective_settings['NO_DECO']\\n if 'NO_DECO' in self._changed_settings:\\n del self._changed_settings['NO_DECO']\\n\\n f = f_or_klass if inspect.isfunction(f_or_klass) else None\\n klass = f_or_klass if inspect.isclass(f_or_klass) else None\\n\\n self.f = f\\n self.cls = klass\\n\\n if klass:\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a class\\\" -- namely, klass\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n\\n self._class__call__(klass) # modifies klass (methods & inner classes) (if not builtin)\\n self._add_class_attrs(klass) # v0.3.0v20 traps TypeError for builtins\\n return klass\\n\\n elif not f:\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a callable but not a function\\\"\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # functools.partial objects are callable, have no __name__ much less __qualname__,\\n # and trying to deco __call__ gets messy.\\n # Callable builtins e.g. len are not functions in the isfunction sense,\\n # can't deco anyway. Just give up (quietly):\\n return f_or_klass\\n\\n else: # not a class, f nonempty is a function of f_or_klass callable\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n # 0.3.0 -- case \\\"f_or_klass is a function\\\" -- namely, f\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\\n\\n #----------------------------------------------------------------\\n # Don't double-decorate -- don't wanna, & it doesn't work anyway!\\n #----------------------------------------------------------------\\n # Note: As with methods of classes,\\n # . if f is deco'd, its existing EXPLICITLY GIVEN settings take precedence.\\n\\n # # From _class__call__, props & methods cases, w/a few name changes\\n deco_obj = getattr(f, self._sentinels['DECO_OF'], None) # type: _deco_base\\n\\n # get a fresh copy for each attr\\n new_settings = self._changed_settings.copy() # updated below\\n\\n # __init__ fixup, a nicety:\\n # By default, don't log retval for __init__.\\n # If user insists on it with 'log_retval=True' in __init__ deco,\\n # that will override this.\\n if f.__name__ == '__init__':\\n self.fixup_for_init(new_settings)\\n\\n if deco_obj: # f is deco'd by this decorator\\n # Yes. Figure out settings for f,\\n ### 0.3.0b18 -- Use self._override\\n self._update_settings(new=new_settings,\\n old=deco_obj._changed_settings,\\n override_existing=self._override)\\n # update func's settings (_force_mutable=True to handle `max_history` properly)\\n deco_obj._settings_mapping.update(new_settings, _force_mutable=True)\\n return f\\n\\n #----------------------------------------------------------------\\n # f is a function & is NOT already deco'd\\n #----------------------------------------------------------------\\n\\n # 0.3.0.x -- f may not have a .__qualname__\\n try:\\n self._classname_of_f = '.'.join( f.__qualname__.split('.')[:-1] )\\n except AttributeError as e:\\n self._classname_of_f = ''\\n\\n # Special-case '__repr__' handling, if deco subclass doesn't allow it.\\n if f.__name__ == '__repr__' and self._classname_of_f and not self.allow_repr():\\n # v0.3.0b23 -- Instead of refusing to deco, use recursive_repr\\n # return f\\n return recursive_repr(fillvalue=\\\"...\\\")(f)\\n\\n # 0.3.0\\n # Use __qualname__ ALL the time, unless user provides `name=display_name_str`\\n # where `display_name_str` is either the name to be used for the fn in logged output,\\n # or is an oldstyle format str into which f.__name__ will be substituted\\n # to obtain the display name.\\n # We require Py3.3+, so __qualname__ is available.\\n\\n # setup f_display_name\\n if self._name_param:\\n try:\\n self.f_display_name = (self._name_param % f.__name__)\\n except TypeError:\\n self.f_display_name = self._name_param\\n else:\\n self.f_display_name = f.__qualname__\\n\\n # TODO TRY THIS -- anything break?\\n # 0.3.1 Inspired by fractions.Fraction.__sub__ et al:\\n # __name__ may be very different from __qualname__;\\n # if so, show both\\n if f.__name__ not in f.__qualname__:\\n self.f_display_name += \\\" (\\\" + f.__name__ + \\\")\\\"\\n\\n #================================================================\\n # 0.3.0 -- Init things (migrated from __init__)\\n #----------------------------------------------------------------\\n # set up pseudo-dict (DecoSettingsMapping),\\n # using settings given by self._effective_settings.\\n #\\n # *** DecoSettingsMapping \\\"API\\\" --\\n # (2) construct DecoSettingsMapping object\\n # that will provide mapping & attribute access to settings, & more\\n #----------------------------------------------------------------\\n self._settings_mapping = DecoSettingsMapping(\\n deco_class=self.__class__,\\n # DecoSettingsMapping calls the rest ** values_dict\\n ** self._effective_settings # 0.3.0 set by __init__\\n )\\n\\n #----------------------------------------------------------------\\n # Init more stuff\\n #----------------------------------------------------------------\\n self._stats = ClassInstanceAttrProxy(\\n class_instance=self,\\n data_descriptor_names=self.__class__._data_descriptor_names,\\n method_descriptor_names=self.__class__._method_descriptor_names)\\n # Accessed by descriptors on the stats obj\\n self._num_calls_total = 0\\n self._num_calls_logged = 0\\n # max_history > 0 --> size of self._call_history; <= 0 --> unbounded\\n # Set before calling _make_call_history\\n\\n # 0.3.0 self._other_values_dict set by __init__\\n self.max_history = self._other_values_dict.get('max_history', 0) # <-- Nota bene\\n self._call_history = self._make_call_history()\\n\\n # Accumulate this (for logged calls only)\\n # even when record_history is false:\\n self._elapsed_secs_logged = 0.0\\n self._process_secs_logged = 0.0\\n\\n # 0.2.2.post1\\n # stack(s), pushed & popped wrapper of deco'd function\\n # by _logging_state_push, _logging_state_pop\\n # 0.3.0 convert to pushing/popping single namedtuples\\n self.logging_state_stack = [] # 0.3.0 stack of LoggingState namedtuples\\n self._enabled_stack = [] # 0.3.0 - um, stack, of 'enabled's\\n\\n #----------------------------------------------------------------\\n # end of Init passage\\n #================================================================\\n\\n # Save signature and parameters of f\\n self.f_signature = inspect.signature(f) # Py >= 3.3\\n self.f_params = self.f_signature.parameters\\n\\n # 0.3.0 We assume Py3.3 so we use perf_counter, process_time all the time\\n wall_time_fn = time.perf_counter\\n process_time_fn = time.process_time\\n\\n #############################\\n # The wrapper of a callable\\n #############################\\n\\n @wraps(f)\\n def _deco_base_f_wrapper_(*args, **kwargs):\\n \\\"\\\"\\\"Wrapper around the wrapped function f.\\n When this runs, f has been called, so we can now resolve\\n any indirect values for the settings/keyword-params\\n of log_calls, using info in kwargs and self.f_params.\\\"\\\"\\\"\\n # *** Part of the DecoSettingsMapping \\\"API\\\" --\\n # (4) using self._settings_mapping.get_final_value in wrapper\\n # [[[ This/these is/are 4th chronologically ]]]\\n\\n # inner/local fn -- save a few cycles and characters -\\n # we call this a lot (<= 9x).\\n def _get_final_value(setting_name):\\n \\\"Use outer scope's kwargs and self.f_params\\\"\\n return self._settings_mapping.get_final_value(\\n setting_name, kwargs, fparams=self.f_params)\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # if nothing to do, hurry up & don't do it.\\n # NOTE: call_chain_to_next_log_calls_fn looks in stack frames\\n # to find (0.2.4) STACKFRAME_HACK_DICT_NAME (really!)\\n # It and its values (the following _XXX variables)\\n # must be set before calling f.\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n _enabled = _get_final_value('enabled')\\n # 0.3.0 in case f calls log_message (no output if f disabled)\\n self._enabled_state_push(_enabled)\\n\\n # 0.2.4.post5 \\\"true bypass\\\": if 'enabled' < 0 then scram\\n if _enabled < 0:\\n ret = f(*args, **kwargs)\\n self._enabled_state_pop()\\n return ret\\n\\n # Bump call counters, before calling fn.\\n # Note: elapsed_secs, process_secs not reflected yet of course\\n self._add_call(logged=_enabled)\\n\\n _log_call_numbers = _get_final_value('log_call_numbers')\\n # counters just got bumped\\n _active_call_number = (self._stats.num_calls_logged\\n if _log_call_numbers else\\n 0)\\n # Get list of callers up to & including first log_call's-deco'd fn\\n # (or just caller, if no such fn)\\n call_list, prev_indent_level = self.call_chain_to_next_log_calls_fn()\\n\\n # Bump _extra_indent_level if last fn on call_list is deco'd AND enabled,\\n # o/w it's the _extra_indent_level which that fn 'inherited'.\\n # _extra_indent_level: prev_indent_level, or prev_indent_level + 1\\n do_indent = _get_final_value('indent')\\n _extra_indent_level = (prev_indent_level +\\n int(not not do_indent and not not _enabled))\\n # 0.3.0\\n ########## prefixed_fname = _get_final_value('prefix') + f.__name__\\n prefixed_fname = _get_final_value('prefix') + self.f_display_name\\n\\n # Stackframe hack:\\n assert '_deco_base__active_call_items__' == STACKFRAME_HACK_DICT_NAME\\n _deco_base__active_call_items__ = {\\n '_enabled': _enabled,\\n '_log_call_numbers': _log_call_numbers,\\n '_prefixed_fname': prefixed_fname, # Hack alert (Pt 1)\\n '_active_call_number': _active_call_number,\\n '_extra_indent_level': _extra_indent_level,\\n # 0.3.0 for _get_own_deco_wrapper\\n '_wrapper_deco': self\\n }\\n\\n # Get logging function IF ANY.\\n # For the benefit of callees further down the call chain,\\n # if this f is not enabled (_enabled <= 0).\\n # Subclass can return None to suppress printed/logged output.\\n logging_fn = self.get_logging_fn(_get_final_value)\\n\\n # Only do global indentation for print, not for loggers\\n global_indent_len = max(_extra_indent_level, 0) * self.INDENT\\n\\n # 0.2.2.post1 - save output_fname for log_message use\\n call_number_str = ((' [%d]' % _active_call_number)\\n if _log_call_numbers else '')\\n output_fname = prefixed_fname + call_number_str\\n\\n # 0.3.0\\n # Note: DON'T combine with global_mute(),\\n # cuz this value will be pushed,\\n # and when popped any realtime changes to global mute\\n # made during call to f would be ignored.\\n mute = _get_final_value('mute')\\n\\n # 0.2.2 -- self._log_message() will use\\n # the logging_fn, indent_len and output_fname at top of these stacks;\\n # thus, verbose functions should use log_calls.print (~ log_message)\\n # to write their blather.\\n # There's a stack of logging-state ,\\n # used by self._log_message(), maintained in this wrapper.\\n self._logging_state_push(logging_fn, global_indent_len, output_fname, mute)\\n\\n # (_xxx variables set, ok to call f)\\n if not _enabled:\\n ret = f(*args, **kwargs)\\n self._logging_state_pop(enabled_too=True)\\n return ret\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Set up context, for pre-call handlers\\n # (after calling f, add to it for post-call handlers)\\n # THIS is the time sink - 23x slower than other 'blocks'\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Key/values of \\\"context\\\" whose values we know so far:\\n context = {\\n 'decorator': self,\\n 'settings': self._settings_mapping,\\n 'stats': self._stats,\\n 'prefixed_fname': prefixed_fname,\\n 'fparams': self.f_params,\\n 'call_list': call_list,\\n 'args': args,\\n 'kwargs': kwargs,\\n 'indent': \\\" \\\" * self.INDENT, # our unit of indentation\\n 'output_fname': output_fname,\\n }\\n\\n # Gather all the things we need (for log output, & for history)\\n # Use inspect module's Signature.bind method.\\n # bound_args.arguments -- contains only explicitly bound arguments\\n # 0.2.4.post5 - using\\n # inspect.signature(f).bind(*args, **kwargs)\\n # took 45% of execution time of entire wrapper; this takes 23%:\\n # 0.3.1 TODO BUG No args is a problem?!\\n bound_args = self.f_signature.bind(*args, **kwargs)\\n \\\"\\\"\\\"\\n File \\\"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/tests/_temp.py\\\", line 12, in <module>\\n g(f())\\n File \\\"/Users/brianoneill/Desktop/Programming/Python-package-staging/log_calls/log_calls/log_calls.py\\\", line 1935, in _deco_base_f_wrapper_\\n bound_args = self.f_signature.bind(*args, **kwargs)\\n File \\\"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\\\", line 2646, in bind\\n return args[0]._bind(args[1:], kwargs)\\n File \\\"/Library/Frameworks/Python.framework/Versions/3.4/lib/python3.4/inspect.py\\\", line 2571, in _bind\\n raise TypeError('too many positional arguments') from None\\n TypeError: too many positional arguments\\n \\\"\\\"\\\"\\n\\n varargs_pos = get_args_pos(self.f_params) # -1 if no *args in signature\\n argcount = varargs_pos if varargs_pos >= 0 else len(args)\\n context['argcount'] = argcount\\n # The first argcount-many things in bound_args\\n context['argnames'] = list(bound_args.arguments)[:argcount]\\n context['argvals'] = args[:argcount]\\n\\n context['varargs'] = args[argcount:]\\n (context['varargs_name'],\\n context['kwargs_name']) = get_args_kwargs_param_names(self.f_params)\\n\\n # These 3 statements = 31% of execution time of wrapper\\n context['defaulted_kwargs'] = get_defaulted_kwargs_OD(self.f_params, bound_args)\\n context['explicit_kwargs'] = get_explicit_kwargs_OD(self.f_params, bound_args, kwargs)\\n # context['implicit_kwargs'] = {\\n # k: kwargs[k] for k in kwargs if k not in context['explicit_kwargs']\\n # }\\n # At least 2x as fast:\\n context['implicit_kwargs'] = \\\\\\n difference_update(kwargs.copy(), context['explicit_kwargs'])\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call pre-call handlers, collect nonempty return values\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # only consult global mute in r/t\\n if not (mute or self.global_mute()): # 0.3.0\\n pre_msgs = []\\n for setting_name in self._settings_mapping._pre_call_handlers: # keys\\n if _get_final_value(setting_name):\\n info = self._settings_mapping._get_DecoSetting(setting_name)\\n msg = info.pre_call_handler(context)\\n if msg:\\n pre_msgs.append(msg)\\n\\n # Write pre-call messages\\n if logging_fn:\\n for msg in pre_msgs:\\n self._log_message(msg, extra_indent_level=0)\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call f(*args, **kwargs) and get its retval; time it.\\n # Add timestamp, elapsed time(s) and retval to context.\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # No dictionary overhead between timer(s) start & stop.\\n t0 = time.time() # for timestamp\\n t0_wall = wall_time_fn()\\n t0_process = process_time_fn()\\n retval = f(*args, **kwargs)\\n t_end_wall = wall_time_fn()\\n t_end_process = process_time_fn()\\n context['elapsed_secs'] = (t_end_wall - t0_wall)\\n context['process_secs'] = (t_end_process - t0_process)\\n context['timestamp'] = t0\\n context['retval'] = retval\\n\\n self._add_to_elapsed(context['elapsed_secs'], context['process_secs'])\\n\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # Call post-call handlers, collect nonempty return values\\n # - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\\n # only consult global mute in r/t\\n if not (mute or self.global_mute()): # 0.3.0\\n post_msgs = []\\n for setting_name in self._settings_mapping._post_call_handlers: # keys\\n if _get_final_value(setting_name):\\n info = self._settings_mapping._get_DecoSetting(setting_name)\\n msg = info.post_call_handler(context)\\n if msg:\\n post_msgs.append(msg)\\n\\n # Write post-call messages\\n if logging_fn:\\n for msg in post_msgs:\\n self._log_message(msg, extra_indent_level=0)\\n # v0.3.0b22 -- if recording history, add record of call even if we're muted(!)\\n elif _get_final_value('record_history'):\\n info = self._settings_mapping._get_DecoSetting('record_history')\\n _ = info.post_call_handler(context)\\n\\n self._logging_state_pop(enabled_too=True)\\n\\n return retval\\n\\n self._add_function_attrs(f, _deco_base_f_wrapper_)\\n return _deco_base_f_wrapper_\\n\\n #-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+\\n # end else (case \\\"f_or_klass is a function\\\",\\n # subcase \\\"f is a function & is NOT already deco'd\\\")\\n #+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*+*\",\n \"def f0(self):\\n return self._f0\",\n \"def getcfix(self):\\n cfix_ = ctypes.c_double()\\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\\n if res != 0:\\n _,msg = self.__getlasterror(res)\\n raise Error(rescode(res),msg)\\n cfix_ = cfix_.value\\n _cfix_return_value = cfix_\\n return (_cfix_return_value)\",\n \"def AceType(self):\\n raw_type = super(AceHeader, self).AceType\\n return ACE_CLASS_TYPE_MAPPER[raw_type]\",\n \"def f_get_class_name(self):\\n return self.__class__.__name__\",\n \"def getElementName(self):\\n return _libsbml.ListOfFunctionDefinitions_getElementName(self)\",\n \"def XPGetWidgetClassFunc(inWidgetClass):\\n pass\",\n \"def get_function(self,attr):\\n func_name = self._user_funcs[attr] \\n if hasattr(func_name,'__call__'):\\n func_name = func_name(self)\\n return func_name\",\n \"def cpf(self):\\n return self._cpf\",\n \"def get_score_function(self, idattr):\\n return self.get_node(\\n '//ScoreFunctions/ScoreFunction[@id=\\\"%s\\\"]' % idattr)\",\n \"def f1(self, option='average'):\\n\\t\\tif option == 'classwise':\\treturn self.class_f1_\\n\\t\\telif option == 'average':\\treturn self.avg_f1_\",\n \"def get_cfft(self):\\n return self.get_rfft().get_cfft()\",\n \"def getTypeCode(self):\\n return _libsbml.CubicBezier_getTypeCode(self)\",\n \"def functional(self):\\n return self.__functional\",\n \"def functional(self):\\n return self.vasprun.run_type\",\n \"def firedrake_fspace(self, shape=None):\\n if shape is None:\\n from firedrake import FunctionSpace\\n return FunctionSpace(self._mesh_geometry,\\n self._ufl_element.family(),\\n degree=self._ufl_element.degree())\\n elif isinstance(shape, int):\\n from firedrake import VectorFunctionSpace\\n return VectorFunctionSpace(self._mesh_geometry,\\n self._ufl_element.family(),\\n degree=self._ufl_element.degree(),\\n dim=shape)\\n elif isinstance(shape, tuple):\\n from firedrake import TensorFunctionSpace\\n return TensorFunctionSpace(self._mesh_geometry,\\n self._ufl_element.family(),\\n degree=self._ufl_element.degree(),\\n shape=shape)\\n else:\\n raise TypeError(\\\"'shape' must be *None*, an integer, \\\"\\n \\\" or a tuple of integers, not of type '%s'.\\\"\\n % type(shape))\",\n \"def class_impl(self):\\n # timing is stored by node, we compute timing by class on demand\\n rval = {}\\n for fgraph, node in self.apply_callcount:\\n typ = type(node.op)\\n if self.apply_cimpl[node]:\\n impl = \\\"C \\\"\\n else:\\n impl = \\\"Py\\\"\\n rval.setdefault(typ, impl)\\n if rval[typ] != impl and len(rval[typ]) == 2:\\n rval[typ] += impl\\n return rval\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.702561","0.63299495","0.5833264","0.5420016","0.53800005","0.5317835","0.52904075","0.5280903","0.52620566","0.5225637","0.5214052","0.51871824","0.5172217","0.5167699","0.51647925","0.5153158","0.51406175","0.5124004","0.51109064","0.5101468","0.5093111","0.5048223","0.5032329","0.5029493","0.50044256","0.49840248","0.4982087","0.49606174","0.49494192","0.49321398","0.49300972","0.49176717","0.49091223","0.48961496","0.48864338","0.48781994","0.4859858","0.4856915","0.4850793","0.48492286","0.48490876","0.48407704","0.48321202","0.48245955","0.4805533","0.48042473","0.48027784","0.47903845","0.47861186","0.47787434","0.47775277","0.4767753","0.47628084","0.47626263","0.47571895","0.4741274","0.47362646","0.47359946","0.47298896","0.47119358","0.4710855","0.46882534","0.46874365","0.46871117","0.4685755","0.46812713","0.46776116","0.4675179","0.46725023","0.4671931","0.46647313","0.4663469","0.46602884","0.46483538","0.46403402","0.4629117","0.46159667","0.46062765","0.46036562","0.4599054","0.4597936","0.45901474","0.45785713","0.45772302","0.45762625","0.4573747","0.4571952","0.45693034","0.45683852","0.4563999","0.4563582","0.45625696","0.4559377","0.45587578","0.45537677","0.45535895","0.4551976","0.4551615","0.45502988","0.45478785"],"string":"[\n \"0.702561\",\n \"0.63299495\",\n \"0.5833264\",\n \"0.5420016\",\n \"0.53800005\",\n \"0.5317835\",\n \"0.52904075\",\n \"0.5280903\",\n \"0.52620566\",\n \"0.5225637\",\n \"0.5214052\",\n \"0.51871824\",\n \"0.5172217\",\n \"0.5167699\",\n \"0.51647925\",\n \"0.5153158\",\n \"0.51406175\",\n \"0.5124004\",\n \"0.51109064\",\n \"0.5101468\",\n \"0.5093111\",\n \"0.5048223\",\n \"0.5032329\",\n \"0.5029493\",\n \"0.50044256\",\n \"0.49840248\",\n \"0.4982087\",\n \"0.49606174\",\n \"0.49494192\",\n \"0.49321398\",\n \"0.49300972\",\n \"0.49176717\",\n \"0.49091223\",\n \"0.48961496\",\n \"0.48864338\",\n \"0.48781994\",\n \"0.4859858\",\n \"0.4856915\",\n \"0.4850793\",\n \"0.48492286\",\n \"0.48490876\",\n \"0.48407704\",\n \"0.48321202\",\n \"0.48245955\",\n \"0.4805533\",\n \"0.48042473\",\n \"0.48027784\",\n \"0.47903845\",\n \"0.47861186\",\n \"0.47787434\",\n \"0.47775277\",\n \"0.4767753\",\n \"0.47628084\",\n \"0.47626263\",\n \"0.47571895\",\n \"0.4741274\",\n \"0.47362646\",\n \"0.47359946\",\n \"0.47298896\",\n \"0.47119358\",\n \"0.4710855\",\n \"0.46882534\",\n \"0.46874365\",\n \"0.46871117\",\n \"0.4685755\",\n \"0.46812713\",\n \"0.46776116\",\n \"0.4675179\",\n \"0.46725023\",\n \"0.4671931\",\n \"0.46647313\",\n \"0.4663469\",\n \"0.46602884\",\n \"0.46483538\",\n \"0.46403402\",\n \"0.4629117\",\n \"0.46159667\",\n \"0.46062765\",\n \"0.46036562\",\n \"0.4599054\",\n \"0.4597936\",\n \"0.45901474\",\n \"0.45785713\",\n \"0.45772302\",\n \"0.45762625\",\n \"0.4573747\",\n \"0.4571952\",\n \"0.45693034\",\n \"0.45683852\",\n \"0.4563999\",\n \"0.4563582\",\n \"0.45625696\",\n \"0.4559377\",\n \"0.45587578\",\n \"0.45537677\",\n \"0.45535895\",\n \"0.4551976\",\n \"0.4551615\",\n \"0.45502988\",\n \"0.45478785\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.8683575"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90891,"cells":{"query":{"kind":"string","value":"Spawn a new FriCAS commandline session."},"document":{"kind":"string","value":"def console(self):\n fricas_console()"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def cli():\n config, auth, execute_now = read_command_line_arguments()\n main(config, auth, execute_now)","def main():\n # Parse arguments.\n parser = argparse.ArgumentParser(description='Console for cluster test.')\n parser.add_argument('-c', dest='command_config_file',\n default='command_config',\n help='set command config file path (default=command_config)')\n args = parser.parse_args()\n\n console = Console()\n console.config(args.command_config_file)\n console.init_proxies()\n console.run()","def __init__(self, name='fricas', command='fricas -nox -noclef',\n script_subdirectory=None, logfile=None,\n server=None, server_tmpdir=None,\n init_code=[')lisp (si::readline-off)']):\n PanAxiom.__init__(self, name, command,\n script_subdirectory, logfile,\n server, server_tmpdir,\n init_code)","def Spawn(proc):\n proc.start()\n return proc","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():\r\n pass","def cli():\n ...","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n\n pass","def spawn(self):\n self._proc = subprocess.Popen(\n self._args, stdout=subprocess.PIPE, stderr=subprocess.PIPE\n )","def main_cli():\n pass","def cli(ctx: click.Context, config_instance: str, config_path: str) -> None:\n ctx.obj = TurbiniaCli(\n config_instance=config_instance, config_path=config_path)\n\n # Set up the tool based on the configuration file parameters.\n ctx.obj.setup()\n\n # Build all the commands based on responses from the API server.\n request_commands = factory.CommandFactory.create_dynamic_objects(\n evidence_mapping=ctx.obj.evidence_mapping,\n request_options=ctx.obj.request_options)\n for command in request_commands:\n groups.submit_group.add_command(command)","def main():\n\n run_manual_session()\n # run_automated_session()","def start(self):\n self.start_time = dt.datetime.now()\n self.call = ' '.join(sys.argv)\n self.commands = []","def launch_cli() -> None:\n app.run(main, flags_parser=_parse_flags)","def cli():\n pass","def cli():\n return","def _cli():\n pass","def cli():\n logger.debug('cli() called')","def _spawn_cli() -> GooeyParser:\n parser = _add_gui_arg(_cli())\n\n return vars(parser.parse_args())","def use_karaf():\n # type: () -> subprocess.Popen\n karaf_dir = os.environ.get(\"KARAF_DIR\")\n\n # Start Karaf\n start = time.time()\n install_karaf(karaf_dir)\n print(\"Karaf installed in\", round(time.time() - start, 3), \"s\")\n karaf_root = find_karaf_root(karaf_dir)\n\n start = time.time()\n with start_karaf(karaf_root) as karaf:\n # Wait for Karaf to start\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- Karaf started\")\n\n # Add the ECF repository\n karaf.stdin.write(b\"feature:repo-add ecf\\n\")\n karaf.stdin.flush()\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- ECF repository added\")\n\n # Install the tutorial sample\n karaf.stdin.write(\n b\"feature:install -v ecf-rs-examples-python.java-hello\\n\"\n )\n karaf.stdin.flush()\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- Feature installed\")\n\n # Give hand to the caller\n try:\n yield karaf\n finally:\n # Exit Karaf\n karaf.stdin.write(b\"logout\\n\")","def cli() -> None:","def cli() -> None:","def spawn(cls, *args, **kwargs):\n g = cls(*args, **kwargs)\n g.start()\n return g","def monkeypatch_spawn():\n ccompiler.CCompiler.spawn = _commandfile_spawn","def spawn():\n if platform.system() == \"Windows\":\n # HACK https://github.com/prompt-toolkit/python-prompt-toolkit/issues/1243#issuecomment-706668723\n # FIXME Use pexpect or wexpect somehow to fix this\n pytest.xfail(\n \"pexpect fails on Windows\",\n )\n # Using PopenSpawn, although probably it would be best to use pexpect.spawn\n # instead. However, it's working fine and it seems easier to fix in the\n # future to work on Windows (where, this way, spawning actually works; it's just\n # python-prompt-toolkit that rejects displaying a TUI)\n return PopenSpawn","def run():\n if len(sys.argv)<2:\n sys.argv.append(\"-h\")\n plac.call(main)","def new_session(self):\n self.command(\"new\")","def start( self ):\n pathCheck( self.command )\n cout = '/tmp/' + self.name + '.log'\n if self.cdir is not None:\n self.cmd( 'cd ' + self.cdir )\n self.cmd( self.command + ' ' + self.cargs % self.port +\n ' 1>' + cout + ' 2>' + cout + '&' )\n self.execed = False","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def spawn(mystring, mysettings, debug=0, free=0, droppriv=0, sesandbox=0, fakeroot=0, **keywords):\n\n\tcheck_config_instance(mysettings)\n\n\tfd_pipes = keywords.get(\"fd_pipes\")\n\tif fd_pipes is None:\n\t\tfd_pipes = {\n\t\t\t0:sys.__stdin__.fileno(),\n\t\t\t1:sys.__stdout__.fileno(),\n\t\t\t2:sys.__stderr__.fileno(),\n\t\t}\n\t# In some cases the above print statements don't flush stdout, so\n\t# it needs to be flushed before allowing a child process to use it\n\t# so that output always shows in the correct order.\n\tstdout_filenos = (sys.__stdout__.fileno(), sys.__stderr__.fileno())\n\tfor fd in fd_pipes.values():\n\t\tif fd in stdout_filenos:\n\t\t\tsys.__stdout__.flush()\n\t\t\tsys.__stderr__.flush()\n\t\t\tbreak\n\n\tfeatures = mysettings.features\n\t# TODO: Enable fakeroot to be used together with droppriv. The\n\t# fake ownership/permissions will have to be converted to real\n\t# permissions in the merge phase.\n\tfakeroot = fakeroot and uid != 0 and portage.process.fakeroot_capable\n\tif droppriv and uid == 0 and portage_gid and portage_uid and \\\n\t\thasattr(os, \"setgroups\"):\n\t\tkeywords.update({\"uid\":portage_uid,\"gid\":portage_gid,\n\t\t\t\"groups\":userpriv_groups,\"umask\":0o02})\n\tif not free:\n\t\tfree=((droppriv and \"usersandbox\" not in features) or \\\n\t\t\t(not droppriv and \"sandbox\" not in features and \\\n\t\t\t\"usersandbox\" not in features and not fakeroot))\n\n\tif not free and not (fakeroot or portage.process.sandbox_capable):\n\t\tfree = True\n\n\tif mysettings.mycpv is not None:\n\t\tkeywords[\"opt_name\"] = \"[%s]\" % mysettings.mycpv\n\telse:\n\t\tkeywords[\"opt_name\"] = \"[%s/%s]\" % \\\n\t\t\t(mysettings.get(\"CATEGORY\",\"\"), mysettings.get(\"PF\",\"\"))\n\n\tif free or \"SANDBOX_ACTIVE\" in os.environ:\n\t\tkeywords[\"opt_name\"] += \" bash\"\n\t\tspawn_func = portage.process.spawn_bash\n\telif fakeroot:\n\t\tkeywords[\"opt_name\"] += \" fakeroot\"\n\t\tkeywords[\"fakeroot_state\"] = os.path.join(mysettings[\"T\"], \"fakeroot.state\")\n\t\tspawn_func = portage.process.spawn_fakeroot\n\telse:\n\t\tkeywords[\"opt_name\"] += \" sandbox\"\n\t\tspawn_func = portage.process.spawn_sandbox\n\n\tif sesandbox:\n\t\tspawn_func = selinux.spawn_wrapper(spawn_func,\n\t\t\tmysettings[\"PORTAGE_SANDBOX_T\"])\n\n\tif keywords.get(\"returnpid\"):\n\t\treturn spawn_func(mystring, env=mysettings.environ(), **keywords)\n\n\tproc = EbuildSpawnProcess(\n\t\tbackground=False, args=mystring,\n\t\tscheduler=SchedulerInterface(EventLoop(main=False)),\n\t\tspawn_func=spawn_func,\n\t\tsettings=mysettings, **keywords)\n\n\tproc.start()\n\tproc.wait()\n\n\treturn proc.returncode","def main():\n\tcli = Cli()\n\tcli.run()","def __init__(self, command, args, prompt, parent=None, timeout=30,context=None ):\n if parent:\n self.spawnProc = parent.spawnProc\n self.isChild = True\n else:\n self.isChild = False\n\t \n self.context = context\n self.command = command\n\n self.args = args\n self.prompt = prompt\n self.sshTimeout = int ( globalVar.sshTimeout)\n if (timeout != 0) :\n self.sshTimeout = timeout\n \n try:\n self._connect() \n # sleeping 2 second to give control to other processes\n # This is done to resolve EOF problem\n time.sleep(2)\n\n # Checking if the spawned process lives. It is observed that the\n # spawned process for gen dies because of TIME_WAIT problem \n # If the process had died, sleep for 60 secs so that the socket\n # is closed or binding is released (BSD implementation) \n # Try spawning the process again, this time it should get thru.\n\n if not self.spawnProc.isalive():\n time.sleep(60)\n self._connect()\n \n except pexpect.EOF:\n if self.spawnProc.isalive():\n raise SessionManagerException(\"error starting %s\" % command)\n else:\n self._postConnect()\n except Exception, e:\n msg = \"session error: %s\" % e\n self.disconnect()\n raise e","def _launch(self):\n annotators = ['tokenize', 'ssplit']\n if 'ner' in self.annotators:\n annotators.extend(['pos', 'lemma', 'ner'])\n elif 'lemma' in self.annotators:\n annotators.extend(['pos', 'lemma'])\n elif 'pos' in self.annotators:\n annotators.extend(['pos'])\n annotators = ','.join(annotators)\n options = ','.join(['untokenizable=noneDelete',\n 'invertible=true'])\n # if you work on English, use this this command\n cmd = ['java', '-mx' + self.mem, '-cp', '\"%s\"' % self.classpath,\n 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-annotators',\n annotators, '-tokenize.options', options,\n '-outputFormat', 'json', '-prettyPrint', 'false']\n \n # if you work on arabic, use this this command\n \n # cmd = ['java', '-mx' + self.mem, '-cp', '\"%s\"' % self.classpath,\n # # 'edu.stanford.nlp.pipeline.StanfordCoreNLP','-annotators',\n # 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-props', 'StanfordCoreNLP-arabic.properties','-annotators',\n # annotators, '-tokenize.options', options, #'-tokenize.whitespace', 'true',\n # '-outputFormat', 'json', '-prettyPrint', 'false']\n print(' '.join(cmd))\n\n # We use pexpect to keep the subprocess alive and feed it commands.\n # Because we don't want to get hit by the max terminal buffer size,\n # we turn off canonical input processing to have unlimited bytes.\n self.corenlp = pexpect.spawn('/bin/bash', maxread=100000, timeout=60)\n self.corenlp.setecho(False)\n self.corenlp.sendline('stty -icanon')\n self.corenlp.sendline(' '.join(cmd))\n self.corenlp.delaybeforesend = 0\n self.corenlp.delayafterread = 0\n self.corenlp.expect_exact('NLP>', searchwindowsize=100)"],"string":"[\n \"def cli():\\n config, auth, execute_now = read_command_line_arguments()\\n main(config, auth, execute_now)\",\n \"def main():\\n # Parse arguments.\\n parser = argparse.ArgumentParser(description='Console for cluster test.')\\n parser.add_argument('-c', dest='command_config_file',\\n default='command_config',\\n help='set command config file path (default=command_config)')\\n args = parser.parse_args()\\n\\n console = Console()\\n console.config(args.command_config_file)\\n console.init_proxies()\\n console.run()\",\n \"def __init__(self, name='fricas', command='fricas -nox -noclef',\\n script_subdirectory=None, logfile=None,\\n server=None, server_tmpdir=None,\\n init_code=[')lisp (si::readline-off)']):\\n PanAxiom.__init__(self, name, command,\\n script_subdirectory, logfile,\\n server, server_tmpdir,\\n init_code)\",\n \"def Spawn(proc):\\n proc.start()\\n return proc\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\\r\\n pass\",\n \"def cli():\\n ...\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n\\n pass\",\n \"def spawn(self):\\n self._proc = subprocess.Popen(\\n self._args, stdout=subprocess.PIPE, stderr=subprocess.PIPE\\n )\",\n \"def main_cli():\\n pass\",\n \"def cli(ctx: click.Context, config_instance: str, config_path: str) -> None:\\n ctx.obj = TurbiniaCli(\\n config_instance=config_instance, config_path=config_path)\\n\\n # Set up the tool based on the configuration file parameters.\\n ctx.obj.setup()\\n\\n # Build all the commands based on responses from the API server.\\n request_commands = factory.CommandFactory.create_dynamic_objects(\\n evidence_mapping=ctx.obj.evidence_mapping,\\n request_options=ctx.obj.request_options)\\n for command in request_commands:\\n groups.submit_group.add_command(command)\",\n \"def main():\\n\\n run_manual_session()\\n # run_automated_session()\",\n \"def start(self):\\n self.start_time = dt.datetime.now()\\n self.call = ' '.join(sys.argv)\\n self.commands = []\",\n \"def launch_cli() -> None:\\n app.run(main, flags_parser=_parse_flags)\",\n \"def cli():\\n pass\",\n \"def cli():\\n return\",\n \"def _cli():\\n pass\",\n \"def cli():\\n logger.debug('cli() called')\",\n \"def _spawn_cli() -> GooeyParser:\\n parser = _add_gui_arg(_cli())\\n\\n return vars(parser.parse_args())\",\n \"def use_karaf():\\n # type: () -> subprocess.Popen\\n karaf_dir = os.environ.get(\\\"KARAF_DIR\\\")\\n\\n # Start Karaf\\n start = time.time()\\n install_karaf(karaf_dir)\\n print(\\\"Karaf installed in\\\", round(time.time() - start, 3), \\\"s\\\")\\n karaf_root = find_karaf_root(karaf_dir)\\n\\n start = time.time()\\n with start_karaf(karaf_root) as karaf:\\n # Wait for Karaf to start\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- Karaf started\\\")\\n\\n # Add the ECF repository\\n karaf.stdin.write(b\\\"feature:repo-add ecf\\\\n\\\")\\n karaf.stdin.flush()\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- ECF repository added\\\")\\n\\n # Install the tutorial sample\\n karaf.stdin.write(\\n b\\\"feature:install -v ecf-rs-examples-python.java-hello\\\\n\\\"\\n )\\n karaf.stdin.flush()\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- Feature installed\\\")\\n\\n # Give hand to the caller\\n try:\\n yield karaf\\n finally:\\n # Exit Karaf\\n karaf.stdin.write(b\\\"logout\\\\n\\\")\",\n \"def cli() -> None:\",\n \"def cli() -> None:\",\n \"def spawn(cls, *args, **kwargs):\\n g = cls(*args, **kwargs)\\n g.start()\\n return g\",\n \"def monkeypatch_spawn():\\n ccompiler.CCompiler.spawn = _commandfile_spawn\",\n \"def spawn():\\n if platform.system() == \\\"Windows\\\":\\n # HACK https://github.com/prompt-toolkit/python-prompt-toolkit/issues/1243#issuecomment-706668723\\n # FIXME Use pexpect or wexpect somehow to fix this\\n pytest.xfail(\\n \\\"pexpect fails on Windows\\\",\\n )\\n # Using PopenSpawn, although probably it would be best to use pexpect.spawn\\n # instead. However, it's working fine and it seems easier to fix in the\\n # future to work on Windows (where, this way, spawning actually works; it's just\\n # python-prompt-toolkit that rejects displaying a TUI)\\n return PopenSpawn\",\n \"def run():\\n if len(sys.argv)<2:\\n sys.argv.append(\\\"-h\\\")\\n plac.call(main)\",\n \"def new_session(self):\\n self.command(\\\"new\\\")\",\n \"def start( self ):\\n pathCheck( self.command )\\n cout = '/tmp/' + self.name + '.log'\\n if self.cdir is not None:\\n self.cmd( 'cd ' + self.cdir )\\n self.cmd( self.command + ' ' + self.cargs % self.port +\\n ' 1>' + cout + ' 2>' + cout + '&' )\\n self.execed = False\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def spawn(mystring, mysettings, debug=0, free=0, droppriv=0, sesandbox=0, fakeroot=0, **keywords):\\n\\n\\tcheck_config_instance(mysettings)\\n\\n\\tfd_pipes = keywords.get(\\\"fd_pipes\\\")\\n\\tif fd_pipes is None:\\n\\t\\tfd_pipes = {\\n\\t\\t\\t0:sys.__stdin__.fileno(),\\n\\t\\t\\t1:sys.__stdout__.fileno(),\\n\\t\\t\\t2:sys.__stderr__.fileno(),\\n\\t\\t}\\n\\t# In some cases the above print statements don't flush stdout, so\\n\\t# it needs to be flushed before allowing a child process to use it\\n\\t# so that output always shows in the correct order.\\n\\tstdout_filenos = (sys.__stdout__.fileno(), sys.__stderr__.fileno())\\n\\tfor fd in fd_pipes.values():\\n\\t\\tif fd in stdout_filenos:\\n\\t\\t\\tsys.__stdout__.flush()\\n\\t\\t\\tsys.__stderr__.flush()\\n\\t\\t\\tbreak\\n\\n\\tfeatures = mysettings.features\\n\\t# TODO: Enable fakeroot to be used together with droppriv. The\\n\\t# fake ownership/permissions will have to be converted to real\\n\\t# permissions in the merge phase.\\n\\tfakeroot = fakeroot and uid != 0 and portage.process.fakeroot_capable\\n\\tif droppriv and uid == 0 and portage_gid and portage_uid and \\\\\\n\\t\\thasattr(os, \\\"setgroups\\\"):\\n\\t\\tkeywords.update({\\\"uid\\\":portage_uid,\\\"gid\\\":portage_gid,\\n\\t\\t\\t\\\"groups\\\":userpriv_groups,\\\"umask\\\":0o02})\\n\\tif not free:\\n\\t\\tfree=((droppriv and \\\"usersandbox\\\" not in features) or \\\\\\n\\t\\t\\t(not droppriv and \\\"sandbox\\\" not in features and \\\\\\n\\t\\t\\t\\\"usersandbox\\\" not in features and not fakeroot))\\n\\n\\tif not free and not (fakeroot or portage.process.sandbox_capable):\\n\\t\\tfree = True\\n\\n\\tif mysettings.mycpv is not None:\\n\\t\\tkeywords[\\\"opt_name\\\"] = \\\"[%s]\\\" % mysettings.mycpv\\n\\telse:\\n\\t\\tkeywords[\\\"opt_name\\\"] = \\\"[%s/%s]\\\" % \\\\\\n\\t\\t\\t(mysettings.get(\\\"CATEGORY\\\",\\\"\\\"), mysettings.get(\\\"PF\\\",\\\"\\\"))\\n\\n\\tif free or \\\"SANDBOX_ACTIVE\\\" in os.environ:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" bash\\\"\\n\\t\\tspawn_func = portage.process.spawn_bash\\n\\telif fakeroot:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" fakeroot\\\"\\n\\t\\tkeywords[\\\"fakeroot_state\\\"] = os.path.join(mysettings[\\\"T\\\"], \\\"fakeroot.state\\\")\\n\\t\\tspawn_func = portage.process.spawn_fakeroot\\n\\telse:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" sandbox\\\"\\n\\t\\tspawn_func = portage.process.spawn_sandbox\\n\\n\\tif sesandbox:\\n\\t\\tspawn_func = selinux.spawn_wrapper(spawn_func,\\n\\t\\t\\tmysettings[\\\"PORTAGE_SANDBOX_T\\\"])\\n\\n\\tif keywords.get(\\\"returnpid\\\"):\\n\\t\\treturn spawn_func(mystring, env=mysettings.environ(), **keywords)\\n\\n\\tproc = EbuildSpawnProcess(\\n\\t\\tbackground=False, args=mystring,\\n\\t\\tscheduler=SchedulerInterface(EventLoop(main=False)),\\n\\t\\tspawn_func=spawn_func,\\n\\t\\tsettings=mysettings, **keywords)\\n\\n\\tproc.start()\\n\\tproc.wait()\\n\\n\\treturn proc.returncode\",\n \"def main():\\n\\tcli = Cli()\\n\\tcli.run()\",\n \"def __init__(self, command, args, prompt, parent=None, timeout=30,context=None ):\\n if parent:\\n self.spawnProc = parent.spawnProc\\n self.isChild = True\\n else:\\n self.isChild = False\\n\\t \\n self.context = context\\n self.command = command\\n\\n self.args = args\\n self.prompt = prompt\\n self.sshTimeout = int ( globalVar.sshTimeout)\\n if (timeout != 0) :\\n self.sshTimeout = timeout\\n \\n try:\\n self._connect() \\n # sleeping 2 second to give control to other processes\\n # This is done to resolve EOF problem\\n time.sleep(2)\\n\\n # Checking if the spawned process lives. It is observed that the\\n # spawned process for gen dies because of TIME_WAIT problem \\n # If the process had died, sleep for 60 secs so that the socket\\n # is closed or binding is released (BSD implementation) \\n # Try spawning the process again, this time it should get thru.\\n\\n if not self.spawnProc.isalive():\\n time.sleep(60)\\n self._connect()\\n \\n except pexpect.EOF:\\n if self.spawnProc.isalive():\\n raise SessionManagerException(\\\"error starting %s\\\" % command)\\n else:\\n self._postConnect()\\n except Exception, e:\\n msg = \\\"session error: %s\\\" % e\\n self.disconnect()\\n raise e\",\n \"def _launch(self):\\n annotators = ['tokenize', 'ssplit']\\n if 'ner' in self.annotators:\\n annotators.extend(['pos', 'lemma', 'ner'])\\n elif 'lemma' in self.annotators:\\n annotators.extend(['pos', 'lemma'])\\n elif 'pos' in self.annotators:\\n annotators.extend(['pos'])\\n annotators = ','.join(annotators)\\n options = ','.join(['untokenizable=noneDelete',\\n 'invertible=true'])\\n # if you work on English, use this this command\\n cmd = ['java', '-mx' + self.mem, '-cp', '\\\"%s\\\"' % self.classpath,\\n 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-annotators',\\n annotators, '-tokenize.options', options,\\n '-outputFormat', 'json', '-prettyPrint', 'false']\\n \\n # if you work on arabic, use this this command\\n \\n # cmd = ['java', '-mx' + self.mem, '-cp', '\\\"%s\\\"' % self.classpath,\\n # # 'edu.stanford.nlp.pipeline.StanfordCoreNLP','-annotators',\\n # 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-props', 'StanfordCoreNLP-arabic.properties','-annotators',\\n # annotators, '-tokenize.options', options, #'-tokenize.whitespace', 'true',\\n # '-outputFormat', 'json', '-prettyPrint', 'false']\\n print(' '.join(cmd))\\n\\n # We use pexpect to keep the subprocess alive and feed it commands.\\n # Because we don't want to get hit by the max terminal buffer size,\\n # we turn off canonical input processing to have unlimited bytes.\\n self.corenlp = pexpect.spawn('/bin/bash', maxread=100000, timeout=60)\\n self.corenlp.setecho(False)\\n self.corenlp.sendline('stty -icanon')\\n self.corenlp.sendline(' '.join(cmd))\\n self.corenlp.delaybeforesend = 0\\n self.corenlp.delayafterread = 0\\n self.corenlp.expect_exact('NLP>', searchwindowsize=100)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.60068893","0.5865383","0.5834904","0.58237004","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5797985","0.5786446","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.57515025","0.5723277","0.5705709","0.57052696","0.57039106","0.56917137","0.56841964","0.56743085","0.56502104","0.56183946","0.55473167","0.5537812","0.553661","0.55204964","0.55204964","0.54925025","0.5487892","0.5433574","0.54125625","0.54105437","0.5393246","0.53773993","0.53773993","0.53773993","0.53773993","0.53773993","0.53486776","0.5340044","0.5328309","0.5316158"],"string":"[\n \"0.60068893\",\n \"0.5865383\",\n \"0.5834904\",\n \"0.58237004\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5797985\",\n \"0.5786446\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.57515025\",\n \"0.5723277\",\n \"0.5705709\",\n \"0.57052696\",\n \"0.57039106\",\n \"0.56917137\",\n \"0.56841964\",\n \"0.56743085\",\n \"0.56502104\",\n \"0.56183946\",\n \"0.55473167\",\n \"0.5537812\",\n \"0.553661\",\n \"0.55204964\",\n \"0.55204964\",\n \"0.54925025\",\n \"0.5487892\",\n \"0.5433574\",\n \"0.54125625\",\n \"0.54105437\",\n \"0.5393246\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53486776\",\n \"0.5340044\",\n \"0.5328309\",\n \"0.5316158\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90892,"cells":{"query":{"kind":"string","value":"A helper function for converting FriCAS domains to the corresponding Sage object."},"document":{"kind":"string","value":"def _sage_domain(self):\n P = self._check_valid()\n name = str(self)\n if name == 'Integer':\n from sage.rings.all import ZZ\n return ZZ\n elif name == 'DoubleFloat':\n from sage.rings.all import RDF\n return RDF\n elif name.startswith('Fraction('):\n name = name.lstrip('Fraction(')\n name = name.rstrip(')')\n return P(name)._sage_domain().fraction_field()\n\n raise NotImplementedError"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def domain(self, domain):","def convert_to(self, domain):\n if domain == self.domain:\n return self.copy()\n elif domain == QQ and self.domain == ZZ:\n return self._new(flint.fmpq_mat(self.rep), self.shape, domain)\n elif domain == ZZ and self.domain == QQ:\n # XXX: python-flint has no fmpz_mat.from_fmpq_mat\n return self.to_ddm().convert_to(domain).to_dfm()\n else:\n # It is the callers responsibility to convert to DDM before calling\n # this method if the domain is not supported by DFM.\n raise NotImplementedError(\"Only ZZ and QQ are supported by DFM\")","def format_domain(domain):\n domain.ns_converted = []\n for ns in domain.ns :\n if isinstance(ns, objects.DomainHostAttr) :\n ns_item = {\n 'hostname' : ns.hostname,\n 'ips' : []\n }\n\n for hostaddr in ns.hostAddr :\n ns_item['ips'].append(hostaddr.ip)\n else :\n ns_item = {\n 'hostname' : ns.name,\n 'ips' : [],\n 'hostobj' : 1\n }\n domain.ns_converted.append(ns_item)\n\n return domain","def transform_domains(points, from_domain, to_domain, use_mp=False, dps=None):\n with mp.workdps(dps or mp.dps):\n fl = mp.mpf if use_mp else float\n a, b = map(fl, to_domain)\n c, d = map(fl, from_domain)\n scale = (b-a)/(d-c)\n trans = a - (b-a)*c/(d-c)\n return [min(b, max(a, scale*p + trans)) for p in points]","def _infer_domain(self, name, domain, elements):\n if '*' not in domain:\n return domain\n debug('guessing a better domain for {}: {}'.format(name, domain))\n\n # Domain as a list of references to Variables in the File/xr.Dataset\n domain_ = [self[d] for d in domain]\n\n for i, d in enumerate(domain_): # Iterate over dimensions\n e = set(elements[i])\n if d.name != '*' or len(e) == 0: # pragma: no cover\n assert set(d.values).issuperset(e)\n continue # The stated domain matches the data; or no data\n # '*' is given\n if (self._state[name]['attrs']['type_code'] == gdxcc.GMS_DT_PAR and\n self._implicit):\n d = '_{}_{}'.format(name, i)\n debug(('Constructing implicit set {} for dimension {} of {}\\n'\n ' {} instead of {} elements')\n .format(d, name, i, len(e), len(self['*'])))\n self.coords[d] = elements[i]\n d = self[d]\n else:\n # try to find a smaller domain for this dimension\n # Iterate over every Set/Coordinate\n for s in self.coords.values():\n if s.ndim == 1 and set(s.values).issuperset(e) and \\\n len(s) < len(d):\n d = s # Found a smaller Set; use this instead\n domain_[i] = d\n\n # Convert the references to names\n inferred = [d.name for d in domain_]\n\n if domain != inferred:\n # Store the result\n self._state[name]['attrs']['domain_inferred'] = inferred\n debug('…inferred {}.'.format(inferred))\n else:\n debug('…failed.')\n\n return inferred","def set_subdomains(self, f):\n s = \"::: setting 2D subdomains :::\"\n print_text(s, cls=self)\n\n self.ff = MeshFunction('size_t', self.mesh)\n self.cf = MeshFunction('size_t', self.mesh)\n self.ff_acc = MeshFunction('size_t', self.mesh)\n f.read(self.ff, 'ff')\n f.read(self.cf, 'cf')\n f.read(self.ff_acc, 'ff_acc')\n \n self.ds = Measure('ds')[self.ff]\n self.dx = Measure('dx')[self.cf]\n \n self.dx_g = self.dx(0) # internal above grounded\n self.dx_f = self.dx(1) # internal above floating\n self.dBed_g = self.ds(3) # grounded bed\n self.dBed_f = self.ds(5) # floating bed\n self.dBed = self.ds(3) + self.ds(5) # bed\n self.dSrf_gu = self.ds(8) # grounded with U observations\n self.dSrf_fu = self.ds(9) # floating with U observations\n self.dSrf_u = self.ds(8) + self.ds(9) # surface with U observations\n self.dSrf_g = self.ds(2) + self.ds(8) # surface of grounded ice\n self.dSrf_f = self.ds(6) + self.ds(9) # surface of floating ice\n self.dSrf = self.ds(6) + self.ds(2) \\\n + self.ds(8) + self.ds(9) # surface\n self.dLat_d = self.ds(7) # lateral divide\n self.dLat_to = self.ds(4) # lateral terminus overwater\n self.dLat_tu = self.ds(10) # lateral terminus underwater\n self.dLat_t = self.ds(4) + self.ds(10) # lateral terminus\n self.dLat = self.ds(4) + self.ds(7) \\\n + self.ds(10) # lateral","def shp2idomain(shp_path,g,idomain,features_type=\"polygon\",layer=0):\n \n if len(idomain.shape) != 1:\n idomain = idomain.reshape(g.nlay*g.nrow*g.ncol) # reshape idomain if not in the right format\n res = g.intersect(shp_path,features_type,layer)\n idomain[res.nodenumber] = 1\n idomain = idomain.reshape(g.nlay,g.nrow,g.ncol)","def fastlydomain(args):\n pprint(api.domain(service_id, args[0], args[1]).attrs)","def _parse_domain(self, f_domain):\n\n parse_tree = PDDL_Tree.create(f_domain)\n\n assert \"domain\" in parse_tree, \"Domain must have a name\"\n self.domain_name = parse_tree [\"domain\"].named_children ()[0]\n\n # must read types before constants\n if \":types\" in parse_tree:\n if \"-\" in parse_tree[\":types\"].named_children():\n type_hierarchy = PDDL_Utils.read_type(parse_tree[\":types\"])\n self.parent_types = {subtype: parent for subtype, parent in type_hierarchy}\n self.types = set(parse_tree[\":types\"].named_children())\n self.types.discard(\"-\")\n else:\n self.types = set(parse_tree[\":types\"].named_children())\n self.parent_types = {t: None for t in self.types}\n else:\n self.types = set([Predicate.OBJECT])\n self.parent_types = {Predicate.OBJECT: None}\n\n self.agents = [a.name for a in parse_tree[\":agents\"].children]\n self.types.add('agent')\n self.parent_types['agent'] = None\n self._add_objects([(ag, 'agent') for ag in self.agents])\n\n # must read in constants before actions or predicates\n if \":constants\" in parse_tree:\n object_list = PDDL_Utils.read_type(parse_tree[\":constants\"])\n self._add_objects(object_list)\n\n #TODO this may not be correct, depending on the type hierarchy\n const_map = dict()\n for const in self.objects:\n if len(self.obj_to_type[const]) == 0:\n raise RuntimeError(\"No type for constant object %s\" % const)\n else:\n const_map[const] = list(self.obj_to_type[const])[0]\n\n self.predicates = [self.to_predicate(c, map=const_map) for c in parse_tree[\":predicates\"].children]\n\n # some predicates have this property: they are untyped.\n for predicate in self.predicates:\n if Predicate.OBJECT not in self.types and any([arg[1] == Predicate.OBJECT for arg in predicate.args]):\n for t in self.types:\n if self.parent_types[t] is None:\n self.parent_types[t] = Predicate.OBJECT\n\n self.parent_types[Predicate.OBJECT] = None\n self.types.add(Predicate.OBJECT)\n self.type_to_obj[Predicate.OBJECT] = set([])\n for obj, type_list in self.obj_to_type.items():\n type_list.add(Predicate.OBJECT)\n self.type_to_obj[Predicate.OBJECT].add(obj)\n\n # only need to do this once, obviously\n break\n\n self.actions = [self.to_action(c) for c in parse_tree.find_all(\":action\")]","def domain_range(domain, _range=[0, 1], return_transform=False):\n\n if not return_transform:\n return interp1d([min(domain), max(domain)], [min(_range), max(_range)], bounds_error=False)\n else:\n m = interp1d([min(domain), max(domain)], [min(_range), max(_range)])\n return [float(m(v)) for v in domain] # Take float, else returns weird numpy.ndarray element","def _deserialize(self, data):\n uri = data[1:-1]\n # We have to retrieve the type to rebuild the object\n attr = self.__dict__['field']\n # Be careful when orig = None !!!!!\n orig = getattr(attr.model, attr.name)\n if None == orig:\n return rdfSubject(rdflib.term.URIRef(uri))\n elif isinstance(orig, list):\n # rdfalchemy mapper gives me the solution\n rt = attr.model.__class__.__dict__[attr.name].range_type\n from rdfalchemy.orm import mapper\n alch_map = mapper()\n try:\n cls = alch_map[str(rt)]\n return cls(rdflib.term.URIRef(uri))\n except:\n rdfSubject(rdflib.term.URIRef(uri))\n else:\n return type(orig)(rdflib.term.URIRef(uri))","def domain_to_geo(domain):\n\n return gi.record_by_name(domain)","def par_domain(self):","def fixDomains(self, domainMin, domainMax, fixToDomain):\n\n return 0","def getDSDomainUnion(gdb_dss):\n li_domain = []\n for g in gdb_dss:\n arcpy.env.workspace = os.path.join(inws, g)\n dss = arcpy.ListDatasets()\n for ds in dss:\n desc = arcpy.Describe(ds)\n sr = desc.spatialReference\n # Notice that sr.domain is a string of (xmin, ymin, xmax, ymax) separated by space.\n li_domain.append([float(x) for x in sr.domain.split(\" \")])\n\n df = pd.DataFrame(li_domain)\n # Return min of xmin, ymin and max of xmax, ymax\n return (df.iloc[:, 0].min(), df.iloc[:, 1].min(), df.iloc[:, 2].max(), df.iloc[:, 3].max())","def ground_to_field(f):\n return f.set_domain(f.dom.get_field())","def as_domain(labeled_docs, labels, issource, domain, unlabeled_docs=None, unlabeled_y=None, tokken_pattern=r\"(?u)\\b\\w\\w+\\b\", min_df=1):\n if issource:\n counter = CountVectorizer(token_pattern=tokken_pattern, min_df=min_df)\n v = counter.fit(labeled_docs).vocabulary_\n tfidf = TfidfVectorizer(sublinear_tf=True, token_pattern=tokken_pattern, vocabulary=v)\n else:\n tfidf = TfidfVectorizer(sublinear_tf=True, token_pattern=tokken_pattern, min_df=min_df)\n if unlabeled_docs is not None:\n X = tfidf.fit_transform(labeled_docs)\n U = tfidf.transform(unlabeled_docs)\n y = np.array(labels)\n V = tfidf.vocabulary_\n U_y = np.array(unlabeled_y)\n domain = Domain(X, y, V, domain, U, U_y)\n return domain\n X = tfidf.fit_transform(labeled_docs)\n y = np.array(labels)\n V = tfidf.vocabulary_\n domain = Domain(X, y, V, domain)\n return domain","def set_domain(f, dom):\n if f.dom == dom:\n return f\n else:\n return f.per(dmp_set_domain(f.rep, f.lev, f.dom, dom), dom=dom)","def order_domain_values(var,assignment,csp):\n #right now it works only as just convert value and return\n #no special black magic yet\n return var.domain","def dictTransform(domain):\n url = urlStringFormat('tomorrow', domain)\n data = requests.get(url)\n dataDict = xmlToJson(data)\n # Check if TimeSeries is dict or not. If it isn't, then it is an array of two TimeSeries\n if type(dataDict['Publication_MarketDocument']['TimeSeries']) is dict:\n url = urlStringFormat('today', domain)\n data = requests.get(url)\n dataDict = xmlToJson(data)\n\n # Transform dict into new dict without first hour and with 25. hour, \n # which is appended from the second TimeSeries \n priceFromSecondTimeSeries = dataDict['Publication_MarketDocument']['TimeSeries'][1]['Period']['Point']['price.amount']['$']\n del dataDict['Publication_MarketDocument']['TimeSeries'][1]\n # Transform array of one dictionary into dictionary\n dataDict['Publication_MarketDocument']['TimeSeries'] = dataDict['Publication_MarketDocument']['TimeSeries'][0]\n dataDict['Publication_MarketDocument']['TimeSeries']['Period']['Point'].append({'position': {'$': 25}, 'price.amount': {'$': priceFromSecondTimeSeries}})\n del dataDict['Publication_MarketDocument']['TimeSeries']['Period']['Point'][0]\n return dataDict","def _translate_freq_domain(self, freq_domain):\n\n lin_spaces = [np.linspace(-0.5, 0.5, x) for x in freq_domain.shape]\n meshgrids = np.meshgrid(*lin_spaces, indexing='ij')\n grid_coords = np.array([mg.flatten() for mg in meshgrids])\n\n phase_shift = np.multiply(grid_coords, self.translations).sum(axis=0) # phase shift is added\n exp_phase_shift = np.exp(-2j * math.pi * phase_shift)\n freq_domain_translated = np.multiply(exp_phase_shift, freq_domain.flatten(order='C')).reshape(freq_domain.shape)\n\n return freq_domain_translated","def stardomain(var, curr_domains):\n\t\tif curr_domains:\n\t\t\tstardomain = sorted(curr_domains[state], key = lambda restaurant: (star_weight*math.log10(restaurant['stars']) + reviews_weight*math.log10(restaurant['review_count']))/(star_weight + reviews_weight))\n\t\t\tstardomain.reverse()\n\t\telse:\n\t\t\tstardomain = sorted(state_domains[state], key = lambda restaurant: (star_weight*math.log10(restaurant['stars']) + reviews_weight*math.log10(restaurant['review_count']))/(star_weight + reviews_weight))\n\t\t\tstardomain.reverse()\n\t\treturn stardomain","def _get_domain(self):\n self.ensure_one()\n domain = []\n return domain","def handle_domains(\n actapi: act.api.Act, content: Text, domains: List[Text]\n) -> List[act.api.fact.Fact]:\n\n feeds_facts: List[act.api.fact.Fact] = []\n\n for domain in domains:\n\n chain = []\n\n chain.append(\n actapi.fact(\"connectsTo\").source(\"content\", content).destination(\"uri\", \"*\")\n )\n chain.append(\n actapi.fact(\"componentOf\").source(\"fqdn\", domain).destination(\"uri\", \"*\")\n )\n\n feeds_facts += act.api.fact.fact_chain(*chain)\n\n return feeds_facts","def _normalize_feature(feature: schema_pb2.Feature,\n schema: schema_pb2.Schema) -> None:\n\n if feature.HasField(\"struct_domain\"):\n for x in feature.struct_domain.feature:\n _normalize_feature(x, schema)\n if feature.HasField(\"domain\"):\n for string_domain in schema.string_domain:\n if string_domain.name == feature.domain:\n feature.string_domain.CopyFrom(string_domain)\n return\n for int_domain in schema.int_domain:\n if int_domain.name == feature.domain:\n feature.int_domain.CopyFrom(int_domain)\n return\n for float_domain in schema.float_domain:\n if float_domain.name == feature.domain:\n feature.float_domain.CopyFrom(float_domain)\n return\n raise ValueError(\"Did not find domain {} in schema {}\".format(\n feature.domain, schema))","def _compue_domain(self, new_position):\n if abs(new_position) <= self._safe_zone: \n return self._domain\n return self.Domain( sign(new_position) )","def domain_decomposition(self, domain_decomposition):\n\n self._domain_decomposition = domain_decomposition","def domain_resolve(self, domain):\n a_result = server.resolve(domain, 'A')\n mx_result = server.resolve(domain, 'MX')\n ns_result = server.resolve(domain, 'NS')\n\n domain_data = {}\n domain_data.setdefault('resolving', {'A': a_result})\n domain_data['resolving']['MX'] = mx_result\n domain_data['resolving']['NS'] = ns_result\n\n return domain_data","def _make_tfx_schema_domain_info_inline(schema: schema_pb2.Schema) -> None:\n # Reference to domains defined at schema level\n domain_ref_to_string_domain = {d.name: d for d in schema.string_domain}\n domain_ref_to_float_domain = {d.name: d for d in schema.float_domain}\n domain_ref_to_int_domain = {d.name: d for d in schema.int_domain}\n\n # With the reference, it is safe to remove the domains defined at schema level\n del schema.string_domain[:]\n del schema.float_domain[:]\n del schema.int_domain[:]\n\n for feature in schema.feature:\n domain_info_case = feature.WhichOneof(\"domain_info\")\n if domain_info_case == \"domain\":\n domain_ref = feature.domain\n if domain_ref in domain_ref_to_string_domain:\n feature.string_domain.MergeFrom(domain_ref_to_string_domain[domain_ref])\n elif domain_ref in domain_ref_to_float_domain:\n feature.float_domain.MergeFrom(domain_ref_to_float_domain[domain_ref])\n elif domain_ref in domain_ref_to_int_domain:\n feature.int_domain.MergeFrom(domain_ref_to_int_domain[domain_ref])","def freqdomain(self):\n \n\n #self.df = self.f[1] - self.f[0]\n #frequency vector\n #fv = fftshift(fftfreq(len(eta),1./fs))\n #fv = fv[len(fv)/2:]\n \n #spectral analysis\n self.sn1 = self.espec1(self.n1)\n self.sn2 = self.espec1(self.n2)\n self.sn3 = self.espec1(self.n3)\n self.sn12 = self.espec2(self.n1,self.n2)\n self.sn13 = self.espec2(self.n1,self.n3)\n self.sn23 = self.espec2(self.n2,self.n3)\n \n #delta freq\n self.df = self.f[3] - self.f[2]\n\n #calculo do numero de onda\n #self.wavenumber()\n #k = numeronda(h,f,len(f))\n #k = np.array(k)\n\n #calculo dos coeficientes de fourier - NDBC 96_01 e Steele (1992)\n c = self.sn2[:,1] + self.sn3[:,1]\n cc = np.sqrt(self.sn1[:,1] * (c))\n \n self.a1 = self.sn12[:,3] / cc\n self.b1 = self.sn13[:,3] / cc\n \n self.a2 = (self.sn2[:,1] - self.sn3[:,1]) / c\n self.b2 = 2 * self.sn12[:,2] / c\n \n #calcula direcao de onda\n #mean direction\n self.dire1 = np.array([np.angle(np.complex(self.b1[i],self.a1[i]),deg=True) for i in range(len(self.a1))])\n \n #principal direction\n self.dire2 = 0.5 * np.array([np.angle(np.complex(self.b2[i],self.a2[i]),deg=True) for i in range(len(self.a2))])\n \n #condicao para valores maiores que 360 e menores que 0\n self.dire1[np.where(self.dire1 < 0)] = self.dire1[np.where(self.dire1 < 0)] + 360\n self.dire1[np.where(self.dire1 > 360)] = self.dire1[np.where(self.dire1 > 360)] - 360\n self.dire2[np.where(self.dire2 < 0)] = self.dire2[np.where(self.dire2 < 0)] + 360\n self.dire2[np.where(self.dire2 > 360)] = self.dire2[np.where(self.dire2 > 360)] - 360\n \n #acha o indice da frequencia de pico\n ind = np.where(self.sn1[:,1] == np.max(self.sn1[:,1]))[0]\n \n #periodo de pico\n self.tp = (1. / self.f[ind])[0]\n \n #momento espectral de ordem zero total - m0\n self.m0 = np.sum(self.sn1[:,1]) * self.df\n \n #calculo da altura significativa\n self.hm0 = 4.01 * np.sqrt(self.m0)\n \n #direcao do periodo de pico\n self.dp = self.dire1[ind][0]\n \n #Espalhamento direcional\n #Formula do sigma1 do livro Tucker&Pitt(2001) \"Waves in Ocean Engineering\" pags 196-198\n c1 = np.sqrt(self.a1 ** 2 + self.b1 ** 2)\n c2 = np.sqrt(self.a2 ** 2 + self.b2 ** 2)\n \n s1 = c1 / (1-c1)\n s2 = (1 + 3 * c2 + np.sqrt(1 + 14 * c2 + c2 ** 2)) / (2 * (1 - c2))\n \n self.sigma1 = np.sqrt(2 - 2 * c1) * 180 / np.pi\n self.sigma2 = np.sqrt((1 - c2) / 2) * 180 / np.pi\n \n self.sigma1p = np.real(self.sigma1[ind])[0]\n self.sigma2p = np.real(self.sigma2[ind])[0]\n \n # pondaf = np.array([hm0, tp, dp, sigma1p, sigma2p])\n \n #hm0, tp, dp, sigma1, sigma2, sigma1p, sigma2p, f, df, k, sn, snx, sny, snn, snnx, snny, snxny, snxnx, snyny, a1, b1, a2, b2, dire1, dire2\n #return hm0, tp, dp, sigma1, sigma2, sigma1p, sigma2p, f, df, k, sn, snx, sny, snn, snnx, snny, snxny, snxnx, snyny, a1, b1, a2, b2, dire1, dire2","def detect_domains (nffg):\n return {infra.domain for infra in nffg.infras}","def _discretize_freestream_domain(self, xmin, xmax, ymin, ymax, zmin, zmax, resolution):\n x = np.linspace(xmin, xmax, int(resolution.x1))\n y = np.linspace(ymin, ymax, int(resolution.x2))\n z = np.linspace(zmin, zmax, int(resolution.x3))\n return np.meshgrid(x, y, z, indexing=\"ij\")","def domain_features(domain, path_set):\n return string_features_v2(domain_string(domain, path_set))","def convert_to_model(self, *args):\n city_xref_service_data, *_ = args\n return [CityXRefService(**city_xref_service) for city_xref_service in city_xref_service_data]","def _convert(self):\n logger.info(\"Converting conformers to density\")\n logger.debug(\"Masking\")\n self._transformer.reset(full=True)\n for n, coor in enumerate(self._coor_set):\n self.conformer.coor = coor\n self._transformer.mask(self._rmask)\n mask = self._transformer.xmap.array > 0\n self._transformer.reset(full=True)\n\n nvalues = mask.sum()\n self._target = self.xmap.array[mask]\n logger.debug(\"Density\")\n nmodels = len(self._coor_set)\n self._models = np.zeros((nmodels, nvalues), float)\n for n, coor in enumerate(self._coor_set):\n self.conformer.coor = coor\n self.conformer.b = self._bs[n]\n self._transformer.density()\n model = self._models[n]\n model[:] = self._transformer.xmap.array[mask]\n np.maximum(model, self.options.bulk_solvent_level, out=model)\n self._transformer.reset(full=True)","def convert(self, values, axes_list, unit, is_norm):\n \n # Convert into right unit\n if unit == self.unit or unit == \"SI\":\n if is_norm:\n try:\n values = values / self.normalizations.get(\"ref\")\n except:\n raise NormError(\n \"ERROR: Reference value not specified for normalization\"\n )\n elif unit in self.normalizations:\n values = values / self.normalizations.get(unit)\n else:\n values = convert(values, self.unit, unit)\n return values","def domain(s):\r\n res = r_domain.findall(s)\r\n domain = (res and res[0]) or s\r\n return domain.lower()","def _unify_domain(cls, *matrices):\n domains = {matrix.domain for matrix in matrices}\n if len(domains) == 1:\n return matrices\n domain = reduce(lambda x, y: x.unify(y), domains)\n return tuple(matrix.convert_to(domain) for matrix in matrices)","def parse_domain(self, domainfile):\n\n with open(domainfile) as dfile:\n dfile_array = self._get_file_as_array(dfile)\n #Deal with front/end define, problem, :domain\n if dfile_array[0:4] != ['(', 'define', '(', 'domain']:\n print('PARSING ERROR: Expected (define (domain ... at start of domain file')\n sys.exit()\n self.domain = dfile_array[4]\n\n dfile_array = dfile_array[6:-1]\n opencounter = 0\n keyword = ''\n obj_list = []\n is_obj_list = True\n for word in dfile_array:\n if word == '(':\n opencounter += 1\n elif word == ')':\n opencounter -= 1\n elif word.startswith(':'):\n if word[1:] not in DFILE_KEYWORDS:\n pass\n elif keyword != 'requirements':\n keyword = word[1:]\n if opencounter == 0:\n if keyword == 'action':\n self.actions.append(obj_list)\n obj_list = []\n if keyword == 'types':\n for element in obj_list:\n self.types.setdefault('object', []).append(element)\n self.type_list.add('object')\n self.type_list.add(element)\n obj_list = []\n keyword = ''\n\n if keyword == 'requirements': #Requirements list\n if word != ':requirements':\n if not word.startswith(':'):\n print('PARSING ERROR: Expected requirement to start with :')\n sys.exit()\n elif word[1:] not in DFILE_REQ_KEYWORDS:\n print('WARNING: Unknown Rquierement ' + word[1:])\n #print 'Requirements must only be: ' + str(DFILE_REQ_KEYWORDS)\n #sys.exit()\n else:\n self.requirements.add(word[1:])\n elif keyword == 'action':\n obj_list.append(word)\n elif not word.startswith(':'):\n if keyword == 'types': #Typed list of objects\n if is_obj_list:\n if word == '-':\n is_obj_list = False\n else:\n obj_list.append(word)\n else:\n #word is type\n for element in obj_list:\n if not word in self.type_list:\n self.types.setdefault('object', []).append(word)\n self.type_list.add(word)\n self.types.setdefault(word, []).append(element)\n self.type_list.add(element)\n self.type_list.add(word)\n is_obj_list = True\n obj_list = []\n elif keyword == 'constants': #Typed list of objects\n if is_obj_list:\n if word == '-':\n is_obj_list = False\n else:\n obj_list.append(word)\n else:\n #word is type\n for element in obj_list:\n if word in self.type_list:\n self.constants.setdefault(word, []).append(element)\n #self.object_list.add(element)\n else:\n print(self.type_list)\n print(\"ERROR unknown type \" + word)\n sys.exit()\n is_obj_list = True\n obj_list = []\n elif keyword == 'predicates' or keyword == 'private': #Internally typed predicates\n if word == ')':\n if keyword == 'private':\n #print \"...skip agent: \" + str(obj_list[:3])\n obj_list = obj_list[3:]\n keyword = 'predicates'\n if len(obj_list) == 0:\n #print \"...skip )\"\n continue\n p_name = obj_list[0]\n #print \"parse predicate: \" + p_name + \" \" + str(obj_list)\n pred_list = self._parse_name_type_pairs(obj_list[1:],self.type_list)\n self.predicates.append(Predicate(p_name, pred_list, True, False))\n obj_list = []\n elif word != '(':\n obj_list.append(word)\n elif keyword == 'functions': #functions\n if word == ')':\n p_name = obj_list[0]\n if obj_list[0] == '-':\n obj_list = obj_list[2:]\n #print \"function: \" + word + \" - \" + str(obj_list)\n self.functions.append(Function(obj_list))\n obj_list = []\n elif word != '(':\n obj_list.append(word)\n\n #Work on the actions\n new_actions = []\n for action in self.actions:\n if action[0] == '-':\n action = action[2:]\n act_name = action[1]\n act = {}\n action = action[2:]\n keyword = ''\n for word in action:\n if word.startswith(':'):\n keyword = word[1:]\n else:\n act.setdefault(keyword, []).append(word)\n self.agent_types.add(act.get('agent')[2])\n agent = self._parse_name_type_pairs(act.get('agent'),self.type_list)\n param_list = agent + self._parse_name_type_pairs(act.get('parameters')[1:-1],self.type_list)\n up_params = Predicate('', param_list, True, False)\n pre_list = self._parse_unground_propositions(act.get('precondition'))\n eff_list = self._parse_unground_propositions(act.get('effect'))\n new_act = Action(act_name, up_params, pre_list, eff_list)\n\n new_actions.append(new_act)\n self.actions = new_actions","def transform():","def __init__(self, domain, range):\n self.domain = domain.cloneSpace()\n self.range = range.cloneSpace()","def dict_of_domains(fc):\r\n # need to find root database (GDB or SDE)\r\n db_root = os.path.dirname(fc)\r\n while db_root[-4:].lower() != '.gdb' and db_root[-4:].lower() != '.sde':\r\n old_db_root = db_root # protect against infinite loop\r\n db_root = os.path.dirname(db_root)\r\n if old_db_root == db_root: # protect against infinite loop\r\n break\r\n arcpy.AddMessage(\"Retrieving Domains from \" + str(db_root))\r\n return {domain.name: domain.codedValues for domain in arcpy.da.ListDomains(db_root)}","def from_ExpressionDomain(K1, a, K0):\n return K1.from_sympy(a.ex)","def _domain_variants(domain):\n parts = domain.split('.')\n for i in range(len(parts), 1, -1):\n yield \".\".join(parts[-i:])","def replace_domain(address, old_domain, new_domain):\n old_domain_pattern = r'' + old_domain + '$'\n address = re.sub(old_domain_pattern, new_domain, address)\n return address","def _identify_domains(self):\n\n domains = [FEMDomain(TR3, MeshPart(self.mesh, labels=(0,)), self.media, self.labels)]\n return domains","def convert_to_model(self, *args):\n rt_xref_service_data, *_ = args\n return [RealtyTypeXRefService(**rt_xref_service) for rt_xref_service in rt_xref_service_data]","def reduceScale(structure,scale,dim,prec=1E-2):\n\n structure.translate_sites(indices=range(structure.num_sites),\n vector=-1*structure.sites[0].frac_coords+[0,0,.5])\n\n specs = structure.species\n cart = [x.coords for x in structure.sites]\n\n lat = np.array(structure.lattice.as_dict()['matrix'])\n if type(scale)==type(1) or type(scale)==type(1.0):\n lat[0]*=scale\n if dim>=2:\n lat[1]*=scale\n if dim==3:\n lat[2]*=scale\n else:\n lat[0]*=scale[0]\n lat[1]*=scale[1]\n lat[2]*=scale[2] \n\n fracs = np.around(abs(np.array(np.linalg.solve(lat.T,np.array(cart).T).T)),4)\n\n specs = []\n u_cart = []\n i=0\n n_fracs = []\n for frac in fracs:\n if not np.any([magni(np.around(frac,4)%1-np.around(x,4)%1)<prec for x in n_fracs]):\n n_fracs.append(frac%1)\n #if frac[0]<1 and frac[1]<1 and frac[2]<1:\n \n specs.append(structure.species[i])\n u_cart.append(cart[i])\n i+=1\n \n new_sites = []\n i=0\n for site in u_cart:\n p = PeriodicSite(lattice = Lattice(lat),coords=site,\n coords_are_cartesian=True,\n species=specs[i])\n new_sites.append(p)\n i+=1\n\n new_struct = Structure.from_sites(new_sites)\n\n return(new_struct)","def norm_it(startdate, enddate, dfin, dfname, l_back=1, toscale=True):\n\n # # subset\n dfin = dfin.loc[startdate:enddate]\n if toscale is True: \n # # normalize\n scaler = MinMaxScaler(feature_range=(0, 1))\n # # add normalized to dataset\n dfin[dfname+'_norm'] = scaler.fit_transform(dfin[\n dfname].to_numpy().reshape(-1, 1))\n # # lag\n dfin[dfname+'_norm'+str(l_back)] = dfin[dfname+'_norm'].shift(l_back)\n return dfin, scaler\n else:\n dfin[dfname+str(l_back)] = dfin[dfname].shift(l_back)\n return dfin","def generate_domain(domain):\n\n if domain.lower() in (\"greenland\", \"gris\", \"gris_ext\"):\n pism_exec = \"pismr\"\n elif domain.lower() in (\"og\"):\n pism_exec = \"pismr -no_model_strip 0 -calving_wrap_around\"\n elif domain.lower() in (\"hia\"):\n x_min = -652200.0\n x_max = -232600.0\n y_min = -1263900.0\n y_max = -943500.0\n pism_exec = \"\"\"pismr -x_range {x_min},{x_max} -y_range {y_min},{y_max} -bootstrap\"\"\".format(\n x_min=x_min, x_max=x_max, y_min=y_min, y_max=y_max\n )\n elif domain.lower() in (\"jakobshavn\", \"jib\"):\n x_min = -280000.0\n x_max = 320000.0\n y_min = -2410000.0\n y_max = -2020000.0\n pism_exec = \"\"\"pismr -regional -x_range {x_min},{x_max} -y_range {y_min},{y_max} -bootstrap -regional.zero_gradient true -regional.no_model_strip 4.5\"\"\".format(\n x_min=x_min, x_max=x_max, y_min=y_min, y_max=y_max\n )\n else:\n print((\"Domain {} not recognized, exiting\".format(domain)))\n import sys\n\n sys.exit(0)\n\n return pism_exec","def parse_domain(domain_str):\n def floatify(latlon):\n \"\"\" Turns a latlon string into a float \"\"\"\n sign = -2. * (latlon[-1].lower() in ['s', 'w']) + 1\n return float(latlon[:-1]) * sign\n corners = domain_str.strip().split(',')\n if not len(corners) == 4:\n raise BadQuery(\"Expected four comma seperated values \"\n \"defining the corners of the domain.\")\n\n is_lat = lambda x: x[-1].lower() in ['n', 's']\n lats = filter(is_lat, corners)\n if not len(lats) == 2:\n raise BadQuery(\"Expected two latitudes (determined by \" +\n \"values ending in 'N' or 'S'\")\n is_lon = lambda x: x[-1].lower() in ['e', 'w']\n lons = filter(is_lon, corners)\n if not len(lons) == 2:\n raise BadQuery(\"Expected two longitudes (determined by \" +\n \"values ending in 'E' or 'W'\")\n lats = np.array([floatify(x) for x in lats])\n lons = np.array([floatify(x) for x in lons])\n # make sure latitudes are in range.\n if np.any(lats > 90.) or np.any(lats < -90):\n raise BadQuery(\"Latitudes must be within -90 and 90, got %s\" %\n ','.join(map(str, lats)))\n # we let the user use either longitudes of 0 to 360\n # or -180 to 180, then convert to nautical (-180 to 180).\n if np.any(lons > 360.) or np.any(lons < -180.):\n raise BadQuery(\"Longitudes must be within -180 and 360, got %s\" %\n ','.join(map(str, lons)))\n # make sure lons end up in -180 to 180.\n lons = np.mod(lons + 180., 360.) - 180.\n\n lats = np.sort(lats)\n # chose the order for lons that minimizes the\n # angle. This means people can't make requests\n # for more than 180 degrees at a time. But that seems\n # more desirable than accidentally sending near global\n # files when the dateline is crossed.\n angular_diff = lambda x, y: np.mod(x - y, 360.)\n if angular_diff(*lons) > angular_diff(*reversed(lons)):\n lons = lons[::-1]\n\n domain = {'N': lats[1], 'S': lats[0],\n 'W': lons[1], 'E': lons[0]}\n return domain","def convert_to(self, K):\n if K == self.domain:\n return self.copy()\n\n rep = self.rep\n\n # The DFM, DDM and SDM types do not do any implicit conversions so we\n # manage switching between DDM and DFM here.\n if rep.is_DFM and not DFM._supports_domain(K):\n rep_K = rep.to_ddm().convert_to(K)\n elif rep.is_DDM and DFM._supports_domain(K):\n rep_K = rep.convert_to(K).to_dfm()\n else:\n rep_K = rep.convert_to(K)\n\n return self.from_rep(rep_K)","def transform_series(obj):\n vals = obj.values\n return transform_array(vals)","def read_fusion_domains(fus_path, wt_path, fusion_annot, min_domain_len=25, min_domain_frac=0.5):\n # read in domains\n domains = pd.read_csv(fus_path, sep='\\t')\n wt_domains = pd.read_csv(wt_path, sep='\\t')\n\n # filter out transmembrane helix\n is_tm = domains[\"Domain_ID\"].str.startswith('TMhelix')\n domains = domains[~is_tm].copy()\n\n # filter out short domains\n domain_len = domains['Protein_end'] - domains['Protein_start']\n domains = domains[domain_len>=min_domain_len].copy()\n\n # merge in info about where the fusion happened\n domains = pd.merge(domains, fusion_annot[['ID', 'CodonPos1', 'CodonPos2']].drop_duplicates('ID'), on='ID', how='left')\n domains['origin of domain'] = \"3'_gene domain\"\n domains.loc[domains['Protein_end']<=domains['CodonPos1'], 'origin of domain'] = \"5'_gene domain\"\n\n # separate out protein ID\n domains['ID_five'] = domains['ID'].str.split('-', expand=True)[0]\n domains['ID_three'] = domains['ID'].str.split('-', expand=True)[1].str.split(':', expand=True)[0]\n\n # check 5' domains\n domains_five = domains[domains[\"origin of domain\"]==\"5'_gene domain\"].copy()\n is_good_domain = []\n for ix, row in domains_five.iterrows():\n domain_end = row['Protein_end']\n fus_codon_pos = row['CodonPos1']\n if fus_codon_pos==domain_end:\n pid = row['ID_five']\n is_pid = wt_domains['PROT_ID']==pid\n is_loc = (fus_codon_pos>=wt_domains['Protein_start']) & (fus_codon_pos<=wt_domains['Protein_end'])\n tmp_df = wt_domains.loc[is_pid & is_loc,:]\n if len(tmp_df)==0:\n is_good_domain.append(True)\n continue\n else:\n tmp = tmp_df.iloc[0]\n frac_domain = (fus_codon_pos - tmp['Protein_start']) / (tmp['Protein_end'] - tmp['Protein_start'])\n if frac_domain>=min_domain_frac:\n is_good_domain.append(True)\n else:\n is_good_domain.append(False)\n else:\n is_good_domain.append(True)\n domains_five = domains_five.loc[is_good_domain,:]\n # check 3' domains\n domains_three = domains[domains[\"origin of domain\"]==\"3'_gene domain\"].copy()\n is_good_domain = []\n for ix, row in domains_three.iterrows():\n domain_start = row['Protein_start']\n fus_codon_pos = row['CodonPos2']\n if fus_codon_pos==domain_start:\n pid = row['ID_three']\n is_pid = wt_domains['PROT_ID']==pid\n is_loc = (fus_codon_pos>=wt_domains['Protein_start']) & (fus_codon_pos<=wt_domains['Protein_end'])\n tmp_df = wt_domains.loc[is_pid & is_loc,:]\n if len(tmp_df)==0:\n is_good_domain.append(True)\n continue\n else:\n tmp = tmp_df.iloc[0]\n frac_domain = (tmp['Protein_end']-fus_codon_pos) / (tmp['Protein_end'] - tmp['Protein_start'])\n if frac_domain>=min_domain_frac:\n is_good_domain.append(True)\n else:\n is_good_domain.append(False)\n else:\n is_good_domain.append(True)\n domains_three = domains_three.loc[is_good_domain,:]\n # append results\n domains = pd.concat([domains_five, domains_three])\n\n # aggregate all domains for one fusion into a single line\n domains = domains[~domains['Domain_name'].isnull()]\n domains_agg = domains.groupby(['ID', 'origin of domain'])['Domain_name'].agg(lambda x: ','.join(x))\n domains_agg = domains_agg.reset_index()\n domains_agg = domains_agg.pivot(index='ID', columns='origin of domain', values='Domain_name')\n domains_agg = domains_agg.reset_index()\n\n return domains_agg","def domain( self ):\n raise NotImplementedError(\"domain\")","def shapely_to_cf(geometries: xr.DataArray | Sequence, grid_mapping: str | None = None):\n # Get all types to call the appropriate translation function.\n types = {\n geom.item().geom_type if isinstance(geom, xr.DataArray) else geom.geom_type\n for geom in geometries\n }\n if types.issubset({\"Point\", \"MultiPoint\"}):\n ds = points_to_cf(geometries)\n elif types.issubset({\"Polygon\", \"MultiPolygon\"}) or types.issubset(\n {\"LineString\", \"MultiLineString\"}\n ):\n raise NotImplementedError(\"Only point geometries conversion is implemented.\")\n else:\n raise ValueError(\n f\"Mixed geometry types are not supported in CF-compliant datasets. Got {types}\"\n )\n\n # Special treatment of selected grid mappings\n if grid_mapping == \"longitude_latitude\":\n # Special case for longitude_latitude grid mapping\n ds = ds.rename(crd_x=\"lon\", crd_y=\"lat\")\n ds.lon.attrs.update(units=\"degrees_east\", standard_name=\"longitude\")\n ds.lat.attrs.update(units=\"degrees_north\", standard_name=\"latitude\")\n ds.geometry_container.attrs.update(coordinates=\"lon lat\")\n ds.x.attrs.update(units=\"degrees_east\", standard_name=\"longitude\")\n ds.y.attrs.update(units=\"degrees_north\", standard_name=\"latitude\")\n elif grid_mapping is not None:\n raise NotImplementedError(\n f\"Only grid mapping longitude_latitude is implemented. Got {grid_mapping}.\"\n )\n\n return ds","def get_domains(graph: Graph, property_to_id: Dict[str, int], entity_type_to_id: Dict[str, int]) -> Dict[int, int]:\n # dictionary pointing from object property id to an entity type id\n domains = {}\n\n # add all domain triples for which the subject is an object property and the object is an entity type\n for subject, predicate, object in graph.triples((None, RDFS.domain, None)):\n if subject in property_to_id and object in entity_type_to_id:\n domains[property_to_id[subject]] = entity_type_to_id[object]\n\n return domains","def get_domain(self, rel_name):\n return self._declaration[rel_name].domain_type","def make_companies():\n logging.info(\"Making CH\")\n companies_address = get_address()\n companies_sector = get_sector()\n\n companies = (\n companies_address[[\"company_number\", \"postcode\"]]\n .merge(\n companies_sector.query(\"rank==1\")[[\"company_number\", \"SIC4_code\"]],\n on=\"company_number\",\n )\n .assign(division=lambda x: [c[:2] for c in x[\"SIC4_code\"]])\n .assign(division_name=lambda x: x[\"division\"].map(_DIV_NAME_LOOKUP))\n .merge(nspl, left_on=\"postcode\", right_on=\"pcds\")\n )\n\n return companies","def resolveOriginalDomains():\n print('[+] Populating Domain Name Resolution for later check ')\n\n try:\n for domain in domains:\n response = dns.resolver.query(domain)\n d = Domain_Poison_Check(domain)\n print('[+] Domain: %s' % domain)\n for record in response:\n print(' |____> maps to %s.' % (record.address))\n d.pushAddr(record)\n check_domain_poison_results.append(d)\n return time.time()\n except Exception as err:\n print('[+] Exception: %s' % err)\n traceback.print_exc()\n return time.time()","def change_domain(self, new_domain):\n self.domain=new_domain\n for pm in self._maps.values():\n pm.change_domain(new_domain)","def get_input_domains():\n df = pandas.read_excel(\"AutoScrapy/files/EBE21 - Top 100 Onlineshops to scrapp.ods\", engine=\"odf\")\n list_of_addresses = df['Domain'].to_list()\n list_of_addresses = [(\"http://\" + address) for address in list_of_addresses]\n print(list_of_addresses)\n return list_of_addresses","def ground_to_exact(f):\n return f.set_domain(f.dom.get_exact())","def createDomain(self):\n if not self.rank:\n logging.info('Creating domain')\n\n if 'box' in self.pargs:\n self.lmp.command('region domain block ' + ('{} ' * len(self.pargs['box'])).format(*self.pargs['box']) + ' units box volume_limit 1e-20')\n elif 'cylinder' in self.pargs:\n self.lmp.command('region domain cylinder ' + ('{} ' * len(self.pargs['cylinder'])).format(*self.pargs['cylinder']) + ' units box volume_limit 1e-20') \n\n self.lmp.command('create_box {} domain'.format(self.pargs['nSS']))","def __learn__(self):\n\n return domains # including 3 values, (begin, end, key)","def cf_to_shapely(ds: xr.Dataset):\n geom_type = ds.geometry_container.attrs[\"geometry_type\"]\n if geom_type == \"point\":\n geometries = cf_to_points(ds)\n elif geom_type in [\"line\", \"polygon\"]:\n raise NotImplementedError(\"Only point geometries conversion is implemented.\")\n else:\n raise ValueError(\n f\"Valid CF geometry types are 'point', 'line' and 'polygon'. Got {geom_type}\"\n )\n\n return geometries.rename(\"geometry\")","def domain_region_grow(counts, ndomain):\n \n # make it so maximum value is 100,000,000 (fits well inside int32)\n norm = 10000.0 / counts.max()\n wts = np.square(norm * counts).astype(int32)\n# wts = ((norm * counts)**1.0).astype(int32)\n \n tot = wts.sum(dtype=int64)\n \n sum_per_domain = tot//ndomain + 1\n \n doms = domain_regiongrow(wts, sum_per_domain)\n return doms","def fixDomains(self, energyMin, energyMax, domainToFix):\n\n return 0","def fixDomains(self, energyMin, energyMax, domainToFix):\n\n return 0","def convert(osis_refs):\n is_string = False\n if isinstance(osis_refs, str):\n osis_refs = [osis_refs]\n is_string = True\n\n usfm_refs = []\n\n for ref in osis_refs:\n parts = ref.split('.')\n\n # If the book part is different, convert otherwise just uppercase it\n if parts[0].lower() in BOOKS.keys():\n parts[0] = BOOKS[parts[0].lower()]\n else:\n parts[0] = parts[0].upper()\n\n usfm_refs.append('.'.join(parts))\n\n return usfm_refs if is_string is False else usfm_refs[0]","def _adddomain(self, domain: Domain):\n\n domain = copy.deepcopy(domain)\n if self.model is not None:\n # Check that model and domain are compatible\n self._validate_model_domain(self.model, domain)\n\n # Add in domain\n self.domain = domain\n\n # Setup base namelists\n self._set_base_namelists()\n else:\n self.domain = domain","def merge_new_domain (cls, base, nffg, log=logging.getLogger(\"MERGE\")):\n # Get new domain name\n domain = cls.detect_domains(nffg=nffg)\n if len(domain) == 0:\n log.error(\"No domain detected in new %s!\" % nffg)\n return\n if len(domain) > 1:\n log.warning(\"Multiple domain name detected in new %s!\" % nffg)\n return\n # Copy infras\n log.debug(\"Merge domain: %s resource info into %s...\" % (domain.pop(),\n base.id))\n # Check if the infra with given id is already exist in the base NFFG\n for infra in nffg.infras:\n if infra.id not in base:\n c_infra = base.add_infra(infra=deepcopy(infra))\n log.debug(\"Copy infra node: %s\" % c_infra)\n else:\n log.warning(\"Infra node: %s does already exist in %s. Skip adding...\" %\n (infra, base))\n # Copy NFs\n for nf in nffg.nfs:\n if nf.id not in base:\n c_nf = base.add_nf(nf=deepcopy(nf))\n log.debug(\"Copy NF node: %s\" % c_nf)\n else:\n log.warning(\"NF node: %s does already exist in %s. Skip adding...\" %\n (nf, base))\n # Copy SAPs\n for sap_id in [s.id for s in nffg.saps]:\n if sap_id in [s.id for s in base.saps]:\n # Found inter-domain SAP\n log.debug(\"Found Inter-domain SAP: %s\" % sap_id)\n # Search outgoing links from SAP, should be only one\n b_links = [l for u, v, l in base.real_out_edges_iter(sap_id)]\n if len(b_links) < 1:\n log.warning(\n \"SAP is not connected to any node! Maybe you forgot to call \"\n \"duplicate_static_links?\")\n return\n elif 1 < len(b_links):\n log.warning(\n \"Inter-domain SAP should have one and only one connection to the \"\n \"domain! Using only the first connection.\")\n continue\n # Get inter-domain port in base NFFG\n domain_port_dov = b_links[0].dst\n sap_port_dov = b_links[0].src\n log.debug(\"Found inter-domain port: %s\" % domain_port_dov)\n # Search outgoing links from SAP, should be only one\n n_links = [l for u, v, l in nffg.real_out_edges_iter(sap_id)]\n if len(n_links) < 1:\n log.warning(\n \"SAP is not connected to any node! Maybe you forgot to call \"\n \"duplicate_static_links?\")\n return\n elif 1 < len(n_links):\n log.warning(\n \"Inter-domain SAP should have one and only one connection to the \"\n \"domain! Using only the first connection.\")\n continue\n # Get port and Infra id's in nffg NFFG\n p_id = n_links[0].dst.id\n n_id = n_links[0].dst.node.id\n # Get the inter-domain port from already copied Infra\n domain_port_nffg = base.network.node[n_id].ports[p_id]\n sap_port_nffg = n_links[0].src\n log.debug(\"Found inter-domain port: %s\" % domain_port_nffg)\n\n # # If the two resource value does not match\n # if sap_port_dov.delay != sap_port_nffg.delay:\n # if sap_port_dov.delay is None:\n # # If first is None the other can not be None\n # s_delay = sap_port_nffg.delay\n # elif sap_port_nffg.delay is None:\n # # If second is None the other can not be None\n # s_delay = sap_port_dov.delay\n # else:\n # # Both values are valid, but different\n # s_delay = max(sap_port_dov.delay, sap_port_nffg.delay)\n # log.warning(\n # \"Inter-domain delay values (%s, %s) are set but do not match!\"\n # \" Use max: %s\" % (sap_port_dov.delay, sap_port_nffg.delay,\n # s_delay))\n # else:\n # # Both value match: ether valid values or Nones --> choose first\n # value\n # s_delay = sap_port_dov.delay\n #\n # # If the two resource value does not match\n # if sap_port_dov.bandwidth != sap_port_nffg.bandwidth:\n # if sap_port_dov.bandwidth is None:\n # # If first is None the other can not be None\n # s_bandwidth = sap_port_nffg.bandwidth\n # elif sap_port_nffg.bandwidth is None:\n # # If second is None the other can not be None\n # s_bandwidth = sap_port_dov.bandwidth\n # else:\n # # Both values are valid, but different\n # s_bandwidth = min(sap_port_dov.bandwidth,\n # sap_port_nffg.bandwidth)\n # log.warning(\n # \"Inter-domain bandwidth values (%s, %s) are set but do not\n # match!\"\n # \" Use min: %s\" % (sap_port_dov.bandwidth,\n # sap_port_nffg.bandwidth, s_bandwidth))\n # else:\n # # Both value match: ether valid values or Nones --> choose first\n # value\n # s_bandwidth = sap_port_dov.bandwidth\n #\n # log.debug(\"Detected inter-domain resource values: delay: %s, \"\n # \"bandwidth: %s\" % (s_delay, s_bandwidth))\n\n # Copy inter-domain port properties/values for redundant storing\n if len(domain_port_nffg.properties) > 0:\n domain_port_dov.properties.update(domain_port_nffg.properties)\n log.debug(\"Copy inter-domain port properties: %s\" %\n domain_port_dov.properties)\n elif len(domain_port_dov.properties) > 0:\n domain_port_nffg.properties.update(domain_port_dov.properties)\n log.debug(\"Copy inter-domain port properties: %s\" %\n domain_port_nffg.properties)\n # Ensure to add sap tag to inter domain ports\n if 'sap' not in domain_port_dov.properties:\n domain_port_dov.add_property(\"sap\", sap_id)\n if 'sap' not in domain_port_nffg.properties:\n domain_port_nffg.add_property(\"sap\", sap_id)\n # Signal Inter-domain port type\n domain_port_dov.add_property(\"type\", \"inter-domain\")\n domain_port_nffg.add_property(\"type\", \"inter-domain\")\n\n # Copy SAP port values into the infra ports\n domain_port_dov.name = sap_port_dov.name\n domain_port_dov.sap = sap_port_dov.sap\n domain_port_dov.capability = sap_port_dov.capability\n domain_port_dov.technology = sap_port_dov.technology\n domain_port_dov.delay = sap_port_dov.delay\n domain_port_dov.bandwidth = sap_port_dov.bandwidth\n domain_port_dov.cost = sap_port_dov.cost\n domain_port_dov.controller = sap_port_dov.controller\n domain_port_dov.orchestrator = sap_port_dov.orchestrator\n domain_port_dov.l2 = sap_port_dov.l2\n domain_port_dov.l4 = sap_port_dov.l4\n for l3 in sap_port_dov.l3:\n domain_port_dov.l3.append(l3.copy())\n domain_port_dov.metadata.update(sap_port_dov.metadata)\n\n domain_port_nffg.name = sap_port_nffg.name\n domain_port_nffg.sap = sap_port_nffg.sap\n domain_port_nffg.capability = sap_port_nffg.capability\n domain_port_nffg.technology = sap_port_nffg.technology\n domain_port_nffg.delay = sap_port_nffg.delay\n domain_port_nffg.bandwidth = sap_port_nffg.bandwidth\n domain_port_nffg.cost = sap_port_nffg.cost\n domain_port_nffg.controller = sap_port_nffg.controller\n domain_port_nffg.orchestrator = sap_port_nffg.orchestrator\n domain_port_nffg.l2 = sap_port_nffg.l2\n domain_port_nffg.l4 = sap_port_nffg.l4\n for l3 in sap_port_nffg.l3:\n domain_port_nffg.l3.append(l3.copy())\n domain_port_nffg.metadata.update(sap_port_nffg.metadata)\n\n # Delete both inter-domain SAP and links connected to them\n base.del_node(sap_id)\n nffg.del_node(sap_id)\n\n # Add the inter-domain links for both ways\n l1, l2 = base.add_undirected_link(\n p1p2id=\"inter-domain-link-%s\" % sap_id,\n p2p1id=\"inter-domain-link-%s-back\" % sap_id,\n port1=domain_port_dov,\n port2=domain_port_nffg)\n # Set delay/bandwidth values for outgoing link port1 -> port2\n l1.delay = domain_port_dov.delay\n l1.bandwidth = domain_port_dov.bandwidth\n # Set delay/bandwidth values for outgoing link port2 -> port2\n l2.delay = domain_port_nffg.delay\n l2.bandwidth = domain_port_nffg.bandwidth\n\n else:\n # Normal SAP --> copy SAP\n c_sap = base.add_sap(sap_obj=deepcopy(nffg.network.node[sap_id]))\n log.debug(\"Copy SAP: %s\" % c_sap)\n # Copy remaining links which should be valid\n for u, v, link in nffg.network.edges_iter(data=True):\n src_port = base.network.node[u].ports[link.src.id]\n dst_port = base.network.node[v].ports[link.dst.id]\n c_link = deepcopy(link)\n c_link.src = src_port\n c_link.dst = dst_port\n base.add_link(src_port=src_port, dst_port=dst_port, link=c_link)\n log.debug(\"Copy Link: %s\" % c_link)\n log.debug(\"Domain merging has been finished!\")\n # Return the updated NFFG\n return base","def setup_domains():\n sdb = boto.connect_sdb()\n sdb.create_domain(\"mls_domain\")","def domain(self, domain=None):\n\n return self.domain_class(apiobj=self, domainname=domain)","def transform(self, series_of_articles):\n # sklearn vectorizer needs an iterable list/series,\n # so convert to list of just a single string\n if isinstance(series_of_articles, str):\n series_of_articles = [series_of_articles] # Convert to list\n result = self.vectorizer.transform(series_of_articles)\n dtm = self._post_process(result)\n return dtm","def translateValues(nadPoints):\n with arcpy.da.UpdateCursor(nadPoints,\n ['StN_PreDir', 'StN_PosDir', 'StN_PosTyp', 'County']) as cursor:\n for row in cursor:\n row[0] = directionDomain.get(row[0], None)\n row[1] = directionDomain.get(row[1], None)\n row[2] = streetDomain.get(row[2], None)\n row[3] = countyFipsDomain.get(row[3], None)\n cursor.updateRow(row)","def transform_atrf2014_to_gda2020(x, y, z, epoch_from, vcv=None):\n return conform14(x, y, z, epoch_from, atrf2014_to_gda2020, vcv=vcv)","def _scipy_transform(fqdn, value):\n import scipy\n return _package_transform(scipy, fqdn, value)","def find_domain_range(record):\n response = {\"domain\": [], \"range\": []}\n if \"http://schema.org/domainIncludes\" in record:\n if isinstance(record[\"http://schema.org/domainIncludes\"], dict):\n response[\"domain\"] = [record[\"http://schema.org/domainIncludes\"][\"@id\"]]\n elif isinstance(record[\"http://schema.org/domainIncludes\"], list):\n response[\"domain\"] = [\n _item[\"@id\"] for _item in record[\"http://schema.org/domainIncludes\"]\n ]\n if \"http://schema.org/rangeIncludes\" in record:\n if isinstance(record[\"http://schema.org/rangeIncludes\"], dict):\n response[\"range\"] = [record[\"http://schema.org/rangeIncludes\"][\"@id\"]]\n elif isinstance(record[\"http://schema.org/rangeIncludes\"], list):\n response[\"range\"] = [\n _item[\"@id\"] for _item in record[\"http://schema.org/rangeIncludes\"]\n ]\n return (response[\"domain\"], response[\"range\"])","def __init__(self, cfg):\n self.domains = []\n\n # process domains in order\n for i in range(1, len(cfg)+1):\n this_dom = cfg[str(i)]\n par_dom = self.domains[this_dom['parent_id']-1] if 'parent_id' in this_dom else None\n self.domains.append(WPSDomainLCC(i, this_dom, par_dom))","def convert_to_model(self, *args):\n state_xref_service_data, *_ = args\n return [StateXRefService(**state_xref_service) for state_xref_service in state_xref_service_data]","def transform_gda2020_to_atrf2014(x, y, z, epoch_to, vcv=None):\n return conform14(x, y, z, epoch_to, -atrf2014_to_gda2020, vcv=vcv)","def get_technique_to_domain():\n global technique_to_domain\n \n if not technique_to_domain:\n technique_to_domain = stixhelpers.get_technique_id_domain_map(get_ms())\n \n return technique_to_domain","def update_domain():\n\n for e in Expr.search() + User.search(): e.set_tld(config.server_name)","def get_feature_domain_dict(self):\n feature_domain_dict = {}\n for feature_index in range(len(self.train_examples[0])):\n domain = set([example[feature_index] for example in self.train_examples])\n feature_domain_dict[self.features[feature_index]] = domain\n\n return feature_domain_dict","def _from_sequence(cls, scalars):\n return cls(pa.array(scalars))","def get_storage_domains(cohesity_client):\n storage_domain_list = cohesity_client.view_boxes.get_view_boxes()\n for domain in storage_domain_list:\n exported_res_dict[\"Storage Domains\"].append(domain.name)\n return storage_domain_list","def normalize(feats, train_nid, dtype=np.float32):\n train_feats = feats[train_nid]\n scaler = preprocessing.StandardScaler()\n scaler.fit(train_feats)\n feats = scaler.transform(feats)\n return feats.astype(dtype)","def normaliseScores(essays):\n normalised_scores = []\n for index, essay in essays.iterrows():\n score = essay['domain1_score']\n # essay_set refers to the prompt(topic) of the essay\n essay_set = essay['essay_set']\n # Perform min-max normalization on the scores to get range in [0-1]\n normalised_score = (score - MIN_SCORES[essay_set]) / (MAX_SCORES[essay_set] - MIN_SCORES[essay_set])\n normalised_scores.append(normalised_score)\n return np.array(normalised_scores)","def transform_geometries(datasource, src_epsg, dst_epsg):\n # Part 1\n src_srs = osr.SpatialReference()\n src_srs.ImportFromEPSG(src_epsg)\n dst_srs = osr.SpatialReference()\n dst_srs.ImportFromEPSG(dst_epsg)\n transformation = osr.CoordinateTransformation(src_srs, dst_srs)\n layer = datasource.GetLayerByIndex(0)\n \n # Part 2\n geoms = []\n layer.ResetReading()\n for feature in layer:\n geom = feature.GetGeometryRef().Clone()\n geom.Transform(transformation)\n geoms.append(geom)\n return geoms","def catalogConvert():\n ret = libxml2mod.xmlCatalogConvert()\n return ret","def domains(self, domains):\n\n self._domains = domains","def convert(self, chain, agebins=None):\n cols = [\"av\", \"logzsol\"]\n sfh, sfh_label = construct_sfh_measure(chain, agebins)\n niter = len(sfh[0])\n cols += sfh_label\n dt = np.dtype([(c, np.float) for c in cols])\n params = np.zeros(niter, dtype=dt)\n\n for c in cols:\n if c in chain.dtype.names:\n params[c] = np.squeeze(chain[c])\n\n # --- dust attenuation\n params[\"av\"] = np.squeeze(1.086 * chain[\"dust2\"])\n\n # --- stellar ---\n for i, c in enumerate(sfh_label):\n params[c] = np.squeeze(sfh[i])\n return params","def get_schaefer2018_yeo(scale, data_dir=None):\n\n # substring in schaefer parcel names indiciating yeo network affiliation\n schaefer_yeo = dict(\n Vis='visual',\n SomMot='somatomotor',\n DorsAttn='dorsal attention',\n SalVentAttn='ventral attention',\n Limbic='limbic',\n Cont='frontoparietal',\n Default='default mode'\n )\n\n # get requested annotation files\n schaefer = nndata.fetch_schaefer2018('fsaverage5',\n data_dir=data_dir)[scale]\n\n network_labels, parcel_labels = [], []\n for hemi in ('lh', 'rh'):\n # read in annotation file for given hemisphere\n annot = getattr(schaefer, hemi)\n labels, ctab, names = nib.freesurfer.read_annot(annot)\n names = [m.decode() for m in names]\n\n # create empty arrays for vertex- and parcel-level affiliations\n networks = np.zeros_like(labels)\n parcels = np.zeros(len(names), dtype=int)\n for n, parcel in enumerate(names):\n for abbrev, net in schaefer_yeo.items():\n # check which network this parcel belongs to by looking for the\n # network substring in the parcel name and assign accordingly\n if abbrev in parcel:\n parcels[n] = YEO_CODES[net]\n networks[labels == n] = YEO_CODES[net]\n\n # store network affiliations for this hemisphere\n network_labels.append(networks)\n parcel_labels.append(parcels)\n\n return NETWORKS(np.hstack(network_labels), np.hstack(parcel_labels))","def get_domain(id):\n return query(WEB_EXAMPLE_BASE + f\"/classical/domain/{id}\")","def get_domain(self, classname):\n subs = classname.split('-')\n sc = ''.join(subs[:2]) # section, class\n scsc = ''.join(subs) # section, class, subclass \n if self.codes.has_key(sc):\n return sc, self.codes[sc]\n return scsc, self.codes.get(scsc, None)","def as_domain(\n domain: DomainLike, input_dim: Optional[IntLike]\n) -> Tuple[DomainType, int]:\n if input_dim is not None and input_dim < 1:\n raise ValueError(\n f\"Input dimension must be positive. Current value is ({input_dim}).\"\n )\n\n if len(domain) != 2:\n raise ValueError(f\"'domain' must be of length 2 ({len(domain)}).\")\n\n # Domain limits must have equal dimensions\n if np.size(domain[0]) != np.size(domain[1]):\n raise ValueError(\n f\"Domain limits must be given either as scalars or arrays \"\n f\"of equal dimension. Current sizes are ({np.size(domain[0])}) \"\n f\"and ({np.size(domain[1])}).\"\n )\n\n domain_dim = int(np.size(domain[0]))\n\n # Input dimension not given, infer it from the domain.\n if input_dim is None:\n input_dim = domain_dim\n domain_a, domain_b = domain\n\n # Bounds are given as scalars: Expand domain limits.\n elif domain_dim == 1:\n domain_a = np.full((input_dim,), domain[0])\n domain_b = np.full((input_dim,), domain[1])\n\n # Bounds are given as arrays\n else:\n # Size of domain and input dimension do not match\n if input_dim != domain_dim:\n raise ValueError(\n f\"If domain limits are not scalars, their lengths \"\n f\"must match the input dimension ({input_dim}).\"\n )\n domain_a = domain[0]\n domain_b = domain[1]\n\n # convert bounds to 1D arrays if necessary\n domain_a = np.atleast_1d(domain_a)\n domain_b = np.atleast_1d(domain_b)\n if domain_a.ndim > 1 or domain_b.ndim > 1:\n raise ValueError(\n f\"Upper ({domain_b.ndim}) or lower ({domain_a.ndim}) \"\n f\"bounds have too many dimensions.\"\n )\n\n # Make sure the domain is non-empty\n if not np.all(domain_a < domain_b):\n raise ValueError(\n \"Integration domain must be non-empty. For example, some \"\n \"lower bounds may be larger than their corresponding \"\n \"upper bounds.\"\n )\n\n domain = (domain_a, domain_b)\n return domain, input_dim","def _numpy_transform(fqdn, value):\n import numpy\n return _package_transform(numpy, fqdn, value)","def _get_domain(self):\n self.ensure_one()\n domain = ['|', ('active', '=', True), ('active', '=', False)]\n # Check active\n if self.active == 'true':\n domain += [('active', '=', True)]\n elif self.active == 'false':\n domain += [('active', '=', False)]\n # Check partner type\n if self.partner_type == 'customer_or_supplier':\n domain += ['|', ('customer', '=', True), ('supplier', '=', True)]\n elif self.partner_type == 'customer_and_supplier':\n domain += [('customer', '=', True), ('supplier', '=', True)]\n elif self.partner_type == 'customer':\n domain += [('customer', '=', True)]\n elif self.partner_type == 'supplier':\n domain += [('supplier', '=', True)]\n # Check category\n if self.category_ids:\n domain += [('category_id', 'in', self.category_ids.ids)]\n return domain","def test_Synthase_serialisation(synthase, domains, tmp_path):\n result_dict = {\n \"header\": \"test\",\n \"sequence\": \"A\" * 200,\n \"domains\": [\n {\n \"start\": 1,\n \"end\": 90,\n \"type\": \"KS\",\n \"domain\": \"PKS_KS\",\n \"evalue\": None,\n \"bitscore\": None,\n \"pssm\": None,\n \"superfamily\": None,\n \"accession\": None,\n },\n {\n \"start\": 100,\n \"end\": 200,\n \"type\": \"AT\",\n \"domain\": \"PKS_AT\",\n \"evalue\": None,\n \"bitscore\": None,\n \"pssm\": None,\n \"superfamily\": None,\n \"accession\": None,\n },\n ],\n \"classification\": []\n }\n\n assert synthase.to_dict() == result_dict\n\n json_file = tmp_path / \"json\"\n json_file.write_text(synthase.to_json())\n\n with json_file.open() as js:\n from_json = Synthase.from_json(js)\n\n assert from_json.header == \"test\"\n assert from_json.sequence == \"A\" * 200\n assert from_json.domains == [domains[0], domains[2]]"],"string":"[\n \"def domain(self, domain):\",\n \"def convert_to(self, domain):\\n if domain == self.domain:\\n return self.copy()\\n elif domain == QQ and self.domain == ZZ:\\n return self._new(flint.fmpq_mat(self.rep), self.shape, domain)\\n elif domain == ZZ and self.domain == QQ:\\n # XXX: python-flint has no fmpz_mat.from_fmpq_mat\\n return self.to_ddm().convert_to(domain).to_dfm()\\n else:\\n # It is the callers responsibility to convert to DDM before calling\\n # this method if the domain is not supported by DFM.\\n raise NotImplementedError(\\\"Only ZZ and QQ are supported by DFM\\\")\",\n \"def format_domain(domain):\\n domain.ns_converted = []\\n for ns in domain.ns :\\n if isinstance(ns, objects.DomainHostAttr) :\\n ns_item = {\\n 'hostname' : ns.hostname,\\n 'ips' : []\\n }\\n\\n for hostaddr in ns.hostAddr :\\n ns_item['ips'].append(hostaddr.ip)\\n else :\\n ns_item = {\\n 'hostname' : ns.name,\\n 'ips' : [],\\n 'hostobj' : 1\\n }\\n domain.ns_converted.append(ns_item)\\n\\n return domain\",\n \"def transform_domains(points, from_domain, to_domain, use_mp=False, dps=None):\\n with mp.workdps(dps or mp.dps):\\n fl = mp.mpf if use_mp else float\\n a, b = map(fl, to_domain)\\n c, d = map(fl, from_domain)\\n scale = (b-a)/(d-c)\\n trans = a - (b-a)*c/(d-c)\\n return [min(b, max(a, scale*p + trans)) for p in points]\",\n \"def _infer_domain(self, name, domain, elements):\\n if '*' not in domain:\\n return domain\\n debug('guessing a better domain for {}: {}'.format(name, domain))\\n\\n # Domain as a list of references to Variables in the File/xr.Dataset\\n domain_ = [self[d] for d in domain]\\n\\n for i, d in enumerate(domain_): # Iterate over dimensions\\n e = set(elements[i])\\n if d.name != '*' or len(e) == 0: # pragma: no cover\\n assert set(d.values).issuperset(e)\\n continue # The stated domain matches the data; or no data\\n # '*' is given\\n if (self._state[name]['attrs']['type_code'] == gdxcc.GMS_DT_PAR and\\n self._implicit):\\n d = '_{}_{}'.format(name, i)\\n debug(('Constructing implicit set {} for dimension {} of {}\\\\n'\\n ' {} instead of {} elements')\\n .format(d, name, i, len(e), len(self['*'])))\\n self.coords[d] = elements[i]\\n d = self[d]\\n else:\\n # try to find a smaller domain for this dimension\\n # Iterate over every Set/Coordinate\\n for s in self.coords.values():\\n if s.ndim == 1 and set(s.values).issuperset(e) and \\\\\\n len(s) < len(d):\\n d = s # Found a smaller Set; use this instead\\n domain_[i] = d\\n\\n # Convert the references to names\\n inferred = [d.name for d in domain_]\\n\\n if domain != inferred:\\n # Store the result\\n self._state[name]['attrs']['domain_inferred'] = inferred\\n debug('…inferred {}.'.format(inferred))\\n else:\\n debug('…failed.')\\n\\n return inferred\",\n \"def set_subdomains(self, f):\\n s = \\\"::: setting 2D subdomains :::\\\"\\n print_text(s, cls=self)\\n\\n self.ff = MeshFunction('size_t', self.mesh)\\n self.cf = MeshFunction('size_t', self.mesh)\\n self.ff_acc = MeshFunction('size_t', self.mesh)\\n f.read(self.ff, 'ff')\\n f.read(self.cf, 'cf')\\n f.read(self.ff_acc, 'ff_acc')\\n \\n self.ds = Measure('ds')[self.ff]\\n self.dx = Measure('dx')[self.cf]\\n \\n self.dx_g = self.dx(0) # internal above grounded\\n self.dx_f = self.dx(1) # internal above floating\\n self.dBed_g = self.ds(3) # grounded bed\\n self.dBed_f = self.ds(5) # floating bed\\n self.dBed = self.ds(3) + self.ds(5) # bed\\n self.dSrf_gu = self.ds(8) # grounded with U observations\\n self.dSrf_fu = self.ds(9) # floating with U observations\\n self.dSrf_u = self.ds(8) + self.ds(9) # surface with U observations\\n self.dSrf_g = self.ds(2) + self.ds(8) # surface of grounded ice\\n self.dSrf_f = self.ds(6) + self.ds(9) # surface of floating ice\\n self.dSrf = self.ds(6) + self.ds(2) \\\\\\n + self.ds(8) + self.ds(9) # surface\\n self.dLat_d = self.ds(7) # lateral divide\\n self.dLat_to = self.ds(4) # lateral terminus overwater\\n self.dLat_tu = self.ds(10) # lateral terminus underwater\\n self.dLat_t = self.ds(4) + self.ds(10) # lateral terminus\\n self.dLat = self.ds(4) + self.ds(7) \\\\\\n + self.ds(10) # lateral\",\n \"def shp2idomain(shp_path,g,idomain,features_type=\\\"polygon\\\",layer=0):\\n \\n if len(idomain.shape) != 1:\\n idomain = idomain.reshape(g.nlay*g.nrow*g.ncol) # reshape idomain if not in the right format\\n res = g.intersect(shp_path,features_type,layer)\\n idomain[res.nodenumber] = 1\\n idomain = idomain.reshape(g.nlay,g.nrow,g.ncol)\",\n \"def fastlydomain(args):\\n pprint(api.domain(service_id, args[0], args[1]).attrs)\",\n \"def _parse_domain(self, f_domain):\\n\\n parse_tree = PDDL_Tree.create(f_domain)\\n\\n assert \\\"domain\\\" in parse_tree, \\\"Domain must have a name\\\"\\n self.domain_name = parse_tree [\\\"domain\\\"].named_children ()[0]\\n\\n # must read types before constants\\n if \\\":types\\\" in parse_tree:\\n if \\\"-\\\" in parse_tree[\\\":types\\\"].named_children():\\n type_hierarchy = PDDL_Utils.read_type(parse_tree[\\\":types\\\"])\\n self.parent_types = {subtype: parent for subtype, parent in type_hierarchy}\\n self.types = set(parse_tree[\\\":types\\\"].named_children())\\n self.types.discard(\\\"-\\\")\\n else:\\n self.types = set(parse_tree[\\\":types\\\"].named_children())\\n self.parent_types = {t: None for t in self.types}\\n else:\\n self.types = set([Predicate.OBJECT])\\n self.parent_types = {Predicate.OBJECT: None}\\n\\n self.agents = [a.name for a in parse_tree[\\\":agents\\\"].children]\\n self.types.add('agent')\\n self.parent_types['agent'] = None\\n self._add_objects([(ag, 'agent') for ag in self.agents])\\n\\n # must read in constants before actions or predicates\\n if \\\":constants\\\" in parse_tree:\\n object_list = PDDL_Utils.read_type(parse_tree[\\\":constants\\\"])\\n self._add_objects(object_list)\\n\\n #TODO this may not be correct, depending on the type hierarchy\\n const_map = dict()\\n for const in self.objects:\\n if len(self.obj_to_type[const]) == 0:\\n raise RuntimeError(\\\"No type for constant object %s\\\" % const)\\n else:\\n const_map[const] = list(self.obj_to_type[const])[0]\\n\\n self.predicates = [self.to_predicate(c, map=const_map) for c in parse_tree[\\\":predicates\\\"].children]\\n\\n # some predicates have this property: they are untyped.\\n for predicate in self.predicates:\\n if Predicate.OBJECT not in self.types and any([arg[1] == Predicate.OBJECT for arg in predicate.args]):\\n for t in self.types:\\n if self.parent_types[t] is None:\\n self.parent_types[t] = Predicate.OBJECT\\n\\n self.parent_types[Predicate.OBJECT] = None\\n self.types.add(Predicate.OBJECT)\\n self.type_to_obj[Predicate.OBJECT] = set([])\\n for obj, type_list in self.obj_to_type.items():\\n type_list.add(Predicate.OBJECT)\\n self.type_to_obj[Predicate.OBJECT].add(obj)\\n\\n # only need to do this once, obviously\\n break\\n\\n self.actions = [self.to_action(c) for c in parse_tree.find_all(\\\":action\\\")]\",\n \"def domain_range(domain, _range=[0, 1], return_transform=False):\\n\\n if not return_transform:\\n return interp1d([min(domain), max(domain)], [min(_range), max(_range)], bounds_error=False)\\n else:\\n m = interp1d([min(domain), max(domain)], [min(_range), max(_range)])\\n return [float(m(v)) for v in domain] # Take float, else returns weird numpy.ndarray element\",\n \"def _deserialize(self, data):\\n uri = data[1:-1]\\n # We have to retrieve the type to rebuild the object\\n attr = self.__dict__['field']\\n # Be careful when orig = None !!!!!\\n orig = getattr(attr.model, attr.name)\\n if None == orig:\\n return rdfSubject(rdflib.term.URIRef(uri))\\n elif isinstance(orig, list):\\n # rdfalchemy mapper gives me the solution\\n rt = attr.model.__class__.__dict__[attr.name].range_type\\n from rdfalchemy.orm import mapper\\n alch_map = mapper()\\n try:\\n cls = alch_map[str(rt)]\\n return cls(rdflib.term.URIRef(uri))\\n except:\\n rdfSubject(rdflib.term.URIRef(uri))\\n else:\\n return type(orig)(rdflib.term.URIRef(uri))\",\n \"def domain_to_geo(domain):\\n\\n return gi.record_by_name(domain)\",\n \"def par_domain(self):\",\n \"def fixDomains(self, domainMin, domainMax, fixToDomain):\\n\\n return 0\",\n \"def getDSDomainUnion(gdb_dss):\\n li_domain = []\\n for g in gdb_dss:\\n arcpy.env.workspace = os.path.join(inws, g)\\n dss = arcpy.ListDatasets()\\n for ds in dss:\\n desc = arcpy.Describe(ds)\\n sr = desc.spatialReference\\n # Notice that sr.domain is a string of (xmin, ymin, xmax, ymax) separated by space.\\n li_domain.append([float(x) for x in sr.domain.split(\\\" \\\")])\\n\\n df = pd.DataFrame(li_domain)\\n # Return min of xmin, ymin and max of xmax, ymax\\n return (df.iloc[:, 0].min(), df.iloc[:, 1].min(), df.iloc[:, 2].max(), df.iloc[:, 3].max())\",\n \"def ground_to_field(f):\\n return f.set_domain(f.dom.get_field())\",\n \"def as_domain(labeled_docs, labels, issource, domain, unlabeled_docs=None, unlabeled_y=None, tokken_pattern=r\\\"(?u)\\\\b\\\\w\\\\w+\\\\b\\\", min_df=1):\\n if issource:\\n counter = CountVectorizer(token_pattern=tokken_pattern, min_df=min_df)\\n v = counter.fit(labeled_docs).vocabulary_\\n tfidf = TfidfVectorizer(sublinear_tf=True, token_pattern=tokken_pattern, vocabulary=v)\\n else:\\n tfidf = TfidfVectorizer(sublinear_tf=True, token_pattern=tokken_pattern, min_df=min_df)\\n if unlabeled_docs is not None:\\n X = tfidf.fit_transform(labeled_docs)\\n U = tfidf.transform(unlabeled_docs)\\n y = np.array(labels)\\n V = tfidf.vocabulary_\\n U_y = np.array(unlabeled_y)\\n domain = Domain(X, y, V, domain, U, U_y)\\n return domain\\n X = tfidf.fit_transform(labeled_docs)\\n y = np.array(labels)\\n V = tfidf.vocabulary_\\n domain = Domain(X, y, V, domain)\\n return domain\",\n \"def set_domain(f, dom):\\n if f.dom == dom:\\n return f\\n else:\\n return f.per(dmp_set_domain(f.rep, f.lev, f.dom, dom), dom=dom)\",\n \"def order_domain_values(var,assignment,csp):\\n #right now it works only as just convert value and return\\n #no special black magic yet\\n return var.domain\",\n \"def dictTransform(domain):\\n url = urlStringFormat('tomorrow', domain)\\n data = requests.get(url)\\n dataDict = xmlToJson(data)\\n # Check if TimeSeries is dict or not. If it isn't, then it is an array of two TimeSeries\\n if type(dataDict['Publication_MarketDocument']['TimeSeries']) is dict:\\n url = urlStringFormat('today', domain)\\n data = requests.get(url)\\n dataDict = xmlToJson(data)\\n\\n # Transform dict into new dict without first hour and with 25. hour, \\n # which is appended from the second TimeSeries \\n priceFromSecondTimeSeries = dataDict['Publication_MarketDocument']['TimeSeries'][1]['Period']['Point']['price.amount']['$']\\n del dataDict['Publication_MarketDocument']['TimeSeries'][1]\\n # Transform array of one dictionary into dictionary\\n dataDict['Publication_MarketDocument']['TimeSeries'] = dataDict['Publication_MarketDocument']['TimeSeries'][0]\\n dataDict['Publication_MarketDocument']['TimeSeries']['Period']['Point'].append({'position': {'$': 25}, 'price.amount': {'$': priceFromSecondTimeSeries}})\\n del dataDict['Publication_MarketDocument']['TimeSeries']['Period']['Point'][0]\\n return dataDict\",\n \"def _translate_freq_domain(self, freq_domain):\\n\\n lin_spaces = [np.linspace(-0.5, 0.5, x) for x in freq_domain.shape]\\n meshgrids = np.meshgrid(*lin_spaces, indexing='ij')\\n grid_coords = np.array([mg.flatten() for mg in meshgrids])\\n\\n phase_shift = np.multiply(grid_coords, self.translations).sum(axis=0) # phase shift is added\\n exp_phase_shift = np.exp(-2j * math.pi * phase_shift)\\n freq_domain_translated = np.multiply(exp_phase_shift, freq_domain.flatten(order='C')).reshape(freq_domain.shape)\\n\\n return freq_domain_translated\",\n \"def stardomain(var, curr_domains):\\n\\t\\tif curr_domains:\\n\\t\\t\\tstardomain = sorted(curr_domains[state], key = lambda restaurant: (star_weight*math.log10(restaurant['stars']) + reviews_weight*math.log10(restaurant['review_count']))/(star_weight + reviews_weight))\\n\\t\\t\\tstardomain.reverse()\\n\\t\\telse:\\n\\t\\t\\tstardomain = sorted(state_domains[state], key = lambda restaurant: (star_weight*math.log10(restaurant['stars']) + reviews_weight*math.log10(restaurant['review_count']))/(star_weight + reviews_weight))\\n\\t\\t\\tstardomain.reverse()\\n\\t\\treturn stardomain\",\n \"def _get_domain(self):\\n self.ensure_one()\\n domain = []\\n return domain\",\n \"def handle_domains(\\n actapi: act.api.Act, content: Text, domains: List[Text]\\n) -> List[act.api.fact.Fact]:\\n\\n feeds_facts: List[act.api.fact.Fact] = []\\n\\n for domain in domains:\\n\\n chain = []\\n\\n chain.append(\\n actapi.fact(\\\"connectsTo\\\").source(\\\"content\\\", content).destination(\\\"uri\\\", \\\"*\\\")\\n )\\n chain.append(\\n actapi.fact(\\\"componentOf\\\").source(\\\"fqdn\\\", domain).destination(\\\"uri\\\", \\\"*\\\")\\n )\\n\\n feeds_facts += act.api.fact.fact_chain(*chain)\\n\\n return feeds_facts\",\n \"def _normalize_feature(feature: schema_pb2.Feature,\\n schema: schema_pb2.Schema) -> None:\\n\\n if feature.HasField(\\\"struct_domain\\\"):\\n for x in feature.struct_domain.feature:\\n _normalize_feature(x, schema)\\n if feature.HasField(\\\"domain\\\"):\\n for string_domain in schema.string_domain:\\n if string_domain.name == feature.domain:\\n feature.string_domain.CopyFrom(string_domain)\\n return\\n for int_domain in schema.int_domain:\\n if int_domain.name == feature.domain:\\n feature.int_domain.CopyFrom(int_domain)\\n return\\n for float_domain in schema.float_domain:\\n if float_domain.name == feature.domain:\\n feature.float_domain.CopyFrom(float_domain)\\n return\\n raise ValueError(\\\"Did not find domain {} in schema {}\\\".format(\\n feature.domain, schema))\",\n \"def _compue_domain(self, new_position):\\n if abs(new_position) <= self._safe_zone: \\n return self._domain\\n return self.Domain( sign(new_position) )\",\n \"def domain_decomposition(self, domain_decomposition):\\n\\n self._domain_decomposition = domain_decomposition\",\n \"def domain_resolve(self, domain):\\n a_result = server.resolve(domain, 'A')\\n mx_result = server.resolve(domain, 'MX')\\n ns_result = server.resolve(domain, 'NS')\\n\\n domain_data = {}\\n domain_data.setdefault('resolving', {'A': a_result})\\n domain_data['resolving']['MX'] = mx_result\\n domain_data['resolving']['NS'] = ns_result\\n\\n return domain_data\",\n \"def _make_tfx_schema_domain_info_inline(schema: schema_pb2.Schema) -> None:\\n # Reference to domains defined at schema level\\n domain_ref_to_string_domain = {d.name: d for d in schema.string_domain}\\n domain_ref_to_float_domain = {d.name: d for d in schema.float_domain}\\n domain_ref_to_int_domain = {d.name: d for d in schema.int_domain}\\n\\n # With the reference, it is safe to remove the domains defined at schema level\\n del schema.string_domain[:]\\n del schema.float_domain[:]\\n del schema.int_domain[:]\\n\\n for feature in schema.feature:\\n domain_info_case = feature.WhichOneof(\\\"domain_info\\\")\\n if domain_info_case == \\\"domain\\\":\\n domain_ref = feature.domain\\n if domain_ref in domain_ref_to_string_domain:\\n feature.string_domain.MergeFrom(domain_ref_to_string_domain[domain_ref])\\n elif domain_ref in domain_ref_to_float_domain:\\n feature.float_domain.MergeFrom(domain_ref_to_float_domain[domain_ref])\\n elif domain_ref in domain_ref_to_int_domain:\\n feature.int_domain.MergeFrom(domain_ref_to_int_domain[domain_ref])\",\n \"def freqdomain(self):\\n \\n\\n #self.df = self.f[1] - self.f[0]\\n #frequency vector\\n #fv = fftshift(fftfreq(len(eta),1./fs))\\n #fv = fv[len(fv)/2:]\\n \\n #spectral analysis\\n self.sn1 = self.espec1(self.n1)\\n self.sn2 = self.espec1(self.n2)\\n self.sn3 = self.espec1(self.n3)\\n self.sn12 = self.espec2(self.n1,self.n2)\\n self.sn13 = self.espec2(self.n1,self.n3)\\n self.sn23 = self.espec2(self.n2,self.n3)\\n \\n #delta freq\\n self.df = self.f[3] - self.f[2]\\n\\n #calculo do numero de onda\\n #self.wavenumber()\\n #k = numeronda(h,f,len(f))\\n #k = np.array(k)\\n\\n #calculo dos coeficientes de fourier - NDBC 96_01 e Steele (1992)\\n c = self.sn2[:,1] + self.sn3[:,1]\\n cc = np.sqrt(self.sn1[:,1] * (c))\\n \\n self.a1 = self.sn12[:,3] / cc\\n self.b1 = self.sn13[:,3] / cc\\n \\n self.a2 = (self.sn2[:,1] - self.sn3[:,1]) / c\\n self.b2 = 2 * self.sn12[:,2] / c\\n \\n #calcula direcao de onda\\n #mean direction\\n self.dire1 = np.array([np.angle(np.complex(self.b1[i],self.a1[i]),deg=True) for i in range(len(self.a1))])\\n \\n #principal direction\\n self.dire2 = 0.5 * np.array([np.angle(np.complex(self.b2[i],self.a2[i]),deg=True) for i in range(len(self.a2))])\\n \\n #condicao para valores maiores que 360 e menores que 0\\n self.dire1[np.where(self.dire1 < 0)] = self.dire1[np.where(self.dire1 < 0)] + 360\\n self.dire1[np.where(self.dire1 > 360)] = self.dire1[np.where(self.dire1 > 360)] - 360\\n self.dire2[np.where(self.dire2 < 0)] = self.dire2[np.where(self.dire2 < 0)] + 360\\n self.dire2[np.where(self.dire2 > 360)] = self.dire2[np.where(self.dire2 > 360)] - 360\\n \\n #acha o indice da frequencia de pico\\n ind = np.where(self.sn1[:,1] == np.max(self.sn1[:,1]))[0]\\n \\n #periodo de pico\\n self.tp = (1. / self.f[ind])[0]\\n \\n #momento espectral de ordem zero total - m0\\n self.m0 = np.sum(self.sn1[:,1]) * self.df\\n \\n #calculo da altura significativa\\n self.hm0 = 4.01 * np.sqrt(self.m0)\\n \\n #direcao do periodo de pico\\n self.dp = self.dire1[ind][0]\\n \\n #Espalhamento direcional\\n #Formula do sigma1 do livro Tucker&Pitt(2001) \\\"Waves in Ocean Engineering\\\" pags 196-198\\n c1 = np.sqrt(self.a1 ** 2 + self.b1 ** 2)\\n c2 = np.sqrt(self.a2 ** 2 + self.b2 ** 2)\\n \\n s1 = c1 / (1-c1)\\n s2 = (1 + 3 * c2 + np.sqrt(1 + 14 * c2 + c2 ** 2)) / (2 * (1 - c2))\\n \\n self.sigma1 = np.sqrt(2 - 2 * c1) * 180 / np.pi\\n self.sigma2 = np.sqrt((1 - c2) / 2) * 180 / np.pi\\n \\n self.sigma1p = np.real(self.sigma1[ind])[0]\\n self.sigma2p = np.real(self.sigma2[ind])[0]\\n \\n # pondaf = np.array([hm0, tp, dp, sigma1p, sigma2p])\\n \\n #hm0, tp, dp, sigma1, sigma2, sigma1p, sigma2p, f, df, k, sn, snx, sny, snn, snnx, snny, snxny, snxnx, snyny, a1, b1, a2, b2, dire1, dire2\\n #return hm0, tp, dp, sigma1, sigma2, sigma1p, sigma2p, f, df, k, sn, snx, sny, snn, snnx, snny, snxny, snxnx, snyny, a1, b1, a2, b2, dire1, dire2\",\n \"def detect_domains (nffg):\\n return {infra.domain for infra in nffg.infras}\",\n \"def _discretize_freestream_domain(self, xmin, xmax, ymin, ymax, zmin, zmax, resolution):\\n x = np.linspace(xmin, xmax, int(resolution.x1))\\n y = np.linspace(ymin, ymax, int(resolution.x2))\\n z = np.linspace(zmin, zmax, int(resolution.x3))\\n return np.meshgrid(x, y, z, indexing=\\\"ij\\\")\",\n \"def domain_features(domain, path_set):\\n return string_features_v2(domain_string(domain, path_set))\",\n \"def convert_to_model(self, *args):\\n city_xref_service_data, *_ = args\\n return [CityXRefService(**city_xref_service) for city_xref_service in city_xref_service_data]\",\n \"def _convert(self):\\n logger.info(\\\"Converting conformers to density\\\")\\n logger.debug(\\\"Masking\\\")\\n self._transformer.reset(full=True)\\n for n, coor in enumerate(self._coor_set):\\n self.conformer.coor = coor\\n self._transformer.mask(self._rmask)\\n mask = self._transformer.xmap.array > 0\\n self._transformer.reset(full=True)\\n\\n nvalues = mask.sum()\\n self._target = self.xmap.array[mask]\\n logger.debug(\\\"Density\\\")\\n nmodels = len(self._coor_set)\\n self._models = np.zeros((nmodels, nvalues), float)\\n for n, coor in enumerate(self._coor_set):\\n self.conformer.coor = coor\\n self.conformer.b = self._bs[n]\\n self._transformer.density()\\n model = self._models[n]\\n model[:] = self._transformer.xmap.array[mask]\\n np.maximum(model, self.options.bulk_solvent_level, out=model)\\n self._transformer.reset(full=True)\",\n \"def convert(self, values, axes_list, unit, is_norm):\\n \\n # Convert into right unit\\n if unit == self.unit or unit == \\\"SI\\\":\\n if is_norm:\\n try:\\n values = values / self.normalizations.get(\\\"ref\\\")\\n except:\\n raise NormError(\\n \\\"ERROR: Reference value not specified for normalization\\\"\\n )\\n elif unit in self.normalizations:\\n values = values / self.normalizations.get(unit)\\n else:\\n values = convert(values, self.unit, unit)\\n return values\",\n \"def domain(s):\\r\\n res = r_domain.findall(s)\\r\\n domain = (res and res[0]) or s\\r\\n return domain.lower()\",\n \"def _unify_domain(cls, *matrices):\\n domains = {matrix.domain for matrix in matrices}\\n if len(domains) == 1:\\n return matrices\\n domain = reduce(lambda x, y: x.unify(y), domains)\\n return tuple(matrix.convert_to(domain) for matrix in matrices)\",\n \"def parse_domain(self, domainfile):\\n\\n with open(domainfile) as dfile:\\n dfile_array = self._get_file_as_array(dfile)\\n #Deal with front/end define, problem, :domain\\n if dfile_array[0:4] != ['(', 'define', '(', 'domain']:\\n print('PARSING ERROR: Expected (define (domain ... at start of domain file')\\n sys.exit()\\n self.domain = dfile_array[4]\\n\\n dfile_array = dfile_array[6:-1]\\n opencounter = 0\\n keyword = ''\\n obj_list = []\\n is_obj_list = True\\n for word in dfile_array:\\n if word == '(':\\n opencounter += 1\\n elif word == ')':\\n opencounter -= 1\\n elif word.startswith(':'):\\n if word[1:] not in DFILE_KEYWORDS:\\n pass\\n elif keyword != 'requirements':\\n keyword = word[1:]\\n if opencounter == 0:\\n if keyword == 'action':\\n self.actions.append(obj_list)\\n obj_list = []\\n if keyword == 'types':\\n for element in obj_list:\\n self.types.setdefault('object', []).append(element)\\n self.type_list.add('object')\\n self.type_list.add(element)\\n obj_list = []\\n keyword = ''\\n\\n if keyword == 'requirements': #Requirements list\\n if word != ':requirements':\\n if not word.startswith(':'):\\n print('PARSING ERROR: Expected requirement to start with :')\\n sys.exit()\\n elif word[1:] not in DFILE_REQ_KEYWORDS:\\n print('WARNING: Unknown Rquierement ' + word[1:])\\n #print 'Requirements must only be: ' + str(DFILE_REQ_KEYWORDS)\\n #sys.exit()\\n else:\\n self.requirements.add(word[1:])\\n elif keyword == 'action':\\n obj_list.append(word)\\n elif not word.startswith(':'):\\n if keyword == 'types': #Typed list of objects\\n if is_obj_list:\\n if word == '-':\\n is_obj_list = False\\n else:\\n obj_list.append(word)\\n else:\\n #word is type\\n for element in obj_list:\\n if not word in self.type_list:\\n self.types.setdefault('object', []).append(word)\\n self.type_list.add(word)\\n self.types.setdefault(word, []).append(element)\\n self.type_list.add(element)\\n self.type_list.add(word)\\n is_obj_list = True\\n obj_list = []\\n elif keyword == 'constants': #Typed list of objects\\n if is_obj_list:\\n if word == '-':\\n is_obj_list = False\\n else:\\n obj_list.append(word)\\n else:\\n #word is type\\n for element in obj_list:\\n if word in self.type_list:\\n self.constants.setdefault(word, []).append(element)\\n #self.object_list.add(element)\\n else:\\n print(self.type_list)\\n print(\\\"ERROR unknown type \\\" + word)\\n sys.exit()\\n is_obj_list = True\\n obj_list = []\\n elif keyword == 'predicates' or keyword == 'private': #Internally typed predicates\\n if word == ')':\\n if keyword == 'private':\\n #print \\\"...skip agent: \\\" + str(obj_list[:3])\\n obj_list = obj_list[3:]\\n keyword = 'predicates'\\n if len(obj_list) == 0:\\n #print \\\"...skip )\\\"\\n continue\\n p_name = obj_list[0]\\n #print \\\"parse predicate: \\\" + p_name + \\\" \\\" + str(obj_list)\\n pred_list = self._parse_name_type_pairs(obj_list[1:],self.type_list)\\n self.predicates.append(Predicate(p_name, pred_list, True, False))\\n obj_list = []\\n elif word != '(':\\n obj_list.append(word)\\n elif keyword == 'functions': #functions\\n if word == ')':\\n p_name = obj_list[0]\\n if obj_list[0] == '-':\\n obj_list = obj_list[2:]\\n #print \\\"function: \\\" + word + \\\" - \\\" + str(obj_list)\\n self.functions.append(Function(obj_list))\\n obj_list = []\\n elif word != '(':\\n obj_list.append(word)\\n\\n #Work on the actions\\n new_actions = []\\n for action in self.actions:\\n if action[0] == '-':\\n action = action[2:]\\n act_name = action[1]\\n act = {}\\n action = action[2:]\\n keyword = ''\\n for word in action:\\n if word.startswith(':'):\\n keyword = word[1:]\\n else:\\n act.setdefault(keyword, []).append(word)\\n self.agent_types.add(act.get('agent')[2])\\n agent = self._parse_name_type_pairs(act.get('agent'),self.type_list)\\n param_list = agent + self._parse_name_type_pairs(act.get('parameters')[1:-1],self.type_list)\\n up_params = Predicate('', param_list, True, False)\\n pre_list = self._parse_unground_propositions(act.get('precondition'))\\n eff_list = self._parse_unground_propositions(act.get('effect'))\\n new_act = Action(act_name, up_params, pre_list, eff_list)\\n\\n new_actions.append(new_act)\\n self.actions = new_actions\",\n \"def transform():\",\n \"def __init__(self, domain, range):\\n self.domain = domain.cloneSpace()\\n self.range = range.cloneSpace()\",\n \"def dict_of_domains(fc):\\r\\n # need to find root database (GDB or SDE)\\r\\n db_root = os.path.dirname(fc)\\r\\n while db_root[-4:].lower() != '.gdb' and db_root[-4:].lower() != '.sde':\\r\\n old_db_root = db_root # protect against infinite loop\\r\\n db_root = os.path.dirname(db_root)\\r\\n if old_db_root == db_root: # protect against infinite loop\\r\\n break\\r\\n arcpy.AddMessage(\\\"Retrieving Domains from \\\" + str(db_root))\\r\\n return {domain.name: domain.codedValues for domain in arcpy.da.ListDomains(db_root)}\",\n \"def from_ExpressionDomain(K1, a, K0):\\n return K1.from_sympy(a.ex)\",\n \"def _domain_variants(domain):\\n parts = domain.split('.')\\n for i in range(len(parts), 1, -1):\\n yield \\\".\\\".join(parts[-i:])\",\n \"def replace_domain(address, old_domain, new_domain):\\n old_domain_pattern = r'' + old_domain + '$'\\n address = re.sub(old_domain_pattern, new_domain, address)\\n return address\",\n \"def _identify_domains(self):\\n\\n domains = [FEMDomain(TR3, MeshPart(self.mesh, labels=(0,)), self.media, self.labels)]\\n return domains\",\n \"def convert_to_model(self, *args):\\n rt_xref_service_data, *_ = args\\n return [RealtyTypeXRefService(**rt_xref_service) for rt_xref_service in rt_xref_service_data]\",\n \"def reduceScale(structure,scale,dim,prec=1E-2):\\n\\n structure.translate_sites(indices=range(structure.num_sites),\\n vector=-1*structure.sites[0].frac_coords+[0,0,.5])\\n\\n specs = structure.species\\n cart = [x.coords for x in structure.sites]\\n\\n lat = np.array(structure.lattice.as_dict()['matrix'])\\n if type(scale)==type(1) or type(scale)==type(1.0):\\n lat[0]*=scale\\n if dim>=2:\\n lat[1]*=scale\\n if dim==3:\\n lat[2]*=scale\\n else:\\n lat[0]*=scale[0]\\n lat[1]*=scale[1]\\n lat[2]*=scale[2] \\n\\n fracs = np.around(abs(np.array(np.linalg.solve(lat.T,np.array(cart).T).T)),4)\\n\\n specs = []\\n u_cart = []\\n i=0\\n n_fracs = []\\n for frac in fracs:\\n if not np.any([magni(np.around(frac,4)%1-np.around(x,4)%1)<prec for x in n_fracs]):\\n n_fracs.append(frac%1)\\n #if frac[0]<1 and frac[1]<1 and frac[2]<1:\\n \\n specs.append(structure.species[i])\\n u_cart.append(cart[i])\\n i+=1\\n \\n new_sites = []\\n i=0\\n for site in u_cart:\\n p = PeriodicSite(lattice = Lattice(lat),coords=site,\\n coords_are_cartesian=True,\\n species=specs[i])\\n new_sites.append(p)\\n i+=1\\n\\n new_struct = Structure.from_sites(new_sites)\\n\\n return(new_struct)\",\n \"def norm_it(startdate, enddate, dfin, dfname, l_back=1, toscale=True):\\n\\n # # subset\\n dfin = dfin.loc[startdate:enddate]\\n if toscale is True: \\n # # normalize\\n scaler = MinMaxScaler(feature_range=(0, 1))\\n # # add normalized to dataset\\n dfin[dfname+'_norm'] = scaler.fit_transform(dfin[\\n dfname].to_numpy().reshape(-1, 1))\\n # # lag\\n dfin[dfname+'_norm'+str(l_back)] = dfin[dfname+'_norm'].shift(l_back)\\n return dfin, scaler\\n else:\\n dfin[dfname+str(l_back)] = dfin[dfname].shift(l_back)\\n return dfin\",\n \"def generate_domain(domain):\\n\\n if domain.lower() in (\\\"greenland\\\", \\\"gris\\\", \\\"gris_ext\\\"):\\n pism_exec = \\\"pismr\\\"\\n elif domain.lower() in (\\\"og\\\"):\\n pism_exec = \\\"pismr -no_model_strip 0 -calving_wrap_around\\\"\\n elif domain.lower() in (\\\"hia\\\"):\\n x_min = -652200.0\\n x_max = -232600.0\\n y_min = -1263900.0\\n y_max = -943500.0\\n pism_exec = \\\"\\\"\\\"pismr -x_range {x_min},{x_max} -y_range {y_min},{y_max} -bootstrap\\\"\\\"\\\".format(\\n x_min=x_min, x_max=x_max, y_min=y_min, y_max=y_max\\n )\\n elif domain.lower() in (\\\"jakobshavn\\\", \\\"jib\\\"):\\n x_min = -280000.0\\n x_max = 320000.0\\n y_min = -2410000.0\\n y_max = -2020000.0\\n pism_exec = \\\"\\\"\\\"pismr -regional -x_range {x_min},{x_max} -y_range {y_min},{y_max} -bootstrap -regional.zero_gradient true -regional.no_model_strip 4.5\\\"\\\"\\\".format(\\n x_min=x_min, x_max=x_max, y_min=y_min, y_max=y_max\\n )\\n else:\\n print((\\\"Domain {} not recognized, exiting\\\".format(domain)))\\n import sys\\n\\n sys.exit(0)\\n\\n return pism_exec\",\n \"def parse_domain(domain_str):\\n def floatify(latlon):\\n \\\"\\\"\\\" Turns a latlon string into a float \\\"\\\"\\\"\\n sign = -2. * (latlon[-1].lower() in ['s', 'w']) + 1\\n return float(latlon[:-1]) * sign\\n corners = domain_str.strip().split(',')\\n if not len(corners) == 4:\\n raise BadQuery(\\\"Expected four comma seperated values \\\"\\n \\\"defining the corners of the domain.\\\")\\n\\n is_lat = lambda x: x[-1].lower() in ['n', 's']\\n lats = filter(is_lat, corners)\\n if not len(lats) == 2:\\n raise BadQuery(\\\"Expected two latitudes (determined by \\\" +\\n \\\"values ending in 'N' or 'S'\\\")\\n is_lon = lambda x: x[-1].lower() in ['e', 'w']\\n lons = filter(is_lon, corners)\\n if not len(lons) == 2:\\n raise BadQuery(\\\"Expected two longitudes (determined by \\\" +\\n \\\"values ending in 'E' or 'W'\\\")\\n lats = np.array([floatify(x) for x in lats])\\n lons = np.array([floatify(x) for x in lons])\\n # make sure latitudes are in range.\\n if np.any(lats > 90.) or np.any(lats < -90):\\n raise BadQuery(\\\"Latitudes must be within -90 and 90, got %s\\\" %\\n ','.join(map(str, lats)))\\n # we let the user use either longitudes of 0 to 360\\n # or -180 to 180, then convert to nautical (-180 to 180).\\n if np.any(lons > 360.) or np.any(lons < -180.):\\n raise BadQuery(\\\"Longitudes must be within -180 and 360, got %s\\\" %\\n ','.join(map(str, lons)))\\n # make sure lons end up in -180 to 180.\\n lons = np.mod(lons + 180., 360.) - 180.\\n\\n lats = np.sort(lats)\\n # chose the order for lons that minimizes the\\n # angle. This means people can't make requests\\n # for more than 180 degrees at a time. But that seems\\n # more desirable than accidentally sending near global\\n # files when the dateline is crossed.\\n angular_diff = lambda x, y: np.mod(x - y, 360.)\\n if angular_diff(*lons) > angular_diff(*reversed(lons)):\\n lons = lons[::-1]\\n\\n domain = {'N': lats[1], 'S': lats[0],\\n 'W': lons[1], 'E': lons[0]}\\n return domain\",\n \"def convert_to(self, K):\\n if K == self.domain:\\n return self.copy()\\n\\n rep = self.rep\\n\\n # The DFM, DDM and SDM types do not do any implicit conversions so we\\n # manage switching between DDM and DFM here.\\n if rep.is_DFM and not DFM._supports_domain(K):\\n rep_K = rep.to_ddm().convert_to(K)\\n elif rep.is_DDM and DFM._supports_domain(K):\\n rep_K = rep.convert_to(K).to_dfm()\\n else:\\n rep_K = rep.convert_to(K)\\n\\n return self.from_rep(rep_K)\",\n \"def transform_series(obj):\\n vals = obj.values\\n return transform_array(vals)\",\n \"def read_fusion_domains(fus_path, wt_path, fusion_annot, min_domain_len=25, min_domain_frac=0.5):\\n # read in domains\\n domains = pd.read_csv(fus_path, sep='\\\\t')\\n wt_domains = pd.read_csv(wt_path, sep='\\\\t')\\n\\n # filter out transmembrane helix\\n is_tm = domains[\\\"Domain_ID\\\"].str.startswith('TMhelix')\\n domains = domains[~is_tm].copy()\\n\\n # filter out short domains\\n domain_len = domains['Protein_end'] - domains['Protein_start']\\n domains = domains[domain_len>=min_domain_len].copy()\\n\\n # merge in info about where the fusion happened\\n domains = pd.merge(domains, fusion_annot[['ID', 'CodonPos1', 'CodonPos2']].drop_duplicates('ID'), on='ID', how='left')\\n domains['origin of domain'] = \\\"3'_gene domain\\\"\\n domains.loc[domains['Protein_end']<=domains['CodonPos1'], 'origin of domain'] = \\\"5'_gene domain\\\"\\n\\n # separate out protein ID\\n domains['ID_five'] = domains['ID'].str.split('-', expand=True)[0]\\n domains['ID_three'] = domains['ID'].str.split('-', expand=True)[1].str.split(':', expand=True)[0]\\n\\n # check 5' domains\\n domains_five = domains[domains[\\\"origin of domain\\\"]==\\\"5'_gene domain\\\"].copy()\\n is_good_domain = []\\n for ix, row in domains_five.iterrows():\\n domain_end = row['Protein_end']\\n fus_codon_pos = row['CodonPos1']\\n if fus_codon_pos==domain_end:\\n pid = row['ID_five']\\n is_pid = wt_domains['PROT_ID']==pid\\n is_loc = (fus_codon_pos>=wt_domains['Protein_start']) & (fus_codon_pos<=wt_domains['Protein_end'])\\n tmp_df = wt_domains.loc[is_pid & is_loc,:]\\n if len(tmp_df)==0:\\n is_good_domain.append(True)\\n continue\\n else:\\n tmp = tmp_df.iloc[0]\\n frac_domain = (fus_codon_pos - tmp['Protein_start']) / (tmp['Protein_end'] - tmp['Protein_start'])\\n if frac_domain>=min_domain_frac:\\n is_good_domain.append(True)\\n else:\\n is_good_domain.append(False)\\n else:\\n is_good_domain.append(True)\\n domains_five = domains_five.loc[is_good_domain,:]\\n # check 3' domains\\n domains_three = domains[domains[\\\"origin of domain\\\"]==\\\"3'_gene domain\\\"].copy()\\n is_good_domain = []\\n for ix, row in domains_three.iterrows():\\n domain_start = row['Protein_start']\\n fus_codon_pos = row['CodonPos2']\\n if fus_codon_pos==domain_start:\\n pid = row['ID_three']\\n is_pid = wt_domains['PROT_ID']==pid\\n is_loc = (fus_codon_pos>=wt_domains['Protein_start']) & (fus_codon_pos<=wt_domains['Protein_end'])\\n tmp_df = wt_domains.loc[is_pid & is_loc,:]\\n if len(tmp_df)==0:\\n is_good_domain.append(True)\\n continue\\n else:\\n tmp = tmp_df.iloc[0]\\n frac_domain = (tmp['Protein_end']-fus_codon_pos) / (tmp['Protein_end'] - tmp['Protein_start'])\\n if frac_domain>=min_domain_frac:\\n is_good_domain.append(True)\\n else:\\n is_good_domain.append(False)\\n else:\\n is_good_domain.append(True)\\n domains_three = domains_three.loc[is_good_domain,:]\\n # append results\\n domains = pd.concat([domains_five, domains_three])\\n\\n # aggregate all domains for one fusion into a single line\\n domains = domains[~domains['Domain_name'].isnull()]\\n domains_agg = domains.groupby(['ID', 'origin of domain'])['Domain_name'].agg(lambda x: ','.join(x))\\n domains_agg = domains_agg.reset_index()\\n domains_agg = domains_agg.pivot(index='ID', columns='origin of domain', values='Domain_name')\\n domains_agg = domains_agg.reset_index()\\n\\n return domains_agg\",\n \"def domain( self ):\\n raise NotImplementedError(\\\"domain\\\")\",\n \"def shapely_to_cf(geometries: xr.DataArray | Sequence, grid_mapping: str | None = None):\\n # Get all types to call the appropriate translation function.\\n types = {\\n geom.item().geom_type if isinstance(geom, xr.DataArray) else geom.geom_type\\n for geom in geometries\\n }\\n if types.issubset({\\\"Point\\\", \\\"MultiPoint\\\"}):\\n ds = points_to_cf(geometries)\\n elif types.issubset({\\\"Polygon\\\", \\\"MultiPolygon\\\"}) or types.issubset(\\n {\\\"LineString\\\", \\\"MultiLineString\\\"}\\n ):\\n raise NotImplementedError(\\\"Only point geometries conversion is implemented.\\\")\\n else:\\n raise ValueError(\\n f\\\"Mixed geometry types are not supported in CF-compliant datasets. Got {types}\\\"\\n )\\n\\n # Special treatment of selected grid mappings\\n if grid_mapping == \\\"longitude_latitude\\\":\\n # Special case for longitude_latitude grid mapping\\n ds = ds.rename(crd_x=\\\"lon\\\", crd_y=\\\"lat\\\")\\n ds.lon.attrs.update(units=\\\"degrees_east\\\", standard_name=\\\"longitude\\\")\\n ds.lat.attrs.update(units=\\\"degrees_north\\\", standard_name=\\\"latitude\\\")\\n ds.geometry_container.attrs.update(coordinates=\\\"lon lat\\\")\\n ds.x.attrs.update(units=\\\"degrees_east\\\", standard_name=\\\"longitude\\\")\\n ds.y.attrs.update(units=\\\"degrees_north\\\", standard_name=\\\"latitude\\\")\\n elif grid_mapping is not None:\\n raise NotImplementedError(\\n f\\\"Only grid mapping longitude_latitude is implemented. Got {grid_mapping}.\\\"\\n )\\n\\n return ds\",\n \"def get_domains(graph: Graph, property_to_id: Dict[str, int], entity_type_to_id: Dict[str, int]) -> Dict[int, int]:\\n # dictionary pointing from object property id to an entity type id\\n domains = {}\\n\\n # add all domain triples for which the subject is an object property and the object is an entity type\\n for subject, predicate, object in graph.triples((None, RDFS.domain, None)):\\n if subject in property_to_id and object in entity_type_to_id:\\n domains[property_to_id[subject]] = entity_type_to_id[object]\\n\\n return domains\",\n \"def get_domain(self, rel_name):\\n return self._declaration[rel_name].domain_type\",\n \"def make_companies():\\n logging.info(\\\"Making CH\\\")\\n companies_address = get_address()\\n companies_sector = get_sector()\\n\\n companies = (\\n companies_address[[\\\"company_number\\\", \\\"postcode\\\"]]\\n .merge(\\n companies_sector.query(\\\"rank==1\\\")[[\\\"company_number\\\", \\\"SIC4_code\\\"]],\\n on=\\\"company_number\\\",\\n )\\n .assign(division=lambda x: [c[:2] for c in x[\\\"SIC4_code\\\"]])\\n .assign(division_name=lambda x: x[\\\"division\\\"].map(_DIV_NAME_LOOKUP))\\n .merge(nspl, left_on=\\\"postcode\\\", right_on=\\\"pcds\\\")\\n )\\n\\n return companies\",\n \"def resolveOriginalDomains():\\n print('[+] Populating Domain Name Resolution for later check ')\\n\\n try:\\n for domain in domains:\\n response = dns.resolver.query(domain)\\n d = Domain_Poison_Check(domain)\\n print('[+] Domain: %s' % domain)\\n for record in response:\\n print(' |____> maps to %s.' % (record.address))\\n d.pushAddr(record)\\n check_domain_poison_results.append(d)\\n return time.time()\\n except Exception as err:\\n print('[+] Exception: %s' % err)\\n traceback.print_exc()\\n return time.time()\",\n \"def change_domain(self, new_domain):\\n self.domain=new_domain\\n for pm in self._maps.values():\\n pm.change_domain(new_domain)\",\n \"def get_input_domains():\\n df = pandas.read_excel(\\\"AutoScrapy/files/EBE21 - Top 100 Onlineshops to scrapp.ods\\\", engine=\\\"odf\\\")\\n list_of_addresses = df['Domain'].to_list()\\n list_of_addresses = [(\\\"http://\\\" + address) for address in list_of_addresses]\\n print(list_of_addresses)\\n return list_of_addresses\",\n \"def ground_to_exact(f):\\n return f.set_domain(f.dom.get_exact())\",\n \"def createDomain(self):\\n if not self.rank:\\n logging.info('Creating domain')\\n\\n if 'box' in self.pargs:\\n self.lmp.command('region domain block ' + ('{} ' * len(self.pargs['box'])).format(*self.pargs['box']) + ' units box volume_limit 1e-20')\\n elif 'cylinder' in self.pargs:\\n self.lmp.command('region domain cylinder ' + ('{} ' * len(self.pargs['cylinder'])).format(*self.pargs['cylinder']) + ' units box volume_limit 1e-20') \\n\\n self.lmp.command('create_box {} domain'.format(self.pargs['nSS']))\",\n \"def __learn__(self):\\n\\n return domains # including 3 values, (begin, end, key)\",\n \"def cf_to_shapely(ds: xr.Dataset):\\n geom_type = ds.geometry_container.attrs[\\\"geometry_type\\\"]\\n if geom_type == \\\"point\\\":\\n geometries = cf_to_points(ds)\\n elif geom_type in [\\\"line\\\", \\\"polygon\\\"]:\\n raise NotImplementedError(\\\"Only point geometries conversion is implemented.\\\")\\n else:\\n raise ValueError(\\n f\\\"Valid CF geometry types are 'point', 'line' and 'polygon'. Got {geom_type}\\\"\\n )\\n\\n return geometries.rename(\\\"geometry\\\")\",\n \"def domain_region_grow(counts, ndomain):\\n \\n # make it so maximum value is 100,000,000 (fits well inside int32)\\n norm = 10000.0 / counts.max()\\n wts = np.square(norm * counts).astype(int32)\\n# wts = ((norm * counts)**1.0).astype(int32)\\n \\n tot = wts.sum(dtype=int64)\\n \\n sum_per_domain = tot//ndomain + 1\\n \\n doms = domain_regiongrow(wts, sum_per_domain)\\n return doms\",\n \"def fixDomains(self, energyMin, energyMax, domainToFix):\\n\\n return 0\",\n \"def fixDomains(self, energyMin, energyMax, domainToFix):\\n\\n return 0\",\n \"def convert(osis_refs):\\n is_string = False\\n if isinstance(osis_refs, str):\\n osis_refs = [osis_refs]\\n is_string = True\\n\\n usfm_refs = []\\n\\n for ref in osis_refs:\\n parts = ref.split('.')\\n\\n # If the book part is different, convert otherwise just uppercase it\\n if parts[0].lower() in BOOKS.keys():\\n parts[0] = BOOKS[parts[0].lower()]\\n else:\\n parts[0] = parts[0].upper()\\n\\n usfm_refs.append('.'.join(parts))\\n\\n return usfm_refs if is_string is False else usfm_refs[0]\",\n \"def _adddomain(self, domain: Domain):\\n\\n domain = copy.deepcopy(domain)\\n if self.model is not None:\\n # Check that model and domain are compatible\\n self._validate_model_domain(self.model, domain)\\n\\n # Add in domain\\n self.domain = domain\\n\\n # Setup base namelists\\n self._set_base_namelists()\\n else:\\n self.domain = domain\",\n \"def merge_new_domain (cls, base, nffg, log=logging.getLogger(\\\"MERGE\\\")):\\n # Get new domain name\\n domain = cls.detect_domains(nffg=nffg)\\n if len(domain) == 0:\\n log.error(\\\"No domain detected in new %s!\\\" % nffg)\\n return\\n if len(domain) > 1:\\n log.warning(\\\"Multiple domain name detected in new %s!\\\" % nffg)\\n return\\n # Copy infras\\n log.debug(\\\"Merge domain: %s resource info into %s...\\\" % (domain.pop(),\\n base.id))\\n # Check if the infra with given id is already exist in the base NFFG\\n for infra in nffg.infras:\\n if infra.id not in base:\\n c_infra = base.add_infra(infra=deepcopy(infra))\\n log.debug(\\\"Copy infra node: %s\\\" % c_infra)\\n else:\\n log.warning(\\\"Infra node: %s does already exist in %s. Skip adding...\\\" %\\n (infra, base))\\n # Copy NFs\\n for nf in nffg.nfs:\\n if nf.id not in base:\\n c_nf = base.add_nf(nf=deepcopy(nf))\\n log.debug(\\\"Copy NF node: %s\\\" % c_nf)\\n else:\\n log.warning(\\\"NF node: %s does already exist in %s. Skip adding...\\\" %\\n (nf, base))\\n # Copy SAPs\\n for sap_id in [s.id for s in nffg.saps]:\\n if sap_id in [s.id for s in base.saps]:\\n # Found inter-domain SAP\\n log.debug(\\\"Found Inter-domain SAP: %s\\\" % sap_id)\\n # Search outgoing links from SAP, should be only one\\n b_links = [l for u, v, l in base.real_out_edges_iter(sap_id)]\\n if len(b_links) < 1:\\n log.warning(\\n \\\"SAP is not connected to any node! Maybe you forgot to call \\\"\\n \\\"duplicate_static_links?\\\")\\n return\\n elif 1 < len(b_links):\\n log.warning(\\n \\\"Inter-domain SAP should have one and only one connection to the \\\"\\n \\\"domain! Using only the first connection.\\\")\\n continue\\n # Get inter-domain port in base NFFG\\n domain_port_dov = b_links[0].dst\\n sap_port_dov = b_links[0].src\\n log.debug(\\\"Found inter-domain port: %s\\\" % domain_port_dov)\\n # Search outgoing links from SAP, should be only one\\n n_links = [l for u, v, l in nffg.real_out_edges_iter(sap_id)]\\n if len(n_links) < 1:\\n log.warning(\\n \\\"SAP is not connected to any node! Maybe you forgot to call \\\"\\n \\\"duplicate_static_links?\\\")\\n return\\n elif 1 < len(n_links):\\n log.warning(\\n \\\"Inter-domain SAP should have one and only one connection to the \\\"\\n \\\"domain! Using only the first connection.\\\")\\n continue\\n # Get port and Infra id's in nffg NFFG\\n p_id = n_links[0].dst.id\\n n_id = n_links[0].dst.node.id\\n # Get the inter-domain port from already copied Infra\\n domain_port_nffg = base.network.node[n_id].ports[p_id]\\n sap_port_nffg = n_links[0].src\\n log.debug(\\\"Found inter-domain port: %s\\\" % domain_port_nffg)\\n\\n # # If the two resource value does not match\\n # if sap_port_dov.delay != sap_port_nffg.delay:\\n # if sap_port_dov.delay is None:\\n # # If first is None the other can not be None\\n # s_delay = sap_port_nffg.delay\\n # elif sap_port_nffg.delay is None:\\n # # If second is None the other can not be None\\n # s_delay = sap_port_dov.delay\\n # else:\\n # # Both values are valid, but different\\n # s_delay = max(sap_port_dov.delay, sap_port_nffg.delay)\\n # log.warning(\\n # \\\"Inter-domain delay values (%s, %s) are set but do not match!\\\"\\n # \\\" Use max: %s\\\" % (sap_port_dov.delay, sap_port_nffg.delay,\\n # s_delay))\\n # else:\\n # # Both value match: ether valid values or Nones --> choose first\\n # value\\n # s_delay = sap_port_dov.delay\\n #\\n # # If the two resource value does not match\\n # if sap_port_dov.bandwidth != sap_port_nffg.bandwidth:\\n # if sap_port_dov.bandwidth is None:\\n # # If first is None the other can not be None\\n # s_bandwidth = sap_port_nffg.bandwidth\\n # elif sap_port_nffg.bandwidth is None:\\n # # If second is None the other can not be None\\n # s_bandwidth = sap_port_dov.bandwidth\\n # else:\\n # # Both values are valid, but different\\n # s_bandwidth = min(sap_port_dov.bandwidth,\\n # sap_port_nffg.bandwidth)\\n # log.warning(\\n # \\\"Inter-domain bandwidth values (%s, %s) are set but do not\\n # match!\\\"\\n # \\\" Use min: %s\\\" % (sap_port_dov.bandwidth,\\n # sap_port_nffg.bandwidth, s_bandwidth))\\n # else:\\n # # Both value match: ether valid values or Nones --> choose first\\n # value\\n # s_bandwidth = sap_port_dov.bandwidth\\n #\\n # log.debug(\\\"Detected inter-domain resource values: delay: %s, \\\"\\n # \\\"bandwidth: %s\\\" % (s_delay, s_bandwidth))\\n\\n # Copy inter-domain port properties/values for redundant storing\\n if len(domain_port_nffg.properties) > 0:\\n domain_port_dov.properties.update(domain_port_nffg.properties)\\n log.debug(\\\"Copy inter-domain port properties: %s\\\" %\\n domain_port_dov.properties)\\n elif len(domain_port_dov.properties) > 0:\\n domain_port_nffg.properties.update(domain_port_dov.properties)\\n log.debug(\\\"Copy inter-domain port properties: %s\\\" %\\n domain_port_nffg.properties)\\n # Ensure to add sap tag to inter domain ports\\n if 'sap' not in domain_port_dov.properties:\\n domain_port_dov.add_property(\\\"sap\\\", sap_id)\\n if 'sap' not in domain_port_nffg.properties:\\n domain_port_nffg.add_property(\\\"sap\\\", sap_id)\\n # Signal Inter-domain port type\\n domain_port_dov.add_property(\\\"type\\\", \\\"inter-domain\\\")\\n domain_port_nffg.add_property(\\\"type\\\", \\\"inter-domain\\\")\\n\\n # Copy SAP port values into the infra ports\\n domain_port_dov.name = sap_port_dov.name\\n domain_port_dov.sap = sap_port_dov.sap\\n domain_port_dov.capability = sap_port_dov.capability\\n domain_port_dov.technology = sap_port_dov.technology\\n domain_port_dov.delay = sap_port_dov.delay\\n domain_port_dov.bandwidth = sap_port_dov.bandwidth\\n domain_port_dov.cost = sap_port_dov.cost\\n domain_port_dov.controller = sap_port_dov.controller\\n domain_port_dov.orchestrator = sap_port_dov.orchestrator\\n domain_port_dov.l2 = sap_port_dov.l2\\n domain_port_dov.l4 = sap_port_dov.l4\\n for l3 in sap_port_dov.l3:\\n domain_port_dov.l3.append(l3.copy())\\n domain_port_dov.metadata.update(sap_port_dov.metadata)\\n\\n domain_port_nffg.name = sap_port_nffg.name\\n domain_port_nffg.sap = sap_port_nffg.sap\\n domain_port_nffg.capability = sap_port_nffg.capability\\n domain_port_nffg.technology = sap_port_nffg.technology\\n domain_port_nffg.delay = sap_port_nffg.delay\\n domain_port_nffg.bandwidth = sap_port_nffg.bandwidth\\n domain_port_nffg.cost = sap_port_nffg.cost\\n domain_port_nffg.controller = sap_port_nffg.controller\\n domain_port_nffg.orchestrator = sap_port_nffg.orchestrator\\n domain_port_nffg.l2 = sap_port_nffg.l2\\n domain_port_nffg.l4 = sap_port_nffg.l4\\n for l3 in sap_port_nffg.l3:\\n domain_port_nffg.l3.append(l3.copy())\\n domain_port_nffg.metadata.update(sap_port_nffg.metadata)\\n\\n # Delete both inter-domain SAP and links connected to them\\n base.del_node(sap_id)\\n nffg.del_node(sap_id)\\n\\n # Add the inter-domain links for both ways\\n l1, l2 = base.add_undirected_link(\\n p1p2id=\\\"inter-domain-link-%s\\\" % sap_id,\\n p2p1id=\\\"inter-domain-link-%s-back\\\" % sap_id,\\n port1=domain_port_dov,\\n port2=domain_port_nffg)\\n # Set delay/bandwidth values for outgoing link port1 -> port2\\n l1.delay = domain_port_dov.delay\\n l1.bandwidth = domain_port_dov.bandwidth\\n # Set delay/bandwidth values for outgoing link port2 -> port2\\n l2.delay = domain_port_nffg.delay\\n l2.bandwidth = domain_port_nffg.bandwidth\\n\\n else:\\n # Normal SAP --> copy SAP\\n c_sap = base.add_sap(sap_obj=deepcopy(nffg.network.node[sap_id]))\\n log.debug(\\\"Copy SAP: %s\\\" % c_sap)\\n # Copy remaining links which should be valid\\n for u, v, link in nffg.network.edges_iter(data=True):\\n src_port = base.network.node[u].ports[link.src.id]\\n dst_port = base.network.node[v].ports[link.dst.id]\\n c_link = deepcopy(link)\\n c_link.src = src_port\\n c_link.dst = dst_port\\n base.add_link(src_port=src_port, dst_port=dst_port, link=c_link)\\n log.debug(\\\"Copy Link: %s\\\" % c_link)\\n log.debug(\\\"Domain merging has been finished!\\\")\\n # Return the updated NFFG\\n return base\",\n \"def setup_domains():\\n sdb = boto.connect_sdb()\\n sdb.create_domain(\\\"mls_domain\\\")\",\n \"def domain(self, domain=None):\\n\\n return self.domain_class(apiobj=self, domainname=domain)\",\n \"def transform(self, series_of_articles):\\n # sklearn vectorizer needs an iterable list/series,\\n # so convert to list of just a single string\\n if isinstance(series_of_articles, str):\\n series_of_articles = [series_of_articles] # Convert to list\\n result = self.vectorizer.transform(series_of_articles)\\n dtm = self._post_process(result)\\n return dtm\",\n \"def translateValues(nadPoints):\\n with arcpy.da.UpdateCursor(nadPoints,\\n ['StN_PreDir', 'StN_PosDir', 'StN_PosTyp', 'County']) as cursor:\\n for row in cursor:\\n row[0] = directionDomain.get(row[0], None)\\n row[1] = directionDomain.get(row[1], None)\\n row[2] = streetDomain.get(row[2], None)\\n row[3] = countyFipsDomain.get(row[3], None)\\n cursor.updateRow(row)\",\n \"def transform_atrf2014_to_gda2020(x, y, z, epoch_from, vcv=None):\\n return conform14(x, y, z, epoch_from, atrf2014_to_gda2020, vcv=vcv)\",\n \"def _scipy_transform(fqdn, value):\\n import scipy\\n return _package_transform(scipy, fqdn, value)\",\n \"def find_domain_range(record):\\n response = {\\\"domain\\\": [], \\\"range\\\": []}\\n if \\\"http://schema.org/domainIncludes\\\" in record:\\n if isinstance(record[\\\"http://schema.org/domainIncludes\\\"], dict):\\n response[\\\"domain\\\"] = [record[\\\"http://schema.org/domainIncludes\\\"][\\\"@id\\\"]]\\n elif isinstance(record[\\\"http://schema.org/domainIncludes\\\"], list):\\n response[\\\"domain\\\"] = [\\n _item[\\\"@id\\\"] for _item in record[\\\"http://schema.org/domainIncludes\\\"]\\n ]\\n if \\\"http://schema.org/rangeIncludes\\\" in record:\\n if isinstance(record[\\\"http://schema.org/rangeIncludes\\\"], dict):\\n response[\\\"range\\\"] = [record[\\\"http://schema.org/rangeIncludes\\\"][\\\"@id\\\"]]\\n elif isinstance(record[\\\"http://schema.org/rangeIncludes\\\"], list):\\n response[\\\"range\\\"] = [\\n _item[\\\"@id\\\"] for _item in record[\\\"http://schema.org/rangeIncludes\\\"]\\n ]\\n return (response[\\\"domain\\\"], response[\\\"range\\\"])\",\n \"def __init__(self, cfg):\\n self.domains = []\\n\\n # process domains in order\\n for i in range(1, len(cfg)+1):\\n this_dom = cfg[str(i)]\\n par_dom = self.domains[this_dom['parent_id']-1] if 'parent_id' in this_dom else None\\n self.domains.append(WPSDomainLCC(i, this_dom, par_dom))\",\n \"def convert_to_model(self, *args):\\n state_xref_service_data, *_ = args\\n return [StateXRefService(**state_xref_service) for state_xref_service in state_xref_service_data]\",\n \"def transform_gda2020_to_atrf2014(x, y, z, epoch_to, vcv=None):\\n return conform14(x, y, z, epoch_to, -atrf2014_to_gda2020, vcv=vcv)\",\n \"def get_technique_to_domain():\\n global technique_to_domain\\n \\n if not technique_to_domain:\\n technique_to_domain = stixhelpers.get_technique_id_domain_map(get_ms())\\n \\n return technique_to_domain\",\n \"def update_domain():\\n\\n for e in Expr.search() + User.search(): e.set_tld(config.server_name)\",\n \"def get_feature_domain_dict(self):\\n feature_domain_dict = {}\\n for feature_index in range(len(self.train_examples[0])):\\n domain = set([example[feature_index] for example in self.train_examples])\\n feature_domain_dict[self.features[feature_index]] = domain\\n\\n return feature_domain_dict\",\n \"def _from_sequence(cls, scalars):\\n return cls(pa.array(scalars))\",\n \"def get_storage_domains(cohesity_client):\\n storage_domain_list = cohesity_client.view_boxes.get_view_boxes()\\n for domain in storage_domain_list:\\n exported_res_dict[\\\"Storage Domains\\\"].append(domain.name)\\n return storage_domain_list\",\n \"def normalize(feats, train_nid, dtype=np.float32):\\n train_feats = feats[train_nid]\\n scaler = preprocessing.StandardScaler()\\n scaler.fit(train_feats)\\n feats = scaler.transform(feats)\\n return feats.astype(dtype)\",\n \"def normaliseScores(essays):\\n normalised_scores = []\\n for index, essay in essays.iterrows():\\n score = essay['domain1_score']\\n # essay_set refers to the prompt(topic) of the essay\\n essay_set = essay['essay_set']\\n # Perform min-max normalization on the scores to get range in [0-1]\\n normalised_score = (score - MIN_SCORES[essay_set]) / (MAX_SCORES[essay_set] - MIN_SCORES[essay_set])\\n normalised_scores.append(normalised_score)\\n return np.array(normalised_scores)\",\n \"def transform_geometries(datasource, src_epsg, dst_epsg):\\n # Part 1\\n src_srs = osr.SpatialReference()\\n src_srs.ImportFromEPSG(src_epsg)\\n dst_srs = osr.SpatialReference()\\n dst_srs.ImportFromEPSG(dst_epsg)\\n transformation = osr.CoordinateTransformation(src_srs, dst_srs)\\n layer = datasource.GetLayerByIndex(0)\\n \\n # Part 2\\n geoms = []\\n layer.ResetReading()\\n for feature in layer:\\n geom = feature.GetGeometryRef().Clone()\\n geom.Transform(transformation)\\n geoms.append(geom)\\n return geoms\",\n \"def catalogConvert():\\n ret = libxml2mod.xmlCatalogConvert()\\n return ret\",\n \"def domains(self, domains):\\n\\n self._domains = domains\",\n \"def convert(self, chain, agebins=None):\\n cols = [\\\"av\\\", \\\"logzsol\\\"]\\n sfh, sfh_label = construct_sfh_measure(chain, agebins)\\n niter = len(sfh[0])\\n cols += sfh_label\\n dt = np.dtype([(c, np.float) for c in cols])\\n params = np.zeros(niter, dtype=dt)\\n\\n for c in cols:\\n if c in chain.dtype.names:\\n params[c] = np.squeeze(chain[c])\\n\\n # --- dust attenuation\\n params[\\\"av\\\"] = np.squeeze(1.086 * chain[\\\"dust2\\\"])\\n\\n # --- stellar ---\\n for i, c in enumerate(sfh_label):\\n params[c] = np.squeeze(sfh[i])\\n return params\",\n \"def get_schaefer2018_yeo(scale, data_dir=None):\\n\\n # substring in schaefer parcel names indiciating yeo network affiliation\\n schaefer_yeo = dict(\\n Vis='visual',\\n SomMot='somatomotor',\\n DorsAttn='dorsal attention',\\n SalVentAttn='ventral attention',\\n Limbic='limbic',\\n Cont='frontoparietal',\\n Default='default mode'\\n )\\n\\n # get requested annotation files\\n schaefer = nndata.fetch_schaefer2018('fsaverage5',\\n data_dir=data_dir)[scale]\\n\\n network_labels, parcel_labels = [], []\\n for hemi in ('lh', 'rh'):\\n # read in annotation file for given hemisphere\\n annot = getattr(schaefer, hemi)\\n labels, ctab, names = nib.freesurfer.read_annot(annot)\\n names = [m.decode() for m in names]\\n\\n # create empty arrays for vertex- and parcel-level affiliations\\n networks = np.zeros_like(labels)\\n parcels = np.zeros(len(names), dtype=int)\\n for n, parcel in enumerate(names):\\n for abbrev, net in schaefer_yeo.items():\\n # check which network this parcel belongs to by looking for the\\n # network substring in the parcel name and assign accordingly\\n if abbrev in parcel:\\n parcels[n] = YEO_CODES[net]\\n networks[labels == n] = YEO_CODES[net]\\n\\n # store network affiliations for this hemisphere\\n network_labels.append(networks)\\n parcel_labels.append(parcels)\\n\\n return NETWORKS(np.hstack(network_labels), np.hstack(parcel_labels))\",\n \"def get_domain(id):\\n return query(WEB_EXAMPLE_BASE + f\\\"/classical/domain/{id}\\\")\",\n \"def get_domain(self, classname):\\n subs = classname.split('-')\\n sc = ''.join(subs[:2]) # section, class\\n scsc = ''.join(subs) # section, class, subclass \\n if self.codes.has_key(sc):\\n return sc, self.codes[sc]\\n return scsc, self.codes.get(scsc, None)\",\n \"def as_domain(\\n domain: DomainLike, input_dim: Optional[IntLike]\\n) -> Tuple[DomainType, int]:\\n if input_dim is not None and input_dim < 1:\\n raise ValueError(\\n f\\\"Input dimension must be positive. Current value is ({input_dim}).\\\"\\n )\\n\\n if len(domain) != 2:\\n raise ValueError(f\\\"'domain' must be of length 2 ({len(domain)}).\\\")\\n\\n # Domain limits must have equal dimensions\\n if np.size(domain[0]) != np.size(domain[1]):\\n raise ValueError(\\n f\\\"Domain limits must be given either as scalars or arrays \\\"\\n f\\\"of equal dimension. Current sizes are ({np.size(domain[0])}) \\\"\\n f\\\"and ({np.size(domain[1])}).\\\"\\n )\\n\\n domain_dim = int(np.size(domain[0]))\\n\\n # Input dimension not given, infer it from the domain.\\n if input_dim is None:\\n input_dim = domain_dim\\n domain_a, domain_b = domain\\n\\n # Bounds are given as scalars: Expand domain limits.\\n elif domain_dim == 1:\\n domain_a = np.full((input_dim,), domain[0])\\n domain_b = np.full((input_dim,), domain[1])\\n\\n # Bounds are given as arrays\\n else:\\n # Size of domain and input dimension do not match\\n if input_dim != domain_dim:\\n raise ValueError(\\n f\\\"If domain limits are not scalars, their lengths \\\"\\n f\\\"must match the input dimension ({input_dim}).\\\"\\n )\\n domain_a = domain[0]\\n domain_b = domain[1]\\n\\n # convert bounds to 1D arrays if necessary\\n domain_a = np.atleast_1d(domain_a)\\n domain_b = np.atleast_1d(domain_b)\\n if domain_a.ndim > 1 or domain_b.ndim > 1:\\n raise ValueError(\\n f\\\"Upper ({domain_b.ndim}) or lower ({domain_a.ndim}) \\\"\\n f\\\"bounds have too many dimensions.\\\"\\n )\\n\\n # Make sure the domain is non-empty\\n if not np.all(domain_a < domain_b):\\n raise ValueError(\\n \\\"Integration domain must be non-empty. For example, some \\\"\\n \\\"lower bounds may be larger than their corresponding \\\"\\n \\\"upper bounds.\\\"\\n )\\n\\n domain = (domain_a, domain_b)\\n return domain, input_dim\",\n \"def _numpy_transform(fqdn, value):\\n import numpy\\n return _package_transform(numpy, fqdn, value)\",\n \"def _get_domain(self):\\n self.ensure_one()\\n domain = ['|', ('active', '=', True), ('active', '=', False)]\\n # Check active\\n if self.active == 'true':\\n domain += [('active', '=', True)]\\n elif self.active == 'false':\\n domain += [('active', '=', False)]\\n # Check partner type\\n if self.partner_type == 'customer_or_supplier':\\n domain += ['|', ('customer', '=', True), ('supplier', '=', True)]\\n elif self.partner_type == 'customer_and_supplier':\\n domain += [('customer', '=', True), ('supplier', '=', True)]\\n elif self.partner_type == 'customer':\\n domain += [('customer', '=', True)]\\n elif self.partner_type == 'supplier':\\n domain += [('supplier', '=', True)]\\n # Check category\\n if self.category_ids:\\n domain += [('category_id', 'in', self.category_ids.ids)]\\n return domain\",\n \"def test_Synthase_serialisation(synthase, domains, tmp_path):\\n result_dict = {\\n \\\"header\\\": \\\"test\\\",\\n \\\"sequence\\\": \\\"A\\\" * 200,\\n \\\"domains\\\": [\\n {\\n \\\"start\\\": 1,\\n \\\"end\\\": 90,\\n \\\"type\\\": \\\"KS\\\",\\n \\\"domain\\\": \\\"PKS_KS\\\",\\n \\\"evalue\\\": None,\\n \\\"bitscore\\\": None,\\n \\\"pssm\\\": None,\\n \\\"superfamily\\\": None,\\n \\\"accession\\\": None,\\n },\\n {\\n \\\"start\\\": 100,\\n \\\"end\\\": 200,\\n \\\"type\\\": \\\"AT\\\",\\n \\\"domain\\\": \\\"PKS_AT\\\",\\n \\\"evalue\\\": None,\\n \\\"bitscore\\\": None,\\n \\\"pssm\\\": None,\\n \\\"superfamily\\\": None,\\n \\\"accession\\\": None,\\n },\\n ],\\n \\\"classification\\\": []\\n }\\n\\n assert synthase.to_dict() == result_dict\\n\\n json_file = tmp_path / \\\"json\\\"\\n json_file.write_text(synthase.to_json())\\n\\n with json_file.open() as js:\\n from_json = Synthase.from_json(js)\\n\\n assert from_json.header == \\\"test\\\"\\n assert from_json.sequence == \\\"A\\\" * 200\\n assert from_json.domains == [domains[0], domains[2]]\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.5323861","0.5196539","0.5191682","0.5174157","0.5057873","0.50186723","0.50181675","0.50177693","0.5009062","0.49185956","0.49113917","0.4876111","0.48465654","0.48391303","0.48309028","0.48160523","0.48104942","0.48038787","0.47553554","0.47521913","0.47490832","0.47213486","0.4720635","0.47030315","0.46894348","0.46744937","0.46522427","0.4642893","0.4628759","0.46279395","0.45869565","0.45822182","0.45781547","0.456119","0.45571175","0.45481655","0.45457375","0.45228675","0.45168248","0.4516644","0.44952947","0.44906297","0.44888037","0.4466618","0.4462686","0.44608626","0.4453564","0.44493926","0.4425691","0.44124705","0.4406465","0.4403211","0.44002366","0.43854263","0.4370997","0.4366409","0.43583685","0.43529704","0.4348016","0.43390635","0.43239483","0.4310939","0.43106982","0.43099666","0.43097547","0.43096763","0.4307653","0.4303357","0.4303357","0.42981085","0.4295782","0.4295773","0.4280262","0.4279888","0.42792562","0.42770967","0.426897","0.4265695","0.4256696","0.42551246","0.4253894","0.42511585","0.42476526","0.42437142","0.42250982","0.42174116","0.421708","0.42120048","0.4211366","0.42009977","0.41975066","0.4196735","0.41945702","0.41887376","0.41816244","0.4176389","0.41736796","0.41723612","0.41679725","0.41670468"],"string":"[\n \"0.5323861\",\n \"0.5196539\",\n \"0.5191682\",\n \"0.5174157\",\n \"0.5057873\",\n \"0.50186723\",\n \"0.50181675\",\n \"0.50177693\",\n \"0.5009062\",\n \"0.49185956\",\n \"0.49113917\",\n \"0.4876111\",\n \"0.48465654\",\n \"0.48391303\",\n \"0.48309028\",\n \"0.48160523\",\n \"0.48104942\",\n \"0.48038787\",\n \"0.47553554\",\n \"0.47521913\",\n \"0.47490832\",\n \"0.47213486\",\n \"0.4720635\",\n \"0.47030315\",\n \"0.46894348\",\n \"0.46744937\",\n \"0.46522427\",\n \"0.4642893\",\n \"0.4628759\",\n \"0.46279395\",\n \"0.45869565\",\n \"0.45822182\",\n \"0.45781547\",\n \"0.456119\",\n \"0.45571175\",\n \"0.45481655\",\n \"0.45457375\",\n \"0.45228675\",\n \"0.45168248\",\n \"0.4516644\",\n \"0.44952947\",\n \"0.44906297\",\n \"0.44888037\",\n \"0.4466618\",\n \"0.4462686\",\n \"0.44608626\",\n \"0.4453564\",\n \"0.44493926\",\n \"0.4425691\",\n \"0.44124705\",\n \"0.4406465\",\n \"0.4403211\",\n \"0.44002366\",\n \"0.43854263\",\n \"0.4370997\",\n \"0.4366409\",\n \"0.43583685\",\n \"0.43529704\",\n \"0.4348016\",\n \"0.43390635\",\n \"0.43239483\",\n \"0.4310939\",\n \"0.43106982\",\n \"0.43099666\",\n \"0.43097547\",\n \"0.43096763\",\n \"0.4307653\",\n \"0.4303357\",\n \"0.4303357\",\n \"0.42981085\",\n \"0.4295782\",\n \"0.4295773\",\n \"0.4280262\",\n \"0.4279888\",\n \"0.42792562\",\n \"0.42770967\",\n \"0.426897\",\n \"0.4265695\",\n \"0.4256696\",\n \"0.42551246\",\n \"0.4253894\",\n \"0.42511585\",\n \"0.42476526\",\n \"0.42437142\",\n \"0.42250982\",\n \"0.42174116\",\n \"0.421708\",\n \"0.42120048\",\n \"0.4211366\",\n \"0.42009977\",\n \"0.41975066\",\n \"0.4196735\",\n \"0.41945702\",\n \"0.41887376\",\n \"0.41816244\",\n \"0.4176389\",\n \"0.41736796\",\n \"0.41723612\",\n \"0.41679725\",\n \"0.41670468\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.5950558"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90893,"cells":{"query":{"kind":"string","value":"Return True if x is of type FriCASElement."},"document":{"kind":"string","value":"def is_FriCASElement(x):\n return isinstance(x, FriCASElement)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def is_(t, x):\n return type(x) is t","def _isinstance(cls, x):\n return isinstance(x, cls.PYTHON_TYPE_CHECK)","def has_x(self):\n return any(map(lambda s: s.is_x, self))","def is_my_case(self, type_):\n return (\n isinstance(self.__apply_sequence(type_), self.declaration_class)\n )","def is_element(obj):\n return isinstance(obj, etree._Element) # noqa","def of_type(self, a):\n return type(a) == type(self.one)","def supports(self, x):\n return True","def is_a(self, t):\n return isinstance(self._, t)","def is_simple(self) -> bool:\n return self.data in ('int', 'bool', 'float', 'str')","def ispoint(x):\n if isvect(x) and x[3] > 0.0:\n return True\n return False","def is_CPRFanoToricVariety(x):\n return isinstance(x, CPRFanoToricVariety_field)","def _is_node_an_element(self, node):\n # Try the simplest approach first, works for plain old ElementTree\n if isinstance(node, BaseET.Element):\n return True\n # For cElementTree we need to be more cunning (or find a better way)\n if hasattr(node, 'makeelement') \\\n and isinstance(node.tag, six.string_types):\n return True","def bool(x) -> bool:\n pass","def isElement(self):\n return _libsbml.XMLToken_isElement(self)","def _is_input_element(se):\n return inspect.isclass(se) and issubclass(se, BaseField)","def _isA(self, elementClass, category = ''):\n if not isinstance(self, elementClass):\n return False\n if category and self.getCategory() != category:\n return False\n return True","def strtype(x):\n if type(x) == str:\n return True\n if type(x) == unicode:\n return True\n return False","def is_function_type(self, objtype):\n # return self.__cfuncptrt == type(objtype)\n return issubclass(objtype, self.__cfuncptrt)\n # return isinstance(objtype, self.__cfuncptrt)","def isx(tree, x, accept_attr=True):\n # WTF, so sometimes there **is** a Captured node, while sometimes there isn't (letdoutil.islet)? At which point are these removed?\n # Captured nodes only come from unpythonic.syntax, and we use from-imports\n # and bare names for anything hq[]'d; but any references that appear\n # explicitly in the user code may use either bare names or somemodule.f.\n ismatch = x if callable(x) else lambda s: s == x\n return ((type(tree) is Name and ismatch(tree.id)) or\n (type(tree) is Captured and ismatch(tree.name)) or\n (accept_attr and type(tree) is Attribute and ismatch(tree.attr)))","def is_of_type(cls, value) -> bool:\n # UTF8 = 'utf-8'\n # UTF16 = 'utf-16'\n # UTF32 = 'utf-32'\n # ASCII = 'ascii'\n # BINARY = 'binary'\n # OCTAL = 'octal'\n # HEXADECIMAL = 'hexadecimal'\n # CP1252 = 'cp1252'\n # WINDOWS1252 = 'windows-1252'\n # UNICODEESCAPE = 'unicode-escape'\n\n v = None\n if cls == cls.UTF8 or cls == cls.UTF16 or cls == cls.UTF32 or cls == cls.UNICODEESCAPE:\n try:\n v = bytes(value)\n except:\n return False\n\n if cls == cls.ASCII:\n try:\n v = ascii(value)\n except:\n return False\n\n if cls == cls.BINARY:\n try:\n v = bin(value)\n except:\n return False\n\n if cls == cls.OCTAL:\n try:\n v = oct(value)\n except:\n return False\n\n if cls == cls.HEXADECIMAL:\n try:\n v = hex(value)\n except:\n return False\n\n if cls == cls.WINDOWS1252 or cls == cls.CP1252:\n try:\n v = str(value)\n except:\n return False\n return True","def isSinglePiece(self):\n return isinstance(self.substrates, Polygon)","def applies(cls, obj):\n return type(obj) in cls.types","def isdatatype(object):\n return isinstance(object, (str, int, bool, float, type(None)))","def predicate(obj):\n return inspect.isclass(obj) and issubclass(obj, MafColumnRecord)","def is_reaction_class(rxn_class):\n return rxn_class in _values(ReactionClass.Typ)","def is_valid(self, attribute: Attribute) -> bool:\n return self.get_data_type() == attribute.type","def isa_binary_element(self, element):\n\n return len(self.above(element)) <= 2 and len(self.under(element)) <= 2","def isNumber(x):\n\treturn type(x) in [int, float]","def isclass(object):\r\n return isinstance(object, (type, types.ClassType))","def canAdapt(self, stru):\n return isinstance(stru, crystal)","def trueifset(xeval, typematch=False):\n if typematch:\n if not (xeval is False or xeval is None): return True\n else: return False\n else:\n if xeval: return True\n else: return False","def __bool__(x):\n if x.value == 1:\n return True\n elif x.value == -1:\n return False\n else:\n raise ValueError('cannot determine boolean value of Unknown')","def isString(x):\n if type(x) == str:\n return True\n else:\n return False","def isone(x):\n try:\n return int(x)==1;\n except TypeError:\n return False","def any(x) -> bool:\n pass","def __contains__(self, a):\n try:\n self.convert(a)\n except CoercionFailed:\n return False\n\n return True","def isElementOnly(self):\n return _libsbml.SBaseExtensionPoint_isElementOnly(self)","def _is_this_color(cls, obj: Any) -> bool:\n\n return type(obj) is cls","def filter_variant(self, x):\n return True","def is_integer(x):\r\n if isinstance(x, float):\r\n return x == int(x)\r\n else:\r\n raise TypeError, \"Input float\"","def is_float(x):\r\n try:\r\n float(x)\r\n except ValueError:\r\n return False\r\n return True","def __isFastener(f):\n\n if type(f) != Fastener:\n raise TypeError(\"FastnerGroups may contain only Fasteners\")\n else:\n return True","def validate(self,value):\r\n return type(value) is self.datatype","def verify_type(self, obj):\n return isinstance(obj, self.type_)","def isNumber(x):\n return isinstance(x, (int, float))","def contains(self, x):\n raise NotImplementedError","def __contains__(self, x):\n # if not isinstance(x, int) or not x % 2:\n if not (isinstance(x, int) and (x % 2)):\n return False\n return True","def is_Gamma0(x):\n return isinstance(x, Gamma0_class)","def __datatype_check(self, record_attribute, attribute_schema):\n if 'INT' in attribute_schema[TYPE_KEY].upper():\n if record_attribute.isdigit():\n return True\n elif attribute_schema[TYPE_KEY].upper() in DECIMAL_TYPES:\n if record_attribute.isdecimal():\n return True\n elif 'CHAR' in attribute_schema[TYPE_KEY].upper() \\\n or 'TEXT' in attribute_schema[TYPE_KEY].upper():\n if type(record_attribute) is str:\n return True\n else:\n IS_VALID_FILE = False\n return False","def contains(self, x):\n raise NotImplementedError()","def compa_type(t_ef, t_f):\n\tif t_ef==1 and t_f==2:\n\t\treturn True\n\tif t_ef in [2,3,4] and t_f ==1:\n\t\treturn True\n\tif t_ef == 5 and t_f == 3:\n\t\treturn True\n\tif t_ef==6 and t_f == 4:\n\t\treturn True\n\treturn False","def is_int(x):\n return int(x) == x","def check_type(X, x_lat_dim, x_lon_dim, x_sample_dim, x_feature_dim):\n\tassert type(X) == xr.DataArray, 'XCast requires a dataset to be of type \"Xarray.DataArray\"'","def is_type_correct(*args):\n return _ida_hexrays.is_type_correct(*args)","def bool(self, obj):\n return True","def bool(self, obj):\n return True","def iterable(x):\n try:\n iter(x)\n return True\n except TypeError:\n return False","def any(self) -> bool:","def is_Hypersurface(self):\n return isinstance(self, (ProjectiveHypersurface, AffineHypersurface))","def check_type(self):\n return True","def isNodeType(self, t):\n return isinstance(self, t)","def is_same_as(self, c_repr):\n return self == Type._from_c_repr(c_repr)","def supports(self, x):\n return 0.0 < x","def supports(self, x):\n return 0.0 < x","def is_x(self, var):\n x_list = ['lon', 'longitude', 'LONGITUDE', 'Longitude', 'x']\n\n if self.get_units(var) == 'degrees_east':\n return True\n if self.get_name(var) in x_list:\n return True\n if self.get_description(var) in x_list:\n return True\n else:\n return False","def is_frontier_node(node):\r\n if isinstance(node, ast.Return):\r\n return True\r\n elif isinstance(node, ast.If):\r\n fields = dict(ast.iter_fields(node))\r\n body = fields['body']\r\n orelse = fields['orelse']\r\n if (ModuleVisitor.is_frontier_node(body) and \r\n ModuleVisitor.is_frontier_node(orelse)):\r\n return True\r\n else: \r\n return False\r\n elif isinstance(node, ast.For):\r\n fields = dict(ast.iter_fields(node))\r\n body = fields['body']\r\n if ModuleVisitor.is_frontier_node(body):\r\n return True\r\n else:\r\n return False\r\n elif isinstance(node, ast.While):\r\n fields = dict(ast.iter_fields(node))\r\n body = fields['body']\r\n if ModuleVisitor.is_frontier_node(body):\r\n return True\r\n else:\r\n return False\r\n elif isinstance(node, list):\r\n num_frontiers = [ModuleVisitor.is_frontier_node(c) \r\n for c in node].count(True)\r\n if num_frontiers > 0:\r\n return True\r\n else:\r\n return False\r\n else:\r\n return False","def is_int(x):\n # boolean are subclasses of integers in Python, so explicitly exclude them\n return isinstance(x, (int, np.integer)) and not isinstance(x, bool)","def isClass(self, className):\n return self.characterClass == className or self.baseClass == className","def isdependent(self, t):\n deptypes = set([k[0] for k in self.refined_types \\\n if not isinstance(k, basestring)])\n if isinstance(t, basestring):\n return t in deptypes\n if isinstance(t, Sequence):\n return self.isdependent(t[0])\n return False","def is_pyxb_d1_type_name(pyxb_obj, expected_pyxb_type_name):\n try:\n return pyxb_get_type_name(pyxb_obj) == expected_pyxb_type_name\n except AttributeError:\n return False","def contains(self, x):\n # need more to assure its a real SSP - ie on right torus\n return (len(x) == self._shape[0])","def ISLOGICAL(value):\n return isinstance(value, bool)","def isFunctionalCompartment(*args):\n return _libsbml.SBO_isFunctionalCompartment(*args)","def is_type(obj):\n return type(obj) is type or type(obj) is types.ClassType","def is_number(x):\n if isinstance(x, (int, float)):\n return True\n else:\n return False","def contains(self, x: object):\n return x in self.items","def __can_be_part_of_assert(cfunc, ctree_item):\n # type: (idaapi.cfunc_t, idaapi.ctree_item_t) -> bool\n\n if ctree_item.citype != idaapi.VDI_EXPR:\n return False\n\n expression = ctree_item.it.to_specific_type\n if expression.op != idaapi.cot_obj:\n return False\n\n parent = cfunc.body.find_parent_of(expression).to_specific_type\n if parent.op != idaapi.cot_call or parent.x.op != idaapi.cot_obj:\n return False\n\n obj_ea = expression.obj_ea\n if not helper.is_code_ea(obj_ea) and idc.get_str_type(obj_ea) == idc.STRTYPE_C:\n str_potential_name = idc.get_strlit_contents(obj_ea)\n if type(str_potential_name) is not str:\n # convert bytes to str (python 3)\n str_potential_name = str_potential_name.decode('ascii')\n return idaapi.is_valid_typename(str_potential_name)\n return False","def is_basic_ctype(self, objtype):\n if objtype in [self.c_char_p, self.c_void_p, self.CString]:\n return False\n # DOC: if <ctypes.c_uint> is not in self.__basic_types, its probably\n # because you are using the wrong ctypes Proxy instance\n return objtype in self.__basic_types","def isbinary(f):\n return f.dtype == bool","def is_etree_element(obj: Any) -> bool:\n return hasattr(obj, 'append') and hasattr(obj, 'tag') and hasattr(obj, 'attrib')","def is_icecube_class(obj: Any) -> bool:\n classname = str(type(obj))\n return \"icecube.\" in classname","def check_predicate(self, predicate, x):\n if isinstance(predicate, dtype.TypeMeta):\n return issubclass(x, predicate)\n elif isinstance(predicate, type):\n return isinstance(x, predicate)\n elif callable(predicate):\n return predicate(self, x)\n else:\n raise ValueError(predicate) # pragma: no cover","def check_type(content):\n return (isinstance(content, Elem) or type(content) == Text or\n (type(content) == list and all([type(elem) == Text or\n isinstance(elem, Elem)\n for elem in content])))","def is_element_only(self) -> bool:\n raise NotImplementedError()","def isiterable(x):\n try:\n iter(x)\n return True\n except TypeError:\n return False","def _is_valid(self, value):\n\n # Entities have an istypeof method that can perform more sophisticated\n # type checking.\n if hasattr(self._type, \"istypeof\"):\n return self._type.istypeof(value)\n else:\n return isinstance(value, self._type)","def bool(self):\n return bool(self.int(2))","def isrefinement(self, t):\n if isinstance(t, basestring):\n return t in self.refined_types\n return self.isdependent(t)","def is_basic_type(self, objtype):\n if not hasattr(objtype, '_type_'):\n # could be python types\n return objtype in [int, long, float, bool]\n return self.is_basic_ctype(objtype)","def __bool__(self):\n return self is TRUE","def test_match_types(self):\n f = lws.match_types\n # assert f(str, u'test') is True\n assert f(str, 'test') is True\n assert f(int, 123) is True\n assert f(int, 123.00) is False\n assert f(bool, [1, 2, 3]) is False","def test_evaluate_is_of_expression(self):\n value = self.evaluate_common(\"isof(2D,'Edm.Double')\")\n self.assertTrue(\n value.type_code == edm.SimpleType.Boolean, \"Expected Boolean\")\n self.assertTrue(value.value is True, \"Expected True\")\n value = self.evaluate_common(\"isof(2M,'Edm.Double')\")\n self.assertTrue(value.value is True, \"Expected True\")\n value = self.evaluate_common(\"isof(2,'Edm.Double')\")\n self.assertTrue(value.value is True, \"Expected True\")\n value = self.evaluate_common(\"isof(2.0D,'Edm.Single')\")\n self.assertTrue(value.value is False, \"Expected False\")\n value = self.evaluate_common(\"isof('x','Edm.String')\")\n self.assertTrue(value.value is True, \"Expected True\")\n value = self.evaluate_common(\"isof(X'DEAD','Edm.String')\")\n self.assertTrue(value.value is False, \"Expected False\")\n value = self.evaluate_common(\"isof(false or true,'Edm.Boolean')\")\n self.assertTrue(value.value is True, \"Expected True\")\n value = self.evaluate_common(\"isof(null,'Edm.String')\")\n self.assertTrue(value.value is False, \"Expected False\")\n value = self.evaluate_common(\"isof('Edm.String')\")\n self.assertTrue(value.value is False, \"Expected False\")","def match(self, cls):\n return isinstance(self, cls)","def accepts(cls, value: Any) -> bool:\n try:\n cls.convert(value)\n return True\n except ValueError:\n return False","def is_enum(self):\n return self.is_complex and not self.is_class","def is_interpretable(self):\n return bool(self.as_date() or self.as_time())","def is_type(obj: Any) -> bool:\n return type(obj).__name__ == \"type\"","def is_chars(obj: Any) -> bool:\n return isinstance(obj, (str, bytes, bytearray))","def check_instrument_type(row):\n instrument_type = row[2]\n return instrument_type == \"DEED\" or \"DEED OF TRUST\"","def _is_run_type(cls, object_):\n # Do a string comparison instead of using isinstance() to avoid needing\n # to import lyse or other modules with these classes.\n return (type(object_).__name__ in cls._RUN_TYPES)"],"string":"[\n \"def is_(t, x):\\n return type(x) is t\",\n \"def _isinstance(cls, x):\\n return isinstance(x, cls.PYTHON_TYPE_CHECK)\",\n \"def has_x(self):\\n return any(map(lambda s: s.is_x, self))\",\n \"def is_my_case(self, type_):\\n return (\\n isinstance(self.__apply_sequence(type_), self.declaration_class)\\n )\",\n \"def is_element(obj):\\n return isinstance(obj, etree._Element) # noqa\",\n \"def of_type(self, a):\\n return type(a) == type(self.one)\",\n \"def supports(self, x):\\n return True\",\n \"def is_a(self, t):\\n return isinstance(self._, t)\",\n \"def is_simple(self) -> bool:\\n return self.data in ('int', 'bool', 'float', 'str')\",\n \"def ispoint(x):\\n if isvect(x) and x[3] > 0.0:\\n return True\\n return False\",\n \"def is_CPRFanoToricVariety(x):\\n return isinstance(x, CPRFanoToricVariety_field)\",\n \"def _is_node_an_element(self, node):\\n # Try the simplest approach first, works for plain old ElementTree\\n if isinstance(node, BaseET.Element):\\n return True\\n # For cElementTree we need to be more cunning (or find a better way)\\n if hasattr(node, 'makeelement') \\\\\\n and isinstance(node.tag, six.string_types):\\n return True\",\n \"def bool(x) -> bool:\\n pass\",\n \"def isElement(self):\\n return _libsbml.XMLToken_isElement(self)\",\n \"def _is_input_element(se):\\n return inspect.isclass(se) and issubclass(se, BaseField)\",\n \"def _isA(self, elementClass, category = ''):\\n if not isinstance(self, elementClass):\\n return False\\n if category and self.getCategory() != category:\\n return False\\n return True\",\n \"def strtype(x):\\n if type(x) == str:\\n return True\\n if type(x) == unicode:\\n return True\\n return False\",\n \"def is_function_type(self, objtype):\\n # return self.__cfuncptrt == type(objtype)\\n return issubclass(objtype, self.__cfuncptrt)\\n # return isinstance(objtype, self.__cfuncptrt)\",\n \"def isx(tree, x, accept_attr=True):\\n # WTF, so sometimes there **is** a Captured node, while sometimes there isn't (letdoutil.islet)? At which point are these removed?\\n # Captured nodes only come from unpythonic.syntax, and we use from-imports\\n # and bare names for anything hq[]'d; but any references that appear\\n # explicitly in the user code may use either bare names or somemodule.f.\\n ismatch = x if callable(x) else lambda s: s == x\\n return ((type(tree) is Name and ismatch(tree.id)) or\\n (type(tree) is Captured and ismatch(tree.name)) or\\n (accept_attr and type(tree) is Attribute and ismatch(tree.attr)))\",\n \"def is_of_type(cls, value) -> bool:\\n # UTF8 = 'utf-8'\\n # UTF16 = 'utf-16'\\n # UTF32 = 'utf-32'\\n # ASCII = 'ascii'\\n # BINARY = 'binary'\\n # OCTAL = 'octal'\\n # HEXADECIMAL = 'hexadecimal'\\n # CP1252 = 'cp1252'\\n # WINDOWS1252 = 'windows-1252'\\n # UNICODEESCAPE = 'unicode-escape'\\n\\n v = None\\n if cls == cls.UTF8 or cls == cls.UTF16 or cls == cls.UTF32 or cls == cls.UNICODEESCAPE:\\n try:\\n v = bytes(value)\\n except:\\n return False\\n\\n if cls == cls.ASCII:\\n try:\\n v = ascii(value)\\n except:\\n return False\\n\\n if cls == cls.BINARY:\\n try:\\n v = bin(value)\\n except:\\n return False\\n\\n if cls == cls.OCTAL:\\n try:\\n v = oct(value)\\n except:\\n return False\\n\\n if cls == cls.HEXADECIMAL:\\n try:\\n v = hex(value)\\n except:\\n return False\\n\\n if cls == cls.WINDOWS1252 or cls == cls.CP1252:\\n try:\\n v = str(value)\\n except:\\n return False\\n return True\",\n \"def isSinglePiece(self):\\n return isinstance(self.substrates, Polygon)\",\n \"def applies(cls, obj):\\n return type(obj) in cls.types\",\n \"def isdatatype(object):\\n return isinstance(object, (str, int, bool, float, type(None)))\",\n \"def predicate(obj):\\n return inspect.isclass(obj) and issubclass(obj, MafColumnRecord)\",\n \"def is_reaction_class(rxn_class):\\n return rxn_class in _values(ReactionClass.Typ)\",\n \"def is_valid(self, attribute: Attribute) -> bool:\\n return self.get_data_type() == attribute.type\",\n \"def isa_binary_element(self, element):\\n\\n return len(self.above(element)) <= 2 and len(self.under(element)) <= 2\",\n \"def isNumber(x):\\n\\treturn type(x) in [int, float]\",\n \"def isclass(object):\\r\\n return isinstance(object, (type, types.ClassType))\",\n \"def canAdapt(self, stru):\\n return isinstance(stru, crystal)\",\n \"def trueifset(xeval, typematch=False):\\n if typematch:\\n if not (xeval is False or xeval is None): return True\\n else: return False\\n else:\\n if xeval: return True\\n else: return False\",\n \"def __bool__(x):\\n if x.value == 1:\\n return True\\n elif x.value == -1:\\n return False\\n else:\\n raise ValueError('cannot determine boolean value of Unknown')\",\n \"def isString(x):\\n if type(x) == str:\\n return True\\n else:\\n return False\",\n \"def isone(x):\\n try:\\n return int(x)==1;\\n except TypeError:\\n return False\",\n \"def any(x) -> bool:\\n pass\",\n \"def __contains__(self, a):\\n try:\\n self.convert(a)\\n except CoercionFailed:\\n return False\\n\\n return True\",\n \"def isElementOnly(self):\\n return _libsbml.SBaseExtensionPoint_isElementOnly(self)\",\n \"def _is_this_color(cls, obj: Any) -> bool:\\n\\n return type(obj) is cls\",\n \"def filter_variant(self, x):\\n return True\",\n \"def is_integer(x):\\r\\n if isinstance(x, float):\\r\\n return x == int(x)\\r\\n else:\\r\\n raise TypeError, \\\"Input float\\\"\",\n \"def is_float(x):\\r\\n try:\\r\\n float(x)\\r\\n except ValueError:\\r\\n return False\\r\\n return True\",\n \"def __isFastener(f):\\n\\n if type(f) != Fastener:\\n raise TypeError(\\\"FastnerGroups may contain only Fasteners\\\")\\n else:\\n return True\",\n \"def validate(self,value):\\r\\n return type(value) is self.datatype\",\n \"def verify_type(self, obj):\\n return isinstance(obj, self.type_)\",\n \"def isNumber(x):\\n return isinstance(x, (int, float))\",\n \"def contains(self, x):\\n raise NotImplementedError\",\n \"def __contains__(self, x):\\n # if not isinstance(x, int) or not x % 2:\\n if not (isinstance(x, int) and (x % 2)):\\n return False\\n return True\",\n \"def is_Gamma0(x):\\n return isinstance(x, Gamma0_class)\",\n \"def __datatype_check(self, record_attribute, attribute_schema):\\n if 'INT' in attribute_schema[TYPE_KEY].upper():\\n if record_attribute.isdigit():\\n return True\\n elif attribute_schema[TYPE_KEY].upper() in DECIMAL_TYPES:\\n if record_attribute.isdecimal():\\n return True\\n elif 'CHAR' in attribute_schema[TYPE_KEY].upper() \\\\\\n or 'TEXT' in attribute_schema[TYPE_KEY].upper():\\n if type(record_attribute) is str:\\n return True\\n else:\\n IS_VALID_FILE = False\\n return False\",\n \"def contains(self, x):\\n raise NotImplementedError()\",\n \"def compa_type(t_ef, t_f):\\n\\tif t_ef==1 and t_f==2:\\n\\t\\treturn True\\n\\tif t_ef in [2,3,4] and t_f ==1:\\n\\t\\treturn True\\n\\tif t_ef == 5 and t_f == 3:\\n\\t\\treturn True\\n\\tif t_ef==6 and t_f == 4:\\n\\t\\treturn True\\n\\treturn False\",\n \"def is_int(x):\\n return int(x) == x\",\n \"def check_type(X, x_lat_dim, x_lon_dim, x_sample_dim, x_feature_dim):\\n\\tassert type(X) == xr.DataArray, 'XCast requires a dataset to be of type \\\"Xarray.DataArray\\\"'\",\n \"def is_type_correct(*args):\\n return _ida_hexrays.is_type_correct(*args)\",\n \"def bool(self, obj):\\n return True\",\n \"def bool(self, obj):\\n return True\",\n \"def iterable(x):\\n try:\\n iter(x)\\n return True\\n except TypeError:\\n return False\",\n \"def any(self) -> bool:\",\n \"def is_Hypersurface(self):\\n return isinstance(self, (ProjectiveHypersurface, AffineHypersurface))\",\n \"def check_type(self):\\n return True\",\n \"def isNodeType(self, t):\\n return isinstance(self, t)\",\n \"def is_same_as(self, c_repr):\\n return self == Type._from_c_repr(c_repr)\",\n \"def supports(self, x):\\n return 0.0 < x\",\n \"def supports(self, x):\\n return 0.0 < x\",\n \"def is_x(self, var):\\n x_list = ['lon', 'longitude', 'LONGITUDE', 'Longitude', 'x']\\n\\n if self.get_units(var) == 'degrees_east':\\n return True\\n if self.get_name(var) in x_list:\\n return True\\n if self.get_description(var) in x_list:\\n return True\\n else:\\n return False\",\n \"def is_frontier_node(node):\\r\\n if isinstance(node, ast.Return):\\r\\n return True\\r\\n elif isinstance(node, ast.If):\\r\\n fields = dict(ast.iter_fields(node))\\r\\n body = fields['body']\\r\\n orelse = fields['orelse']\\r\\n if (ModuleVisitor.is_frontier_node(body) and \\r\\n ModuleVisitor.is_frontier_node(orelse)):\\r\\n return True\\r\\n else: \\r\\n return False\\r\\n elif isinstance(node, ast.For):\\r\\n fields = dict(ast.iter_fields(node))\\r\\n body = fields['body']\\r\\n if ModuleVisitor.is_frontier_node(body):\\r\\n return True\\r\\n else:\\r\\n return False\\r\\n elif isinstance(node, ast.While):\\r\\n fields = dict(ast.iter_fields(node))\\r\\n body = fields['body']\\r\\n if ModuleVisitor.is_frontier_node(body):\\r\\n return True\\r\\n else:\\r\\n return False\\r\\n elif isinstance(node, list):\\r\\n num_frontiers = [ModuleVisitor.is_frontier_node(c) \\r\\n for c in node].count(True)\\r\\n if num_frontiers > 0:\\r\\n return True\\r\\n else:\\r\\n return False\\r\\n else:\\r\\n return False\",\n \"def is_int(x):\\n # boolean are subclasses of integers in Python, so explicitly exclude them\\n return isinstance(x, (int, np.integer)) and not isinstance(x, bool)\",\n \"def isClass(self, className):\\n return self.characterClass == className or self.baseClass == className\",\n \"def isdependent(self, t):\\n deptypes = set([k[0] for k in self.refined_types \\\\\\n if not isinstance(k, basestring)])\\n if isinstance(t, basestring):\\n return t in deptypes\\n if isinstance(t, Sequence):\\n return self.isdependent(t[0])\\n return False\",\n \"def is_pyxb_d1_type_name(pyxb_obj, expected_pyxb_type_name):\\n try:\\n return pyxb_get_type_name(pyxb_obj) == expected_pyxb_type_name\\n except AttributeError:\\n return False\",\n \"def contains(self, x):\\n # need more to assure its a real SSP - ie on right torus\\n return (len(x) == self._shape[0])\",\n \"def ISLOGICAL(value):\\n return isinstance(value, bool)\",\n \"def isFunctionalCompartment(*args):\\n return _libsbml.SBO_isFunctionalCompartment(*args)\",\n \"def is_type(obj):\\n return type(obj) is type or type(obj) is types.ClassType\",\n \"def is_number(x):\\n if isinstance(x, (int, float)):\\n return True\\n else:\\n return False\",\n \"def contains(self, x: object):\\n return x in self.items\",\n \"def __can_be_part_of_assert(cfunc, ctree_item):\\n # type: (idaapi.cfunc_t, idaapi.ctree_item_t) -> bool\\n\\n if ctree_item.citype != idaapi.VDI_EXPR:\\n return False\\n\\n expression = ctree_item.it.to_specific_type\\n if expression.op != idaapi.cot_obj:\\n return False\\n\\n parent = cfunc.body.find_parent_of(expression).to_specific_type\\n if parent.op != idaapi.cot_call or parent.x.op != idaapi.cot_obj:\\n return False\\n\\n obj_ea = expression.obj_ea\\n if not helper.is_code_ea(obj_ea) and idc.get_str_type(obj_ea) == idc.STRTYPE_C:\\n str_potential_name = idc.get_strlit_contents(obj_ea)\\n if type(str_potential_name) is not str:\\n # convert bytes to str (python 3)\\n str_potential_name = str_potential_name.decode('ascii')\\n return idaapi.is_valid_typename(str_potential_name)\\n return False\",\n \"def is_basic_ctype(self, objtype):\\n if objtype in [self.c_char_p, self.c_void_p, self.CString]:\\n return False\\n # DOC: if <ctypes.c_uint> is not in self.__basic_types, its probably\\n # because you are using the wrong ctypes Proxy instance\\n return objtype in self.__basic_types\",\n \"def isbinary(f):\\n return f.dtype == bool\",\n \"def is_etree_element(obj: Any) -> bool:\\n return hasattr(obj, 'append') and hasattr(obj, 'tag') and hasattr(obj, 'attrib')\",\n \"def is_icecube_class(obj: Any) -> bool:\\n classname = str(type(obj))\\n return \\\"icecube.\\\" in classname\",\n \"def check_predicate(self, predicate, x):\\n if isinstance(predicate, dtype.TypeMeta):\\n return issubclass(x, predicate)\\n elif isinstance(predicate, type):\\n return isinstance(x, predicate)\\n elif callable(predicate):\\n return predicate(self, x)\\n else:\\n raise ValueError(predicate) # pragma: no cover\",\n \"def check_type(content):\\n return (isinstance(content, Elem) or type(content) == Text or\\n (type(content) == list and all([type(elem) == Text or\\n isinstance(elem, Elem)\\n for elem in content])))\",\n \"def is_element_only(self) -> bool:\\n raise NotImplementedError()\",\n \"def isiterable(x):\\n try:\\n iter(x)\\n return True\\n except TypeError:\\n return False\",\n \"def _is_valid(self, value):\\n\\n # Entities have an istypeof method that can perform more sophisticated\\n # type checking.\\n if hasattr(self._type, \\\"istypeof\\\"):\\n return self._type.istypeof(value)\\n else:\\n return isinstance(value, self._type)\",\n \"def bool(self):\\n return bool(self.int(2))\",\n \"def isrefinement(self, t):\\n if isinstance(t, basestring):\\n return t in self.refined_types\\n return self.isdependent(t)\",\n \"def is_basic_type(self, objtype):\\n if not hasattr(objtype, '_type_'):\\n # could be python types\\n return objtype in [int, long, float, bool]\\n return self.is_basic_ctype(objtype)\",\n \"def __bool__(self):\\n return self is TRUE\",\n \"def test_match_types(self):\\n f = lws.match_types\\n # assert f(str, u'test') is True\\n assert f(str, 'test') is True\\n assert f(int, 123) is True\\n assert f(int, 123.00) is False\\n assert f(bool, [1, 2, 3]) is False\",\n \"def test_evaluate_is_of_expression(self):\\n value = self.evaluate_common(\\\"isof(2D,'Edm.Double')\\\")\\n self.assertTrue(\\n value.type_code == edm.SimpleType.Boolean, \\\"Expected Boolean\\\")\\n self.assertTrue(value.value is True, \\\"Expected True\\\")\\n value = self.evaluate_common(\\\"isof(2M,'Edm.Double')\\\")\\n self.assertTrue(value.value is True, \\\"Expected True\\\")\\n value = self.evaluate_common(\\\"isof(2,'Edm.Double')\\\")\\n self.assertTrue(value.value is True, \\\"Expected True\\\")\\n value = self.evaluate_common(\\\"isof(2.0D,'Edm.Single')\\\")\\n self.assertTrue(value.value is False, \\\"Expected False\\\")\\n value = self.evaluate_common(\\\"isof('x','Edm.String')\\\")\\n self.assertTrue(value.value is True, \\\"Expected True\\\")\\n value = self.evaluate_common(\\\"isof(X'DEAD','Edm.String')\\\")\\n self.assertTrue(value.value is False, \\\"Expected False\\\")\\n value = self.evaluate_common(\\\"isof(false or true,'Edm.Boolean')\\\")\\n self.assertTrue(value.value is True, \\\"Expected True\\\")\\n value = self.evaluate_common(\\\"isof(null,'Edm.String')\\\")\\n self.assertTrue(value.value is False, \\\"Expected False\\\")\\n value = self.evaluate_common(\\\"isof('Edm.String')\\\")\\n self.assertTrue(value.value is False, \\\"Expected False\\\")\",\n \"def match(self, cls):\\n return isinstance(self, cls)\",\n \"def accepts(cls, value: Any) -> bool:\\n try:\\n cls.convert(value)\\n return True\\n except ValueError:\\n return False\",\n \"def is_enum(self):\\n return self.is_complex and not self.is_class\",\n \"def is_interpretable(self):\\n return bool(self.as_date() or self.as_time())\",\n \"def is_type(obj: Any) -> bool:\\n return type(obj).__name__ == \\\"type\\\"\",\n \"def is_chars(obj: Any) -> bool:\\n return isinstance(obj, (str, bytes, bytearray))\",\n \"def check_instrument_type(row):\\n instrument_type = row[2]\\n return instrument_type == \\\"DEED\\\" or \\\"DEED OF TRUST\\\"\",\n \"def _is_run_type(cls, object_):\\n # Do a string comparison instead of using isinstance() to avoid needing\\n # to import lyse or other modules with these classes.\\n return (type(object_).__name__ in cls._RUN_TYPES)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.6683116","0.6371762","0.6274012","0.6056215","0.5898661","0.58613616","0.58362883","0.57680345","0.56762767","0.5658424","0.56527305","0.5597607","0.5569221","0.55410606","0.55354494","0.55278945","0.552147","0.54687995","0.54323465","0.54302615","0.54199487","0.5406785","0.538738","0.5381316","0.5359967","0.53474826","0.5331598","0.5330459","0.5325735","0.53213865","0.5313828","0.5302398","0.5296101","0.5294096","0.5277669","0.5276891","0.52679956","0.52566326","0.5256551","0.5251687","0.52491564","0.52402973","0.52353877","0.5235181","0.5233259","0.5206837","0.5205336","0.51993","0.5197546","0.5194697","0.51881385","0.5181997","0.5180138","0.51792","0.5168748","0.5168748","0.5161026","0.516019","0.51504546","0.5138394","0.5133982","0.5128797","0.5125044","0.5125044","0.51216197","0.5115733","0.5113726","0.51114947","0.511061","0.5103895","0.5098585","0.5095152","0.5091778","0.5089054","0.50879836","0.5083286","0.5079605","0.5077377","0.5073769","0.5059972","0.50535744","0.5050823","0.50479347","0.5044594","0.5044424","0.50439817","0.5042128","0.5035871","0.5030318","0.502556","0.5021209","0.5016339","0.50131524","0.5012558","0.50125223","0.5008855","0.50088227","0.5004764","0.50035334","0.5002456"],"string":"[\n \"0.6683116\",\n \"0.6371762\",\n \"0.6274012\",\n \"0.6056215\",\n \"0.5898661\",\n \"0.58613616\",\n \"0.58362883\",\n \"0.57680345\",\n \"0.56762767\",\n \"0.5658424\",\n \"0.56527305\",\n \"0.5597607\",\n \"0.5569221\",\n \"0.55410606\",\n \"0.55354494\",\n \"0.55278945\",\n \"0.552147\",\n \"0.54687995\",\n \"0.54323465\",\n \"0.54302615\",\n \"0.54199487\",\n \"0.5406785\",\n \"0.538738\",\n \"0.5381316\",\n \"0.5359967\",\n \"0.53474826\",\n \"0.5331598\",\n \"0.5330459\",\n \"0.5325735\",\n \"0.53213865\",\n \"0.5313828\",\n \"0.5302398\",\n \"0.5296101\",\n \"0.5294096\",\n \"0.5277669\",\n \"0.5276891\",\n \"0.52679956\",\n \"0.52566326\",\n \"0.5256551\",\n \"0.5251687\",\n \"0.52491564\",\n \"0.52402973\",\n \"0.52353877\",\n \"0.5235181\",\n \"0.5233259\",\n \"0.5206837\",\n \"0.5205336\",\n \"0.51993\",\n \"0.5197546\",\n \"0.5194697\",\n \"0.51881385\",\n \"0.5181997\",\n \"0.5180138\",\n \"0.51792\",\n \"0.5168748\",\n \"0.5168748\",\n \"0.5161026\",\n \"0.516019\",\n \"0.51504546\",\n \"0.5138394\",\n \"0.5133982\",\n \"0.5128797\",\n \"0.5125044\",\n \"0.5125044\",\n \"0.51216197\",\n \"0.5115733\",\n \"0.5113726\",\n \"0.51114947\",\n \"0.511061\",\n \"0.5103895\",\n \"0.5098585\",\n \"0.5095152\",\n \"0.5091778\",\n \"0.5089054\",\n \"0.50879836\",\n \"0.5083286\",\n \"0.5079605\",\n \"0.5077377\",\n \"0.5073769\",\n \"0.5059972\",\n \"0.50535744\",\n \"0.5050823\",\n \"0.50479347\",\n \"0.5044594\",\n \"0.5044424\",\n \"0.50439817\",\n \"0.5042128\",\n \"0.5035871\",\n \"0.5030318\",\n \"0.502556\",\n \"0.5021209\",\n \"0.5016339\",\n \"0.50131524\",\n \"0.5012558\",\n \"0.50125223\",\n \"0.5008855\",\n \"0.50088227\",\n \"0.5004764\",\n \"0.50035334\",\n \"0.5002456\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.8895811"},"document_rank":{"kind":"string","value":"0"}}},{"rowIdx":90894,"cells":{"query":{"kind":"string","value":"Returns the FriCAS interface object defined in sage.interfaces.fricas."},"document":{"kind":"string","value":"def reduce_load_fricas():\n return fricas"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def get_fcigraph(self) -> 'FciGraph':\n return self._core","def ICRFFile_Instance():\n return _gmat_py.ICRFFile_Instance()","def copy(self):\n return FlorisInterface(self.floris.as_dict())","def _get_fcoe_intf_port_type(self):\n return self.__fcoe_intf_port_type","def get_fsurface(self, path):\n raise NotImplementedError","def fortran_interface(self) -> str:\n return ''.join([i.fortran_interface() for i in self.instances])","def obtem_fila(self):\n\n return self.fila","def classFactory(iface): # pylint: disable=invalid-name\n #\n from .indicadores_geosaude import IndicadoresGeosaude\n return IndicadoresGeosaude(iface)","def _get_cal_interface(self):\n plugin = rw_peas.PeasPlugin('rwcal_openstack', 'RwCal-1.0')\n engine, info, extension = plugin()\n cal = plugin.get_interface(\"Cloud\")\n rwloggerctx = rwlogger.RwLog.Ctx.new(\"Cal-Log\")\n rc = cal.init(rwloggerctx)\n assert rc == RwTypes.RwStatus.SUCCESS\n\n return cal","def get_cfi(data):\n if data is None:\n raise EmptyDataError('[!] Invalid data value')\n\n result = TA.CFI(data)\n if result is None:\n raise IndicatorException\n return result","def get_interface(name=''):\n global interfaces\n return interfaces[name]","def _get_factory(self):\n if self._factory_ref is None:\n self._factory_ref = get_factory(self.PLUGGAGE_FACTORY_NAME)\n return self._factory_ref","def getFactura(self): \n return self.caja","def getFactura(self): \n return self.caja","def getFactorys(self) -> List[ghidra.app.util.viewer.field.FieldFactory]:\n ...","def Instance():\n return _gmat_py.ICRFFile_Instance()","def recon_fista_tv(im, angles, lam, fista_iter, iter):\t\n\n\t# Create ASTRA geometries:\n\tvol_geom = astra.create_vol_geom(im.shape[1] , im.shape[1])\n\tproj_geom = astra.create_proj_geom('parallel', 1.0, im.shape[1], linspace(0, angles, im.shape[0], False))\n\n\t# Create the ASTRA projector:\n\tp = tvtomo.ProjectorASTRA2D(proj_geom,vol_geom)\t\n\t\n\t# Define parameters and FISTA object that performs reconstruction:\n\t#lam = 1**-14\n\tf = tvtomo.FISTA(p, double(lam), fista_iter)\n\n\t# Actual reconstruction (takes time):\n\tim_rec = f.reconstruct(im.astype(float32), iter)\n\t\n\treturn im_rec.astype(float32)","def classFactory(iface): # pylint: disable=invalid-name\n #\n from .CADDigitize import CADDigitize\n return CADDigitize(iface)","def _function_element_class(self):\n return FriCASFunctionElement","def getcfix(self): # 3\n res,resargs = self.__obj.getcfix()\n if res != 0:\n result,msg = self.__getlasterror(res)\n raise Error(rescode(res),msg)\n _cfix_return_value = resargs\n return _cfix_return_value","def fcvs(self): \n return self._link_reg.fcvs","def _readFAI(self, fai):\n\t\t#FAI Format http://www.biostars.org/p/1495/\n #chrName chrLen chrSeek lineBases lineLen\n #Chr1 30427671 6 79 80\n #Line len is bases+\\n\n\t\twith open(fai, 'r') as FAI:\n\t\t\tlines = [line.rstrip('\\n').split() for line in FAI]\n\t\tself.sorted_chroms = sorted([line[0] for line in lines])\n\t\tself.chrom_dict = {line[0]:int(line[1]) for line in lines}","def GetCallAccCashFlow(leg, run_date, cf_type):\n query = acm.CreateFASQLQuery('FCashFlow', 'AND')\n query.AddAttrNode('Leg.Oid', 'EQUAL', leg.Oid())\n query.AddAttrNode('PayDate', 'EQUAL', run_date)\n query.AddAttrNode('CashFlowType', 'EQUAL', \"Fixed Amount\")\n query.AddAttrNode('AdditionalInfo.PS_DepositType', 'EQUAL', cf_type)\n cashFlows = query.Select()\n if cashFlows:\n return cashFlows[0]\n return None","def get_factory(self):\n if self.factory is None:\n this_dir, this_filename = os.path.split(__file__)\n data_path = os.path.join(this_dir, \"RDKitPh4.fdef\")\n self.factory = ChemicalFeatures.BuildFeatureFactory(data_path)\n return self.factory","def fota(self):\n return FotaCollection(client=self)","def FaceToFace(\n movablePlane: str, fixedPlane: str, flip: Boolean, clearance: float\n ) -> \"Feature\":\n return Feature()","def get(self, id):\n return Freigabe.find_by_id(id)","def _get_fcoe_intf_ifindex(self):\n return self.__fcoe_intf_ifindex","def get_fish(data):\n if data is None:\n raise EmptyDataError('[!] Invalid data value')\n\n result = TA.FISH(data)\n if result is None:\n raise IndicatorException\n return result","def ICSClientFactory(full_config, slave_device, master_device, \n icsiface_index=0, clientiface_index=0):\n slave_config = full_config['vdevs'][slave_device]\n master_config = full_config['vdevs'][master_device]\n if not slave_config['icsifaces'] or not master_config['clientifaces']:\n return None #NOTE: We may want an exception thrown here\n \n to_config = slave_config['icsifaces'][icsiface_index]\n from_config = master_config['clientifaces'][clientiface_index]\n\n className = to_config['typ']\n cls = globals()[className] #Gets class from this module\n return cls(to_config, slave_config['points'], from_config)","def harvest_corr_fact(self):\n # Search in the file #\n lines = [line for line in self.paths.sas if \"if _2='\" in str(line)]\n # Do each line #\n query = \"if _2='([A-Z][A-Z])' then CF=([0-9].[0-9]+);\"\n extract = lambda line: re.findall(query, str(line))\n result = list(map(extract, lines))\n result = [found[0] for found in result if found]\n # Make a data frame #\n df = pandas.DataFrame(result, columns=['forest_type', 'corr_fact'])\n # Write back into a CSV #\n df.to_csv(str(self.paths.corr_fact), index=False)","def interface(self) -> type:\n return self.get_interface()","def get_family(self):\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\n return self._catalog","def get_family(self):\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\n return self._catalog","def get_family(self):\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\n return self._catalog","def getFacility(name):\n return Cuebot.getStub('facility').Get(\n facility_pb2.FacilityGetRequest(name=name), timeout=Cuebot.Timeout).facility","def _get_fcoe_intf_port_state(self):\n return self.__fcoe_intf_port_state","def to_acf(self):\n rho = 0.5*np.fft.irfft(self) / self.delta_t\n return AutoCovariance(rho, delta_t=self.delta_t)","def _get_interface_ref(self):\n return self.__interface_ref","def _get_interface_ref(self):\n return self.__interface_ref","def _get_interface_ref(self):\n return self.__interface_ref","def _get_interface_ref(self):\n return self.__interface_ref","def _get_interface_ref(self):\n return self.__interface_ref","def _get_interface_ref(self):\n return self.__interface_ref","def Interface(self):\n from ixnetwork_restpy.testplatform.sessions.ixnetwork.vport.protocols.interface_21173f93b01472152dbb5ea4d71785b0 import Interface\n return Interface(self)","def getService(self, interfaceClass: java.lang.Class) -> object:\n ...","def _get_fcoe_intf_config_port_type(self):\n return self.__fcoe_intf_config_port_type","def getInterface(self):\n\t\tquery = ''\n\t\tconn = self.get_connection()\n\t\theaders = { 'Content-type' : 'application/json', 'Authorization' : 'A10 %s' %self.sessionid}\n\t\tconn.request('GET', self.get_path() + '/' + query, headers=headers)\n\t\tresponse = conn.getresponse()\n\t\texpected_status = 200\n\t\terrors = {500: 'An unexpected runtime exception', 404: 'Specified interface does not exist'}\n\t\tpayload = self.get_output(response, expected_status, errors)\n\t\tconn.close()\n\t\tif self.debug:\n\t\t\tprint 'payload:', payload\n\t\tif payload == '':\n\t\t\tpayload = None\n\t\tif payload is not None:\n\t\t\tdata = json.loads(payload)\n\t\t\tpayload= data.get('interface')\n\t\treturn deserialize_Interface_json(payload)","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def _get_interface(self):\n return self.__interface","def get_vfi(data):\n if data is None:\n raise EmptyDataError('[!] Invalid data value')\n\n result = TA.VFI(data)\n if result is None:\n raise IndicatorException\n return result","def ReadAIFF(file):\n\ts = aifc.open(file,'r')\n\tnFrames = s.getnframes()\n\tstrSig = s.readframes(nFrames)\n\treturn np.fromstring(strSig, np.short).byteswap()","def _get_surfaces(idf):\n surface_types = [\n 'BUILDINGSURFACE:DETAILED',\n 'FENESTRATIONSURFACE:DETAILED',\n ]\n surfaces = []\n for surface_type in surface_types:\n surfaces.extend(idf.idfobjects[surface_type])\n\n return surfaces","def get_ctf(ima):\n\tfrom EMAN2 import EMAN2Ctf\n\tctf_params = ima.get_attr(\"ctf\")\t\n\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang","def get_interface(self) -> object:\n return get_analysis_version_interface(self)","def get_tgis_c_library_interface():\n global c_library_interface\n return c_library_interface","def Interface(self):\n unpublishedMethods = (\"roots\", \"Interface\",\n \"WishIdBeenAGirlie\")\n methodType = type(self.Interface)\n ifList = []\n for i in dir(self):\n if (type(getattr(self, i)) == methodType\n and not i.startswith('_')):\n ifList.append(i)\n for i in unpublishedMethods:\n ifList.remove(i)\n return LumberjackInterface(ifList)","def estimate_F_ransac(corr, num_iter, inlier_thresh):\n _, inlier_mask = ransac(corr, estimate_F, sym_epipolar_dist, 8, num_iter, inlier_thresh)\n # inlier_mask = np.ones(9)\n # inlier_mask[0] = 0\n F = estimate_F(corr[inlier_mask.astype(np.bool)])\n return F","def factory(self):\n return self._factory","def factory(self):\n return self._factory","def get_far_crefs_from( ea ):\r\n\tret = []\r\n\txrf = get_first_fcref_from( ea )\r\n\tif xrf != BADADDR:\r\n\t\tret.append( xrf )\r\n\txrf = get_next_fcref_from( ea, xrf )\r\n\twhile xrf != BADADDR:\r\n\t\tret.append( xrf )\r\n\t\txrf = get_next_fcref_from( ea, xrf )\r\n\treturn ret","def scheme_from_aif(obj: aif.Node, config: Config) -> t.Optional[SchemeNode]:\n\n aif_type = obj[\"type\"]\n aif_scheme: str = obj.get(\"scheme\", obj[\"text\"])\n\n if aif_type in aif2scheme:\n scheme = aif2scheme[t.cast(aif.SchemeType, aif_type)]\n\n # TODO: Handle formatting like capitalization, spaces, underscores, etc.\n # TODO: Araucaria does not use spaces between scheme names\n # aif_scheme = re.sub(\"([A-Z])\", r\" \\1\", aif_scheme)\n if scheme and (found_scheme := text2scheme[type(scheme)].get(aif_scheme)):\n scheme = found_scheme\n\n timestamp = (\n dt.from_format(obj.get(\"timestamp\"), aif.DATE_FORMAT) or pendulum.now()\n )\n\n return config.SchemeNodeClass(\n id=obj[\"nodeID\"],\n metadata=config.MetadataClass(timestamp, timestamp),\n scheme=scheme,\n )\n\n return None","def affaires_factures_view(request):\n # Check connected\n if not Utils.check_connected(request):\n raise exc.HTTPForbidden()\n\n affaire_id = request.matchdict[\"id\"]\n\n query = request.dbsession.query(VFactures).filter(\n VFactures.affaire_id == affaire_id\n ).all()\n return Utils.serialize_many(query)","def getFactorys(self, row: int) -> List[ghidra.app.util.viewer.field.FieldFactory]:\n ...","def getCalibFromHeader(self):\n calibs = ['DARK', 'FLAT', 'PHAS', 'VISI', 'WAVE']\n fsus = ['FSUA', 'FSUB']\n channels = ['W', '1', '2', '3', '4', '5']\n try:\n self.fsu_calib = {}\n for fsu in fsus:\n for calib in calibs:\n self.fsu_calib[(fsu, calib)] = np.zeros((6,4))\n for k, chan in enumerate(channels):\n self.fsu_calib[(fsu, calib)][k,:] =\\\n self.read4num('OCS '+fsu+' K'+\\\n chan+calib)\n\n return True\n except:\n if self.verbose:\n print '*WARNING* there do not seem to be calibrations'\n return False","def getcfix(self):\n cfix_ = ctypes.c_double()\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\n if res != 0:\n _,msg = self.__getlasterror(res)\n raise Error(rescode(res),msg)\n cfix_ = cfix_.value\n _cfix_return_value = cfix_\n return (_cfix_return_value)","def _get_fcoe_intf_fabric_map_name(self):\n return self.__fcoe_intf_fabric_map_name","def crea_falla_fosa(lats, lons, prof, dip, strike, latini, latfin, area_sf, razon_aspecto):\n # ruta del archivo de la fosa\n ruta_fosa = \"../Slab/\"\n # archivo fosa ( primera columna: longitudes, segunda columna: latitudes)\n arc_fosa = ruta_fosa + \"SAM2.txt\"\n lonfosa, latfosa = carga_fosa(arc_fosa)\n # largo de falla\n largo_falla_completa = Geodesic.WGS84.Inverse(limnorte, -72, limsur, -72)[ \"s12\" ]\n # tamano subfallas (subfallas cuadradas)\n tamano_subfalla = np.sqrt(area_sf)\n # cantidad de fallas\n cant_subfallas_lats = largo_falla_completa // tamano_subfalla\n cant_subfallas_lons = largo_falla_completa // razon_aspecto\n # latitudes subfallas\n lats_subfallas = np.flip(np.linspace(limsur, limnorte, int(cant_subfallas_lats))) \n\n # longitudes subfallas\n ancho_falla = tamano_subfalla * cant_subfallas_lons\n # interpolacion de longitud de la fosa para las latitudes limites\n interpolador_lons_fosa = interp1d(latfosa, lonfosa)\n interpolador_lats_fosa = interp1d(lonfosa, latfosa)\n # longitud de la fosa a la latitud de cada subfalla\n lons_fosa_para_subfallas = interpolador_lons_fosa(lats_subfallas)\n # teniendo las longitudes de la fosa para las latitudes de las subfallas se tiene\n # el limite oeste de la falla, falta encontrar el limite este. Como se conoce el ancho de la falla, \n # basta encontrar la longitud de este ancho para cada latitud\n lons_limite_este = np.ones(np.shape(lons_fosa_para_subfallas))\n\n for ilon in range(len(lons_limite_este)):\n lons_limite_este[ilon] = Geodesic.WGS84.Direct(lats_subfallas[ilon], lons_fosa_para_subfallas[ilon], 90, ancho_falla)[ \"lon2\" ]\n\n # teniendo los limites este y oeste, basta encontrar las longitudes de las subfallas intermedias\n\n array_lons = np.ones((int(n_subfallas_lats),int(n_subfallas_lons))) # LONS SUBFALLAS\n for jlat in range(int(n_subfallas_lats)):\n array_lons[jlat,:] = np.linspace(lons_fosa_para_subfallas[jlat],lons_limite_este[jlat],int(n_subfallas_lons))\n\n array_lats = np.tile(np.reshape(lats_subfallas,(int(n_subfallas_lats),1)),int(n_subfallas_lons)) # LATS SUBFALLAS","def fortran_interface(self) -> str:\n if self.public is None:\n return ''\n\n result = ' interface\\n\\n'\n result += \"\".join(member.fortran_interface() for member in self.members)\n result += ' end interface\\n\\n'\n\n for member in self.members:\n result += member.fortran_overloads()\n return result","def load_irac_psf(channel): # pragma: no cover\n\n channel = int(channel)\n if channel < 1 or channel > 4:\n raise ValueError('channel must be 1, 2, 3, or 4')\n\n fn = 'irac_ch{0}_flight.fits'.format(channel)\n path = get_path(fn, location='remote')\n hdu = fits.open(path)[0]\n\n return hdu","def frtype(self):\n return self._frtype","def RicercaPerCodiceFiscale(self, CodiceFiscale, SoloProtocollo=False, **kw):\n\n defaults = dict(\n Utente = UTENTE,\n Ruolo = RUOLO,\n CodiceFiscale=CodiceFiscale,\n SoloProtocollo=SoloProtocollo\n )\n\n request = self.build_xml('RicercaPerCodiceFiscale', **dict(defaults, **kw))\n\n return self.query_service('RicercaPerCodiceFiscale', request)","def getFencerByUSFAId(table_conn=None, fencer_id=None):\n\n if table_conn is None or fencer_id is None:\n return None\n\n table_conn.execute('SELECT * FROM fencers WHERE membership_id=?', fencer_id) \n\n row = table_conn.fetchall()\n if len(row) > 1:\n print((\"Duplicate fencer record:\", row[0][1], \" \", row[0][2]))\n\n if len(row) == 0:\n return None\n\n return Fencer(row[0][0], row[0][1], row[0][2], row[0][3], row[0][4])","def interface(self):\n return self._interface","def fortran_interface(self) -> str:\n result = ''\n for member in self.members:\n result += member.fortran_interface()\n return result","def fobj(self):\n return self._fobj","def retrieve_REFC(\n ds,\n variable=\"zFactorFinal\",\n radar_frequency=\"Ku\",\n mask_bright_band=False,\n mask_solid_phase=False,\n mask_liquid_phase=False,\n):\n if mask_solid_phase and mask_liquid_phase:\n raise ValueError(\"Either specify 'mask_solid_phase' or 'mask_liquid_phase'.\")\n # Retrieve required DataArrays\n da = get_variable_dataarray(ds, variable=variable)\n if len(da[\"radar_frequency\"].data) != 1:\n da = da.sel({\"radar_frequency\": radar_frequency})\n # Mask bright band region\n if mask_bright_band:\n da_bright_band = get_bright_band_mask(ds)\n da = da.where(~da_bright_band)\n # Mask ice phase region\n if mask_solid_phase:\n da_mask = get_solid_phase_mask(ds)\n da = da.where(da_mask)\n # Mask liquid phase region\n if mask_liquid_phase:\n da_mask = get_liquid_phase_mask(ds)\n da = da.where(da_mask)\n # Compute maximum\n da_max = da.max(dim=\"range\")\n # Add attributes\n if mask_solid_phase:\n da_max.name = \"REFC_liquid\"\n elif mask_liquid_phase:\n da_max.name = \"REFC_solid\"\n else:\n da_max.name = \"REFC\"\n da_max.attrs[\"units\"] = \"dBZ\"\n return da_max","def classFactory(iface):\n from qgbif.qgbif import QGBIFPlugin\n return QGBIFPlugin(iface)","def __init__(self, caldb_name, irf_name, verbose=False):\n\n self.caldb_path = os.environ['CALDB']\n self.caldb_name = caldb_name\n self.irf = irf_name\n self.verbose = verbose\n\n self.am_ok = True\n\n self._aeff = dict()\n self._psf = dict()\n self._edips = dict()\n\n self._check_available_irfs()\n\n self.input_irf_file_name = '{path:s}/data/cta/{caldb:s}/bcf/{irf:s}/irf_file.fits'.format(path=self.caldb_path,\n caldb=self.caldb_name,\n irf=irf_name)","def isc(rxn_class):\n return rxn_class[3]","def _FinditForFracas():\n return FinditForFracas(MOCK_GET_REPOSITORY)","def getServices(self, interfaceClass: java.lang.Class) -> List[object]:\n ...","def GetInterface(interface, dummy_interface=DummyInterface, is_dummy=False):\n return dummy_interface if is_dummy else interface","def classFactory(iface):\n from .os_translator_ii import OsTranslatorII\n return OsTranslatorII(iface)","def classFactory(iface): # pylint: disable=invalid-name\n #\n from .CCD_Plugin import CCD_Plugin\n return CCD_Plugin(iface)","def load_dlib_frontal_face_detector():\n return DlibDetector(dlib.get_frontal_face_detector())","def facility_code(self) -> FacilityCode:\n return self._facility_code"],"string":"[\n \"def get_fcigraph(self) -> 'FciGraph':\\n return self._core\",\n \"def ICRFFile_Instance():\\n return _gmat_py.ICRFFile_Instance()\",\n \"def copy(self):\\n return FlorisInterface(self.floris.as_dict())\",\n \"def _get_fcoe_intf_port_type(self):\\n return self.__fcoe_intf_port_type\",\n \"def get_fsurface(self, path):\\n raise NotImplementedError\",\n \"def fortran_interface(self) -> str:\\n return ''.join([i.fortran_interface() for i in self.instances])\",\n \"def obtem_fila(self):\\n\\n return self.fila\",\n \"def classFactory(iface): # pylint: disable=invalid-name\\n #\\n from .indicadores_geosaude import IndicadoresGeosaude\\n return IndicadoresGeosaude(iface)\",\n \"def _get_cal_interface(self):\\n plugin = rw_peas.PeasPlugin('rwcal_openstack', 'RwCal-1.0')\\n engine, info, extension = plugin()\\n cal = plugin.get_interface(\\\"Cloud\\\")\\n rwloggerctx = rwlogger.RwLog.Ctx.new(\\\"Cal-Log\\\")\\n rc = cal.init(rwloggerctx)\\n assert rc == RwTypes.RwStatus.SUCCESS\\n\\n return cal\",\n \"def get_cfi(data):\\n if data is None:\\n raise EmptyDataError('[!] Invalid data value')\\n\\n result = TA.CFI(data)\\n if result is None:\\n raise IndicatorException\\n return result\",\n \"def get_interface(name=''):\\n global interfaces\\n return interfaces[name]\",\n \"def _get_factory(self):\\n if self._factory_ref is None:\\n self._factory_ref = get_factory(self.PLUGGAGE_FACTORY_NAME)\\n return self._factory_ref\",\n \"def getFactura(self): \\n return self.caja\",\n \"def getFactura(self): \\n return self.caja\",\n \"def getFactorys(self) -> List[ghidra.app.util.viewer.field.FieldFactory]:\\n ...\",\n \"def Instance():\\n return _gmat_py.ICRFFile_Instance()\",\n \"def recon_fista_tv(im, angles, lam, fista_iter, iter):\\t\\n\\n\\t# Create ASTRA geometries:\\n\\tvol_geom = astra.create_vol_geom(im.shape[1] , im.shape[1])\\n\\tproj_geom = astra.create_proj_geom('parallel', 1.0, im.shape[1], linspace(0, angles, im.shape[0], False))\\n\\n\\t# Create the ASTRA projector:\\n\\tp = tvtomo.ProjectorASTRA2D(proj_geom,vol_geom)\\t\\n\\t\\n\\t# Define parameters and FISTA object that performs reconstruction:\\n\\t#lam = 1**-14\\n\\tf = tvtomo.FISTA(p, double(lam), fista_iter)\\n\\n\\t# Actual reconstruction (takes time):\\n\\tim_rec = f.reconstruct(im.astype(float32), iter)\\n\\t\\n\\treturn im_rec.astype(float32)\",\n \"def classFactory(iface): # pylint: disable=invalid-name\\n #\\n from .CADDigitize import CADDigitize\\n return CADDigitize(iface)\",\n \"def _function_element_class(self):\\n return FriCASFunctionElement\",\n \"def getcfix(self): # 3\\n res,resargs = self.__obj.getcfix()\\n if res != 0:\\n result,msg = self.__getlasterror(res)\\n raise Error(rescode(res),msg)\\n _cfix_return_value = resargs\\n return _cfix_return_value\",\n \"def fcvs(self): \\n return self._link_reg.fcvs\",\n \"def _readFAI(self, fai):\\n\\t\\t#FAI Format http://www.biostars.org/p/1495/\\n #chrName chrLen chrSeek lineBases lineLen\\n #Chr1 30427671 6 79 80\\n #Line len is bases+\\\\n\\n\\t\\twith open(fai, 'r') as FAI:\\n\\t\\t\\tlines = [line.rstrip('\\\\n').split() for line in FAI]\\n\\t\\tself.sorted_chroms = sorted([line[0] for line in lines])\\n\\t\\tself.chrom_dict = {line[0]:int(line[1]) for line in lines}\",\n \"def GetCallAccCashFlow(leg, run_date, cf_type):\\n query = acm.CreateFASQLQuery('FCashFlow', 'AND')\\n query.AddAttrNode('Leg.Oid', 'EQUAL', leg.Oid())\\n query.AddAttrNode('PayDate', 'EQUAL', run_date)\\n query.AddAttrNode('CashFlowType', 'EQUAL', \\\"Fixed Amount\\\")\\n query.AddAttrNode('AdditionalInfo.PS_DepositType', 'EQUAL', cf_type)\\n cashFlows = query.Select()\\n if cashFlows:\\n return cashFlows[0]\\n return None\",\n \"def get_factory(self):\\n if self.factory is None:\\n this_dir, this_filename = os.path.split(__file__)\\n data_path = os.path.join(this_dir, \\\"RDKitPh4.fdef\\\")\\n self.factory = ChemicalFeatures.BuildFeatureFactory(data_path)\\n return self.factory\",\n \"def fota(self):\\n return FotaCollection(client=self)\",\n \"def FaceToFace(\\n movablePlane: str, fixedPlane: str, flip: Boolean, clearance: float\\n ) -> \\\"Feature\\\":\\n return Feature()\",\n \"def get(self, id):\\n return Freigabe.find_by_id(id)\",\n \"def _get_fcoe_intf_ifindex(self):\\n return self.__fcoe_intf_ifindex\",\n \"def get_fish(data):\\n if data is None:\\n raise EmptyDataError('[!] Invalid data value')\\n\\n result = TA.FISH(data)\\n if result is None:\\n raise IndicatorException\\n return result\",\n \"def ICSClientFactory(full_config, slave_device, master_device, \\n icsiface_index=0, clientiface_index=0):\\n slave_config = full_config['vdevs'][slave_device]\\n master_config = full_config['vdevs'][master_device]\\n if not slave_config['icsifaces'] or not master_config['clientifaces']:\\n return None #NOTE: We may want an exception thrown here\\n \\n to_config = slave_config['icsifaces'][icsiface_index]\\n from_config = master_config['clientifaces'][clientiface_index]\\n\\n className = to_config['typ']\\n cls = globals()[className] #Gets class from this module\\n return cls(to_config, slave_config['points'], from_config)\",\n \"def harvest_corr_fact(self):\\n # Search in the file #\\n lines = [line for line in self.paths.sas if \\\"if _2='\\\" in str(line)]\\n # Do each line #\\n query = \\\"if _2='([A-Z][A-Z])' then CF=([0-9].[0-9]+);\\\"\\n extract = lambda line: re.findall(query, str(line))\\n result = list(map(extract, lines))\\n result = [found[0] for found in result if found]\\n # Make a data frame #\\n df = pandas.DataFrame(result, columns=['forest_type', 'corr_fact'])\\n # Write back into a CSV #\\n df.to_csv(str(self.paths.corr_fact), index=False)\",\n \"def interface(self) -> type:\\n return self.get_interface()\",\n \"def get_family(self):\\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\\n return self._catalog\",\n \"def get_family(self):\\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\\n return self._catalog\",\n \"def get_family(self):\\n # Implemented from template for osid.resource.ResourceLookupSession.get_bin\\n return self._catalog\",\n \"def getFacility(name):\\n return Cuebot.getStub('facility').Get(\\n facility_pb2.FacilityGetRequest(name=name), timeout=Cuebot.Timeout).facility\",\n \"def _get_fcoe_intf_port_state(self):\\n return self.__fcoe_intf_port_state\",\n \"def to_acf(self):\\n rho = 0.5*np.fft.irfft(self) / self.delta_t\\n return AutoCovariance(rho, delta_t=self.delta_t)\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def _get_interface_ref(self):\\n return self.__interface_ref\",\n \"def Interface(self):\\n from ixnetwork_restpy.testplatform.sessions.ixnetwork.vport.protocols.interface_21173f93b01472152dbb5ea4d71785b0 import Interface\\n return Interface(self)\",\n \"def getService(self, interfaceClass: java.lang.Class) -> object:\\n ...\",\n \"def _get_fcoe_intf_config_port_type(self):\\n return self.__fcoe_intf_config_port_type\",\n \"def getInterface(self):\\n\\t\\tquery = ''\\n\\t\\tconn = self.get_connection()\\n\\t\\theaders = { 'Content-type' : 'application/json', 'Authorization' : 'A10 %s' %self.sessionid}\\n\\t\\tconn.request('GET', self.get_path() + '/' + query, headers=headers)\\n\\t\\tresponse = conn.getresponse()\\n\\t\\texpected_status = 200\\n\\t\\terrors = {500: 'An unexpected runtime exception', 404: 'Specified interface does not exist'}\\n\\t\\tpayload = self.get_output(response, expected_status, errors)\\n\\t\\tconn.close()\\n\\t\\tif self.debug:\\n\\t\\t\\tprint 'payload:', payload\\n\\t\\tif payload == '':\\n\\t\\t\\tpayload = None\\n\\t\\tif payload is not None:\\n\\t\\t\\tdata = json.loads(payload)\\n\\t\\t\\tpayload= data.get('interface')\\n\\t\\treturn deserialize_Interface_json(payload)\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def _get_interface(self):\\n return self.__interface\",\n \"def get_vfi(data):\\n if data is None:\\n raise EmptyDataError('[!] Invalid data value')\\n\\n result = TA.VFI(data)\\n if result is None:\\n raise IndicatorException\\n return result\",\n \"def ReadAIFF(file):\\n\\ts = aifc.open(file,'r')\\n\\tnFrames = s.getnframes()\\n\\tstrSig = s.readframes(nFrames)\\n\\treturn np.fromstring(strSig, np.short).byteswap()\",\n \"def _get_surfaces(idf):\\n surface_types = [\\n 'BUILDINGSURFACE:DETAILED',\\n 'FENESTRATIONSURFACE:DETAILED',\\n ]\\n surfaces = []\\n for surface_type in surface_types:\\n surfaces.extend(idf.idfobjects[surface_type])\\n\\n return surfaces\",\n \"def get_ctf(ima):\\n\\tfrom EMAN2 import EMAN2Ctf\\n\\tctf_params = ima.get_attr(\\\"ctf\\\")\\t\\n\\treturn ctf_params.defocus, ctf_params.cs, ctf_params.voltage, ctf_params.apix, ctf_params.bfactor, ctf_params.ampcont, ctf_params.dfdiff, ctf_params.dfang\",\n \"def get_interface(self) -> object:\\n return get_analysis_version_interface(self)\",\n \"def get_tgis_c_library_interface():\\n global c_library_interface\\n return c_library_interface\",\n \"def Interface(self):\\n unpublishedMethods = (\\\"roots\\\", \\\"Interface\\\",\\n \\\"WishIdBeenAGirlie\\\")\\n methodType = type(self.Interface)\\n ifList = []\\n for i in dir(self):\\n if (type(getattr(self, i)) == methodType\\n and not i.startswith('_')):\\n ifList.append(i)\\n for i in unpublishedMethods:\\n ifList.remove(i)\\n return LumberjackInterface(ifList)\",\n \"def estimate_F_ransac(corr, num_iter, inlier_thresh):\\n _, inlier_mask = ransac(corr, estimate_F, sym_epipolar_dist, 8, num_iter, inlier_thresh)\\n # inlier_mask = np.ones(9)\\n # inlier_mask[0] = 0\\n F = estimate_F(corr[inlier_mask.astype(np.bool)])\\n return F\",\n \"def factory(self):\\n return self._factory\",\n \"def factory(self):\\n return self._factory\",\n \"def get_far_crefs_from( ea ):\\r\\n\\tret = []\\r\\n\\txrf = get_first_fcref_from( ea )\\r\\n\\tif xrf != BADADDR:\\r\\n\\t\\tret.append( xrf )\\r\\n\\txrf = get_next_fcref_from( ea, xrf )\\r\\n\\twhile xrf != BADADDR:\\r\\n\\t\\tret.append( xrf )\\r\\n\\t\\txrf = get_next_fcref_from( ea, xrf )\\r\\n\\treturn ret\",\n \"def scheme_from_aif(obj: aif.Node, config: Config) -> t.Optional[SchemeNode]:\\n\\n aif_type = obj[\\\"type\\\"]\\n aif_scheme: str = obj.get(\\\"scheme\\\", obj[\\\"text\\\"])\\n\\n if aif_type in aif2scheme:\\n scheme = aif2scheme[t.cast(aif.SchemeType, aif_type)]\\n\\n # TODO: Handle formatting like capitalization, spaces, underscores, etc.\\n # TODO: Araucaria does not use spaces between scheme names\\n # aif_scheme = re.sub(\\\"([A-Z])\\\", r\\\" \\\\1\\\", aif_scheme)\\n if scheme and (found_scheme := text2scheme[type(scheme)].get(aif_scheme)):\\n scheme = found_scheme\\n\\n timestamp = (\\n dt.from_format(obj.get(\\\"timestamp\\\"), aif.DATE_FORMAT) or pendulum.now()\\n )\\n\\n return config.SchemeNodeClass(\\n id=obj[\\\"nodeID\\\"],\\n metadata=config.MetadataClass(timestamp, timestamp),\\n scheme=scheme,\\n )\\n\\n return None\",\n \"def affaires_factures_view(request):\\n # Check connected\\n if not Utils.check_connected(request):\\n raise exc.HTTPForbidden()\\n\\n affaire_id = request.matchdict[\\\"id\\\"]\\n\\n query = request.dbsession.query(VFactures).filter(\\n VFactures.affaire_id == affaire_id\\n ).all()\\n return Utils.serialize_many(query)\",\n \"def getFactorys(self, row: int) -> List[ghidra.app.util.viewer.field.FieldFactory]:\\n ...\",\n \"def getCalibFromHeader(self):\\n calibs = ['DARK', 'FLAT', 'PHAS', 'VISI', 'WAVE']\\n fsus = ['FSUA', 'FSUB']\\n channels = ['W', '1', '2', '3', '4', '5']\\n try:\\n self.fsu_calib = {}\\n for fsu in fsus:\\n for calib in calibs:\\n self.fsu_calib[(fsu, calib)] = np.zeros((6,4))\\n for k, chan in enumerate(channels):\\n self.fsu_calib[(fsu, calib)][k,:] =\\\\\\n self.read4num('OCS '+fsu+' K'+\\\\\\n chan+calib)\\n\\n return True\\n except:\\n if self.verbose:\\n print '*WARNING* there do not seem to be calibrations'\\n return False\",\n \"def getcfix(self):\\n cfix_ = ctypes.c_double()\\n res = __library__.MSK_XX_getcfix(self.__nativep,ctypes.byref(cfix_))\\n if res != 0:\\n _,msg = self.__getlasterror(res)\\n raise Error(rescode(res),msg)\\n cfix_ = cfix_.value\\n _cfix_return_value = cfix_\\n return (_cfix_return_value)\",\n \"def _get_fcoe_intf_fabric_map_name(self):\\n return self.__fcoe_intf_fabric_map_name\",\n \"def crea_falla_fosa(lats, lons, prof, dip, strike, latini, latfin, area_sf, razon_aspecto):\\n # ruta del archivo de la fosa\\n ruta_fosa = \\\"../Slab/\\\"\\n # archivo fosa ( primera columna: longitudes, segunda columna: latitudes)\\n arc_fosa = ruta_fosa + \\\"SAM2.txt\\\"\\n lonfosa, latfosa = carga_fosa(arc_fosa)\\n # largo de falla\\n largo_falla_completa = Geodesic.WGS84.Inverse(limnorte, -72, limsur, -72)[ \\\"s12\\\" ]\\n # tamano subfallas (subfallas cuadradas)\\n tamano_subfalla = np.sqrt(area_sf)\\n # cantidad de fallas\\n cant_subfallas_lats = largo_falla_completa // tamano_subfalla\\n cant_subfallas_lons = largo_falla_completa // razon_aspecto\\n # latitudes subfallas\\n lats_subfallas = np.flip(np.linspace(limsur, limnorte, int(cant_subfallas_lats))) \\n\\n # longitudes subfallas\\n ancho_falla = tamano_subfalla * cant_subfallas_lons\\n # interpolacion de longitud de la fosa para las latitudes limites\\n interpolador_lons_fosa = interp1d(latfosa, lonfosa)\\n interpolador_lats_fosa = interp1d(lonfosa, latfosa)\\n # longitud de la fosa a la latitud de cada subfalla\\n lons_fosa_para_subfallas = interpolador_lons_fosa(lats_subfallas)\\n # teniendo las longitudes de la fosa para las latitudes de las subfallas se tiene\\n # el limite oeste de la falla, falta encontrar el limite este. Como se conoce el ancho de la falla, \\n # basta encontrar la longitud de este ancho para cada latitud\\n lons_limite_este = np.ones(np.shape(lons_fosa_para_subfallas))\\n\\n for ilon in range(len(lons_limite_este)):\\n lons_limite_este[ilon] = Geodesic.WGS84.Direct(lats_subfallas[ilon], lons_fosa_para_subfallas[ilon], 90, ancho_falla)[ \\\"lon2\\\" ]\\n\\n # teniendo los limites este y oeste, basta encontrar las longitudes de las subfallas intermedias\\n\\n array_lons = np.ones((int(n_subfallas_lats),int(n_subfallas_lons))) # LONS SUBFALLAS\\n for jlat in range(int(n_subfallas_lats)):\\n array_lons[jlat,:] = np.linspace(lons_fosa_para_subfallas[jlat],lons_limite_este[jlat],int(n_subfallas_lons))\\n\\n array_lats = np.tile(np.reshape(lats_subfallas,(int(n_subfallas_lats),1)),int(n_subfallas_lons)) # LATS SUBFALLAS\",\n \"def fortran_interface(self) -> str:\\n if self.public is None:\\n return ''\\n\\n result = ' interface\\\\n\\\\n'\\n result += \\\"\\\".join(member.fortran_interface() for member in self.members)\\n result += ' end interface\\\\n\\\\n'\\n\\n for member in self.members:\\n result += member.fortran_overloads()\\n return result\",\n \"def load_irac_psf(channel): # pragma: no cover\\n\\n channel = int(channel)\\n if channel < 1 or channel > 4:\\n raise ValueError('channel must be 1, 2, 3, or 4')\\n\\n fn = 'irac_ch{0}_flight.fits'.format(channel)\\n path = get_path(fn, location='remote')\\n hdu = fits.open(path)[0]\\n\\n return hdu\",\n \"def frtype(self):\\n return self._frtype\",\n \"def RicercaPerCodiceFiscale(self, CodiceFiscale, SoloProtocollo=False, **kw):\\n\\n defaults = dict(\\n Utente = UTENTE,\\n Ruolo = RUOLO,\\n CodiceFiscale=CodiceFiscale,\\n SoloProtocollo=SoloProtocollo\\n )\\n\\n request = self.build_xml('RicercaPerCodiceFiscale', **dict(defaults, **kw))\\n\\n return self.query_service('RicercaPerCodiceFiscale', request)\",\n \"def getFencerByUSFAId(table_conn=None, fencer_id=None):\\n\\n if table_conn is None or fencer_id is None:\\n return None\\n\\n table_conn.execute('SELECT * FROM fencers WHERE membership_id=?', fencer_id) \\n\\n row = table_conn.fetchall()\\n if len(row) > 1:\\n print((\\\"Duplicate fencer record:\\\", row[0][1], \\\" \\\", row[0][2]))\\n\\n if len(row) == 0:\\n return None\\n\\n return Fencer(row[0][0], row[0][1], row[0][2], row[0][3], row[0][4])\",\n \"def interface(self):\\n return self._interface\",\n \"def fortran_interface(self) -> str:\\n result = ''\\n for member in self.members:\\n result += member.fortran_interface()\\n return result\",\n \"def fobj(self):\\n return self._fobj\",\n \"def retrieve_REFC(\\n ds,\\n variable=\\\"zFactorFinal\\\",\\n radar_frequency=\\\"Ku\\\",\\n mask_bright_band=False,\\n mask_solid_phase=False,\\n mask_liquid_phase=False,\\n):\\n if mask_solid_phase and mask_liquid_phase:\\n raise ValueError(\\\"Either specify 'mask_solid_phase' or 'mask_liquid_phase'.\\\")\\n # Retrieve required DataArrays\\n da = get_variable_dataarray(ds, variable=variable)\\n if len(da[\\\"radar_frequency\\\"].data) != 1:\\n da = da.sel({\\\"radar_frequency\\\": radar_frequency})\\n # Mask bright band region\\n if mask_bright_band:\\n da_bright_band = get_bright_band_mask(ds)\\n da = da.where(~da_bright_band)\\n # Mask ice phase region\\n if mask_solid_phase:\\n da_mask = get_solid_phase_mask(ds)\\n da = da.where(da_mask)\\n # Mask liquid phase region\\n if mask_liquid_phase:\\n da_mask = get_liquid_phase_mask(ds)\\n da = da.where(da_mask)\\n # Compute maximum\\n da_max = da.max(dim=\\\"range\\\")\\n # Add attributes\\n if mask_solid_phase:\\n da_max.name = \\\"REFC_liquid\\\"\\n elif mask_liquid_phase:\\n da_max.name = \\\"REFC_solid\\\"\\n else:\\n da_max.name = \\\"REFC\\\"\\n da_max.attrs[\\\"units\\\"] = \\\"dBZ\\\"\\n return da_max\",\n \"def classFactory(iface):\\n from qgbif.qgbif import QGBIFPlugin\\n return QGBIFPlugin(iface)\",\n \"def __init__(self, caldb_name, irf_name, verbose=False):\\n\\n self.caldb_path = os.environ['CALDB']\\n self.caldb_name = caldb_name\\n self.irf = irf_name\\n self.verbose = verbose\\n\\n self.am_ok = True\\n\\n self._aeff = dict()\\n self._psf = dict()\\n self._edips = dict()\\n\\n self._check_available_irfs()\\n\\n self.input_irf_file_name = '{path:s}/data/cta/{caldb:s}/bcf/{irf:s}/irf_file.fits'.format(path=self.caldb_path,\\n caldb=self.caldb_name,\\n irf=irf_name)\",\n \"def isc(rxn_class):\\n return rxn_class[3]\",\n \"def _FinditForFracas():\\n return FinditForFracas(MOCK_GET_REPOSITORY)\",\n \"def getServices(self, interfaceClass: java.lang.Class) -> List[object]:\\n ...\",\n \"def GetInterface(interface, dummy_interface=DummyInterface, is_dummy=False):\\n return dummy_interface if is_dummy else interface\",\n \"def classFactory(iface):\\n from .os_translator_ii import OsTranslatorII\\n return OsTranslatorII(iface)\",\n \"def classFactory(iface): # pylint: disable=invalid-name\\n #\\n from .CCD_Plugin import CCD_Plugin\\n return CCD_Plugin(iface)\",\n \"def load_dlib_frontal_face_detector():\\n return DlibDetector(dlib.get_frontal_face_detector())\",\n \"def facility_code(self) -> FacilityCode:\\n return self._facility_code\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.57734627","0.52523446","0.5242085","0.52116334","0.51110476","0.5084472","0.50731456","0.5066771","0.5064217","0.5060581","0.4997012","0.49558657","0.4905208","0.4905208","0.48911813","0.4859921","0.48593447","0.48388115","0.4777076","0.4754972","0.47447875","0.47428632","0.473576","0.4733572","0.4727232","0.47237876","0.47088814","0.46954492","0.46950984","0.46860653","0.46849108","0.46642938","0.46580824","0.46580824","0.46580824","0.4644518","0.46429032","0.46333286","0.46200213","0.46200213","0.46200213","0.46200213","0.46200213","0.46200213","0.4612544","0.46043766","0.4597288","0.45934203","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45926738","0.45891532","0.45817783","0.4568277","0.45682758","0.45599502","0.45520836","0.4551573","0.45401937","0.45373988","0.45373988","0.45286733","0.45211953","0.45183173","0.44973543","0.44951206","0.44915313","0.4491325","0.44895568","0.4485285","0.44846496","0.4470758","0.44700408","0.44673312","0.44615316","0.44612172","0.44598877","0.44586793","0.44585338","0.4444979","0.44384146","0.44375184","0.4435116","0.44345728","0.44342297","0.44298333","0.442753","0.44243452"],"string":"[\n \"0.57734627\",\n \"0.52523446\",\n \"0.5242085\",\n \"0.52116334\",\n \"0.51110476\",\n \"0.5084472\",\n \"0.50731456\",\n \"0.5066771\",\n \"0.5064217\",\n \"0.5060581\",\n \"0.4997012\",\n \"0.49558657\",\n \"0.4905208\",\n \"0.4905208\",\n \"0.48911813\",\n \"0.4859921\",\n \"0.48593447\",\n \"0.48388115\",\n \"0.4777076\",\n \"0.4754972\",\n \"0.47447875\",\n \"0.47428632\",\n \"0.473576\",\n \"0.4733572\",\n \"0.4727232\",\n \"0.47237876\",\n \"0.47088814\",\n \"0.46954492\",\n \"0.46950984\",\n \"0.46860653\",\n \"0.46849108\",\n \"0.46642938\",\n \"0.46580824\",\n \"0.46580824\",\n \"0.46580824\",\n \"0.4644518\",\n \"0.46429032\",\n \"0.46333286\",\n \"0.46200213\",\n \"0.46200213\",\n \"0.46200213\",\n \"0.46200213\",\n \"0.46200213\",\n \"0.46200213\",\n \"0.4612544\",\n \"0.46043766\",\n \"0.4597288\",\n \"0.45934203\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45926738\",\n \"0.45891532\",\n \"0.45817783\",\n \"0.4568277\",\n \"0.45682758\",\n \"0.45599502\",\n \"0.45520836\",\n \"0.4551573\",\n \"0.45401937\",\n \"0.45373988\",\n \"0.45373988\",\n \"0.45286733\",\n \"0.45211953\",\n \"0.45183173\",\n \"0.44973543\",\n \"0.44951206\",\n \"0.44915313\",\n \"0.4491325\",\n \"0.44895568\",\n \"0.4485285\",\n \"0.44846496\",\n \"0.4470758\",\n \"0.44700408\",\n \"0.44673312\",\n \"0.44615316\",\n \"0.44612172\",\n \"0.44598877\",\n \"0.44586793\",\n \"0.44585338\",\n \"0.4444979\",\n \"0.44384146\",\n \"0.44375184\",\n \"0.4435116\",\n \"0.44345728\",\n \"0.44342297\",\n \"0.44298333\",\n \"0.442753\",\n \"0.44243452\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.4545254"},"document_rank":{"kind":"string","value":"70"}}},{"rowIdx":90895,"cells":{"query":{"kind":"string","value":"Spawn a new FriCAS commandline session."},"document":{"kind":"string","value":"def fricas_console():\n from sage.repl.rich_output.display_manager import get_display_manager\n if not get_display_manager().is_in_terminal():\n raise RuntimeError('Can use the console only in the terminal. Try %%fricas magics instead.')\n os.system('fricas -nox')"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def cli():\n config, auth, execute_now = read_command_line_arguments()\n main(config, auth, execute_now)","def main():\n # Parse arguments.\n parser = argparse.ArgumentParser(description='Console for cluster test.')\n parser.add_argument('-c', dest='command_config_file',\n default='command_config',\n help='set command config file path (default=command_config)')\n args = parser.parse_args()\n\n console = Console()\n console.config(args.command_config_file)\n console.init_proxies()\n console.run()","def __init__(self, name='fricas', command='fricas -nox -noclef',\n script_subdirectory=None, logfile=None,\n server=None, server_tmpdir=None,\n init_code=[')lisp (si::readline-off)']):\n PanAxiom.__init__(self, name, command,\n script_subdirectory, logfile,\n server, server_tmpdir,\n init_code)","def Spawn(proc):\n proc.start()\n return proc","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():","def cli():\r\n pass","def cli():\n ...","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n pass","def cli():\n\n pass","def spawn(self):\n self._proc = subprocess.Popen(\n self._args, stdout=subprocess.PIPE, stderr=subprocess.PIPE\n )","def main_cli():\n pass","def cli(ctx: click.Context, config_instance: str, config_path: str) -> None:\n ctx.obj = TurbiniaCli(\n config_instance=config_instance, config_path=config_path)\n\n # Set up the tool based on the configuration file parameters.\n ctx.obj.setup()\n\n # Build all the commands based on responses from the API server.\n request_commands = factory.CommandFactory.create_dynamic_objects(\n evidence_mapping=ctx.obj.evidence_mapping,\n request_options=ctx.obj.request_options)\n for command in request_commands:\n groups.submit_group.add_command(command)","def main():\n\n run_manual_session()\n # run_automated_session()","def start(self):\n self.start_time = dt.datetime.now()\n self.call = ' '.join(sys.argv)\n self.commands = []","def launch_cli() -> None:\n app.run(main, flags_parser=_parse_flags)","def cli():\n pass","def cli():\n return","def _cli():\n pass","def cli():\n logger.debug('cli() called')","def _spawn_cli() -> GooeyParser:\n parser = _add_gui_arg(_cli())\n\n return vars(parser.parse_args())","def use_karaf():\n # type: () -> subprocess.Popen\n karaf_dir = os.environ.get(\"KARAF_DIR\")\n\n # Start Karaf\n start = time.time()\n install_karaf(karaf_dir)\n print(\"Karaf installed in\", round(time.time() - start, 3), \"s\")\n karaf_root = find_karaf_root(karaf_dir)\n\n start = time.time()\n with start_karaf(karaf_root) as karaf:\n # Wait for Karaf to start\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- Karaf started\")\n\n # Add the ECF repository\n karaf.stdin.write(b\"feature:repo-add ecf\\n\")\n karaf.stdin.flush()\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- ECF repository added\")\n\n # Install the tutorial sample\n karaf.stdin.write(\n b\"feature:install -v ecf-rs-examples-python.java-hello\\n\"\n )\n karaf.stdin.flush()\n wait_for_prompt(karaf)\n print(round(time.time() - start, 3), \"- Feature installed\")\n\n # Give hand to the caller\n try:\n yield karaf\n finally:\n # Exit Karaf\n karaf.stdin.write(b\"logout\\n\")","def cli() -> None:","def cli() -> None:","def spawn(cls, *args, **kwargs):\n g = cls(*args, **kwargs)\n g.start()\n return g","def monkeypatch_spawn():\n ccompiler.CCompiler.spawn = _commandfile_spawn","def spawn():\n if platform.system() == \"Windows\":\n # HACK https://github.com/prompt-toolkit/python-prompt-toolkit/issues/1243#issuecomment-706668723\n # FIXME Use pexpect or wexpect somehow to fix this\n pytest.xfail(\n \"pexpect fails on Windows\",\n )\n # Using PopenSpawn, although probably it would be best to use pexpect.spawn\n # instead. However, it's working fine and it seems easier to fix in the\n # future to work on Windows (where, this way, spawning actually works; it's just\n # python-prompt-toolkit that rejects displaying a TUI)\n return PopenSpawn","def run():\n if len(sys.argv)<2:\n sys.argv.append(\"-h\")\n plac.call(main)","def new_session(self):\n self.command(\"new\")","def start( self ):\n pathCheck( self.command )\n cout = '/tmp/' + self.name + '.log'\n if self.cdir is not None:\n self.cmd( 'cd ' + self.cdir )\n self.cmd( self.command + ' ' + self.cargs % self.port +\n ' 1>' + cout + ' 2>' + cout + '&' )\n self.execed = False","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def cli() -> None:\n pass","def spawn(mystring, mysettings, debug=0, free=0, droppriv=0, sesandbox=0, fakeroot=0, **keywords):\n\n\tcheck_config_instance(mysettings)\n\n\tfd_pipes = keywords.get(\"fd_pipes\")\n\tif fd_pipes is None:\n\t\tfd_pipes = {\n\t\t\t0:sys.__stdin__.fileno(),\n\t\t\t1:sys.__stdout__.fileno(),\n\t\t\t2:sys.__stderr__.fileno(),\n\t\t}\n\t# In some cases the above print statements don't flush stdout, so\n\t# it needs to be flushed before allowing a child process to use it\n\t# so that output always shows in the correct order.\n\tstdout_filenos = (sys.__stdout__.fileno(), sys.__stderr__.fileno())\n\tfor fd in fd_pipes.values():\n\t\tif fd in stdout_filenos:\n\t\t\tsys.__stdout__.flush()\n\t\t\tsys.__stderr__.flush()\n\t\t\tbreak\n\n\tfeatures = mysettings.features\n\t# TODO: Enable fakeroot to be used together with droppriv. The\n\t# fake ownership/permissions will have to be converted to real\n\t# permissions in the merge phase.\n\tfakeroot = fakeroot and uid != 0 and portage.process.fakeroot_capable\n\tif droppriv and uid == 0 and portage_gid and portage_uid and \\\n\t\thasattr(os, \"setgroups\"):\n\t\tkeywords.update({\"uid\":portage_uid,\"gid\":portage_gid,\n\t\t\t\"groups\":userpriv_groups,\"umask\":0o02})\n\tif not free:\n\t\tfree=((droppriv and \"usersandbox\" not in features) or \\\n\t\t\t(not droppriv and \"sandbox\" not in features and \\\n\t\t\t\"usersandbox\" not in features and not fakeroot))\n\n\tif not free and not (fakeroot or portage.process.sandbox_capable):\n\t\tfree = True\n\n\tif mysettings.mycpv is not None:\n\t\tkeywords[\"opt_name\"] = \"[%s]\" % mysettings.mycpv\n\telse:\n\t\tkeywords[\"opt_name\"] = \"[%s/%s]\" % \\\n\t\t\t(mysettings.get(\"CATEGORY\",\"\"), mysettings.get(\"PF\",\"\"))\n\n\tif free or \"SANDBOX_ACTIVE\" in os.environ:\n\t\tkeywords[\"opt_name\"] += \" bash\"\n\t\tspawn_func = portage.process.spawn_bash\n\telif fakeroot:\n\t\tkeywords[\"opt_name\"] += \" fakeroot\"\n\t\tkeywords[\"fakeroot_state\"] = os.path.join(mysettings[\"T\"], \"fakeroot.state\")\n\t\tspawn_func = portage.process.spawn_fakeroot\n\telse:\n\t\tkeywords[\"opt_name\"] += \" sandbox\"\n\t\tspawn_func = portage.process.spawn_sandbox\n\n\tif sesandbox:\n\t\tspawn_func = selinux.spawn_wrapper(spawn_func,\n\t\t\tmysettings[\"PORTAGE_SANDBOX_T\"])\n\n\tif keywords.get(\"returnpid\"):\n\t\treturn spawn_func(mystring, env=mysettings.environ(), **keywords)\n\n\tproc = EbuildSpawnProcess(\n\t\tbackground=False, args=mystring,\n\t\tscheduler=SchedulerInterface(EventLoop(main=False)),\n\t\tspawn_func=spawn_func,\n\t\tsettings=mysettings, **keywords)\n\n\tproc.start()\n\tproc.wait()\n\n\treturn proc.returncode","def main():\n\tcli = Cli()\n\tcli.run()","def __init__(self, command, args, prompt, parent=None, timeout=30,context=None ):\n if parent:\n self.spawnProc = parent.spawnProc\n self.isChild = True\n else:\n self.isChild = False\n\t \n self.context = context\n self.command = command\n\n self.args = args\n self.prompt = prompt\n self.sshTimeout = int ( globalVar.sshTimeout)\n if (timeout != 0) :\n self.sshTimeout = timeout\n \n try:\n self._connect() \n # sleeping 2 second to give control to other processes\n # This is done to resolve EOF problem\n time.sleep(2)\n\n # Checking if the spawned process lives. It is observed that the\n # spawned process for gen dies because of TIME_WAIT problem \n # If the process had died, sleep for 60 secs so that the socket\n # is closed or binding is released (BSD implementation) \n # Try spawning the process again, this time it should get thru.\n\n if not self.spawnProc.isalive():\n time.sleep(60)\n self._connect()\n \n except pexpect.EOF:\n if self.spawnProc.isalive():\n raise SessionManagerException(\"error starting %s\" % command)\n else:\n self._postConnect()\n except Exception, e:\n msg = \"session error: %s\" % e\n self.disconnect()\n raise e","def _launch(self):\n annotators = ['tokenize', 'ssplit']\n if 'ner' in self.annotators:\n annotators.extend(['pos', 'lemma', 'ner'])\n elif 'lemma' in self.annotators:\n annotators.extend(['pos', 'lemma'])\n elif 'pos' in self.annotators:\n annotators.extend(['pos'])\n annotators = ','.join(annotators)\n options = ','.join(['untokenizable=noneDelete',\n 'invertible=true'])\n # if you work on English, use this this command\n cmd = ['java', '-mx' + self.mem, '-cp', '\"%s\"' % self.classpath,\n 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-annotators',\n annotators, '-tokenize.options', options,\n '-outputFormat', 'json', '-prettyPrint', 'false']\n \n # if you work on arabic, use this this command\n \n # cmd = ['java', '-mx' + self.mem, '-cp', '\"%s\"' % self.classpath,\n # # 'edu.stanford.nlp.pipeline.StanfordCoreNLP','-annotators',\n # 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-props', 'StanfordCoreNLP-arabic.properties','-annotators',\n # annotators, '-tokenize.options', options, #'-tokenize.whitespace', 'true',\n # '-outputFormat', 'json', '-prettyPrint', 'false']\n print(' '.join(cmd))\n\n # We use pexpect to keep the subprocess alive and feed it commands.\n # Because we don't want to get hit by the max terminal buffer size,\n # we turn off canonical input processing to have unlimited bytes.\n self.corenlp = pexpect.spawn('/bin/bash', maxread=100000, timeout=60)\n self.corenlp.setecho(False)\n self.corenlp.sendline('stty -icanon')\n self.corenlp.sendline(' '.join(cmd))\n self.corenlp.delaybeforesend = 0\n self.corenlp.delayafterread = 0\n self.corenlp.expect_exact('NLP>', searchwindowsize=100)"],"string":"[\n \"def cli():\\n config, auth, execute_now = read_command_line_arguments()\\n main(config, auth, execute_now)\",\n \"def main():\\n # Parse arguments.\\n parser = argparse.ArgumentParser(description='Console for cluster test.')\\n parser.add_argument('-c', dest='command_config_file',\\n default='command_config',\\n help='set command config file path (default=command_config)')\\n args = parser.parse_args()\\n\\n console = Console()\\n console.config(args.command_config_file)\\n console.init_proxies()\\n console.run()\",\n \"def __init__(self, name='fricas', command='fricas -nox -noclef',\\n script_subdirectory=None, logfile=None,\\n server=None, server_tmpdir=None,\\n init_code=[')lisp (si::readline-off)']):\\n PanAxiom.__init__(self, name, command,\\n script_subdirectory, logfile,\\n server, server_tmpdir,\\n init_code)\",\n \"def Spawn(proc):\\n proc.start()\\n return proc\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\",\n \"def cli():\\r\\n pass\",\n \"def cli():\\n ...\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n pass\",\n \"def cli():\\n\\n pass\",\n \"def spawn(self):\\n self._proc = subprocess.Popen(\\n self._args, stdout=subprocess.PIPE, stderr=subprocess.PIPE\\n )\",\n \"def main_cli():\\n pass\",\n \"def cli(ctx: click.Context, config_instance: str, config_path: str) -> None:\\n ctx.obj = TurbiniaCli(\\n config_instance=config_instance, config_path=config_path)\\n\\n # Set up the tool based on the configuration file parameters.\\n ctx.obj.setup()\\n\\n # Build all the commands based on responses from the API server.\\n request_commands = factory.CommandFactory.create_dynamic_objects(\\n evidence_mapping=ctx.obj.evidence_mapping,\\n request_options=ctx.obj.request_options)\\n for command in request_commands:\\n groups.submit_group.add_command(command)\",\n \"def main():\\n\\n run_manual_session()\\n # run_automated_session()\",\n \"def start(self):\\n self.start_time = dt.datetime.now()\\n self.call = ' '.join(sys.argv)\\n self.commands = []\",\n \"def launch_cli() -> None:\\n app.run(main, flags_parser=_parse_flags)\",\n \"def cli():\\n pass\",\n \"def cli():\\n return\",\n \"def _cli():\\n pass\",\n \"def cli():\\n logger.debug('cli() called')\",\n \"def _spawn_cli() -> GooeyParser:\\n parser = _add_gui_arg(_cli())\\n\\n return vars(parser.parse_args())\",\n \"def use_karaf():\\n # type: () -> subprocess.Popen\\n karaf_dir = os.environ.get(\\\"KARAF_DIR\\\")\\n\\n # Start Karaf\\n start = time.time()\\n install_karaf(karaf_dir)\\n print(\\\"Karaf installed in\\\", round(time.time() - start, 3), \\\"s\\\")\\n karaf_root = find_karaf_root(karaf_dir)\\n\\n start = time.time()\\n with start_karaf(karaf_root) as karaf:\\n # Wait for Karaf to start\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- Karaf started\\\")\\n\\n # Add the ECF repository\\n karaf.stdin.write(b\\\"feature:repo-add ecf\\\\n\\\")\\n karaf.stdin.flush()\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- ECF repository added\\\")\\n\\n # Install the tutorial sample\\n karaf.stdin.write(\\n b\\\"feature:install -v ecf-rs-examples-python.java-hello\\\\n\\\"\\n )\\n karaf.stdin.flush()\\n wait_for_prompt(karaf)\\n print(round(time.time() - start, 3), \\\"- Feature installed\\\")\\n\\n # Give hand to the caller\\n try:\\n yield karaf\\n finally:\\n # Exit Karaf\\n karaf.stdin.write(b\\\"logout\\\\n\\\")\",\n \"def cli() -> None:\",\n \"def cli() -> None:\",\n \"def spawn(cls, *args, **kwargs):\\n g = cls(*args, **kwargs)\\n g.start()\\n return g\",\n \"def monkeypatch_spawn():\\n ccompiler.CCompiler.spawn = _commandfile_spawn\",\n \"def spawn():\\n if platform.system() == \\\"Windows\\\":\\n # HACK https://github.com/prompt-toolkit/python-prompt-toolkit/issues/1243#issuecomment-706668723\\n # FIXME Use pexpect or wexpect somehow to fix this\\n pytest.xfail(\\n \\\"pexpect fails on Windows\\\",\\n )\\n # Using PopenSpawn, although probably it would be best to use pexpect.spawn\\n # instead. However, it's working fine and it seems easier to fix in the\\n # future to work on Windows (where, this way, spawning actually works; it's just\\n # python-prompt-toolkit that rejects displaying a TUI)\\n return PopenSpawn\",\n \"def run():\\n if len(sys.argv)<2:\\n sys.argv.append(\\\"-h\\\")\\n plac.call(main)\",\n \"def new_session(self):\\n self.command(\\\"new\\\")\",\n \"def start( self ):\\n pathCheck( self.command )\\n cout = '/tmp/' + self.name + '.log'\\n if self.cdir is not None:\\n self.cmd( 'cd ' + self.cdir )\\n self.cmd( self.command + ' ' + self.cargs % self.port +\\n ' 1>' + cout + ' 2>' + cout + '&' )\\n self.execed = False\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def cli() -> None:\\n pass\",\n \"def spawn(mystring, mysettings, debug=0, free=0, droppriv=0, sesandbox=0, fakeroot=0, **keywords):\\n\\n\\tcheck_config_instance(mysettings)\\n\\n\\tfd_pipes = keywords.get(\\\"fd_pipes\\\")\\n\\tif fd_pipes is None:\\n\\t\\tfd_pipes = {\\n\\t\\t\\t0:sys.__stdin__.fileno(),\\n\\t\\t\\t1:sys.__stdout__.fileno(),\\n\\t\\t\\t2:sys.__stderr__.fileno(),\\n\\t\\t}\\n\\t# In some cases the above print statements don't flush stdout, so\\n\\t# it needs to be flushed before allowing a child process to use it\\n\\t# so that output always shows in the correct order.\\n\\tstdout_filenos = (sys.__stdout__.fileno(), sys.__stderr__.fileno())\\n\\tfor fd in fd_pipes.values():\\n\\t\\tif fd in stdout_filenos:\\n\\t\\t\\tsys.__stdout__.flush()\\n\\t\\t\\tsys.__stderr__.flush()\\n\\t\\t\\tbreak\\n\\n\\tfeatures = mysettings.features\\n\\t# TODO: Enable fakeroot to be used together with droppriv. The\\n\\t# fake ownership/permissions will have to be converted to real\\n\\t# permissions in the merge phase.\\n\\tfakeroot = fakeroot and uid != 0 and portage.process.fakeroot_capable\\n\\tif droppriv and uid == 0 and portage_gid and portage_uid and \\\\\\n\\t\\thasattr(os, \\\"setgroups\\\"):\\n\\t\\tkeywords.update({\\\"uid\\\":portage_uid,\\\"gid\\\":portage_gid,\\n\\t\\t\\t\\\"groups\\\":userpriv_groups,\\\"umask\\\":0o02})\\n\\tif not free:\\n\\t\\tfree=((droppriv and \\\"usersandbox\\\" not in features) or \\\\\\n\\t\\t\\t(not droppriv and \\\"sandbox\\\" not in features and \\\\\\n\\t\\t\\t\\\"usersandbox\\\" not in features and not fakeroot))\\n\\n\\tif not free and not (fakeroot or portage.process.sandbox_capable):\\n\\t\\tfree = True\\n\\n\\tif mysettings.mycpv is not None:\\n\\t\\tkeywords[\\\"opt_name\\\"] = \\\"[%s]\\\" % mysettings.mycpv\\n\\telse:\\n\\t\\tkeywords[\\\"opt_name\\\"] = \\\"[%s/%s]\\\" % \\\\\\n\\t\\t\\t(mysettings.get(\\\"CATEGORY\\\",\\\"\\\"), mysettings.get(\\\"PF\\\",\\\"\\\"))\\n\\n\\tif free or \\\"SANDBOX_ACTIVE\\\" in os.environ:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" bash\\\"\\n\\t\\tspawn_func = portage.process.spawn_bash\\n\\telif fakeroot:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" fakeroot\\\"\\n\\t\\tkeywords[\\\"fakeroot_state\\\"] = os.path.join(mysettings[\\\"T\\\"], \\\"fakeroot.state\\\")\\n\\t\\tspawn_func = portage.process.spawn_fakeroot\\n\\telse:\\n\\t\\tkeywords[\\\"opt_name\\\"] += \\\" sandbox\\\"\\n\\t\\tspawn_func = portage.process.spawn_sandbox\\n\\n\\tif sesandbox:\\n\\t\\tspawn_func = selinux.spawn_wrapper(spawn_func,\\n\\t\\t\\tmysettings[\\\"PORTAGE_SANDBOX_T\\\"])\\n\\n\\tif keywords.get(\\\"returnpid\\\"):\\n\\t\\treturn spawn_func(mystring, env=mysettings.environ(), **keywords)\\n\\n\\tproc = EbuildSpawnProcess(\\n\\t\\tbackground=False, args=mystring,\\n\\t\\tscheduler=SchedulerInterface(EventLoop(main=False)),\\n\\t\\tspawn_func=spawn_func,\\n\\t\\tsettings=mysettings, **keywords)\\n\\n\\tproc.start()\\n\\tproc.wait()\\n\\n\\treturn proc.returncode\",\n \"def main():\\n\\tcli = Cli()\\n\\tcli.run()\",\n \"def __init__(self, command, args, prompt, parent=None, timeout=30,context=None ):\\n if parent:\\n self.spawnProc = parent.spawnProc\\n self.isChild = True\\n else:\\n self.isChild = False\\n\\t \\n self.context = context\\n self.command = command\\n\\n self.args = args\\n self.prompt = prompt\\n self.sshTimeout = int ( globalVar.sshTimeout)\\n if (timeout != 0) :\\n self.sshTimeout = timeout\\n \\n try:\\n self._connect() \\n # sleeping 2 second to give control to other processes\\n # This is done to resolve EOF problem\\n time.sleep(2)\\n\\n # Checking if the spawned process lives. It is observed that the\\n # spawned process for gen dies because of TIME_WAIT problem \\n # If the process had died, sleep for 60 secs so that the socket\\n # is closed or binding is released (BSD implementation) \\n # Try spawning the process again, this time it should get thru.\\n\\n if not self.spawnProc.isalive():\\n time.sleep(60)\\n self._connect()\\n \\n except pexpect.EOF:\\n if self.spawnProc.isalive():\\n raise SessionManagerException(\\\"error starting %s\\\" % command)\\n else:\\n self._postConnect()\\n except Exception, e:\\n msg = \\\"session error: %s\\\" % e\\n self.disconnect()\\n raise e\",\n \"def _launch(self):\\n annotators = ['tokenize', 'ssplit']\\n if 'ner' in self.annotators:\\n annotators.extend(['pos', 'lemma', 'ner'])\\n elif 'lemma' in self.annotators:\\n annotators.extend(['pos', 'lemma'])\\n elif 'pos' in self.annotators:\\n annotators.extend(['pos'])\\n annotators = ','.join(annotators)\\n options = ','.join(['untokenizable=noneDelete',\\n 'invertible=true'])\\n # if you work on English, use this this command\\n cmd = ['java', '-mx' + self.mem, '-cp', '\\\"%s\\\"' % self.classpath,\\n 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-annotators',\\n annotators, '-tokenize.options', options,\\n '-outputFormat', 'json', '-prettyPrint', 'false']\\n \\n # if you work on arabic, use this this command\\n \\n # cmd = ['java', '-mx' + self.mem, '-cp', '\\\"%s\\\"' % self.classpath,\\n # # 'edu.stanford.nlp.pipeline.StanfordCoreNLP','-annotators',\\n # 'edu.stanford.nlp.pipeline.StanfordCoreNLP', '-props', 'StanfordCoreNLP-arabic.properties','-annotators',\\n # annotators, '-tokenize.options', options, #'-tokenize.whitespace', 'true',\\n # '-outputFormat', 'json', '-prettyPrint', 'false']\\n print(' '.join(cmd))\\n\\n # We use pexpect to keep the subprocess alive and feed it commands.\\n # Because we don't want to get hit by the max terminal buffer size,\\n # we turn off canonical input processing to have unlimited bytes.\\n self.corenlp = pexpect.spawn('/bin/bash', maxread=100000, timeout=60)\\n self.corenlp.setecho(False)\\n self.corenlp.sendline('stty -icanon')\\n self.corenlp.sendline(' '.join(cmd))\\n self.corenlp.delaybeforesend = 0\\n self.corenlp.delayafterread = 0\\n self.corenlp.expect_exact('NLP>', searchwindowsize=100)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.60068893","0.5865383","0.5834904","0.58237004","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5808352","0.5797985","0.5786446","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.5762459","0.57515025","0.5723277","0.5705709","0.57052696","0.57039106","0.56917137","0.56841964","0.56743085","0.56502104","0.56183946","0.55473167","0.5537812","0.553661","0.55204964","0.55204964","0.54925025","0.5487892","0.5433574","0.54125625","0.54105437","0.5393246","0.53773993","0.53773993","0.53773993","0.53773993","0.53773993","0.53486776","0.5340044","0.5328309","0.5316158"],"string":"[\n \"0.60068893\",\n \"0.5865383\",\n \"0.5834904\",\n \"0.58237004\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5808352\",\n \"0.5797985\",\n \"0.5786446\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.5762459\",\n \"0.57515025\",\n \"0.5723277\",\n \"0.5705709\",\n \"0.57052696\",\n \"0.57039106\",\n \"0.56917137\",\n \"0.56841964\",\n \"0.56743085\",\n \"0.56502104\",\n \"0.56183946\",\n \"0.55473167\",\n \"0.5537812\",\n \"0.553661\",\n \"0.55204964\",\n \"0.55204964\",\n \"0.54925025\",\n \"0.5487892\",\n \"0.5433574\",\n \"0.54125625\",\n \"0.54105437\",\n \"0.5393246\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53773993\",\n \"0.53486776\",\n \"0.5340044\",\n \"0.5328309\",\n \"0.5316158\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90896,"cells":{"query":{"kind":"string","value":"optimise the order of matrixchain multiplication"},"document":{"kind":"string","value":"def MatMulOrder(D):\r\n\tnum = len(D)-1 # number of matrix in the chain\r\n\tprint(f\"There are {num} matrix to multiply\")\r\n\tM = [[0 for _ in range(num)] for _ in range(num)]\r\n\tP = [[0 for _ in range(num)] for _ in range(num)]\r\n\r\n\t# i要从大到小\r\n\t# i == j时, M[i][j]=0,所以不用更新\r\n\t# i-th矩阵到j-th矩阵的乘的最优值初始化为inf\r\n\tfor i in range(num-2, -1, -1):\r\n\t\tfor j in range(i+1, num):\r\n\t\t\tM[i][j] = 100000000\r\n\t\t\tfor k in range(i, j):\r\n\t\t\t\tnew = M[i][k] + M[k+1][j] + D[i]*D[k+1]*D[j+1]\r\n\t\t\t\tif new < M[i][j]:\r\n\t\t\t\t\tM[i][j] = new \r\n\t\t\t\t\tP[i][j] = k\r\n\treturn M, P"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def _multi_matmul_chain_order(arrays):\n n = len(arrays)\n # p stores the dimensions of the matrices\n # Example for p: A_{10x100}, B_{100x5}, C_{5x50} --> p = [10, 100, 5, 50]\n # Using -2 to generalize for shapes that are more than 2 dimmensions\n p = [a.shape[-2] for a in arrays] + [arrays[-1].shape[-1]]\n # m is a matrix of costs of the subproblems\n # m[i,j]: min number of scalar multiplications needed to compute A_{i..j}\n m = np.zeros((n, n), dtype=np.double)\n # s is the actual ordering\n # s[i, j] is the value of k at which we split the product A_i..A_j\n s = np.empty((n, n), dtype=np.intp)\n\n for l in range(1, n):\n for i in range(n - l):\n j = i + l\n m[i, j] = np.inf\n for k in range(i, j):\n q = m[i, k] + m[k + 1, j] + p[i] * p[k + 1] * p[j + 1]\n if q < m[i, j]:\n m[i, j] = q\n s[i, j] = k # Note that Cormen uses 1-based index\n return s","def chain_matmul_square(As):\n\n As_matmul = As\n while As_matmul.shape[0] > 1:\n if As_matmul.shape[0] % 2:\n A_last = As_matmul[-1:]\n else:\n A_last = None\n \n As_matmul = torch.matmul(As_matmul[0:-1:2], As_matmul[1::2])\n if A_last is not None:\n As_matmul = torch.cat([As_matmul, A_last], dim=0)\n \n return As_matmul.squeeze(0)","def recursive_multiply(a, b):\n if len(a) == 2:\n return naive_multiply(a, b)\n\n a11 = a[0:int(len(a) / 2)]\n for index, row in enumerate(a11):\n a11[index] = row[0:int(len(row) / 2)]\n\n a12 = a[0:int(len(a) / 2)]\n for index, row in enumerate(a12):\n a12[index] = row[int(len(a) / 2):len(a)]\n\n a21 = a[int(len(a) / 2):len(a)]\n for index, row in enumerate(a21):\n a21[index] = row[0:int(len(row) / 2)]\n\n a22 = a[int(len(a) / 2):len(a)]\n for index, row in enumerate(a22):\n a22[index] = row[int(len(a) / 2):len(a)]\n\n b11 = b[0:int(len(b) / 2)]\n for index, row in enumerate(b11):\n b11[index] = row[0:int(len(row) / 2)]\n\n b12 = b[0:int(len(b) / 2)]\n for index, row in enumerate(b12):\n b12[index] = row[int(len(b) / 2):len(b)]\n\n b21 = b[int(len(b) / 2):len(b)]\n for index, row in enumerate(b21):\n b21[index] = row[0:int(len(row) / 2)]\n\n b22 = b[int(len(b) / 2):len(b)]\n for index, row in enumerate(b22):\n b22[index] = row[int(len(b) / 2):len(b)]\n\n c11 = matrix_add(recursive_multiply(a11, b11), recursive_multiply(a12, b21)) # C11 = A11*B11 + A12*B21\n c12 = matrix_add(recursive_multiply(a11, b12), recursive_multiply(a12, b22)) # C12 = A11*B12 + A12*B22\n c21 = matrix_add(recursive_multiply(a21, b11), recursive_multiply(a22, b21)) # C21 = A21*B11 + A22*B21\n c22 = matrix_add(recursive_multiply(a21, b12), recursive_multiply(a22, b22)) # C22 = A21*B12 + A22*B22\n\n # Append c12 to c11\n for row_index, row in enumerate(c11):\n for col_index, col in enumerate(c12):\n row.append(c12[row_index][col_index])\n\n # Append c22 to c21\n for row_index, row in enumerate(c21):\n for col_index, col in enumerate(c12):\n row.append(c22[row_index][col_index])\n\n # Append c21 to c11\n for i in c21:\n c11.append(i)\n\n return c11","def lazy_matrix_mul(m_a, m_b):\n return np.dot(m_a, m_b)","def __mul__(self, other):\n #\n # TODO - your code here\n #\n final_matrix = []\n for i in range(self.h):\n temp_row = []\n for j in range(other.w):\n # take dot-product of row of\n # matrix in 1st arg with col of\n # matrix in 2nd arg\n temp_row.append(dot_product(get_row(self.g, i), get_col(other.g, j)))\n final_matrix.append(temp_row)\n return Matrix(final_matrix)\n # TODO - your code here","def matmul():\n\n if RESULT_IN_NVRAM:\n matrix_c = ResultMatrixInDaos()\n else:\n matrix_c = ResultMatrixInMemory()\n\n # This could be trivially optimized by reordering indexes\n # and caching either a_block or b_block (assuming C in-memory).\n # *However* it would result in unfair comparisons with the \n # previous implementation used elsewhere.\n # Using the naive algorithm makes sense for a raw comparison.\n for i in range(MATRIXSIZE):\n for j in range(MATRIXSIZE):\n partial_result_block = np.zeros((BLOCKSIZE, BLOCKSIZE))\n\n for k in range(MATRIXSIZE):\n a_block = np.fromstring(\n DAOS_KV[\"A%02d%02d\" % (i, k)],\n dtype=NP_FROMSTRING_DTYPE\n ).reshape((BLOCKSIZE, BLOCKSIZE))\n\n b_block = np.fromstring(\n DAOS_KV[\"B%02d%02d\" % (k, j)],\n dtype=NP_FROMSTRING_DTYPE\n ).reshape((BLOCKSIZE, BLOCKSIZE))\n\n partial_result_block += a_block @ b_block\n \n matrix_c[i,j] = partial_result_block\n\n return matrix_c","def lazy_matrix_mul(m_a, m_b):\n return (np.matmul(m_a, m_b))","def matrix_mult(m1, m2):\n pass","def _multi_matmul(arrays, order, i, j, constant=False) -> Tensor:\n if i == j:\n return arrays[i]\n else:\n return matmul(\n _multi_matmul(arrays, order, i, order[i, j], constant),\n _multi_matmul(arrays, order, order[i, j] + 1, j, constant),\n constant,\n )","def lazy_matrix_mul(m_a, m_b):\n return np.matmul(np.array(m_a), np.array(m_b))","def matrixMul(self, matrix, matrix2):\n matrix0 = matrix[:]\n matrix[0] = matrix0[0] * matrix2[0] + matrix0[2]*matrix2[1] # + matrix0[4]*0\n matrix[1] = matrix0[1] * matrix2[0] + matrix0[3]*matrix2[1] # + matrix0[5]*0\n matrix[2] = matrix0[0] * matrix2[2] + matrix0[2]*matrix2[3] # + matrix0[4]*0\n matrix[3] = matrix0[1] * matrix2[2] + matrix0[3]*matrix2[3] # + matrix0[5]*0\n matrix[4] = matrix0[0] * matrix2[4] + matrix0[2]*matrix2[5] + matrix0[4]\n matrix[5] = matrix0[1] * matrix2[4] + matrix0[3]*matrix2[5] + matrix0[5]","def lazy_matrix_mul(m_a, m_b):\n m_a = np.array(m_a)\n m_b = np.array(m_b)\n\n return m_a.dot(m_b)","def Multiply(M1,M2):\r\n M3=[]\r\n w=0\r\n while w<len(M2[0]):\r\n tap=[]\r\n t=0\r\n while t<len(M2):\r\n tap.append(M2[t][w])\r\n t=t+1\r\n M3.append(tap)\r\n w=w+1\r\n M=[]\r\n # Multiplying matrices\r\n k=0\r\n sums=0\r\n while k<len(M1):\r\n j=0\r\n mpy=[]\r\n while j<len(M3):\r\n p=0\r\n sums=0\r\n while p<len(M3[j]):\r\n temp = (M1[k][p])*(M3[j][p])\r\n sums=sums+temp\r\n p=p+1\r\n mpy.append(sums)\r\n j=j+1\r\n M.append(mpy)\r\n k=k+1\r\n return M","def _z2matmul(self, left, right):\n prod = np.mod(np.dot(left, right), 2)\n return prod","def __matmul__(self, q: np.ndarray) -> np.ndarray:\n return self.product(q)","def matrix_chain_dynamic(dimensions, n):\n\n m = [[-1 for _ in range(n)] for _ in range(n)]\n s = [[0 for _ in range(n)] for _ in range(n)]\n\n # multiplying matrix by itself\n for i in range(1, n):\n m[i][i] = 0\n\n for length in range(2, n):\n for i in range(1, n - length + 1):\n j = i + length - 1\n for k in range(i, j):\n cost = m[i][k] + m[k + 1][j] + dimensions[i - 1] * dimensions[k] * dimensions[j]\n if cost > m[i][j]:\n m[i][j] = cost\n # index if splitting\n s[i][j] = k\n return m, s","def combined_costs(matrix_MSLL_IO):\r\n return","def __matmul__(self, tensor):\n return self.matmul(tensor)","def matrix_mult(M, vector1, vector2):\n out = None\n ### YOUR CODE HERE\n pass\n ### END YOUR CODE\n\n return out","def matrix_mult(M, vector1, vector2):\n out = None\n ### YOUR CODE HERE\n pass\n ### END YOUR CODE\n\n return out","def Mult(A, B, C_, IM, M):\n for i in range(M):\n for j in range(M):\n for k in range(M):\n C_[i, j] = A[i, j] * B[i, j]","def __mul__(self, other):\r\n T = type(other)\r\n # mat4*scalar\r\n if T==types.FloatType or T==types.IntType or T==types.LongType:\r\n return mat4(map(lambda x,other=other: x*other, self.mlist))\r\n # mat4*vec3\r\n if isinstance(other, _vec3):\r\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\r\n w = float(m41*other.x + m42*other.y + m43*other.z + m44)\r\n return _vec3(m11*other.x + m12*other.y + m13*other.z + m14, \r\n m21*other.x + m22*other.y + m23*other.z + m24, \r\n m31*other.x + m32*other.y + m33*other.z + m34)/w\r\n # mat4*vec4\r\n if isinstance(other, _vec4):\r\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\r\n return _vec4(m11*other.x + m12*other.y + m13*other.z + m14*other.w, \r\n m21*other.x + m22*other.y + m23*other.z + m24*other.w, \r\n m31*other.x + m32*other.y + m33*other.z + m34*other.w,\r\n m41*other.x + m42*other.y + m43*other.z + m44*other.w)\r\n # mat4*mat4\r\n if isinstance(other, mat4):\r\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\r\n n11,n12,n13,n14,n21,n22,n23,n24,n31,n32,n33,n34,n41,n42,n43,n44 = other.mlist\r\n return mat4( m11*n11+m12*n21+m13*n31+m14*n41,\r\n m11*n12+m12*n22+m13*n32+m14*n42,\r\n m11*n13+m12*n23+m13*n33+m14*n43,\r\n m11*n14+m12*n24+m13*n34+m14*n44,\r\n\r\n m21*n11+m22*n21+m23*n31+m24*n41,\r\n m21*n12+m22*n22+m23*n32+m24*n42,\r\n m21*n13+m22*n23+m23*n33+m24*n43,\r\n m21*n14+m22*n24+m23*n34+m24*n44,\r\n\r\n m31*n11+m32*n21+m33*n31+m34*n41,\r\n m31*n12+m32*n22+m33*n32+m34*n42,\r\n m31*n13+m32*n23+m33*n33+m34*n43,\r\n m31*n14+m32*n24+m33*n34+m34*n44,\r\n\r\n m41*n11+m42*n21+m43*n31+m44*n41,\r\n m41*n12+m42*n22+m43*n32+m44*n42,\r\n m41*n13+m42*n23+m43*n33+m44*n43,\r\n m41*n14+m42*n24+m43*n34+m44*n44)\r\n # unsupported\r\n else:\r\n raise TypeError, \"unsupported operand type for *\"","def fast_matmul_2d(self, other):\n if isinstance(other, np.ndarray):\n mat_tensor = PaillierTensor(other, partitions=self.partitions)\n return self.fast_matmul_2d(mat_tensor)\n\n if isinstance(other, CTableABC):\n other = PaillierTensor(other)\n\n func = self._vector_mul\n ret_mat = self._obj.join(other.get_obj(), lambda vec1, vec2: (vec1, vec2)).applyPartitions(func).reduce(\n lambda mat1, mat2: mat1 + mat2)\n\n return ret_mat","def mmultiply(self, matrix):\n try:\n result_matrix = [[0 for row in range(len(self.matrix))] for col in range(len(matrix[0]))]\n for i in range(len(self.matrix)):\n for j in range(len(matrix[0])):\n for k in range(len(matrix)):\n result_matrix[i][j] += self.matrix[i][k] * matrix[k][j]\n self.matrix = result_matrix\n except IndexError:\n pass\n pass","def test_matrix_product(self, use_cache):\n\n key = jrandom.PRNGKey(0)\n dim = 50\n max_power = 25\n\n matrix = jrandom.normal(key, (dim, dim)) / 10\n vector = jnp.ones((dim,), dtype=jnp.float32)\n\n if use_cache:\n mpstate = model_utils.CachedMatrixPowerState.precompute(matrix, max_power)\n else:\n mpstate = model_utils.LazyMatrixPowerState(matrix)\n\n for t in range(max_power):\n result = mpstate.matrix_power_multiply(vector, t)\n expected = np.linalg.matrix_power(matrix, t) @ vector\n\n np.testing.assert_array_almost_equal(result, expected, decimal=1)","def my_matmul(x, y):\n ##\n cmd = getattr(th, \"matmul\")\n x1, x2 = my_cut(x)\n y1, y2 = my_cut(y)\n x2y1 = cmd(x2, y1)\n x1y2 = cmd(x1, y2)\n x2y2 = cmd(x2, y2)\n ret = (x2y1 + x1y2) % int24field * int24field + x2y2\n ret = int48module(ret)\n return ret","def multiply_matrices(list):\n # Section 1: Start matrix product using 1st matrix in list\n matrix_product = list[0]\n\n # Section 2: Loop thru list to create product\n for matrix in list[1:]:\n matrix_product = matrix_multiply(matrix_product, matrix)\n\n return matrix_product","def mat_mul(mat1, mat2):\n\n if len(mat1[0]) == len(mat2):\n\n mat2 = matrix_transpose(mat2)\n response = []\n\n for row in range(len(mat1)):\n response.append(\n [\n sum(dot_product(mat1[row], mat2[column]))\n for column in range(len(mat2))\n ]\n )\n\n return response\n\n else:\n return None","def u_update(eta_0, eta, eta_lin, w_0, w, w_lin, eta_T_H_L_stacked, premultiplied_lhs = None, nnls_max_iter=50): \n # PREMULTIPLIED LHS IS AN EXTRA ARGUMENT! Set it to None and add solver! \n \"\"\"In the following +[[]] and [:-1] are added to keep thing 1dim array of objects and still multiply it elemtwisely\"\"\" \n# #B.append([]) #THIS IS WRONG, CHANGES THE LIST \n# B_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(B+[[]])[:-1], axis = 0) \n# A_ls = np.concatenate([(1/np.sqrt(2*eta0))*A, B_concat], axis = 0) \n# #print(np.array(B).shape) \n# #print(w[0].shape) \n# #print(w, eta) \n# #w.append([]) THIS IS WRONG, CHANGES THE LIST \n# w_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(w+[[]])[:-1], axis = 0) #[:-1] Added as a hack to keep it one-dim array of objects \n# eta_w = np.expand_dims(1/np.sqrt(2*eta),1)*np.array(w) \n# print(eta_w.shape) \n# b_ls = np.concatenate([(1/np.sqrt(2*eta_0))*w_0, eta_w.flatten()], axis = 0) \n #Use correct broadcasting?\n w_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(w+[[]])[:-1], axis = 0) #[:-1] Added as a hack to keep it one-dim array of objects \n b_ls = np.concatenate([(1/np.sqrt(2*eta_0))*w_0, w_concat, (1/np.sqrt(2*eta_lin))*w_lin], axis = 0) \n# print(np.sum(eta_w.flatten() != w_concat)) \n# premultiplied_time_start = time.time() \n# premultiplied_lhs = eta_T_H_stacked.T.dot(eta_T_H_stacked).toarray() \n# premultiplied_time_end = time.time() \n# print('premultiplying took {}'.format(premultiplied_time_end - premultiplied_time_start)) \n# premultiplied_rhs = eta_T_H_stacked.T.dot(b_ls) \n# u_next = nnls_predotted(premultiplied_lhs, premultiplied_rhs, tol=1e-5) \n# print(eta_T_H_stacked.shape, b_ls.shape) \n# A_ls_t_b = eta_T_H_stacked.T.dot(b_ls) \n# w =scipy.sparse.linalg.spsolve_triangular(RT, A_ls_t_b, lower = True) \n# x = scipy.sparse.linalg.spsolve_triangular(R, w, lower = False) \n# u_next = x \n u_next = scipy.optimize.lsq_linear(eta_T_H_L_stacked, b_ls, bounds = (0, np.inf), tol=1e-3, lsmr_tol=1e-3, max_iter=nnls_max_iter, verbose=1).x \n# u = scipy.optimize.lsq_linear(premultiplied_lhs, premultiplied_rhs, bounds = (0, np.inf), tol=1e-5).x \n return u_next","def python_nonsquare_matrix_mult(matrix):\n\n transposed_matrix = np.zeros([matrix.shape[1],matrix.shape[0]])\n start = time.time()\n # for i in range(matrix.shape[0]):\n # for j in range(matrix.shape[1]):\n # transposed_matrix[j,i] = matrix[i,j]\n\n transposed_matrix = np.transpose(matrix)\n product = matrix.dot(transposed_matrix)\n\n # transposed_matrix = np.transpose(matrix)\n end = time.time()-start\n\n # print(\"Python Golden Transpose: %s\" % product)\n # print('python transpose time: %.2E' % end)\n return [product, end]","def multiM(*args):\r\n filas_1,filas_2 = len(args[0]),len(args[1])\r\n columnas_1,columnas_2 = len(args[0][0]),len(args[1][0])\r\n matriz_r = []\r\n for k in range(filas_1):\r\n matriz_r.append([0]*columnas_2)\r\n for i in range(columnas_2):\r\n matriz_r[k][i] = 0\r\n for i in range(filas_1):\r\n for j in range(columnas_1):\r\n for k in range(columnas_2):\r\n matriz_r[i][k] = matriz_r[i][k] + args[0][i][j] * args[1][j][k]\r\n return matriz_r","def matrix_mult(m1, m2):\n output = []\n for rowIndex, row in enumerate(m1): #go through rows in m1\n new_row = []\n for columnIndex in range(len(m2[0])): #go through indices for each column of m2\n sum = 0\n for index3 in range(len(row)):\n product = m1[rowIndex][index3] * m2[index3][columnIndex]\n sum += product\n new_row.append(sum)\n output.append(new_row)\n return output\n \n \n #output = []\n #first for loop corresponds to the rows of my output matrix and loops through the rows of m1 (enumerate)\n #create an empty new row\n # second for loop, loops through columns of m2\n # create sum variable, initialize it with zero\n # third for loop, multiplies the index of the row in m1 times the index of the column in m2\n # add sum to product and assign this to the sum variable\n # append sum to new row\n # append new row to output\n # return output","def matrix_add():","def it_matrixpower(p,t,n,root_field=RR):\n assert n>=2, \"Carleman matrix must at least be of size 2 to retrieve the coefficients. But given was \" + repr(n)\n CM = p.carleman_matrix(n)\n ev = CM.charpoly().roots(root_field)\n assert len(ev) == n, \"Carleman matrix must have exactly \" + repr(n) + \"eigenvalues, but has \" + repr(len(ev))\n\n Char = [0]*n\n for k in range(n):\n #here is possibility for improvement of precision\n #to separate the fractional from the root parts\n #expanding the product\n Char[k] = CM - ev[k][0]*identity_matrix(n)\n\n #we want to have the first row of the product of the matrices\n #thatswhy we mulitply in front with:\n prod = vector(p.K,[0,1]+[0]*(n-2))\n prodwo = [0]*n\n for k in range(n):\n prodwo[k]=prod #these are the first terms until k-1\n\n #no need to continue\n if k == n-1:\n break\n\n #and we add the terms starting with k+1\n for i in range(k+1,n):\n prodwo[k] = prodwo[k] * Char[i]\n\n prod = prod * Char[k]\n\n sprodwo = [0]*n\n for k in range(n):\n if k==0:\n sprodwo[k] = ev[k][0] - ev[1][0]\n start = 2\n else:\n sprodwo[k] = ev[k][0] - ev[0][0]\n start = 1\n\n for i in range(start,n):\n if i != k:\n sprodwo[k] = sprodwo[k] * (ev[k][0] - ev[i][0])\n\n res = ev[0][0]**t/sprodwo[0] * prodwo[0]\n for k in range(1,n):\n res += ev[k][0]**t/sprodwo[k]*prodwo[k]\n\n return res.list()","def _mul(*args):\n\treturn functools.reduce(numpy.dot, args)","def matrixMultiplication(self, n, id, context):\n\n print(\"id: {}\".format(id))\n # Create one matrix\n f = 1\n m1 = []\n for x in range(n):\n row = []\n for y in range(n):\n row.append(f)\n f = f + 1\n m1.append(row)\n # The second matrix is equal to the first matrix\n m2 = m1\n print(\"m2: {}\".format(m2))\n\n # Multiply matrices\n m3 = []\n for i in range(n):\n row = []\n for j in range(n):\n sum = 0\n for k in range(n):\n sum = sum + m1[i][k] * m2[k][j]\n row.append(sum)\n m3.append(row)\n\n sum = 0\n # add the entries\n for i in range(n):\n for j in range(n):\n sum = sum + m3[i][j]\n\n print(\"Result of multiplication is {}\".format(sum))\n return sum","def mat_mul(mat1, mat2):\n\n rows1 = len(mat1)\n cols1 = len(mat1[0])\n rows2 = len(mat2)\n cols2 = len(mat2[0])\n\n if cols1 != rows2:\n return None\n else:\n new_matrix = []\n for x in range(rows1):\n aux_row = []\n for y in range(cols2):\n aux_sum = []\n for z in range(cols1):\n aux_sum.append(mat1[x][z] * mat2[z][y])\n aux_row.append(sum(aux_sum))\n new_matrix.append(aux_row)\n\n return new_matrix","def mul(Z,X,Y):","def mul(self,mat1,mat2):\n if(isinstance(mat2,int)==True):\n result = [[mat1[i][j] * mat2 for j in range(len(mat1[0]))] for i in range(len(mat1))]\n self.out = result\n return self.out\n elif(len(mat1[0])==len(mat2)):\n result = [[sum(a*b for a,b in zip(i,j)) for j in zip(*mat2)] for i in mat1]\n self.out = result\n return self.out","def __matmul__(self, other):\n return F.MatMul.apply(self, other)","def __mul__(self,m):\n if type(m) != Matrix:\n raise TypeError('The second argument is not a matrix lol')\n if self.ncols != m.nrows:\n raise ValueError('matrix dot argument has incorrect number of rows')\n new = Matrix(self.nrows,m.ncols)\n columns = m.getCols()\n rowindex = 0\n colindex = 0 \n for row in self.matrix:\n colindex = 0 \n for col in columns:\n summ = 0\n for i,j in zip(row,col):\n summ+= i*j \n new.matrix[rowindex][colindex] = summ\n print new.matrix\n colindex += 1 \n rowindex+=1\n return new","def __mul__(self, other):\n # \n # TODO - your code here\n #\n \n result = [];\n row_result = [];\n product = 0;\n \n if(self.w != other.h):\n raise(ValueError, \"Matrices can not multiply for their dimesion doesn't match\"); \n \n for row in self.g:\n row_result = [];\n for j in range(other.w):\n product = dot_product(row,other.get_column(j));\n row_result.append(product);\n result.append(row_result);\n \n return Matrix(result);","def __mul__(self, othertr):\n res = self.dot(othertr)\n return res","def nonsquare_matrix_mult_opt2(matrix):\n\n #Setup openCL\n dev, ctx, queue = setup_CL()\n\n #openCL Kernel\n #Naive approach with local/private memory\n #Naive approach reworked to use local memory and tiling\n #Modified boundary condition tiling kernel in lecture\n kernel_code = \"\"\"\n #define MATRIX_ROW_SIZE {}\n #define MATRIX_COL_SIZE {}\n #define TILE_WIDTH {}\n #define n {}\n __kernel void func(__global float* a, __global float* b) {{\n\n __local float M[TILE_WIDTH][TILE_WIDTH];\n __local float N[TILE_WIDTH][TILE_WIDTH];\n\n int bx = get_group_id(0); int by = get_group_id(1);\n int tx = get_local_id(0); int ty = get_local_id(1);\n int Row = by * get_local_size(1) + ty;\n int Col = bx * get_local_size(0) + tx;\n float Cvalue = 0;\n\n // Loop over the A and B tiles required to compute the C element\n for (int t = 0; t < (n-1)/TILE_WIDTH + 1;++t) {{\n\n //Assign rows of input\n if(t*TILE_WIDTH+tx < MATRIX_COL_SIZE && tx < MATRIX_COL_SIZE && (Row*MATRIX_COL_SIZE + t*TILE_WIDTH + tx)<MATRIX_COL_SIZE*MATRIX_ROW_SIZE) {{\n M[ty][tx] = a[Row*MATRIX_COL_SIZE + t*TILE_WIDTH + tx];\n }} else {{\n M[ty][tx] = 0.0;\n }}\n\n //Assign columns of transpose\n if (t*TILE_WIDTH+ty < n && Col < MATRIX_ROW_SIZE) {{\n N[ty][tx] = a[t*TILE_WIDTH + MATRIX_COL_SIZE*Col + ty];\n }} else {{\n N[ty][tx] = 0.0;\n }}\n\n barrier(CLK_LOCAL_MEM_FENCE);\n\n //Sum tile\n for (int i = 0; i < TILE_WIDTH; ++i) {{\n Cvalue += M[ty][i] * N[i][tx];\n }}\n\n barrier(CLK_LOCAL_MEM_FENCE);\n\n //Assign values to output\n if(Row<MATRIX_ROW_SIZE && Col<MATRIX_ROW_SIZE) {{\n b[Row*MATRIX_ROW_SIZE + Col] = Cvalue;\n\n }}\n }}\n }}\n \"\"\"\n\n #Move data to device\n matrix_float = matrix.astype(np.float32)\n matrix_gpu = cl.array.to_device(queue, matrix_float)\n transposeMult_gpu = cl.array.empty(queue, (matrix.shape[0], matrix.shape[0]), np.float32)\n transposed_gpu = cl.array.empty(queue, (matrix.shape[1],matrix.shape[0]), np.float32)\n\n matrix_row_size = matrix.shape[0]\n matrix_col_size = matrix.shape[1]\n TILE_WIDTH = 2\n\n #Calculate workItems, workGroup size, workGroups for input\n matrix_val_count = matrix_float.shape[0]*matrix_float.shape[1]\n xWorkItems = int(int(matrix_row_size-1)/TILE_WIDTH)+1\n yWorkItems = int(int(matrix_row_size-1)/TILE_WIDTH)+1\n totalWorkItems = float(TILE_WIDTH*TILE_WIDTH)\n groups = np.int(max(np.ceil(matrix_val_count / xWorkItems),1))\n\n # print(\"workItems: %s, matrix_val_count: %s, groups: %s\" % (totalWorkItems, matrix_val_count, groups))\n\n # update template with current runtime requirements\n kernel = kernel_code.format(matrix_row_size, matrix_col_size, TILE_WIDTH, max(matrix_col_size, matrix_row_size))\n\n #Launch kernel and time it\n #Set global ID, workItems, workGroups\n prg = cl.Program(ctx, kernel).build()\n start = time.time()\n event = prg.func(queue, (xWorkItems*TILE_WIDTH,yWorkItems*TILE_WIDTH,1),(TILE_WIDTH,TILE_WIDTH,1), matrix_gpu.data, transposeMult_gpu.data)\n runtime = time.time()-start\n\n #Save output\n transposedMult = transposeMult_gpu.get()\n transposed = transposed_gpu.get()\n\n # print('openCL_opt2 %d x %d transpose-mult time: %.2E' % (matrix.shape[0], matrix.shape[1], runtime))\n # print('openCL_opt2_transposed==goldenTransposed: %s' % np.allclose(transposed, np.transpose(matrix)))\n # print('openCL_opt2_mult==goldenMult: %s' % np.allclose(transposedMult, matrix.dot(np.transpose(matrix))))\n if not(np.allclose(transposedMult, matrix.dot(np.transpose(matrix)))):\n # print('Original Matrix:\\n %s' % matrix)\n print('openCL_opt2 transposed val:\\n %s' % transposed)\n print('golden transpose-mult:\\n %s' % matrix.dot(np.transpose(matrix)))\n transposedMult[(transposedMult>0) & (transposedMult<1)] = -1\n print('openCL_opt2 mult val:\\n %s' % transposedMult)\n print('openCL_opt2 transpose-mult:\\n %s' % np.isclose(transposedMult,matrix.dot(np.transpose(matrix))))\n # print('--------------------')\n\n return [transposedMult, runtime]","def __mul__(self,that):\n return self.__opExpand2(that, np.multiply)","def matrix_mult(A,B):\n\n m = len(A)\n p = len(B)\n n = len(B[0])\n AB = []\n for i in range(m):\n AB.append([])\n for j in range(n):\n total = 0\n for k in range(p):\n total += A[i][k] * B[k][j]\n AB[i].append(total)\n return AB","def _multi_matmul_three(A, B, C, constant=False) -> Tensor:\n a0, a1b0 = A.shape[-2:]\n b1c0, c1 = C.shape[-2:]\n cost1 = a0 * b1c0 * (a1b0 + c1)\n cost2 = a1b0 * c1 * (a0 + b1c0)\n\n if cost1 < cost2:\n return matmul(matmul(A, B, constant), C, constant)\n else:\n return matmul(A, matmul(B, C, constant), constant)","def _generate_mult_process(X, mat, inits):\n M = np.empty_like(X, dtype=float)\n M[..., 0] = inits[X[..., 0]]\n M[..., 1:] = mat[X[..., :-1], X[..., 1:]]\n np.cumprod(M, axis=-1, out=M)\n return M","def matrixMul(a, b):\n # Initializing Empty Matrix\n c = [[0, 0], [0, 0]]\n # 2x2 matrix multiplication. Essentially O(1)\n for i in range(2):\n for j in range(2):\n for k in range(2):\n c[i][j] = (c[i][j] + (a[i][k] * b[k][j]))\n\n # Returning the products\n return c","def __matmul__(self, other):\n if isinstance(other, types.Vector):\n return self.apply(target=other)\n if isinstance(other, MatrixLieGroup):\n return self.multiply(other=other)\n else:\n assert False, \"Invalid argument\"","def contract(tensor):\n temp = np.einsum('ikma, jlan', tensor, tensor)\n M = np.zeros((tensor.shape[0]**2, tensor.shape[1]**2, tensor.shape[2], tensor.shape[3]))\n for i,j,k,l,m,n in it.product(*[range(x) for x in temp.shape]):\n M[i + tensor.shape[0]*j, k + tensor.shape[1]*l, m, n] = temp[i,j,k,l,m,n]\n return M","def matrixMultiplication(firstMatrix, secondMatrix):\n if len(firstMatrix[0]) == len(secondMatrix): # Checks whether the matrices can be multiplied or not or not\n finalMatrix = []\n for y in range(len(firstMatrix)): # 2\n currentMatrix = []\n for i in range(len(secondMatrix[0])):\n currentSum = 0\n for j in range(len(secondMatrix)):\n currentSum += secondMatrix[j][i] * firstMatrix[y][j]\n currentMatrix.append(currentSum)\n print(\"This is my current matrix: \" + str(currentMatrix))\n finalMatrix.append(currentMatrix)\n print(\"This product of the two matrices is :) \" + str(finalMatrix))\n else:\n print(\"This operation cannot be done, make sure the rows of the first matrix is the same as the number of columns in the second matrix\")","def __matmul__(self, B):\n m, n = self.shape\n n_, r = B.shape\n assert n == n_, (\"Cannot multiply shapes \"\n \"({}, {}) and ({}, {})\".format(m, n, n_, r))\n mul_ = dict()\n # compute A_ik = sum_j A_ij*B_jk\n for i in range(m):\n for k in range(r):\n prod = mpfr(0)\n for j in range(n):\n prod += self[i, j] * B[j, k]\n mul_[i, k] = prod\n return MPMatrix((m, r), mul_)","def matmul(x, y, _pub):\n if x.shape[-1] != y.shape[-2]:\n pass # TODO: REPORT ERROR\n res = paillier_gpu.matmul_impl(x.flatten(), y.flatten(order='F'), x.shape, y.shape)\n\n return res","def test_matmul_mm(self):\n self.check_dot_mm(matmul_usecase, None, \"'@'\")","def matrix_mult(A,B,mod=10**9+7):\n C = [[1,1],[1,1]]\n C[0][0] = ((A[0][0]*B[0][0])%mod + (A[0][1]*B[1][0])%mod)%mod\n C[0][1] = ((A[0][0]*B[0][1])%mod + (A[0][1]*B[1][1])%mod)%mod\n C[1][0] = ((A[1][0]*B[0][0])%mod + (A[1][1]*B[1][0])%mod)%mod\n C[1][1] = ((A[1][0]*B[0][1])%mod + (A[1][1]*B[1][1])%mod)%mod\n return C","def product_on_basis(self, t1, t2):\n return tensor( (module.monomial(x1)*module.monomial(x2) for (module, x1, x2) in zip(self._sets, t1, t2)) ) #.","def __mul__(self, other):\n return Matrix3(\n self.i * other,\n self.j * other,\n self.k * other,\n )","def matmul(A, B):\n # type: (Optional[Tensor], Tensor) -> Tensor\n if A is None:\n return B\n if is_sparse(A):\n return torch.sparse.mm(A, B)\n return torch.matmul(A, B)","def __imatmul__(self, tensor):\n # Note: Matching PyTorch convention, which is not in-place here.\n return self.matmul(tensor)","def unfolded_product(m1, m2, coords = None, mx = None):\n if coords is None:\n mx = np.max( np.array([np.max(np.abs(m1.coords), axis = 0), \n np.max(np.abs(m2.coords), axis = 0)]), axis = 0)\n\n coords = lattice_coords(mx)\n else:\n if mx is not None:\n coords = lattice_coords(mx)\n #print(coords, coords[int(len(coords)/2)])\n M1 = m1.tofull(m1,coords,coords)\n M2 = m2.tofull(m2,coords,coords)\n \n RET = np.dot(M1, M2)\n rnx = m1.blocks.shape[1]\n rny = m2.blocks.shape[2]\n rc = int(len(coords)/2)\n retblocks = RET[rnx*rc:rnx*(rc+1),:].reshape(rnx, coords.shape[0], rny).swapaxes(0,1)\n \n ret = tmat()\n ret.load_nparray(retblocks, coords)\n return ret","def bc_matrix(params):\r\n w = params['w']\r\n kx = params['kx']\r\n d_list = params['d_list']\r\n ex_list = params['ex_list']\r\n ez_list = params['ez_list']\r\n kz_list = params['kz_list']\r\n N = len(d_list)\r\n assert N == len(d_list) == len(ex_list) == len(ez_list) == len(kz_list)\r\n assert N >= 2\r\n assert d_list[0] == d_list[-1] == inf\r\n \r\n # delta = e^{i * kz * d}, i.e. phase change across each layer\r\n # delta[0] and delta[-1] are undefined and are not used.\r\n delta_list = [cmath.exp(1j * kz_list[i] * d_list[i]) for i in range(N)]\r\n \r\n Ex_up_over_H_up_list = [kz_list[i] / (w * ex_list[i] * nu.eps0)\r\n for i in range(N)]\r\n Ex_down_over_H_down_list = [-a for a in Ex_up_over_H_up_list]\r\n Ez_up_over_H_up_list = [-kx / (w * ez_list[i] * nu.eps0) for i in range(N)]\r\n Ez_down_over_H_down_list = Ez_up_over_H_up_list[:]\r\n \r\n mat = np.zeros((2*N-2, 2*N-2), dtype=complex)\r\n \r\n for row_now in range(N-1):\r\n # This row concerns continuity of Ex across the boundary between\r\n # layer_under and layer_over (under and over the boundary respectively)\r\n layer_under = row_now\r\n layer_over = layer_under + 1\r\n # up_under_index is the column index in mat that gets multiplied by\r\n # H_{up} in layer_under.\r\n up_under_index = 2 * layer_under - 1\r\n down_under_index = 2 * layer_under\r\n up_over_index = 2 * layer_over - 1\r\n down_over_index = 2 * layer_over\r\n \r\n if layer_under != 0:\r\n assert 0 <= up_under_index < 2*N-2\r\n mat[row_now, up_under_index] = (\r\n Ex_up_over_H_up_list[layer_under] * delta_list[layer_under])\r\n mat[row_now, down_under_index] = Ex_down_over_H_down_list[layer_under]\r\n mat[row_now, up_over_index] = -Ex_up_over_H_up_list[layer_over]\r\n if layer_over != N-1:\r\n assert 0 <= down_over_index < 2*N-2\r\n mat[row_now, down_over_index] = (\r\n -Ex_down_over_H_down_list[layer_over] * delta_list[layer_over])\r\n\r\n for row_now in range(N-1, 2*N-2):\r\n # This row concerns continuity of eps_z * Ez across the boundary between\r\n # layer_under and layer_over (under and over the boundary respectively)\r\n layer_under = row_now - (N-1)\r\n layer_over = layer_under + 1\r\n # up_under_index is the column index in mat that gets multiplied by\r\n # H_{up} in layer_under.\r\n up_under_index = 2 * layer_under - 1\r\n down_under_index = 2 * layer_under\r\n up_over_index = 2 * layer_over - 1\r\n down_over_index = 2 * layer_over\r\n \r\n if layer_under != 0:\r\n assert 0 <= up_under_index < 2*N-2\r\n mat[row_now, up_under_index] = (ez_list[layer_under] *\r\n Ez_up_over_H_up_list[layer_under] * delta_list[layer_under])\r\n mat[row_now, down_under_index] = (ez_list[layer_under] *\r\n Ez_down_over_H_down_list[layer_under])\r\n mat[row_now, up_over_index] = (-ez_list[layer_over] * \r\n Ez_up_over_H_up_list[layer_over])\r\n if layer_over != N-1:\r\n assert 0 <= down_over_index < 2*N-2\r\n mat[row_now, down_over_index] = (-ez_list[layer_over] *\r\n Ez_down_over_H_down_list[layer_over] * delta_list[layer_over])\r\n \r\n return mat","def build_M(K, R, T):\n external = np.append(R, T, axis=1)\n M = K.dot(external)\n return M","def combine_powers(mul):\n factor, args = mul.as_coeff_matrices()\n new_args = [args[0]]\n\n for i in range(1, len(args)):\n A = new_args[-1]\n B = args[i]\n\n if isinstance(B, Inverse) and isinstance(B.arg, MatMul):\n Bargs = B.arg.args\n l = len(Bargs)\n if list(Bargs) == new_args[-l:]:\n new_args = new_args[:-l] + [Identity(B.shape[0])]\n continue\n\n if isinstance(A, Inverse) and isinstance(A.arg, MatMul):\n Aargs = A.arg.args\n l = len(Aargs)\n if list(Aargs) == args[i:i+l]:\n identity = Identity(A.shape[0])\n new_args[-1] = identity\n for j in range(i, i+l):\n args[j] = identity\n continue\n\n if A.is_square == False or B.is_square == False:\n new_args.append(B)\n continue\n\n if isinstance(A, MatPow):\n A_base, A_exp = A.args\n else:\n A_base, A_exp = A, S.One\n\n if isinstance(B, MatPow):\n B_base, B_exp = B.args\n else:\n B_base, B_exp = B, S.One\n\n if A_base == B_base:\n new_exp = A_exp + B_exp\n new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)\n continue\n elif not isinstance(B_base, MatrixBase):\n try:\n B_base_inv = B_base.inverse()\n except NonInvertibleMatrixError:\n B_base_inv = None\n if B_base_inv is not None and A_base == B_base_inv:\n new_exp = A_exp - B_exp\n new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)\n continue\n new_args.append(B)\n\n return newmul(factor, *new_args)","def matrix_mult(m1, m2):\n\ttemp = []\n\tfor i in range(len(m1)):\n\t\te = []\n\t\tfor j in range(len(m2[0])):\n\t\t\te.append(row_times_column(m1,i,m2,j))\n\t\ttemp.append(e)\n\treturn temp","def matmul(a, b):\n raise NotImplementedError","def mul(self):\n a = self.pop()\n b = self.pop()\n c= a*b\n self.push(c)","def applyMat(my_map, linsys_setup):\n \n datamaps, ninvs, beams, freqs, power_2d, precond_2d, clumaps, g_nu, \\\n map_prop = linsys_setup\n \n \n\n nx, ny, pixScaleX, pixScaleY = map_prop\n nFreq = len(g_nu); nCluster = len(clumaps[0])\n\n #Always apply beam * precond\n beam_prec=[]\n\n for f in range(nFreq):\n beam_prec+=[beams[f][:,:ny/2+1]*precond_2d[:,:ny/2+1]]\n precond_2d=precond_2d[:,:ny/2+1]\n power_2d=power_2d[:,:ny/2+1]\n \n ksz = False\n if len(clumaps) == 2: ksz = True\n \n # Routines to perform block matrix multiplication defined in Eriksen Eq. 19\n \n def apply_cmb_cmb(d0):\n \"\"\"\n Apply (S^-1 + A N^-1 A) x\n \"\"\"\n d1 = d0.copy()\n d1 = numpy.reshape(d1,(nx,ny))\n a_l = fft.rfft(d1,axes=[-2,-1])\n \n c_l = 0\n for f in range(nFreq):\n\n b_l = a_l * beam_prec[f]\n d2 = fft.irfft(b_l,axes=[-2,-1],normalize=True)\n d2 *= ninvs[f]\n b_l = fft.rfft(d2,axes=[-2,-1])\n c_l += b_l * beam_prec[f]\n \n d2 = fft.irfft(c_l,axes=[-2,-1],normalize=True)\n d1 = fft.irfft(precond_2d**2 * a_l/power_2d,axes=[-2,-1],normalize=True)\n \n d2 += d1\n \n return d2.reshape((nx*ny,))\n \n \"\"\"\n def apply_tsz_tsz(d0): # DONE\n \\\"\"\"\n Apply (F^T A^T N^-1 A F) x\n \\\"\"\"\n d1 = numpy.zeros(nCluster)\n mat = numpy.zeros((nCluster, nCluster))\n # TODO: This could probably be more efficient (e.g. using np.outer)\n for freq in range(nFreq):\n for ic in range(nCluster):\n for jc in range(0, ic+1):\n mat[ic,jc] = numpy.sum( ninvs[freq] * g_nu[freq]**2. \\\n * clumaps[0][ic][freq] * clumaps[0][jc][freq] )\n if ic != jc: mat[jc,ic] = mat[ic,jc]\n d1 += numpy.dot(mat, d0)\n return d1\n \n def apply_ksz_ksz(d0): # DONE\n \\\"\"\"\n Apply (K^T A^T N^-1 A K) x\n \\\"\"\"\n # FIXME: Missing factor of ivcov\n d1 = numpy.zeros(nCluster)\n mat = numpy.zeros((nCluster, nCluster))\n # TODO: This could probably be more efficient (e.g. using np.outer)\n for freq in range(nFreq):\n for ic in range(nCluster):\n for jc in range(0, ic+1):\n mat[ic,jc] = numpy.sum( ninvs[freq] \\\n * clumaps[1][ic][freq] * clumaps[1][jc][freq] )\n if ic != jc: mat[jc,ic] = mat[ic,jc]\n d1 += numpy.dot(mat, d0)\n d1 += numpy.dot(ivcov, d0) # Add prior term\n return d1\n \"\"\"\n \n def apply_cmb_foreground_block(dc, dm, dt, dk=None):\n \"\"\"\n Apply the CMB x (Monopole + TSZ + KSZ) terms in one block:\n (A^T N^-1 A T) x_mono\n (A^T N^-1 A F) x_tsz\n (A^T N^-1 A K) x_ksz\n \n (T^T A^T N^-1 A) x_cmb\n (F^T A^T N^-1 A) x_cmb\n (K^T A^T N^-1 A) x_cmb\n \"\"\"\n ksz = False\n if dk is not None: ksz = True\n \n # (A^T N^-1 A T) x_mono; (A^T N^-1 A F) x_tsz; (A^T N^-1 A K) x_ksz\n b_lt = 0; b_lk = 0; b_lm = 0\n for f in range(nFreq):\n mct = 0; mck = 0\n for ic in range(nCluster):\n mct += dt[ic] * ninvs[f] * clumaps[0][ic][f] * g_nu[f]\n if ksz: mck += dk[ic] * ninvs[f] * clumaps[1][ic][f]\n \n b_lm += fft.rfft(dm * ninvs[f],axes=[-2,-1]) * beam_prec[f]\n b_lt += fft.rfft(mct,axes=[-2,-1]) * beam_prec[f]\n if ksz: b_lk += fft.rfft(mck,axes=[-2,-1]) * beam_prec[f]\n\n mcm = fft.irfft(b_lm,axes=[-2,-1],normalize=True).reshape((nx*ny,))\n mct = fft.irfft(b_lt,axes=[-2,-1],normalize=True).reshape((nx*ny,))\n if ksz: mck = fft.irfft(b_lk,axes=[-2,-1],normalize=True).reshape((nx*ny,))\n \n # (T^T A^T N^-1 A) x_cmb; (F^T A^T N^-1 A) x_cmb; (K^T A^T N^-1 A) x_cmb\n mc = dc.copy().reshape((nx,ny))\n a_l = fft.rfft(mc,axes=[-2,-1])\n mtc = numpy.zeros(nCluster)\n mkc = numpy.zeros(nCluster)\n mmc = 0\n for f in range(nFreq):\n b_l = a_l * beam_prec[f]\n mc = fft.irfft(b_l,axes=[-2,-1],normalize=True)\n mmc += numpy.sum(mc * ninvs[f])\n for ic in range(nCluster):\n mtc[ic] += numpy.sum(mc * ninvs[f] * clumaps[0][ic][f] * g_nu[f])\n if ksz: mkc[ic] += numpy.sum(mc * ninvs[f] * clumaps[1][ic][f])\n \n if ksz: return mct, mcm, mck, mtc, mmc, mkc\n return mct, mcm, mtc, mmc\n \n \n def apply_foreground_block(m0, t0, k0=None):\n \"\"\"\n Apply the TSZ + KSZ + Monopole terms in one block:\n [ (T^T A^T N^-1 A F) (T^T A^T N^-1 A K) (T^T A^T N^-1 A T) ] (x_mono)\n [ (F^T A^T N^-1 A F) (F^T A^T N^-1 A K) (F^T A^T N^-1 A T) ] (x_tsz)\n [ (K^T A^T N^-1 A F) (K^T A^T N^-1 A K) (K^T A^T N^-1 A T) ] (x_ksz)\n \"\"\"\n ksz = True if k0 is not None else False\n \n dtt, dkk, dtk, dkt = [numpy.zeros(nCluster) for i in range(4)]\n mtt, mkk, mtk, mkt = [numpy.zeros((nCluster, nCluster)) for i in range(4)]\n dmm, dmk, dmt = [0 for i in range(3)]\n dkm, dtm = [numpy.zeros(nCluster) for i in range(2)]\n \n # TODO: This could probably be more efficient (e.g. using np.outer)\n for f in range(nFreq):\n dmm += numpy.sum(ninvs[f]) * m0\n \n # Loop through clusters\n for ic in range(nCluster):\n dmt += numpy.sum( ninvs[f] * g_nu[f] * clumaps[0][ic][f] * t0[ic] )\n dtm[ic] += numpy.sum( ninvs[f] * g_nu[f] * clumaps[0][ic][f] * m0 )\n if ksz:\n dmk += numpy.sum( ninvs[f] * clumaps[1][ic][f] * k0[ic] )\n dkm[ic] += numpy.sum( ninvs[f] * clumaps[1][ic][f] * m0 )\n \n for jc in range(0, ic+1):\n mtt[ic,jc] = numpy.sum( ninvs[f] * g_nu[f]**2. \\\n * clumaps[0][ic][f] * clumaps[0][jc][f] )\n if ksz:\n mkk[ic,jc] = numpy.sum( ninvs[f] \\\n * clumaps[1][ic][f] * clumaps[1][jc][f] )\n mtk[ic,jc] = numpy.sum( ninvs[f] * g_nu[f] \\\n * clumaps[0][ic][f] * clumaps[1][jc][f] )\n mkt[ic,jc] = numpy.sum( ninvs[f] * g_nu[f] \\\n * clumaps[1][ic][f] * clumaps[0][jc][f] )\n # Mirror indices\n mtt[jc,ic] = mtt[ic,jc]\n if ksz:\n mkk[jc,ic] = mkk[ic,jc]\n mtk[jc,ic] = mtk[ic,jc]\n mkt[jc,ic] = mkt[ic,jc]\n \n # Add total contribs. for this band\n dtt += numpy.dot(mtt, t0)\n if ksz:\n dkk += numpy.dot(mkk, k0)\n dtk += numpy.dot(mtk, k0)\n dkt += numpy.dot(mkt, t0)\n \n if ksz: return dtt, dkk, dmm, dtk, dkt, dmk, dkm, dtm, dmt\n return dtt, dmm, dtm, dmt\n \n # Apply block matrix multiplications and return\n # FIXME: What if KSZ not used?\n x0 = my_map[:nx*ny]\n x1 = my_map[nx*ny:nx*ny+1]\n x2 = my_map[nx*ny+1:nx*ny+nCluster+1]\n if ksz: x3 = my_map[nx*ny+nCluster+1:nx*ny+2*nCluster+1]\n \n # Multiply input vector in blocks\n #t=time.time()\n dcc = apply_cmb_cmb(x0)\n #print 'CMB', time.time()-t\n if ksz:\n dct, dcm, dck, dtc, dmc, dkc = apply_cmb_foreground_block(x0, x1, x2, x3)\n dtt, dkk, dmm, dtk, dkt, dmk, dkm, dtm, dmt = apply_foreground_block(x1, x2, x3)\n x_new_0 = dcc + dct + dck + dcm\n x_new_1= dmc + dmt + dmk + dmm\n x_new_2 = dtc + dtt + dtk + dtm\n x_new_3 = dkc + dkt + dkk + dkm\n x_new = numpy.concatenate((x_new_0, x_new_1, x_new_2, x_new_3))\n else:\n #t=time.time()\n dct, dcm, dtc, dmc = apply_cmb_foreground_block(x0, x1, x2)\n #print 'CMB-F', time.time()-t\n \n #t=time.time()\n dtt, dmm, dtm, dmt = apply_foreground_block(x1, x2)\n #print 'F', time.time()-t\n \n x_new_0 = dcc + dct + dcm\n x_new_1 = dmc + dmt + dmm\n x_new_2 = dtc + dtt + dtm\n x_new = numpy.concatenate((x_new_0, x_new_1, x_new_2))\n\n\n#sys.exit()\n return x_new","def __matmul__(self, csys):\n self._transform(csys)\n return self","def matrix_mult_matrix(matrix_a, matrix_b):\n m = len(matrix_a)\n n = len(matrix_b)\n result = []\n matrix_b_t = transpose_matrix(matrix_b)\n for i in xrange(m):\n row = []\n\tfor j in xrange(m):\n row.append(dot_product(matrix_a[i], matrix_b_t[j]))\n\tresult.append(row)\n return result","def distribute_monom(mul):\n args = mul.args\n if len(args) == 2:\n from .matadd import MatAdd\n if args[0].is_MatAdd and args[1].is_Rational:\n return MatAdd(*[MatMul(mat, args[1]).doit() for mat in args[0].args])\n if args[1].is_MatAdd and args[0].is_Rational:\n return MatAdd(*[MatMul(args[0], mat).doit() for mat in args[1].args])\n return mul","def power_matrix(A, k):\n nrow = np.shape(A)[0]\n A0 = np.identity(nrow) \n for k in range(q):\n A0 = np.dot(A0, A)\n \n return A0","def test_matmul_adjoint(self):\n x = tf.constant([[9, 4, 1, 3], [1, 3, 1, 6], [7, 2, 2, 8]],\n dtype=tf.float32)\n self.assertAllClose(\n tf.linalg.matmul(self.linop, x, adjoint_a=True),\n tf.linalg.matmul(self.matrix, x, adjoint_a=True))\n self.assertAllClose(\n tf.linalg.matmul(self.linop, x, adjoint_a=False, adjoint_b=True),\n tf.linalg.matmul(self.matrix, x, adjoint_a=False, adjoint_b=True))","def __matmul__(self, B):\n if isinstance(B, dict):\n X = self.metacopy()\n\n if B['leg'] not in X.indexes:\n raise ValueError('Leg of singular values not an indexes '\n 'of self')\n\n if B['symmetries'] != X.symmetries:\n raise ValueError('Not same symmetries')\n\n x, y = X.coupling_id(B['leg'])\n for k in self:\n newshape = [1] * len(self[k].shape)\n newshape[X.indexes.index(B['leg'])] = -1\n X[k] = self[k] * B[k[x][y]].reshape(newshape)\n\n return X\n\n connections = self.connections(B)\n if not connections:\n raise ValueError(f'No connections found between {self} and {B}')\n\n return self.contract(B, (list(connections),) * 2).simplify()","def kkMul(*args):\n if (None in args):\n return None\n product = 1\n for arg in args:\n product *= arg\n return product","def matMul(a, b):\n sa=matShape(a)\n sb=matShape(b)\n if sa[1]!=sb[0]: raise ValueError\n ret=matZeros((sa[0],sb[1]))\n for i in range(sa[0]):\n for j in range(sb[1]):\n val=0.0\n for k in range(sa[1]):\n val+=matGet(a,i,k)*matGet(b,k,j)\n matSet(ret,i,j,val)\n return ret","def matmul(self, x, work_buffer):\n\n x = asarray(x)\n space = work_buffer.flat\n\n if (x.ndim == 0):\n ValueError(\n \"matmul: Input operand 1 does not have enough dimensions \"\n \"(has 0, gufunc core with signature (n?,k),(k,m?)->(n?,m?) \"\n \"requires 1\"\n )\n\n if x.ndim == 1 and self.ndim == 1:\n # Dot product\n if x.shape[0] == self._size:\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n out = matmul(self[self._begin:], x[:k]).view(ndarray)\n out += matmul(self[:self._end], x[k:])\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n out = matmul(part, x).view(ndarray)\n\n return(out)\n else:\n raise ValueError(\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\"\n \" (size {n} is different from {m})\".format(\n n=self._size,\n m=x.shape[0]\n )\n )\n elif self.ndim == 1 and x.ndim > 1:\n if self._size == x.shape[-2]:\n out_shape = *x.shape[:-2], x.shape[-1]\n out = empty(out_shape)\n out2 = space[:reduce(operator.mul, out_shape)].reshape(\n out_shape\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(self[self._begin:], x[..., :k, :], out)\n out += matmul(self[:self._end], x[..., k:, :], out2)\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(part, x, out).view(ndarray)\n\n return(out)\n else:\n raise ValueError(\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\"\n \" (size {n} is different from {m})\".format(\n n=self.shape[-2],\n m=x.shape[0]\n )\n )\n elif self.ndim == 2:\n if (self.shape[-1] == x.shape[-2]):\n out = empty(\n (*x.shape[:-2], self.shape[-1], x.shape[-2])\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(self[self._begin:], x, out[..., :k, :])\n matmul(self[:self._end], x, out[..., k:, :])\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(part, x, out)\n\n return(out.view(ndarray))\n\n else:\n raise ValueError(\n (\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->\"\n \"(n?,m?) (size {n} is different from {m})\"\n ).format(\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )\n else:\n if (self.shape[-1] == x.shape[-2]):\n self_shape = (self._size, *self.shape[1:-2])\n\n starexpr = tuple(\n zip_longest(self_shape, x.shape[:-2], fillvalue=1)\n )\n if star_can_broadcast(starexpr):\n broadcast_shape = tuple(\n starmap(\n lambda a, b: max(a, b),\n starexpr\n )\n )\n\n out = empty(\n (*broadcast_shape, self.shape[-2], x.shape[-1])\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n if x.ndim > 2:\n matmul(self[self._begin:], x[:k], out[:k])\n matmul(self[:self._end], x[k:], out[k:])\n else:\n matmul(self[self._begin:], x, out[:k])\n matmul(self[:self._end], x, out[k:])\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(part, x, out)\n\n return(out.view(ndarray))\n else:\n raise ValueError(\n (\n \"operands could not be broadcast together with\"\n \"remapped shapes [original->remapped]: \"\n \"{shape_b}->({shape_bn}, newaxis,newaxis) \"\n \"{shape_a}->({shape_an}, newaxis,newaxis) \"\n \"and requested shape ({n},{m})\"\n ).format(\n shape_a=self_shape,\n shape_b=x.shape,\n shape_an=self.shape[:-2].__str__()[:-1],\n shape_bn=x.shape[:-2].__str__()[:-1],\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )\n else:\n raise ValueError(\n (\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->\"\n \"(n?,m?) (size {n} is different from {m})\"\n ).format(\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )","def __mul__(self, other):\n if self.n != other.m:\n raise TypeError(\"Illegal dimensions for mul operator\")\n tmp = [[0 for _ in xrange(self.n)] for _ in xrange(other.m)]\n for i in xrange(self.n):\n for j in xrange(other.m):\n for k in xrange(other.n):\n tmp[i][j] += self.values[i][k] * other.values[k][j]\n res = []\n for i in tmp:\n res += i\n return simplematrix(self.n, other.m, res)","def naive_multiply(a, b):\n m = len(a) # Number of rows in first matrix\n k = len(b) # Number of rows in the second matrix\n res = []\n p = len(b[0])\n n = k\n for q in range(m):\n res.append([0])\n for q in range(m):\n for w in range(p - 1):\n res[q].append(0)\n for i in range(m):\n for j in range(p):\n for r in range(n):\n res[i][j] = a[i][r] * b[r][j] + res[i][j]\n return res","def combine_one_matrices(mul):\n factor, args = mul.as_coeff_matrices()\n new_args = [args[0]]\n\n for B in args[1:]:\n A = new_args[-1]\n if not isinstance(A, OneMatrix) or not isinstance(B, OneMatrix):\n new_args.append(B)\n continue\n new_args.pop()\n new_args.append(OneMatrix(A.shape[0], B.shape[1]))\n factor *= A.shape[1]\n\n return newmul(factor, *new_args)","def nonsquare_matrix_mult_opt1(matrix):\n\n #Setup openCL\n dev, ctx, queue = setup_CL()\n\n #openCL Kernel\n #Naive approach with local/private memory\n kernel_code = \"\"\"\n #define MATRIX_ROW_SIZE %(matrix_row_size)s\n #define MATRIX_COL_SIZE %(matrix_col_size)s\n\n __kernel void func(__global float* a, __global float* b, __global float* transposed) {\n\n unsigned int i = get_local_id(0);\n __local float tmp[MATRIX_ROW_SIZE*MATRIX_COL_SIZE];\n\n //Initialize tmp to 0\n //Initialize output b to 0 for this thread\n for(int k=0; k<MATRIX_COL_SIZE*MATRIX_ROW_SIZE; k++){\n tmp[k] = 0;\n }\n\n for(int k=0; k<MATRIX_ROW_SIZE; k++){\n b[k + MATRIX_ROW_SIZE*get_group_id(0)] = 0;\n }\n\n float localMatrix[MATRIX_ROW_SIZE*MATRIX_COL_SIZE];\n //Copy matrix to local\n for(int j=0; j < MATRIX_COL_SIZE; j++){\n localMatrix[i+get_local_size(0)*get_group_id(0)]=a[i+get_local_size(0)*get_group_id(0)];\n }\n\n\n //Transpose output\n transposed[i*MATRIX_ROW_SIZE+get_group_id(0)]=localMatrix[i+get_local_size(0)*get_group_id(0)];\n\n for(int j=0; j < MATRIX_ROW_SIZE; j++){\n tmp[j+MATRIX_ROW_SIZE*i] = localMatrix[i+get_local_size(0)*get_group_id(0)]*a[i+j*MATRIX_COL_SIZE];\n }\n\n // Store to output\n for(int j=0; j < MATRIX_ROW_SIZE; j++){\n for(int k=0; k < MATRIX_COL_SIZE; k++){\n if(i==0){\n b[j + MATRIX_ROW_SIZE*get_group_id(0)] += tmp[j+MATRIX_ROW_SIZE*k];\n }\n }\n }\n barrier(CLK_LOCAL_MEM_FENCE);\n }\n \"\"\"\n\n #Move data to device\n matrix_float = matrix.astype(np.float32)\n matrix_gpu = cl.array.to_device(queue, matrix_float)\n transposeMult_gpu = cl.array.empty(queue, (matrix.shape[0], matrix.shape[0]), np.float32)\n transposed_gpu = cl.array.empty(queue, (matrix.shape[1],matrix.shape[0]), np.float32)\n\n matrix_row_size = np.int32(matrix.shape[0])\n matrix_col_size = np.int32(matrix.shape[1])\n\n #Calculate workItems, workGroup size, workGroups for input\n matrix_val_count = matrix_float.shape[0]*matrix_float.shape[1]\n xWorkItems = min(int(matrix_row_size),1024)\n yWorkItems = min(int(matrix_col_size),1024)\n totalWorkItems = float(xWorkItems*yWorkItems)\n groups = np.int(max(np.ceil(matrix_val_count / xWorkItems),1))\n\n # print(\"workItems: %s, matrix_val_count: %s, groups: %s\" % (totalWorkItems, matrix_val_count, groups))\n\n # update template with current runtime requirements\n kernel = kernel_code % {\n 'matrix_row_size': matrix_row_size,\n 'matrix_col_size': matrix_col_size\n }\n\n #Launch kernel and time it\n #Set global ID, workItems, workGroups\n prg = cl.Program(ctx, kernel).build()\n start = time.time()\n event = prg.func(queue, (xWorkItems*yWorkItems,1),(groups,1), matrix_gpu.data, transposeMult_gpu.data, transposed_gpu.data)\n\n #event.wait()\n runtime = time.time()-start\n\n #Save output\n transposedMult = transposeMult_gpu.get()\n transposed = transposed_gpu.get()\n\n # print('openCL_opt1 %d x %d transpose-mult time: %.2E' % (matrix.shape[0], matrix.shape[1], runtime))\n # print('openCL_opt1_transposed==goldenTransposed: %s' % np.allclose(transposed, np.transpose(matrix)))\n # print('openCL_opt1_mult==goldenMult: %s' % np.allclose(transposedMult, matrix.dot(np.transpose(matrix))))\n if not(np.allclose(transposedMult, matrix.dot(np.transpose(matrix)))):\n # print('Original Matrix:\\n %s' % matrix)\n print('openCL_opt1 transposed val:\\n %s' % transposed)\n print('golden transpose-mult:\\n %s' % matrix.dot(np.transpose(matrix)))\n transposedMult[(transposedMult>0) & (transposedMult<1)] = -1\n print('openCL_opt1 mult val:\\n %s' % transposedMult)\n print('openCL_opt1 transpose-mult:\\n %s' % np.isclose(transposedMult,matrix.dot(np.transpose(matrix))))\n # print('--------------------')\n\n return [transposedMult, runtime]","def prod_mat(self,other):\n [rs,cs],[ro,co] = self.D,other.D\n assert cs == ro, \"tailles incompatibles\"\n return Mat([rs,co], lambda i,j : prod_scal(self.ligne(i),other.col(j)))","def norm2(x):\n if iscomplexobj(x):\n# t1=time.time()\n# mat1=x.real**2 + x.imag**2\n# t2=time.time()\n mat2=multiply(x.real,x.real) + multiply(x.imag,x.imag)\n# t3=time.time()\n# print('---------------------------')\n# print('pow time='+str(t2-t1))\n# print('multiply time='+str(t3-t2))\n# print('pow time/multiply time='+str((t2-t1)/(t3-t2)))\n# print('shape(x)='+str(shape(x)))\n# print('x.typecode='+str(x.typecode()))\n# print('mat1.typecode='+str(mat1.typecode()))\n# print('mat2.typecode='+str(mat2.typecode()))\n# if len(shape(x))==1:\n# print('type(x[0])='+str(type(x[0])))\n# print('type(mat1[0])='+str(type(mat1[0])))\n# print('type(mat2[0])='+str(type(mat2[0])))\n# else:\n# print('type(x[0,0])='+str(type(x[0,0])))\n# print('type(mat1[0,0])='+str(type(mat1[0,0])))\n# print('type(mat2[0,0])='+str(type(mat2[0,0])))\n return mat2\n else:\n return multiply(x,x)","def __matmul__(self, qubit):\n if isinstance(qubit, str):\n qubit = self.get_index(qubit)\n return self.compiled[qubit].y","def rmatmul(self, x, work_buffer) -> ndarray:\n x = asarray(x)\n space = asarray(work_buffer).flat\n\n if (x.ndim == 0):\n ValueError(\n \"matmul: Input operand 1 does not have enough dimensions \"\n \"(has 0, gufunc core with signature (n?,k),(k,m?)->(n?,m?) \"\n \"requires 1\"\n )\n\n if x.ndim == 1 and self.ndim == 1:\n # Dot product\n if x.shape[0] == self._size:\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n out = matmul(x[:k], self[self._begin:]).view(ndarray)\n out += matmul(x[k:], self[:self._end]).view(ndarray)\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n out = matmul(x, part)\n\n return(out.view(ndarray))\n else:\n raise ValueError(\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\"\n \" (size {n} is different from {m})\".format(\n n=self._size,\n m=x.shape[0]\n )\n )\n elif x.ndim == 1 and self.ndim > 1:\n if x.shape[0] == self.shape[-2]:\n if self.ndim == 2:\n out = empty(self.shape[-1])\n out2 = space[:self.shape[-1]]\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(x[:k], self[self._begin:], out)\n out += matmul(x[k:], self[:self._end], out2)\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(x, part, out)\n\n return(out)\n else:\n out = empty(\n (self._size, *self.shape[1:-2], self.shape[-1])\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(x, self[self._begin:], out[:k])\n matmul(x, self[:self._end], out[k:])\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(x, part, out)\n\n return(out)\n else:\n raise ValueError(\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\"\n \" (size {n} is different from {m})\".format(\n n=self.shape[-2],\n m=x.shape[0]\n )\n )\n elif x.ndim > 1 and self.ndim == 1:\n if x.shape[-1] == self.shape[0]:\n out = empty(x.shape[:-1])\n out2 = space[:reduce(operator.mul, x.shape[:-1])].reshape(\n x.shape[:-1]\n )\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(x[..., :, :k], self[self._begin:], out)\n out += matmul(x[..., :, k:], self[:self._end], out2)\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(x, part, out)\n\n return(out)\n else:\n raise ValueError(\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\"\n \" (size {n} is different from {m})\".format(\n n=self.shape[-2],\n m=x.shape[0]\n )\n )\n elif self.ndim == 2:\n if (x.shape[-1] == self.shape[-2]):\n out_shape = (*x.shape[:-1], self.shape[-2])\n out = empty(out_shape)\n out2 = space[:reduce(operator.mul, out_shape)].reshape(\n out_shape\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n matmul(x[..., :, :k], self[self._begin:], out)\n out += matmul(x[..., :, k:], self[:self._end], out2)\n\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(x, part, out)\n\n return(out.view(ndarray))\n\n else:\n raise ValueError(\n (\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->\"\n \"(n?,m?) (size {n} is different from {m})\"\n ).format(\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )\n else:\n if (x.shape[-1] == self.shape[-2]):\n self_shape = (self._size, *self.shape[1:-2])\n\n starexpr = tuple(\n zip_longest(self_shape, x.shape[:-2], fillvalue=1)\n )\n if star_can_broadcast(starexpr):\n broadcast_shape = tuple(\n starmap(\n lambda a, b: max(a, b),\n starexpr\n )\n )\n\n out = empty(\n (*broadcast_shape, x.shape[-2], self.shape[-1])\n )\n\n if self.fragmented:\n k = self._capacity - self._begin # fragmentation index\n\n if x.ndim > 2:\n matmul(x[:k], self[self._begin:], out[:k])\n matmul(x[k:], self[:self._end], out[k:])\n else:\n matmul(x, self[self._begin:], out[:k])\n matmul(x, self[:self._end], out[k:])\n else:\n if self._begin < self._end:\n part = self[self._begin:self._end]\n elif self._end == 0:\n part = self[self._begin:]\n\n matmul(x, part, out)\n\n return(out.view(ndarray))\n else:\n raise ValueError(\n (\n \"operands could not be broadcast together with\"\n \"remapped shapes [original->remapped]: \"\n \"{shape_b}->({shape_bn}, newaxis,newaxis) \"\n \"{shape_a}->({shape_an}, newaxis,newaxis) \"\n \"and requested shape ({n},{m})\"\n ).format(\n shape_a=self_shape,\n shape_b=x.shape,\n shape_an=self.shape[:-2].__str__()[:-1],\n shape_bn=x.shape[:-2].__str__()[:-1],\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )\n else:\n raise ValueError(\n (\n \"matmul: Input operand 1 has a mismatch in its core \"\n \"dimension 0, with gufunc signature (n?,k),(k,m?)->\"\n \"(n?,m?) (size {n} is different from {m})\"\n ).format(\n n=self.shape[-1],\n m=x.shape[-2]\n )\n )","def matmul(x, y):\n return np.matmul(x, y)","def compute(self, node, input_vals):\n mat_A = input_vals[0]\n mat_B = input_vals[1]\n if node.matmul_attr_trans_A:\n mat_A = mat_A.T\n if node.matmul_attr_trans_B:\n mat_B = mat_B.T\n return np.matmul(mat_A, mat_B)","def __imul__(self, other):\n\t\t#print(\"imul\")\n\t\t# merge other branch\n\t\tself.graph.update(other.graph)\n\t\tself.bottoms.update(other.bottoms)\n\t\tself.output_shape.update(other.output_shape)\n\t\tlayer_name = \"mul_{}\".format(len(self.graph))\n\t\tself.graph[layer_name] = layer_name\n\t\tself.bottoms[layer_name] = [self.cur_id, other.cur_id]\n\t\tself.output_shape[layer_name] = self.cur_tensor.size()\n\t\tself.cur_id = layer_name\n\t\t# save memory\n\t\tdel other\n\t\treturn self","def __mul__(self, A):\n pass","def test_gemm_opt0():\r\n X, Y, Z, a, b = XYZab()\r\n\r\n just_gemm([X, Y, Z, a, b], [T.dot(X, Y) * a + Z * b])\r\n just_gemm([X, Y, Z, a, b], [a * T.dot(X, Y) + b * Z])\r\n just_gemm([X, Y, Z, a, b], [b * Z + a * T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [T.dot(X, Y) * a - Z * b])\r\n just_gemm([X, Y, Z, a, b], [a * T.dot(X, Y) - b * Z])\r\n just_gemm([X, Y, Z, a, b], [b * Z - a * T.dot(X, Y)])\r\n\r\n #with transposes (transposes should be pushed through dot in canonicalize)\r\n just_gemm([X, Y, Z, a, b], [b * Z.T - a * T.dot(Y.T, X.T)])\r\n just_gemm([X, Y, Z, a, b], [b * Z.T + a * b * T.dot(X, Y).T])\r\n just_gemm([X, Y, Z, a, b], [b * Z + a * T.dot(X, Y).T],\r\n ishapes=[(5, 3), (3, 4), (4, 5), (), ()])\r\n\r\n #with N multiplications instead of just one\r\n just_gemm([X, Y, Z, a, b], [(b * b) * Z * a + (a * a) * T.dot(X, Y) * b])\r\n just_gemm([X, Y, Z, a, b], [Z + T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [Z * b + T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [Z + a * b * a * T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [(b * b) * Z * a - (a * a) * T.dot(X, Y) * b])\r\n just_gemm([X, Y, Z, a, b], [Z - T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [Z * b - T.dot(X, Y)])\r\n just_gemm([X, Y, Z, a, b], [Z - a * b * a * T.dot(X, Y)])","def sparse_matmul(A: SparseTensor, B: SparseTensor, out: torch.Tensor) -> torch.Tensor:\n if A.nnz() == 0 or B.nnz() == 0:\n return out\n\n if A.is_cuda:\n return _sparse_matmul_cuda(A, B, out)\n else:\n return _sparse_matmul_cpu(A, B, out)","def matmulreduce(arr, reverse=False):\n if reverse:\n items = range(len(arr) - 1, -1, -1)\n else:\n items = range(len(arr))\n P = arr[items[0]]\n for i in items[1:]:\n P = logical_matmul(P, arr[i]) * 1\n return P","def matrix_multiply_mapper(m, element):\n\n name, i, j, value = element\n\n if name == \"A\":\n for k in range(m):\n yield ((i, k), (j, value))\n \n else:\n for k in range(m):\n yield((k, j), (i, value))","def __mul__(self, other):\n if isinstance(other, Vector):\n # Matrix vector product\n v = Vector(list())\n for n in range(len(other.vectors)):\n v += scale(other.vectors[n][n], self.vectors[n])\n return v\n elif isinstance(other, Matrix):\n # Matrix matrix product\n if self.n != other.m:\n raise ValueError(\"Wrong fucking sizes, nøøb\")\n\n selfVectors = self.vectors\n selfColVectors = self.transpose()\n otherVectors = other.vectors\n otherColVectors = other.transpose()\n vectors = list()\n for col in range(other.n):\n cordinator = []\n\n for row in range(self.m):\n coord = 0\n\n for k in range(other.m):\n coord += (\n selfVectors[row].coords[k]\n * otherColVectors.vectors[col].coords[k]\n )\n\n cordinator.append(coord)\n\n v = Vector(cordinator)\n vectors.append(v)\n matrix = Matrix(vectors)\n matrix = matrix.transpose()\n return matrix\n elif isinstance(other, int) or isinstance(other, float): # Skalering af matrix\n for i in range(len(self.vectors)):\n self.vectors[i] *= other\n else:\n raise ValueError(\n \"Can only multiply Matrix with Matrix, Vector, Integer or Float\"\n )","def tt_dense_matmul(tt_matrix_a, matrix_b, convert_to_tensors=False):\n\n ndims = tt_matrix_a.ndims\n a_columns = tt_matrix_a.shape[1]\n b_rows = matrix_b.shape[0]\n if a_columns is not None and b_rows is not None:\n if a_columns != b_rows:\n raise ValueError('Arguments shapes should align got %d and %d instead.' % (tt_matrix_a.shape, matrix_b.shape))\n\n a_shape = tt_matrix_a.shape\n a_raw_shape = tt_matrix_a.raw_shape\n b_shape = matrix_b.shape\n a_ranks = tt_matrix_a.ranks\n\n # If A is (i0, ..., id-1) x (j0, ..., jd-1) and B is (j0, ..., jd-1) x K,\n # data is (K, j0, ..., jd-2) x jd-1 x 1\n data = matrix_b.transpose(0, 1)\n #Liancheng +contiguous\n data = data.contiguous().view(-1, a_raw_shape[1][-1], 1)\n\n for core_idx in reversed(range(ndims)):\n curr_core = tt_matrix_a.tt_cores[core_idx]\n #print('curr_core device:',curr_core.device)\n\n #Liancheng\n curr_core = curr_core.to(data.device)\n\n # On the k = core_idx iteration, after applying einsum the shape of data\n # becomes ik x (ik-1..., id-1, K, j0, ..., jk-1) x rank_k\n #print('curr_core device:',curr_core.device)\n #print('data device:',data.device)\n data = torch.einsum('aijb,rjb->ira', curr_core, data)\n if core_idx > 0:\n # After reshape the shape of data becomes\n # (ik, ..., id-1, K, j0, ..., jk-2) x jk-1 x rank_k\n new_data_shape = (-1, a_raw_shape[1][core_idx - 1], a_ranks[core_idx])\n data = data.contiguous().view(new_data_shape)\n\n # At the end the shape of the data is (i0, ..., id-1) x K\n return data.view(a_shape[0], b_shape[1])","def Chain(A, B):\n return _prodOperator(B, A)","def __mul__(self, other):\n if hasattr(other, 'as_homogenous_transformation'):\n return basis(homogenous_transformation = self.as_homogenous_transformation() * other.as_homogenous_transformation())\n elif hasattr(other, 'n'):\n if other.n == (3,1):\n b = matrix.col((other[0], other[1], other[2], 1))\n elif other.n == (4,1):\n b = other\n else:\n raise TypeError(b, \"Incompatible matrices\")\n p = self.as_homogenous_transformation() * b\n if other.n == (3,1):\n return matrix.col(p[0:3])\n else:\n return p\n else:\n raise TypeError(b)","def inner_product_batch(pattern_stack_one, pattern_num_one, pattern_stack_two, pattern_num_two):\n\n \"\"\"\n Notice that the two stacks can be different. So we can not deduce the lower triangular pattern from the \n other half.\n \"\"\"\n holder = np.zeros((pattern_num_one, pattern_num_two))\n for l in range(pattern_num_one):\n for m in range(pattern_num_two):\n holder[l, m] = np.sum(np.multiply(pattern_stack_one[l], pattern_stack_two[m]))\n\n return holder","def multiply(traj, result_list):\n z=traj.x*traj.y\n result_list[traj.v_idx] = z","def complex_multiplication(c1,c2,cr):\n cr[0] = c1[0]*c2[0] - c1[1]*c2[1]\n cr[1] = c1[0]*c2[1] + c1[1]*c2[0]\n return cr"],"string":"[\n \"def _multi_matmul_chain_order(arrays):\\n n = len(arrays)\\n # p stores the dimensions of the matrices\\n # Example for p: A_{10x100}, B_{100x5}, C_{5x50} --> p = [10, 100, 5, 50]\\n # Using -2 to generalize for shapes that are more than 2 dimmensions\\n p = [a.shape[-2] for a in arrays] + [arrays[-1].shape[-1]]\\n # m is a matrix of costs of the subproblems\\n # m[i,j]: min number of scalar multiplications needed to compute A_{i..j}\\n m = np.zeros((n, n), dtype=np.double)\\n # s is the actual ordering\\n # s[i, j] is the value of k at which we split the product A_i..A_j\\n s = np.empty((n, n), dtype=np.intp)\\n\\n for l in range(1, n):\\n for i in range(n - l):\\n j = i + l\\n m[i, j] = np.inf\\n for k in range(i, j):\\n q = m[i, k] + m[k + 1, j] + p[i] * p[k + 1] * p[j + 1]\\n if q < m[i, j]:\\n m[i, j] = q\\n s[i, j] = k # Note that Cormen uses 1-based index\\n return s\",\n \"def chain_matmul_square(As):\\n\\n As_matmul = As\\n while As_matmul.shape[0] > 1:\\n if As_matmul.shape[0] % 2:\\n A_last = As_matmul[-1:]\\n else:\\n A_last = None\\n \\n As_matmul = torch.matmul(As_matmul[0:-1:2], As_matmul[1::2])\\n if A_last is not None:\\n As_matmul = torch.cat([As_matmul, A_last], dim=0)\\n \\n return As_matmul.squeeze(0)\",\n \"def recursive_multiply(a, b):\\n if len(a) == 2:\\n return naive_multiply(a, b)\\n\\n a11 = a[0:int(len(a) / 2)]\\n for index, row in enumerate(a11):\\n a11[index] = row[0:int(len(row) / 2)]\\n\\n a12 = a[0:int(len(a) / 2)]\\n for index, row in enumerate(a12):\\n a12[index] = row[int(len(a) / 2):len(a)]\\n\\n a21 = a[int(len(a) / 2):len(a)]\\n for index, row in enumerate(a21):\\n a21[index] = row[0:int(len(row) / 2)]\\n\\n a22 = a[int(len(a) / 2):len(a)]\\n for index, row in enumerate(a22):\\n a22[index] = row[int(len(a) / 2):len(a)]\\n\\n b11 = b[0:int(len(b) / 2)]\\n for index, row in enumerate(b11):\\n b11[index] = row[0:int(len(row) / 2)]\\n\\n b12 = b[0:int(len(b) / 2)]\\n for index, row in enumerate(b12):\\n b12[index] = row[int(len(b) / 2):len(b)]\\n\\n b21 = b[int(len(b) / 2):len(b)]\\n for index, row in enumerate(b21):\\n b21[index] = row[0:int(len(row) / 2)]\\n\\n b22 = b[int(len(b) / 2):len(b)]\\n for index, row in enumerate(b22):\\n b22[index] = row[int(len(b) / 2):len(b)]\\n\\n c11 = matrix_add(recursive_multiply(a11, b11), recursive_multiply(a12, b21)) # C11 = A11*B11 + A12*B21\\n c12 = matrix_add(recursive_multiply(a11, b12), recursive_multiply(a12, b22)) # C12 = A11*B12 + A12*B22\\n c21 = matrix_add(recursive_multiply(a21, b11), recursive_multiply(a22, b21)) # C21 = A21*B11 + A22*B21\\n c22 = matrix_add(recursive_multiply(a21, b12), recursive_multiply(a22, b22)) # C22 = A21*B12 + A22*B22\\n\\n # Append c12 to c11\\n for row_index, row in enumerate(c11):\\n for col_index, col in enumerate(c12):\\n row.append(c12[row_index][col_index])\\n\\n # Append c22 to c21\\n for row_index, row in enumerate(c21):\\n for col_index, col in enumerate(c12):\\n row.append(c22[row_index][col_index])\\n\\n # Append c21 to c11\\n for i in c21:\\n c11.append(i)\\n\\n return c11\",\n \"def lazy_matrix_mul(m_a, m_b):\\n return np.dot(m_a, m_b)\",\n \"def __mul__(self, other):\\n #\\n # TODO - your code here\\n #\\n final_matrix = []\\n for i in range(self.h):\\n temp_row = []\\n for j in range(other.w):\\n # take dot-product of row of\\n # matrix in 1st arg with col of\\n # matrix in 2nd arg\\n temp_row.append(dot_product(get_row(self.g, i), get_col(other.g, j)))\\n final_matrix.append(temp_row)\\n return Matrix(final_matrix)\\n # TODO - your code here\",\n \"def matmul():\\n\\n if RESULT_IN_NVRAM:\\n matrix_c = ResultMatrixInDaos()\\n else:\\n matrix_c = ResultMatrixInMemory()\\n\\n # This could be trivially optimized by reordering indexes\\n # and caching either a_block or b_block (assuming C in-memory).\\n # *However* it would result in unfair comparisons with the \\n # previous implementation used elsewhere.\\n # Using the naive algorithm makes sense for a raw comparison.\\n for i in range(MATRIXSIZE):\\n for j in range(MATRIXSIZE):\\n partial_result_block = np.zeros((BLOCKSIZE, BLOCKSIZE))\\n\\n for k in range(MATRIXSIZE):\\n a_block = np.fromstring(\\n DAOS_KV[\\\"A%02d%02d\\\" % (i, k)],\\n dtype=NP_FROMSTRING_DTYPE\\n ).reshape((BLOCKSIZE, BLOCKSIZE))\\n\\n b_block = np.fromstring(\\n DAOS_KV[\\\"B%02d%02d\\\" % (k, j)],\\n dtype=NP_FROMSTRING_DTYPE\\n ).reshape((BLOCKSIZE, BLOCKSIZE))\\n\\n partial_result_block += a_block @ b_block\\n \\n matrix_c[i,j] = partial_result_block\\n\\n return matrix_c\",\n \"def lazy_matrix_mul(m_a, m_b):\\n return (np.matmul(m_a, m_b))\",\n \"def matrix_mult(m1, m2):\\n pass\",\n \"def _multi_matmul(arrays, order, i, j, constant=False) -> Tensor:\\n if i == j:\\n return arrays[i]\\n else:\\n return matmul(\\n _multi_matmul(arrays, order, i, order[i, j], constant),\\n _multi_matmul(arrays, order, order[i, j] + 1, j, constant),\\n constant,\\n )\",\n \"def lazy_matrix_mul(m_a, m_b):\\n return np.matmul(np.array(m_a), np.array(m_b))\",\n \"def matrixMul(self, matrix, matrix2):\\n matrix0 = matrix[:]\\n matrix[0] = matrix0[0] * matrix2[0] + matrix0[2]*matrix2[1] # + matrix0[4]*0\\n matrix[1] = matrix0[1] * matrix2[0] + matrix0[3]*matrix2[1] # + matrix0[5]*0\\n matrix[2] = matrix0[0] * matrix2[2] + matrix0[2]*matrix2[3] # + matrix0[4]*0\\n matrix[3] = matrix0[1] * matrix2[2] + matrix0[3]*matrix2[3] # + matrix0[5]*0\\n matrix[4] = matrix0[0] * matrix2[4] + matrix0[2]*matrix2[5] + matrix0[4]\\n matrix[5] = matrix0[1] * matrix2[4] + matrix0[3]*matrix2[5] + matrix0[5]\",\n \"def lazy_matrix_mul(m_a, m_b):\\n m_a = np.array(m_a)\\n m_b = np.array(m_b)\\n\\n return m_a.dot(m_b)\",\n \"def Multiply(M1,M2):\\r\\n M3=[]\\r\\n w=0\\r\\n while w<len(M2[0]):\\r\\n tap=[]\\r\\n t=0\\r\\n while t<len(M2):\\r\\n tap.append(M2[t][w])\\r\\n t=t+1\\r\\n M3.append(tap)\\r\\n w=w+1\\r\\n M=[]\\r\\n # Multiplying matrices\\r\\n k=0\\r\\n sums=0\\r\\n while k<len(M1):\\r\\n j=0\\r\\n mpy=[]\\r\\n while j<len(M3):\\r\\n p=0\\r\\n sums=0\\r\\n while p<len(M3[j]):\\r\\n temp = (M1[k][p])*(M3[j][p])\\r\\n sums=sums+temp\\r\\n p=p+1\\r\\n mpy.append(sums)\\r\\n j=j+1\\r\\n M.append(mpy)\\r\\n k=k+1\\r\\n return M\",\n \"def _z2matmul(self, left, right):\\n prod = np.mod(np.dot(left, right), 2)\\n return prod\",\n \"def __matmul__(self, q: np.ndarray) -> np.ndarray:\\n return self.product(q)\",\n \"def matrix_chain_dynamic(dimensions, n):\\n\\n m = [[-1 for _ in range(n)] for _ in range(n)]\\n s = [[0 for _ in range(n)] for _ in range(n)]\\n\\n # multiplying matrix by itself\\n for i in range(1, n):\\n m[i][i] = 0\\n\\n for length in range(2, n):\\n for i in range(1, n - length + 1):\\n j = i + length - 1\\n for k in range(i, j):\\n cost = m[i][k] + m[k + 1][j] + dimensions[i - 1] * dimensions[k] * dimensions[j]\\n if cost > m[i][j]:\\n m[i][j] = cost\\n # index if splitting\\n s[i][j] = k\\n return m, s\",\n \"def combined_costs(matrix_MSLL_IO):\\r\\n return\",\n \"def __matmul__(self, tensor):\\n return self.matmul(tensor)\",\n \"def matrix_mult(M, vector1, vector2):\\n out = None\\n ### YOUR CODE HERE\\n pass\\n ### END YOUR CODE\\n\\n return out\",\n \"def matrix_mult(M, vector1, vector2):\\n out = None\\n ### YOUR CODE HERE\\n pass\\n ### END YOUR CODE\\n\\n return out\",\n \"def Mult(A, B, C_, IM, M):\\n for i in range(M):\\n for j in range(M):\\n for k in range(M):\\n C_[i, j] = A[i, j] * B[i, j]\",\n \"def __mul__(self, other):\\r\\n T = type(other)\\r\\n # mat4*scalar\\r\\n if T==types.FloatType or T==types.IntType or T==types.LongType:\\r\\n return mat4(map(lambda x,other=other: x*other, self.mlist))\\r\\n # mat4*vec3\\r\\n if isinstance(other, _vec3):\\r\\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\\r\\n w = float(m41*other.x + m42*other.y + m43*other.z + m44)\\r\\n return _vec3(m11*other.x + m12*other.y + m13*other.z + m14, \\r\\n m21*other.x + m22*other.y + m23*other.z + m24, \\r\\n m31*other.x + m32*other.y + m33*other.z + m34)/w\\r\\n # mat4*vec4\\r\\n if isinstance(other, _vec4):\\r\\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\\r\\n return _vec4(m11*other.x + m12*other.y + m13*other.z + m14*other.w, \\r\\n m21*other.x + m22*other.y + m23*other.z + m24*other.w, \\r\\n m31*other.x + m32*other.y + m33*other.z + m34*other.w,\\r\\n m41*other.x + m42*other.y + m43*other.z + m44*other.w)\\r\\n # mat4*mat4\\r\\n if isinstance(other, mat4):\\r\\n m11,m12,m13,m14,m21,m22,m23,m24,m31,m32,m33,m34,m41,m42,m43,m44 = self.mlist\\r\\n n11,n12,n13,n14,n21,n22,n23,n24,n31,n32,n33,n34,n41,n42,n43,n44 = other.mlist\\r\\n return mat4( m11*n11+m12*n21+m13*n31+m14*n41,\\r\\n m11*n12+m12*n22+m13*n32+m14*n42,\\r\\n m11*n13+m12*n23+m13*n33+m14*n43,\\r\\n m11*n14+m12*n24+m13*n34+m14*n44,\\r\\n\\r\\n m21*n11+m22*n21+m23*n31+m24*n41,\\r\\n m21*n12+m22*n22+m23*n32+m24*n42,\\r\\n m21*n13+m22*n23+m23*n33+m24*n43,\\r\\n m21*n14+m22*n24+m23*n34+m24*n44,\\r\\n\\r\\n m31*n11+m32*n21+m33*n31+m34*n41,\\r\\n m31*n12+m32*n22+m33*n32+m34*n42,\\r\\n m31*n13+m32*n23+m33*n33+m34*n43,\\r\\n m31*n14+m32*n24+m33*n34+m34*n44,\\r\\n\\r\\n m41*n11+m42*n21+m43*n31+m44*n41,\\r\\n m41*n12+m42*n22+m43*n32+m44*n42,\\r\\n m41*n13+m42*n23+m43*n33+m44*n43,\\r\\n m41*n14+m42*n24+m43*n34+m44*n44)\\r\\n # unsupported\\r\\n else:\\r\\n raise TypeError, \\\"unsupported operand type for *\\\"\",\n \"def fast_matmul_2d(self, other):\\n if isinstance(other, np.ndarray):\\n mat_tensor = PaillierTensor(other, partitions=self.partitions)\\n return self.fast_matmul_2d(mat_tensor)\\n\\n if isinstance(other, CTableABC):\\n other = PaillierTensor(other)\\n\\n func = self._vector_mul\\n ret_mat = self._obj.join(other.get_obj(), lambda vec1, vec2: (vec1, vec2)).applyPartitions(func).reduce(\\n lambda mat1, mat2: mat1 + mat2)\\n\\n return ret_mat\",\n \"def mmultiply(self, matrix):\\n try:\\n result_matrix = [[0 for row in range(len(self.matrix))] for col in range(len(matrix[0]))]\\n for i in range(len(self.matrix)):\\n for j in range(len(matrix[0])):\\n for k in range(len(matrix)):\\n result_matrix[i][j] += self.matrix[i][k] * matrix[k][j]\\n self.matrix = result_matrix\\n except IndexError:\\n pass\\n pass\",\n \"def test_matrix_product(self, use_cache):\\n\\n key = jrandom.PRNGKey(0)\\n dim = 50\\n max_power = 25\\n\\n matrix = jrandom.normal(key, (dim, dim)) / 10\\n vector = jnp.ones((dim,), dtype=jnp.float32)\\n\\n if use_cache:\\n mpstate = model_utils.CachedMatrixPowerState.precompute(matrix, max_power)\\n else:\\n mpstate = model_utils.LazyMatrixPowerState(matrix)\\n\\n for t in range(max_power):\\n result = mpstate.matrix_power_multiply(vector, t)\\n expected = np.linalg.matrix_power(matrix, t) @ vector\\n\\n np.testing.assert_array_almost_equal(result, expected, decimal=1)\",\n \"def my_matmul(x, y):\\n ##\\n cmd = getattr(th, \\\"matmul\\\")\\n x1, x2 = my_cut(x)\\n y1, y2 = my_cut(y)\\n x2y1 = cmd(x2, y1)\\n x1y2 = cmd(x1, y2)\\n x2y2 = cmd(x2, y2)\\n ret = (x2y1 + x1y2) % int24field * int24field + x2y2\\n ret = int48module(ret)\\n return ret\",\n \"def multiply_matrices(list):\\n # Section 1: Start matrix product using 1st matrix in list\\n matrix_product = list[0]\\n\\n # Section 2: Loop thru list to create product\\n for matrix in list[1:]:\\n matrix_product = matrix_multiply(matrix_product, matrix)\\n\\n return matrix_product\",\n \"def mat_mul(mat1, mat2):\\n\\n if len(mat1[0]) == len(mat2):\\n\\n mat2 = matrix_transpose(mat2)\\n response = []\\n\\n for row in range(len(mat1)):\\n response.append(\\n [\\n sum(dot_product(mat1[row], mat2[column]))\\n for column in range(len(mat2))\\n ]\\n )\\n\\n return response\\n\\n else:\\n return None\",\n \"def u_update(eta_0, eta, eta_lin, w_0, w, w_lin, eta_T_H_L_stacked, premultiplied_lhs = None, nnls_max_iter=50): \\n # PREMULTIPLIED LHS IS AN EXTRA ARGUMENT! Set it to None and add solver! \\n \\\"\\\"\\\"In the following +[[]] and [:-1] are added to keep thing 1dim array of objects and still multiply it elemtwisely\\\"\\\"\\\" \\n# #B.append([]) #THIS IS WRONG, CHANGES THE LIST \\n# B_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(B+[[]])[:-1], axis = 0) \\n# A_ls = np.concatenate([(1/np.sqrt(2*eta0))*A, B_concat], axis = 0) \\n# #print(np.array(B).shape) \\n# #print(w[0].shape) \\n# #print(w, eta) \\n# #w.append([]) THIS IS WRONG, CHANGES THE LIST \\n# w_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(w+[[]])[:-1], axis = 0) #[:-1] Added as a hack to keep it one-dim array of objects \\n# eta_w = np.expand_dims(1/np.sqrt(2*eta),1)*np.array(w) \\n# print(eta_w.shape) \\n# b_ls = np.concatenate([(1/np.sqrt(2*eta_0))*w_0, eta_w.flatten()], axis = 0) \\n #Use correct broadcasting?\\n w_concat = np.concatenate((1/np.sqrt(2*eta))*np.array(w+[[]])[:-1], axis = 0) #[:-1] Added as a hack to keep it one-dim array of objects \\n b_ls = np.concatenate([(1/np.sqrt(2*eta_0))*w_0, w_concat, (1/np.sqrt(2*eta_lin))*w_lin], axis = 0) \\n# print(np.sum(eta_w.flatten() != w_concat)) \\n# premultiplied_time_start = time.time() \\n# premultiplied_lhs = eta_T_H_stacked.T.dot(eta_T_H_stacked).toarray() \\n# premultiplied_time_end = time.time() \\n# print('premultiplying took {}'.format(premultiplied_time_end - premultiplied_time_start)) \\n# premultiplied_rhs = eta_T_H_stacked.T.dot(b_ls) \\n# u_next = nnls_predotted(premultiplied_lhs, premultiplied_rhs, tol=1e-5) \\n# print(eta_T_H_stacked.shape, b_ls.shape) \\n# A_ls_t_b = eta_T_H_stacked.T.dot(b_ls) \\n# w =scipy.sparse.linalg.spsolve_triangular(RT, A_ls_t_b, lower = True) \\n# x = scipy.sparse.linalg.spsolve_triangular(R, w, lower = False) \\n# u_next = x \\n u_next = scipy.optimize.lsq_linear(eta_T_H_L_stacked, b_ls, bounds = (0, np.inf), tol=1e-3, lsmr_tol=1e-3, max_iter=nnls_max_iter, verbose=1).x \\n# u = scipy.optimize.lsq_linear(premultiplied_lhs, premultiplied_rhs, bounds = (0, np.inf), tol=1e-5).x \\n return u_next\",\n \"def python_nonsquare_matrix_mult(matrix):\\n\\n transposed_matrix = np.zeros([matrix.shape[1],matrix.shape[0]])\\n start = time.time()\\n # for i in range(matrix.shape[0]):\\n # for j in range(matrix.shape[1]):\\n # transposed_matrix[j,i] = matrix[i,j]\\n\\n transposed_matrix = np.transpose(matrix)\\n product = matrix.dot(transposed_matrix)\\n\\n # transposed_matrix = np.transpose(matrix)\\n end = time.time()-start\\n\\n # print(\\\"Python Golden Transpose: %s\\\" % product)\\n # print('python transpose time: %.2E' % end)\\n return [product, end]\",\n \"def multiM(*args):\\r\\n filas_1,filas_2 = len(args[0]),len(args[1])\\r\\n columnas_1,columnas_2 = len(args[0][0]),len(args[1][0])\\r\\n matriz_r = []\\r\\n for k in range(filas_1):\\r\\n matriz_r.append([0]*columnas_2)\\r\\n for i in range(columnas_2):\\r\\n matriz_r[k][i] = 0\\r\\n for i in range(filas_1):\\r\\n for j in range(columnas_1):\\r\\n for k in range(columnas_2):\\r\\n matriz_r[i][k] = matriz_r[i][k] + args[0][i][j] * args[1][j][k]\\r\\n return matriz_r\",\n \"def matrix_mult(m1, m2):\\n output = []\\n for rowIndex, row in enumerate(m1): #go through rows in m1\\n new_row = []\\n for columnIndex in range(len(m2[0])): #go through indices for each column of m2\\n sum = 0\\n for index3 in range(len(row)):\\n product = m1[rowIndex][index3] * m2[index3][columnIndex]\\n sum += product\\n new_row.append(sum)\\n output.append(new_row)\\n return output\\n \\n \\n #output = []\\n #first for loop corresponds to the rows of my output matrix and loops through the rows of m1 (enumerate)\\n #create an empty new row\\n # second for loop, loops through columns of m2\\n # create sum variable, initialize it with zero\\n # third for loop, multiplies the index of the row in m1 times the index of the column in m2\\n # add sum to product and assign this to the sum variable\\n # append sum to new row\\n # append new row to output\\n # return output\",\n \"def matrix_add():\",\n \"def it_matrixpower(p,t,n,root_field=RR):\\n assert n>=2, \\\"Carleman matrix must at least be of size 2 to retrieve the coefficients. But given was \\\" + repr(n)\\n CM = p.carleman_matrix(n)\\n ev = CM.charpoly().roots(root_field)\\n assert len(ev) == n, \\\"Carleman matrix must have exactly \\\" + repr(n) + \\\"eigenvalues, but has \\\" + repr(len(ev))\\n\\n Char = [0]*n\\n for k in range(n):\\n #here is possibility for improvement of precision\\n #to separate the fractional from the root parts\\n #expanding the product\\n Char[k] = CM - ev[k][0]*identity_matrix(n)\\n\\n #we want to have the first row of the product of the matrices\\n #thatswhy we mulitply in front with:\\n prod = vector(p.K,[0,1]+[0]*(n-2))\\n prodwo = [0]*n\\n for k in range(n):\\n prodwo[k]=prod #these are the first terms until k-1\\n\\n #no need to continue\\n if k == n-1:\\n break\\n\\n #and we add the terms starting with k+1\\n for i in range(k+1,n):\\n prodwo[k] = prodwo[k] * Char[i]\\n\\n prod = prod * Char[k]\\n\\n sprodwo = [0]*n\\n for k in range(n):\\n if k==0:\\n sprodwo[k] = ev[k][0] - ev[1][0]\\n start = 2\\n else:\\n sprodwo[k] = ev[k][0] - ev[0][0]\\n start = 1\\n\\n for i in range(start,n):\\n if i != k:\\n sprodwo[k] = sprodwo[k] * (ev[k][0] - ev[i][0])\\n\\n res = ev[0][0]**t/sprodwo[0] * prodwo[0]\\n for k in range(1,n):\\n res += ev[k][0]**t/sprodwo[k]*prodwo[k]\\n\\n return res.list()\",\n \"def _mul(*args):\\n\\treturn functools.reduce(numpy.dot, args)\",\n \"def matrixMultiplication(self, n, id, context):\\n\\n print(\\\"id: {}\\\".format(id))\\n # Create one matrix\\n f = 1\\n m1 = []\\n for x in range(n):\\n row = []\\n for y in range(n):\\n row.append(f)\\n f = f + 1\\n m1.append(row)\\n # The second matrix is equal to the first matrix\\n m2 = m1\\n print(\\\"m2: {}\\\".format(m2))\\n\\n # Multiply matrices\\n m3 = []\\n for i in range(n):\\n row = []\\n for j in range(n):\\n sum = 0\\n for k in range(n):\\n sum = sum + m1[i][k] * m2[k][j]\\n row.append(sum)\\n m3.append(row)\\n\\n sum = 0\\n # add the entries\\n for i in range(n):\\n for j in range(n):\\n sum = sum + m3[i][j]\\n\\n print(\\\"Result of multiplication is {}\\\".format(sum))\\n return sum\",\n \"def mat_mul(mat1, mat2):\\n\\n rows1 = len(mat1)\\n cols1 = len(mat1[0])\\n rows2 = len(mat2)\\n cols2 = len(mat2[0])\\n\\n if cols1 != rows2:\\n return None\\n else:\\n new_matrix = []\\n for x in range(rows1):\\n aux_row = []\\n for y in range(cols2):\\n aux_sum = []\\n for z in range(cols1):\\n aux_sum.append(mat1[x][z] * mat2[z][y])\\n aux_row.append(sum(aux_sum))\\n new_matrix.append(aux_row)\\n\\n return new_matrix\",\n \"def mul(Z,X,Y):\",\n \"def mul(self,mat1,mat2):\\n if(isinstance(mat2,int)==True):\\n result = [[mat1[i][j] * mat2 for j in range(len(mat1[0]))] for i in range(len(mat1))]\\n self.out = result\\n return self.out\\n elif(len(mat1[0])==len(mat2)):\\n result = [[sum(a*b for a,b in zip(i,j)) for j in zip(*mat2)] for i in mat1]\\n self.out = result\\n return self.out\",\n \"def __matmul__(self, other):\\n return F.MatMul.apply(self, other)\",\n \"def __mul__(self,m):\\n if type(m) != Matrix:\\n raise TypeError('The second argument is not a matrix lol')\\n if self.ncols != m.nrows:\\n raise ValueError('matrix dot argument has incorrect number of rows')\\n new = Matrix(self.nrows,m.ncols)\\n columns = m.getCols()\\n rowindex = 0\\n colindex = 0 \\n for row in self.matrix:\\n colindex = 0 \\n for col in columns:\\n summ = 0\\n for i,j in zip(row,col):\\n summ+= i*j \\n new.matrix[rowindex][colindex] = summ\\n print new.matrix\\n colindex += 1 \\n rowindex+=1\\n return new\",\n \"def __mul__(self, other):\\n # \\n # TODO - your code here\\n #\\n \\n result = [];\\n row_result = [];\\n product = 0;\\n \\n if(self.w != other.h):\\n raise(ValueError, \\\"Matrices can not multiply for their dimesion doesn't match\\\"); \\n \\n for row in self.g:\\n row_result = [];\\n for j in range(other.w):\\n product = dot_product(row,other.get_column(j));\\n row_result.append(product);\\n result.append(row_result);\\n \\n return Matrix(result);\",\n \"def __mul__(self, othertr):\\n res = self.dot(othertr)\\n return res\",\n \"def nonsquare_matrix_mult_opt2(matrix):\\n\\n #Setup openCL\\n dev, ctx, queue = setup_CL()\\n\\n #openCL Kernel\\n #Naive approach with local/private memory\\n #Naive approach reworked to use local memory and tiling\\n #Modified boundary condition tiling kernel in lecture\\n kernel_code = \\\"\\\"\\\"\\n #define MATRIX_ROW_SIZE {}\\n #define MATRIX_COL_SIZE {}\\n #define TILE_WIDTH {}\\n #define n {}\\n __kernel void func(__global float* a, __global float* b) {{\\n\\n __local float M[TILE_WIDTH][TILE_WIDTH];\\n __local float N[TILE_WIDTH][TILE_WIDTH];\\n\\n int bx = get_group_id(0); int by = get_group_id(1);\\n int tx = get_local_id(0); int ty = get_local_id(1);\\n int Row = by * get_local_size(1) + ty;\\n int Col = bx * get_local_size(0) + tx;\\n float Cvalue = 0;\\n\\n // Loop over the A and B tiles required to compute the C element\\n for (int t = 0; t < (n-1)/TILE_WIDTH + 1;++t) {{\\n\\n //Assign rows of input\\n if(t*TILE_WIDTH+tx < MATRIX_COL_SIZE && tx < MATRIX_COL_SIZE && (Row*MATRIX_COL_SIZE + t*TILE_WIDTH + tx)<MATRIX_COL_SIZE*MATRIX_ROW_SIZE) {{\\n M[ty][tx] = a[Row*MATRIX_COL_SIZE + t*TILE_WIDTH + tx];\\n }} else {{\\n M[ty][tx] = 0.0;\\n }}\\n\\n //Assign columns of transpose\\n if (t*TILE_WIDTH+ty < n && Col < MATRIX_ROW_SIZE) {{\\n N[ty][tx] = a[t*TILE_WIDTH + MATRIX_COL_SIZE*Col + ty];\\n }} else {{\\n N[ty][tx] = 0.0;\\n }}\\n\\n barrier(CLK_LOCAL_MEM_FENCE);\\n\\n //Sum tile\\n for (int i = 0; i < TILE_WIDTH; ++i) {{\\n Cvalue += M[ty][i] * N[i][tx];\\n }}\\n\\n barrier(CLK_LOCAL_MEM_FENCE);\\n\\n //Assign values to output\\n if(Row<MATRIX_ROW_SIZE && Col<MATRIX_ROW_SIZE) {{\\n b[Row*MATRIX_ROW_SIZE + Col] = Cvalue;\\n\\n }}\\n }}\\n }}\\n \\\"\\\"\\\"\\n\\n #Move data to device\\n matrix_float = matrix.astype(np.float32)\\n matrix_gpu = cl.array.to_device(queue, matrix_float)\\n transposeMult_gpu = cl.array.empty(queue, (matrix.shape[0], matrix.shape[0]), np.float32)\\n transposed_gpu = cl.array.empty(queue, (matrix.shape[1],matrix.shape[0]), np.float32)\\n\\n matrix_row_size = matrix.shape[0]\\n matrix_col_size = matrix.shape[1]\\n TILE_WIDTH = 2\\n\\n #Calculate workItems, workGroup size, workGroups for input\\n matrix_val_count = matrix_float.shape[0]*matrix_float.shape[1]\\n xWorkItems = int(int(matrix_row_size-1)/TILE_WIDTH)+1\\n yWorkItems = int(int(matrix_row_size-1)/TILE_WIDTH)+1\\n totalWorkItems = float(TILE_WIDTH*TILE_WIDTH)\\n groups = np.int(max(np.ceil(matrix_val_count / xWorkItems),1))\\n\\n # print(\\\"workItems: %s, matrix_val_count: %s, groups: %s\\\" % (totalWorkItems, matrix_val_count, groups))\\n\\n # update template with current runtime requirements\\n kernel = kernel_code.format(matrix_row_size, matrix_col_size, TILE_WIDTH, max(matrix_col_size, matrix_row_size))\\n\\n #Launch kernel and time it\\n #Set global ID, workItems, workGroups\\n prg = cl.Program(ctx, kernel).build()\\n start = time.time()\\n event = prg.func(queue, (xWorkItems*TILE_WIDTH,yWorkItems*TILE_WIDTH,1),(TILE_WIDTH,TILE_WIDTH,1), matrix_gpu.data, transposeMult_gpu.data)\\n runtime = time.time()-start\\n\\n #Save output\\n transposedMult = transposeMult_gpu.get()\\n transposed = transposed_gpu.get()\\n\\n # print('openCL_opt2 %d x %d transpose-mult time: %.2E' % (matrix.shape[0], matrix.shape[1], runtime))\\n # print('openCL_opt2_transposed==goldenTransposed: %s' % np.allclose(transposed, np.transpose(matrix)))\\n # print('openCL_opt2_mult==goldenMult: %s' % np.allclose(transposedMult, matrix.dot(np.transpose(matrix))))\\n if not(np.allclose(transposedMult, matrix.dot(np.transpose(matrix)))):\\n # print('Original Matrix:\\\\n %s' % matrix)\\n print('openCL_opt2 transposed val:\\\\n %s' % transposed)\\n print('golden transpose-mult:\\\\n %s' % matrix.dot(np.transpose(matrix)))\\n transposedMult[(transposedMult>0) & (transposedMult<1)] = -1\\n print('openCL_opt2 mult val:\\\\n %s' % transposedMult)\\n print('openCL_opt2 transpose-mult:\\\\n %s' % np.isclose(transposedMult,matrix.dot(np.transpose(matrix))))\\n # print('--------------------')\\n\\n return [transposedMult, runtime]\",\n \"def __mul__(self,that):\\n return self.__opExpand2(that, np.multiply)\",\n \"def matrix_mult(A,B):\\n\\n m = len(A)\\n p = len(B)\\n n = len(B[0])\\n AB = []\\n for i in range(m):\\n AB.append([])\\n for j in range(n):\\n total = 0\\n for k in range(p):\\n total += A[i][k] * B[k][j]\\n AB[i].append(total)\\n return AB\",\n \"def _multi_matmul_three(A, B, C, constant=False) -> Tensor:\\n a0, a1b0 = A.shape[-2:]\\n b1c0, c1 = C.shape[-2:]\\n cost1 = a0 * b1c0 * (a1b0 + c1)\\n cost2 = a1b0 * c1 * (a0 + b1c0)\\n\\n if cost1 < cost2:\\n return matmul(matmul(A, B, constant), C, constant)\\n else:\\n return matmul(A, matmul(B, C, constant), constant)\",\n \"def _generate_mult_process(X, mat, inits):\\n M = np.empty_like(X, dtype=float)\\n M[..., 0] = inits[X[..., 0]]\\n M[..., 1:] = mat[X[..., :-1], X[..., 1:]]\\n np.cumprod(M, axis=-1, out=M)\\n return M\",\n \"def matrixMul(a, b):\\n # Initializing Empty Matrix\\n c = [[0, 0], [0, 0]]\\n # 2x2 matrix multiplication. Essentially O(1)\\n for i in range(2):\\n for j in range(2):\\n for k in range(2):\\n c[i][j] = (c[i][j] + (a[i][k] * b[k][j]))\\n\\n # Returning the products\\n return c\",\n \"def __matmul__(self, other):\\n if isinstance(other, types.Vector):\\n return self.apply(target=other)\\n if isinstance(other, MatrixLieGroup):\\n return self.multiply(other=other)\\n else:\\n assert False, \\\"Invalid argument\\\"\",\n \"def contract(tensor):\\n temp = np.einsum('ikma, jlan', tensor, tensor)\\n M = np.zeros((tensor.shape[0]**2, tensor.shape[1]**2, tensor.shape[2], tensor.shape[3]))\\n for i,j,k,l,m,n in it.product(*[range(x) for x in temp.shape]):\\n M[i + tensor.shape[0]*j, k + tensor.shape[1]*l, m, n] = temp[i,j,k,l,m,n]\\n return M\",\n \"def matrixMultiplication(firstMatrix, secondMatrix):\\n if len(firstMatrix[0]) == len(secondMatrix): # Checks whether the matrices can be multiplied or not or not\\n finalMatrix = []\\n for y in range(len(firstMatrix)): # 2\\n currentMatrix = []\\n for i in range(len(secondMatrix[0])):\\n currentSum = 0\\n for j in range(len(secondMatrix)):\\n currentSum += secondMatrix[j][i] * firstMatrix[y][j]\\n currentMatrix.append(currentSum)\\n print(\\\"This is my current matrix: \\\" + str(currentMatrix))\\n finalMatrix.append(currentMatrix)\\n print(\\\"This product of the two matrices is :) \\\" + str(finalMatrix))\\n else:\\n print(\\\"This operation cannot be done, make sure the rows of the first matrix is the same as the number of columns in the second matrix\\\")\",\n \"def __matmul__(self, B):\\n m, n = self.shape\\n n_, r = B.shape\\n assert n == n_, (\\\"Cannot multiply shapes \\\"\\n \\\"({}, {}) and ({}, {})\\\".format(m, n, n_, r))\\n mul_ = dict()\\n # compute A_ik = sum_j A_ij*B_jk\\n for i in range(m):\\n for k in range(r):\\n prod = mpfr(0)\\n for j in range(n):\\n prod += self[i, j] * B[j, k]\\n mul_[i, k] = prod\\n return MPMatrix((m, r), mul_)\",\n \"def matmul(x, y, _pub):\\n if x.shape[-1] != y.shape[-2]:\\n pass # TODO: REPORT ERROR\\n res = paillier_gpu.matmul_impl(x.flatten(), y.flatten(order='F'), x.shape, y.shape)\\n\\n return res\",\n \"def test_matmul_mm(self):\\n self.check_dot_mm(matmul_usecase, None, \\\"'@'\\\")\",\n \"def matrix_mult(A,B,mod=10**9+7):\\n C = [[1,1],[1,1]]\\n C[0][0] = ((A[0][0]*B[0][0])%mod + (A[0][1]*B[1][0])%mod)%mod\\n C[0][1] = ((A[0][0]*B[0][1])%mod + (A[0][1]*B[1][1])%mod)%mod\\n C[1][0] = ((A[1][0]*B[0][0])%mod + (A[1][1]*B[1][0])%mod)%mod\\n C[1][1] = ((A[1][0]*B[0][1])%mod + (A[1][1]*B[1][1])%mod)%mod\\n return C\",\n \"def product_on_basis(self, t1, t2):\\n return tensor( (module.monomial(x1)*module.monomial(x2) for (module, x1, x2) in zip(self._sets, t1, t2)) ) #.\",\n \"def __mul__(self, other):\\n return Matrix3(\\n self.i * other,\\n self.j * other,\\n self.k * other,\\n )\",\n \"def matmul(A, B):\\n # type: (Optional[Tensor], Tensor) -> Tensor\\n if A is None:\\n return B\\n if is_sparse(A):\\n return torch.sparse.mm(A, B)\\n return torch.matmul(A, B)\",\n \"def __imatmul__(self, tensor):\\n # Note: Matching PyTorch convention, which is not in-place here.\\n return self.matmul(tensor)\",\n \"def unfolded_product(m1, m2, coords = None, mx = None):\\n if coords is None:\\n mx = np.max( np.array([np.max(np.abs(m1.coords), axis = 0), \\n np.max(np.abs(m2.coords), axis = 0)]), axis = 0)\\n\\n coords = lattice_coords(mx)\\n else:\\n if mx is not None:\\n coords = lattice_coords(mx)\\n #print(coords, coords[int(len(coords)/2)])\\n M1 = m1.tofull(m1,coords,coords)\\n M2 = m2.tofull(m2,coords,coords)\\n \\n RET = np.dot(M1, M2)\\n rnx = m1.blocks.shape[1]\\n rny = m2.blocks.shape[2]\\n rc = int(len(coords)/2)\\n retblocks = RET[rnx*rc:rnx*(rc+1),:].reshape(rnx, coords.shape[0], rny).swapaxes(0,1)\\n \\n ret = tmat()\\n ret.load_nparray(retblocks, coords)\\n return ret\",\n \"def bc_matrix(params):\\r\\n w = params['w']\\r\\n kx = params['kx']\\r\\n d_list = params['d_list']\\r\\n ex_list = params['ex_list']\\r\\n ez_list = params['ez_list']\\r\\n kz_list = params['kz_list']\\r\\n N = len(d_list)\\r\\n assert N == len(d_list) == len(ex_list) == len(ez_list) == len(kz_list)\\r\\n assert N >= 2\\r\\n assert d_list[0] == d_list[-1] == inf\\r\\n \\r\\n # delta = e^{i * kz * d}, i.e. phase change across each layer\\r\\n # delta[0] and delta[-1] are undefined and are not used.\\r\\n delta_list = [cmath.exp(1j * kz_list[i] * d_list[i]) for i in range(N)]\\r\\n \\r\\n Ex_up_over_H_up_list = [kz_list[i] / (w * ex_list[i] * nu.eps0)\\r\\n for i in range(N)]\\r\\n Ex_down_over_H_down_list = [-a for a in Ex_up_over_H_up_list]\\r\\n Ez_up_over_H_up_list = [-kx / (w * ez_list[i] * nu.eps0) for i in range(N)]\\r\\n Ez_down_over_H_down_list = Ez_up_over_H_up_list[:]\\r\\n \\r\\n mat = np.zeros((2*N-2, 2*N-2), dtype=complex)\\r\\n \\r\\n for row_now in range(N-1):\\r\\n # This row concerns continuity of Ex across the boundary between\\r\\n # layer_under and layer_over (under and over the boundary respectively)\\r\\n layer_under = row_now\\r\\n layer_over = layer_under + 1\\r\\n # up_under_index is the column index in mat that gets multiplied by\\r\\n # H_{up} in layer_under.\\r\\n up_under_index = 2 * layer_under - 1\\r\\n down_under_index = 2 * layer_under\\r\\n up_over_index = 2 * layer_over - 1\\r\\n down_over_index = 2 * layer_over\\r\\n \\r\\n if layer_under != 0:\\r\\n assert 0 <= up_under_index < 2*N-2\\r\\n mat[row_now, up_under_index] = (\\r\\n Ex_up_over_H_up_list[layer_under] * delta_list[layer_under])\\r\\n mat[row_now, down_under_index] = Ex_down_over_H_down_list[layer_under]\\r\\n mat[row_now, up_over_index] = -Ex_up_over_H_up_list[layer_over]\\r\\n if layer_over != N-1:\\r\\n assert 0 <= down_over_index < 2*N-2\\r\\n mat[row_now, down_over_index] = (\\r\\n -Ex_down_over_H_down_list[layer_over] * delta_list[layer_over])\\r\\n\\r\\n for row_now in range(N-1, 2*N-2):\\r\\n # This row concerns continuity of eps_z * Ez across the boundary between\\r\\n # layer_under and layer_over (under and over the boundary respectively)\\r\\n layer_under = row_now - (N-1)\\r\\n layer_over = layer_under + 1\\r\\n # up_under_index is the column index in mat that gets multiplied by\\r\\n # H_{up} in layer_under.\\r\\n up_under_index = 2 * layer_under - 1\\r\\n down_under_index = 2 * layer_under\\r\\n up_over_index = 2 * layer_over - 1\\r\\n down_over_index = 2 * layer_over\\r\\n \\r\\n if layer_under != 0:\\r\\n assert 0 <= up_under_index < 2*N-2\\r\\n mat[row_now, up_under_index] = (ez_list[layer_under] *\\r\\n Ez_up_over_H_up_list[layer_under] * delta_list[layer_under])\\r\\n mat[row_now, down_under_index] = (ez_list[layer_under] *\\r\\n Ez_down_over_H_down_list[layer_under])\\r\\n mat[row_now, up_over_index] = (-ez_list[layer_over] * \\r\\n Ez_up_over_H_up_list[layer_over])\\r\\n if layer_over != N-1:\\r\\n assert 0 <= down_over_index < 2*N-2\\r\\n mat[row_now, down_over_index] = (-ez_list[layer_over] *\\r\\n Ez_down_over_H_down_list[layer_over] * delta_list[layer_over])\\r\\n \\r\\n return mat\",\n \"def build_M(K, R, T):\\n external = np.append(R, T, axis=1)\\n M = K.dot(external)\\n return M\",\n \"def combine_powers(mul):\\n factor, args = mul.as_coeff_matrices()\\n new_args = [args[0]]\\n\\n for i in range(1, len(args)):\\n A = new_args[-1]\\n B = args[i]\\n\\n if isinstance(B, Inverse) and isinstance(B.arg, MatMul):\\n Bargs = B.arg.args\\n l = len(Bargs)\\n if list(Bargs) == new_args[-l:]:\\n new_args = new_args[:-l] + [Identity(B.shape[0])]\\n continue\\n\\n if isinstance(A, Inverse) and isinstance(A.arg, MatMul):\\n Aargs = A.arg.args\\n l = len(Aargs)\\n if list(Aargs) == args[i:i+l]:\\n identity = Identity(A.shape[0])\\n new_args[-1] = identity\\n for j in range(i, i+l):\\n args[j] = identity\\n continue\\n\\n if A.is_square == False or B.is_square == False:\\n new_args.append(B)\\n continue\\n\\n if isinstance(A, MatPow):\\n A_base, A_exp = A.args\\n else:\\n A_base, A_exp = A, S.One\\n\\n if isinstance(B, MatPow):\\n B_base, B_exp = B.args\\n else:\\n B_base, B_exp = B, S.One\\n\\n if A_base == B_base:\\n new_exp = A_exp + B_exp\\n new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)\\n continue\\n elif not isinstance(B_base, MatrixBase):\\n try:\\n B_base_inv = B_base.inverse()\\n except NonInvertibleMatrixError:\\n B_base_inv = None\\n if B_base_inv is not None and A_base == B_base_inv:\\n new_exp = A_exp - B_exp\\n new_args[-1] = MatPow(A_base, new_exp).doit(deep=False)\\n continue\\n new_args.append(B)\\n\\n return newmul(factor, *new_args)\",\n \"def matrix_mult(m1, m2):\\n\\ttemp = []\\n\\tfor i in range(len(m1)):\\n\\t\\te = []\\n\\t\\tfor j in range(len(m2[0])):\\n\\t\\t\\te.append(row_times_column(m1,i,m2,j))\\n\\t\\ttemp.append(e)\\n\\treturn temp\",\n \"def matmul(a, b):\\n raise NotImplementedError\",\n \"def mul(self):\\n a = self.pop()\\n b = self.pop()\\n c= a*b\\n self.push(c)\",\n \"def applyMat(my_map, linsys_setup):\\n \\n datamaps, ninvs, beams, freqs, power_2d, precond_2d, clumaps, g_nu, \\\\\\n map_prop = linsys_setup\\n \\n \\n\\n nx, ny, pixScaleX, pixScaleY = map_prop\\n nFreq = len(g_nu); nCluster = len(clumaps[0])\\n\\n #Always apply beam * precond\\n beam_prec=[]\\n\\n for f in range(nFreq):\\n beam_prec+=[beams[f][:,:ny/2+1]*precond_2d[:,:ny/2+1]]\\n precond_2d=precond_2d[:,:ny/2+1]\\n power_2d=power_2d[:,:ny/2+1]\\n \\n ksz = False\\n if len(clumaps) == 2: ksz = True\\n \\n # Routines to perform block matrix multiplication defined in Eriksen Eq. 19\\n \\n def apply_cmb_cmb(d0):\\n \\\"\\\"\\\"\\n Apply (S^-1 + A N^-1 A) x\\n \\\"\\\"\\\"\\n d1 = d0.copy()\\n d1 = numpy.reshape(d1,(nx,ny))\\n a_l = fft.rfft(d1,axes=[-2,-1])\\n \\n c_l = 0\\n for f in range(nFreq):\\n\\n b_l = a_l * beam_prec[f]\\n d2 = fft.irfft(b_l,axes=[-2,-1],normalize=True)\\n d2 *= ninvs[f]\\n b_l = fft.rfft(d2,axes=[-2,-1])\\n c_l += b_l * beam_prec[f]\\n \\n d2 = fft.irfft(c_l,axes=[-2,-1],normalize=True)\\n d1 = fft.irfft(precond_2d**2 * a_l/power_2d,axes=[-2,-1],normalize=True)\\n \\n d2 += d1\\n \\n return d2.reshape((nx*ny,))\\n \\n \\\"\\\"\\\"\\n def apply_tsz_tsz(d0): # DONE\\n \\\\\\\"\\\"\\\"\\n Apply (F^T A^T N^-1 A F) x\\n \\\\\\\"\\\"\\\"\\n d1 = numpy.zeros(nCluster)\\n mat = numpy.zeros((nCluster, nCluster))\\n # TODO: This could probably be more efficient (e.g. using np.outer)\\n for freq in range(nFreq):\\n for ic in range(nCluster):\\n for jc in range(0, ic+1):\\n mat[ic,jc] = numpy.sum( ninvs[freq] * g_nu[freq]**2. \\\\\\n * clumaps[0][ic][freq] * clumaps[0][jc][freq] )\\n if ic != jc: mat[jc,ic] = mat[ic,jc]\\n d1 += numpy.dot(mat, d0)\\n return d1\\n \\n def apply_ksz_ksz(d0): # DONE\\n \\\\\\\"\\\"\\\"\\n Apply (K^T A^T N^-1 A K) x\\n \\\\\\\"\\\"\\\"\\n # FIXME: Missing factor of ivcov\\n d1 = numpy.zeros(nCluster)\\n mat = numpy.zeros((nCluster, nCluster))\\n # TODO: This could probably be more efficient (e.g. using np.outer)\\n for freq in range(nFreq):\\n for ic in range(nCluster):\\n for jc in range(0, ic+1):\\n mat[ic,jc] = numpy.sum( ninvs[freq] \\\\\\n * clumaps[1][ic][freq] * clumaps[1][jc][freq] )\\n if ic != jc: mat[jc,ic] = mat[ic,jc]\\n d1 += numpy.dot(mat, d0)\\n d1 += numpy.dot(ivcov, d0) # Add prior term\\n return d1\\n \\\"\\\"\\\"\\n \\n def apply_cmb_foreground_block(dc, dm, dt, dk=None):\\n \\\"\\\"\\\"\\n Apply the CMB x (Monopole + TSZ + KSZ) terms in one block:\\n (A^T N^-1 A T) x_mono\\n (A^T N^-1 A F) x_tsz\\n (A^T N^-1 A K) x_ksz\\n \\n (T^T A^T N^-1 A) x_cmb\\n (F^T A^T N^-1 A) x_cmb\\n (K^T A^T N^-1 A) x_cmb\\n \\\"\\\"\\\"\\n ksz = False\\n if dk is not None: ksz = True\\n \\n # (A^T N^-1 A T) x_mono; (A^T N^-1 A F) x_tsz; (A^T N^-1 A K) x_ksz\\n b_lt = 0; b_lk = 0; b_lm = 0\\n for f in range(nFreq):\\n mct = 0; mck = 0\\n for ic in range(nCluster):\\n mct += dt[ic] * ninvs[f] * clumaps[0][ic][f] * g_nu[f]\\n if ksz: mck += dk[ic] * ninvs[f] * clumaps[1][ic][f]\\n \\n b_lm += fft.rfft(dm * ninvs[f],axes=[-2,-1]) * beam_prec[f]\\n b_lt += fft.rfft(mct,axes=[-2,-1]) * beam_prec[f]\\n if ksz: b_lk += fft.rfft(mck,axes=[-2,-1]) * beam_prec[f]\\n\\n mcm = fft.irfft(b_lm,axes=[-2,-1],normalize=True).reshape((nx*ny,))\\n mct = fft.irfft(b_lt,axes=[-2,-1],normalize=True).reshape((nx*ny,))\\n if ksz: mck = fft.irfft(b_lk,axes=[-2,-1],normalize=True).reshape((nx*ny,))\\n \\n # (T^T A^T N^-1 A) x_cmb; (F^T A^T N^-1 A) x_cmb; (K^T A^T N^-1 A) x_cmb\\n mc = dc.copy().reshape((nx,ny))\\n a_l = fft.rfft(mc,axes=[-2,-1])\\n mtc = numpy.zeros(nCluster)\\n mkc = numpy.zeros(nCluster)\\n mmc = 0\\n for f in range(nFreq):\\n b_l = a_l * beam_prec[f]\\n mc = fft.irfft(b_l,axes=[-2,-1],normalize=True)\\n mmc += numpy.sum(mc * ninvs[f])\\n for ic in range(nCluster):\\n mtc[ic] += numpy.sum(mc * ninvs[f] * clumaps[0][ic][f] * g_nu[f])\\n if ksz: mkc[ic] += numpy.sum(mc * ninvs[f] * clumaps[1][ic][f])\\n \\n if ksz: return mct, mcm, mck, mtc, mmc, mkc\\n return mct, mcm, mtc, mmc\\n \\n \\n def apply_foreground_block(m0, t0, k0=None):\\n \\\"\\\"\\\"\\n Apply the TSZ + KSZ + Monopole terms in one block:\\n [ (T^T A^T N^-1 A F) (T^T A^T N^-1 A K) (T^T A^T N^-1 A T) ] (x_mono)\\n [ (F^T A^T N^-1 A F) (F^T A^T N^-1 A K) (F^T A^T N^-1 A T) ] (x_tsz)\\n [ (K^T A^T N^-1 A F) (K^T A^T N^-1 A K) (K^T A^T N^-1 A T) ] (x_ksz)\\n \\\"\\\"\\\"\\n ksz = True if k0 is not None else False\\n \\n dtt, dkk, dtk, dkt = [numpy.zeros(nCluster) for i in range(4)]\\n mtt, mkk, mtk, mkt = [numpy.zeros((nCluster, nCluster)) for i in range(4)]\\n dmm, dmk, dmt = [0 for i in range(3)]\\n dkm, dtm = [numpy.zeros(nCluster) for i in range(2)]\\n \\n # TODO: This could probably be more efficient (e.g. using np.outer)\\n for f in range(nFreq):\\n dmm += numpy.sum(ninvs[f]) * m0\\n \\n # Loop through clusters\\n for ic in range(nCluster):\\n dmt += numpy.sum( ninvs[f] * g_nu[f] * clumaps[0][ic][f] * t0[ic] )\\n dtm[ic] += numpy.sum( ninvs[f] * g_nu[f] * clumaps[0][ic][f] * m0 )\\n if ksz:\\n dmk += numpy.sum( ninvs[f] * clumaps[1][ic][f] * k0[ic] )\\n dkm[ic] += numpy.sum( ninvs[f] * clumaps[1][ic][f] * m0 )\\n \\n for jc in range(0, ic+1):\\n mtt[ic,jc] = numpy.sum( ninvs[f] * g_nu[f]**2. \\\\\\n * clumaps[0][ic][f] * clumaps[0][jc][f] )\\n if ksz:\\n mkk[ic,jc] = numpy.sum( ninvs[f] \\\\\\n * clumaps[1][ic][f] * clumaps[1][jc][f] )\\n mtk[ic,jc] = numpy.sum( ninvs[f] * g_nu[f] \\\\\\n * clumaps[0][ic][f] * clumaps[1][jc][f] )\\n mkt[ic,jc] = numpy.sum( ninvs[f] * g_nu[f] \\\\\\n * clumaps[1][ic][f] * clumaps[0][jc][f] )\\n # Mirror indices\\n mtt[jc,ic] = mtt[ic,jc]\\n if ksz:\\n mkk[jc,ic] = mkk[ic,jc]\\n mtk[jc,ic] = mtk[ic,jc]\\n mkt[jc,ic] = mkt[ic,jc]\\n \\n # Add total contribs. for this band\\n dtt += numpy.dot(mtt, t0)\\n if ksz:\\n dkk += numpy.dot(mkk, k0)\\n dtk += numpy.dot(mtk, k0)\\n dkt += numpy.dot(mkt, t0)\\n \\n if ksz: return dtt, dkk, dmm, dtk, dkt, dmk, dkm, dtm, dmt\\n return dtt, dmm, dtm, dmt\\n \\n # Apply block matrix multiplications and return\\n # FIXME: What if KSZ not used?\\n x0 = my_map[:nx*ny]\\n x1 = my_map[nx*ny:nx*ny+1]\\n x2 = my_map[nx*ny+1:nx*ny+nCluster+1]\\n if ksz: x3 = my_map[nx*ny+nCluster+1:nx*ny+2*nCluster+1]\\n \\n # Multiply input vector in blocks\\n #t=time.time()\\n dcc = apply_cmb_cmb(x0)\\n #print 'CMB', time.time()-t\\n if ksz:\\n dct, dcm, dck, dtc, dmc, dkc = apply_cmb_foreground_block(x0, x1, x2, x3)\\n dtt, dkk, dmm, dtk, dkt, dmk, dkm, dtm, dmt = apply_foreground_block(x1, x2, x3)\\n x_new_0 = dcc + dct + dck + dcm\\n x_new_1= dmc + dmt + dmk + dmm\\n x_new_2 = dtc + dtt + dtk + dtm\\n x_new_3 = dkc + dkt + dkk + dkm\\n x_new = numpy.concatenate((x_new_0, x_new_1, x_new_2, x_new_3))\\n else:\\n #t=time.time()\\n dct, dcm, dtc, dmc = apply_cmb_foreground_block(x0, x1, x2)\\n #print 'CMB-F', time.time()-t\\n \\n #t=time.time()\\n dtt, dmm, dtm, dmt = apply_foreground_block(x1, x2)\\n #print 'F', time.time()-t\\n \\n x_new_0 = dcc + dct + dcm\\n x_new_1 = dmc + dmt + dmm\\n x_new_2 = dtc + dtt + dtm\\n x_new = numpy.concatenate((x_new_0, x_new_1, x_new_2))\\n\\n\\n#sys.exit()\\n return x_new\",\n \"def __matmul__(self, csys):\\n self._transform(csys)\\n return self\",\n \"def matrix_mult_matrix(matrix_a, matrix_b):\\n m = len(matrix_a)\\n n = len(matrix_b)\\n result = []\\n matrix_b_t = transpose_matrix(matrix_b)\\n for i in xrange(m):\\n row = []\\n\\tfor j in xrange(m):\\n row.append(dot_product(matrix_a[i], matrix_b_t[j]))\\n\\tresult.append(row)\\n return result\",\n \"def distribute_monom(mul):\\n args = mul.args\\n if len(args) == 2:\\n from .matadd import MatAdd\\n if args[0].is_MatAdd and args[1].is_Rational:\\n return MatAdd(*[MatMul(mat, args[1]).doit() for mat in args[0].args])\\n if args[1].is_MatAdd and args[0].is_Rational:\\n return MatAdd(*[MatMul(args[0], mat).doit() for mat in args[1].args])\\n return mul\",\n \"def power_matrix(A, k):\\n nrow = np.shape(A)[0]\\n A0 = np.identity(nrow) \\n for k in range(q):\\n A0 = np.dot(A0, A)\\n \\n return A0\",\n \"def test_matmul_adjoint(self):\\n x = tf.constant([[9, 4, 1, 3], [1, 3, 1, 6], [7, 2, 2, 8]],\\n dtype=tf.float32)\\n self.assertAllClose(\\n tf.linalg.matmul(self.linop, x, adjoint_a=True),\\n tf.linalg.matmul(self.matrix, x, adjoint_a=True))\\n self.assertAllClose(\\n tf.linalg.matmul(self.linop, x, adjoint_a=False, adjoint_b=True),\\n tf.linalg.matmul(self.matrix, x, adjoint_a=False, adjoint_b=True))\",\n \"def __matmul__(self, B):\\n if isinstance(B, dict):\\n X = self.metacopy()\\n\\n if B['leg'] not in X.indexes:\\n raise ValueError('Leg of singular values not an indexes '\\n 'of self')\\n\\n if B['symmetries'] != X.symmetries:\\n raise ValueError('Not same symmetries')\\n\\n x, y = X.coupling_id(B['leg'])\\n for k in self:\\n newshape = [1] * len(self[k].shape)\\n newshape[X.indexes.index(B['leg'])] = -1\\n X[k] = self[k] * B[k[x][y]].reshape(newshape)\\n\\n return X\\n\\n connections = self.connections(B)\\n if not connections:\\n raise ValueError(f'No connections found between {self} and {B}')\\n\\n return self.contract(B, (list(connections),) * 2).simplify()\",\n \"def kkMul(*args):\\n if (None in args):\\n return None\\n product = 1\\n for arg in args:\\n product *= arg\\n return product\",\n \"def matMul(a, b):\\n sa=matShape(a)\\n sb=matShape(b)\\n if sa[1]!=sb[0]: raise ValueError\\n ret=matZeros((sa[0],sb[1]))\\n for i in range(sa[0]):\\n for j in range(sb[1]):\\n val=0.0\\n for k in range(sa[1]):\\n val+=matGet(a,i,k)*matGet(b,k,j)\\n matSet(ret,i,j,val)\\n return ret\",\n \"def matmul(self, x, work_buffer):\\n\\n x = asarray(x)\\n space = work_buffer.flat\\n\\n if (x.ndim == 0):\\n ValueError(\\n \\\"matmul: Input operand 1 does not have enough dimensions \\\"\\n \\\"(has 0, gufunc core with signature (n?,k),(k,m?)->(n?,m?) \\\"\\n \\\"requires 1\\\"\\n )\\n\\n if x.ndim == 1 and self.ndim == 1:\\n # Dot product\\n if x.shape[0] == self._size:\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n out = matmul(self[self._begin:], x[:k]).view(ndarray)\\n out += matmul(self[:self._end], x[k:])\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n out = matmul(part, x).view(ndarray)\\n\\n return(out)\\n else:\\n raise ValueError(\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\\\"\\n \\\" (size {n} is different from {m})\\\".format(\\n n=self._size,\\n m=x.shape[0]\\n )\\n )\\n elif self.ndim == 1 and x.ndim > 1:\\n if self._size == x.shape[-2]:\\n out_shape = *x.shape[:-2], x.shape[-1]\\n out = empty(out_shape)\\n out2 = space[:reduce(operator.mul, out_shape)].reshape(\\n out_shape\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(self[self._begin:], x[..., :k, :], out)\\n out += matmul(self[:self._end], x[..., k:, :], out2)\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(part, x, out).view(ndarray)\\n\\n return(out)\\n else:\\n raise ValueError(\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\\\"\\n \\\" (size {n} is different from {m})\\\".format(\\n n=self.shape[-2],\\n m=x.shape[0]\\n )\\n )\\n elif self.ndim == 2:\\n if (self.shape[-1] == x.shape[-2]):\\n out = empty(\\n (*x.shape[:-2], self.shape[-1], x.shape[-2])\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(self[self._begin:], x, out[..., :k, :])\\n matmul(self[:self._end], x, out[..., k:, :])\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(part, x, out)\\n\\n return(out.view(ndarray))\\n\\n else:\\n raise ValueError(\\n (\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->\\\"\\n \\\"(n?,m?) (size {n} is different from {m})\\\"\\n ).format(\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\\n else:\\n if (self.shape[-1] == x.shape[-2]):\\n self_shape = (self._size, *self.shape[1:-2])\\n\\n starexpr = tuple(\\n zip_longest(self_shape, x.shape[:-2], fillvalue=1)\\n )\\n if star_can_broadcast(starexpr):\\n broadcast_shape = tuple(\\n starmap(\\n lambda a, b: max(a, b),\\n starexpr\\n )\\n )\\n\\n out = empty(\\n (*broadcast_shape, self.shape[-2], x.shape[-1])\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n if x.ndim > 2:\\n matmul(self[self._begin:], x[:k], out[:k])\\n matmul(self[:self._end], x[k:], out[k:])\\n else:\\n matmul(self[self._begin:], x, out[:k])\\n matmul(self[:self._end], x, out[k:])\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(part, x, out)\\n\\n return(out.view(ndarray))\\n else:\\n raise ValueError(\\n (\\n \\\"operands could not be broadcast together with\\\"\\n \\\"remapped shapes [original->remapped]: \\\"\\n \\\"{shape_b}->({shape_bn}, newaxis,newaxis) \\\"\\n \\\"{shape_a}->({shape_an}, newaxis,newaxis) \\\"\\n \\\"and requested shape ({n},{m})\\\"\\n ).format(\\n shape_a=self_shape,\\n shape_b=x.shape,\\n shape_an=self.shape[:-2].__str__()[:-1],\\n shape_bn=x.shape[:-2].__str__()[:-1],\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\\n else:\\n raise ValueError(\\n (\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->\\\"\\n \\\"(n?,m?) (size {n} is different from {m})\\\"\\n ).format(\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\",\n \"def __mul__(self, other):\\n if self.n != other.m:\\n raise TypeError(\\\"Illegal dimensions for mul operator\\\")\\n tmp = [[0 for _ in xrange(self.n)] for _ in xrange(other.m)]\\n for i in xrange(self.n):\\n for j in xrange(other.m):\\n for k in xrange(other.n):\\n tmp[i][j] += self.values[i][k] * other.values[k][j]\\n res = []\\n for i in tmp:\\n res += i\\n return simplematrix(self.n, other.m, res)\",\n \"def naive_multiply(a, b):\\n m = len(a) # Number of rows in first matrix\\n k = len(b) # Number of rows in the second matrix\\n res = []\\n p = len(b[0])\\n n = k\\n for q in range(m):\\n res.append([0])\\n for q in range(m):\\n for w in range(p - 1):\\n res[q].append(0)\\n for i in range(m):\\n for j in range(p):\\n for r in range(n):\\n res[i][j] = a[i][r] * b[r][j] + res[i][j]\\n return res\",\n \"def combine_one_matrices(mul):\\n factor, args = mul.as_coeff_matrices()\\n new_args = [args[0]]\\n\\n for B in args[1:]:\\n A = new_args[-1]\\n if not isinstance(A, OneMatrix) or not isinstance(B, OneMatrix):\\n new_args.append(B)\\n continue\\n new_args.pop()\\n new_args.append(OneMatrix(A.shape[0], B.shape[1]))\\n factor *= A.shape[1]\\n\\n return newmul(factor, *new_args)\",\n \"def nonsquare_matrix_mult_opt1(matrix):\\n\\n #Setup openCL\\n dev, ctx, queue = setup_CL()\\n\\n #openCL Kernel\\n #Naive approach with local/private memory\\n kernel_code = \\\"\\\"\\\"\\n #define MATRIX_ROW_SIZE %(matrix_row_size)s\\n #define MATRIX_COL_SIZE %(matrix_col_size)s\\n\\n __kernel void func(__global float* a, __global float* b, __global float* transposed) {\\n\\n unsigned int i = get_local_id(0);\\n __local float tmp[MATRIX_ROW_SIZE*MATRIX_COL_SIZE];\\n\\n //Initialize tmp to 0\\n //Initialize output b to 0 for this thread\\n for(int k=0; k<MATRIX_COL_SIZE*MATRIX_ROW_SIZE; k++){\\n tmp[k] = 0;\\n }\\n\\n for(int k=0; k<MATRIX_ROW_SIZE; k++){\\n b[k + MATRIX_ROW_SIZE*get_group_id(0)] = 0;\\n }\\n\\n float localMatrix[MATRIX_ROW_SIZE*MATRIX_COL_SIZE];\\n //Copy matrix to local\\n for(int j=0; j < MATRIX_COL_SIZE; j++){\\n localMatrix[i+get_local_size(0)*get_group_id(0)]=a[i+get_local_size(0)*get_group_id(0)];\\n }\\n\\n\\n //Transpose output\\n transposed[i*MATRIX_ROW_SIZE+get_group_id(0)]=localMatrix[i+get_local_size(0)*get_group_id(0)];\\n\\n for(int j=0; j < MATRIX_ROW_SIZE; j++){\\n tmp[j+MATRIX_ROW_SIZE*i] = localMatrix[i+get_local_size(0)*get_group_id(0)]*a[i+j*MATRIX_COL_SIZE];\\n }\\n\\n // Store to output\\n for(int j=0; j < MATRIX_ROW_SIZE; j++){\\n for(int k=0; k < MATRIX_COL_SIZE; k++){\\n if(i==0){\\n b[j + MATRIX_ROW_SIZE*get_group_id(0)] += tmp[j+MATRIX_ROW_SIZE*k];\\n }\\n }\\n }\\n barrier(CLK_LOCAL_MEM_FENCE);\\n }\\n \\\"\\\"\\\"\\n\\n #Move data to device\\n matrix_float = matrix.astype(np.float32)\\n matrix_gpu = cl.array.to_device(queue, matrix_float)\\n transposeMult_gpu = cl.array.empty(queue, (matrix.shape[0], matrix.shape[0]), np.float32)\\n transposed_gpu = cl.array.empty(queue, (matrix.shape[1],matrix.shape[0]), np.float32)\\n\\n matrix_row_size = np.int32(matrix.shape[0])\\n matrix_col_size = np.int32(matrix.shape[1])\\n\\n #Calculate workItems, workGroup size, workGroups for input\\n matrix_val_count = matrix_float.shape[0]*matrix_float.shape[1]\\n xWorkItems = min(int(matrix_row_size),1024)\\n yWorkItems = min(int(matrix_col_size),1024)\\n totalWorkItems = float(xWorkItems*yWorkItems)\\n groups = np.int(max(np.ceil(matrix_val_count / xWorkItems),1))\\n\\n # print(\\\"workItems: %s, matrix_val_count: %s, groups: %s\\\" % (totalWorkItems, matrix_val_count, groups))\\n\\n # update template with current runtime requirements\\n kernel = kernel_code % {\\n 'matrix_row_size': matrix_row_size,\\n 'matrix_col_size': matrix_col_size\\n }\\n\\n #Launch kernel and time it\\n #Set global ID, workItems, workGroups\\n prg = cl.Program(ctx, kernel).build()\\n start = time.time()\\n event = prg.func(queue, (xWorkItems*yWorkItems,1),(groups,1), matrix_gpu.data, transposeMult_gpu.data, transposed_gpu.data)\\n\\n #event.wait()\\n runtime = time.time()-start\\n\\n #Save output\\n transposedMult = transposeMult_gpu.get()\\n transposed = transposed_gpu.get()\\n\\n # print('openCL_opt1 %d x %d transpose-mult time: %.2E' % (matrix.shape[0], matrix.shape[1], runtime))\\n # print('openCL_opt1_transposed==goldenTransposed: %s' % np.allclose(transposed, np.transpose(matrix)))\\n # print('openCL_opt1_mult==goldenMult: %s' % np.allclose(transposedMult, matrix.dot(np.transpose(matrix))))\\n if not(np.allclose(transposedMult, matrix.dot(np.transpose(matrix)))):\\n # print('Original Matrix:\\\\n %s' % matrix)\\n print('openCL_opt1 transposed val:\\\\n %s' % transposed)\\n print('golden transpose-mult:\\\\n %s' % matrix.dot(np.transpose(matrix)))\\n transposedMult[(transposedMult>0) & (transposedMult<1)] = -1\\n print('openCL_opt1 mult val:\\\\n %s' % transposedMult)\\n print('openCL_opt1 transpose-mult:\\\\n %s' % np.isclose(transposedMult,matrix.dot(np.transpose(matrix))))\\n # print('--------------------')\\n\\n return [transposedMult, runtime]\",\n \"def prod_mat(self,other):\\n [rs,cs],[ro,co] = self.D,other.D\\n assert cs == ro, \\\"tailles incompatibles\\\"\\n return Mat([rs,co], lambda i,j : prod_scal(self.ligne(i),other.col(j)))\",\n \"def norm2(x):\\n if iscomplexobj(x):\\n# t1=time.time()\\n# mat1=x.real**2 + x.imag**2\\n# t2=time.time()\\n mat2=multiply(x.real,x.real) + multiply(x.imag,x.imag)\\n# t3=time.time()\\n# print('---------------------------')\\n# print('pow time='+str(t2-t1))\\n# print('multiply time='+str(t3-t2))\\n# print('pow time/multiply time='+str((t2-t1)/(t3-t2)))\\n# print('shape(x)='+str(shape(x)))\\n# print('x.typecode='+str(x.typecode()))\\n# print('mat1.typecode='+str(mat1.typecode()))\\n# print('mat2.typecode='+str(mat2.typecode()))\\n# if len(shape(x))==1:\\n# print('type(x[0])='+str(type(x[0])))\\n# print('type(mat1[0])='+str(type(mat1[0])))\\n# print('type(mat2[0])='+str(type(mat2[0])))\\n# else:\\n# print('type(x[0,0])='+str(type(x[0,0])))\\n# print('type(mat1[0,0])='+str(type(mat1[0,0])))\\n# print('type(mat2[0,0])='+str(type(mat2[0,0])))\\n return mat2\\n else:\\n return multiply(x,x)\",\n \"def __matmul__(self, qubit):\\n if isinstance(qubit, str):\\n qubit = self.get_index(qubit)\\n return self.compiled[qubit].y\",\n \"def rmatmul(self, x, work_buffer) -> ndarray:\\n x = asarray(x)\\n space = asarray(work_buffer).flat\\n\\n if (x.ndim == 0):\\n ValueError(\\n \\\"matmul: Input operand 1 does not have enough dimensions \\\"\\n \\\"(has 0, gufunc core with signature (n?,k),(k,m?)->(n?,m?) \\\"\\n \\\"requires 1\\\"\\n )\\n\\n if x.ndim == 1 and self.ndim == 1:\\n # Dot product\\n if x.shape[0] == self._size:\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n out = matmul(x[:k], self[self._begin:]).view(ndarray)\\n out += matmul(x[k:], self[:self._end]).view(ndarray)\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n out = matmul(x, part)\\n\\n return(out.view(ndarray))\\n else:\\n raise ValueError(\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\\\"\\n \\\" (size {n} is different from {m})\\\".format(\\n n=self._size,\\n m=x.shape[0]\\n )\\n )\\n elif x.ndim == 1 and self.ndim > 1:\\n if x.shape[0] == self.shape[-2]:\\n if self.ndim == 2:\\n out = empty(self.shape[-1])\\n out2 = space[:self.shape[-1]]\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(x[:k], self[self._begin:], out)\\n out += matmul(x[k:], self[:self._end], out2)\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(x, part, out)\\n\\n return(out)\\n else:\\n out = empty(\\n (self._size, *self.shape[1:-2], self.shape[-1])\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(x, self[self._begin:], out[:k])\\n matmul(x, self[:self._end], out[k:])\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(x, part, out)\\n\\n return(out)\\n else:\\n raise ValueError(\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\\\"\\n \\\" (size {n} is different from {m})\\\".format(\\n n=self.shape[-2],\\n m=x.shape[0]\\n )\\n )\\n elif x.ndim > 1 and self.ndim == 1:\\n if x.shape[-1] == self.shape[0]:\\n out = empty(x.shape[:-1])\\n out2 = space[:reduce(operator.mul, x.shape[:-1])].reshape(\\n x.shape[:-1]\\n )\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(x[..., :, :k], self[self._begin:], out)\\n out += matmul(x[..., :, k:], self[:self._end], out2)\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(x, part, out)\\n\\n return(out)\\n else:\\n raise ValueError(\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?)\\\"\\n \\\" (size {n} is different from {m})\\\".format(\\n n=self.shape[-2],\\n m=x.shape[0]\\n )\\n )\\n elif self.ndim == 2:\\n if (x.shape[-1] == self.shape[-2]):\\n out_shape = (*x.shape[:-1], self.shape[-2])\\n out = empty(out_shape)\\n out2 = space[:reduce(operator.mul, out_shape)].reshape(\\n out_shape\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n matmul(x[..., :, :k], self[self._begin:], out)\\n out += matmul(x[..., :, k:], self[:self._end], out2)\\n\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(x, part, out)\\n\\n return(out.view(ndarray))\\n\\n else:\\n raise ValueError(\\n (\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->\\\"\\n \\\"(n?,m?) (size {n} is different from {m})\\\"\\n ).format(\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\\n else:\\n if (x.shape[-1] == self.shape[-2]):\\n self_shape = (self._size, *self.shape[1:-2])\\n\\n starexpr = tuple(\\n zip_longest(self_shape, x.shape[:-2], fillvalue=1)\\n )\\n if star_can_broadcast(starexpr):\\n broadcast_shape = tuple(\\n starmap(\\n lambda a, b: max(a, b),\\n starexpr\\n )\\n )\\n\\n out = empty(\\n (*broadcast_shape, x.shape[-2], self.shape[-1])\\n )\\n\\n if self.fragmented:\\n k = self._capacity - self._begin # fragmentation index\\n\\n if x.ndim > 2:\\n matmul(x[:k], self[self._begin:], out[:k])\\n matmul(x[k:], self[:self._end], out[k:])\\n else:\\n matmul(x, self[self._begin:], out[:k])\\n matmul(x, self[:self._end], out[k:])\\n else:\\n if self._begin < self._end:\\n part = self[self._begin:self._end]\\n elif self._end == 0:\\n part = self[self._begin:]\\n\\n matmul(x, part, out)\\n\\n return(out.view(ndarray))\\n else:\\n raise ValueError(\\n (\\n \\\"operands could not be broadcast together with\\\"\\n \\\"remapped shapes [original->remapped]: \\\"\\n \\\"{shape_b}->({shape_bn}, newaxis,newaxis) \\\"\\n \\\"{shape_a}->({shape_an}, newaxis,newaxis) \\\"\\n \\\"and requested shape ({n},{m})\\\"\\n ).format(\\n shape_a=self_shape,\\n shape_b=x.shape,\\n shape_an=self.shape[:-2].__str__()[:-1],\\n shape_bn=x.shape[:-2].__str__()[:-1],\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\\n else:\\n raise ValueError(\\n (\\n \\\"matmul: Input operand 1 has a mismatch in its core \\\"\\n \\\"dimension 0, with gufunc signature (n?,k),(k,m?)->\\\"\\n \\\"(n?,m?) (size {n} is different from {m})\\\"\\n ).format(\\n n=self.shape[-1],\\n m=x.shape[-2]\\n )\\n )\",\n \"def matmul(x, y):\\n return np.matmul(x, y)\",\n \"def compute(self, node, input_vals):\\n mat_A = input_vals[0]\\n mat_B = input_vals[1]\\n if node.matmul_attr_trans_A:\\n mat_A = mat_A.T\\n if node.matmul_attr_trans_B:\\n mat_B = mat_B.T\\n return np.matmul(mat_A, mat_B)\",\n \"def __imul__(self, other):\\n\\t\\t#print(\\\"imul\\\")\\n\\t\\t# merge other branch\\n\\t\\tself.graph.update(other.graph)\\n\\t\\tself.bottoms.update(other.bottoms)\\n\\t\\tself.output_shape.update(other.output_shape)\\n\\t\\tlayer_name = \\\"mul_{}\\\".format(len(self.graph))\\n\\t\\tself.graph[layer_name] = layer_name\\n\\t\\tself.bottoms[layer_name] = [self.cur_id, other.cur_id]\\n\\t\\tself.output_shape[layer_name] = self.cur_tensor.size()\\n\\t\\tself.cur_id = layer_name\\n\\t\\t# save memory\\n\\t\\tdel other\\n\\t\\treturn self\",\n \"def __mul__(self, A):\\n pass\",\n \"def test_gemm_opt0():\\r\\n X, Y, Z, a, b = XYZab()\\r\\n\\r\\n just_gemm([X, Y, Z, a, b], [T.dot(X, Y) * a + Z * b])\\r\\n just_gemm([X, Y, Z, a, b], [a * T.dot(X, Y) + b * Z])\\r\\n just_gemm([X, Y, Z, a, b], [b * Z + a * T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [T.dot(X, Y) * a - Z * b])\\r\\n just_gemm([X, Y, Z, a, b], [a * T.dot(X, Y) - b * Z])\\r\\n just_gemm([X, Y, Z, a, b], [b * Z - a * T.dot(X, Y)])\\r\\n\\r\\n #with transposes (transposes should be pushed through dot in canonicalize)\\r\\n just_gemm([X, Y, Z, a, b], [b * Z.T - a * T.dot(Y.T, X.T)])\\r\\n just_gemm([X, Y, Z, a, b], [b * Z.T + a * b * T.dot(X, Y).T])\\r\\n just_gemm([X, Y, Z, a, b], [b * Z + a * T.dot(X, Y).T],\\r\\n ishapes=[(5, 3), (3, 4), (4, 5), (), ()])\\r\\n\\r\\n #with N multiplications instead of just one\\r\\n just_gemm([X, Y, Z, a, b], [(b * b) * Z * a + (a * a) * T.dot(X, Y) * b])\\r\\n just_gemm([X, Y, Z, a, b], [Z + T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [Z * b + T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [Z + a * b * a * T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [(b * b) * Z * a - (a * a) * T.dot(X, Y) * b])\\r\\n just_gemm([X, Y, Z, a, b], [Z - T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [Z * b - T.dot(X, Y)])\\r\\n just_gemm([X, Y, Z, a, b], [Z - a * b * a * T.dot(X, Y)])\",\n \"def sparse_matmul(A: SparseTensor, B: SparseTensor, out: torch.Tensor) -> torch.Tensor:\\n if A.nnz() == 0 or B.nnz() == 0:\\n return out\\n\\n if A.is_cuda:\\n return _sparse_matmul_cuda(A, B, out)\\n else:\\n return _sparse_matmul_cpu(A, B, out)\",\n \"def matmulreduce(arr, reverse=False):\\n if reverse:\\n items = range(len(arr) - 1, -1, -1)\\n else:\\n items = range(len(arr))\\n P = arr[items[0]]\\n for i in items[1:]:\\n P = logical_matmul(P, arr[i]) * 1\\n return P\",\n \"def matrix_multiply_mapper(m, element):\\n\\n name, i, j, value = element\\n\\n if name == \\\"A\\\":\\n for k in range(m):\\n yield ((i, k), (j, value))\\n \\n else:\\n for k in range(m):\\n yield((k, j), (i, value))\",\n \"def __mul__(self, other):\\n if isinstance(other, Vector):\\n # Matrix vector product\\n v = Vector(list())\\n for n in range(len(other.vectors)):\\n v += scale(other.vectors[n][n], self.vectors[n])\\n return v\\n elif isinstance(other, Matrix):\\n # Matrix matrix product\\n if self.n != other.m:\\n raise ValueError(\\\"Wrong fucking sizes, nøøb\\\")\\n\\n selfVectors = self.vectors\\n selfColVectors = self.transpose()\\n otherVectors = other.vectors\\n otherColVectors = other.transpose()\\n vectors = list()\\n for col in range(other.n):\\n cordinator = []\\n\\n for row in range(self.m):\\n coord = 0\\n\\n for k in range(other.m):\\n coord += (\\n selfVectors[row].coords[k]\\n * otherColVectors.vectors[col].coords[k]\\n )\\n\\n cordinator.append(coord)\\n\\n v = Vector(cordinator)\\n vectors.append(v)\\n matrix = Matrix(vectors)\\n matrix = matrix.transpose()\\n return matrix\\n elif isinstance(other, int) or isinstance(other, float): # Skalering af matrix\\n for i in range(len(self.vectors)):\\n self.vectors[i] *= other\\n else:\\n raise ValueError(\\n \\\"Can only multiply Matrix with Matrix, Vector, Integer or Float\\\"\\n )\",\n \"def tt_dense_matmul(tt_matrix_a, matrix_b, convert_to_tensors=False):\\n\\n ndims = tt_matrix_a.ndims\\n a_columns = tt_matrix_a.shape[1]\\n b_rows = matrix_b.shape[0]\\n if a_columns is not None and b_rows is not None:\\n if a_columns != b_rows:\\n raise ValueError('Arguments shapes should align got %d and %d instead.' % (tt_matrix_a.shape, matrix_b.shape))\\n\\n a_shape = tt_matrix_a.shape\\n a_raw_shape = tt_matrix_a.raw_shape\\n b_shape = matrix_b.shape\\n a_ranks = tt_matrix_a.ranks\\n\\n # If A is (i0, ..., id-1) x (j0, ..., jd-1) and B is (j0, ..., jd-1) x K,\\n # data is (K, j0, ..., jd-2) x jd-1 x 1\\n data = matrix_b.transpose(0, 1)\\n #Liancheng +contiguous\\n data = data.contiguous().view(-1, a_raw_shape[1][-1], 1)\\n\\n for core_idx in reversed(range(ndims)):\\n curr_core = tt_matrix_a.tt_cores[core_idx]\\n #print('curr_core device:',curr_core.device)\\n\\n #Liancheng\\n curr_core = curr_core.to(data.device)\\n\\n # On the k = core_idx iteration, after applying einsum the shape of data\\n # becomes ik x (ik-1..., id-1, K, j0, ..., jk-1) x rank_k\\n #print('curr_core device:',curr_core.device)\\n #print('data device:',data.device)\\n data = torch.einsum('aijb,rjb->ira', curr_core, data)\\n if core_idx > 0:\\n # After reshape the shape of data becomes\\n # (ik, ..., id-1, K, j0, ..., jk-2) x jk-1 x rank_k\\n new_data_shape = (-1, a_raw_shape[1][core_idx - 1], a_ranks[core_idx])\\n data = data.contiguous().view(new_data_shape)\\n\\n # At the end the shape of the data is (i0, ..., id-1) x K\\n return data.view(a_shape[0], b_shape[1])\",\n \"def Chain(A, B):\\n return _prodOperator(B, A)\",\n \"def __mul__(self, other):\\n if hasattr(other, 'as_homogenous_transformation'):\\n return basis(homogenous_transformation = self.as_homogenous_transformation() * other.as_homogenous_transformation())\\n elif hasattr(other, 'n'):\\n if other.n == (3,1):\\n b = matrix.col((other[0], other[1], other[2], 1))\\n elif other.n == (4,1):\\n b = other\\n else:\\n raise TypeError(b, \\\"Incompatible matrices\\\")\\n p = self.as_homogenous_transformation() * b\\n if other.n == (3,1):\\n return matrix.col(p[0:3])\\n else:\\n return p\\n else:\\n raise TypeError(b)\",\n \"def inner_product_batch(pattern_stack_one, pattern_num_one, pattern_stack_two, pattern_num_two):\\n\\n \\\"\\\"\\\"\\n Notice that the two stacks can be different. So we can not deduce the lower triangular pattern from the \\n other half.\\n \\\"\\\"\\\"\\n holder = np.zeros((pattern_num_one, pattern_num_two))\\n for l in range(pattern_num_one):\\n for m in range(pattern_num_two):\\n holder[l, m] = np.sum(np.multiply(pattern_stack_one[l], pattern_stack_two[m]))\\n\\n return holder\",\n \"def multiply(traj, result_list):\\n z=traj.x*traj.y\\n result_list[traj.v_idx] = z\",\n \"def complex_multiplication(c1,c2,cr):\\n cr[0] = c1[0]*c2[0] - c1[1]*c2[1]\\n cr[1] = c1[0]*c2[1] + c1[1]*c2[0]\\n return cr\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.72562563","0.66338646","0.65753275","0.6564074","0.65081275","0.64550495","0.644735","0.64060277","0.63821524","0.6356483","0.6325638","0.62876886","0.6280246","0.6212008","0.6144578","0.6132298","0.6067564","0.6057658","0.5975746","0.5975746","0.59674287","0.5912267","0.5890951","0.5888945","0.5880107","0.5878204","0.5866108","0.5822242","0.5801993","0.5793156","0.57852644","0.5772086","0.57630086","0.57591915","0.57460207","0.573553","0.5732338","0.57283354","0.57049","0.5689399","0.56611454","0.565712","0.5640302","0.56329966","0.56246424","0.5600505","0.5598132","0.55815077","0.55793417","0.5579124","0.5573368","0.55731297","0.5567568","0.55626273","0.5557004","0.5553699","0.55510336","0.55415654","0.5534582","0.5527443","0.55255866","0.5524619","0.55181885","0.5506518","0.5506002","0.5505821","0.5500965","0.54999834","0.54998493","0.54762787","0.5474103","0.54677033","0.5467628","0.5467399","0.544828","0.5442598","0.5435893","0.54341817","0.5431291","0.54310644","0.54274416","0.54176986","0.54120034","0.5409243","0.540409","0.54020643","0.53997606","0.53933364","0.5383516","0.5370474","0.5370216","0.53686154","0.5349183","0.53444606","0.53389335","0.5335912","0.5334275","0.5333455","0.5326514","0.53168464"],"string":"[\n \"0.72562563\",\n \"0.66338646\",\n \"0.65753275\",\n \"0.6564074\",\n \"0.65081275\",\n \"0.64550495\",\n \"0.644735\",\n \"0.64060277\",\n \"0.63821524\",\n \"0.6356483\",\n \"0.6325638\",\n \"0.62876886\",\n \"0.6280246\",\n \"0.6212008\",\n \"0.6144578\",\n \"0.6132298\",\n \"0.6067564\",\n \"0.6057658\",\n \"0.5975746\",\n \"0.5975746\",\n \"0.59674287\",\n \"0.5912267\",\n \"0.5890951\",\n \"0.5888945\",\n \"0.5880107\",\n \"0.5878204\",\n \"0.5866108\",\n \"0.5822242\",\n \"0.5801993\",\n \"0.5793156\",\n \"0.57852644\",\n \"0.5772086\",\n \"0.57630086\",\n \"0.57591915\",\n \"0.57460207\",\n \"0.573553\",\n \"0.5732338\",\n \"0.57283354\",\n \"0.57049\",\n \"0.5689399\",\n \"0.56611454\",\n \"0.565712\",\n \"0.5640302\",\n \"0.56329966\",\n \"0.56246424\",\n \"0.5600505\",\n \"0.5598132\",\n \"0.55815077\",\n \"0.55793417\",\n \"0.5579124\",\n \"0.5573368\",\n \"0.55731297\",\n \"0.5567568\",\n \"0.55626273\",\n \"0.5557004\",\n \"0.5553699\",\n \"0.55510336\",\n \"0.55415654\",\n \"0.5534582\",\n \"0.5527443\",\n \"0.55255866\",\n \"0.5524619\",\n \"0.55181885\",\n \"0.5506518\",\n \"0.5506002\",\n \"0.5505821\",\n \"0.5500965\",\n \"0.54999834\",\n \"0.54998493\",\n \"0.54762787\",\n \"0.5474103\",\n \"0.54677033\",\n \"0.5467628\",\n \"0.5467399\",\n \"0.544828\",\n \"0.5442598\",\n \"0.5435893\",\n \"0.54341817\",\n \"0.5431291\",\n \"0.54310644\",\n \"0.54274416\",\n \"0.54176986\",\n \"0.54120034\",\n \"0.5409243\",\n \"0.540409\",\n \"0.54020643\",\n \"0.53997606\",\n \"0.53933364\",\n \"0.5383516\",\n \"0.5370474\",\n \"0.5370216\",\n \"0.53686154\",\n \"0.5349183\",\n \"0.53444606\",\n \"0.53389335\",\n \"0.5335912\",\n \"0.5334275\",\n \"0.5333455\",\n \"0.5326514\",\n \"0.53168464\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.697833"},"document_rank":{"kind":"string","value":"1"}}},{"rowIdx":90897,"cells":{"query":{"kind":"string","value":"Pick GPU if available, else CPU"},"document":{"kind":"string","value":"def get_default_device():\n if torch.cuda.is_available():\n return torch.device('cuda')\n else:\n return torch.device('cpu')"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def create_gpu_device_if_present():\n d = dpctl.SyclDevice(\"gpu,cpu\")\n print(\"Selected \" + (\"GPU\" if d.is_gpu else \"CPU\") + \" device\")","def return_free_GPU():\r\n if torch.cuda.is_available():\r\n gpu_num = torch.cuda.device_count()\r\n device = torch.device('cuda:{}'.format(gpu_num-1))\r\n print('Using GPU:[{}]/[{}] for training...'.format(gpu_num-1,gpu_num-1))\r\n return device\r\n \r\n raise ValueError('GPU not available for training. Check CUDA env with function \"check_cuda_env\"')","def try_gpu(i=0): #@save\n if len(tf.config.experimental.list_physical_devices('GPU')) >= i + 1:\n return tf.device(f'/GPU:{i}')\n return tf.device('/CPU:0')","def setup_device(n_gpus: int) -> object:\n if n_gpus >= 1 and torch.cuda.is_available():\n LOG.info('\\n CUDA is available! using GPU...')\n return torch.device('cuda')\n else:\n LOG.info('\\n Using CPU...')\n return torch.device('cpu')","def set_device(gpu_arg):\n\n dev = 'cpu'\n if gpu_arg and torch.cuda.is_available():\n dev = 'cuda'\n elif gpu_arg:\n print('Not gpu found. Using cpu instead.') \n\n return torch.device(dev)","def GetGPU():\n return option['device_id']","def get_device(arguments):\n\n # Checks if the GPU is available to be used and sets the .\n if arguments.gpu and torch.cuda.is_available():\n return torch.device(f\"cuda:{torch.cuda.device_count() - 1}\")\n\n # Sets the device to CPU.\n else:\n return torch.device(\"cpu\")","def set_device(in_arg): \n \n return torch.device(\"cuda\" if torch.cuda.is_available() and in_arg.gpu == 1 else \"cpu\")","def cuda_if_gpu(T):\n\n return T.cuda() if use_cuda else T","def try_gpu(i=0):\n if torch.cuda.device_count() >= i + 1:\n return torch.device(f'cuda:{i}')\n return torch.device('cpu')","def try_gpu(i=0):\n if torch.cuda.device_count() >= i + 1:\n return torch.device(f'cuda:{i}')\n return torch.device('cpu')","def try_gpu(i=0):\n if torch.cuda.device_count() >= i + 1:\n return torch.device(f'cuda:{i}')\n return torch.device('cpu')","def try_gpu(x):\n global _GPUS_EXIST\n if _GPUS_EXIST:\n try:\n return x.cuda()\n except (AssertionError, RuntimeError):\n print('No GPUs detected. Sticking with CPUs.')\n _GPUS_EXIST = False\n return x","def finddevice():\n\n return next((device for device in [\"xpu\"] if hasattr(torch, device) and getattr(torch, device).is_available()), None)","def try_gpu():\n try:\n ctx = mx.gpu()\n _ = nd.array([0], ctx=ctx)\n except:\n ctx = mx.cpu()\n return ctx","def try_gpu():\n try:\n ctx = mx.gpu()\n _ = nd.array([0], ctx=ctx)\n except:\n ctx = mx.cpu()\n return ctx","def device():\n return torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')","def try_gpu(x):\n global _GPUS_EXIST\n\n if _GPUS_EXIST:\n try:\n return x.cuda()\n except (AssertionError, RuntimeError):\n # actually, GPUs don't exist\n print 'No GPUs detected. Sticking with CPUs.'\n _GPUS_EXIST = False\n return x\n else:\n return x","def try_gpu():\r\n try:\r\n ctx = mx.gpu()\r\n _ = nd.array([0], ctx=ctx)\r\n except mx.base.MXNetError:\r\n ctx = mx.cpu()\r\n return ctx","def detect_gpus():\n def worker(q):\n # `device_lib` will not release the memory it took,\n # so we run it in a sub-process.\n try:\n from tensorflow.python.client import device_lib\n\n if is_tensorflow_version_higher_or_equal('1.8.0'):\n config = tf.ConfigProto()\n config.gpu_options.allow_growth = True\n devices = list(device_lib.list_local_devices(config))\n else:\n devices = list(device_lib.list_local_devices())\n gpus = [\n (device.name, device)\n for device in devices\n if device.device_type == 'GPU'\n ]\n union_set = {i: i for i in range(len(gpus))}\n\n for i, (name, device) in enumerate(gpus):\n assert (device.name == '/device:GPU:{}'.format(i))\n for link in device.locality.links.link:\n if link.device_id != i:\n union_set[i] = union_set[link.device_id]\n\n for i in six.iterkeys(union_set):\n while union_set[i] != union_set[union_set[i]]:\n union_set[i] = union_set[union_set[i]]\n\n root_devices = sorted(set(union_set.values()))\n gpu_groups = [[] for _ in range(len(root_devices))]\n dev_to_group = {j: i for i, j in enumerate(root_devices)}\n for i, (name, device) in enumerate(gpus):\n gpu_groups[dev_to_group[union_set[i]]].append(name)\n\n q.put((1, gpu_groups))\n except Exception:\n q.put((0, traceback.format_exc()))\n\n q = mp.Queue()\n p = mp.Process(target=worker, args=(q,))\n\n try:\n p.start()\n result = q.get()\n if result[0] == 1:\n return result[1]\n else:\n raise RuntimeError(\n 'Failed to retrieve GPU information, the traceback of '\n 'sub-process is:\\n {}'.\n format('\\n '.join(result[1].split('\\n')))\n )\n finally:\n p.terminate()\n p.join()","def get_device():\n import torch\n\n if torch.cuda.is_available():\n return torch.device('cuda')\n return torch.device('cpu')","def setup_gpu(use_gpu: int, silent=None) -> None:\n if silent is None:\n local_msg = Printer()\n else:\n local_msg = Printer(no_print=silent, pretty=not silent)\n if use_gpu >= 0:\n local_msg.info(f\"Using GPU: {use_gpu}\")\n require_gpu(use_gpu)\n else:\n local_msg.info(\"Using CPU\")\n if gpu_is_available():\n local_msg.info(\"To switch to GPU 0, use the option: --gpu-id 0\")","def _create_device(self):\n return torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")","def create_gpu_device():\n d1 = dpctl.SyclDevice(\"gpu\")\n d2 = dpctl.select_gpu_device()\n assert d1 == d2\n print_device(d1)\n return d1","def get_available_device():\n if torch.cuda.is_available():\n free_mem, device_idx = 0.0, 0\n for d in range(torch.cuda.device_count()):\n mem = torch.cuda.get_device_properties(d).total_memory - torch.cuda.memory_allocated(d)\n if mem > free_mem:\n device_idx = d\n free_mem = mem\n return torch.device(f'cuda:{device_idx}')\n else:\n return torch.device('cpu')","def get_default_device():\n if torch.cuda.is_available():\n return torch.device('cuda:0')\n else:\n return torch.device('cpu')","def get_default_device():\n device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n return device","def __getGPUthread(self, interactive, force) :\n #first try to import Reikna\n try :\n import reikna as rk \n except ModuleNotFoundError :\n if force: raise\n self.logger.warningglobal(\"Reikna isn't installed. Please install with 'pip install reikna' to use GPU devices.\")\n return None\n #create an API\n #try :\n # api = rk.cluda.cuda_api()\n #except Exception :\n # logger.info('CUDA-based GPU API not available, will try to get one based on OpenCL instead.')\n # try :\n # api = rk.cluda.ocl_api()\n # except Exception :\n # logger.warningglobal('WARNING: Failed to create an OpenCL API, no GPU computation will be available!!')\n # return None\n try :\n api = rk.cluda.ocl_api()\n #return a thread from the API\n return api.Thread.create(interactive=interactive)\n except Exception :\n if force: raise\n self.logger.warningglobal('Failed to create an OpenCL API, no GPU computation will be available!!')\n return None","def ConvertGpuToVendorName(gpu):\n if not gpu:\n return 'No GPU'\n elif '8086' in gpu:\n return 'Intel'\n elif '10de' in gpu:\n return 'NVIDIA'\n elif '1002' in gpu:\n return 'AMD'\n return gpu","def try_all_gpus(): #@save\n num_gpus = len(tf.config.experimental.list_physical_devices('GPU'))\n devices = [tf.device(f'/GPU:{i}') for i in range(num_gpus)]\n return devices if devices else [tf.device('/CPU:0')]","def is_gpu_available() -> bool:\n return torch.cuda.is_available()","def maybe_cuda(t):\n if torch.cuda.is_available():\n return t\n return t","def get_device(args: dict) -> torch.device:\n\n if is_distributed(args):\n device = torch.device(\"cuda\", args.local_rank)\n else:\n if torch.cuda.is_available():\n device = torch.device(\"cuda\", 0)\n else:\n device = torch.device(\"cpu\")\n return device","def detect_available():\n global _CUDA_AVAILABLE\n if _CUDA_AVAILABLE is not None: return _CUDA_AVAILABLE\n _CUDA_AVAILABLE = shell.run('{} -c \"import torch;print(torch.cuda.is_available())\"'.format(sys.executable)).strip('\\n') == 'True'\n return _CUDA_AVAILABLE","def get_device(i=0):\n if torch.cuda.is_available():\n return torch.device(\"cuda:%d\" % i)\n else:\n return torch.device(\"cpu\")","def set_max_gpu(self):\r\n gpu_getter = GPUGetter()\r\n gpu = str(gpu_getter.get_free_gpu())\r\n\r\n if gpu:\r\n print(\"Using GPU: %s\" % gpu)\r\n os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\" # see issue #152\r\n os.environ['CUDA_VISIBLE_DEVICES'] = gpu\r\n\r\n if not gpu:\r\n print('No GPU detected')","def cpu_fallback(flag=True):\n gpu_phy_devices = tf.config.list_physical_devices(\"GPU\")\n cpu_phy_devices = tf.config.list_physical_devices(\"CPU\")\n\n general_warning_msg = (\n f\"Tensorflow has already been initialized, {inspect.currentframe().f_code.co_name}() needs \"\n f\"to be called before any Tensorflow operation, as a result this function will have no effect\"\n )\n\n if flag is True:\n try:\n tf.config.set_visible_devices([], \"GPU\")\n except RuntimeError:\n warnings.warn(general_warning_msg)\n elif flag is False:\n try:\n tf.config.set_visible_devices(gpu_phy_devices, \"GPU\")\n except RuntimeError:\n warnings.warn(general_warning_msg)\n else:\n raise ValueError(\"Unknown flag, can only be True of False!\")","def check_gpus(_: None, gpus_optional: bool = False) -> None:\n gpu_devices = tf.config.list_physical_devices(\"GPU\")\n if gpu_devices:\n logging.info(f\"Using GPU: {gpu_devices}\")\n elif gpus_optional:\n logging.warning(\"No GPUs found, defaulting to CPU.\")\n else:\n raise RuntimeError(\"No GPUs found.\")","def set_device(device, backend='autograd'):\n if backend == 'autograd':\n return None\n elif backend == 'pytorch':\n try:\n tc.cuda.set_device(device)\n except:\n pass","def device() -> str:\n import torch\n\n if torch.cuda.is_available() and torch.cuda.device_count() > 0:\n if hasattr(Config().trainer,\n 'parallelized') and Config().trainer.parallelized:\n device = 'cuda'\n else:\n device = 'cuda:' + str(\n random.randint(0,\n torch.cuda.device_count() - 1))\n else:\n device = 'cpu'\n\n return device","def prepare_device(n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n print(\"Warning: There\\'s no GPU available on this machine,\"\n \"training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n print(\"Warning: The number of GPU\\'s configured to use is {}, but only {} are available \"\n \"on this machine.\".format(n_gpu_use, n_gpu))\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n return device, list_ids","def device(self) -> torch.device:\n for param in self.parameters():\n return param.device\n return get_device(\"cpu\")","def get_free_gpu(self):\r\n output = subprocess.Popen('nvidia-smi -q -d Memory |grep -A4 GPU|grep Free', stdout=subprocess.PIPE,\r\n shell=True).communicate()[0]\r\n output = output.decode(\"ascii\")\r\n\r\n # assumes that it is on the popiah server and the last gpu is not used\r\n memory_available = [int(x.split()[2]) for x in output.split(\"\\n\")[:-2]]\r\n\r\n if memory_available:\r\n print(\"Setting GPU to use to PID {}\".format(np.argmax(memory_available)))\r\n return np.argmax(memory_available)\r\n\r\n if not memory_available:\r\n print('No GPU memory available')","def prepare_device(n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n print(\"Warning: There\\'s no GPU available on this machine,\"\n \"training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n print(f\"Warning: The number of GPU\\'s configured to use is {n_gpu_use}, but only {n_gpu} are \"\n \"available on this machine.\")\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n return device, list_ids","def prepare_device(n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n print(\"Warning: There\\'s no GPU available on this machine,\"\n \"training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n print(f\"Warning: The number of GPU\\'s configured to use is {n_gpu_use}, but only {n_gpu} are \"\n \"available on this machine.\")\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n return device, list_ids","def cpu(device_id=0):\n return Context('cpu', device_id)","def cpu(device_id=0):\n return Context('cpu', device_id)","def configure_gpu_tf():\n\n try:\n # locate available devices & set required environment variables\n available_device_ids = GPUtil.getFirstAvailable(order='first', maxLoad=0.7, maxMemory=0.7, attempts=1, interval=10)\n available_device_id = available_device_ids[0]\n os.environ['CUDA_DEVICE_ORDER'] = 'PCI_BUS_ID'\n os.environ['CUDA_VISIBLE_DEVICES'] = str(available_device_id)\n print(f\"\\n GPU Found! running on GPU:{available_device_id}\\n\")\n\n # set GPU configuration (use all GPU memory if device 0, else use <50% of memory)\n tf.debugging.set_log_device_placement(False)\n physical_gpu = tf.config.experimental.list_physical_devices('GPU')[0]\n\n if available_device_id == 0:\n tf.config.experimental.set_memory_growth(physical_gpu, True)\n else:\n tf.config.experimental.set_virtual_device_configuration(\n physical_gpu,\n [tf.config.experimental.VirtualDeviceConfiguration(memory_limit=4500)]\n )\n logical_gpus = tf.config.experimental.list_logical_devices('GPU')\n assert len(logical_gpus) == 1, \"error creating virtual GPU to fractionally use memory\"\n\n # if we can't find a GPU, or they are all busy, default to using CPU\n except RuntimeError:\n print(\"\\n No GPUs available... running on CPU\\n\")\n os.environ['CUDA_VISIBLE_DEVICES'] = '-1'","def gpu(device_id=0):\n return Context('gpu', device_id)","def gpu(device_id=0):\n return Context('gpu', device_id)","def SetGPU(id):\n global option\n option['device_id'] = id","def EnableCPU():\n global option\n option['device'] = 'CPU'","def model_device(model):\n # Source: https://discuss.pytorch.org/t/how-to-check-if-model-is-on-cuda/180\n try:\n return str(next(model.parameters()).device)\n except StopIteration:\n # Model has no parameters\n pass\n return 'cpu'","def get_computation_devices(\n preferred_gpu_list: Optional[List[int]],\n multi_gpu_flag: bool,\n) -> List[Device]:\n\n # use CPU when GPUs are not preferred or not available\n if (preferred_gpu_list is None) \\\n or (len(preferred_gpu_list) == 0) \\\n or (not torch.cuda.is_available()):\n return [Device('cpu'), ]\n\n # else GPUs are preferred and available\n # get all available GPU indexes\n _available_gpu_list: List[int]\n if getAvailable:\n # by default, use GPU utility package with load and memory usage\n # specification so that the 'available' GPUs are actually ready\n # for deep learning runs (https://github.com/anderskm/gputil)\n _available_gpu_list = getAvailable(\n limit=_MAX_NUM_GPUS,\n maxLoad=_MAX_GPU_LOAD,\n maxMemory=_MAX_GPU_MEM_USED,\n )\n else:\n # assume all GPUs are good to use without GPUtil package\n _available_gpu_list = list(range(torch.cuda.device_count()))\n _warning_msg = \\\n f'GPUtil (https://github.com/anderskm/gputil) not installed.' \\\n f'Assuming all GPUs ({_available_gpu_list}) are available ' \\\n f'and ready for training ... '\n _LOGGER.warning(_warning_msg)\n\n # get the overlap between the preferred and the available GPUs\n _gpus = \\\n [_g for _g in _available_gpu_list if _g in preferred_gpu_list]\n\n # use CPU if there is no preferred GPUs that are available\n if len(_gpus) == 0:\n return [Device('cpu'), ]\n\n # otherwise return one or all GPUs depending on the multi-GPU flag\n return [Device(f'cuda:{_g}') for _g in _gpus] \\\n if multi_gpu_flag else [Device(f'cuda:{_gpus[0]}'), ]","def get_cpu(self):\n pass","def get_device(model):\n\tif next(model.parameters()).is_cuda:\n\t\treturn 'cuda:{}'.format(torch.cuda.current_device())\n\telse:\n\t\treturn 'cpu'","def setup_device(self, conf: DictConfig) -> device:\n device = torch.device(conf.runner.device) if torch.cuda.is_available() else torch.device('cpu')\n\n return device","def _next_device(self):\n if self._num_gpus == 0:\n return ''\n dev = '/gpu:%d' % self._cur_gpu\n if self._num_gpus > 1:\n self._cur_gpu = (self._cur_gpu + 1) % (self._num_gpus-1)\n return dev","def setGPU(state):\n\n\timport tensorflow as tf\n\tfrom keras import backend as K\n\n\tcheckGPU()\n\n\tnum_cores = 1\n\tnum_CPU = 1\n\tnum_GPU = 0\n\tif state:\n\t\tnum_GPU = 1\n\n\tconfig = tf.ConfigProto(intra_op_parallelism_threads=num_cores,\\\n\t inter_op_parallelism_threads=num_cores, allow_soft_placement=True,\\\n\t device_count = {'CPU' : num_CPU, 'GPU' : num_GPU})\n\tsession = tf.Session(config=config)\n\tK.set_session(session)","def cpu(self) -> Optional[Any]:\n return pulumi.get(self, \"cpu\")","def cpu(self) -> Optional[Any]:\n return pulumi.get(self, \"cpu\")","def _get_available_gpus():\n #global _LOCAL_DEVICES\n if tfback._LOCAL_DEVICES is None:\n devices = tf.config.list_logical_devices()\n tfback._LOCAL_DEVICES = [x.name for x in devices]\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def _get_available_gpus():\n #global _LOCAL_DEVICES\n if tfback._LOCAL_DEVICES is None:\n devices = tf.config.list_logical_devices()\n tfback._LOCAL_DEVICES = [x.name for x in devices]\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def setup_device(gpuid=None):\n\n if gpuid is not None and not isinstance(gpuid, str):\n gpuid = str(gpuid)\n\n if gpuid is not None:\n nb_devices = len(gpuid.split(','))\n else:\n nb_devices = 1\n\n if gpuid is not None and (gpuid != '-1'):\n device = '/gpu:' + gpuid\n os.environ['CUDA_VISIBLE_DEVICES'] = gpuid\n\n # GPU memory configuration differs between TF 1 and 2\n if hasattr(tf, 'ConfigProto'):\n config = tf.ConfigProto()\n config.gpu_options.allow_growth = True\n config.allow_soft_placement = True\n tf.keras.backend.set_session(tf.Session(config=config))\n else:\n tf.config.set_soft_device_placement(True)\n for pd in tf.config.list_physical_devices('GPU'):\n tf.config.experimental.set_memory_growth(pd, True)\n else:\n device = '/cpu:0'\n os.environ['CUDA_VISIBLE_DEVICES'] = '-1'\n\n return device, nb_devices","def create_cpu():\n return CPU()","def set_device(device: Union[str, torch.device]) -> torch.device:\n err_msg = None\n if isinstance(device, torch.device):\n pass\n elif device == 'auto':\n cuda = torch.cuda.is_available()\n device = torch.device('cuda' if cuda else 'cpu')\n elif device == 'gpu':\n cuda = torch.cuda.is_available()\n if cuda:\n device = torch.device('cuda')\n else:\n err_msg = ('Device set to \"gpu\", but could not access '\n 'any CUDA-enabled GPU. Please make sure that '\n 'a GPU is available and CUDA is installed '\n 'on this machine.')\n elif device == 'cpu':\n device = torch.device('cpu')\n else:\n err_msg = f'Unknown device \"{device}\". Try \"auto\".'\n if err_msg is not None:\n logger = get_logger(__name__, verbose=0)\n logger.error(f'Unknown device \"{device}\". Try \"auto\".')\n import sys\n sys.exit(1)\n return device","def _get_device(self, n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n self.logger.warning(\"Warning: There\\'s no GPU available on this machine,\"\n \"training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n self.logger.warning(f\"Warning: The number of GPU\\'s configured to use is {n_gpu_use}, \"\n f\"but only {n_gpu} are available on this machine.\")\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n self.logger.info(f'Using device: {device}, {list_ids}')\n return device, list_ids","def get_gpu_info(**kwargs):\n # Set GPU info fields\n conn_gpu_count = None\n source_db_gpu_count = None\n source_db_gpu_mem = None\n source_db_gpu_driver_ver = \"\"\n source_db_gpu_name = \"\"\n if kwargs[\"no_gather_conn_gpu_info\"]:\n logging.debug(\n \"--no-gather-conn-gpu-info passed, \"\n + \"using blank values for source database GPU info fields \"\n + \"[run_gpu_count, run_gpu_mem_mb] \"\n )\n else:\n logging.debug(\n \"Gathering source database GPU info fields \"\n + \"[run_gpu_count, run_gpu_mem_mb] \"\n + \"using pymapd connection info. \"\n )\n conn_hardware_info = kwargs[\"con\"]._client.get_hardware_info(\n kwargs[\"con\"]._session\n )\n conn_gpu_count = conn_hardware_info.hardware_info[0].num_gpu_allocated\n if conn_gpu_count == 0 or conn_gpu_count is None:\n no_gather_nvml_gpu_info = True\n if conn_gpu_count == 0:\n logging.warning(\n \"0 GPUs detected from connection info, \"\n + \"using blank values for source database GPU info fields \"\n + \"If running against cpu-only server, make sure to set \"\n + \"--no-gather-nvml-gpu-info and --no-gather-conn-gpu-info.\"\n )\n else:\n no_gather_nvml_gpu_info = kwargs[\"no_gather_nvml_gpu_info\"]\n source_db_gpu_count = conn_gpu_count\n try:\n source_db_gpu_mem = int(\n conn_hardware_info.hardware_info[0].gpu_info[0].memory\n / 1000000\n )\n except IndexError:\n logging.error(\"GPU memory info not available from connection.\")\n if no_gather_nvml_gpu_info:\n logging.debug(\n \"--no-gather-nvml-gpu-info passed, \"\n + \"using blank values for source database GPU info fields \"\n + \"[gpu_driver_ver, run_gpu_name] \"\n )\n elif (\n kwargs[\"conn_machine_name\"] == \"localhost\"\n or kwargs[\"gather_nvml_gpu_info\"]\n ):\n logging.debug(\n \"Gathering source database GPU info fields \"\n + \"[gpu_driver_ver, run_gpu_name] \"\n + \"from local GPU using pynvml. \"\n )\n import pynvml\n\n pynvml.nvmlInit()\n source_db_gpu_driver_ver = pynvml.nvmlSystemGetDriverVersion().decode()\n for i in range(source_db_gpu_count):\n handle = pynvml.nvmlDeviceGetHandleByIndex(i)\n # Assume all cards are the same, overwrite name value\n source_db_gpu_name = pynvml.nvmlDeviceGetName(handle).decode()\n pynvml.nvmlShutdown()\n # If gpu_count argument passed in, override gathered value\n if kwargs[\"gpu_count\"]:\n source_db_gpu_count = kwargs[\"gpu_count\"]\n if kwargs[\"gpu_name\"]:\n source_db_gpu_name = kwargs[\"gpu_name\"]\n gpu_info = {\n \"conn_gpu_count\": conn_gpu_count,\n \"source_db_gpu_count\": source_db_gpu_count,\n \"source_db_gpu_mem\": source_db_gpu_mem,\n \"source_db_gpu_driver_ver\": source_db_gpu_driver_ver,\n \"source_db_gpu_name\": source_db_gpu_name,\n }\n return gpu_info","def setup_CL():\n\n #Set up openCL platform\n NAME = 'NVIDIA CUDA'\n platforms = cl.get_platforms()\n\n dev = None\n for p in platforms:\n #Easy switching for local vs remote machine\n if p.name == 'Apple':\n NAME = 'Apple'\n if p.name == NAME:\n dev = p.get_devices()\n\n # Command queue, enable GPU profiling\n ctx = cl.Context(dev)\n queue = cl.CommandQueue(ctx,properties=cl.command_queue_properties.PROFILING_ENABLE)\n\n return [dev,ctx,queue]","def _get_available_gpus():\r\n #global _LOCAL_DEVICES\r\n if tfback._LOCAL_DEVICES is None:\r\n devices = tf.config.list_logical_devices()\r\n tfback._LOCAL_DEVICES = [x.name for x in devices]\r\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def set_current_device_id():\n import paddle\n curr_dev = paddle.device.get_device()\n select_gpu = os.getenv(\"FLAGS_selected_gpus\", \"0\")\n paddle.set_flags({\n 'FLAGS_selected_gpus': os.getenv(\"FLAGS_selected_gpus\", \"0\")\n })\n if \"gpu\" in curr_dev and select_gpu != curr_dev.split(\":\")[-1]:\n paddle.set_device(\"gpu:\" + select_gpu)\n\n curr_dev_id = paddle.framework.core.get_cuda_current_device_id()\n if \"gpu\" in curr_dev and select_gpu != str(curr_dev_id):\n paddle.zeros([])","def cpu(self):\n self.reader.model.cpu()\n self.reader.device = torch.device(\"cpu\")\n return self","def get_free_gpu():\n\tos.system('nvidia-smi -q -d Memory |grep -A4 GPU|grep Free >tmp')\n\tif os.path.exists('tmp'):\n\t\tmemory_available = [int(x.split()[2]) for x in open('tmp', 'r').readlines()]\n\t\tos.remove('tmp')\n\t\treturn np.argmax(memory_available)\n\treturn 0","def check_cuda():\n if OS_VERSION[0] == \"Linux\":\n check_cuda_linux()\n elif OS_VERSION[0] == \"Windows\":\n check_cuda_windows()","def _get_available_gpus():\n # global _LOCAL_DEVICES\n if tfback._LOCAL_DEVICES is None:\n devices = tf.config.list_logical_devices()\n tfback._LOCAL_DEVICES = [x.name for x in devices]\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def min_gpu(self) -> Optional[pulumi.Input[int]]:\n return pulumi.get(self, \"min_gpu\")","def get_test_devices():\n\n # Assumption: CPU is always available\n devices = ['cpu']\n\n if torch.cuda.is_available():\n devices.append('cuda')\n\n return devices","def _get_available_gpus():\n # global _LOCAL_DEVICES\n if tf_back._LOCAL_DEVICES is None:\n devices = tf.config.list_logical_devices()\n tf_back._LOCAL_DEVICES = [x.name for x in devices]\n return [x for x in tf_back._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def device(request):\n d = request.param()\n\n # enable GPU error checking\n if isinstance(d, hoomd.device.GPU):\n d.gpu_error_checking = True\n\n return d","def OnGPU(gpu_id):\n device_option = caffe2_pb2.DeviceOption()\n device_option.device_type = workspace.GpuDeviceType\n device_option.device_id = gpu_id\n return device_option","def only_gpu(request):\n if request.node.get_closest_marker('gpu'):\n if 'device' in request.fixturenames:\n if not isinstance(request.getfixturevalue('device'),\n hoomd.device.GPU):\n pytest.skip('Test is run only on GPU(s).')\n else:\n raise ValueError('only_gpu requires the *device* fixture')","def maybe_cuda(x: Union[torch.Tensor, nn.Module]) -> Union[torch.Tensor, nn.Module]:\n if torch.cuda.is_available():\n return x.cuda()\n return x","def _get_available_gpus():\n #global _LOCAL_DEVICES\n if tfback._LOCAL_DEVICES is None:\n devices = tf.config.list_logical_devices()\n tfback._LOCAL_DEVICES = [x.name for x in devices]\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]","def is_gpu_available():\n ret = get_gpu_count() > 0\n if _HAS_PADDLE:\n import paddle\n if ret is True and not paddle.is_compiled_with_cuda():\n logger.warning(\"Found non-empty CUDA_VISIBLE_DEVICES. \\\n But PARL found that Paddle was not complied with CUDA, which may cause issues. \\\n Thus PARL will not use GPU.\")\n return False\n if _HAS_FLUID:\n from paddle import fluid\n if ret is True and not fluid.is_compiled_with_cuda():\n logger.warning(\"Found non-empty CUDA_VISIBLE_DEVICES. \\\n But PARL found that Paddle was not complied with CUDA, which may cause issues. \\\n Thus PARL will not use GPU.\")\n return False\n return ret","def get_default_device():\n global _default_device\n\n if _default_device is None:\n import wgpu.backends.rs # noqa\n\n adapter = wgpu.request_adapter(canvas=None, power_preference=\"high-performance\")\n _default_device = adapter.request_device()\n return _default_device","def _on_gpu(self) -> bool:\n return self._current_device_index != CPU_INDEX","def _prepare_device(self, n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n self.logger.warning(\n \"Warning: There\\'s no GPU available on this machine, training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n self.logger.warning(\n \"Warning: The number of GPU\\'s configured to use is {}, but only {} are available on this machine.\".format(\n n_gpu_use, n_gpu))\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n return device, list_ids","def use_cuda():\n return torch.cuda.is_available() and os.getenv('AICROWD_CUDA', True)","def gpu_availability():\n # assume if using tensorflow-gpu, then Nvidia GPU is available\n if is_built_with_cuda():\n return len(tf.config.list_physical_devices(\"GPU\")) > 0\n else:\n return False","def get_cpu_info():\n try:\n cpu_info = subprocess.check_output('lscpu')\n return cpu_info\n except OSError:\n return None","def test_gpu_cuda_code() -> None:\n if get_from_environ(\"DISABLE_GPU_FOR_TESTING\") is not None:\n print(\"GPU payload disabled for testing\")\n return\n\n # if the command exists it can run on the hardware below\n proc = subprocess.Popen([\"nvidia-smi\"], stdout=subprocess.PIPE)\n stdout, _ = proc.communicate()\n str_stdout = stdout.decode()\n assert \"NVIDIA-SMI\" in str_stdout, str_stdout\n assert proc.returncode == 0\n # search the history for the CUDA implementation","def get_test_devices():\n devices = [\"cpu\"]\n if torch.cuda.is_available():\n devices.append(\"cuda\")\n return devices","def _prepare_device(self, n_gpu_use):\n n_gpu = torch.cuda.device_count()\n if n_gpu_use > 0 and n_gpu == 0:\n self.logger.warning(\"Warning: There\\'s no GPU available on this machine,\"\n \"training will be performed on CPU.\")\n n_gpu_use = 0\n if n_gpu_use > n_gpu:\n self.logger.warning(\"Warning: The number of GPU\\'s configured to use is {}, but only {} are available \"\n \"on this machine.\".format(n_gpu_use, n_gpu))\n n_gpu_use = n_gpu\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\n list_ids = list(range(n_gpu_use))\n return device, list_ids","def with_cpu(ops, model):\n ...","def check_device(use_gpu, use_xpu=False, use_npu=False, use_mlu=False):\n err = \"Config {} cannot be set as true while your paddle \" \\\n \"is not compiled with {} ! \\nPlease try: \\n\" \\\n \"\\t1. Install paddlepaddle to run model on {} \\n\" \\\n \"\\t2. Set {} as false in config file to run \" \\\n \"model on CPU\"\n\n try:\n if use_gpu and use_xpu:\n print(\"use_xpu and use_gpu can not both be ture.\")\n if use_gpu and not paddle.is_compiled_with_cuda():\n print(err.format(\"use_gpu\", \"cuda\", \"gpu\", \"use_gpu\"))\n sys.exit(1)\n if use_xpu and not paddle.device.is_compiled_with_xpu():\n print(err.format(\"use_xpu\", \"xpu\", \"xpu\", \"use_xpu\"))\n sys.exit(1)\n if use_npu:\n if int(paddle.version.major) != 0 and int(\n paddle.version.major) <= 2 and int(\n paddle.version.minor) <= 4:\n if not paddle.device.is_compiled_with_npu():\n print(err.format(\"use_npu\", \"npu\", \"npu\", \"use_npu\"))\n sys.exit(1)\n # is_compiled_with_npu() has been updated after paddle-2.4\n else:\n if not paddle.device.is_compiled_with_custom_device(\"npu\"):\n print(err.format(\"use_npu\", \"npu\", \"npu\", \"use_npu\"))\n sys.exit(1)\n if use_mlu and not paddle.device.is_compiled_with_mlu():\n print(err.format(\"use_mlu\", \"mlu\", \"mlu\", \"use_mlu\"))\n sys.exit(1)\n except Exception as e:\n pass","def device_placement(self):\n if is_tf_available():\n import tensorflow as tf\n with tf.device('/CPU:0' if self.device == -1 else '/device:GPU:{}'.format(self.device)):\n yield\n else:\n import torch\n if self.device >= 0:\n torch.cuda.set_device(self.device)\n\n yield","def InitCL(GPU=0):\n platform, device, context, queue = None, None, None, None\n for iplatform in range(3):\n try:\n platform = cl.get_platforms()[iplatform]\n if (GPU>0):\n device = platform.get_devices(cl.device_type.GPU)\n else:\n device = platform.get_devices(cl.device_type.CPU)\n context = cl.Context(device)\n queue = cl.CommandQueue(context)\n break\n except:\n pass\n return platform, device, context, queue, cl.mem_flags"],"string":"[\n \"def create_gpu_device_if_present():\\n d = dpctl.SyclDevice(\\\"gpu,cpu\\\")\\n print(\\\"Selected \\\" + (\\\"GPU\\\" if d.is_gpu else \\\"CPU\\\") + \\\" device\\\")\",\n \"def return_free_GPU():\\r\\n if torch.cuda.is_available():\\r\\n gpu_num = torch.cuda.device_count()\\r\\n device = torch.device('cuda:{}'.format(gpu_num-1))\\r\\n print('Using GPU:[{}]/[{}] for training...'.format(gpu_num-1,gpu_num-1))\\r\\n return device\\r\\n \\r\\n raise ValueError('GPU not available for training. Check CUDA env with function \\\"check_cuda_env\\\"')\",\n \"def try_gpu(i=0): #@save\\n if len(tf.config.experimental.list_physical_devices('GPU')) >= i + 1:\\n return tf.device(f'/GPU:{i}')\\n return tf.device('/CPU:0')\",\n \"def setup_device(n_gpus: int) -> object:\\n if n_gpus >= 1 and torch.cuda.is_available():\\n LOG.info('\\\\n CUDA is available! using GPU...')\\n return torch.device('cuda')\\n else:\\n LOG.info('\\\\n Using CPU...')\\n return torch.device('cpu')\",\n \"def set_device(gpu_arg):\\n\\n dev = 'cpu'\\n if gpu_arg and torch.cuda.is_available():\\n dev = 'cuda'\\n elif gpu_arg:\\n print('Not gpu found. Using cpu instead.') \\n\\n return torch.device(dev)\",\n \"def GetGPU():\\n return option['device_id']\",\n \"def get_device(arguments):\\n\\n # Checks if the GPU is available to be used and sets the .\\n if arguments.gpu and torch.cuda.is_available():\\n return torch.device(f\\\"cuda:{torch.cuda.device_count() - 1}\\\")\\n\\n # Sets the device to CPU.\\n else:\\n return torch.device(\\\"cpu\\\")\",\n \"def set_device(in_arg): \\n \\n return torch.device(\\\"cuda\\\" if torch.cuda.is_available() and in_arg.gpu == 1 else \\\"cpu\\\")\",\n \"def cuda_if_gpu(T):\\n\\n return T.cuda() if use_cuda else T\",\n \"def try_gpu(i=0):\\n if torch.cuda.device_count() >= i + 1:\\n return torch.device(f'cuda:{i}')\\n return torch.device('cpu')\",\n \"def try_gpu(i=0):\\n if torch.cuda.device_count() >= i + 1:\\n return torch.device(f'cuda:{i}')\\n return torch.device('cpu')\",\n \"def try_gpu(i=0):\\n if torch.cuda.device_count() >= i + 1:\\n return torch.device(f'cuda:{i}')\\n return torch.device('cpu')\",\n \"def try_gpu(x):\\n global _GPUS_EXIST\\n if _GPUS_EXIST:\\n try:\\n return x.cuda()\\n except (AssertionError, RuntimeError):\\n print('No GPUs detected. Sticking with CPUs.')\\n _GPUS_EXIST = False\\n return x\",\n \"def finddevice():\\n\\n return next((device for device in [\\\"xpu\\\"] if hasattr(torch, device) and getattr(torch, device).is_available()), None)\",\n \"def try_gpu():\\n try:\\n ctx = mx.gpu()\\n _ = nd.array([0], ctx=ctx)\\n except:\\n ctx = mx.cpu()\\n return ctx\",\n \"def try_gpu():\\n try:\\n ctx = mx.gpu()\\n _ = nd.array([0], ctx=ctx)\\n except:\\n ctx = mx.cpu()\\n return ctx\",\n \"def device():\\n return torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')\",\n \"def try_gpu(x):\\n global _GPUS_EXIST\\n\\n if _GPUS_EXIST:\\n try:\\n return x.cuda()\\n except (AssertionError, RuntimeError):\\n # actually, GPUs don't exist\\n print 'No GPUs detected. Sticking with CPUs.'\\n _GPUS_EXIST = False\\n return x\\n else:\\n return x\",\n \"def try_gpu():\\r\\n try:\\r\\n ctx = mx.gpu()\\r\\n _ = nd.array([0], ctx=ctx)\\r\\n except mx.base.MXNetError:\\r\\n ctx = mx.cpu()\\r\\n return ctx\",\n \"def detect_gpus():\\n def worker(q):\\n # `device_lib` will not release the memory it took,\\n # so we run it in a sub-process.\\n try:\\n from tensorflow.python.client import device_lib\\n\\n if is_tensorflow_version_higher_or_equal('1.8.0'):\\n config = tf.ConfigProto()\\n config.gpu_options.allow_growth = True\\n devices = list(device_lib.list_local_devices(config))\\n else:\\n devices = list(device_lib.list_local_devices())\\n gpus = [\\n (device.name, device)\\n for device in devices\\n if device.device_type == 'GPU'\\n ]\\n union_set = {i: i for i in range(len(gpus))}\\n\\n for i, (name, device) in enumerate(gpus):\\n assert (device.name == '/device:GPU:{}'.format(i))\\n for link in device.locality.links.link:\\n if link.device_id != i:\\n union_set[i] = union_set[link.device_id]\\n\\n for i in six.iterkeys(union_set):\\n while union_set[i] != union_set[union_set[i]]:\\n union_set[i] = union_set[union_set[i]]\\n\\n root_devices = sorted(set(union_set.values()))\\n gpu_groups = [[] for _ in range(len(root_devices))]\\n dev_to_group = {j: i for i, j in enumerate(root_devices)}\\n for i, (name, device) in enumerate(gpus):\\n gpu_groups[dev_to_group[union_set[i]]].append(name)\\n\\n q.put((1, gpu_groups))\\n except Exception:\\n q.put((0, traceback.format_exc()))\\n\\n q = mp.Queue()\\n p = mp.Process(target=worker, args=(q,))\\n\\n try:\\n p.start()\\n result = q.get()\\n if result[0] == 1:\\n return result[1]\\n else:\\n raise RuntimeError(\\n 'Failed to retrieve GPU information, the traceback of '\\n 'sub-process is:\\\\n {}'.\\n format('\\\\n '.join(result[1].split('\\\\n')))\\n )\\n finally:\\n p.terminate()\\n p.join()\",\n \"def get_device():\\n import torch\\n\\n if torch.cuda.is_available():\\n return torch.device('cuda')\\n return torch.device('cpu')\",\n \"def setup_gpu(use_gpu: int, silent=None) -> None:\\n if silent is None:\\n local_msg = Printer()\\n else:\\n local_msg = Printer(no_print=silent, pretty=not silent)\\n if use_gpu >= 0:\\n local_msg.info(f\\\"Using GPU: {use_gpu}\\\")\\n require_gpu(use_gpu)\\n else:\\n local_msg.info(\\\"Using CPU\\\")\\n if gpu_is_available():\\n local_msg.info(\\\"To switch to GPU 0, use the option: --gpu-id 0\\\")\",\n \"def _create_device(self):\\n return torch.device(\\\"cuda\\\" if torch.cuda.is_available() else \\\"cpu\\\")\",\n \"def create_gpu_device():\\n d1 = dpctl.SyclDevice(\\\"gpu\\\")\\n d2 = dpctl.select_gpu_device()\\n assert d1 == d2\\n print_device(d1)\\n return d1\",\n \"def get_available_device():\\n if torch.cuda.is_available():\\n free_mem, device_idx = 0.0, 0\\n for d in range(torch.cuda.device_count()):\\n mem = torch.cuda.get_device_properties(d).total_memory - torch.cuda.memory_allocated(d)\\n if mem > free_mem:\\n device_idx = d\\n free_mem = mem\\n return torch.device(f'cuda:{device_idx}')\\n else:\\n return torch.device('cpu')\",\n \"def get_default_device():\\n if torch.cuda.is_available():\\n return torch.device('cuda:0')\\n else:\\n return torch.device('cpu')\",\n \"def get_default_device():\\n device = torch.device(\\\"cuda\\\" if torch.cuda.is_available() else \\\"cpu\\\")\\n return device\",\n \"def __getGPUthread(self, interactive, force) :\\n #first try to import Reikna\\n try :\\n import reikna as rk \\n except ModuleNotFoundError :\\n if force: raise\\n self.logger.warningglobal(\\\"Reikna isn't installed. Please install with 'pip install reikna' to use GPU devices.\\\")\\n return None\\n #create an API\\n #try :\\n # api = rk.cluda.cuda_api()\\n #except Exception :\\n # logger.info('CUDA-based GPU API not available, will try to get one based on OpenCL instead.')\\n # try :\\n # api = rk.cluda.ocl_api()\\n # except Exception :\\n # logger.warningglobal('WARNING: Failed to create an OpenCL API, no GPU computation will be available!!')\\n # return None\\n try :\\n api = rk.cluda.ocl_api()\\n #return a thread from the API\\n return api.Thread.create(interactive=interactive)\\n except Exception :\\n if force: raise\\n self.logger.warningglobal('Failed to create an OpenCL API, no GPU computation will be available!!')\\n return None\",\n \"def ConvertGpuToVendorName(gpu):\\n if not gpu:\\n return 'No GPU'\\n elif '8086' in gpu:\\n return 'Intel'\\n elif '10de' in gpu:\\n return 'NVIDIA'\\n elif '1002' in gpu:\\n return 'AMD'\\n return gpu\",\n \"def try_all_gpus(): #@save\\n num_gpus = len(tf.config.experimental.list_physical_devices('GPU'))\\n devices = [tf.device(f'/GPU:{i}') for i in range(num_gpus)]\\n return devices if devices else [tf.device('/CPU:0')]\",\n \"def is_gpu_available() -> bool:\\n return torch.cuda.is_available()\",\n \"def maybe_cuda(t):\\n if torch.cuda.is_available():\\n return t\\n return t\",\n \"def get_device(args: dict) -> torch.device:\\n\\n if is_distributed(args):\\n device = torch.device(\\\"cuda\\\", args.local_rank)\\n else:\\n if torch.cuda.is_available():\\n device = torch.device(\\\"cuda\\\", 0)\\n else:\\n device = torch.device(\\\"cpu\\\")\\n return device\",\n \"def detect_available():\\n global _CUDA_AVAILABLE\\n if _CUDA_AVAILABLE is not None: return _CUDA_AVAILABLE\\n _CUDA_AVAILABLE = shell.run('{} -c \\\"import torch;print(torch.cuda.is_available())\\\"'.format(sys.executable)).strip('\\\\n') == 'True'\\n return _CUDA_AVAILABLE\",\n \"def get_device(i=0):\\n if torch.cuda.is_available():\\n return torch.device(\\\"cuda:%d\\\" % i)\\n else:\\n return torch.device(\\\"cpu\\\")\",\n \"def set_max_gpu(self):\\r\\n gpu_getter = GPUGetter()\\r\\n gpu = str(gpu_getter.get_free_gpu())\\r\\n\\r\\n if gpu:\\r\\n print(\\\"Using GPU: %s\\\" % gpu)\\r\\n os.environ[\\\"CUDA_DEVICE_ORDER\\\"] = \\\"PCI_BUS_ID\\\" # see issue #152\\r\\n os.environ['CUDA_VISIBLE_DEVICES'] = gpu\\r\\n\\r\\n if not gpu:\\r\\n print('No GPU detected')\",\n \"def cpu_fallback(flag=True):\\n gpu_phy_devices = tf.config.list_physical_devices(\\\"GPU\\\")\\n cpu_phy_devices = tf.config.list_physical_devices(\\\"CPU\\\")\\n\\n general_warning_msg = (\\n f\\\"Tensorflow has already been initialized, {inspect.currentframe().f_code.co_name}() needs \\\"\\n f\\\"to be called before any Tensorflow operation, as a result this function will have no effect\\\"\\n )\\n\\n if flag is True:\\n try:\\n tf.config.set_visible_devices([], \\\"GPU\\\")\\n except RuntimeError:\\n warnings.warn(general_warning_msg)\\n elif flag is False:\\n try:\\n tf.config.set_visible_devices(gpu_phy_devices, \\\"GPU\\\")\\n except RuntimeError:\\n warnings.warn(general_warning_msg)\\n else:\\n raise ValueError(\\\"Unknown flag, can only be True of False!\\\")\",\n \"def check_gpus(_: None, gpus_optional: bool = False) -> None:\\n gpu_devices = tf.config.list_physical_devices(\\\"GPU\\\")\\n if gpu_devices:\\n logging.info(f\\\"Using GPU: {gpu_devices}\\\")\\n elif gpus_optional:\\n logging.warning(\\\"No GPUs found, defaulting to CPU.\\\")\\n else:\\n raise RuntimeError(\\\"No GPUs found.\\\")\",\n \"def set_device(device, backend='autograd'):\\n if backend == 'autograd':\\n return None\\n elif backend == 'pytorch':\\n try:\\n tc.cuda.set_device(device)\\n except:\\n pass\",\n \"def device() -> str:\\n import torch\\n\\n if torch.cuda.is_available() and torch.cuda.device_count() > 0:\\n if hasattr(Config().trainer,\\n 'parallelized') and Config().trainer.parallelized:\\n device = 'cuda'\\n else:\\n device = 'cuda:' + str(\\n random.randint(0,\\n torch.cuda.device_count() - 1))\\n else:\\n device = 'cpu'\\n\\n return device\",\n \"def prepare_device(n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n print(\\\"Warning: There\\\\'s no GPU available on this machine,\\\"\\n \\\"training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n print(\\\"Warning: The number of GPU\\\\'s configured to use is {}, but only {} are available \\\"\\n \\\"on this machine.\\\".format(n_gpu_use, n_gpu))\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n return device, list_ids\",\n \"def device(self) -> torch.device:\\n for param in self.parameters():\\n return param.device\\n return get_device(\\\"cpu\\\")\",\n \"def get_free_gpu(self):\\r\\n output = subprocess.Popen('nvidia-smi -q -d Memory |grep -A4 GPU|grep Free', stdout=subprocess.PIPE,\\r\\n shell=True).communicate()[0]\\r\\n output = output.decode(\\\"ascii\\\")\\r\\n\\r\\n # assumes that it is on the popiah server and the last gpu is not used\\r\\n memory_available = [int(x.split()[2]) for x in output.split(\\\"\\\\n\\\")[:-2]]\\r\\n\\r\\n if memory_available:\\r\\n print(\\\"Setting GPU to use to PID {}\\\".format(np.argmax(memory_available)))\\r\\n return np.argmax(memory_available)\\r\\n\\r\\n if not memory_available:\\r\\n print('No GPU memory available')\",\n \"def prepare_device(n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n print(\\\"Warning: There\\\\'s no GPU available on this machine,\\\"\\n \\\"training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n print(f\\\"Warning: The number of GPU\\\\'s configured to use is {n_gpu_use}, but only {n_gpu} are \\\"\\n \\\"available on this machine.\\\")\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n return device, list_ids\",\n \"def prepare_device(n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n print(\\\"Warning: There\\\\'s no GPU available on this machine,\\\"\\n \\\"training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n print(f\\\"Warning: The number of GPU\\\\'s configured to use is {n_gpu_use}, but only {n_gpu} are \\\"\\n \\\"available on this machine.\\\")\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n return device, list_ids\",\n \"def cpu(device_id=0):\\n return Context('cpu', device_id)\",\n \"def cpu(device_id=0):\\n return Context('cpu', device_id)\",\n \"def configure_gpu_tf():\\n\\n try:\\n # locate available devices & set required environment variables\\n available_device_ids = GPUtil.getFirstAvailable(order='first', maxLoad=0.7, maxMemory=0.7, attempts=1, interval=10)\\n available_device_id = available_device_ids[0]\\n os.environ['CUDA_DEVICE_ORDER'] = 'PCI_BUS_ID'\\n os.environ['CUDA_VISIBLE_DEVICES'] = str(available_device_id)\\n print(f\\\"\\\\n GPU Found! running on GPU:{available_device_id}\\\\n\\\")\\n\\n # set GPU configuration (use all GPU memory if device 0, else use <50% of memory)\\n tf.debugging.set_log_device_placement(False)\\n physical_gpu = tf.config.experimental.list_physical_devices('GPU')[0]\\n\\n if available_device_id == 0:\\n tf.config.experimental.set_memory_growth(physical_gpu, True)\\n else:\\n tf.config.experimental.set_virtual_device_configuration(\\n physical_gpu,\\n [tf.config.experimental.VirtualDeviceConfiguration(memory_limit=4500)]\\n )\\n logical_gpus = tf.config.experimental.list_logical_devices('GPU')\\n assert len(logical_gpus) == 1, \\\"error creating virtual GPU to fractionally use memory\\\"\\n\\n # if we can't find a GPU, or they are all busy, default to using CPU\\n except RuntimeError:\\n print(\\\"\\\\n No GPUs available... running on CPU\\\\n\\\")\\n os.environ['CUDA_VISIBLE_DEVICES'] = '-1'\",\n \"def gpu(device_id=0):\\n return Context('gpu', device_id)\",\n \"def gpu(device_id=0):\\n return Context('gpu', device_id)\",\n \"def SetGPU(id):\\n global option\\n option['device_id'] = id\",\n \"def EnableCPU():\\n global option\\n option['device'] = 'CPU'\",\n \"def model_device(model):\\n # Source: https://discuss.pytorch.org/t/how-to-check-if-model-is-on-cuda/180\\n try:\\n return str(next(model.parameters()).device)\\n except StopIteration:\\n # Model has no parameters\\n pass\\n return 'cpu'\",\n \"def get_computation_devices(\\n preferred_gpu_list: Optional[List[int]],\\n multi_gpu_flag: bool,\\n) -> List[Device]:\\n\\n # use CPU when GPUs are not preferred or not available\\n if (preferred_gpu_list is None) \\\\\\n or (len(preferred_gpu_list) == 0) \\\\\\n or (not torch.cuda.is_available()):\\n return [Device('cpu'), ]\\n\\n # else GPUs are preferred and available\\n # get all available GPU indexes\\n _available_gpu_list: List[int]\\n if getAvailable:\\n # by default, use GPU utility package with load and memory usage\\n # specification so that the 'available' GPUs are actually ready\\n # for deep learning runs (https://github.com/anderskm/gputil)\\n _available_gpu_list = getAvailable(\\n limit=_MAX_NUM_GPUS,\\n maxLoad=_MAX_GPU_LOAD,\\n maxMemory=_MAX_GPU_MEM_USED,\\n )\\n else:\\n # assume all GPUs are good to use without GPUtil package\\n _available_gpu_list = list(range(torch.cuda.device_count()))\\n _warning_msg = \\\\\\n f'GPUtil (https://github.com/anderskm/gputil) not installed.' \\\\\\n f'Assuming all GPUs ({_available_gpu_list}) are available ' \\\\\\n f'and ready for training ... '\\n _LOGGER.warning(_warning_msg)\\n\\n # get the overlap between the preferred and the available GPUs\\n _gpus = \\\\\\n [_g for _g in _available_gpu_list if _g in preferred_gpu_list]\\n\\n # use CPU if there is no preferred GPUs that are available\\n if len(_gpus) == 0:\\n return [Device('cpu'), ]\\n\\n # otherwise return one or all GPUs depending on the multi-GPU flag\\n return [Device(f'cuda:{_g}') for _g in _gpus] \\\\\\n if multi_gpu_flag else [Device(f'cuda:{_gpus[0]}'), ]\",\n \"def get_cpu(self):\\n pass\",\n \"def get_device(model):\\n\\tif next(model.parameters()).is_cuda:\\n\\t\\treturn 'cuda:{}'.format(torch.cuda.current_device())\\n\\telse:\\n\\t\\treturn 'cpu'\",\n \"def setup_device(self, conf: DictConfig) -> device:\\n device = torch.device(conf.runner.device) if torch.cuda.is_available() else torch.device('cpu')\\n\\n return device\",\n \"def _next_device(self):\\n if self._num_gpus == 0:\\n return ''\\n dev = '/gpu:%d' % self._cur_gpu\\n if self._num_gpus > 1:\\n self._cur_gpu = (self._cur_gpu + 1) % (self._num_gpus-1)\\n return dev\",\n \"def setGPU(state):\\n\\n\\timport tensorflow as tf\\n\\tfrom keras import backend as K\\n\\n\\tcheckGPU()\\n\\n\\tnum_cores = 1\\n\\tnum_CPU = 1\\n\\tnum_GPU = 0\\n\\tif state:\\n\\t\\tnum_GPU = 1\\n\\n\\tconfig = tf.ConfigProto(intra_op_parallelism_threads=num_cores,\\\\\\n\\t inter_op_parallelism_threads=num_cores, allow_soft_placement=True,\\\\\\n\\t device_count = {'CPU' : num_CPU, 'GPU' : num_GPU})\\n\\tsession = tf.Session(config=config)\\n\\tK.set_session(session)\",\n \"def cpu(self) -> Optional[Any]:\\n return pulumi.get(self, \\\"cpu\\\")\",\n \"def cpu(self) -> Optional[Any]:\\n return pulumi.get(self, \\\"cpu\\\")\",\n \"def _get_available_gpus():\\n #global _LOCAL_DEVICES\\n if tfback._LOCAL_DEVICES is None:\\n devices = tf.config.list_logical_devices()\\n tfback._LOCAL_DEVICES = [x.name for x in devices]\\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def _get_available_gpus():\\n #global _LOCAL_DEVICES\\n if tfback._LOCAL_DEVICES is None:\\n devices = tf.config.list_logical_devices()\\n tfback._LOCAL_DEVICES = [x.name for x in devices]\\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def setup_device(gpuid=None):\\n\\n if gpuid is not None and not isinstance(gpuid, str):\\n gpuid = str(gpuid)\\n\\n if gpuid is not None:\\n nb_devices = len(gpuid.split(','))\\n else:\\n nb_devices = 1\\n\\n if gpuid is not None and (gpuid != '-1'):\\n device = '/gpu:' + gpuid\\n os.environ['CUDA_VISIBLE_DEVICES'] = gpuid\\n\\n # GPU memory configuration differs between TF 1 and 2\\n if hasattr(tf, 'ConfigProto'):\\n config = tf.ConfigProto()\\n config.gpu_options.allow_growth = True\\n config.allow_soft_placement = True\\n tf.keras.backend.set_session(tf.Session(config=config))\\n else:\\n tf.config.set_soft_device_placement(True)\\n for pd in tf.config.list_physical_devices('GPU'):\\n tf.config.experimental.set_memory_growth(pd, True)\\n else:\\n device = '/cpu:0'\\n os.environ['CUDA_VISIBLE_DEVICES'] = '-1'\\n\\n return device, nb_devices\",\n \"def create_cpu():\\n return CPU()\",\n \"def set_device(device: Union[str, torch.device]) -> torch.device:\\n err_msg = None\\n if isinstance(device, torch.device):\\n pass\\n elif device == 'auto':\\n cuda = torch.cuda.is_available()\\n device = torch.device('cuda' if cuda else 'cpu')\\n elif device == 'gpu':\\n cuda = torch.cuda.is_available()\\n if cuda:\\n device = torch.device('cuda')\\n else:\\n err_msg = ('Device set to \\\"gpu\\\", but could not access '\\n 'any CUDA-enabled GPU. Please make sure that '\\n 'a GPU is available and CUDA is installed '\\n 'on this machine.')\\n elif device == 'cpu':\\n device = torch.device('cpu')\\n else:\\n err_msg = f'Unknown device \\\"{device}\\\". Try \\\"auto\\\".'\\n if err_msg is not None:\\n logger = get_logger(__name__, verbose=0)\\n logger.error(f'Unknown device \\\"{device}\\\". Try \\\"auto\\\".')\\n import sys\\n sys.exit(1)\\n return device\",\n \"def _get_device(self, n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n self.logger.warning(\\\"Warning: There\\\\'s no GPU available on this machine,\\\"\\n \\\"training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n self.logger.warning(f\\\"Warning: The number of GPU\\\\'s configured to use is {n_gpu_use}, \\\"\\n f\\\"but only {n_gpu} are available on this machine.\\\")\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n self.logger.info(f'Using device: {device}, {list_ids}')\\n return device, list_ids\",\n \"def get_gpu_info(**kwargs):\\n # Set GPU info fields\\n conn_gpu_count = None\\n source_db_gpu_count = None\\n source_db_gpu_mem = None\\n source_db_gpu_driver_ver = \\\"\\\"\\n source_db_gpu_name = \\\"\\\"\\n if kwargs[\\\"no_gather_conn_gpu_info\\\"]:\\n logging.debug(\\n \\\"--no-gather-conn-gpu-info passed, \\\"\\n + \\\"using blank values for source database GPU info fields \\\"\\n + \\\"[run_gpu_count, run_gpu_mem_mb] \\\"\\n )\\n else:\\n logging.debug(\\n \\\"Gathering source database GPU info fields \\\"\\n + \\\"[run_gpu_count, run_gpu_mem_mb] \\\"\\n + \\\"using pymapd connection info. \\\"\\n )\\n conn_hardware_info = kwargs[\\\"con\\\"]._client.get_hardware_info(\\n kwargs[\\\"con\\\"]._session\\n )\\n conn_gpu_count = conn_hardware_info.hardware_info[0].num_gpu_allocated\\n if conn_gpu_count == 0 or conn_gpu_count is None:\\n no_gather_nvml_gpu_info = True\\n if conn_gpu_count == 0:\\n logging.warning(\\n \\\"0 GPUs detected from connection info, \\\"\\n + \\\"using blank values for source database GPU info fields \\\"\\n + \\\"If running against cpu-only server, make sure to set \\\"\\n + \\\"--no-gather-nvml-gpu-info and --no-gather-conn-gpu-info.\\\"\\n )\\n else:\\n no_gather_nvml_gpu_info = kwargs[\\\"no_gather_nvml_gpu_info\\\"]\\n source_db_gpu_count = conn_gpu_count\\n try:\\n source_db_gpu_mem = int(\\n conn_hardware_info.hardware_info[0].gpu_info[0].memory\\n / 1000000\\n )\\n except IndexError:\\n logging.error(\\\"GPU memory info not available from connection.\\\")\\n if no_gather_nvml_gpu_info:\\n logging.debug(\\n \\\"--no-gather-nvml-gpu-info passed, \\\"\\n + \\\"using blank values for source database GPU info fields \\\"\\n + \\\"[gpu_driver_ver, run_gpu_name] \\\"\\n )\\n elif (\\n kwargs[\\\"conn_machine_name\\\"] == \\\"localhost\\\"\\n or kwargs[\\\"gather_nvml_gpu_info\\\"]\\n ):\\n logging.debug(\\n \\\"Gathering source database GPU info fields \\\"\\n + \\\"[gpu_driver_ver, run_gpu_name] \\\"\\n + \\\"from local GPU using pynvml. \\\"\\n )\\n import pynvml\\n\\n pynvml.nvmlInit()\\n source_db_gpu_driver_ver = pynvml.nvmlSystemGetDriverVersion().decode()\\n for i in range(source_db_gpu_count):\\n handle = pynvml.nvmlDeviceGetHandleByIndex(i)\\n # Assume all cards are the same, overwrite name value\\n source_db_gpu_name = pynvml.nvmlDeviceGetName(handle).decode()\\n pynvml.nvmlShutdown()\\n # If gpu_count argument passed in, override gathered value\\n if kwargs[\\\"gpu_count\\\"]:\\n source_db_gpu_count = kwargs[\\\"gpu_count\\\"]\\n if kwargs[\\\"gpu_name\\\"]:\\n source_db_gpu_name = kwargs[\\\"gpu_name\\\"]\\n gpu_info = {\\n \\\"conn_gpu_count\\\": conn_gpu_count,\\n \\\"source_db_gpu_count\\\": source_db_gpu_count,\\n \\\"source_db_gpu_mem\\\": source_db_gpu_mem,\\n \\\"source_db_gpu_driver_ver\\\": source_db_gpu_driver_ver,\\n \\\"source_db_gpu_name\\\": source_db_gpu_name,\\n }\\n return gpu_info\",\n \"def setup_CL():\\n\\n #Set up openCL platform\\n NAME = 'NVIDIA CUDA'\\n platforms = cl.get_platforms()\\n\\n dev = None\\n for p in platforms:\\n #Easy switching for local vs remote machine\\n if p.name == 'Apple':\\n NAME = 'Apple'\\n if p.name == NAME:\\n dev = p.get_devices()\\n\\n # Command queue, enable GPU profiling\\n ctx = cl.Context(dev)\\n queue = cl.CommandQueue(ctx,properties=cl.command_queue_properties.PROFILING_ENABLE)\\n\\n return [dev,ctx,queue]\",\n \"def _get_available_gpus():\\r\\n #global _LOCAL_DEVICES\\r\\n if tfback._LOCAL_DEVICES is None:\\r\\n devices = tf.config.list_logical_devices()\\r\\n tfback._LOCAL_DEVICES = [x.name for x in devices]\\r\\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def set_current_device_id():\\n import paddle\\n curr_dev = paddle.device.get_device()\\n select_gpu = os.getenv(\\\"FLAGS_selected_gpus\\\", \\\"0\\\")\\n paddle.set_flags({\\n 'FLAGS_selected_gpus': os.getenv(\\\"FLAGS_selected_gpus\\\", \\\"0\\\")\\n })\\n if \\\"gpu\\\" in curr_dev and select_gpu != curr_dev.split(\\\":\\\")[-1]:\\n paddle.set_device(\\\"gpu:\\\" + select_gpu)\\n\\n curr_dev_id = paddle.framework.core.get_cuda_current_device_id()\\n if \\\"gpu\\\" in curr_dev and select_gpu != str(curr_dev_id):\\n paddle.zeros([])\",\n \"def cpu(self):\\n self.reader.model.cpu()\\n self.reader.device = torch.device(\\\"cpu\\\")\\n return self\",\n \"def get_free_gpu():\\n\\tos.system('nvidia-smi -q -d Memory |grep -A4 GPU|grep Free >tmp')\\n\\tif os.path.exists('tmp'):\\n\\t\\tmemory_available = [int(x.split()[2]) for x in open('tmp', 'r').readlines()]\\n\\t\\tos.remove('tmp')\\n\\t\\treturn np.argmax(memory_available)\\n\\treturn 0\",\n \"def check_cuda():\\n if OS_VERSION[0] == \\\"Linux\\\":\\n check_cuda_linux()\\n elif OS_VERSION[0] == \\\"Windows\\\":\\n check_cuda_windows()\",\n \"def _get_available_gpus():\\n # global _LOCAL_DEVICES\\n if tfback._LOCAL_DEVICES is None:\\n devices = tf.config.list_logical_devices()\\n tfback._LOCAL_DEVICES = [x.name for x in devices]\\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def min_gpu(self) -> Optional[pulumi.Input[int]]:\\n return pulumi.get(self, \\\"min_gpu\\\")\",\n \"def get_test_devices():\\n\\n # Assumption: CPU is always available\\n devices = ['cpu']\\n\\n if torch.cuda.is_available():\\n devices.append('cuda')\\n\\n return devices\",\n \"def _get_available_gpus():\\n # global _LOCAL_DEVICES\\n if tf_back._LOCAL_DEVICES is None:\\n devices = tf.config.list_logical_devices()\\n tf_back._LOCAL_DEVICES = [x.name for x in devices]\\n return [x for x in tf_back._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def device(request):\\n d = request.param()\\n\\n # enable GPU error checking\\n if isinstance(d, hoomd.device.GPU):\\n d.gpu_error_checking = True\\n\\n return d\",\n \"def OnGPU(gpu_id):\\n device_option = caffe2_pb2.DeviceOption()\\n device_option.device_type = workspace.GpuDeviceType\\n device_option.device_id = gpu_id\\n return device_option\",\n \"def only_gpu(request):\\n if request.node.get_closest_marker('gpu'):\\n if 'device' in request.fixturenames:\\n if not isinstance(request.getfixturevalue('device'),\\n hoomd.device.GPU):\\n pytest.skip('Test is run only on GPU(s).')\\n else:\\n raise ValueError('only_gpu requires the *device* fixture')\",\n \"def maybe_cuda(x: Union[torch.Tensor, nn.Module]) -> Union[torch.Tensor, nn.Module]:\\n if torch.cuda.is_available():\\n return x.cuda()\\n return x\",\n \"def _get_available_gpus():\\n #global _LOCAL_DEVICES\\n if tfback._LOCAL_DEVICES is None:\\n devices = tf.config.list_logical_devices()\\n tfback._LOCAL_DEVICES = [x.name for x in devices]\\n return [x for x in tfback._LOCAL_DEVICES if 'device:gpu' in x.lower()]\",\n \"def is_gpu_available():\\n ret = get_gpu_count() > 0\\n if _HAS_PADDLE:\\n import paddle\\n if ret is True and not paddle.is_compiled_with_cuda():\\n logger.warning(\\\"Found non-empty CUDA_VISIBLE_DEVICES. \\\\\\n But PARL found that Paddle was not complied with CUDA, which may cause issues. \\\\\\n Thus PARL will not use GPU.\\\")\\n return False\\n if _HAS_FLUID:\\n from paddle import fluid\\n if ret is True and not fluid.is_compiled_with_cuda():\\n logger.warning(\\\"Found non-empty CUDA_VISIBLE_DEVICES. \\\\\\n But PARL found that Paddle was not complied with CUDA, which may cause issues. \\\\\\n Thus PARL will not use GPU.\\\")\\n return False\\n return ret\",\n \"def get_default_device():\\n global _default_device\\n\\n if _default_device is None:\\n import wgpu.backends.rs # noqa\\n\\n adapter = wgpu.request_adapter(canvas=None, power_preference=\\\"high-performance\\\")\\n _default_device = adapter.request_device()\\n return _default_device\",\n \"def _on_gpu(self) -> bool:\\n return self._current_device_index != CPU_INDEX\",\n \"def _prepare_device(self, n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n self.logger.warning(\\n \\\"Warning: There\\\\'s no GPU available on this machine, training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n self.logger.warning(\\n \\\"Warning: The number of GPU\\\\'s configured to use is {}, but only {} are available on this machine.\\\".format(\\n n_gpu_use, n_gpu))\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n return device, list_ids\",\n \"def use_cuda():\\n return torch.cuda.is_available() and os.getenv('AICROWD_CUDA', True)\",\n \"def gpu_availability():\\n # assume if using tensorflow-gpu, then Nvidia GPU is available\\n if is_built_with_cuda():\\n return len(tf.config.list_physical_devices(\\\"GPU\\\")) > 0\\n else:\\n return False\",\n \"def get_cpu_info():\\n try:\\n cpu_info = subprocess.check_output('lscpu')\\n return cpu_info\\n except OSError:\\n return None\",\n \"def test_gpu_cuda_code() -> None:\\n if get_from_environ(\\\"DISABLE_GPU_FOR_TESTING\\\") is not None:\\n print(\\\"GPU payload disabled for testing\\\")\\n return\\n\\n # if the command exists it can run on the hardware below\\n proc = subprocess.Popen([\\\"nvidia-smi\\\"], stdout=subprocess.PIPE)\\n stdout, _ = proc.communicate()\\n str_stdout = stdout.decode()\\n assert \\\"NVIDIA-SMI\\\" in str_stdout, str_stdout\\n assert proc.returncode == 0\\n # search the history for the CUDA implementation\",\n \"def get_test_devices():\\n devices = [\\\"cpu\\\"]\\n if torch.cuda.is_available():\\n devices.append(\\\"cuda\\\")\\n return devices\",\n \"def _prepare_device(self, n_gpu_use):\\n n_gpu = torch.cuda.device_count()\\n if n_gpu_use > 0 and n_gpu == 0:\\n self.logger.warning(\\\"Warning: There\\\\'s no GPU available on this machine,\\\"\\n \\\"training will be performed on CPU.\\\")\\n n_gpu_use = 0\\n if n_gpu_use > n_gpu:\\n self.logger.warning(\\\"Warning: The number of GPU\\\\'s configured to use is {}, but only {} are available \\\"\\n \\\"on this machine.\\\".format(n_gpu_use, n_gpu))\\n n_gpu_use = n_gpu\\n device = torch.device('cuda:0' if n_gpu_use > 0 else 'cpu')\\n list_ids = list(range(n_gpu_use))\\n return device, list_ids\",\n \"def with_cpu(ops, model):\\n ...\",\n \"def check_device(use_gpu, use_xpu=False, use_npu=False, use_mlu=False):\\n err = \\\"Config {} cannot be set as true while your paddle \\\" \\\\\\n \\\"is not compiled with {} ! \\\\nPlease try: \\\\n\\\" \\\\\\n \\\"\\\\t1. Install paddlepaddle to run model on {} \\\\n\\\" \\\\\\n \\\"\\\\t2. Set {} as false in config file to run \\\" \\\\\\n \\\"model on CPU\\\"\\n\\n try:\\n if use_gpu and use_xpu:\\n print(\\\"use_xpu and use_gpu can not both be ture.\\\")\\n if use_gpu and not paddle.is_compiled_with_cuda():\\n print(err.format(\\\"use_gpu\\\", \\\"cuda\\\", \\\"gpu\\\", \\\"use_gpu\\\"))\\n sys.exit(1)\\n if use_xpu and not paddle.device.is_compiled_with_xpu():\\n print(err.format(\\\"use_xpu\\\", \\\"xpu\\\", \\\"xpu\\\", \\\"use_xpu\\\"))\\n sys.exit(1)\\n if use_npu:\\n if int(paddle.version.major) != 0 and int(\\n paddle.version.major) <= 2 and int(\\n paddle.version.minor) <= 4:\\n if not paddle.device.is_compiled_with_npu():\\n print(err.format(\\\"use_npu\\\", \\\"npu\\\", \\\"npu\\\", \\\"use_npu\\\"))\\n sys.exit(1)\\n # is_compiled_with_npu() has been updated after paddle-2.4\\n else:\\n if not paddle.device.is_compiled_with_custom_device(\\\"npu\\\"):\\n print(err.format(\\\"use_npu\\\", \\\"npu\\\", \\\"npu\\\", \\\"use_npu\\\"))\\n sys.exit(1)\\n if use_mlu and not paddle.device.is_compiled_with_mlu():\\n print(err.format(\\\"use_mlu\\\", \\\"mlu\\\", \\\"mlu\\\", \\\"use_mlu\\\"))\\n sys.exit(1)\\n except Exception as e:\\n pass\",\n \"def device_placement(self):\\n if is_tf_available():\\n import tensorflow as tf\\n with tf.device('/CPU:0' if self.device == -1 else '/device:GPU:{}'.format(self.device)):\\n yield\\n else:\\n import torch\\n if self.device >= 0:\\n torch.cuda.set_device(self.device)\\n\\n yield\",\n \"def InitCL(GPU=0):\\n platform, device, context, queue = None, None, None, None\\n for iplatform in range(3):\\n try:\\n platform = cl.get_platforms()[iplatform]\\n if (GPU>0):\\n device = platform.get_devices(cl.device_type.GPU)\\n else:\\n device = platform.get_devices(cl.device_type.CPU)\\n context = cl.Context(device)\\n queue = cl.CommandQueue(context)\\n break\\n except:\\n pass\\n return platform, device, context, queue, cl.mem_flags\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.764121","0.7275213","0.72637","0.7186788","0.7095117","0.70836645","0.7030326","0.7010705","0.6991861","0.698923","0.698923","0.698923","0.6968887","0.6954786","0.6902639","0.6902639","0.68919474","0.6876222","0.6868331","0.6805084","0.67924947","0.6769179","0.67539823","0.66931343","0.66306907","0.6611765","0.6529878","0.65268654","0.6508832","0.6504441","0.6503202","0.64771694","0.64720714","0.6467823","0.6433878","0.6412888","0.6402619","0.6380536","0.6374527","0.6288381","0.6287262","0.6281002","0.6267664","0.6255833","0.6255833","0.62453264","0.62453264","0.6211627","0.6204074","0.6204074","0.6179717","0.6156944","0.6146286","0.61455554","0.6136982","0.6132767","0.61292213","0.6122363","0.6101238","0.6090462","0.6090462","0.6088164","0.6088164","0.60812515","0.60795003","0.6070867","0.6067297","0.60623527","0.6062345","0.6054775","0.60491824","0.60379684","0.6037647","0.60319453","0.60207146","0.60176617","0.60147995","0.60136306","0.6012459","0.60059935","0.6004722","0.5991201","0.59855413","0.5984648","0.59832627","0.59818316","0.5980316","0.59706664","0.59640956","0.59538597","0.59469277","0.59322274","0.5932119","0.59172475","0.59165823","0.5891552","0.58832514"],"string":"[\n \"0.764121\",\n \"0.7275213\",\n \"0.72637\",\n \"0.7186788\",\n \"0.7095117\",\n \"0.70836645\",\n \"0.7030326\",\n \"0.7010705\",\n \"0.6991861\",\n \"0.698923\",\n \"0.698923\",\n \"0.698923\",\n \"0.6968887\",\n \"0.6954786\",\n \"0.6902639\",\n \"0.6902639\",\n \"0.68919474\",\n \"0.6876222\",\n \"0.6868331\",\n \"0.6805084\",\n \"0.67924947\",\n \"0.6769179\",\n \"0.67539823\",\n \"0.66931343\",\n \"0.66306907\",\n \"0.6611765\",\n \"0.6529878\",\n \"0.65268654\",\n \"0.6508832\",\n \"0.6504441\",\n \"0.6503202\",\n \"0.64771694\",\n \"0.64720714\",\n \"0.6467823\",\n \"0.6433878\",\n \"0.6412888\",\n \"0.6402619\",\n \"0.6380536\",\n \"0.6374527\",\n \"0.6288381\",\n \"0.6287262\",\n \"0.6281002\",\n \"0.6267664\",\n \"0.6255833\",\n \"0.6255833\",\n \"0.62453264\",\n \"0.62453264\",\n \"0.6211627\",\n \"0.6204074\",\n \"0.6204074\",\n \"0.6179717\",\n \"0.6156944\",\n \"0.6146286\",\n \"0.61455554\",\n \"0.6136982\",\n \"0.6132767\",\n \"0.61292213\",\n \"0.6122363\",\n \"0.6101238\",\n \"0.6090462\",\n \"0.6090462\",\n \"0.6088164\",\n \"0.6088164\",\n \"0.60812515\",\n \"0.60795003\",\n \"0.6070867\",\n \"0.6067297\",\n \"0.60623527\",\n \"0.6062345\",\n \"0.6054775\",\n \"0.60491824\",\n \"0.60379684\",\n \"0.6037647\",\n \"0.60319453\",\n \"0.60207146\",\n \"0.60176617\",\n \"0.60147995\",\n \"0.60136306\",\n \"0.6012459\",\n \"0.60059935\",\n \"0.6004722\",\n \"0.5991201\",\n \"0.59855413\",\n \"0.5984648\",\n \"0.59832627\",\n \"0.59818316\",\n \"0.5980316\",\n \"0.59706664\",\n \"0.59640956\",\n \"0.59538597\",\n \"0.59469277\",\n \"0.59322274\",\n \"0.5932119\",\n \"0.59172475\",\n \"0.59165823\",\n \"0.5891552\",\n \"0.58832514\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.6680545"},"document_rank":{"kind":"string","value":"26"}}},{"rowIdx":90898,"cells":{"query":{"kind":"string","value":"Move tensor(s) to chosen device"},"document":{"kind":"string","value":"def to_device(data, device):\n if isinstance(data, (list,tuple)):\n return [to_device(x, device) for x in data]\n return data.to(device, non_blocking=True)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def move_to_device(input_tensors: List[torch.Tensor],\n target_device: Optional[torch.device],\n non_blocking: bool = False) -> Iterable[torch.Tensor]:\n return (tensor if tensor.device == target_device or target_device is None\n else tensor.to(target_device, non_blocking=non_blocking)\n for tensor in input_tensors)","def move2gpu(self):\n if self.generator.backend.upper() == \"LOOPY\":\n try:\n import pyopencl as cl\n import pyopencl.array # pylint: disable=unused-variable\n from .context import queue\n except ImportError:\n raise ImportError(\"Please install loo.py\")\n\n self.rhs = cl.array.to_device(queue, self.rhs)\n if hasattr(self, \"s\"):\n self.s = cl.array.to_device(\n queue, self.s\n ) # pylint: disable=attribute-defined-outside-init\n self.istore = cl.array.to_device(queue, self.istore)\n for i in range(len(self.iload)):\n self.iload[i] = cl.array.to_device(queue, self.iload[i])","def to(self, device):\n for item in self.data:\n if torch.is_tensor(item):\n item.to(item)\n else:\n for subitem in item:\n subitem.to(device)\n return self","def move_to_device(obj, cuda_device: int):\n\n if cuda_device < 0 or not has_tensor(obj):\n return obj\n elif isinstance(obj, torch.Tensor):\n return obj.cuda(cuda_device)\n elif isinstance(obj, dict):\n return {key: move_to_device(value, cuda_device) for key, value in obj.items()}\n elif isinstance(obj, list):\n return [move_to_device(item, cuda_device) for item in obj]\n elif isinstance(obj, tuple) and hasattr(obj, \"_fields\"):\n # This is the best way to detect a NamedTuple, it turns out.\n return obj.__class__(*(move_to_device(item, cuda_device) for item in obj))\n elif isinstance(obj, tuple):\n return tuple(move_to_device(item, cuda_device) for item in obj)\n else:\n return obj","def move_and_process_input(batch):\n x, y = batch\n x = x.to(device).float()\n y = torch.as_tensor(y).to(device)\n x = x.permute(0, -1, 1, 2, 3)\n return x, y","def batch_to_device(batch):\n for key in batch:\n if isinstance(batch[key], torch.Tensor):\n batch[key] = batch[key].to(device)\n return batch","def to_device(data, device):\n if isinstance(data, (list,tuple)): # allows to apply function to lists or tuples of tensors\n return [to_device(x, device) for x in data]\n return data.to(device, non_blocking=True)","def deploy_to_device(self):\n if self.device_ids is not None and len(self.device_ids) > 1:\n if not isinstance(self.model, torch.nn.DataParallel):\n self.model = torch.nn.DataParallel(self.model, self.device_ids)\n\n self.model = self.model.to(self.device)\n self.criterion = self.criterion.to(self.device)","def set_device(self, device: torch.Tensor) -> None:\n raise NotImplementedError","def _send(x, dst=0):\n x = torch.tensor(x)\n x = to_device(x)\n dist.send(x, dst)\n del x \n torch.cuda.empty_cache()","def _batch_to_device(batch, target_device):\n tensor = _getattr(\"torch\", \"Tensor\")\n for key in batch:\n if isinstance(batch[key], tensor):\n batch[key] = batch[key].to(target_device)\n return batch","def device_placement(self):\n if is_tf_available():\n import tensorflow as tf\n with tf.device('/CPU:0' if self.device == -1 else '/device:GPU:{}'.format(self.device)):\n yield\n else:\n import torch\n if self.device >= 0:\n torch.cuda.set_device(self.device)\n\n yield","def __move_sample_to_device(self, sample):\n _in, _target = sample\n _in = _in.to(self.device).float()\n _target = _target.to(self.device).long()\n\n return _in, _target","def move_devices(self, data):\n data = clean(data, self.move_parameters)\n return self.put(\"/devices/move\", data)","def switch_to_tuned_inputs(self):\n \n self.h_e=self.inputs_flat.T\n self.h=np.vstack([self.h_e,self.h_i])","def to_device(self, move_optim_data=True):\n if move_optim_data:\n self.optimizer_values = self.optimizer_values.to(\n self._embeddings.weight.device\n )","def train(self):\n if len(self.buffer) >= self.batch_size:\n with torch.no_grad():\n states, actions, rewards, next_states, dones = self.buffer.sample(self.batch_size)\n\n # Send data to GPU\n states = torch.stack(states).to(self.device, dtype=torch.float)\n actions = torch.stack(actions).to(self.device, dtype=torch.float)\n rewards = torch.stack(rewards).to(self.device, dtype=torch.float)\n rewards = torch.reshape(rewards, (self.batch_size, 1))\n\n next_states = torch.stack(next_states).to(self.device, dtype=torch.float)\n dones = torch.stack(dones).to(self.device, dtype=torch.float)\n\n #TODO\n\n # Calculate target Q values using the Target Network\n selection = torch.argmax(self.main_dqn(next_states), dim = 1).unsqueeze(1)\n\n evaluation = self.target_dqn(next_states)\n evaluation = evaluation.gather(1, selection.long()) #size [256,1]\n\n #Create Done mask\n nonzero_indices = torch.nonzero(dones).reshape(-1).tolist()\n dones_mask = torch.eye(self.batch_size)\n for index in nonzero_indices:\n dones_mask[index,index] = 0\n dones_mask = dones_mask.to(self.device, dtype=torch.float)\n\n # Calculte target\n target = rewards + torch.matmul(dones_mask, evaluation*self.gamma)\n target = target.detach()\n\n # Calculate Q values using the Main Network\n if self.env.freely_moving:\n n_classes = self.env.number_of_action_channels * self.env.number_of_rows * self.env.number_of_columns\n else:\n n_classes = self.env.number_of_action_channels * 1 * self.env.nA\n\n n_samples = self.batch_size\n labels = torch.flatten(actions.type(torch.LongTensor), start_dim=0)\n labels_tensor = torch.as_tensor(labels)\n action_masks = torch.nn.functional.one_hot(labels_tensor, num_classes=n_classes).to(self.device, dtype=torch.float)\n\n q_value = action_masks * self.main_dqn(states)\n q_value = torch.sum(q_value, dim=-1).reshape((self.batch_size, 1))\n\n # Calculate loss\n loss = self.mse(target, q_value)\n\n # Optimize the model\n self.optimizer.zero_grad()\n loss.backward()\n torch.nn.utils.clip_grad_norm_(self.main_dqn.parameters(), 5)\n self.optimizer.step()\n\n # Soft Copy the Main Network's weights to the Target Network\n self.soft_update_of_target_network(self.main_dqn, self.target_dqn,tau=1e-3)\n\n return loss\n return 0","def change_device(self, device=None):\n\n if device is None:\n # If the function is called without a device, use the current device\n device = self.device\n\n # Create the appropriate device object\n device = torch.device(f'cuda:{device}'\n if torch.cuda.is_available() else 'cpu')\n\n # Change device field\n self.device = device\n # Load the transcription model onto the device\n self.to(self.device)","def move(model, origin, dest):\n model.move(origin, dest)","def move(model, origin, dest):\n model.move(origin, dest)","def to_device(model, device):\n p = next(model.parameters())\n if p.device == device:\n return\n model.to(device)","def to_device(m: torch.nn.Module, x:torch.Tensor):\n if isinstance(m, torch.nn.Module):\n device = next(m.parameters()).device\n elif isinstance(m, torch.Tensor):\n device = m.device\n else:\n raise TypeError(\n \"Expected torch.nn.Module or torch.tensor, \" f\"bot got: {type(m)}\"\n )\n return x.to(device)","def _move_module_to_device(\n self,\n module: nn.Module,\n ignored_params: Set[nn.Parameter],\n device_from_device_id: Optional[torch.device],\n ):\n cpu_device = torch.device(\"cpu\")\n param = next(self._get_orig_params(module, ignored_params), None)\n if param is None:\n return # no original parameters to manage\n if device_from_device_id is not None:\n if param.device == cpu_device:\n # NOTE: This includes moving ignored modules' parameters.\n module = module.to(device_from_device_id)\n # TODO: This is a temporary fix to move already- constructed\n # `FlatParameter`s back to CPU if needed. This is needed to\n # make CPU offload work with `device_id`.\n for submodule in module.modules():\n if (\n isinstance(submodule, FullyShardedDataParallel)\n and submodule.cpu_offload.offload_params\n ):\n with torch.no_grad():\n for handle in submodule._handles:\n handle._flat_param_to(torch.device(\"cpu\"))\n elif param.device == cpu_device:\n warnings.warn(\n \"Module is put on CPU and will thus have flattening and sharding\"\n \" run on CPU, which is less efficient than on GPU. We recommend passing in \"\n \"`device_id` argument which will enable FSDP to put module on GPU device,\"\n \" module must also be on GPU device to work with `sync_module_states=True` flag\"\n \" which requires GPU communication.\"\n )","def swap_xy(tensor: tf.Tensor) -> tf.Tensor:\n return moveaxis(tensor, -2, -1)","def place_on_device(x, device):\n if isinstance(x, list):\n return [place_on_device(xi, device) for xi in x]\n elif isinstance(x, dict):\n return {k: place_on_device(v, device) for k, v in x.items()}\n elif isinstance(x, set):\n return {place_on_device(xi, device) for xi in x}\n elif isinstance(x, tuple):\n return tuple(place_on_device(xi, device) for xi in x)\n elif isinstance(x, torch.Tensor):\n return x\n else:\n return x","def forward(self, observation: Tensor) -> Tensor:\n pass","def set_src_dst_tensor(self, tik_instance):\n src_element_number = functools_reduce(lambda x1, x2: x1 * x2,\n self.src_shape_ori[:])\n dst_element_number = functools_reduce(lambda x1, x2: x1 * x2,\n self.dst_shape[:])\n self.src_gm = tik_instance.Tensor(self.dtype,\n (src_element_number,),\n name=\"src_gm\",\n scope=tik.scope_gm)\n self.dst_gm = tik_instance.Tensor(self.dtype,\n (dst_element_number,),\n name=\"dst_gm\",\n scope=tik.scope_gm)","def set_src_dst_tensor(self, tik_instance):\n src_element_number = functools_reduce(lambda x1, x2: x1 * x2,\n self.src_shape_ori[:])\n dst_element_number = functools_reduce(lambda x1, x2: x1 * x2,\n self.dst_shape[:])\n self.src_gm = tik_instance.Tensor(self.dtype,\n (src_element_number,),\n name=\"src_gm\",\n scope=tik.scope_gm)\n self.dst_gm = tik_instance.Tensor(self.dtype,\n (dst_element_number,),\n name=\"dst_gm\",\n scope=tik.scope_gm)","def set_src_dst_tensor(self, tik_instance):\n src_element_number = functools_reduce(lambda x, y: x * y,\n self.src_shape[:])\n dst_element_number = functools_reduce(lambda x, y: x * y,\n self.dst_shape[:])\n self.src_gm = tik_instance.Tensor(self.dtype,\n (src_element_number,),\n name=\"src_gm\",\n scope=tik.scope_gm)\n self.dst_gm = tik_instance.Tensor(self.dtype,\n (dst_element_number,),\n name=\"dst_gm\",\n scope=tik.scope_gm)","def set_device(sys_device_id):\n device_id = -1\n cuda = (sys_device_id != -1)\n if cuda:\n # CUDA_VISIBLE_DEVICE is a list, and device_id is the index of its members.\n import os\n os.environ['CUDA_VISIBLE_DEVICES'] = str(sys_device_id)\n device_id = 0\n TVT = TransferVarTensor(device_id)\n TMO = TransferModulesOptims(device_id)\n return TVT, TMO","def distribute_tensor(\n tensor: torch.Tensor,\n device_mesh: Optional[DeviceMesh] = None,\n placements: Optional[Sequence[Placement]] = None,\n) -> DTensor:\n\n torch._C._log_api_usage_once(\"torch.dtensor.distribute_tensor\")\n\n # get default device mesh if there's nothing specified\n device_mesh = device_mesh or mesh_resources.get_current_mesh()\n\n # instantiate a RNG tracker if haven't. By default DTensor uses an\n # OffsetBasedRNGTracker to perform random operators.\n # TODO: the value assignment to global variable is not the ideal solution\n # we can replace it in future.\n if is_rng_supported_mesh(device_mesh) and not random._rng_tracker:\n random._rng_tracker = OffsetBasedRNGTracker()\n\n if not tensor.is_leaf:\n raise RuntimeError(\n \"`distribute_tensor` should be used to distribute leaf tensors! but found non-leaf tensor!\"\n )\n\n # convert tensor to the corresponding device type if it's not in that device type\n if not tensor.is_meta:\n tensor = tensor.to(device_mesh.device_type)\n # set default placements to replicated if not specified\n if placements is None:\n placements = [Replicate() for _ in range(device_mesh.ndim)]\n\n if len(placements) != device_mesh.ndim:\n raise ValueError(\n f\"`placements` must have the same length as `device_mesh.ndim`! \"\n f\"Found placements length: {len(placements)}, and device_mesh.ndim: {device_mesh.ndim}.\"\n )\n\n if isinstance(tensor, DTensor):\n # if the tensor is already a DTensor, we just need to check if the\n # device mesh and placements are the same\n if tensor.device_mesh != device_mesh:\n raise ValueError(\n f\"Cannot distribute a DTensor with device mesh {tensor.device_mesh} \"\n f\"to a different device mesh {device_mesh}.\"\n )\n if tensor.placements != placements:\n raise ValueError(\n f\"Cannot distribute a DTensor with placements {tensor.placements} \"\n f\"to a different placements {placements}. do you want to call \"\n f\"`redistribute` instead?\"\n )\n return tensor\n\n local_tensor = tensor\n\n # distribute the tensor according to the placements.\n for idx, placement in enumerate(placements):\n if placement.is_shard():\n placement = cast(Shard, placement)\n output = placement._shard_tensor(local_tensor, device_mesh, idx)\n # scatter call could not return a tensor with correct requires_grad\n # field, as ProcessGroupNCCL refuse to take a tensor with requires_grad\n # to do inplace update! So we manually set it here\n output.requires_grad_(tensor.requires_grad)\n local_tensor = output\n elif placement.is_replicate():\n placement = cast(Replicate, placement)\n local_tensor = placement._replicate_tensor(local_tensor, device_mesh, idx)\n else:\n raise RuntimeError(\n f\"Trying to distribute tensor with unsupported placements {placement} on device mesh dimension {idx}!\"\n )\n\n assert local_tensor is not None, \"distributing a tensor should not be None\"\n # detach the local tensor passed to DTensor since after the construction\n # of DTensor, autograd would work on top of DTensor instead of local tensor\n return DTensor(\n local_tensor.detach(),\n device_mesh,\n placements,\n shape=tensor.size(),\n dtype=tensor.dtype,\n requires_grad=tensor.requires_grad,\n stride=tensor.stride(),\n )","def batch_to_device(batch, target_device: device):\n moved_batch = {}\n for key, val in batch.items():\n val = val.to(target_device)\n moved_batch[key] = val\n return moved_batch","def copy_tensor(src, dest):\n dest.copy_(src)","def to(self, device):\n self.device = device\n self.model.to(self.device)","def _recursive_to_cuda(self, tensors: Union[Tuple[torch.Tensor], torch.Tensor]) \\\n -> Union[List[torch.Tensor], torch.Tensor]:\n if self.device is None: # keep on cpu\n return tensors\n\n if type(tensors) != list and type(tensors) != tuple: # not only for torch.Tensor\n return tensors.to(device=self.device)\n\n cuda_tensors = list()\n for i in range(len(tensors)):\n cuda_tensors.append(self._recursive_to_cuda(tensors[i]))\n return cuda_tensors","def module_transfer_to_device(self) -> None:\n for name, module in self.modules.items():\n module.to(self.device)\n if self.device.type == 'cuda':\n self.modules[name] = torch.nn.DataParallel(module, self.gpu_ids)\n return","def forward(self, bottom, top):\n blobs = self._get_next_minibatch() # Get a blob\n\n for blob_id, blob in enumerate(blobs): # blob[0] for image, blob[1] for mask\n # Reshape net's input blobs\n top[blob_id].reshape(*(blob.shape)) # Reshape top tensor/blob\n top[blob_id].data[...] = blob.astype(np.float32, copy=False) # Feed corresponding data into top tensor/blob","def mv_step(self):\n # def mv_all(self):\n self.device_reg_data &= ~(0x1 << 3)\n bus.write_byte_data(self.device_address, self.device_reg_mode1, self.device_reg_data)","def _eunn_loop(state, capacity, diag_vec_list, off_vec_list, diag, fft):\n i = 0\n def layer_tunable(x, i):\n\n diag_vec = diag_vec_list.read(i)\n off_vec = off_vec_list.read(i)\n\n diag = tf.multiply(x, diag_vec)\n off = tf.multiply(x, off_vec)\n\n def even_input(off, size):\n\n def even_s(off, size):\n off = tf.reshape(off, [-1, size//2, 2])\n off = tf.reshape(tf.reverse(off, [2]), [-1, size])\n return off\n\n def odd_s(off, size):\n off, helper = tf.split(off, [size-1, 1], 1)\n size -= 1\n off = even_s(off, size)\n off = tf.concat([off, helper], 1)\n return off\n\n off = tf.cond(tf.equal(tf.mod(size, 2), 0), lambda: even_s(off, size), lambda: odd_s(off, size))\n return off\n\n def odd_input(off, size):\n helper, off = tf.split(off, [1, size-1], 1)\n size -= 1\n off = even_input(off, size)\n off = tf.concat([helper, off], 1)\n return off\n\n size = int(off.get_shape()[1])\n off = tf.cond(tf.equal(tf.mod(i, 2), 0), lambda: even_input(off, size), lambda: odd_input(off, size))\n\n layer_output = diag + off\n i += 1\n\n return layer_output, i\n\n def layer_fft(state, i):\n\n diag_vec = diag_vec_list.read(i)\n off_vec = off_vec_list.read(i)\n diag = tf.multiply(state, diag_vec)\n off = tf.multiply(state, off_vec)\n\n hidden_size = int(off.get_shape()[1])\n # size = 2**i\n dist = capacity - i\n normal_size = (hidden_size // (2**dist)) * (2**(dist-1))\n normal_size *= 2\n extra_size = tf.maximum(0, (hidden_size % (2**dist)) - (2**(dist-1)))\n hidden_size -= normal_size\n\n def modify(off_normal, dist, normal_size):\n off_normal = tf.reshape(tf.reverse(tf.reshape(off_normal, [-1, normal_size//(2**dist), 2, (2**(dist-1))]), [2]), [-1, normal_size])\n return off_normal\n\n def do_nothing(off_normal):\n return off_normal\n\n off_normal, off_extra = tf.split(off, [normal_size, hidden_size], 1)\n off_normal = tf.cond(tf.equal(normal_size, 0), lambda: do_nothing(off_normal), lambda: modify(off_normal, dist, normal_size))\n helper1, helper2 = tf.split(off_extra, [hidden_size-extra_size, extra_size], 1)\n off_extra = tf.concat([helper2, helper1], 1)\n off = tf.concat([off_normal, off_extra], 1)\n\n layer_output = diag + off\n i += 1\n\n return layer_output, i\n\n if fft:\n layer_function = layer_fft\n else:\n layer_function = layer_tunable\n output, _ = tf.while_loop(lambda state, i: tf.less(i, capacity), layer_function, [state, i])\n\n if not diag is None:\n output = tf.multiply(output, diag)\n\n\n return output","def set_devices(args):\n global devices\n if args is not None:\n devices = [torch.device(i) for i in ast.literal_eval('[' + args + ']')]\n torch.cuda.set_device(devices[0])\n else:\n devices = [torch.device('cpu')]","def forward_batch(model, batch, device):\n inputs, labels = (batch, batch)\n inputs, labels = inputs.to(device), labels.to(device)\n\n outputs = model(inputs, labels=labels)\n\n return outputs[:2]","def _update(self):\n self.parametrize_beam()\n self.update_ranks()\n self._points = tf.reshape(self._endpoint, (1, 2)) * tf.reshape(self._ranks, (-1, 1))","def to_device(self, device):\n for i in range(self.num_layers):\n getattr(self, \"conv{}\".format(i+1)).to_device(device)\n self.to(device)\n return self","def update_target_network(self):\n variables = self.online_network.trainable_variables\n variables_copy = [tf.Variable(v) for v in variables]\n self.target_network.trainable_variables = variables_copy","def forward_target(self, img, **kwargs):\n assert img.dim() == 5, \\\n \"Input must have 5 dims, got: {}\".format(img.dim())\n\n img_v_l = []\n idx_unshuffle_v_l = []\n for idx in range(img.shape[1]):\n img_vi = img[:, idx, ...].contiguous()\n img_vi, idx_unshuffle_vi = self._batch_shuffle_ddp(img_vi)\n img_v_l.append(img_vi)\n idx_unshuffle_v_l.append(idx_unshuffle_vi)\n\n if self.start_block > 0:\n for idx in range(img.shape[1]):\n for i, best_path in enumerate(self.best_paths):\n img_v_l[idx] = self.target_backbone(img_v_l[idx],\n start_block=i,\n forward_op=best_path,\n block_op=True)[0]\n\n self.forward_op_target = self.forward_op_online\n proj_target_v1 = 0\n proj_target_v2 = 0\n v2_idx = img.shape[1]//2\n with torch.no_grad():\n for op_idx, forward_singleop_target in enumerate(self.forward_op_target):\n temp_v1 = self.target_neck(self.target_backbone(img_v_l[op_idx],\n start_block=self.start_block,\n forward_op=forward_singleop_target))[\n 0].clone().detach()\n temp_v2 = self.target_neck(self.target_backbone(img_v_l[v2_idx + op_idx],\n start_block=self.start_block,\n forward_op=forward_singleop_target))[\n 0].clone().detach()\n temp_v1 = nn.functional.normalize(temp_v1, dim=1)\n temp_v1 = self._batch_unshuffle_ddp(temp_v1, idx_unshuffle_v_l[op_idx])\n\n temp_v2 = nn.functional.normalize(temp_v2, dim=1)\n temp_v2 = self._batch_unshuffle_ddp(temp_v2, idx_unshuffle_v_l[v2_idx + op_idx])\n\n proj_target_v1 += temp_v1\n proj_target_v2 += temp_v2\n\n self.proj_target_v1 = proj_target_v1 / (len(self.forward_op_target))\n self.proj_target_v2 = proj_target_v2 / (len(self.forward_op_target))","def forward(self, in_tensors: List[Tensor], out_tensors: List[Tensor]):\n pass","def forward(self, in_tensors: List[Tensor], out_tensors: List[Tensor]):\n pass","def moveaxis(tensor: tf.Tensor, source: int, destination: int) -> tf.Tensor:\n ndim = len(tensor.shape)\n source = _normalize_axis(source, ndim)\n destination = _normalize_axis(destination, ndim)\n order = [n for n in range(ndim) if n != source]\n order.insert(destination, source)\n return tf.transpose(tensor, order)","def move_to(self, T):\n self.camera_coords = coordinates.Coordinates(\n pos=np.array(T),\n rot=self.camera_coords.worldrot())\n pybullet.resetBasePositionAndOrientation(\n self.camera_id,\n self.camera_coords.worldpos(),\n coordinates.math.wxyz2xyzw(\n coordinates.math.matrix2quaternion(\n self.camera_coords.worldrot())))","def move(self):\n self.pos += self.direc\n self.nearest_node = self.pixel_to_node()","def move_data_to_device(self, data: Tuple) -> Tuple:\n tmp = []\n for dv in data:\n tmp.append(dv.to(self.device))\n \n return tuple(tmp)","def restore_leading_dims(tensors, lead_dim, T=1, B=1):\n is_seq = isinstance(tensors, (tuple, list))\n tensors = tensors if is_seq else (tensors,)\n if lead_dim == 2: # (Put T dim.)\n tensors = tuple(t.view((T, B) + t.shape[1:]) for t in tensors)\n if lead_dim == 0: # (Remove B=1 dim.)\n assert B == 1\n tensors = tuple(t.squeeze(0) for t in tensors)\n return tensors if is_seq else tensors[0]","def sample_batch(pid, args, batch_queue, port_dict, device, actor_id_to_ip_dataport, local_size, cache_array):\n def recv_data(k, data_stream, actor_set, real_data_tasks_i):\n for real_data in data_stream:\n tmp = []\n tmp.append(real_data.state)\n tmp.append(real_data.action)\n tmp.append(real_data.reward)\n tmp.append(real_data.next_state)\n tmp.append(real_data.done)\n tmp.append(actor_set[k]['w'][real_data.idx])\n tmp.append(actor_set[k]['i'][real_data.idx])\n tmp.append(actor_set[k]['t'][real_data.idx])\n tmp.append(real_data.timestamp)\n local_dict[actor_set[k]['i'][real_data.idx]] = tmp\n cache_array[actor_set[k]['i'][real_data.idx]] |= 2**pid\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.state), dtype=np.uint8).reshape((1, 4, 84, 84)))\n real_data_tasks_i['states'].append(decom_state) #.to(device))\n real_data_tasks_i['actions'].append(torch.LongTensor([real_data.action])) #.to(device))\n real_data_tasks_i['rewards'].append(torch.FloatTensor([real_data.reward])) #.to(device))\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.next_state), dtype=np.uint8).reshape((1, 4, 84, 84)))\n real_data_tasks_i['next_states'].append(decom_next_state) #.to(device))\n real_data_tasks_i['dones'].append(torch.FloatTensor([real_data.done])) #.to(device))\n real_data_tasks_i['batch_weights'].append(torch.FloatTensor([actor_set[k]['w'][real_data.idx]])) #.to(device))\n real_data_tasks_i['batch_idxes'].append(actor_set[k]['i'][real_data.idx])\n # is the data overwrited?\n real_data_tasks_i['batch_timestamp_store'].append(actor_set[k]['t'][real_data.idx])\n real_data_tasks_i['batch_timestamp_real'].append(real_data.timestamp)\n conn = grpc.insecure_channel(port_dict['replay_ip'] + ':' + port_dict['sampleDataPort'])\n client = apex_data_pb2_grpc.SampleDataStub(channel=conn)\n local_dict = {}\n while True:\n batch_timestamp_real = []\n batch_timestamp_store = []\n batch_weights = []\n batch_idxes = []\n\n states, actions, rewards, next_states, dones = [], [], [], [], []\n\n res_batch = client.Send(apex_data_pb2.SampleDataRequest(batch_size=args.batch_size, beta = args.beta))\n actor_ids, data_ids, timestamps, weights, idxes = res_batch.actor_ids, res_batch.data_ids, res_batch.timestamp, res_batch.weights, res_batch.idxes\n actor_set = {}\n cached_value = {'states':{},'actions':{},'rewards':{},'next_states':{},'dones':{},'batch_weights':{},'batch_idxes':{},'batch_timestamp_store':{},'batch_timestamp_real':{}}\n for i in range(len(actor_ids)):\n set_a = actor_set.get(actor_ids[i], False)\n if set_a == False:\n actor_set[actor_ids[i]] = {}\n set_a = actor_set[actor_ids[i]]\n set_a['d'] = []\n set_a['w'] = []\n set_a['i'] = []\n set_a['t'] = []\n cached_value['states'][actor_ids[i]] = []\n cached_value['actions'][actor_ids[i]] = []\n cached_value['rewards'][actor_ids[i]] = []\n cached_value['next_states'][actor_ids[i]] = []\n cached_value['dones'][actor_ids[i]] = []\n cached_value['batch_weights'][actor_ids[i]] = []\n cached_value['batch_idxes'][actor_ids[i]] = []\n cached_value['batch_timestamp_store'][actor_ids[i]] = []\n cached_value['batch_timestamp_real'][actor_ids[i]] = []\n cache_id = actor_ids[i]*local_size+data_ids[i]\n cache_trans = cache_array[cache_id]\n if cache_trans & 2**pid == 0:\n set_a['d'].append(data_ids[i])\n set_a['w'].append(weights[i])\n set_a['i'].append(idxes[i])\n set_a['t'].append(timestamps[i])\n if cache_trans == 0 and local_dict.get(cache_id, False) != False:\n del local_dict[cache_id]\n else:\n try:\n state_tmp = local_dict[cache_id][0]\n action_tmp = local_dict[cache_id][1]\n reward_tmp = local_dict[cache_id][2] \n next_state_tmp = local_dict[cache_id][3] \n done_tmp = local_dict[cache_id][4] \n batch_weight_tmp = local_dict[cache_id][5] \n batch_idx_tmp = local_dict[cache_id][6] \n batch_store_tmp = local_dict[cache_id][7] \n batch_real_tmp = local_dict[cache_id][8] \n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\n cached_value['states'][actor_ids[i]].append(decom_state)\n cached_value['actions'][actor_ids[i]].append(torch.LongTensor([action_tmp]))\n cached_value['rewards'][actor_ids[i]].append(torch.FloatTensor([reward_tmp]))\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(next_state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\n cached_value['next_states'][actor_ids[i]].append(decom_next_state)\n cached_value['dones'][actor_ids[i]].append(torch.FloatTensor([done_tmp]))\n cached_value['batch_weights'][actor_ids[i]].append(torch.FloatTensor([batch_weight_tmp]))\n cached_value['batch_idxes'][actor_ids[i]].append(batch_idx_tmp)\n cached_value['batch_timestamp_store'][actor_ids[i]].append(batch_store_tmp)\n cached_value['batch_timestamp_real'][actor_ids[i]].append(batch_real_tmp)\n except:\n set_a['d'].append(data_ids[i])\n set_a['w'].append(weights[i])\n set_a['i'].append(idxes[i])\n set_a['t'].append(timestamps[i])\n real_data_links = {}\n real_data_tasks = {}\n for k, v in actor_set.items():\n actor_ip, data_port = actor_id_to_ip_dataport[k]\n conn_actor = grpc.insecure_channel(actor_ip + ':' + data_port)\n client_actor = apex_data_pb2_grpc.SendRealDataStub(channel=conn_actor)\n real_data_links[k] = client_actor.Send(apex_data_pb2.RealBatchRequest(idxes=v['d']))\n real_data_tasks[k] = {}\n real_data_tasks[k]['states'] = cached_value['states'][k]\n real_data_tasks[k]['actions'] = cached_value['actions'][k]\n real_data_tasks[k]['rewards'] = cached_value['rewards'][k]\n real_data_tasks[k]['next_states'] = cached_value['next_states'][k]\n real_data_tasks[k]['dones'] = cached_value['dones'][k]\n real_data_tasks[k]['batch_weights'] = cached_value['batch_weights'][k]\n real_data_tasks[k]['batch_idxes'] = cached_value['batch_idxes'][k]\n real_data_tasks[k]['batch_timestamp_store'] = cached_value['batch_timestamp_store'][k]\n real_data_tasks[k]['batch_timestamp_real'] = cached_value['batch_timestamp_real'][k]\n threads = []\n for k, v in real_data_links.items():\n t = threading.Thread(target=recv_data, args=(k, v, actor_set, real_data_tasks[k],))\n threads.append(t)\n t.start()\n\n for t in threads:\n t.join()\n\n for k, v in real_data_tasks.items():\n states += v['states']\n actions += v['actions']\n rewards += v['rewards']\n next_states += v['next_states']\n dones += v['dones']\n batch_weights += v['batch_weights']\n batch_idxes += v['batch_idxes']\n batch_timestamp_real += v['batch_timestamp_real']\n batch_timestamp_store += v['batch_timestamp_store']\n\n states = torch.cat(states,0).to(device)\n actions = torch.cat(actions,0).to(device)\n rewards = torch.cat(rewards,0).to(device)\n next_states = torch.cat(next_states,0).to(device)\n dones = torch.cat(dones,0).to(device)\n batch_weights = torch.cat(batch_weights,0).to(device)\n\n batch = [states, actions, rewards, next_states, dones, batch_weights, batch_idxes]\n batch_queue.put(batch)\n data, batch = None, None","def sync_target_network(self):\n for t, e in zip(\n self.target_network.trainable_variables, self.online_network.trainable_variables\n ):\n t.assign(e)","def _hadamard_step(x, dim):\n x_shape = x.shape.as_list()\n x = tf.reshape(x, [-1, 2]) # Reshape so that we have a matrix.\n x = tf.matmul(x, h_core) # Multiply.\n x = tf.reshape(x, [-1, dim // 2, 2]) # Reshape to rank-3.\n x = tf.transpose(x, perm=permutation) # Swap last two dimensions.\n x.set_shape(x_shape) # Failed shape inference in tf.while_loop.\n return x","def _copy_to_gpu(self):\n self.dispatch('on_texture')","def cuda(self, device=None) -> MultiIndicesEmbedding:\n super().cuda(device=device)\n\n self.offsets = self.offsets.cuda(device)\n\n return self","def device(self):\n return self._tensor.device","def _4d_tensor(self, name, tensor, **kwargs):\r\n h, w, c, num_tensor = tensor.shape.as_list()\r\n\r\n # Try to visualise convolutional filters or feature maps\r\n # See: https://gist.github.com/kukuruza/03731dc494603ceab0c5\r\n # input shape: (Y, X, C, N)\r\n if c != 1 and c != 3:\r\n tensor = tf.reduce_mean(tensor, axis=2, keep_dims=True)\r\n c = 1\r\n # shape is now: (Y, X, 1|C, N)\r\n v_min = tf.reduce_min(tensor)\r\n v_max = tf.reduce_max(tensor)\r\n tensor -= v_min\r\n tensor *= 1.0 / (v_max - v_min)\r\n tensor = tf.pad(tensor, [[1, 0], [1, 0], [0, 0], [0, 0]], 'CONSTANT')\r\n tensor = tf.transpose(tensor, perm=(3, 0, 1, 2))\r\n # shape is now: (N, Y, X, C)\r\n # place tensor on grid\r\n num_tensor_x = int(np.round(np.sqrt(num_tensor)))\r\n num_tensor_y = num_tensor / num_tensor_x\r\n while not num_tensor_y.is_integer():\r\n num_tensor_x += 1\r\n num_tensor_y = num_tensor / num_tensor_x\r\n num_tensor_y = int(num_tensor_y)\r\n h += 1\r\n w += 1\r\n tensor = tf.reshape(tensor, (num_tensor_x, h * num_tensor_y, w, c))\r\n # shape is now: (N_x, Y * N_y, X, c)\r\n tensor = tf.transpose(tensor, (0, 2, 1, 3))\r\n # shape is now: (N_x, X, Y * N_y, c)\r\n tensor = tf.reshape(tensor, (1, w * num_tensor_x, h * num_tensor_y, c))\r\n # shape is now: (1, X * N_x, Y * N_y, c)\r\n tensor = tf.transpose(tensor, (0, 2, 1, 3))\r\n # shape is now: (1, Y * N_y, X * N_x, c)\r\n tensor = tf.pad(tensor, [[0, 0], [0, 1], [0, 1], [0, 0]], 'CONSTANT')\r\n\r\n self.image(name, tensor, **kwargs)","def from_numpy_to_tensor(record, device_id):\n [X, A,\n mol_ids_rep, rep_ids_rep, iw_ids,\n last_append_mask,\n NX, NX_rep,\n action_0, actions,\n log_p] = record\n\n X = nd.array(X, ctx=mx.gpu(device_id), dtype='int32')\n A_sparse = []\n for A_i in A:\n if A_i.shape[0] == 0:\n A_sparse.append(None)\n else:\n # transpose may not be supported in gpu\n A_i = np.concatenate([A_i, A_i[:, [1, 0]]], axis=0)\n\n # construct csr matrix ...\n data = np.ones((A_i.shape[0], ), dtype=np.float32)\n row, col = A_i[:, 0], A_i[:, 1]\n A_sparse_i = nd.sparse.csr_matrix((data, (row, col)),\n shape=tuple([int(X.shape[0]), ]*2),\n ctx=mx.gpu(device_id),\n dtype='float32')\n\n # append to list\n A_sparse.append(A_sparse_i)\n\n batch_size, iw_size = np.asscalar(mol_ids_rep.max() + 1), \\\n np.asscalar(rep_ids_rep.max() + 1)\n\n mol_ids_rep, rep_ids_rep, iw_ids, \\\n last_append_mask, \\\n NX, NX_rep, action_0, actions = [nd.array(_x, ctx=mx.gpu(device_id), dtype='int32')\n for _x in [mol_ids_rep, rep_ids_rep, iw_ids,\n last_append_mask,\n NX, NX_rep, action_0, actions]]\n\n log_p = nd.array(log_p, ctx=mx.gpu(device_id), dtype='float32')\n\n record = [X, A_sparse, iw_ids, last_append_mask,\n NX, NX_rep, action_0, actions, log_p,\n batch_size, iw_size]\n\n\n return record","def move(self):\n for agent in self.agents:\n if not agent.fidelity:\n options = agent.get_move_options(agent.hex, self.kernel_size, None, extend=True)\n target = random36.choices(population=options,weights=[x.quality**2 for x in options])\n agent.move(target[0])","def make_sharded_output_tensor(\n output: DTensor, _device_mesh: Optional[DeviceMesh] = None\n) -> torch.Tensor:\n\n return output.to_local() # type: ignore[call-arg]","def hard_update_target_network(self,step):\n \n if step % self.C == 0:\n pars = self.model.get_weights()\n self.target_model.set_weights(pars)","def hard_update_target_network(self,step):\n \n if step % self.C == 0:\n pars = self.model.get_weights()\n self.target_model.set_weights(pars)","def forward_tensor(self, x):\n pass","def _fuse_back(self):\n with tf.variable_scope('fusion'):\n #self.fuse_p_encodes, _ = rnn('lstm', self.match_p_encodes, self.p_length,\n # self.hidden_size, layer_num=1)\n self.fuse_p_encodes = tc.layers.fully_connected(self.match_p_encodes, self.hidden_size)\n if self.use_dropout:\n self.fuse_p_encodes = tf.nn.dropout(self.fuse_p_encodes, self.dropout_keep_prob)","def set_devices(sys_device_ids):\n # Set the CUDA_VISIBLE_DEVICES environment variable\n import os\n visible_devices = ''\n for i in sys_device_ids:\n visible_devices += '{}, '.format(i)\n os.environ['CUDA_VISIBLE_DEVICES'] = visible_devices\n # Return wrappers.\n # Models and user defined Variables/Tensors would be transferred to the\n # first device.\n device_id = 0 if len(sys_device_ids) > 0 else -1\n TVT = TransferVarTensor(device_id)\n TMO = TransferModulesOptims(device_id)\n return TVT, TMO","def _move_in_one_more_block():\n with tik_inst.for_range(0, sub_h_align_block_size) as sub_h_idx:\n tik_inst.data_move(dst[sub_w_block * data_cnt_one_block * sub_h_idx],\n src[w_offset + w_size * sub_h_idx], 0, 1, sub_w_block, 0, 0)\n # in order to avoid dirty data when multiple core\n with tik_inst.for_range(0, data_cnt_one_block) as sub_h_idx_1:\n tik_inst.data_move(dst[sub_w_block * data_cnt_one_block *\n (sub_h_align_block_size + sub_h_idx_1)],\n src[w_offset +\n w_size * (sub_h_size - data_cnt_one_block + sub_h_idx_1)],\n 0, 1, sub_w_block, 0, 0)","def to_device(device, x):\n if device is None:\n return x\n elif device < 0:\n return cuda.to_cpu(x)\n else:\n return cuda.to_gpu(x, device)","def set_devices_for_ml(sys_device_ids):\n import os\n\n all_ids = []\n for ids in sys_device_ids:\n all_ids += ids\n unique_sys_device_ids = list(set(all_ids))\n unique_sys_device_ids.sort()\n if -1 in unique_sys_device_ids:\n unique_sys_device_ids.remove(-1)\n\n # Set the CUDA_VISIBLE_DEVICES environment variable\n\n visible_devices = ''\n for i in unique_sys_device_ids:\n visible_devices += '{}, '.format(i)\n os.environ['CUDA_VISIBLE_DEVICES'] = visible_devices\n\n # Return wrappers\n\n relative_device_ids = []\n TVTs, TMOs = [], []\n for ids in sys_device_ids:\n relative_ids = []\n for id in ids:\n if id != -1:\n id = find_index(unique_sys_device_ids, id)\n relative_ids.append(id)\n relative_device_ids.append(relative_ids)\n\n # Models and user defined Variables/Tensors would be transferred to the\n # first device.\n TVTs.append(TransferVarTensor(relative_ids[0]))\n TMOs.append(TransferModulesOptims(relative_ids[0]))\n return TVTs, TMOs, relative_device_ids","def movetomiddle(pidevice, axes=None):\n if not isdeviceavailable([GCS2Commands, GCS21Commands], pidevice):\n raise TypeError('Type %s of pidevice is not supported!' % type(pidevice).__name__)\n\n axes = getaxeslist(pidevice, axes)\n if not axes:\n return\n rangemin = pidevice.qTMN(axes)\n rangemax = pidevice.qTMX(axes)\n targets = {}\n for axis in axes:\n targets[axis] = rangemin[axis] + (rangemax[axis] - rangemin[axis]) / 2.0\n pidevice.MOV(targets)","def set_default_tensor_type(device):\r\n if device in [torch.device(\"cpu\"), \"cpu\"]:\r\n torch.set_default_tensor_type(torch.FloatTensor)\r\n else:\r\n torch.set_default_tensor_type(torch.cuda.FloatTensor)","def test_setitem_rightvalue_ndarray_fails():\n a = numpy.arange(3 * 4 * 5)\n a.resize((3, 4, 5))\n a = theano._asarray(a, dtype='float32')\n _a = cuda_ndarray.CudaNdarray(a)\n\n b = theano._asarray([7, 8, 9, 10], dtype='float32')\n _b = cuda_ndarray.CudaNdarray(b)\n b5 = theano._asarray([7, 8, 9, 10, 11], dtype='float32')\n _b5 = cuda_ndarray.CudaNdarray(b)\n\n # attempt to assign the ndarray b with setitem\n _a[:, :, 1] = _b\n a[:, :, 1] = b\n assert numpy.allclose(numpy.asarray(_a), a)\n\n # test direct transfert from numpy to contiguous region\n # attempt to assign the ndarray b with setitem\n # same number of dim\n mat = numpy.random.rand(4, 5).astype('float32')\n _a[2, :, :] = mat\n a[2, :, :] = mat\n assert numpy.allclose(numpy.asarray(_a), a)\n\n # without same number of dim\n try:\n _a[0, :, :] = mat\n #a[0, :, :] = mat\n #assert numpy.allclose(numpy.asarray(_a), a)\n except ValueError as e:\n pass\n\n # test direct transfert from numpy with broadcast\n _a[0, :, :] = b5\n a[0, :, :] = b5\n assert numpy.allclose(numpy.asarray(_a), a)\n\n # test direct transfert from numpy to not contiguous region\n # attempt to assign the ndarray b with setitem\n _a[:, :, 2] = b\n a[:, :, 2] = b\n assert numpy.allclose(numpy.asarray(_a), a)","def update_target_net(self, sess):\n sess.run(self.update_target_net_op)","def to(self, device):\n self.detector.to(device)\n # self.recognizer.to(device)\n self.shared_conv.to(device)","def moveaxis(self, tensor, source, destination):\n\n axes = list(range(self.ndim(tensor)))\n if source < 0:\n source = axes[source]\n if destination < 0:\n destination = axes[destination]\n try:\n axes.pop(source)\n except IndexError:\n raise ValueError(\n \"Source should verify 0 <= source < tensor.ndim\" \"Got %d\" % source\n )\n try:\n axes.insert(destination, source)\n except IndexError:\n raise ValueError(\n \"Destination should verify 0 <= destination < tensor.ndim\"\n \"Got %d\" % destination\n )\n return self.transpose(tensor, axes)","def forward(model: nn.Module, inputs: torch.Tensor, device: torch.device):\n\n model.eval()\n model.to(device)\n\n with torch.no_grad():\n inputs = inputs.to(device)\n return model(inputs)","def make_output_reshard_tensor(\n output: DTensor,\n device_mesh: Optional[DeviceMesh] = None,\n) -> torch.Tensor:\n\n return make_output_shard_1d(output, device_mesh).to_local() # type: ignore[call-arg, attr-defined]","def partition(data, device, train_size=0.8):\n # Start with a copy, will be training\n train = copy.deepcopy(data)\n test = []\n\n # Get distribution of message indices, keep ordering\n test_len = int((1 - train_size) * len(data))\n test_indices = sorted(random.sample(range(len(data)), test_len), reverse=True)\n\n # For each index, remove from train and append to test\n for i in test_indices:\n test.append(train.pop(i))\n\n # Need to reverse test now\n test.reverse()\n\n # Now label each set individually (performed in place)\n Labeler(train)\n Labeler(test)\n\n # Rescale data as well\n train = DataRescaler(train).scaled_data\n test = DataRescaler(test).scaled_data\n\n # Convert to tensors\n # Inputs\n Xtrain = torch.tensor([[s.get('time')] + list(s.get('composite').values()) for s in train], dtype=torch.double).to(device)\n Xtest = torch.tensor([[s.get('time')] + list(s.get('composite').values()) for s in test], dtype=torch.double).to(device)\n\n # Targets\n Ttrain = torch.tensor([[s.get('distinct')] for s in train], dtype=torch.long).to(device)\n Ttest = torch.tensor([[s.get('distinct')] for s in test], dtype=torch.long).to(device)\n return(Xtrain, Ttrain, Xtest, Ttest)","def to(self, device) -> None:\n self.obs_buffer = self.obs_buffer.to(device)\n self.hid_buffer = self.hid_buffer.to(device)\n self.rew_buffer = self.rew_buffer.to(device)\n self.act_buffer = self.act_buffer.to(device)\n self.don_buffer = self.don_buffer.to(device)\n self.true_termin = self.true_termin.to(device)\n\n self.device = device","def forward(self, src, device):\n\n src = torch.as_tensor(src).float().to(device)\n\n\n # Set initial hidden and cell states \n h0 = torch.zeros(self.num_layers, src.shape[0], self.hidden_dim).to(device)\n c0 = torch.zeros(self.num_layers, src.shape[0], self.hidden_dim).to(device)\n\n # shape of lstm_out: [batch_size, input_size, hidden_dim]\n # shape of self.hidden: (a, b), where a and b both have shape (num_layers, batch_size, hidden_dim).\n lstm_out, self.hidden = self.lstm(src, (h0, c0)) \n \n # Only take the output from the final timetep\n # Can pass on the entirety of lstm_out to the next layer if it is a seq2seq prediction\n #print(lstm_out.size())\n y_pred = self.linear(lstm_out[:, -1, :].view(src.shape[0], -1))\n return y_pred","def assign_to_device(device, ps_device=None):\n ps_ops = ['Variable', 'VariableV2', 'AutoReloadVariable', 'MutableHashTable',\n 'MutableHashTableOfTensors', 'MutableDenseHashTable']\n\n def _assign(op):\n node_def = op if isinstance(op, tf.NodeDef) else op.node_def\n if node_def.op in ps_ops:\n return \"/\" + ps_device\n else:\n return device\n\n return _assign","def adjust_net_distribution(self):\r\n for net in self.nets:\r\n if self.block.name == \"A\": # \"A\" after move, so the cell moved to \"A\"\r\n net.cell_to_blockA(self)\r\n else:\r\n assert self.block.name == \"B\" # \"B\" after move, so the cell moved to \"B\"\r\n net.cell_to_blockB(self)","def set_device(device, backend='autograd'):\n if backend == 'autograd':\n return None\n elif backend == 'pytorch':\n try:\n tc.cuda.set_device(device)\n except:\n pass","def balanced_device_setter(op):\n if op.type in {'Variable', 'VariableV2', 'VarHandleOp'}:\n # return self._sync_device\n min_load = min(load.values())\n min_load_devices = [d for d in load if load[d] == min_load]\n chosen_device = random.choice(min_load_devices)\n load[chosen_device] += op.outputs[0].get_shape().num_elements()\n return chosen_device\n return device","def to(self, device):\n if device not in (\"cpu\", \"cuda\"):\n raise ValueError(\"Attribute device should be 'cpu' or 'cuda'.\")\n\n self.reader.model.to(device)\n self.reader.device = torch.device(device)\n return self","def coord(self, tensor: Union[Tensor, np.ndarray]) -> None:\n try:\n tensor = tensor.reshape(self.shape[0], 3)\n except (RuntimeError, ValueError): # for torch.Tensor and np.ndarray\n raise ValueError(f'got unexpected shape {tensor.shape}')\n if not isinstance(tensor, Tensor):\n tensor = self.tensor.new_tensor(tensor)\n self.tensor[:, :3] = tensor","def send(\n self,\n *location,\n inplace: bool = False,\n # local_autograd=False,\n # preinitialize_grad=False,\n no_wrap=False,\n garbage_collect_data=True,\n ):\n\n # If you send a pointer p1, you want the pointer to pointer p2\n # to control the garbage collection and not the remaining old\n # p1 (here self). Because if p2 is not GCed, GCing p1 shouldn't delete\n # the remote tensor, but if you want to do so, as p2 is not GCed,\n # you can still do `del p2`. This allows to chain multiple\n # .send().send() calls.\n\n if len(location) == 1:\n\n location = location[0]\n\n if hasattr(self, \"child\") and isinstance(\n self.child, PointerTensor\n ):\n self.child.garbage_collect_data = False\n\n ptr = self.owner.send(\n self, location, garbage_collect_data=garbage_collect_data\n )\n\n ptr.description = self.description\n ptr.tags = self.tags\n\n # The last pointer should control remote GC,\n # not the previous self.ptr\n if hasattr(self, \"ptr\") and self.ptr is not None:\n ptr_ = self.ptr()\n if ptr_ is not None:\n ptr_.garbage_collect_data = False\n\n # we need to cache this weak reference to the pointer so that\n # if this method gets called multiple times we can simply re-use\n # the same pointer which was previously created\n self.ptr = weakref.ref(ptr)\n\n if inplace:\n self.set_() # TODO[jason]: pretty sure this is torch specific\n self.child = ptr\n return self\n else:\n output = ptr if no_wrap else ptr.wrap(type=tf.constant, value=[])\n\n else:\n\n children = list()\n for loc in location:\n children.append(self.send(loc, no_wrap=True))\n\n output = syft.MultiPointerTensor(children=children)\n\n if not no_wrap:\n output = output.wrap(type=tf.constant, value=[])\n\n return output","def cli(ctx: click.Context,\n experiment: str,\n devices: List[int],\n ) -> None:\n f = EXPERIMENTS[experiment]\n try:\n model, B, C, _devices = f(devices)\n except ValueError as exc:\n # Examples:\n # ValueError: too few devices to hold given partitions (devices: 1, paritions: 2)\n ctx.fail(str(exc))\n\n optimizer = SGD(model.parameters(), lr=0.1)\n\n in_device = _devices[0]\n out_device = _devices[-1]\n torch.cuda.set_device(in_device)\n\n input = torch.rand(32, 3, 192, 192, device=in_device)\n target = torch.rand(32, 1, 192, 192, device=out_device)\n\n # HEADER ======================================================================================\n\n title = f'{experiment}, U-Net ({B}, {C})'\n click.echo(title)\n\n if isinstance(model, GPipe):\n click.echo(f'balance: {model.balance}')\n\n click.echo('torchgpipe: %s, python: %s, torch: %s, cudnn: %s, cuda: %s, gpu: %s' % (\n torchgpipe.__version__,\n platform.python_version(),\n torch.__version__,\n torch.backends.cudnn.version(),\n torch.version.cuda,\n torch.cuda.get_device_name(in_device)))\n\n hr()\n\n # PARAMETERS ==================================================================================\n\n param_count = sum(p.storage().size() for p in model.parameters())\n param_size = sum(p.storage().size() * p.storage().element_size() for p in model.parameters())\n param_scale = 2 # param + grad\n\n click.echo(f'# of Model Parameters: {param_count:,}')\n click.echo(f'Total Model Parameter Memory: {param_size*param_scale:,} Bytes')\n\n # ACTIVATIONS =================================================================================\n\n try:\n torch.cuda.empty_cache()\n for d in _devices:\n torch.cuda.reset_max_memory_cached(d)\n\n for _ in range(2):\n output = model(input)\n output = cast(Tensor, output)\n loss = F.binary_cross_entropy_with_logits(output, target)\n loss.backward()\n optimizer.step()\n\n max_memory = 0\n for d in _devices:\n torch.cuda.synchronize(d)\n max_memory += torch.cuda.max_memory_cached(d)\n\n latent_size = max_memory - param_size*param_scale\n click.echo(f'Peak Activation Memory: {latent_size:,} Bytes')\n click.echo(f'Total Memory: {max_memory:,} Bytes')\n\n # MAX MEMORY PER DEVICE =======================================================================\n\n finally:\n hr()\n\n for d in _devices:\n memory_usage = torch.cuda.memory_cached(d)\n click.echo(f'{d!s}: {memory_usage:,} Bytes')","def move(self, t, s):\n raise NotImplementedError","def _move(self, at, to):\r\n copy = self.copy()\r\n i, j = at\r\n r, c = to\r\n copy.board[i][j], copy.board[r][c] = copy.board[r][c], copy.board[i][j]\r\n return copy","def target_tensors(self):\n return None","def move_to(var, device):\n if isinstance(var, dict):\n return {k: move_to(v, device) for k, v in var.items()}\n elif isinstance(var, list):\n return [move_to(v, device) for v in var]\n elif isinstance(var, tuple):\n return (move_to(v, device) for v in var)\n return var.to(device)","def move(self, newPath):\n\t\tif self.hasUdim:\n\t\t\tfor a in self.udimPaths:\n\t\t\t\ta.move(newPath)\n\t\telse:\n\t\t\tsuper( textureFile, self ).move( newPath )","def __call__(self, A, B):\n assert len(A.shape) == len(B.shape)\n # A = A.contiguous()\n # B = B.contiguous()\n if len(A.shape) == 2 and len(B.shape) == 2:\n A = A[None]\n B = B[None]\n two_dimentional = True\n dim += 1\n elif len(A.shape) == 3 and len(B.shape) == 3:\n two_dimentional = False\n pass\n else:\n raise ValueError(\"A and B need to be 2d or 3d\")\n assert A.shape[0] == B.shape[0]\n assert A.shape[2] == B.shape[1]\n assert A.dtype == B.dtype\n assert A.dtype in [torch.float, torch.half]\n assert A.device.type == B.device.type == \"cuda\"\n\n mode = self.get_mode(A, B)\n\n if mode == \"nn\":\n kernel_fn = self._fn_nn\n elif mode == \"tt\":\n kernel_fn = self._fn_tt\n elif mode == \"tn\":\n kernel_fn = self._fn_tn\n elif mode == \"nt\":\n kernel_fn = self._fn_nt\n\n l, m, k = A.shape\n l, k, n = B.shape\n\n C = torch.zeros([l, m, n], device=A.device, dtype=A.dtype)\n\n threads_per_block = (256,)\n #blocks_per_grid = (math.ceil(n/128), math.ceil(m/128), l)\n \n n_ = math.ceil(n/(128*self.patch_n))\n m_ = math.ceil(m/(128*self.patch_m))\n blocks_per_grid = (self.patch_n*self.patch_m, n_ * m_, l)\n\n self._fn_nn(\n grid=blocks_per_grid,\n block=threads_per_block,\n args=[\n A.data_ptr(),\n B.data_ptr(),\n C.data_ptr(),\n m, n, k,\n ],\n stream=self.stream\n )\n\n if two_dimentional:\n C = C[0]\n return C","def mv_all(self):\n # def mv_step(self):\n self.device_reg_data &= ~(0x1 << 2)\n bus.write_byte_data(self.device_address, self.device_reg_mode1, self.device_reg_data)","def switch_to_untuned_inputs(self):\n\n self.h_e=self.inputs.noise_flat.T\n self.h=np.vstack([self.h_e,self.h_i])","def forward(self,bottom,top):\n # imgmaps = self.batch_loader.mixup_gen()\n # print(\"DataLayer forward!!\")\n trainX, trainY = self.batch_loader.batch_imgs()\n # print(\"trainX:\",trainX.shape)\n # print(\"trainY:\",trainY.shape)\n # print(\"trainY:\", trainY)\n # print(\"top[0].data.shape:\",top[0].data.shape)\n # print(\"top[1].data.shape:\", top[1].data.shape)\n top[0].data[:, ...] = trainX\n top[1].data[:, ...] = trainY\n # print(\"DataLayer forward!!\")","def test_auto_transfer_correct_device(ray_start_4_cpus_2_gpus):\n import nvidia_smi\n\n nvidia_smi.nvmlInit()\n\n def get_gpu_used_mem(i):\n handle = nvidia_smi.nvmlDeviceGetHandleByIndex(i)\n info = nvidia_smi.nvmlDeviceGetMemoryInfo(handle)\n return info.used\n\n start_gpu_memory = get_gpu_used_mem(1)\n\n device = torch.device(\"cuda:1\")\n small_dataloader = [(torch.randn((1024 * 4, 1024 * 4)),) for _ in range(10)]\n wrapped_dataloader = ( # noqa: F841\n ray.train.torch.train_loop_utils._WrappedDataLoader(\n small_dataloader, device, True\n )\n )\n\n end_gpu_memory = get_gpu_used_mem(1)\n\n # Verify GPU memory usage increases on the right cuda device\n assert end_gpu_memory > start_gpu_memory","def set_data(self, source, target):\n if self.use_cuda:\n self.source = source.cuda(async=True)\n self.target = target.cuda(async=True)\n else:\n self.source = source\n self.target = target","def _data_move_out_last_dim_lt_one_block(tik_inst, dst, src, data_pos_info):\n sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, out_offset = data_pos_info\n data_size_one_block = _get_elment_cnt_one_block(src.dtype)\n\n with tik_inst.if_scope(sub_axis_1 == 1):\n with tik_inst.if_scope(sub_axis_0 * axis_2 > data_size_one_block):\n sub_axis_0_2 = sub_axis_0 * axis_2\n sub_axis_0_block_align = sub_axis_0_2 // data_size_one_block\n left_data = sub_axis_0_2 % data_size_one_block\n tik_inst.data_move(dst[out_offset], src, 0, 1, sub_axis_0_block_align, 0, 0)\n with tik_inst.if_scope(left_data > 0):\n with tik_inst.new_stmt_scope():\n reg_temp = [tik_inst.Scalar(src.dtype)\n for i in ADDR_IDX_LIST[:data_size_one_block]]\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n reg_temp[idx].set_as(src[sub_axis_0_2 - data_size_one_block + idx])\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n src[idx].set_as(reg_temp[idx])\n tik_inst.data_move(dst[out_offset + sub_axis_0_2 - data_size_one_block],\n src, 0, 1, 1, 0, 0)\n with tik_inst.else_scope():\n # for case sub_axis_0 * axis_2 < data_size_one_block\n with tik_inst.if_scope(axis_0 * axis_1 * axis_2 > data_size_one_block):\n m_sub_axis_0 = _ceil_div(data_size_one_block, axis_2)\n sub_axis_0_2 = m_sub_axis_0 * axis_2\n sub_axis_0_block_align = sub_axis_0_2 // data_size_one_block\n left_data = sub_axis_0_2 % data_size_one_block\n tik_inst.data_move(dst[out_offset + (sub_axis_0 - m_sub_axis_0) * axis_2],\n src, 0, 1, sub_axis_0_block_align, 0, 0)\n with tik_inst.if_scope(left_data > 0):\n with tik_inst.new_stmt_scope():\n reg_temp = [tik_inst.Scalar(src.dtype)\n for i in ADDR_IDX_LIST[:data_size_one_block]]\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n reg_temp[idx].set_as(src[sub_axis_0_2 - data_size_one_block + idx])\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n src[idx].set_as(reg_temp[idx])\n tik_inst.data_move(dst[out_offset + sub_axis_0 * axis_2 -\n data_size_one_block],\n src, 0, 1, 1, 0, 0)\n with tik_inst.else_scope():\n # the shape size is not bigger than one block size\n tik_inst.data_move(dst[out_offset], src, 0, 1,\n _ceil_div(axis_0 * axis_1 * axis_2, data_size_one_block), 0, 0)\n\n with tik_inst.else_scope():\n with tik_inst.if_scope(sub_axis_1 * axis_0 * axis_2 > data_size_one_block):\n sub_axis_1_0_2 = sub_axis_1 * axis_0 * axis_2\n sub_axis_1_0_2_block_align = sub_axis_1_0_2 // data_size_one_block\n left_data = sub_axis_1_0_2 % data_size_one_block\n tik_inst.data_move(dst[out_offset], src, 0, 1, sub_axis_1_0_2_block_align, 0, 0)\n with tik_inst.if_scope(left_data > 0):\n with tik_inst.new_stmt_scope():\n reg_temp = [tik_inst.Scalar(src.dtype)\n for i in ADDR_IDX_LIST[:data_size_one_block]]\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n reg_temp[idx].set_as(src[sub_axis_1_0_2 - data_size_one_block + idx])\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n src[idx].set_as(reg_temp[idx])\n tik_inst.data_move(dst[out_offset + sub_axis_1_0_2 - data_size_one_block],\n src, 0, 1, 1, 0, 0)\n with tik_inst.else_scope():\n # to make sure move in data size is not less than one block\n with tik_inst.if_scope(axis_0 * axis_1 * axis_2 > data_size_one_block):\n m_sub_axis_1 = _ceil_div(data_size_one_block, axis_0 * axis_2)\n sub_axis_1_0_2 = m_sub_axis_1 * axis_0 * axis_2\n sub_axis_0_block_align = sub_axis_1_0_2 // data_size_one_block\n left_data = sub_axis_1_0_2 % data_size_one_block\n tik_inst.data_move(dst[out_offset + (sub_axis_1 - m_sub_axis_1) * axis_0 * axis_2],\n src, 0, 1, sub_axis_0_block_align, 0, 0)\n with tik_inst.if_scope(left_data > 0):\n with tik_inst.new_stmt_scope():\n reg_temp = [tik_inst.Scalar(src.dtype)\n for i in ADDR_IDX_LIST[:data_size_one_block]]\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n reg_temp[idx].set_as(src[sub_axis_1_0_2 - data_size_one_block + idx])\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\n src[idx].set_as(reg_temp[idx])\n tik_inst.data_move(dst[out_offset + sub_axis_1 * axis_0 * axis_2 -\n data_size_one_block],\n src, 0, 1, 1, 0, 0)\n with tik_inst.else_scope():\n # the shape size is not bigger than one block size\n tik_inst.data_move(dst[out_offset], src, 0, 1,\n _ceil_div(axis_0 * axis_1 * axis_2, data_size_one_block), 0, 0)"],"string":"[\n \"def move_to_device(input_tensors: List[torch.Tensor],\\n target_device: Optional[torch.device],\\n non_blocking: bool = False) -> Iterable[torch.Tensor]:\\n return (tensor if tensor.device == target_device or target_device is None\\n else tensor.to(target_device, non_blocking=non_blocking)\\n for tensor in input_tensors)\",\n \"def move2gpu(self):\\n if self.generator.backend.upper() == \\\"LOOPY\\\":\\n try:\\n import pyopencl as cl\\n import pyopencl.array # pylint: disable=unused-variable\\n from .context import queue\\n except ImportError:\\n raise ImportError(\\\"Please install loo.py\\\")\\n\\n self.rhs = cl.array.to_device(queue, self.rhs)\\n if hasattr(self, \\\"s\\\"):\\n self.s = cl.array.to_device(\\n queue, self.s\\n ) # pylint: disable=attribute-defined-outside-init\\n self.istore = cl.array.to_device(queue, self.istore)\\n for i in range(len(self.iload)):\\n self.iload[i] = cl.array.to_device(queue, self.iload[i])\",\n \"def to(self, device):\\n for item in self.data:\\n if torch.is_tensor(item):\\n item.to(item)\\n else:\\n for subitem in item:\\n subitem.to(device)\\n return self\",\n \"def move_to_device(obj, cuda_device: int):\\n\\n if cuda_device < 0 or not has_tensor(obj):\\n return obj\\n elif isinstance(obj, torch.Tensor):\\n return obj.cuda(cuda_device)\\n elif isinstance(obj, dict):\\n return {key: move_to_device(value, cuda_device) for key, value in obj.items()}\\n elif isinstance(obj, list):\\n return [move_to_device(item, cuda_device) for item in obj]\\n elif isinstance(obj, tuple) and hasattr(obj, \\\"_fields\\\"):\\n # This is the best way to detect a NamedTuple, it turns out.\\n return obj.__class__(*(move_to_device(item, cuda_device) for item in obj))\\n elif isinstance(obj, tuple):\\n return tuple(move_to_device(item, cuda_device) for item in obj)\\n else:\\n return obj\",\n \"def move_and_process_input(batch):\\n x, y = batch\\n x = x.to(device).float()\\n y = torch.as_tensor(y).to(device)\\n x = x.permute(0, -1, 1, 2, 3)\\n return x, y\",\n \"def batch_to_device(batch):\\n for key in batch:\\n if isinstance(batch[key], torch.Tensor):\\n batch[key] = batch[key].to(device)\\n return batch\",\n \"def to_device(data, device):\\n if isinstance(data, (list,tuple)): # allows to apply function to lists or tuples of tensors\\n return [to_device(x, device) for x in data]\\n return data.to(device, non_blocking=True)\",\n \"def deploy_to_device(self):\\n if self.device_ids is not None and len(self.device_ids) > 1:\\n if not isinstance(self.model, torch.nn.DataParallel):\\n self.model = torch.nn.DataParallel(self.model, self.device_ids)\\n\\n self.model = self.model.to(self.device)\\n self.criterion = self.criterion.to(self.device)\",\n \"def set_device(self, device: torch.Tensor) -> None:\\n raise NotImplementedError\",\n \"def _send(x, dst=0):\\n x = torch.tensor(x)\\n x = to_device(x)\\n dist.send(x, dst)\\n del x \\n torch.cuda.empty_cache()\",\n \"def _batch_to_device(batch, target_device):\\n tensor = _getattr(\\\"torch\\\", \\\"Tensor\\\")\\n for key in batch:\\n if isinstance(batch[key], tensor):\\n batch[key] = batch[key].to(target_device)\\n return batch\",\n \"def device_placement(self):\\n if is_tf_available():\\n import tensorflow as tf\\n with tf.device('/CPU:0' if self.device == -1 else '/device:GPU:{}'.format(self.device)):\\n yield\\n else:\\n import torch\\n if self.device >= 0:\\n torch.cuda.set_device(self.device)\\n\\n yield\",\n \"def __move_sample_to_device(self, sample):\\n _in, _target = sample\\n _in = _in.to(self.device).float()\\n _target = _target.to(self.device).long()\\n\\n return _in, _target\",\n \"def move_devices(self, data):\\n data = clean(data, self.move_parameters)\\n return self.put(\\\"/devices/move\\\", data)\",\n \"def switch_to_tuned_inputs(self):\\n \\n self.h_e=self.inputs_flat.T\\n self.h=np.vstack([self.h_e,self.h_i])\",\n \"def to_device(self, move_optim_data=True):\\n if move_optim_data:\\n self.optimizer_values = self.optimizer_values.to(\\n self._embeddings.weight.device\\n )\",\n \"def train(self):\\n if len(self.buffer) >= self.batch_size:\\n with torch.no_grad():\\n states, actions, rewards, next_states, dones = self.buffer.sample(self.batch_size)\\n\\n # Send data to GPU\\n states = torch.stack(states).to(self.device, dtype=torch.float)\\n actions = torch.stack(actions).to(self.device, dtype=torch.float)\\n rewards = torch.stack(rewards).to(self.device, dtype=torch.float)\\n rewards = torch.reshape(rewards, (self.batch_size, 1))\\n\\n next_states = torch.stack(next_states).to(self.device, dtype=torch.float)\\n dones = torch.stack(dones).to(self.device, dtype=torch.float)\\n\\n #TODO\\n\\n # Calculate target Q values using the Target Network\\n selection = torch.argmax(self.main_dqn(next_states), dim = 1).unsqueeze(1)\\n\\n evaluation = self.target_dqn(next_states)\\n evaluation = evaluation.gather(1, selection.long()) #size [256,1]\\n\\n #Create Done mask\\n nonzero_indices = torch.nonzero(dones).reshape(-1).tolist()\\n dones_mask = torch.eye(self.batch_size)\\n for index in nonzero_indices:\\n dones_mask[index,index] = 0\\n dones_mask = dones_mask.to(self.device, dtype=torch.float)\\n\\n # Calculte target\\n target = rewards + torch.matmul(dones_mask, evaluation*self.gamma)\\n target = target.detach()\\n\\n # Calculate Q values using the Main Network\\n if self.env.freely_moving:\\n n_classes = self.env.number_of_action_channels * self.env.number_of_rows * self.env.number_of_columns\\n else:\\n n_classes = self.env.number_of_action_channels * 1 * self.env.nA\\n\\n n_samples = self.batch_size\\n labels = torch.flatten(actions.type(torch.LongTensor), start_dim=0)\\n labels_tensor = torch.as_tensor(labels)\\n action_masks = torch.nn.functional.one_hot(labels_tensor, num_classes=n_classes).to(self.device, dtype=torch.float)\\n\\n q_value = action_masks * self.main_dqn(states)\\n q_value = torch.sum(q_value, dim=-1).reshape((self.batch_size, 1))\\n\\n # Calculate loss\\n loss = self.mse(target, q_value)\\n\\n # Optimize the model\\n self.optimizer.zero_grad()\\n loss.backward()\\n torch.nn.utils.clip_grad_norm_(self.main_dqn.parameters(), 5)\\n self.optimizer.step()\\n\\n # Soft Copy the Main Network's weights to the Target Network\\n self.soft_update_of_target_network(self.main_dqn, self.target_dqn,tau=1e-3)\\n\\n return loss\\n return 0\",\n \"def change_device(self, device=None):\\n\\n if device is None:\\n # If the function is called without a device, use the current device\\n device = self.device\\n\\n # Create the appropriate device object\\n device = torch.device(f'cuda:{device}'\\n if torch.cuda.is_available() else 'cpu')\\n\\n # Change device field\\n self.device = device\\n # Load the transcription model onto the device\\n self.to(self.device)\",\n \"def move(model, origin, dest):\\n model.move(origin, dest)\",\n \"def move(model, origin, dest):\\n model.move(origin, dest)\",\n \"def to_device(model, device):\\n p = next(model.parameters())\\n if p.device == device:\\n return\\n model.to(device)\",\n \"def to_device(m: torch.nn.Module, x:torch.Tensor):\\n if isinstance(m, torch.nn.Module):\\n device = next(m.parameters()).device\\n elif isinstance(m, torch.Tensor):\\n device = m.device\\n else:\\n raise TypeError(\\n \\\"Expected torch.nn.Module or torch.tensor, \\\" f\\\"bot got: {type(m)}\\\"\\n )\\n return x.to(device)\",\n \"def _move_module_to_device(\\n self,\\n module: nn.Module,\\n ignored_params: Set[nn.Parameter],\\n device_from_device_id: Optional[torch.device],\\n ):\\n cpu_device = torch.device(\\\"cpu\\\")\\n param = next(self._get_orig_params(module, ignored_params), None)\\n if param is None:\\n return # no original parameters to manage\\n if device_from_device_id is not None:\\n if param.device == cpu_device:\\n # NOTE: This includes moving ignored modules' parameters.\\n module = module.to(device_from_device_id)\\n # TODO: This is a temporary fix to move already- constructed\\n # `FlatParameter`s back to CPU if needed. This is needed to\\n # make CPU offload work with `device_id`.\\n for submodule in module.modules():\\n if (\\n isinstance(submodule, FullyShardedDataParallel)\\n and submodule.cpu_offload.offload_params\\n ):\\n with torch.no_grad():\\n for handle in submodule._handles:\\n handle._flat_param_to(torch.device(\\\"cpu\\\"))\\n elif param.device == cpu_device:\\n warnings.warn(\\n \\\"Module is put on CPU and will thus have flattening and sharding\\\"\\n \\\" run on CPU, which is less efficient than on GPU. We recommend passing in \\\"\\n \\\"`device_id` argument which will enable FSDP to put module on GPU device,\\\"\\n \\\" module must also be on GPU device to work with `sync_module_states=True` flag\\\"\\n \\\" which requires GPU communication.\\\"\\n )\",\n \"def swap_xy(tensor: tf.Tensor) -> tf.Tensor:\\n return moveaxis(tensor, -2, -1)\",\n \"def place_on_device(x, device):\\n if isinstance(x, list):\\n return [place_on_device(xi, device) for xi in x]\\n elif isinstance(x, dict):\\n return {k: place_on_device(v, device) for k, v in x.items()}\\n elif isinstance(x, set):\\n return {place_on_device(xi, device) for xi in x}\\n elif isinstance(x, tuple):\\n return tuple(place_on_device(xi, device) for xi in x)\\n elif isinstance(x, torch.Tensor):\\n return x\\n else:\\n return x\",\n \"def forward(self, observation: Tensor) -> Tensor:\\n pass\",\n \"def set_src_dst_tensor(self, tik_instance):\\n src_element_number = functools_reduce(lambda x1, x2: x1 * x2,\\n self.src_shape_ori[:])\\n dst_element_number = functools_reduce(lambda x1, x2: x1 * x2,\\n self.dst_shape[:])\\n self.src_gm = tik_instance.Tensor(self.dtype,\\n (src_element_number,),\\n name=\\\"src_gm\\\",\\n scope=tik.scope_gm)\\n self.dst_gm = tik_instance.Tensor(self.dtype,\\n (dst_element_number,),\\n name=\\\"dst_gm\\\",\\n scope=tik.scope_gm)\",\n \"def set_src_dst_tensor(self, tik_instance):\\n src_element_number = functools_reduce(lambda x1, x2: x1 * x2,\\n self.src_shape_ori[:])\\n dst_element_number = functools_reduce(lambda x1, x2: x1 * x2,\\n self.dst_shape[:])\\n self.src_gm = tik_instance.Tensor(self.dtype,\\n (src_element_number,),\\n name=\\\"src_gm\\\",\\n scope=tik.scope_gm)\\n self.dst_gm = tik_instance.Tensor(self.dtype,\\n (dst_element_number,),\\n name=\\\"dst_gm\\\",\\n scope=tik.scope_gm)\",\n \"def set_src_dst_tensor(self, tik_instance):\\n src_element_number = functools_reduce(lambda x, y: x * y,\\n self.src_shape[:])\\n dst_element_number = functools_reduce(lambda x, y: x * y,\\n self.dst_shape[:])\\n self.src_gm = tik_instance.Tensor(self.dtype,\\n (src_element_number,),\\n name=\\\"src_gm\\\",\\n scope=tik.scope_gm)\\n self.dst_gm = tik_instance.Tensor(self.dtype,\\n (dst_element_number,),\\n name=\\\"dst_gm\\\",\\n scope=tik.scope_gm)\",\n \"def set_device(sys_device_id):\\n device_id = -1\\n cuda = (sys_device_id != -1)\\n if cuda:\\n # CUDA_VISIBLE_DEVICE is a list, and device_id is the index of its members.\\n import os\\n os.environ['CUDA_VISIBLE_DEVICES'] = str(sys_device_id)\\n device_id = 0\\n TVT = TransferVarTensor(device_id)\\n TMO = TransferModulesOptims(device_id)\\n return TVT, TMO\",\n \"def distribute_tensor(\\n tensor: torch.Tensor,\\n device_mesh: Optional[DeviceMesh] = None,\\n placements: Optional[Sequence[Placement]] = None,\\n) -> DTensor:\\n\\n torch._C._log_api_usage_once(\\\"torch.dtensor.distribute_tensor\\\")\\n\\n # get default device mesh if there's nothing specified\\n device_mesh = device_mesh or mesh_resources.get_current_mesh()\\n\\n # instantiate a RNG tracker if haven't. By default DTensor uses an\\n # OffsetBasedRNGTracker to perform random operators.\\n # TODO: the value assignment to global variable is not the ideal solution\\n # we can replace it in future.\\n if is_rng_supported_mesh(device_mesh) and not random._rng_tracker:\\n random._rng_tracker = OffsetBasedRNGTracker()\\n\\n if not tensor.is_leaf:\\n raise RuntimeError(\\n \\\"`distribute_tensor` should be used to distribute leaf tensors! but found non-leaf tensor!\\\"\\n )\\n\\n # convert tensor to the corresponding device type if it's not in that device type\\n if not tensor.is_meta:\\n tensor = tensor.to(device_mesh.device_type)\\n # set default placements to replicated if not specified\\n if placements is None:\\n placements = [Replicate() for _ in range(device_mesh.ndim)]\\n\\n if len(placements) != device_mesh.ndim:\\n raise ValueError(\\n f\\\"`placements` must have the same length as `device_mesh.ndim`! \\\"\\n f\\\"Found placements length: {len(placements)}, and device_mesh.ndim: {device_mesh.ndim}.\\\"\\n )\\n\\n if isinstance(tensor, DTensor):\\n # if the tensor is already a DTensor, we just need to check if the\\n # device mesh and placements are the same\\n if tensor.device_mesh != device_mesh:\\n raise ValueError(\\n f\\\"Cannot distribute a DTensor with device mesh {tensor.device_mesh} \\\"\\n f\\\"to a different device mesh {device_mesh}.\\\"\\n )\\n if tensor.placements != placements:\\n raise ValueError(\\n f\\\"Cannot distribute a DTensor with placements {tensor.placements} \\\"\\n f\\\"to a different placements {placements}. do you want to call \\\"\\n f\\\"`redistribute` instead?\\\"\\n )\\n return tensor\\n\\n local_tensor = tensor\\n\\n # distribute the tensor according to the placements.\\n for idx, placement in enumerate(placements):\\n if placement.is_shard():\\n placement = cast(Shard, placement)\\n output = placement._shard_tensor(local_tensor, device_mesh, idx)\\n # scatter call could not return a tensor with correct requires_grad\\n # field, as ProcessGroupNCCL refuse to take a tensor with requires_grad\\n # to do inplace update! So we manually set it here\\n output.requires_grad_(tensor.requires_grad)\\n local_tensor = output\\n elif placement.is_replicate():\\n placement = cast(Replicate, placement)\\n local_tensor = placement._replicate_tensor(local_tensor, device_mesh, idx)\\n else:\\n raise RuntimeError(\\n f\\\"Trying to distribute tensor with unsupported placements {placement} on device mesh dimension {idx}!\\\"\\n )\\n\\n assert local_tensor is not None, \\\"distributing a tensor should not be None\\\"\\n # detach the local tensor passed to DTensor since after the construction\\n # of DTensor, autograd would work on top of DTensor instead of local tensor\\n return DTensor(\\n local_tensor.detach(),\\n device_mesh,\\n placements,\\n shape=tensor.size(),\\n dtype=tensor.dtype,\\n requires_grad=tensor.requires_grad,\\n stride=tensor.stride(),\\n )\",\n \"def batch_to_device(batch, target_device: device):\\n moved_batch = {}\\n for key, val in batch.items():\\n val = val.to(target_device)\\n moved_batch[key] = val\\n return moved_batch\",\n \"def copy_tensor(src, dest):\\n dest.copy_(src)\",\n \"def to(self, device):\\n self.device = device\\n self.model.to(self.device)\",\n \"def _recursive_to_cuda(self, tensors: Union[Tuple[torch.Tensor], torch.Tensor]) \\\\\\n -> Union[List[torch.Tensor], torch.Tensor]:\\n if self.device is None: # keep on cpu\\n return tensors\\n\\n if type(tensors) != list and type(tensors) != tuple: # not only for torch.Tensor\\n return tensors.to(device=self.device)\\n\\n cuda_tensors = list()\\n for i in range(len(tensors)):\\n cuda_tensors.append(self._recursive_to_cuda(tensors[i]))\\n return cuda_tensors\",\n \"def module_transfer_to_device(self) -> None:\\n for name, module in self.modules.items():\\n module.to(self.device)\\n if self.device.type == 'cuda':\\n self.modules[name] = torch.nn.DataParallel(module, self.gpu_ids)\\n return\",\n \"def forward(self, bottom, top):\\n blobs = self._get_next_minibatch() # Get a blob\\n\\n for blob_id, blob in enumerate(blobs): # blob[0] for image, blob[1] for mask\\n # Reshape net's input blobs\\n top[blob_id].reshape(*(blob.shape)) # Reshape top tensor/blob\\n top[blob_id].data[...] = blob.astype(np.float32, copy=False) # Feed corresponding data into top tensor/blob\",\n \"def mv_step(self):\\n # def mv_all(self):\\n self.device_reg_data &= ~(0x1 << 3)\\n bus.write_byte_data(self.device_address, self.device_reg_mode1, self.device_reg_data)\",\n \"def _eunn_loop(state, capacity, diag_vec_list, off_vec_list, diag, fft):\\n i = 0\\n def layer_tunable(x, i):\\n\\n diag_vec = diag_vec_list.read(i)\\n off_vec = off_vec_list.read(i)\\n\\n diag = tf.multiply(x, diag_vec)\\n off = tf.multiply(x, off_vec)\\n\\n def even_input(off, size):\\n\\n def even_s(off, size):\\n off = tf.reshape(off, [-1, size//2, 2])\\n off = tf.reshape(tf.reverse(off, [2]), [-1, size])\\n return off\\n\\n def odd_s(off, size):\\n off, helper = tf.split(off, [size-1, 1], 1)\\n size -= 1\\n off = even_s(off, size)\\n off = tf.concat([off, helper], 1)\\n return off\\n\\n off = tf.cond(tf.equal(tf.mod(size, 2), 0), lambda: even_s(off, size), lambda: odd_s(off, size))\\n return off\\n\\n def odd_input(off, size):\\n helper, off = tf.split(off, [1, size-1], 1)\\n size -= 1\\n off = even_input(off, size)\\n off = tf.concat([helper, off], 1)\\n return off\\n\\n size = int(off.get_shape()[1])\\n off = tf.cond(tf.equal(tf.mod(i, 2), 0), lambda: even_input(off, size), lambda: odd_input(off, size))\\n\\n layer_output = diag + off\\n i += 1\\n\\n return layer_output, i\\n\\n def layer_fft(state, i):\\n\\n diag_vec = diag_vec_list.read(i)\\n off_vec = off_vec_list.read(i)\\n diag = tf.multiply(state, diag_vec)\\n off = tf.multiply(state, off_vec)\\n\\n hidden_size = int(off.get_shape()[1])\\n # size = 2**i\\n dist = capacity - i\\n normal_size = (hidden_size // (2**dist)) * (2**(dist-1))\\n normal_size *= 2\\n extra_size = tf.maximum(0, (hidden_size % (2**dist)) - (2**(dist-1)))\\n hidden_size -= normal_size\\n\\n def modify(off_normal, dist, normal_size):\\n off_normal = tf.reshape(tf.reverse(tf.reshape(off_normal, [-1, normal_size//(2**dist), 2, (2**(dist-1))]), [2]), [-1, normal_size])\\n return off_normal\\n\\n def do_nothing(off_normal):\\n return off_normal\\n\\n off_normal, off_extra = tf.split(off, [normal_size, hidden_size], 1)\\n off_normal = tf.cond(tf.equal(normal_size, 0), lambda: do_nothing(off_normal), lambda: modify(off_normal, dist, normal_size))\\n helper1, helper2 = tf.split(off_extra, [hidden_size-extra_size, extra_size], 1)\\n off_extra = tf.concat([helper2, helper1], 1)\\n off = tf.concat([off_normal, off_extra], 1)\\n\\n layer_output = diag + off\\n i += 1\\n\\n return layer_output, i\\n\\n if fft:\\n layer_function = layer_fft\\n else:\\n layer_function = layer_tunable\\n output, _ = tf.while_loop(lambda state, i: tf.less(i, capacity), layer_function, [state, i])\\n\\n if not diag is None:\\n output = tf.multiply(output, diag)\\n\\n\\n return output\",\n \"def set_devices(args):\\n global devices\\n if args is not None:\\n devices = [torch.device(i) for i in ast.literal_eval('[' + args + ']')]\\n torch.cuda.set_device(devices[0])\\n else:\\n devices = [torch.device('cpu')]\",\n \"def forward_batch(model, batch, device):\\n inputs, labels = (batch, batch)\\n inputs, labels = inputs.to(device), labels.to(device)\\n\\n outputs = model(inputs, labels=labels)\\n\\n return outputs[:2]\",\n \"def _update(self):\\n self.parametrize_beam()\\n self.update_ranks()\\n self._points = tf.reshape(self._endpoint, (1, 2)) * tf.reshape(self._ranks, (-1, 1))\",\n \"def to_device(self, device):\\n for i in range(self.num_layers):\\n getattr(self, \\\"conv{}\\\".format(i+1)).to_device(device)\\n self.to(device)\\n return self\",\n \"def update_target_network(self):\\n variables = self.online_network.trainable_variables\\n variables_copy = [tf.Variable(v) for v in variables]\\n self.target_network.trainable_variables = variables_copy\",\n \"def forward_target(self, img, **kwargs):\\n assert img.dim() == 5, \\\\\\n \\\"Input must have 5 dims, got: {}\\\".format(img.dim())\\n\\n img_v_l = []\\n idx_unshuffle_v_l = []\\n for idx in range(img.shape[1]):\\n img_vi = img[:, idx, ...].contiguous()\\n img_vi, idx_unshuffle_vi = self._batch_shuffle_ddp(img_vi)\\n img_v_l.append(img_vi)\\n idx_unshuffle_v_l.append(idx_unshuffle_vi)\\n\\n if self.start_block > 0:\\n for idx in range(img.shape[1]):\\n for i, best_path in enumerate(self.best_paths):\\n img_v_l[idx] = self.target_backbone(img_v_l[idx],\\n start_block=i,\\n forward_op=best_path,\\n block_op=True)[0]\\n\\n self.forward_op_target = self.forward_op_online\\n proj_target_v1 = 0\\n proj_target_v2 = 0\\n v2_idx = img.shape[1]//2\\n with torch.no_grad():\\n for op_idx, forward_singleop_target in enumerate(self.forward_op_target):\\n temp_v1 = self.target_neck(self.target_backbone(img_v_l[op_idx],\\n start_block=self.start_block,\\n forward_op=forward_singleop_target))[\\n 0].clone().detach()\\n temp_v2 = self.target_neck(self.target_backbone(img_v_l[v2_idx + op_idx],\\n start_block=self.start_block,\\n forward_op=forward_singleop_target))[\\n 0].clone().detach()\\n temp_v1 = nn.functional.normalize(temp_v1, dim=1)\\n temp_v1 = self._batch_unshuffle_ddp(temp_v1, idx_unshuffle_v_l[op_idx])\\n\\n temp_v2 = nn.functional.normalize(temp_v2, dim=1)\\n temp_v2 = self._batch_unshuffle_ddp(temp_v2, idx_unshuffle_v_l[v2_idx + op_idx])\\n\\n proj_target_v1 += temp_v1\\n proj_target_v2 += temp_v2\\n\\n self.proj_target_v1 = proj_target_v1 / (len(self.forward_op_target))\\n self.proj_target_v2 = proj_target_v2 / (len(self.forward_op_target))\",\n \"def forward(self, in_tensors: List[Tensor], out_tensors: List[Tensor]):\\n pass\",\n \"def forward(self, in_tensors: List[Tensor], out_tensors: List[Tensor]):\\n pass\",\n \"def moveaxis(tensor: tf.Tensor, source: int, destination: int) -> tf.Tensor:\\n ndim = len(tensor.shape)\\n source = _normalize_axis(source, ndim)\\n destination = _normalize_axis(destination, ndim)\\n order = [n for n in range(ndim) if n != source]\\n order.insert(destination, source)\\n return tf.transpose(tensor, order)\",\n \"def move_to(self, T):\\n self.camera_coords = coordinates.Coordinates(\\n pos=np.array(T),\\n rot=self.camera_coords.worldrot())\\n pybullet.resetBasePositionAndOrientation(\\n self.camera_id,\\n self.camera_coords.worldpos(),\\n coordinates.math.wxyz2xyzw(\\n coordinates.math.matrix2quaternion(\\n self.camera_coords.worldrot())))\",\n \"def move(self):\\n self.pos += self.direc\\n self.nearest_node = self.pixel_to_node()\",\n \"def move_data_to_device(self, data: Tuple) -> Tuple:\\n tmp = []\\n for dv in data:\\n tmp.append(dv.to(self.device))\\n \\n return tuple(tmp)\",\n \"def restore_leading_dims(tensors, lead_dim, T=1, B=1):\\n is_seq = isinstance(tensors, (tuple, list))\\n tensors = tensors if is_seq else (tensors,)\\n if lead_dim == 2: # (Put T dim.)\\n tensors = tuple(t.view((T, B) + t.shape[1:]) for t in tensors)\\n if lead_dim == 0: # (Remove B=1 dim.)\\n assert B == 1\\n tensors = tuple(t.squeeze(0) for t in tensors)\\n return tensors if is_seq else tensors[0]\",\n \"def sample_batch(pid, args, batch_queue, port_dict, device, actor_id_to_ip_dataport, local_size, cache_array):\\n def recv_data(k, data_stream, actor_set, real_data_tasks_i):\\n for real_data in data_stream:\\n tmp = []\\n tmp.append(real_data.state)\\n tmp.append(real_data.action)\\n tmp.append(real_data.reward)\\n tmp.append(real_data.next_state)\\n tmp.append(real_data.done)\\n tmp.append(actor_set[k]['w'][real_data.idx])\\n tmp.append(actor_set[k]['i'][real_data.idx])\\n tmp.append(actor_set[k]['t'][real_data.idx])\\n tmp.append(real_data.timestamp)\\n local_dict[actor_set[k]['i'][real_data.idx]] = tmp\\n cache_array[actor_set[k]['i'][real_data.idx]] |= 2**pid\\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.state), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n real_data_tasks_i['states'].append(decom_state) #.to(device))\\n real_data_tasks_i['actions'].append(torch.LongTensor([real_data.action])) #.to(device))\\n real_data_tasks_i['rewards'].append(torch.FloatTensor([real_data.reward])) #.to(device))\\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.next_state), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n real_data_tasks_i['next_states'].append(decom_next_state) #.to(device))\\n real_data_tasks_i['dones'].append(torch.FloatTensor([real_data.done])) #.to(device))\\n real_data_tasks_i['batch_weights'].append(torch.FloatTensor([actor_set[k]['w'][real_data.idx]])) #.to(device))\\n real_data_tasks_i['batch_idxes'].append(actor_set[k]['i'][real_data.idx])\\n # is the data overwrited?\\n real_data_tasks_i['batch_timestamp_store'].append(actor_set[k]['t'][real_data.idx])\\n real_data_tasks_i['batch_timestamp_real'].append(real_data.timestamp)\\n conn = grpc.insecure_channel(port_dict['replay_ip'] + ':' + port_dict['sampleDataPort'])\\n client = apex_data_pb2_grpc.SampleDataStub(channel=conn)\\n local_dict = {}\\n while True:\\n batch_timestamp_real = []\\n batch_timestamp_store = []\\n batch_weights = []\\n batch_idxes = []\\n\\n states, actions, rewards, next_states, dones = [], [], [], [], []\\n\\n res_batch = client.Send(apex_data_pb2.SampleDataRequest(batch_size=args.batch_size, beta = args.beta))\\n actor_ids, data_ids, timestamps, weights, idxes = res_batch.actor_ids, res_batch.data_ids, res_batch.timestamp, res_batch.weights, res_batch.idxes\\n actor_set = {}\\n cached_value = {'states':{},'actions':{},'rewards':{},'next_states':{},'dones':{},'batch_weights':{},'batch_idxes':{},'batch_timestamp_store':{},'batch_timestamp_real':{}}\\n for i in range(len(actor_ids)):\\n set_a = actor_set.get(actor_ids[i], False)\\n if set_a == False:\\n actor_set[actor_ids[i]] = {}\\n set_a = actor_set[actor_ids[i]]\\n set_a['d'] = []\\n set_a['w'] = []\\n set_a['i'] = []\\n set_a['t'] = []\\n cached_value['states'][actor_ids[i]] = []\\n cached_value['actions'][actor_ids[i]] = []\\n cached_value['rewards'][actor_ids[i]] = []\\n cached_value['next_states'][actor_ids[i]] = []\\n cached_value['dones'][actor_ids[i]] = []\\n cached_value['batch_weights'][actor_ids[i]] = []\\n cached_value['batch_idxes'][actor_ids[i]] = []\\n cached_value['batch_timestamp_store'][actor_ids[i]] = []\\n cached_value['batch_timestamp_real'][actor_ids[i]] = []\\n cache_id = actor_ids[i]*local_size+data_ids[i]\\n cache_trans = cache_array[cache_id]\\n if cache_trans & 2**pid == 0:\\n set_a['d'].append(data_ids[i])\\n set_a['w'].append(weights[i])\\n set_a['i'].append(idxes[i])\\n set_a['t'].append(timestamps[i])\\n if cache_trans == 0 and local_dict.get(cache_id, False) != False:\\n del local_dict[cache_id]\\n else:\\n try:\\n state_tmp = local_dict[cache_id][0]\\n action_tmp = local_dict[cache_id][1]\\n reward_tmp = local_dict[cache_id][2] \\n next_state_tmp = local_dict[cache_id][3] \\n done_tmp = local_dict[cache_id][4] \\n batch_weight_tmp = local_dict[cache_id][5] \\n batch_idx_tmp = local_dict[cache_id][6] \\n batch_store_tmp = local_dict[cache_id][7] \\n batch_real_tmp = local_dict[cache_id][8] \\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n cached_value['states'][actor_ids[i]].append(decom_state)\\n cached_value['actions'][actor_ids[i]].append(torch.LongTensor([action_tmp]))\\n cached_value['rewards'][actor_ids[i]].append(torch.FloatTensor([reward_tmp]))\\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(next_state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n cached_value['next_states'][actor_ids[i]].append(decom_next_state)\\n cached_value['dones'][actor_ids[i]].append(torch.FloatTensor([done_tmp]))\\n cached_value['batch_weights'][actor_ids[i]].append(torch.FloatTensor([batch_weight_tmp]))\\n cached_value['batch_idxes'][actor_ids[i]].append(batch_idx_tmp)\\n cached_value['batch_timestamp_store'][actor_ids[i]].append(batch_store_tmp)\\n cached_value['batch_timestamp_real'][actor_ids[i]].append(batch_real_tmp)\\n except:\\n set_a['d'].append(data_ids[i])\\n set_a['w'].append(weights[i])\\n set_a['i'].append(idxes[i])\\n set_a['t'].append(timestamps[i])\\n real_data_links = {}\\n real_data_tasks = {}\\n for k, v in actor_set.items():\\n actor_ip, data_port = actor_id_to_ip_dataport[k]\\n conn_actor = grpc.insecure_channel(actor_ip + ':' + data_port)\\n client_actor = apex_data_pb2_grpc.SendRealDataStub(channel=conn_actor)\\n real_data_links[k] = client_actor.Send(apex_data_pb2.RealBatchRequest(idxes=v['d']))\\n real_data_tasks[k] = {}\\n real_data_tasks[k]['states'] = cached_value['states'][k]\\n real_data_tasks[k]['actions'] = cached_value['actions'][k]\\n real_data_tasks[k]['rewards'] = cached_value['rewards'][k]\\n real_data_tasks[k]['next_states'] = cached_value['next_states'][k]\\n real_data_tasks[k]['dones'] = cached_value['dones'][k]\\n real_data_tasks[k]['batch_weights'] = cached_value['batch_weights'][k]\\n real_data_tasks[k]['batch_idxes'] = cached_value['batch_idxes'][k]\\n real_data_tasks[k]['batch_timestamp_store'] = cached_value['batch_timestamp_store'][k]\\n real_data_tasks[k]['batch_timestamp_real'] = cached_value['batch_timestamp_real'][k]\\n threads = []\\n for k, v in real_data_links.items():\\n t = threading.Thread(target=recv_data, args=(k, v, actor_set, real_data_tasks[k],))\\n threads.append(t)\\n t.start()\\n\\n for t in threads:\\n t.join()\\n\\n for k, v in real_data_tasks.items():\\n states += v['states']\\n actions += v['actions']\\n rewards += v['rewards']\\n next_states += v['next_states']\\n dones += v['dones']\\n batch_weights += v['batch_weights']\\n batch_idxes += v['batch_idxes']\\n batch_timestamp_real += v['batch_timestamp_real']\\n batch_timestamp_store += v['batch_timestamp_store']\\n\\n states = torch.cat(states,0).to(device)\\n actions = torch.cat(actions,0).to(device)\\n rewards = torch.cat(rewards,0).to(device)\\n next_states = torch.cat(next_states,0).to(device)\\n dones = torch.cat(dones,0).to(device)\\n batch_weights = torch.cat(batch_weights,0).to(device)\\n\\n batch = [states, actions, rewards, next_states, dones, batch_weights, batch_idxes]\\n batch_queue.put(batch)\\n data, batch = None, None\",\n \"def sync_target_network(self):\\n for t, e in zip(\\n self.target_network.trainable_variables, self.online_network.trainable_variables\\n ):\\n t.assign(e)\",\n \"def _hadamard_step(x, dim):\\n x_shape = x.shape.as_list()\\n x = tf.reshape(x, [-1, 2]) # Reshape so that we have a matrix.\\n x = tf.matmul(x, h_core) # Multiply.\\n x = tf.reshape(x, [-1, dim // 2, 2]) # Reshape to rank-3.\\n x = tf.transpose(x, perm=permutation) # Swap last two dimensions.\\n x.set_shape(x_shape) # Failed shape inference in tf.while_loop.\\n return x\",\n \"def _copy_to_gpu(self):\\n self.dispatch('on_texture')\",\n \"def cuda(self, device=None) -> MultiIndicesEmbedding:\\n super().cuda(device=device)\\n\\n self.offsets = self.offsets.cuda(device)\\n\\n return self\",\n \"def device(self):\\n return self._tensor.device\",\n \"def _4d_tensor(self, name, tensor, **kwargs):\\r\\n h, w, c, num_tensor = tensor.shape.as_list()\\r\\n\\r\\n # Try to visualise convolutional filters or feature maps\\r\\n # See: https://gist.github.com/kukuruza/03731dc494603ceab0c5\\r\\n # input shape: (Y, X, C, N)\\r\\n if c != 1 and c != 3:\\r\\n tensor = tf.reduce_mean(tensor, axis=2, keep_dims=True)\\r\\n c = 1\\r\\n # shape is now: (Y, X, 1|C, N)\\r\\n v_min = tf.reduce_min(tensor)\\r\\n v_max = tf.reduce_max(tensor)\\r\\n tensor -= v_min\\r\\n tensor *= 1.0 / (v_max - v_min)\\r\\n tensor = tf.pad(tensor, [[1, 0], [1, 0], [0, 0], [0, 0]], 'CONSTANT')\\r\\n tensor = tf.transpose(tensor, perm=(3, 0, 1, 2))\\r\\n # shape is now: (N, Y, X, C)\\r\\n # place tensor on grid\\r\\n num_tensor_x = int(np.round(np.sqrt(num_tensor)))\\r\\n num_tensor_y = num_tensor / num_tensor_x\\r\\n while not num_tensor_y.is_integer():\\r\\n num_tensor_x += 1\\r\\n num_tensor_y = num_tensor / num_tensor_x\\r\\n num_tensor_y = int(num_tensor_y)\\r\\n h += 1\\r\\n w += 1\\r\\n tensor = tf.reshape(tensor, (num_tensor_x, h * num_tensor_y, w, c))\\r\\n # shape is now: (N_x, Y * N_y, X, c)\\r\\n tensor = tf.transpose(tensor, (0, 2, 1, 3))\\r\\n # shape is now: (N_x, X, Y * N_y, c)\\r\\n tensor = tf.reshape(tensor, (1, w * num_tensor_x, h * num_tensor_y, c))\\r\\n # shape is now: (1, X * N_x, Y * N_y, c)\\r\\n tensor = tf.transpose(tensor, (0, 2, 1, 3))\\r\\n # shape is now: (1, Y * N_y, X * N_x, c)\\r\\n tensor = tf.pad(tensor, [[0, 0], [0, 1], [0, 1], [0, 0]], 'CONSTANT')\\r\\n\\r\\n self.image(name, tensor, **kwargs)\",\n \"def from_numpy_to_tensor(record, device_id):\\n [X, A,\\n mol_ids_rep, rep_ids_rep, iw_ids,\\n last_append_mask,\\n NX, NX_rep,\\n action_0, actions,\\n log_p] = record\\n\\n X = nd.array(X, ctx=mx.gpu(device_id), dtype='int32')\\n A_sparse = []\\n for A_i in A:\\n if A_i.shape[0] == 0:\\n A_sparse.append(None)\\n else:\\n # transpose may not be supported in gpu\\n A_i = np.concatenate([A_i, A_i[:, [1, 0]]], axis=0)\\n\\n # construct csr matrix ...\\n data = np.ones((A_i.shape[0], ), dtype=np.float32)\\n row, col = A_i[:, 0], A_i[:, 1]\\n A_sparse_i = nd.sparse.csr_matrix((data, (row, col)),\\n shape=tuple([int(X.shape[0]), ]*2),\\n ctx=mx.gpu(device_id),\\n dtype='float32')\\n\\n # append to list\\n A_sparse.append(A_sparse_i)\\n\\n batch_size, iw_size = np.asscalar(mol_ids_rep.max() + 1), \\\\\\n np.asscalar(rep_ids_rep.max() + 1)\\n\\n mol_ids_rep, rep_ids_rep, iw_ids, \\\\\\n last_append_mask, \\\\\\n NX, NX_rep, action_0, actions = [nd.array(_x, ctx=mx.gpu(device_id), dtype='int32')\\n for _x in [mol_ids_rep, rep_ids_rep, iw_ids,\\n last_append_mask,\\n NX, NX_rep, action_0, actions]]\\n\\n log_p = nd.array(log_p, ctx=mx.gpu(device_id), dtype='float32')\\n\\n record = [X, A_sparse, iw_ids, last_append_mask,\\n NX, NX_rep, action_0, actions, log_p,\\n batch_size, iw_size]\\n\\n\\n return record\",\n \"def move(self):\\n for agent in self.agents:\\n if not agent.fidelity:\\n options = agent.get_move_options(agent.hex, self.kernel_size, None, extend=True)\\n target = random36.choices(population=options,weights=[x.quality**2 for x in options])\\n agent.move(target[0])\",\n \"def make_sharded_output_tensor(\\n output: DTensor, _device_mesh: Optional[DeviceMesh] = None\\n) -> torch.Tensor:\\n\\n return output.to_local() # type: ignore[call-arg]\",\n \"def hard_update_target_network(self,step):\\n \\n if step % self.C == 0:\\n pars = self.model.get_weights()\\n self.target_model.set_weights(pars)\",\n \"def hard_update_target_network(self,step):\\n \\n if step % self.C == 0:\\n pars = self.model.get_weights()\\n self.target_model.set_weights(pars)\",\n \"def forward_tensor(self, x):\\n pass\",\n \"def _fuse_back(self):\\n with tf.variable_scope('fusion'):\\n #self.fuse_p_encodes, _ = rnn('lstm', self.match_p_encodes, self.p_length,\\n # self.hidden_size, layer_num=1)\\n self.fuse_p_encodes = tc.layers.fully_connected(self.match_p_encodes, self.hidden_size)\\n if self.use_dropout:\\n self.fuse_p_encodes = tf.nn.dropout(self.fuse_p_encodes, self.dropout_keep_prob)\",\n \"def set_devices(sys_device_ids):\\n # Set the CUDA_VISIBLE_DEVICES environment variable\\n import os\\n visible_devices = ''\\n for i in sys_device_ids:\\n visible_devices += '{}, '.format(i)\\n os.environ['CUDA_VISIBLE_DEVICES'] = visible_devices\\n # Return wrappers.\\n # Models and user defined Variables/Tensors would be transferred to the\\n # first device.\\n device_id = 0 if len(sys_device_ids) > 0 else -1\\n TVT = TransferVarTensor(device_id)\\n TMO = TransferModulesOptims(device_id)\\n return TVT, TMO\",\n \"def _move_in_one_more_block():\\n with tik_inst.for_range(0, sub_h_align_block_size) as sub_h_idx:\\n tik_inst.data_move(dst[sub_w_block * data_cnt_one_block * sub_h_idx],\\n src[w_offset + w_size * sub_h_idx], 0, 1, sub_w_block, 0, 0)\\n # in order to avoid dirty data when multiple core\\n with tik_inst.for_range(0, data_cnt_one_block) as sub_h_idx_1:\\n tik_inst.data_move(dst[sub_w_block * data_cnt_one_block *\\n (sub_h_align_block_size + sub_h_idx_1)],\\n src[w_offset +\\n w_size * (sub_h_size - data_cnt_one_block + sub_h_idx_1)],\\n 0, 1, sub_w_block, 0, 0)\",\n \"def to_device(device, x):\\n if device is None:\\n return x\\n elif device < 0:\\n return cuda.to_cpu(x)\\n else:\\n return cuda.to_gpu(x, device)\",\n \"def set_devices_for_ml(sys_device_ids):\\n import os\\n\\n all_ids = []\\n for ids in sys_device_ids:\\n all_ids += ids\\n unique_sys_device_ids = list(set(all_ids))\\n unique_sys_device_ids.sort()\\n if -1 in unique_sys_device_ids:\\n unique_sys_device_ids.remove(-1)\\n\\n # Set the CUDA_VISIBLE_DEVICES environment variable\\n\\n visible_devices = ''\\n for i in unique_sys_device_ids:\\n visible_devices += '{}, '.format(i)\\n os.environ['CUDA_VISIBLE_DEVICES'] = visible_devices\\n\\n # Return wrappers\\n\\n relative_device_ids = []\\n TVTs, TMOs = [], []\\n for ids in sys_device_ids:\\n relative_ids = []\\n for id in ids:\\n if id != -1:\\n id = find_index(unique_sys_device_ids, id)\\n relative_ids.append(id)\\n relative_device_ids.append(relative_ids)\\n\\n # Models and user defined Variables/Tensors would be transferred to the\\n # first device.\\n TVTs.append(TransferVarTensor(relative_ids[0]))\\n TMOs.append(TransferModulesOptims(relative_ids[0]))\\n return TVTs, TMOs, relative_device_ids\",\n \"def movetomiddle(pidevice, axes=None):\\n if not isdeviceavailable([GCS2Commands, GCS21Commands], pidevice):\\n raise TypeError('Type %s of pidevice is not supported!' % type(pidevice).__name__)\\n\\n axes = getaxeslist(pidevice, axes)\\n if not axes:\\n return\\n rangemin = pidevice.qTMN(axes)\\n rangemax = pidevice.qTMX(axes)\\n targets = {}\\n for axis in axes:\\n targets[axis] = rangemin[axis] + (rangemax[axis] - rangemin[axis]) / 2.0\\n pidevice.MOV(targets)\",\n \"def set_default_tensor_type(device):\\r\\n if device in [torch.device(\\\"cpu\\\"), \\\"cpu\\\"]:\\r\\n torch.set_default_tensor_type(torch.FloatTensor)\\r\\n else:\\r\\n torch.set_default_tensor_type(torch.cuda.FloatTensor)\",\n \"def test_setitem_rightvalue_ndarray_fails():\\n a = numpy.arange(3 * 4 * 5)\\n a.resize((3, 4, 5))\\n a = theano._asarray(a, dtype='float32')\\n _a = cuda_ndarray.CudaNdarray(a)\\n\\n b = theano._asarray([7, 8, 9, 10], dtype='float32')\\n _b = cuda_ndarray.CudaNdarray(b)\\n b5 = theano._asarray([7, 8, 9, 10, 11], dtype='float32')\\n _b5 = cuda_ndarray.CudaNdarray(b)\\n\\n # attempt to assign the ndarray b with setitem\\n _a[:, :, 1] = _b\\n a[:, :, 1] = b\\n assert numpy.allclose(numpy.asarray(_a), a)\\n\\n # test direct transfert from numpy to contiguous region\\n # attempt to assign the ndarray b with setitem\\n # same number of dim\\n mat = numpy.random.rand(4, 5).astype('float32')\\n _a[2, :, :] = mat\\n a[2, :, :] = mat\\n assert numpy.allclose(numpy.asarray(_a), a)\\n\\n # without same number of dim\\n try:\\n _a[0, :, :] = mat\\n #a[0, :, :] = mat\\n #assert numpy.allclose(numpy.asarray(_a), a)\\n except ValueError as e:\\n pass\\n\\n # test direct transfert from numpy with broadcast\\n _a[0, :, :] = b5\\n a[0, :, :] = b5\\n assert numpy.allclose(numpy.asarray(_a), a)\\n\\n # test direct transfert from numpy to not contiguous region\\n # attempt to assign the ndarray b with setitem\\n _a[:, :, 2] = b\\n a[:, :, 2] = b\\n assert numpy.allclose(numpy.asarray(_a), a)\",\n \"def update_target_net(self, sess):\\n sess.run(self.update_target_net_op)\",\n \"def to(self, device):\\n self.detector.to(device)\\n # self.recognizer.to(device)\\n self.shared_conv.to(device)\",\n \"def moveaxis(self, tensor, source, destination):\\n\\n axes = list(range(self.ndim(tensor)))\\n if source < 0:\\n source = axes[source]\\n if destination < 0:\\n destination = axes[destination]\\n try:\\n axes.pop(source)\\n except IndexError:\\n raise ValueError(\\n \\\"Source should verify 0 <= source < tensor.ndim\\\" \\\"Got %d\\\" % source\\n )\\n try:\\n axes.insert(destination, source)\\n except IndexError:\\n raise ValueError(\\n \\\"Destination should verify 0 <= destination < tensor.ndim\\\"\\n \\\"Got %d\\\" % destination\\n )\\n return self.transpose(tensor, axes)\",\n \"def forward(model: nn.Module, inputs: torch.Tensor, device: torch.device):\\n\\n model.eval()\\n model.to(device)\\n\\n with torch.no_grad():\\n inputs = inputs.to(device)\\n return model(inputs)\",\n \"def make_output_reshard_tensor(\\n output: DTensor,\\n device_mesh: Optional[DeviceMesh] = None,\\n) -> torch.Tensor:\\n\\n return make_output_shard_1d(output, device_mesh).to_local() # type: ignore[call-arg, attr-defined]\",\n \"def partition(data, device, train_size=0.8):\\n # Start with a copy, will be training\\n train = copy.deepcopy(data)\\n test = []\\n\\n # Get distribution of message indices, keep ordering\\n test_len = int((1 - train_size) * len(data))\\n test_indices = sorted(random.sample(range(len(data)), test_len), reverse=True)\\n\\n # For each index, remove from train and append to test\\n for i in test_indices:\\n test.append(train.pop(i))\\n\\n # Need to reverse test now\\n test.reverse()\\n\\n # Now label each set individually (performed in place)\\n Labeler(train)\\n Labeler(test)\\n\\n # Rescale data as well\\n train = DataRescaler(train).scaled_data\\n test = DataRescaler(test).scaled_data\\n\\n # Convert to tensors\\n # Inputs\\n Xtrain = torch.tensor([[s.get('time')] + list(s.get('composite').values()) for s in train], dtype=torch.double).to(device)\\n Xtest = torch.tensor([[s.get('time')] + list(s.get('composite').values()) for s in test], dtype=torch.double).to(device)\\n\\n # Targets\\n Ttrain = torch.tensor([[s.get('distinct')] for s in train], dtype=torch.long).to(device)\\n Ttest = torch.tensor([[s.get('distinct')] for s in test], dtype=torch.long).to(device)\\n return(Xtrain, Ttrain, Xtest, Ttest)\",\n \"def to(self, device) -> None:\\n self.obs_buffer = self.obs_buffer.to(device)\\n self.hid_buffer = self.hid_buffer.to(device)\\n self.rew_buffer = self.rew_buffer.to(device)\\n self.act_buffer = self.act_buffer.to(device)\\n self.don_buffer = self.don_buffer.to(device)\\n self.true_termin = self.true_termin.to(device)\\n\\n self.device = device\",\n \"def forward(self, src, device):\\n\\n src = torch.as_tensor(src).float().to(device)\\n\\n\\n # Set initial hidden and cell states \\n h0 = torch.zeros(self.num_layers, src.shape[0], self.hidden_dim).to(device)\\n c0 = torch.zeros(self.num_layers, src.shape[0], self.hidden_dim).to(device)\\n\\n # shape of lstm_out: [batch_size, input_size, hidden_dim]\\n # shape of self.hidden: (a, b), where a and b both have shape (num_layers, batch_size, hidden_dim).\\n lstm_out, self.hidden = self.lstm(src, (h0, c0)) \\n \\n # Only take the output from the final timetep\\n # Can pass on the entirety of lstm_out to the next layer if it is a seq2seq prediction\\n #print(lstm_out.size())\\n y_pred = self.linear(lstm_out[:, -1, :].view(src.shape[0], -1))\\n return y_pred\",\n \"def assign_to_device(device, ps_device=None):\\n ps_ops = ['Variable', 'VariableV2', 'AutoReloadVariable', 'MutableHashTable',\\n 'MutableHashTableOfTensors', 'MutableDenseHashTable']\\n\\n def _assign(op):\\n node_def = op if isinstance(op, tf.NodeDef) else op.node_def\\n if node_def.op in ps_ops:\\n return \\\"/\\\" + ps_device\\n else:\\n return device\\n\\n return _assign\",\n \"def adjust_net_distribution(self):\\r\\n for net in self.nets:\\r\\n if self.block.name == \\\"A\\\": # \\\"A\\\" after move, so the cell moved to \\\"A\\\"\\r\\n net.cell_to_blockA(self)\\r\\n else:\\r\\n assert self.block.name == \\\"B\\\" # \\\"B\\\" after move, so the cell moved to \\\"B\\\"\\r\\n net.cell_to_blockB(self)\",\n \"def set_device(device, backend='autograd'):\\n if backend == 'autograd':\\n return None\\n elif backend == 'pytorch':\\n try:\\n tc.cuda.set_device(device)\\n except:\\n pass\",\n \"def balanced_device_setter(op):\\n if op.type in {'Variable', 'VariableV2', 'VarHandleOp'}:\\n # return self._sync_device\\n min_load = min(load.values())\\n min_load_devices = [d for d in load if load[d] == min_load]\\n chosen_device = random.choice(min_load_devices)\\n load[chosen_device] += op.outputs[0].get_shape().num_elements()\\n return chosen_device\\n return device\",\n \"def to(self, device):\\n if device not in (\\\"cpu\\\", \\\"cuda\\\"):\\n raise ValueError(\\\"Attribute device should be 'cpu' or 'cuda'.\\\")\\n\\n self.reader.model.to(device)\\n self.reader.device = torch.device(device)\\n return self\",\n \"def coord(self, tensor: Union[Tensor, np.ndarray]) -> None:\\n try:\\n tensor = tensor.reshape(self.shape[0], 3)\\n except (RuntimeError, ValueError): # for torch.Tensor and np.ndarray\\n raise ValueError(f'got unexpected shape {tensor.shape}')\\n if not isinstance(tensor, Tensor):\\n tensor = self.tensor.new_tensor(tensor)\\n self.tensor[:, :3] = tensor\",\n \"def send(\\n self,\\n *location,\\n inplace: bool = False,\\n # local_autograd=False,\\n # preinitialize_grad=False,\\n no_wrap=False,\\n garbage_collect_data=True,\\n ):\\n\\n # If you send a pointer p1, you want the pointer to pointer p2\\n # to control the garbage collection and not the remaining old\\n # p1 (here self). Because if p2 is not GCed, GCing p1 shouldn't delete\\n # the remote tensor, but if you want to do so, as p2 is not GCed,\\n # you can still do `del p2`. This allows to chain multiple\\n # .send().send() calls.\\n\\n if len(location) == 1:\\n\\n location = location[0]\\n\\n if hasattr(self, \\\"child\\\") and isinstance(\\n self.child, PointerTensor\\n ):\\n self.child.garbage_collect_data = False\\n\\n ptr = self.owner.send(\\n self, location, garbage_collect_data=garbage_collect_data\\n )\\n\\n ptr.description = self.description\\n ptr.tags = self.tags\\n\\n # The last pointer should control remote GC,\\n # not the previous self.ptr\\n if hasattr(self, \\\"ptr\\\") and self.ptr is not None:\\n ptr_ = self.ptr()\\n if ptr_ is not None:\\n ptr_.garbage_collect_data = False\\n\\n # we need to cache this weak reference to the pointer so that\\n # if this method gets called multiple times we can simply re-use\\n # the same pointer which was previously created\\n self.ptr = weakref.ref(ptr)\\n\\n if inplace:\\n self.set_() # TODO[jason]: pretty sure this is torch specific\\n self.child = ptr\\n return self\\n else:\\n output = ptr if no_wrap else ptr.wrap(type=tf.constant, value=[])\\n\\n else:\\n\\n children = list()\\n for loc in location:\\n children.append(self.send(loc, no_wrap=True))\\n\\n output = syft.MultiPointerTensor(children=children)\\n\\n if not no_wrap:\\n output = output.wrap(type=tf.constant, value=[])\\n\\n return output\",\n \"def cli(ctx: click.Context,\\n experiment: str,\\n devices: List[int],\\n ) -> None:\\n f = EXPERIMENTS[experiment]\\n try:\\n model, B, C, _devices = f(devices)\\n except ValueError as exc:\\n # Examples:\\n # ValueError: too few devices to hold given partitions (devices: 1, paritions: 2)\\n ctx.fail(str(exc))\\n\\n optimizer = SGD(model.parameters(), lr=0.1)\\n\\n in_device = _devices[0]\\n out_device = _devices[-1]\\n torch.cuda.set_device(in_device)\\n\\n input = torch.rand(32, 3, 192, 192, device=in_device)\\n target = torch.rand(32, 1, 192, 192, device=out_device)\\n\\n # HEADER ======================================================================================\\n\\n title = f'{experiment}, U-Net ({B}, {C})'\\n click.echo(title)\\n\\n if isinstance(model, GPipe):\\n click.echo(f'balance: {model.balance}')\\n\\n click.echo('torchgpipe: %s, python: %s, torch: %s, cudnn: %s, cuda: %s, gpu: %s' % (\\n torchgpipe.__version__,\\n platform.python_version(),\\n torch.__version__,\\n torch.backends.cudnn.version(),\\n torch.version.cuda,\\n torch.cuda.get_device_name(in_device)))\\n\\n hr()\\n\\n # PARAMETERS ==================================================================================\\n\\n param_count = sum(p.storage().size() for p in model.parameters())\\n param_size = sum(p.storage().size() * p.storage().element_size() for p in model.parameters())\\n param_scale = 2 # param + grad\\n\\n click.echo(f'# of Model Parameters: {param_count:,}')\\n click.echo(f'Total Model Parameter Memory: {param_size*param_scale:,} Bytes')\\n\\n # ACTIVATIONS =================================================================================\\n\\n try:\\n torch.cuda.empty_cache()\\n for d in _devices:\\n torch.cuda.reset_max_memory_cached(d)\\n\\n for _ in range(2):\\n output = model(input)\\n output = cast(Tensor, output)\\n loss = F.binary_cross_entropy_with_logits(output, target)\\n loss.backward()\\n optimizer.step()\\n\\n max_memory = 0\\n for d in _devices:\\n torch.cuda.synchronize(d)\\n max_memory += torch.cuda.max_memory_cached(d)\\n\\n latent_size = max_memory - param_size*param_scale\\n click.echo(f'Peak Activation Memory: {latent_size:,} Bytes')\\n click.echo(f'Total Memory: {max_memory:,} Bytes')\\n\\n # MAX MEMORY PER DEVICE =======================================================================\\n\\n finally:\\n hr()\\n\\n for d in _devices:\\n memory_usage = torch.cuda.memory_cached(d)\\n click.echo(f'{d!s}: {memory_usage:,} Bytes')\",\n \"def move(self, t, s):\\n raise NotImplementedError\",\n \"def _move(self, at, to):\\r\\n copy = self.copy()\\r\\n i, j = at\\r\\n r, c = to\\r\\n copy.board[i][j], copy.board[r][c] = copy.board[r][c], copy.board[i][j]\\r\\n return copy\",\n \"def target_tensors(self):\\n return None\",\n \"def move_to(var, device):\\n if isinstance(var, dict):\\n return {k: move_to(v, device) for k, v in var.items()}\\n elif isinstance(var, list):\\n return [move_to(v, device) for v in var]\\n elif isinstance(var, tuple):\\n return (move_to(v, device) for v in var)\\n return var.to(device)\",\n \"def move(self, newPath):\\n\\t\\tif self.hasUdim:\\n\\t\\t\\tfor a in self.udimPaths:\\n\\t\\t\\t\\ta.move(newPath)\\n\\t\\telse:\\n\\t\\t\\tsuper( textureFile, self ).move( newPath )\",\n \"def __call__(self, A, B):\\n assert len(A.shape) == len(B.shape)\\n # A = A.contiguous()\\n # B = B.contiguous()\\n if len(A.shape) == 2 and len(B.shape) == 2:\\n A = A[None]\\n B = B[None]\\n two_dimentional = True\\n dim += 1\\n elif len(A.shape) == 3 and len(B.shape) == 3:\\n two_dimentional = False\\n pass\\n else:\\n raise ValueError(\\\"A and B need to be 2d or 3d\\\")\\n assert A.shape[0] == B.shape[0]\\n assert A.shape[2] == B.shape[1]\\n assert A.dtype == B.dtype\\n assert A.dtype in [torch.float, torch.half]\\n assert A.device.type == B.device.type == \\\"cuda\\\"\\n\\n mode = self.get_mode(A, B)\\n\\n if mode == \\\"nn\\\":\\n kernel_fn = self._fn_nn\\n elif mode == \\\"tt\\\":\\n kernel_fn = self._fn_tt\\n elif mode == \\\"tn\\\":\\n kernel_fn = self._fn_tn\\n elif mode == \\\"nt\\\":\\n kernel_fn = self._fn_nt\\n\\n l, m, k = A.shape\\n l, k, n = B.shape\\n\\n C = torch.zeros([l, m, n], device=A.device, dtype=A.dtype)\\n\\n threads_per_block = (256,)\\n #blocks_per_grid = (math.ceil(n/128), math.ceil(m/128), l)\\n \\n n_ = math.ceil(n/(128*self.patch_n))\\n m_ = math.ceil(m/(128*self.patch_m))\\n blocks_per_grid = (self.patch_n*self.patch_m, n_ * m_, l)\\n\\n self._fn_nn(\\n grid=blocks_per_grid,\\n block=threads_per_block,\\n args=[\\n A.data_ptr(),\\n B.data_ptr(),\\n C.data_ptr(),\\n m, n, k,\\n ],\\n stream=self.stream\\n )\\n\\n if two_dimentional:\\n C = C[0]\\n return C\",\n \"def mv_all(self):\\n # def mv_step(self):\\n self.device_reg_data &= ~(0x1 << 2)\\n bus.write_byte_data(self.device_address, self.device_reg_mode1, self.device_reg_data)\",\n \"def switch_to_untuned_inputs(self):\\n\\n self.h_e=self.inputs.noise_flat.T\\n self.h=np.vstack([self.h_e,self.h_i])\",\n \"def forward(self,bottom,top):\\n # imgmaps = self.batch_loader.mixup_gen()\\n # print(\\\"DataLayer forward!!\\\")\\n trainX, trainY = self.batch_loader.batch_imgs()\\n # print(\\\"trainX:\\\",trainX.shape)\\n # print(\\\"trainY:\\\",trainY.shape)\\n # print(\\\"trainY:\\\", trainY)\\n # print(\\\"top[0].data.shape:\\\",top[0].data.shape)\\n # print(\\\"top[1].data.shape:\\\", top[1].data.shape)\\n top[0].data[:, ...] = trainX\\n top[1].data[:, ...] = trainY\\n # print(\\\"DataLayer forward!!\\\")\",\n \"def test_auto_transfer_correct_device(ray_start_4_cpus_2_gpus):\\n import nvidia_smi\\n\\n nvidia_smi.nvmlInit()\\n\\n def get_gpu_used_mem(i):\\n handle = nvidia_smi.nvmlDeviceGetHandleByIndex(i)\\n info = nvidia_smi.nvmlDeviceGetMemoryInfo(handle)\\n return info.used\\n\\n start_gpu_memory = get_gpu_used_mem(1)\\n\\n device = torch.device(\\\"cuda:1\\\")\\n small_dataloader = [(torch.randn((1024 * 4, 1024 * 4)),) for _ in range(10)]\\n wrapped_dataloader = ( # noqa: F841\\n ray.train.torch.train_loop_utils._WrappedDataLoader(\\n small_dataloader, device, True\\n )\\n )\\n\\n end_gpu_memory = get_gpu_used_mem(1)\\n\\n # Verify GPU memory usage increases on the right cuda device\\n assert end_gpu_memory > start_gpu_memory\",\n \"def set_data(self, source, target):\\n if self.use_cuda:\\n self.source = source.cuda(async=True)\\n self.target = target.cuda(async=True)\\n else:\\n self.source = source\\n self.target = target\",\n \"def _data_move_out_last_dim_lt_one_block(tik_inst, dst, src, data_pos_info):\\n sub_axis_1, sub_axis_0, axis_0, axis_1, axis_2, out_offset = data_pos_info\\n data_size_one_block = _get_elment_cnt_one_block(src.dtype)\\n\\n with tik_inst.if_scope(sub_axis_1 == 1):\\n with tik_inst.if_scope(sub_axis_0 * axis_2 > data_size_one_block):\\n sub_axis_0_2 = sub_axis_0 * axis_2\\n sub_axis_0_block_align = sub_axis_0_2 // data_size_one_block\\n left_data = sub_axis_0_2 % data_size_one_block\\n tik_inst.data_move(dst[out_offset], src, 0, 1, sub_axis_0_block_align, 0, 0)\\n with tik_inst.if_scope(left_data > 0):\\n with tik_inst.new_stmt_scope():\\n reg_temp = [tik_inst.Scalar(src.dtype)\\n for i in ADDR_IDX_LIST[:data_size_one_block]]\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n reg_temp[idx].set_as(src[sub_axis_0_2 - data_size_one_block + idx])\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n src[idx].set_as(reg_temp[idx])\\n tik_inst.data_move(dst[out_offset + sub_axis_0_2 - data_size_one_block],\\n src, 0, 1, 1, 0, 0)\\n with tik_inst.else_scope():\\n # for case sub_axis_0 * axis_2 < data_size_one_block\\n with tik_inst.if_scope(axis_0 * axis_1 * axis_2 > data_size_one_block):\\n m_sub_axis_0 = _ceil_div(data_size_one_block, axis_2)\\n sub_axis_0_2 = m_sub_axis_0 * axis_2\\n sub_axis_0_block_align = sub_axis_0_2 // data_size_one_block\\n left_data = sub_axis_0_2 % data_size_one_block\\n tik_inst.data_move(dst[out_offset + (sub_axis_0 - m_sub_axis_0) * axis_2],\\n src, 0, 1, sub_axis_0_block_align, 0, 0)\\n with tik_inst.if_scope(left_data > 0):\\n with tik_inst.new_stmt_scope():\\n reg_temp = [tik_inst.Scalar(src.dtype)\\n for i in ADDR_IDX_LIST[:data_size_one_block]]\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n reg_temp[idx].set_as(src[sub_axis_0_2 - data_size_one_block + idx])\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n src[idx].set_as(reg_temp[idx])\\n tik_inst.data_move(dst[out_offset + sub_axis_0 * axis_2 -\\n data_size_one_block],\\n src, 0, 1, 1, 0, 0)\\n with tik_inst.else_scope():\\n # the shape size is not bigger than one block size\\n tik_inst.data_move(dst[out_offset], src, 0, 1,\\n _ceil_div(axis_0 * axis_1 * axis_2, data_size_one_block), 0, 0)\\n\\n with tik_inst.else_scope():\\n with tik_inst.if_scope(sub_axis_1 * axis_0 * axis_2 > data_size_one_block):\\n sub_axis_1_0_2 = sub_axis_1 * axis_0 * axis_2\\n sub_axis_1_0_2_block_align = sub_axis_1_0_2 // data_size_one_block\\n left_data = sub_axis_1_0_2 % data_size_one_block\\n tik_inst.data_move(dst[out_offset], src, 0, 1, sub_axis_1_0_2_block_align, 0, 0)\\n with tik_inst.if_scope(left_data > 0):\\n with tik_inst.new_stmt_scope():\\n reg_temp = [tik_inst.Scalar(src.dtype)\\n for i in ADDR_IDX_LIST[:data_size_one_block]]\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n reg_temp[idx].set_as(src[sub_axis_1_0_2 - data_size_one_block + idx])\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n src[idx].set_as(reg_temp[idx])\\n tik_inst.data_move(dst[out_offset + sub_axis_1_0_2 - data_size_one_block],\\n src, 0, 1, 1, 0, 0)\\n with tik_inst.else_scope():\\n # to make sure move in data size is not less than one block\\n with tik_inst.if_scope(axis_0 * axis_1 * axis_2 > data_size_one_block):\\n m_sub_axis_1 = _ceil_div(data_size_one_block, axis_0 * axis_2)\\n sub_axis_1_0_2 = m_sub_axis_1 * axis_0 * axis_2\\n sub_axis_0_block_align = sub_axis_1_0_2 // data_size_one_block\\n left_data = sub_axis_1_0_2 % data_size_one_block\\n tik_inst.data_move(dst[out_offset + (sub_axis_1 - m_sub_axis_1) * axis_0 * axis_2],\\n src, 0, 1, sub_axis_0_block_align, 0, 0)\\n with tik_inst.if_scope(left_data > 0):\\n with tik_inst.new_stmt_scope():\\n reg_temp = [tik_inst.Scalar(src.dtype)\\n for i in ADDR_IDX_LIST[:data_size_one_block]]\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n reg_temp[idx].set_as(src[sub_axis_1_0_2 - data_size_one_block + idx])\\n for idx in ADDR_IDX_LIST[:data_size_one_block]:\\n src[idx].set_as(reg_temp[idx])\\n tik_inst.data_move(dst[out_offset + sub_axis_1 * axis_0 * axis_2 -\\n data_size_one_block],\\n src, 0, 1, 1, 0, 0)\\n with tik_inst.else_scope():\\n # the shape size is not bigger than one block size\\n tik_inst.data_move(dst[out_offset], src, 0, 1,\\n _ceil_div(axis_0 * axis_1 * axis_2, data_size_one_block), 0, 0)\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.6268374","0.5992489","0.58285815","0.58246756","0.5760445","0.5746895","0.56957227","0.5695672","0.5674205","0.5668232","0.5656621","0.55133724","0.5512487","0.55118686","0.54767984","0.5452559","0.54362","0.5378659","0.5353823","0.5353823","0.53118145","0.5307625","0.5304481","0.52880967","0.52782726","0.5271374","0.5249795","0.5249795","0.52397645","0.5220466","0.52202785","0.52083164","0.51938784","0.51798767","0.5136923","0.5131578","0.51213586","0.5105061","0.5101024","0.50970006","0.5087366","0.50847584","0.507706","0.50396264","0.5039476","0.50324905","0.50324905","0.50184715","0.49968368","0.49800575","0.49782437","0.49747035","0.49650806","0.4959419","0.4953591","0.49452943","0.49448436","0.4931954","0.4930093","0.49253044","0.4912059","0.4909279","0.4903383","0.4903383","0.49006224","0.4899867","0.48957187","0.48817652","0.48806912","0.4875072","0.48735133","0.4873187","0.48686087","0.48647654","0.4862426","0.48619366","0.48591033","0.4858729","0.4852133","0.48482397","0.48464713","0.48448375","0.4841419","0.48378778","0.48372066","0.4836198","0.48342308","0.48338732","0.48310903","0.48309663","0.4828195","0.48248386","0.48191294","0.48181754","0.481407","0.48114648","0.4807979","0.48013735","0.48012412","0.4800905","0.4800539"],"string":"[\n \"0.6268374\",\n \"0.5992489\",\n \"0.58285815\",\n \"0.58246756\",\n \"0.5760445\",\n \"0.5746895\",\n \"0.56957227\",\n \"0.5695672\",\n \"0.5674205\",\n \"0.5668232\",\n \"0.5656621\",\n \"0.55133724\",\n \"0.5512487\",\n \"0.55118686\",\n \"0.54767984\",\n \"0.5452559\",\n \"0.54362\",\n \"0.5378659\",\n \"0.5353823\",\n \"0.5353823\",\n \"0.53118145\",\n \"0.5307625\",\n \"0.5304481\",\n \"0.52880967\",\n \"0.52782726\",\n \"0.5271374\",\n \"0.5249795\",\n \"0.5249795\",\n \"0.52397645\",\n \"0.5220466\",\n \"0.52202785\",\n \"0.52083164\",\n \"0.51938784\",\n \"0.51798767\",\n \"0.5136923\",\n \"0.5131578\",\n \"0.51213586\",\n \"0.5105061\",\n \"0.5101024\",\n \"0.50970006\",\n \"0.5087366\",\n \"0.50847584\",\n \"0.507706\",\n \"0.50396264\",\n \"0.5039476\",\n \"0.50324905\",\n \"0.50324905\",\n \"0.50184715\",\n \"0.49968368\",\n \"0.49800575\",\n \"0.49782437\",\n \"0.49747035\",\n \"0.49650806\",\n \"0.4959419\",\n \"0.4953591\",\n \"0.49452943\",\n \"0.49448436\",\n \"0.4931954\",\n \"0.4930093\",\n \"0.49253044\",\n \"0.4912059\",\n \"0.4909279\",\n \"0.4903383\",\n \"0.4903383\",\n \"0.49006224\",\n \"0.4899867\",\n \"0.48957187\",\n \"0.48817652\",\n \"0.48806912\",\n \"0.4875072\",\n \"0.48735133\",\n \"0.4873187\",\n \"0.48686087\",\n \"0.48647654\",\n \"0.4862426\",\n \"0.48619366\",\n \"0.48591033\",\n \"0.4858729\",\n \"0.4852133\",\n \"0.48482397\",\n \"0.48464713\",\n \"0.48448375\",\n \"0.4841419\",\n \"0.48378778\",\n \"0.48372066\",\n \"0.4836198\",\n \"0.48342308\",\n \"0.48338732\",\n \"0.48310903\",\n \"0.48309663\",\n \"0.4828195\",\n \"0.48248386\",\n \"0.48191294\",\n \"0.48181754\",\n \"0.481407\",\n \"0.48114648\",\n \"0.4807979\",\n \"0.48013735\",\n \"0.48012412\",\n \"0.4800905\",\n \"0.4800539\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.0"},"document_rank":{"kind":"string","value":"-1"}}},{"rowIdx":90899,"cells":{"query":{"kind":"string","value":"Yield a batch of data after moving it to device"},"document":{"kind":"string","value":"def __iter__(self):\n for b in self.dl: \n yield to_device(b, self.device)"},"metadata":{"kind":"string","value":"{\n \"objective\": {\n \"self\": [],\n \"paired\": [],\n \"triplet\": [\n [\n \"query\",\n \"document\",\n \"negatives\"\n ]\n ]\n }\n}"},"negatives":{"kind":"list like","value":["def __iter__(self):\n for b in self.dl: \n yield to_device(b, self.device) # yield pauses the execution, not store values in memory, forgets about them once iterated\n # no need to remove batch of data from device, done automatically","def __iter__(self):\n while True:\n if self.stop:\n return\n for item in self.get_next_batch():\n yield item","def __iter__(self):\n batch = []\n for sample in self.dataset:\n batch.append(sample)\n if len(batch) == self.size:\n yield self.transform(batch)\n batch = []\n if batch:\n # the last batch may be less then batch size.\n yield self.transform(batch)","def __iter__(self):\n for b in self.dl:\n yield to_device(b, self.device)","def __iter__(self):\r\n for b in self.dl:\r\n yield to_device(b, self.device)","def one_shot_iterator(dataloader):\n while True:\n for data in dataloader:\n yield data","def __iter__(self):\n for batch in self.iterator:\n yield Batch.from_iterator_batch(batch, self.pad_index, self.sos_index, self.eos_index)","def cycle_loader(loader, device):\n while True:\n for batch in loader:\n # NOTE this is an adhoc solution\n batch.src = (batch.src[0].to(device), batch.src[1].to(device))\n batch.tgt = batch.tgt.to(device)\n logit, indices = batch.bert_topk\n batch.bert_topk = (logit.to(device), indices.to(device))\n yield batch","def batch_data(data, state_queue):\n while True:\n time.sleep(BATCH_SLEEP_TIME)\n state_queue.put((\"device_values\", [data]))","def data_gen(\n v: int, batch: int, nbatches: int, device: torch.device = torch.device(\"cpu\")\n) -> Iterator[Batch]: # TODO bad name\n for i in range(nbatches):\n data = np.random.randint(1, v, size=(batch, 10))\n data[:, 0] = 1\n src: LongTensorType = torch.from_numpy(data)\n tgt: LongTensorType = torch.from_numpy(data)\n src, tgt = src.to(device), tgt.to(device)\n yield Batch(src, tgt, 0)","def next_batch(self):\n\n while self.cap.isOpened():\n flag, frame = self.cap.read()\n yield frame","def _train_batch(self):\n\n # start epoch\n for i, (source, target) in enumerate(self.train_dataset):\n result = self._batch_iter(source, target, i)\n\n # yield\n yield result","def next_batch(self):\n for nb in xrange(self.num_batches):\n if self.batch_end < self.full_len:\n batch_X_raw = self.full_X[self.batch_start:self.batch_end]\n batch_y_raw = self.full_y[self.batch_start:self.batch_end]\n else:\n batch_X_raw = self.full_X[self.batch_start:]\n batch_y_raw = self.full_y[self.batch_start:]\n batch_X, batch_y = pad_sort_data(batch_X_raw, batch_y_raw)\n self.batch_start = self.batch_end\n self.batch_end += self.batch_size\n yield batch_X, batch_y","def data_generator(delta=1, batch_size=32):\n while True:\n yield generate_samples(delta=delta, n=batch_size)","def data_parallel(self, batch_size, inputs):\n inputs = list(inputs)\n\n # quick path: only one device, do not slice\n if len(self.work_devices) == 1:\n assert(self.main_device == self.work_devices[0])\n yield self.main_device, False, tuple(inputs)\n\n # slow path: multi-GPUs\n else:\n # the GPUs are not in the same group, place variables on CPU\n if self.main_device not in self.work_devices:\n yield self.main_device, True, tuple(inputs)\n\n # build the paralleled computation graph for each device\n with tf.name_scope('data_parallel') as ns:\n pass # generate a name scope to place our data slicing ops\n\n k = len(self.work_devices)\n for i, device in enumerate(self.work_devices):\n dev_inputs = []\n with tf.name_scope(ns + 'tower_gpu_{}'.format(i)):\n for inp in inputs:\n slice_len = (batch_size + k - 1) // k\n low, high = slice_len * i, slice_len * (i + 1)\n dev_inputs.append(inp[low: high])\n yield device, False, tuple(dev_inputs)","def batch_iter(data, batch_size):\n random.shuffle(data)\n num_data = len(data)\n num_steps = int(math.ceil(num_data/float(batch_size)))\n rem = num_data%batch_size\n for step in range(num_steps):\n if step == (num_steps-1) and rem !=0:\n data_batch = data[(step)*batch_size:]\n random.shuffle(data)\n data_batch.extend(data[:(batch_size - rem)])\n board, target, ko = process_mini_batch(data_batch)\n yield (board, target, ko)\n else:\n data_batch = data[step*batch_size:(step+1)*batch_size]\n board, target, ko = process_mini_batch(data_batch)\n yield (board, target, ko)","def __iter__(self):\n\n # collector will fetch chunksize array for each 'get' call\n collector = FIFOArray(self.chunksize, self.axis)\n\n # make tmp array to hold generated subarrs\n tmp = []\n tmp_size = 0\n for subarr in self.data(**self.kwargs):\n\n tmp.append(subarr)\n tmp_size += subarr.shape[self.axis]\n\n # if tmp exceeds chunksize put in collector\n if tmp_size >= self.chunksize:\n arr = np.concatenate(tmp, axis=self.axis)\n collector.put(arr)\n\n # fetch chunksize till not full\n while collector.full():\n yield collector.get()\n\n # place leftover back into tmp and empty collector\n tmp = [collector.queue]\n tmp_size = collector.qsize()\n collector.queue = np.array([])\n\n else:\n\n # append to tmp again\n continue\n\n # else runs after normal loop exit -- required here\n else: #pylint: disable=useless-else-on-loop\n\n # yield whatever is left in tmp (its below chunksize)\n remaining = np.concatenate(tmp, axis=self.axis)\n if remaining.size > 0:\n yield remaining","def _next_batch(self, loader: CustomIterator) -> list:\n return self.mover.move(loader.__next__())","def data_generator(self, data):\n X, y = [], []\n while 1:\n np.random.shuffle(data)\n for line in data:\n img = Image.open(line[0])\n img = img.resize((32, 16))\n img = np.asarray(img, dtype=np.float32)\n img = img / 128. - 1.\n img = np.transpose(img, (2, 0, 1)) \n X.append(img)\n y.append(line[1])\n if len(X) == self.config.batch_size:\n batch = (np.asarray(X), np.asarray(y))\n X = []\n y = []\n yield batch","def __iter__(self):\n for sample in self.data:\n yield sample","def test_minibatches(self):\n batch_size = self.params['batch_size']\n start_index = 0\n while start_index + batch_size < 500:\n end_index = start_index + batch_size\n yield self.input[start_index:end_index], self.y[start_index:end_index]\n start_index = end_index","def batch_data(cls, train_data, train_labels, batch_size):\n for batch in range(int(np.ceil(train_data.shape[0] / batch_size))):\n start = batch_size * batch\n end = start + batch_size\n if end > train_data.shape[0]:\n yield batch, (train_data[start:train_data.shape[0]], \\\n train_labels[start:train_data.shape[0]])\n else:\n yield batch, (train_data[start:end], \\\n train_labels[start:end])","def __iter__(self):\n for i, data in enumerate(self.dataloader):\n if i * self.opt.batch_size >= self.opt.max_dataset_size:\n break\n yield data","def next_batch(self, batch_size):\n raise NotImplementedError","def __iter__(self):\n\t\tfor i, data in enumerate(self.dataloader):\n\t\t\tif i * self.opt.batch_size >= self.opt.max_dataset_size:\n\t\t\t\tbreak\n\t\t\tyield data","def _next(self):\n batch_start, batch_end = self.batch_start, self.batch_start + self.batch_size\n X_batch, y_batch = self.X[batch_start:batch_end], self.y[batch_start:batch_end]\n X_batch, y_batch = self.process_batch(X_batch, y_batch)\n if batch_end > self.X.shape[0]:\n self.batch_start = 0\n else:\n self.batch_start = batch_end\n return X_batch, y_batch","def __iter__(self):\n batch = []\n for idx in self._sampler:\n batch.append(idx)\n if len(batch) == self._batch_size:\n batch = sum(batch, [])\n yield batch\n batch = []\n if len(batch) > 0:\n batch = sum(batch, [])\n yield batch","def sample_batch(pid, args, batch_queue, port_dict, device, actor_id_to_ip_dataport, local_size, cache_array):\n def recv_data(k, data_stream, actor_set, real_data_tasks_i):\n for real_data in data_stream:\n tmp = []\n tmp.append(real_data.state)\n tmp.append(real_data.action)\n tmp.append(real_data.reward)\n tmp.append(real_data.next_state)\n tmp.append(real_data.done)\n tmp.append(actor_set[k]['w'][real_data.idx])\n tmp.append(actor_set[k]['i'][real_data.idx])\n tmp.append(actor_set[k]['t'][real_data.idx])\n tmp.append(real_data.timestamp)\n local_dict[actor_set[k]['i'][real_data.idx]] = tmp\n cache_array[actor_set[k]['i'][real_data.idx]] |= 2**pid\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.state), dtype=np.uint8).reshape((1, 4, 84, 84)))\n real_data_tasks_i['states'].append(decom_state) #.to(device))\n real_data_tasks_i['actions'].append(torch.LongTensor([real_data.action])) #.to(device))\n real_data_tasks_i['rewards'].append(torch.FloatTensor([real_data.reward])) #.to(device))\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.next_state), dtype=np.uint8).reshape((1, 4, 84, 84)))\n real_data_tasks_i['next_states'].append(decom_next_state) #.to(device))\n real_data_tasks_i['dones'].append(torch.FloatTensor([real_data.done])) #.to(device))\n real_data_tasks_i['batch_weights'].append(torch.FloatTensor([actor_set[k]['w'][real_data.idx]])) #.to(device))\n real_data_tasks_i['batch_idxes'].append(actor_set[k]['i'][real_data.idx])\n # is the data overwrited?\n real_data_tasks_i['batch_timestamp_store'].append(actor_set[k]['t'][real_data.idx])\n real_data_tasks_i['batch_timestamp_real'].append(real_data.timestamp)\n conn = grpc.insecure_channel(port_dict['replay_ip'] + ':' + port_dict['sampleDataPort'])\n client = apex_data_pb2_grpc.SampleDataStub(channel=conn)\n local_dict = {}\n while True:\n batch_timestamp_real = []\n batch_timestamp_store = []\n batch_weights = []\n batch_idxes = []\n\n states, actions, rewards, next_states, dones = [], [], [], [], []\n\n res_batch = client.Send(apex_data_pb2.SampleDataRequest(batch_size=args.batch_size, beta = args.beta))\n actor_ids, data_ids, timestamps, weights, idxes = res_batch.actor_ids, res_batch.data_ids, res_batch.timestamp, res_batch.weights, res_batch.idxes\n actor_set = {}\n cached_value = {'states':{},'actions':{},'rewards':{},'next_states':{},'dones':{},'batch_weights':{},'batch_idxes':{},'batch_timestamp_store':{},'batch_timestamp_real':{}}\n for i in range(len(actor_ids)):\n set_a = actor_set.get(actor_ids[i], False)\n if set_a == False:\n actor_set[actor_ids[i]] = {}\n set_a = actor_set[actor_ids[i]]\n set_a['d'] = []\n set_a['w'] = []\n set_a['i'] = []\n set_a['t'] = []\n cached_value['states'][actor_ids[i]] = []\n cached_value['actions'][actor_ids[i]] = []\n cached_value['rewards'][actor_ids[i]] = []\n cached_value['next_states'][actor_ids[i]] = []\n cached_value['dones'][actor_ids[i]] = []\n cached_value['batch_weights'][actor_ids[i]] = []\n cached_value['batch_idxes'][actor_ids[i]] = []\n cached_value['batch_timestamp_store'][actor_ids[i]] = []\n cached_value['batch_timestamp_real'][actor_ids[i]] = []\n cache_id = actor_ids[i]*local_size+data_ids[i]\n cache_trans = cache_array[cache_id]\n if cache_trans & 2**pid == 0:\n set_a['d'].append(data_ids[i])\n set_a['w'].append(weights[i])\n set_a['i'].append(idxes[i])\n set_a['t'].append(timestamps[i])\n if cache_trans == 0 and local_dict.get(cache_id, False) != False:\n del local_dict[cache_id]\n else:\n try:\n state_tmp = local_dict[cache_id][0]\n action_tmp = local_dict[cache_id][1]\n reward_tmp = local_dict[cache_id][2] \n next_state_tmp = local_dict[cache_id][3] \n done_tmp = local_dict[cache_id][4] \n batch_weight_tmp = local_dict[cache_id][5] \n batch_idx_tmp = local_dict[cache_id][6] \n batch_store_tmp = local_dict[cache_id][7] \n batch_real_tmp = local_dict[cache_id][8] \n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\n cached_value['states'][actor_ids[i]].append(decom_state)\n cached_value['actions'][actor_ids[i]].append(torch.LongTensor([action_tmp]))\n cached_value['rewards'][actor_ids[i]].append(torch.FloatTensor([reward_tmp]))\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(next_state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\n cached_value['next_states'][actor_ids[i]].append(decom_next_state)\n cached_value['dones'][actor_ids[i]].append(torch.FloatTensor([done_tmp]))\n cached_value['batch_weights'][actor_ids[i]].append(torch.FloatTensor([batch_weight_tmp]))\n cached_value['batch_idxes'][actor_ids[i]].append(batch_idx_tmp)\n cached_value['batch_timestamp_store'][actor_ids[i]].append(batch_store_tmp)\n cached_value['batch_timestamp_real'][actor_ids[i]].append(batch_real_tmp)\n except:\n set_a['d'].append(data_ids[i])\n set_a['w'].append(weights[i])\n set_a['i'].append(idxes[i])\n set_a['t'].append(timestamps[i])\n real_data_links = {}\n real_data_tasks = {}\n for k, v in actor_set.items():\n actor_ip, data_port = actor_id_to_ip_dataport[k]\n conn_actor = grpc.insecure_channel(actor_ip + ':' + data_port)\n client_actor = apex_data_pb2_grpc.SendRealDataStub(channel=conn_actor)\n real_data_links[k] = client_actor.Send(apex_data_pb2.RealBatchRequest(idxes=v['d']))\n real_data_tasks[k] = {}\n real_data_tasks[k]['states'] = cached_value['states'][k]\n real_data_tasks[k]['actions'] = cached_value['actions'][k]\n real_data_tasks[k]['rewards'] = cached_value['rewards'][k]\n real_data_tasks[k]['next_states'] = cached_value['next_states'][k]\n real_data_tasks[k]['dones'] = cached_value['dones'][k]\n real_data_tasks[k]['batch_weights'] = cached_value['batch_weights'][k]\n real_data_tasks[k]['batch_idxes'] = cached_value['batch_idxes'][k]\n real_data_tasks[k]['batch_timestamp_store'] = cached_value['batch_timestamp_store'][k]\n real_data_tasks[k]['batch_timestamp_real'] = cached_value['batch_timestamp_real'][k]\n threads = []\n for k, v in real_data_links.items():\n t = threading.Thread(target=recv_data, args=(k, v, actor_set, real_data_tasks[k],))\n threads.append(t)\n t.start()\n\n for t in threads:\n t.join()\n\n for k, v in real_data_tasks.items():\n states += v['states']\n actions += v['actions']\n rewards += v['rewards']\n next_states += v['next_states']\n dones += v['dones']\n batch_weights += v['batch_weights']\n batch_idxes += v['batch_idxes']\n batch_timestamp_real += v['batch_timestamp_real']\n batch_timestamp_store += v['batch_timestamp_store']\n\n states = torch.cat(states,0).to(device)\n actions = torch.cat(actions,0).to(device)\n rewards = torch.cat(rewards,0).to(device)\n next_states = torch.cat(next_states,0).to(device)\n dones = torch.cat(dones,0).to(device)\n batch_weights = torch.cat(batch_weights,0).to(device)\n\n batch = [states, actions, rewards, next_states, dones, batch_weights, batch_idxes]\n batch_queue.put(batch)\n data, batch = None, None","def next_simple_dataset(dataset, batch_size: int, datatype):\n while True:\n x_batch = []\n y_batch = []\n for i in range(batch_size):\n try:\n x, y, data_unit, index = create_xy(dataset, datatype)\n # x = normalize(x)\n x_batch.append(x)\n y_batch.append(y)\n except StopIteration:\n break\n x_batch, y_batch = np.array(x_batch), np.array(y_batch)\n if datatype != DataType.test:\n x_batch = SEQ_CVXTZ.augment_images(x_batch).astype(\"float32\")\n x_batch = np.array([normalize(x) for x in x_batch])\n # org_shape = x_batch.shape\n # org_width = x_batch.shape[1]\n # corner = int((org_width - ROI_IMAGE_SIZE) // 2)\n # print(f\"0: org_shape:{org_shape} x_batch:{x_batch.shape} corner:{corner}\")\n # x_batch = x_batch[:, corner:(org_width - corner), corner:(org_width - corner), :]\n # resized_x_batch = []\n # for x in x_batch:\n # img = Image.fromarray(np.uint8(x))\n # img = img.resize((IMAGE_SIZE, IMAGE_SIZE), Image.LANCZOS)\n # resized_x_batch.append(normalize(np.array(img)))\n # print(f\"1: org_shape:{org_shape} corner:{corner} x_batch:{x_batch.shape}\")\n # yield np.array(resized_x_batch), y_batch\n yield np.array(x_batch), y_batch","def __iter__(self):\n for batch in self.data:\n batch_size = len(batch)\n X, e1, e2, dist1, dist2, e1_pos, e2_pos, y = list(zip(*batch))\n\n x_len = max(len(x) for x in X)\n x_ids = torch.LongTensor(batch_size, x_len).fill_(0)\n dist1_padded = torch.LongTensor(batch_size, x_len).fill_(0)\n dist2_padded = torch.LongTensor(batch_size, x_len).fill_(0)\n for i, doc in enumerate(X):\n x_ids[i, :len(doc)] = torch.LongTensor(doc)\n\n dist1_padded[i, :len(doc)] = torch.LongTensor(dist1[i])\n dist1_padded[i, len(doc):] = torch.LongTensor([pos(e1_pos[i][1] - idx) for idx, _ in enumerate(x_ids[i][len(doc):], start=len(doc))])\n\n dist2_padded[i, :len(doc)] = torch.LongTensor(dist2[i])\n dist2_padded[i, len(doc):] = torch.LongTensor([pos(e2_pos[i][1] - idx) for idx, _ in enumerate(x_ids[i][len(doc):], start=len(doc))])\n\n e1_tensor = torch.LongTensor(e1)\n e2_tensor = torch.LongTensor(e2)\n\n y_tensor = torch.LongTensor(y)\n\n if self.gpu:\n x_ids = x_ids.pin_memory()\n e1_tensor = e1_tensor.pin_memory()\n e2_tensor = e2_tensor.pin_memory()\n dist1_padded = dist1_padded.pin_memory()\n dist2_padded = dist2_padded.pin_memory()\n y_tensor = y_tensor.pin_memory()\n\n yield (x_ids, e1_tensor, e2_tensor, dist1_padded, dist2_padded, y_tensor)","def next(self):\n\t\t# keep looping until we reach our batch size\n\t\twhile True:\n\t\t\tret = self.get_batch()\n\t\t\tself.index += self.batch_size\n\t\t\tif self.index >= len(self.texts) - self.batch_size:\n\t\t\t\tself.index = 0\n\t\t\tyield ret","def get_batches(self, batch_size):\n if self.data.shape[0] % batch_size != 0:\n raise RuntimeError('num of data tuples is not a multiple of batch size')\n num_batch = self.data.shape[0] // batch_size\n for b in range(num_batch):\n yield self.data[b*batch_size:(b+1)*batch_size, :], \\\n self.target[b*batch_size:(b+1)*batch_size, :]","def _chunk_data(self):\n for n in range(0, len(self.data) + 1, len(self.data) //\n self.num_of_chunks):\n yield self.data[0 + n:len(self.data) // self.num_of_chunks + n]","def batch_iter(input_data,batch_size):\r\n batch_ids,batch_mask,batch_segment,batch_label=[],[],[],[]\r\n for features in input_data:\r\n if len(batch_ids) == batch_size:\r\n yield batch_ids,batch_mask,batch_segment,batch_label\r\n batch_ids, batch_mask, batch_segment, batch_label = [], [], [], []\r\n\r\n batch_ids.append(features['input_ids'])\r\n batch_mask.append(features['input_mask'])\r\n batch_segment.append(features['segment_ids'])\r\n batch_label.append(features['label_ids'])\r\n\r\n if len(batch_ids) != 0:\r\n yield batch_ids, batch_mask, batch_segment, batch_label","def __iter__(self):\n batch = []\n for i_batch in range(self.episode_num):\n classes = torch.randperm(len(self.idx_list))[: self.way_num]\n for c in classes:\n idxes = self.idx_list[c.item()]\n pos = torch.randperm(idxes.size(0))[: self.image_num]\n batch.append(idxes[pos])\n if len(batch) == self.episode_size * self.way_num:\n batch = torch.stack(batch).reshape(-1)\n yield batch\n batch = []","def _next_batch(self, loader) -> list:\n return self.mover.move(loader.__next__())","def unchunk():\n\n @filters\n def _dagpype_internal_fn_act(target):\n try:\n while True:\n a = (yield)\n if len(a) == 0:\n continue\n if a.ndim == 1:\n for i in range(a.shape[0]):\n target.send(a[i]) \n else:\n for i in range(a.shape[0]):\n target.send(tuple(a[i]))\n except GeneratorExit:\n if len(l) > 0:\n target.send(numpy.array(l, dtype = dtype_)) \n \n return _dagpype_internal_fn_act","def next(self) -> Iterable[RecordBatch]:\n for batch in self._parent_operator.next():\n args = self._process_arguments(self._arguments, batch=batch)\n yield self._kernel(batch, args)","def create_batch(client, generator: DataGenerator):\n try:\n event_data_batch = client.create_batch()\n for device in generator.devices:\n # event_data_batch.add(EventData(gen.generate_payload(device)))\n event_data_batch.add(EventData(generator.generate_payload(device)))\n return event_data_batch\n except Exception as e:\n print(str(e))","def next_batch(self, batch_size):\r\n raise NotImplementedError","def forward(self, batch):\n raise NotImplementedError","def forward(self, batch):\n raise NotImplementedError","def gen_batches(data, batch_size):\n data = np.array(data)\n\n for i in range(0, data.shape[0], batch_size):\n yield data[i:i+batch_size]","def generate_batches(data, labels, batch_size):\n for start in range(0, len(data), batch_size):\n yield Tensor(data[start:start+batch_size, ...]), Tensor(labels[start:start+batch_size, ...])","def batches(data, batch_size) -> list:\n rv = []\n for idx, line in enumerate(data):\n if idx != 0 and idx % batch_size == 0:\n yield rv\n rv = []\n rv.append(line)\n yield rv","def data_iter(data, batch_size, shuffle=True, sort=True):\n\n batched_data = []\n if shuffle: np.random.shuffle(data)\n batched_data.extend(list(batch_slice(data, batch_size, sort)))\n\n if shuffle: np.random.shuffle(batched_data)\n for batch in batched_data:\n yield batch","def static_batch(data, batch_size=16):\n buf = []\n for sample in data:\n buf.append(sample)\n if len(buf) >= batch_size:\n yield buf\n buf = []\n if len(buf) > 0:\n yield buf","def enhancer_iterator(self, data, labels, batch_size, num_steps):\n def seq_to_ints(seq):\n return [self.vocab.word_to_index[c] for c in seq]\n\n # Map raw data to array of ints. if all sequences are the same length L, \n # raw_data will be N-by-L\n mdata = np.array([seq_to_ints(i) for i in data], dtype=np.int32)\n num_batches = len(mdata) // batch_size\n \n # data will have batch_len elements, each of size batch_size\n # ASSUME FIXED SEQUENCE LENGTHS OFF 1000 FOR NOW (5/20/16)\n # Just grab middle self.config.num_steps nucleotides\n a = int(len(mdata[0,:])/2-self.config.num_steps/2)\n b = int(len(mdata[0,:])/2+self.config.num_steps/2)\n for i in range(num_batches):\n x = mdata[batch_size*i:batch_size*(i+1),a:b]\n if labels is not None:\n y = labels[batch_size*i:batch_size*(i+1)]\n else:\n y = None\n yield(x,y)","def batch_generator(data, batch_size, config, target=False):\n listsize = min(len(data.pilist), 10)\n while 1:\n nrd = np.arange(len(data.pilist))\n np.random.shuffle(nrd)\n for ri in range(len(data.pilist) // listsize):\n cpilist = [data.pilist[nrd[rid]] for rid in range(listsize * ri, listsize * (ri + 1))]\n data.loadpatch(config, cpilist)\n if target:\n data_l = [data.xarray, np.zeros(shape=(len(data.xarray), 1))]\n else:\n data_l = [data.xarray, data.yarray - 1]\n\n for di in range(len(data_l[0])):\n data_l[0][di][:, :, :, 0] = data_l[0][di][:, :, :, 0] / np.max(data_l[0][di][:, :, :, 0])\n\n for repi in range(3):\n data_l = shuffle_aligned_list(data_l)\n\n batch_count = 0\n while True:\n if batch_count * batch_size + batch_size >= len(data_l[0]):\n if len(data.pilist) // listsize != 1:\n print('list end', ri * listsize)\n break\n else:\n batch_count = 0\n\n start = batch_count * batch_size\n end = start + batch_size\n batch_count += 1\n yield [d[start:end] for d in data_l]","def next(self):\n if self.curr_idx == len(self.idx):\n raise StopIteration\n\n # Fetch the index\n i = self.idx[self.curr_idx]\n self.curr_idx += 1\n\n # Get labels & ids\n labels = self.ndlabels[i:i+self.batch_size]\n users = self.ndusers[i:i + self.batch_size]\n items = self.nditems[i:i + self.batch_size]\n\n # Get feature arrays\n if self.create_batches:\n user_features = self.nduserfeatures[i:i+self.batch_size]\n item_features = self.nditemfeatures[i:i+self.batch_size]\n else:\n # Create user feature arrays\n user_features = mx.ndarray.take(a=self.unique_user_features, indices=users)\n item_features = mx.ndarray.take(a=self.unique_item_features, indices=items)\n\n return mx.io.DataBatch([user_features, item_features], [labels], index = users, pad=0,\n provide_data=[mx.io.DataDesc(name=self.data_names[0], shape=user_features.shape),\n mx.io.DataDesc(name=self.data_names[1], shape=item_features.shape)],\n provide_label=[mx.io.DataDesc(name=self.label_names[0], shape=labels.shape)])","def next_batch(self, batch_size):\n batch_data = np.zeros([batch_size,] + list(self.example_shape))\n for i in range(batch_size):\n index = self.q.pop()\n batch_data[i,...] = self.data[index]\n if len(self.q)==0:\n self.__new_epoch()\n\n return batch_data","def chunk_generator( callback, request, c ):","def next(self):\n #print('next')\n batch_size = self.batch_size\n batch_data = nd.empty((batch_size,)+self.data_shape)\n batch_label = nd.empty((batch_size,)+self.label_shape)\n i = 0\n #self.cutoff = random.randint(800,1280)\n try:\n while i < batch_size:\n #print('N', i)\n data, label, annot = self.next_sample()\n R = self.get_data(data, label, annot)\n if R is None:\n continue\n data_out, label_out, flip_data_out, flip_label_out = R\n if not self.use_coherent:\n data = nd.array(data_out)\n data = nd.transpose(data, axes=(2, 0, 1))\n label = nd.array(label_out)\n #print(data.shape, label.shape)\n batch_data[i][:] = data\n batch_label[i][:] = label\n i += 1\n else:\n data = nd.array(data_out)\n data = nd.transpose(data, axes=(2, 0, 1))\n label = nd.array(label_out)\n data2 = nd.array(flip_data_out)\n data2 = nd.transpose(data2, axes=(2, 0, 1))\n label2 = nd.array(flip_label_out)\n #M = nd.array(M)\n #print(data.shape, label.shape)\n batch_data[i][:] = data\n batch_label[i][:] = label\n #i+=1\n j = i+self.per_batch_size//2\n batch_data[j][:] = data2\n batch_label[j][:] = label2\n i += 1\n if j%self.per_batch_size==self.per_batch_size-1:\n i = j+1\n except StopIteration:\n if i<batch_size:\n raise StopIteration\n\n #return {self.data_name : batch_data,\n # self.label_name : batch_label}\n #print(batch_data.shape, batch_label.shape)\n return mx.io.DataBatch([batch_data], [batch_label], batch_size - i)","def _batch_iter(self, source, target, i: int):\n # send data to device\n source = source.to(self.device)\n target = target.to(self.device)\n\n # the result and loss\n result = self.model(source)\n loss = self.criterion(result, target)\n\n # optimization and backward\n self.optimizer.zero_grad()\n loss.backward()\n self.optimizer.step()\n\n # update the loss\n self.epoch_loss.update(loss.item(), source.size(0))\n\n # print the information\n if self.info:\n print(f\"\\rEpoch: { self.epoch } | Batch: { i } | loss: { self.epoch_loss.avg }\", end=\"\")\n\n # clean the data\n del source, target\n\n return result","def process(self, data_itr):\n for data in data_itr:\n self.update(data)\n while True:\n try:\n out = self.next()\n yield out\n except StopIteration:\n break","def data_gen(voc_size, batch, nbatches, seq_len = 15):\r\n for i in range(nbatches):\r\n # (batch_size, seq_len)\r\n data = torch.from_numpy(\r\n np.random.randint(1, voc_size, size=(batch, seq_len)))\r\n data[:, 0] = 1 # add start token\r\n src = Variable(data, requires_grad=False)\r\n tgt = Variable(data, requires_grad=False)\r\n yield Batch(src, tgt, 0) # Accessed by next function one by one\r","def prepare_batch(batchdata, device=None, non_blocking=False):\n return batchdata[\"hand\"].to(device=device, non_blocking=non_blocking)","def _batcher(self, rows):\n row_count = 0\n batch = []\n batch_count = 1\n\n total_rows_modified = 0\n throttle_count = 0\n\n i = 0\n for row in rows:\n if row_count > self.batch_size - 1:\n logger.debug(f\"row_count={row_count} batch_size={self.batch_size} and batch={len(batch)}\")\n # Yield the previous batch\n yield batch\n\n # Start the new batch\n batch = []\n batch.append(row)\n row_count = 1\n\n batch_count += 1\n # break # toggle to load one batch only\n else:\n row_count += 1\n batch.append(row)\n\n # Put in a sleep timer to throttle how hard we hit the database\n if self.throttle_time and self.throttle_size and (throttle_count > self.throttle_size - 1):\n logger.info(f\"Sleeping for {self.throttle_time} seconds... row: {i}\")\n time.sleep(int(self.throttle_time))\n throttle_count = 0\n elif self.throttle_time and self.throttle_size:\n throttle_count += 1\n i += 1\n\n yield batch","def TransferMemorySequence():\r\n pass","def get_batch(iterator, batch_size):\n while True:\n center_batch = np.zeros(batch_size, dtype = np.uint32)\n target_batch = np.zeros((batch_size, 1), dtype = np.uint32)\n for index in range(batch_size):\n center_batch[index], target_batch[index] = next(iterator)\n\n yield center_batch, target_batch","def data_generator(batch_size, preprocessor, x, y):\n num_examples = len(x)\n examples = zip(x, y)\n examples = sorted(examples, key = lambda x: x[0].shape[0])\n end = num_examples - batch_size + 1\n batches = [examples[i:i+batch_size]\n for i in range(0, end, batch_size)]\n random.shuffle(batches)\n while True:\n for batch in batches:\n x, y = zip(*batch)\n yield preprocessor.process(x, y)","def __iter__(self):\n worker_info = torch.utils.data.get_worker_info()\n num_batches = len(self.coords_batcher)\n if worker_info is None:\n # In single-processing mode\n start, end = 0, num_batches\n else:\n worker_id = worker_info.id\n num_workers = worker_info.num_workers\n shard_size = int(np.ceil(num_batches / num_workers))\n start = shard_size * worker_id\n end = min(start + shard_size, num_batches)\n return (self.get_batch(i) for i in range(start, end))","def next_batch(self, batch_size=8):\n raise NotImplementedError()","def next_batch(x, y, batch_size):\n\n def as_batch(data, start, count):\n part = []\n for i in range(start, start + count):\n part.append(data[i])\n return np.array(part)\n\n for i in range(0, len(x)-batch_size, batch_size):\n yield as_batch(x, i, batch_size), as_batch(y, i, batch_size)","def _next(self):\n batch_start, batch_end = self.batch_start, self.batch_start + self.batch_size\n if batch_end > self.X.shape[0]:\n self.shuffle()\n return self._next()\n else:\n batch_indices = self.indices[batch_start:batch_end]\n X_batch, y_batch = self.X[batch_indices], self.y[batch_indices]\n X_batch, y_batch = self.process_batch(X_batch, y_batch)\n self.batch_start = batch_end\n return X_batch, y_batch","def next_batch(self, batch_size, shuffle=True):","def __iter__(self):\n\n collector = FIFOArray(self.chunksize, self.axis)\n for arr, maskarr in zip(self.data, self.mask):\n\n if not np.any(maskarr):\n continue\n\n filtered = np.take(arr, np.flatnonzero(maskarr), axis=self.axis)\n collector.put(filtered)\n\n while collector.full():\n\n yield collector.get()\n\n # else runs after normal loop exit -- required here\n else: #pylint: disable=useless-else-on-loop\n\n if collector.qsize() > 0:\n\n yield collector.get()","def _simulation_data_iterator(res, path, data):\n res.append(data)","def serialize_batches():\n # a set of variables for the state of current batch which will be converted to Arrow\n # RecordBatch.\n pdfs = []\n state_pdfs = []\n pdf_data_cnt = 0\n state_data_cnt = 0\n\n return_schema = None\n\n for data in iterator:\n # data represents the result of each call of user function\n packaged_result = data[0]\n\n # There are two results from the call of user function:\n # 1) iterator of pandas DataFrame (output)\n # 2) updated state instance\n pdf_iter = packaged_result[0][0]\n state = packaged_result[0][1]\n\n # This is static and won't change across batches.\n return_schema = packaged_result[1]\n\n for pdf in pdf_iter:\n # We ignore empty pandas DataFrame.\n if len(pdf) > 0:\n pdf_data_cnt += len(pdf)\n pdfs.append(pdf)\n\n # If the total number of records in current batch exceeds the configured\n # threshold, time to construct the Arrow RecordBatch from the batch.\n if pdf_data_cnt > self.arrow_max_records_per_batch:\n batch = construct_record_batch(\n pdfs, pdf_data_cnt, return_schema, state_pdfs, state_data_cnt\n )\n\n # Reset the variables to start with new batch for further data.\n pdfs = []\n state_pdfs = []\n pdf_data_cnt = 0\n state_data_cnt = 0\n\n yield batch\n\n # This has to be performed 'after' evaluating all elements in iterator, so that\n # the user function has been completed and the state is guaranteed to be updated.\n state_pdf = construct_state_pdf(state)\n\n state_pdfs.append(state_pdf)\n state_data_cnt += 1\n\n # processed all output, but current batch may not be flushed yet.\n if pdf_data_cnt > 0 or state_data_cnt > 0:\n batch = construct_record_batch(\n pdfs, pdf_data_cnt, return_schema, state_pdfs, state_data_cnt\n )\n\n yield batch","def __call__(self):\n if self.numbatches is None:\n pool = self.pooler()\n if self.batchsize is None:\n self.batchsize = self.pooler.nInPool()\n self.numbatches = self.pooler.nInPool()//self.batchsize\n for i in xrange(self.numbatches):\n pool = self.pooler()\n self._reset_batch()\n if self.samplemethod == 'balance' and len(self.keysamplers)>0:\n batchinds,keyids = self._samplebalanced(pool)\n elif self.samplemethod == 'uniform':\n batchinds,keyids = self._sampleuniform(pool)\n else:\n batchinds,keyids = self._samplesequential(i)\n batch = self._extractInds(pool,batchinds,keyids)\n for k in batch:\n batch[k][np.isnan(batch[k])] = self.nanreplacement\n yield batch","def _next(self):\n i = 0\n while i < self.size:\n yield self.data[i]\n i += 1","def __iter__(self):\n if not self.big:\n yield self.D[:self.P, :self.d], self.labels[:self.P]\n else:\n for i in range(int(self.P / self.batch_size)):\n yield self.D[i*self.batch_size:min(self.P, (i+1)*self.batch_size), :self.d].cuda(), \\\n self.labels[i*self.batch_size:min(self.P, (i+1)*self.batch_size), :].cuda()","def in_memory_rechunk(\n inputs: List[Tuple[core.ChunkKey, xarray.Dataset]],\n target_chunks: Mapping[str, int],\n) -> Iterator[Tuple[core.ChunkKey, xarray.Dataset]]:\n key, dataset = consolidate_chunks(inputs)\n yield from split_chunks(key, dataset, target_chunks)","def batch_data(source, target, batch_size):\n for batch_i in range(0, len(source)//batch_size):\n start_i = batch_i * batch_size\n source_batch = source[start_i:start_i + batch_size]\n target_batch = target[start_i:start_i + batch_size]\n yield np.array(pad_sentence_batch(source_batch)), np.array(pad_sentence_batch(target_batch))","def train_minibatches(self):\n batch_size = self.params['batch_size']\n start_index = 0\n while start_index + batch_size < 500:\n end_index = start_index + batch_size\n yield self.input[start_index:end_index], self.y[start_index:end_index]\n start_index = end_index","def next(self):\n # Keeps under lock only the mechanism which advances\n # the indexing of each batch.\n with self.lock:\n index_array = next(self.index_generator)\n # The transformation of images is not under thread lock\n # so it can be done in parallel\n return self._get_batches_of_transformed_samples(index_array)","def data_generator(dataset, config, shuffle=True, augment=False, augmentation=None, batch_size=1):\n b = 0 # batch item index\n image_index = -1\n image_ids = np.copy(dataset.image_ids)\n error_count = 0\n\n # Keras requires a generator to run indefinately.\n while True:\n try:\n # Increment index to pick next image. Shuffle if at the start of an epoch.\n image_index = (image_index + 1) % len(image_ids)\n if shuffle and image_index == 0:\n np.random.shuffle(image_ids)\n\n # Get GT bounding boxes and masks for image.\n image_id = image_ids[image_index]\n image, gt_class_ids = load_image_gt(dataset, config, image_id, augment=augment,\n augmentation=augmentation)\n\n # Init batch arrays\n if b == 0:\n batch_images = np.zeros(\n (batch_size,) + image.shape, dtype=np.float32)\n batch_gt_class_ids = np.zeros(\n (batch_size, config.MAX_GT_INSTANCES), dtype=np.int32)\n\n # Add to batch\n batch_images[b] = mold_image(image.astype(np.float32), config)\n batch_gt_class_ids[b, gt_class_ids] = 1\n b += 1\n\n # Batch full?\n if b >= batch_size:\n inputs = [batch_images, batch_gt_class_ids]\n outputs = []\n\n yield inputs, outputs\n\n # start a new batch\n b = 0\n except (GeneratorExit, KeyboardInterrupt):\n raise\n except:\n # Log it and skip the image\n logging.exception(\"Error processing image {}\".format(\n dataset.image_info[image_id]))\n error_count += 1\n if error_count > 5:\n raise","def data_iter(data, vocab, batch_size, shuffle=True, cuda=False):\n\n buckets = defaultdict(list)\n for pair in data:\n src_sent = pair[0]\n buckets[len(src_sent)].append(pair)\n\n batched_data = []\n for src_len in buckets:\n tuples = buckets[src_len]\n if shuffle: np.random.shuffle(tuples)\n batched_data.extend(list(HNCMDataLoader.batch_slice(tuples, batch_size)))\n\n if shuffle:\n np.random.shuffle(batched_data)\n for src_sents, trg_sents, fact_sents in batched_data:\n num_trg_word = sum(len(s[:-1]) for s in trg_sents)\n src_lengths = [len(s) for s in src_sents]\n src_seqs_var = to_input_var(src_sents, vocab.src, cuda)\n trg_seqs_var = to_input_var(trg_sents, vocab.trg, cuda)\n fact_lengths = [[len (s) for s in fact_sent] for fact_sent in fact_sents]\n fact_seqs_var = to_input_var_2d(fact_sents, vocab.src, cuda)\n\n yield {\n 'src_seq': src_seqs_var, 'src_lengths': src_lengths,\n 'fact_seq': fact_seqs_var, 'fact_lengths': fact_lengths,\n 'trg_seq': trg_seqs_var[:, :-1],\n 'target': trg_seqs_var[:, 1:],\n 'num_trg_word': num_trg_word, 'num_trg_seq': len(trg_sents)\n }","def example_generator(data_path, single_pass, device_rank,data_as_tf_example=True):\n\t\t\n\trandom.seed(device_rank+1)\n\tif data_as_tf_example:\n\t\tepoch = 0\n\t\twhile True:\n\t\t\tfilelist = glob.glob(data_path) # get the list of datafiles\n\t\t\tassert filelist, ('Error: Empty filelist at %s' % data_path) # check filelist isn't empty\n\t\t\tif single_pass:\n\t\t\t\tfilelist = sorted(filelist)\n\t\t\telse:\n\t\t\t\trandom.shuffle(filelist)\n\t\t\t#tf.logging.info(filelist)\t\t\t\t\n\t\t\tfor file_no, f in enumerate(filelist):\n\t\t\t\treader = open(f, 'rb')\n\t\t\t\tall_examples = []\n\t\t\t\twhile True:\n\t\t\t\t\tlen_bytes = reader.read(8)\n\t\t\t\t\tif not len_bytes: \n\t\t\t\t\t\tif not single_pass:\n\t\t\t\t\t\t\trandom.shuffle(all_examples)\n\t\t\t\t\t\tfor k in all_examples:\n\t\t\t\t\t\t\tyield example_pb2.Example.FromString(k), epoch\n\t\t\t\t\t\tbreak # finished reading this file\n\t\t\t\t\t\n\t\t\t\t\tstr_len = struct.unpack('q', len_bytes)[0]\n\t\t\t\t\texample_str = struct.unpack('%ds' % str_len, reader.read(str_len))[0]\n\t\t\t\t\tall_examples.append(example_str)\n\n\t\t\tif single_pass:\n\t\t\t\tprint \"example_generator completed reading all datafiles. No more data.\"\n\t\t\t\tbreak\n\n\t\t\n\telse:\n\t\t#pickle format\n\t\twhile True:\n\t\t\tif single_pass:\n\t\t\t\tfor data_ in data_path:\n\t\t\t\t\tfor i in data_:\n\t\t\t\t\t\tyield i\n\t\t\telse:\n\t\t\t\trandom.shuffle(data_path)\n\t\t\t\tfor data_ in data_path:\n\t\t\t\t\tnew_data = data_\n\t\t\t\t\tx = np.arange(len(new_data))\n\t\t\t\t\tnp.random.shuffle(x)\n\t\t\t\t\t# random.shuffle(new_data)\n\t\t\t\t\tfor i in x:\n\t\t\t\t\t\tyield new_data[i]\n\t\t\tif single_pass:\n\t\t\t\tbreak","def __iter__(self):\n\n # Open the data reader\n self.data.open()\n\n starts = np.arange(self.start, self.stop, self.chunksize)\n for a, b in zip_longest(starts, starts[1:], fillvalue=self.stop):\n yield self.data.read(a, b, **self.kwargs)","def batch_iter(X, *tensors, batch_size=256):\n idxs = torch.randperm(X.size(0))\n if X.is_cuda:\n idxs = idxs.cuda()\n for batch_idxs in idxs.split(batch_size):\n res = [X[batch_idxs]]\n for tensor in tensors:\n res.append(tensor[batch_idxs])\n yield res","def __call__(self, batch_size=20, shuffle=True, augment=True):\r\n\r\n if batch_size < 1:\r\n raise ValueError(\"batch_size must be more than 1.\")\r\n if shuffle:\r\n self.shuffle()\r\n\r\n for start in range(0, self.length, batch_size):\r\n batch = self.perm(start, start+batch_size)\r\n if augment:\r\n assert self._augmenter is not None, \"you have to set an augmenter.\"\r\n yield self._augmenter.augment_dataset(batch, method=[ia.ImageAugmenter.NONE, ia.ImageAugmenter.FLIP])\r\n else:\r\n yield batch","def get_batch(self):\n result = self.buffer_.get()\n self._async_next()\n return result","def get_batch(self):\n result = self.buffer_.get()\n self._async_next()\n return result","def make_data_iterator(dataset, batch_size):\n \n if dataset == '8gaussians':\n scale = 2.\n centers = [\n (1, 0),\n (-1, 0),\n (0, 1),\n (0, -1),\n (1. / np.sqrt(2), 1. / np.sqrt(2)),\n (1. / np.sqrt(2), -1. / np.sqrt(2)),\n (-1. / np.sqrt(2), 1. / np.sqrt(2)),\n (-1. / np.sqrt(2), -1. / np.sqrt(2))\n ]\n centers = [(scale * x, scale * y) for x, y in centers]\n while True:\n dataset = []\n for i in range(batch_size):\n point = np.random.randn(2) * .2\n center = random.choice(centers)\n point[0] += center[0]\n point[1] += center[1]\n dataset.append(point)\n dataset = torch.Tensor(dataset)\n dataset /= 1.414 # stdev\n yield dataset\n \n elif dataset == 'sine':\n while True:\n noise = 0.2\n x = torch.linspace(-4, 4, batch_size, dtype=torch.float32)\n y = np.sin(x) + noise*np.random.randn(*x.shape)\n yield torch.stack([x, y], dim=1)\n \n elif dataset == 'heteroscedastic':\n theta = torch.linspace(0, 2, batch_size)\n x = np.exp(theta)*np.tan(0.1*theta)\n while True:\n b = (0.001 + 0.5 * np.abs(x)) * np.random.normal(1, 1, batch_size)\n y = np.exp(theta)*np.sin(0.1*theta) + b\n yield torch.stack([x, y], dim=1)\n \n elif dataset == 'moon':\n noise = 0.1\n while True:\n data, _ = sklearn.datasets.make_moons(n_samples=batch_size,\n noise=noise)\n yield torch.Tensor(data)\n \n elif dataset == 'helix':\n noise = 0.2\n while True:\n t = torch.linspace(0, 20, batch_size)\n x = np.cos(t)\n x2 = np.sin(t) + noise * np.random.randn(*x.shape)\n \n yield torch.stack([x, x2, t], dim=1)\n \n elif dataset == 'circle':\n while True:\n t = np.random.random(batch_size) * 2 * np.pi - np.pi\n length = 1 - np.random.random(batch_size)*0.4\n x = torch.Tensor(np.multiply(np.cos(t), length))\n y = torch.Tensor(np.multiply(np.sin(t), length))\n \n yield torch.stack([x, y], dim=1)\n\n elif dataset == '2spirals':\n while True:\n z = torch.randn(batch_size, 2)\n n = torch.sqrt(torch.rand(batch_size // 2)) * 540 * (2 * math.pi) / 360\n d1x = - torch.cos(n) * n + torch.rand(batch_size // 2) * 0.5\n d1y = torch.sin(n) * n + torch.rand(batch_size // 2) * 0.5\n x = torch.cat([torch.stack([ d1x, d1y], dim=1),\n torch.stack([-d1x, -d1y], dim=1)], dim=0) / 3\n yield x + 0.1*z","def forward_batch(model, batch, device):\n inputs, labels = (batch, batch)\n inputs, labels = inputs.to(device), labels.to(device)\n\n outputs = model(inputs, labels=labels)\n\n return outputs[:2]","def _batch_iter(line_iter, n_objects, batchsize, dtype):\n new_dt = np.dtype(dtype.descr + [('mask',bool), ('weight', float)])\n batch = np.zeros((batchsize, n_objects), dtype=new_dt)\n sample_n = 0\n for array, wt in line_iter:\n\n n_missing = n_objects - array.size\n\n array.resize(n_objects, refcheck=False)\n batch[sample_n, :] = array\n batch['weight'][sample_n, :] = wt\n if n_missing > 0:\n batch['mask'][sample_n, -n_missing:] = True\n\n sample_n += 1\n if sample_n == batchsize:\n yield batch\n sample_n = 0\n batch.fill(0)\n yield batch[:sample_n, ...]","def next(self):\n if self._curr_batch + 1 > self.num_batches:\n # no more batches in current iteration through data set so start\n # new epoch ready for another pass and indicate iteration is at end\n self.new_epoch()\n raise StopIteration()\n # create an index slice corresponding to current batch number\n batch_slice = slice(self._curr_batch * self.batch_size,\n (self._curr_batch + 1) * self.batch_size)\n inputs_batch = self.inputs[batch_slice]\n targets_batch = self.targets[batch_slice]\n # target_ids_global = self.target_ids[batch_slice]\n target_ids_batch = self.target_ids[batch_slice]\n self._curr_batch += 1\n\n batch_inputs, batch_target_ids, batch_targets = \\\n self.transform_batch(inputs_batch, target_ids_batch, targets_batch)\n\n return batch_inputs, batch_targets, batch_target_ids","def next_batch(self):\n # Whether an epoch is done.\n done = False\n samples = []\n for _ in range(self.batch_size):\n # Indeed, `>` will not occur.\n if self.ptr >= self.dataset_size:\n done = True\n break\n else:\n self.ptr += 1\n sample = self.enqueuer.queue.get()\n samples.append(sample)\n # print 'queue size: {}'.format(self.enqueuer.queue.qsize())\n # Indeed, `>` will not occur.\n if self.ptr >= self.dataset_size:\n done = True\n return samples, done","def process(self, data_batch: Any, data_samples: Sequence[dict]) -> None:","def next(self, batch_size):\n if self.batch_id == len(self.data):\n self.batch_id = 0\n batch_data = (self.data[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n batch_labels = (self.labels[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n batch_seqlen = (self.seqlen[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n self.batch_id = min(self.batch_id + batch_size, len(self.data))\n\n\n return batch_data, batch_labels, batch_seqlen","def next(self, batch_size):\n if self.batch_id == len(self.data):\n self.batch_id = 0\n batch_data = (self.data[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n batch_labels = (self.labels[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n batch_seqlen = (self.seqlen[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\n self.batch_id = min(self.batch_id + batch_size, len(self.data))\n\n\n return batch_data, batch_labels, batch_seqlen","def csv_data():\r\n for row in data:\r\n writer.writerow(row)\r\n csv_data = read_and_flush()\r\n yield csv_data","def data_generator():\n msg = Message(Message.ADD, queue.uuid, queue)\n PROVIDER_MQ.put(msg)\n keep_running = True\n while keep_running:\n try:\n chunk = queue.get()\n yield chunk\n except Empty:\n app.logger.info('Queue empty. Ending stream')\n keep_running = False","def get_batch(self, data, batch_size):\n minibatch, size_so_far = [], 0\n for ex in data:\n minibatch.append(ex)\n size_so_far = self.simple_batch_size_fn(ex, len(minibatch))\n if size_so_far == batch_size:\n yield minibatch\n minibatch, size_so_far = [], 0\n elif size_so_far > batch_size:\n yield minibatch[:-1]\n minibatch, size_so_far = minibatch[-1:], self.simple_batch_size_fn(ex, 1)\n if minibatch:\n yield minibatch","def prepare_epoch(dataset):\n print(\"[-] Epoch Start\")\n\n i = 0\n for sample in range(len(dataset)):\n if sample <= i + BATCH_SIZE-1:\n continue\n\n batch = []\n for i in range(i, i+BATCH_SIZE):\n batch.append(get_image(dataset[i], OUT_SIZE, CHANNELS))\n\n i += BATCH_SIZE + 1\n\n batch_images = np.array(batch).astype(np.float32)\n yield (batch_images, batch_images)\n print(\"i: {}, s: {}\".format(i, sample))\n\n print(\"[+] Epoch complete\")","def next_batch(memory_dim, batch_length, task):\n batch_inputs = []\n batch_targets = []\n difficulty = 0.0\n for _ in xrange(batch_length):\n # Values as integers\n random_input = task.generate_random_input(difficulty, memory_dim)\n target_output = task.run(random_input)\n\n # Values as MxM memories\n batch_inputs.append(values_to_distribution_memory(random_input))\n batch_targets.append(values_to_distribution_memory(target_output))\n\n return batch_inputs, batch_targets"],"string":"[\n \"def __iter__(self):\\n for b in self.dl: \\n yield to_device(b, self.device) # yield pauses the execution, not store values in memory, forgets about them once iterated\\n # no need to remove batch of data from device, done automatically\",\n \"def __iter__(self):\\n while True:\\n if self.stop:\\n return\\n for item in self.get_next_batch():\\n yield item\",\n \"def __iter__(self):\\n batch = []\\n for sample in self.dataset:\\n batch.append(sample)\\n if len(batch) == self.size:\\n yield self.transform(batch)\\n batch = []\\n if batch:\\n # the last batch may be less then batch size.\\n yield self.transform(batch)\",\n \"def __iter__(self):\\n for b in self.dl:\\n yield to_device(b, self.device)\",\n \"def __iter__(self):\\r\\n for b in self.dl:\\r\\n yield to_device(b, self.device)\",\n \"def one_shot_iterator(dataloader):\\n while True:\\n for data in dataloader:\\n yield data\",\n \"def __iter__(self):\\n for batch in self.iterator:\\n yield Batch.from_iterator_batch(batch, self.pad_index, self.sos_index, self.eos_index)\",\n \"def cycle_loader(loader, device):\\n while True:\\n for batch in loader:\\n # NOTE this is an adhoc solution\\n batch.src = (batch.src[0].to(device), batch.src[1].to(device))\\n batch.tgt = batch.tgt.to(device)\\n logit, indices = batch.bert_topk\\n batch.bert_topk = (logit.to(device), indices.to(device))\\n yield batch\",\n \"def batch_data(data, state_queue):\\n while True:\\n time.sleep(BATCH_SLEEP_TIME)\\n state_queue.put((\\\"device_values\\\", [data]))\",\n \"def data_gen(\\n v: int, batch: int, nbatches: int, device: torch.device = torch.device(\\\"cpu\\\")\\n) -> Iterator[Batch]: # TODO bad name\\n for i in range(nbatches):\\n data = np.random.randint(1, v, size=(batch, 10))\\n data[:, 0] = 1\\n src: LongTensorType = torch.from_numpy(data)\\n tgt: LongTensorType = torch.from_numpy(data)\\n src, tgt = src.to(device), tgt.to(device)\\n yield Batch(src, tgt, 0)\",\n \"def next_batch(self):\\n\\n while self.cap.isOpened():\\n flag, frame = self.cap.read()\\n yield frame\",\n \"def _train_batch(self):\\n\\n # start epoch\\n for i, (source, target) in enumerate(self.train_dataset):\\n result = self._batch_iter(source, target, i)\\n\\n # yield\\n yield result\",\n \"def next_batch(self):\\n for nb in xrange(self.num_batches):\\n if self.batch_end < self.full_len:\\n batch_X_raw = self.full_X[self.batch_start:self.batch_end]\\n batch_y_raw = self.full_y[self.batch_start:self.batch_end]\\n else:\\n batch_X_raw = self.full_X[self.batch_start:]\\n batch_y_raw = self.full_y[self.batch_start:]\\n batch_X, batch_y = pad_sort_data(batch_X_raw, batch_y_raw)\\n self.batch_start = self.batch_end\\n self.batch_end += self.batch_size\\n yield batch_X, batch_y\",\n \"def data_generator(delta=1, batch_size=32):\\n while True:\\n yield generate_samples(delta=delta, n=batch_size)\",\n \"def data_parallel(self, batch_size, inputs):\\n inputs = list(inputs)\\n\\n # quick path: only one device, do not slice\\n if len(self.work_devices) == 1:\\n assert(self.main_device == self.work_devices[0])\\n yield self.main_device, False, tuple(inputs)\\n\\n # slow path: multi-GPUs\\n else:\\n # the GPUs are not in the same group, place variables on CPU\\n if self.main_device not in self.work_devices:\\n yield self.main_device, True, tuple(inputs)\\n\\n # build the paralleled computation graph for each device\\n with tf.name_scope('data_parallel') as ns:\\n pass # generate a name scope to place our data slicing ops\\n\\n k = len(self.work_devices)\\n for i, device in enumerate(self.work_devices):\\n dev_inputs = []\\n with tf.name_scope(ns + 'tower_gpu_{}'.format(i)):\\n for inp in inputs:\\n slice_len = (batch_size + k - 1) // k\\n low, high = slice_len * i, slice_len * (i + 1)\\n dev_inputs.append(inp[low: high])\\n yield device, False, tuple(dev_inputs)\",\n \"def batch_iter(data, batch_size):\\n random.shuffle(data)\\n num_data = len(data)\\n num_steps = int(math.ceil(num_data/float(batch_size)))\\n rem = num_data%batch_size\\n for step in range(num_steps):\\n if step == (num_steps-1) and rem !=0:\\n data_batch = data[(step)*batch_size:]\\n random.shuffle(data)\\n data_batch.extend(data[:(batch_size - rem)])\\n board, target, ko = process_mini_batch(data_batch)\\n yield (board, target, ko)\\n else:\\n data_batch = data[step*batch_size:(step+1)*batch_size]\\n board, target, ko = process_mini_batch(data_batch)\\n yield (board, target, ko)\",\n \"def __iter__(self):\\n\\n # collector will fetch chunksize array for each 'get' call\\n collector = FIFOArray(self.chunksize, self.axis)\\n\\n # make tmp array to hold generated subarrs\\n tmp = []\\n tmp_size = 0\\n for subarr in self.data(**self.kwargs):\\n\\n tmp.append(subarr)\\n tmp_size += subarr.shape[self.axis]\\n\\n # if tmp exceeds chunksize put in collector\\n if tmp_size >= self.chunksize:\\n arr = np.concatenate(tmp, axis=self.axis)\\n collector.put(arr)\\n\\n # fetch chunksize till not full\\n while collector.full():\\n yield collector.get()\\n\\n # place leftover back into tmp and empty collector\\n tmp = [collector.queue]\\n tmp_size = collector.qsize()\\n collector.queue = np.array([])\\n\\n else:\\n\\n # append to tmp again\\n continue\\n\\n # else runs after normal loop exit -- required here\\n else: #pylint: disable=useless-else-on-loop\\n\\n # yield whatever is left in tmp (its below chunksize)\\n remaining = np.concatenate(tmp, axis=self.axis)\\n if remaining.size > 0:\\n yield remaining\",\n \"def _next_batch(self, loader: CustomIterator) -> list:\\n return self.mover.move(loader.__next__())\",\n \"def data_generator(self, data):\\n X, y = [], []\\n while 1:\\n np.random.shuffle(data)\\n for line in data:\\n img = Image.open(line[0])\\n img = img.resize((32, 16))\\n img = np.asarray(img, dtype=np.float32)\\n img = img / 128. - 1.\\n img = np.transpose(img, (2, 0, 1)) \\n X.append(img)\\n y.append(line[1])\\n if len(X) == self.config.batch_size:\\n batch = (np.asarray(X), np.asarray(y))\\n X = []\\n y = []\\n yield batch\",\n \"def __iter__(self):\\n for sample in self.data:\\n yield sample\",\n \"def test_minibatches(self):\\n batch_size = self.params['batch_size']\\n start_index = 0\\n while start_index + batch_size < 500:\\n end_index = start_index + batch_size\\n yield self.input[start_index:end_index], self.y[start_index:end_index]\\n start_index = end_index\",\n \"def batch_data(cls, train_data, train_labels, batch_size):\\n for batch in range(int(np.ceil(train_data.shape[0] / batch_size))):\\n start = batch_size * batch\\n end = start + batch_size\\n if end > train_data.shape[0]:\\n yield batch, (train_data[start:train_data.shape[0]], \\\\\\n train_labels[start:train_data.shape[0]])\\n else:\\n yield batch, (train_data[start:end], \\\\\\n train_labels[start:end])\",\n \"def __iter__(self):\\n for i, data in enumerate(self.dataloader):\\n if i * self.opt.batch_size >= self.opt.max_dataset_size:\\n break\\n yield data\",\n \"def next_batch(self, batch_size):\\n raise NotImplementedError\",\n \"def __iter__(self):\\n\\t\\tfor i, data in enumerate(self.dataloader):\\n\\t\\t\\tif i * self.opt.batch_size >= self.opt.max_dataset_size:\\n\\t\\t\\t\\tbreak\\n\\t\\t\\tyield data\",\n \"def _next(self):\\n batch_start, batch_end = self.batch_start, self.batch_start + self.batch_size\\n X_batch, y_batch = self.X[batch_start:batch_end], self.y[batch_start:batch_end]\\n X_batch, y_batch = self.process_batch(X_batch, y_batch)\\n if batch_end > self.X.shape[0]:\\n self.batch_start = 0\\n else:\\n self.batch_start = batch_end\\n return X_batch, y_batch\",\n \"def __iter__(self):\\n batch = []\\n for idx in self._sampler:\\n batch.append(idx)\\n if len(batch) == self._batch_size:\\n batch = sum(batch, [])\\n yield batch\\n batch = []\\n if len(batch) > 0:\\n batch = sum(batch, [])\\n yield batch\",\n \"def sample_batch(pid, args, batch_queue, port_dict, device, actor_id_to_ip_dataport, local_size, cache_array):\\n def recv_data(k, data_stream, actor_set, real_data_tasks_i):\\n for real_data in data_stream:\\n tmp = []\\n tmp.append(real_data.state)\\n tmp.append(real_data.action)\\n tmp.append(real_data.reward)\\n tmp.append(real_data.next_state)\\n tmp.append(real_data.done)\\n tmp.append(actor_set[k]['w'][real_data.idx])\\n tmp.append(actor_set[k]['i'][real_data.idx])\\n tmp.append(actor_set[k]['t'][real_data.idx])\\n tmp.append(real_data.timestamp)\\n local_dict[actor_set[k]['i'][real_data.idx]] = tmp\\n cache_array[actor_set[k]['i'][real_data.idx]] |= 2**pid\\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.state), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n real_data_tasks_i['states'].append(decom_state) #.to(device))\\n real_data_tasks_i['actions'].append(torch.LongTensor([real_data.action])) #.to(device))\\n real_data_tasks_i['rewards'].append(torch.FloatTensor([real_data.reward])) #.to(device))\\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(real_data.next_state), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n real_data_tasks_i['next_states'].append(decom_next_state) #.to(device))\\n real_data_tasks_i['dones'].append(torch.FloatTensor([real_data.done])) #.to(device))\\n real_data_tasks_i['batch_weights'].append(torch.FloatTensor([actor_set[k]['w'][real_data.idx]])) #.to(device))\\n real_data_tasks_i['batch_idxes'].append(actor_set[k]['i'][real_data.idx])\\n # is the data overwrited?\\n real_data_tasks_i['batch_timestamp_store'].append(actor_set[k]['t'][real_data.idx])\\n real_data_tasks_i['batch_timestamp_real'].append(real_data.timestamp)\\n conn = grpc.insecure_channel(port_dict['replay_ip'] + ':' + port_dict['sampleDataPort'])\\n client = apex_data_pb2_grpc.SampleDataStub(channel=conn)\\n local_dict = {}\\n while True:\\n batch_timestamp_real = []\\n batch_timestamp_store = []\\n batch_weights = []\\n batch_idxes = []\\n\\n states, actions, rewards, next_states, dones = [], [], [], [], []\\n\\n res_batch = client.Send(apex_data_pb2.SampleDataRequest(batch_size=args.batch_size, beta = args.beta))\\n actor_ids, data_ids, timestamps, weights, idxes = res_batch.actor_ids, res_batch.data_ids, res_batch.timestamp, res_batch.weights, res_batch.idxes\\n actor_set = {}\\n cached_value = {'states':{},'actions':{},'rewards':{},'next_states':{},'dones':{},'batch_weights':{},'batch_idxes':{},'batch_timestamp_store':{},'batch_timestamp_real':{}}\\n for i in range(len(actor_ids)):\\n set_a = actor_set.get(actor_ids[i], False)\\n if set_a == False:\\n actor_set[actor_ids[i]] = {}\\n set_a = actor_set[actor_ids[i]]\\n set_a['d'] = []\\n set_a['w'] = []\\n set_a['i'] = []\\n set_a['t'] = []\\n cached_value['states'][actor_ids[i]] = []\\n cached_value['actions'][actor_ids[i]] = []\\n cached_value['rewards'][actor_ids[i]] = []\\n cached_value['next_states'][actor_ids[i]] = []\\n cached_value['dones'][actor_ids[i]] = []\\n cached_value['batch_weights'][actor_ids[i]] = []\\n cached_value['batch_idxes'][actor_ids[i]] = []\\n cached_value['batch_timestamp_store'][actor_ids[i]] = []\\n cached_value['batch_timestamp_real'][actor_ids[i]] = []\\n cache_id = actor_ids[i]*local_size+data_ids[i]\\n cache_trans = cache_array[cache_id]\\n if cache_trans & 2**pid == 0:\\n set_a['d'].append(data_ids[i])\\n set_a['w'].append(weights[i])\\n set_a['i'].append(idxes[i])\\n set_a['t'].append(timestamps[i])\\n if cache_trans == 0 and local_dict.get(cache_id, False) != False:\\n del local_dict[cache_id]\\n else:\\n try:\\n state_tmp = local_dict[cache_id][0]\\n action_tmp = local_dict[cache_id][1]\\n reward_tmp = local_dict[cache_id][2] \\n next_state_tmp = local_dict[cache_id][3] \\n done_tmp = local_dict[cache_id][4] \\n batch_weight_tmp = local_dict[cache_id][5] \\n batch_idx_tmp = local_dict[cache_id][6] \\n batch_store_tmp = local_dict[cache_id][7] \\n batch_real_tmp = local_dict[cache_id][8] \\n decom_state = torch.FloatTensor(np.frombuffer(zlib.decompress(state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n cached_value['states'][actor_ids[i]].append(decom_state)\\n cached_value['actions'][actor_ids[i]].append(torch.LongTensor([action_tmp]))\\n cached_value['rewards'][actor_ids[i]].append(torch.FloatTensor([reward_tmp]))\\n decom_next_state = torch.FloatTensor(np.frombuffer(zlib.decompress(next_state_tmp), dtype=np.uint8).reshape((1, 4, 84, 84)))\\n cached_value['next_states'][actor_ids[i]].append(decom_next_state)\\n cached_value['dones'][actor_ids[i]].append(torch.FloatTensor([done_tmp]))\\n cached_value['batch_weights'][actor_ids[i]].append(torch.FloatTensor([batch_weight_tmp]))\\n cached_value['batch_idxes'][actor_ids[i]].append(batch_idx_tmp)\\n cached_value['batch_timestamp_store'][actor_ids[i]].append(batch_store_tmp)\\n cached_value['batch_timestamp_real'][actor_ids[i]].append(batch_real_tmp)\\n except:\\n set_a['d'].append(data_ids[i])\\n set_a['w'].append(weights[i])\\n set_a['i'].append(idxes[i])\\n set_a['t'].append(timestamps[i])\\n real_data_links = {}\\n real_data_tasks = {}\\n for k, v in actor_set.items():\\n actor_ip, data_port = actor_id_to_ip_dataport[k]\\n conn_actor = grpc.insecure_channel(actor_ip + ':' + data_port)\\n client_actor = apex_data_pb2_grpc.SendRealDataStub(channel=conn_actor)\\n real_data_links[k] = client_actor.Send(apex_data_pb2.RealBatchRequest(idxes=v['d']))\\n real_data_tasks[k] = {}\\n real_data_tasks[k]['states'] = cached_value['states'][k]\\n real_data_tasks[k]['actions'] = cached_value['actions'][k]\\n real_data_tasks[k]['rewards'] = cached_value['rewards'][k]\\n real_data_tasks[k]['next_states'] = cached_value['next_states'][k]\\n real_data_tasks[k]['dones'] = cached_value['dones'][k]\\n real_data_tasks[k]['batch_weights'] = cached_value['batch_weights'][k]\\n real_data_tasks[k]['batch_idxes'] = cached_value['batch_idxes'][k]\\n real_data_tasks[k]['batch_timestamp_store'] = cached_value['batch_timestamp_store'][k]\\n real_data_tasks[k]['batch_timestamp_real'] = cached_value['batch_timestamp_real'][k]\\n threads = []\\n for k, v in real_data_links.items():\\n t = threading.Thread(target=recv_data, args=(k, v, actor_set, real_data_tasks[k],))\\n threads.append(t)\\n t.start()\\n\\n for t in threads:\\n t.join()\\n\\n for k, v in real_data_tasks.items():\\n states += v['states']\\n actions += v['actions']\\n rewards += v['rewards']\\n next_states += v['next_states']\\n dones += v['dones']\\n batch_weights += v['batch_weights']\\n batch_idxes += v['batch_idxes']\\n batch_timestamp_real += v['batch_timestamp_real']\\n batch_timestamp_store += v['batch_timestamp_store']\\n\\n states = torch.cat(states,0).to(device)\\n actions = torch.cat(actions,0).to(device)\\n rewards = torch.cat(rewards,0).to(device)\\n next_states = torch.cat(next_states,0).to(device)\\n dones = torch.cat(dones,0).to(device)\\n batch_weights = torch.cat(batch_weights,0).to(device)\\n\\n batch = [states, actions, rewards, next_states, dones, batch_weights, batch_idxes]\\n batch_queue.put(batch)\\n data, batch = None, None\",\n \"def next_simple_dataset(dataset, batch_size: int, datatype):\\n while True:\\n x_batch = []\\n y_batch = []\\n for i in range(batch_size):\\n try:\\n x, y, data_unit, index = create_xy(dataset, datatype)\\n # x = normalize(x)\\n x_batch.append(x)\\n y_batch.append(y)\\n except StopIteration:\\n break\\n x_batch, y_batch = np.array(x_batch), np.array(y_batch)\\n if datatype != DataType.test:\\n x_batch = SEQ_CVXTZ.augment_images(x_batch).astype(\\\"float32\\\")\\n x_batch = np.array([normalize(x) for x in x_batch])\\n # org_shape = x_batch.shape\\n # org_width = x_batch.shape[1]\\n # corner = int((org_width - ROI_IMAGE_SIZE) // 2)\\n # print(f\\\"0: org_shape:{org_shape} x_batch:{x_batch.shape} corner:{corner}\\\")\\n # x_batch = x_batch[:, corner:(org_width - corner), corner:(org_width - corner), :]\\n # resized_x_batch = []\\n # for x in x_batch:\\n # img = Image.fromarray(np.uint8(x))\\n # img = img.resize((IMAGE_SIZE, IMAGE_SIZE), Image.LANCZOS)\\n # resized_x_batch.append(normalize(np.array(img)))\\n # print(f\\\"1: org_shape:{org_shape} corner:{corner} x_batch:{x_batch.shape}\\\")\\n # yield np.array(resized_x_batch), y_batch\\n yield np.array(x_batch), y_batch\",\n \"def __iter__(self):\\n for batch in self.data:\\n batch_size = len(batch)\\n X, e1, e2, dist1, dist2, e1_pos, e2_pos, y = list(zip(*batch))\\n\\n x_len = max(len(x) for x in X)\\n x_ids = torch.LongTensor(batch_size, x_len).fill_(0)\\n dist1_padded = torch.LongTensor(batch_size, x_len).fill_(0)\\n dist2_padded = torch.LongTensor(batch_size, x_len).fill_(0)\\n for i, doc in enumerate(X):\\n x_ids[i, :len(doc)] = torch.LongTensor(doc)\\n\\n dist1_padded[i, :len(doc)] = torch.LongTensor(dist1[i])\\n dist1_padded[i, len(doc):] = torch.LongTensor([pos(e1_pos[i][1] - idx) for idx, _ in enumerate(x_ids[i][len(doc):], start=len(doc))])\\n\\n dist2_padded[i, :len(doc)] = torch.LongTensor(dist2[i])\\n dist2_padded[i, len(doc):] = torch.LongTensor([pos(e2_pos[i][1] - idx) for idx, _ in enumerate(x_ids[i][len(doc):], start=len(doc))])\\n\\n e1_tensor = torch.LongTensor(e1)\\n e2_tensor = torch.LongTensor(e2)\\n\\n y_tensor = torch.LongTensor(y)\\n\\n if self.gpu:\\n x_ids = x_ids.pin_memory()\\n e1_tensor = e1_tensor.pin_memory()\\n e2_tensor = e2_tensor.pin_memory()\\n dist1_padded = dist1_padded.pin_memory()\\n dist2_padded = dist2_padded.pin_memory()\\n y_tensor = y_tensor.pin_memory()\\n\\n yield (x_ids, e1_tensor, e2_tensor, dist1_padded, dist2_padded, y_tensor)\",\n \"def next(self):\\n\\t\\t# keep looping until we reach our batch size\\n\\t\\twhile True:\\n\\t\\t\\tret = self.get_batch()\\n\\t\\t\\tself.index += self.batch_size\\n\\t\\t\\tif self.index >= len(self.texts) - self.batch_size:\\n\\t\\t\\t\\tself.index = 0\\n\\t\\t\\tyield ret\",\n \"def get_batches(self, batch_size):\\n if self.data.shape[0] % batch_size != 0:\\n raise RuntimeError('num of data tuples is not a multiple of batch size')\\n num_batch = self.data.shape[0] // batch_size\\n for b in range(num_batch):\\n yield self.data[b*batch_size:(b+1)*batch_size, :], \\\\\\n self.target[b*batch_size:(b+1)*batch_size, :]\",\n \"def _chunk_data(self):\\n for n in range(0, len(self.data) + 1, len(self.data) //\\n self.num_of_chunks):\\n yield self.data[0 + n:len(self.data) // self.num_of_chunks + n]\",\n \"def batch_iter(input_data,batch_size):\\r\\n batch_ids,batch_mask,batch_segment,batch_label=[],[],[],[]\\r\\n for features in input_data:\\r\\n if len(batch_ids) == batch_size:\\r\\n yield batch_ids,batch_mask,batch_segment,batch_label\\r\\n batch_ids, batch_mask, batch_segment, batch_label = [], [], [], []\\r\\n\\r\\n batch_ids.append(features['input_ids'])\\r\\n batch_mask.append(features['input_mask'])\\r\\n batch_segment.append(features['segment_ids'])\\r\\n batch_label.append(features['label_ids'])\\r\\n\\r\\n if len(batch_ids) != 0:\\r\\n yield batch_ids, batch_mask, batch_segment, batch_label\",\n \"def __iter__(self):\\n batch = []\\n for i_batch in range(self.episode_num):\\n classes = torch.randperm(len(self.idx_list))[: self.way_num]\\n for c in classes:\\n idxes = self.idx_list[c.item()]\\n pos = torch.randperm(idxes.size(0))[: self.image_num]\\n batch.append(idxes[pos])\\n if len(batch) == self.episode_size * self.way_num:\\n batch = torch.stack(batch).reshape(-1)\\n yield batch\\n batch = []\",\n \"def _next_batch(self, loader) -> list:\\n return self.mover.move(loader.__next__())\",\n \"def unchunk():\\n\\n @filters\\n def _dagpype_internal_fn_act(target):\\n try:\\n while True:\\n a = (yield)\\n if len(a) == 0:\\n continue\\n if a.ndim == 1:\\n for i in range(a.shape[0]):\\n target.send(a[i]) \\n else:\\n for i in range(a.shape[0]):\\n target.send(tuple(a[i]))\\n except GeneratorExit:\\n if len(l) > 0:\\n target.send(numpy.array(l, dtype = dtype_)) \\n \\n return _dagpype_internal_fn_act\",\n \"def next(self) -> Iterable[RecordBatch]:\\n for batch in self._parent_operator.next():\\n args = self._process_arguments(self._arguments, batch=batch)\\n yield self._kernel(batch, args)\",\n \"def create_batch(client, generator: DataGenerator):\\n try:\\n event_data_batch = client.create_batch()\\n for device in generator.devices:\\n # event_data_batch.add(EventData(gen.generate_payload(device)))\\n event_data_batch.add(EventData(generator.generate_payload(device)))\\n return event_data_batch\\n except Exception as e:\\n print(str(e))\",\n \"def next_batch(self, batch_size):\\r\\n raise NotImplementedError\",\n \"def forward(self, batch):\\n raise NotImplementedError\",\n \"def forward(self, batch):\\n raise NotImplementedError\",\n \"def gen_batches(data, batch_size):\\n data = np.array(data)\\n\\n for i in range(0, data.shape[0], batch_size):\\n yield data[i:i+batch_size]\",\n \"def generate_batches(data, labels, batch_size):\\n for start in range(0, len(data), batch_size):\\n yield Tensor(data[start:start+batch_size, ...]), Tensor(labels[start:start+batch_size, ...])\",\n \"def batches(data, batch_size) -> list:\\n rv = []\\n for idx, line in enumerate(data):\\n if idx != 0 and idx % batch_size == 0:\\n yield rv\\n rv = []\\n rv.append(line)\\n yield rv\",\n \"def data_iter(data, batch_size, shuffle=True, sort=True):\\n\\n batched_data = []\\n if shuffle: np.random.shuffle(data)\\n batched_data.extend(list(batch_slice(data, batch_size, sort)))\\n\\n if shuffle: np.random.shuffle(batched_data)\\n for batch in batched_data:\\n yield batch\",\n \"def static_batch(data, batch_size=16):\\n buf = []\\n for sample in data:\\n buf.append(sample)\\n if len(buf) >= batch_size:\\n yield buf\\n buf = []\\n if len(buf) > 0:\\n yield buf\",\n \"def enhancer_iterator(self, data, labels, batch_size, num_steps):\\n def seq_to_ints(seq):\\n return [self.vocab.word_to_index[c] for c in seq]\\n\\n # Map raw data to array of ints. if all sequences are the same length L, \\n # raw_data will be N-by-L\\n mdata = np.array([seq_to_ints(i) for i in data], dtype=np.int32)\\n num_batches = len(mdata) // batch_size\\n \\n # data will have batch_len elements, each of size batch_size\\n # ASSUME FIXED SEQUENCE LENGTHS OFF 1000 FOR NOW (5/20/16)\\n # Just grab middle self.config.num_steps nucleotides\\n a = int(len(mdata[0,:])/2-self.config.num_steps/2)\\n b = int(len(mdata[0,:])/2+self.config.num_steps/2)\\n for i in range(num_batches):\\n x = mdata[batch_size*i:batch_size*(i+1),a:b]\\n if labels is not None:\\n y = labels[batch_size*i:batch_size*(i+1)]\\n else:\\n y = None\\n yield(x,y)\",\n \"def batch_generator(data, batch_size, config, target=False):\\n listsize = min(len(data.pilist), 10)\\n while 1:\\n nrd = np.arange(len(data.pilist))\\n np.random.shuffle(nrd)\\n for ri in range(len(data.pilist) // listsize):\\n cpilist = [data.pilist[nrd[rid]] for rid in range(listsize * ri, listsize * (ri + 1))]\\n data.loadpatch(config, cpilist)\\n if target:\\n data_l = [data.xarray, np.zeros(shape=(len(data.xarray), 1))]\\n else:\\n data_l = [data.xarray, data.yarray - 1]\\n\\n for di in range(len(data_l[0])):\\n data_l[0][di][:, :, :, 0] = data_l[0][di][:, :, :, 0] / np.max(data_l[0][di][:, :, :, 0])\\n\\n for repi in range(3):\\n data_l = shuffle_aligned_list(data_l)\\n\\n batch_count = 0\\n while True:\\n if batch_count * batch_size + batch_size >= len(data_l[0]):\\n if len(data.pilist) // listsize != 1:\\n print('list end', ri * listsize)\\n break\\n else:\\n batch_count = 0\\n\\n start = batch_count * batch_size\\n end = start + batch_size\\n batch_count += 1\\n yield [d[start:end] for d in data_l]\",\n \"def next(self):\\n if self.curr_idx == len(self.idx):\\n raise StopIteration\\n\\n # Fetch the index\\n i = self.idx[self.curr_idx]\\n self.curr_idx += 1\\n\\n # Get labels & ids\\n labels = self.ndlabels[i:i+self.batch_size]\\n users = self.ndusers[i:i + self.batch_size]\\n items = self.nditems[i:i + self.batch_size]\\n\\n # Get feature arrays\\n if self.create_batches:\\n user_features = self.nduserfeatures[i:i+self.batch_size]\\n item_features = self.nditemfeatures[i:i+self.batch_size]\\n else:\\n # Create user feature arrays\\n user_features = mx.ndarray.take(a=self.unique_user_features, indices=users)\\n item_features = mx.ndarray.take(a=self.unique_item_features, indices=items)\\n\\n return mx.io.DataBatch([user_features, item_features], [labels], index = users, pad=0,\\n provide_data=[mx.io.DataDesc(name=self.data_names[0], shape=user_features.shape),\\n mx.io.DataDesc(name=self.data_names[1], shape=item_features.shape)],\\n provide_label=[mx.io.DataDesc(name=self.label_names[0], shape=labels.shape)])\",\n \"def next_batch(self, batch_size):\\n batch_data = np.zeros([batch_size,] + list(self.example_shape))\\n for i in range(batch_size):\\n index = self.q.pop()\\n batch_data[i,...] = self.data[index]\\n if len(self.q)==0:\\n self.__new_epoch()\\n\\n return batch_data\",\n \"def chunk_generator( callback, request, c ):\",\n \"def next(self):\\n #print('next')\\n batch_size = self.batch_size\\n batch_data = nd.empty((batch_size,)+self.data_shape)\\n batch_label = nd.empty((batch_size,)+self.label_shape)\\n i = 0\\n #self.cutoff = random.randint(800,1280)\\n try:\\n while i < batch_size:\\n #print('N', i)\\n data, label, annot = self.next_sample()\\n R = self.get_data(data, label, annot)\\n if R is None:\\n continue\\n data_out, label_out, flip_data_out, flip_label_out = R\\n if not self.use_coherent:\\n data = nd.array(data_out)\\n data = nd.transpose(data, axes=(2, 0, 1))\\n label = nd.array(label_out)\\n #print(data.shape, label.shape)\\n batch_data[i][:] = data\\n batch_label[i][:] = label\\n i += 1\\n else:\\n data = nd.array(data_out)\\n data = nd.transpose(data, axes=(2, 0, 1))\\n label = nd.array(label_out)\\n data2 = nd.array(flip_data_out)\\n data2 = nd.transpose(data2, axes=(2, 0, 1))\\n label2 = nd.array(flip_label_out)\\n #M = nd.array(M)\\n #print(data.shape, label.shape)\\n batch_data[i][:] = data\\n batch_label[i][:] = label\\n #i+=1\\n j = i+self.per_batch_size//2\\n batch_data[j][:] = data2\\n batch_label[j][:] = label2\\n i += 1\\n if j%self.per_batch_size==self.per_batch_size-1:\\n i = j+1\\n except StopIteration:\\n if i<batch_size:\\n raise StopIteration\\n\\n #return {self.data_name : batch_data,\\n # self.label_name : batch_label}\\n #print(batch_data.shape, batch_label.shape)\\n return mx.io.DataBatch([batch_data], [batch_label], batch_size - i)\",\n \"def _batch_iter(self, source, target, i: int):\\n # send data to device\\n source = source.to(self.device)\\n target = target.to(self.device)\\n\\n # the result and loss\\n result = self.model(source)\\n loss = self.criterion(result, target)\\n\\n # optimization and backward\\n self.optimizer.zero_grad()\\n loss.backward()\\n self.optimizer.step()\\n\\n # update the loss\\n self.epoch_loss.update(loss.item(), source.size(0))\\n\\n # print the information\\n if self.info:\\n print(f\\\"\\\\rEpoch: { self.epoch } | Batch: { i } | loss: { self.epoch_loss.avg }\\\", end=\\\"\\\")\\n\\n # clean the data\\n del source, target\\n\\n return result\",\n \"def process(self, data_itr):\\n for data in data_itr:\\n self.update(data)\\n while True:\\n try:\\n out = self.next()\\n yield out\\n except StopIteration:\\n break\",\n \"def data_gen(voc_size, batch, nbatches, seq_len = 15):\\r\\n for i in range(nbatches):\\r\\n # (batch_size, seq_len)\\r\\n data = torch.from_numpy(\\r\\n np.random.randint(1, voc_size, size=(batch, seq_len)))\\r\\n data[:, 0] = 1 # add start token\\r\\n src = Variable(data, requires_grad=False)\\r\\n tgt = Variable(data, requires_grad=False)\\r\\n yield Batch(src, tgt, 0) # Accessed by next function one by one\\r\",\n \"def prepare_batch(batchdata, device=None, non_blocking=False):\\n return batchdata[\\\"hand\\\"].to(device=device, non_blocking=non_blocking)\",\n \"def _batcher(self, rows):\\n row_count = 0\\n batch = []\\n batch_count = 1\\n\\n total_rows_modified = 0\\n throttle_count = 0\\n\\n i = 0\\n for row in rows:\\n if row_count > self.batch_size - 1:\\n logger.debug(f\\\"row_count={row_count} batch_size={self.batch_size} and batch={len(batch)}\\\")\\n # Yield the previous batch\\n yield batch\\n\\n # Start the new batch\\n batch = []\\n batch.append(row)\\n row_count = 1\\n\\n batch_count += 1\\n # break # toggle to load one batch only\\n else:\\n row_count += 1\\n batch.append(row)\\n\\n # Put in a sleep timer to throttle how hard we hit the database\\n if self.throttle_time and self.throttle_size and (throttle_count > self.throttle_size - 1):\\n logger.info(f\\\"Sleeping for {self.throttle_time} seconds... row: {i}\\\")\\n time.sleep(int(self.throttle_time))\\n throttle_count = 0\\n elif self.throttle_time and self.throttle_size:\\n throttle_count += 1\\n i += 1\\n\\n yield batch\",\n \"def TransferMemorySequence():\\r\\n pass\",\n \"def get_batch(iterator, batch_size):\\n while True:\\n center_batch = np.zeros(batch_size, dtype = np.uint32)\\n target_batch = np.zeros((batch_size, 1), dtype = np.uint32)\\n for index in range(batch_size):\\n center_batch[index], target_batch[index] = next(iterator)\\n\\n yield center_batch, target_batch\",\n \"def data_generator(batch_size, preprocessor, x, y):\\n num_examples = len(x)\\n examples = zip(x, y)\\n examples = sorted(examples, key = lambda x: x[0].shape[0])\\n end = num_examples - batch_size + 1\\n batches = [examples[i:i+batch_size]\\n for i in range(0, end, batch_size)]\\n random.shuffle(batches)\\n while True:\\n for batch in batches:\\n x, y = zip(*batch)\\n yield preprocessor.process(x, y)\",\n \"def __iter__(self):\\n worker_info = torch.utils.data.get_worker_info()\\n num_batches = len(self.coords_batcher)\\n if worker_info is None:\\n # In single-processing mode\\n start, end = 0, num_batches\\n else:\\n worker_id = worker_info.id\\n num_workers = worker_info.num_workers\\n shard_size = int(np.ceil(num_batches / num_workers))\\n start = shard_size * worker_id\\n end = min(start + shard_size, num_batches)\\n return (self.get_batch(i) for i in range(start, end))\",\n \"def next_batch(self, batch_size=8):\\n raise NotImplementedError()\",\n \"def next_batch(x, y, batch_size):\\n\\n def as_batch(data, start, count):\\n part = []\\n for i in range(start, start + count):\\n part.append(data[i])\\n return np.array(part)\\n\\n for i in range(0, len(x)-batch_size, batch_size):\\n yield as_batch(x, i, batch_size), as_batch(y, i, batch_size)\",\n \"def _next(self):\\n batch_start, batch_end = self.batch_start, self.batch_start + self.batch_size\\n if batch_end > self.X.shape[0]:\\n self.shuffle()\\n return self._next()\\n else:\\n batch_indices = self.indices[batch_start:batch_end]\\n X_batch, y_batch = self.X[batch_indices], self.y[batch_indices]\\n X_batch, y_batch = self.process_batch(X_batch, y_batch)\\n self.batch_start = batch_end\\n return X_batch, y_batch\",\n \"def next_batch(self, batch_size, shuffle=True):\",\n \"def __iter__(self):\\n\\n collector = FIFOArray(self.chunksize, self.axis)\\n for arr, maskarr in zip(self.data, self.mask):\\n\\n if not np.any(maskarr):\\n continue\\n\\n filtered = np.take(arr, np.flatnonzero(maskarr), axis=self.axis)\\n collector.put(filtered)\\n\\n while collector.full():\\n\\n yield collector.get()\\n\\n # else runs after normal loop exit -- required here\\n else: #pylint: disable=useless-else-on-loop\\n\\n if collector.qsize() > 0:\\n\\n yield collector.get()\",\n \"def _simulation_data_iterator(res, path, data):\\n res.append(data)\",\n \"def serialize_batches():\\n # a set of variables for the state of current batch which will be converted to Arrow\\n # RecordBatch.\\n pdfs = []\\n state_pdfs = []\\n pdf_data_cnt = 0\\n state_data_cnt = 0\\n\\n return_schema = None\\n\\n for data in iterator:\\n # data represents the result of each call of user function\\n packaged_result = data[0]\\n\\n # There are two results from the call of user function:\\n # 1) iterator of pandas DataFrame (output)\\n # 2) updated state instance\\n pdf_iter = packaged_result[0][0]\\n state = packaged_result[0][1]\\n\\n # This is static and won't change across batches.\\n return_schema = packaged_result[1]\\n\\n for pdf in pdf_iter:\\n # We ignore empty pandas DataFrame.\\n if len(pdf) > 0:\\n pdf_data_cnt += len(pdf)\\n pdfs.append(pdf)\\n\\n # If the total number of records in current batch exceeds the configured\\n # threshold, time to construct the Arrow RecordBatch from the batch.\\n if pdf_data_cnt > self.arrow_max_records_per_batch:\\n batch = construct_record_batch(\\n pdfs, pdf_data_cnt, return_schema, state_pdfs, state_data_cnt\\n )\\n\\n # Reset the variables to start with new batch for further data.\\n pdfs = []\\n state_pdfs = []\\n pdf_data_cnt = 0\\n state_data_cnt = 0\\n\\n yield batch\\n\\n # This has to be performed 'after' evaluating all elements in iterator, so that\\n # the user function has been completed and the state is guaranteed to be updated.\\n state_pdf = construct_state_pdf(state)\\n\\n state_pdfs.append(state_pdf)\\n state_data_cnt += 1\\n\\n # processed all output, but current batch may not be flushed yet.\\n if pdf_data_cnt > 0 or state_data_cnt > 0:\\n batch = construct_record_batch(\\n pdfs, pdf_data_cnt, return_schema, state_pdfs, state_data_cnt\\n )\\n\\n yield batch\",\n \"def __call__(self):\\n if self.numbatches is None:\\n pool = self.pooler()\\n if self.batchsize is None:\\n self.batchsize = self.pooler.nInPool()\\n self.numbatches = self.pooler.nInPool()//self.batchsize\\n for i in xrange(self.numbatches):\\n pool = self.pooler()\\n self._reset_batch()\\n if self.samplemethod == 'balance' and len(self.keysamplers)>0:\\n batchinds,keyids = self._samplebalanced(pool)\\n elif self.samplemethod == 'uniform':\\n batchinds,keyids = self._sampleuniform(pool)\\n else:\\n batchinds,keyids = self._samplesequential(i)\\n batch = self._extractInds(pool,batchinds,keyids)\\n for k in batch:\\n batch[k][np.isnan(batch[k])] = self.nanreplacement\\n yield batch\",\n \"def _next(self):\\n i = 0\\n while i < self.size:\\n yield self.data[i]\\n i += 1\",\n \"def __iter__(self):\\n if not self.big:\\n yield self.D[:self.P, :self.d], self.labels[:self.P]\\n else:\\n for i in range(int(self.P / self.batch_size)):\\n yield self.D[i*self.batch_size:min(self.P, (i+1)*self.batch_size), :self.d].cuda(), \\\\\\n self.labels[i*self.batch_size:min(self.P, (i+1)*self.batch_size), :].cuda()\",\n \"def in_memory_rechunk(\\n inputs: List[Tuple[core.ChunkKey, xarray.Dataset]],\\n target_chunks: Mapping[str, int],\\n) -> Iterator[Tuple[core.ChunkKey, xarray.Dataset]]:\\n key, dataset = consolidate_chunks(inputs)\\n yield from split_chunks(key, dataset, target_chunks)\",\n \"def batch_data(source, target, batch_size):\\n for batch_i in range(0, len(source)//batch_size):\\n start_i = batch_i * batch_size\\n source_batch = source[start_i:start_i + batch_size]\\n target_batch = target[start_i:start_i + batch_size]\\n yield np.array(pad_sentence_batch(source_batch)), np.array(pad_sentence_batch(target_batch))\",\n \"def train_minibatches(self):\\n batch_size = self.params['batch_size']\\n start_index = 0\\n while start_index + batch_size < 500:\\n end_index = start_index + batch_size\\n yield self.input[start_index:end_index], self.y[start_index:end_index]\\n start_index = end_index\",\n \"def next(self):\\n # Keeps under lock only the mechanism which advances\\n # the indexing of each batch.\\n with self.lock:\\n index_array = next(self.index_generator)\\n # The transformation of images is not under thread lock\\n # so it can be done in parallel\\n return self._get_batches_of_transformed_samples(index_array)\",\n \"def data_generator(dataset, config, shuffle=True, augment=False, augmentation=None, batch_size=1):\\n b = 0 # batch item index\\n image_index = -1\\n image_ids = np.copy(dataset.image_ids)\\n error_count = 0\\n\\n # Keras requires a generator to run indefinately.\\n while True:\\n try:\\n # Increment index to pick next image. Shuffle if at the start of an epoch.\\n image_index = (image_index + 1) % len(image_ids)\\n if shuffle and image_index == 0:\\n np.random.shuffle(image_ids)\\n\\n # Get GT bounding boxes and masks for image.\\n image_id = image_ids[image_index]\\n image, gt_class_ids = load_image_gt(dataset, config, image_id, augment=augment,\\n augmentation=augmentation)\\n\\n # Init batch arrays\\n if b == 0:\\n batch_images = np.zeros(\\n (batch_size,) + image.shape, dtype=np.float32)\\n batch_gt_class_ids = np.zeros(\\n (batch_size, config.MAX_GT_INSTANCES), dtype=np.int32)\\n\\n # Add to batch\\n batch_images[b] = mold_image(image.astype(np.float32), config)\\n batch_gt_class_ids[b, gt_class_ids] = 1\\n b += 1\\n\\n # Batch full?\\n if b >= batch_size:\\n inputs = [batch_images, batch_gt_class_ids]\\n outputs = []\\n\\n yield inputs, outputs\\n\\n # start a new batch\\n b = 0\\n except (GeneratorExit, KeyboardInterrupt):\\n raise\\n except:\\n # Log it and skip the image\\n logging.exception(\\\"Error processing image {}\\\".format(\\n dataset.image_info[image_id]))\\n error_count += 1\\n if error_count > 5:\\n raise\",\n \"def data_iter(data, vocab, batch_size, shuffle=True, cuda=False):\\n\\n buckets = defaultdict(list)\\n for pair in data:\\n src_sent = pair[0]\\n buckets[len(src_sent)].append(pair)\\n\\n batched_data = []\\n for src_len in buckets:\\n tuples = buckets[src_len]\\n if shuffle: np.random.shuffle(tuples)\\n batched_data.extend(list(HNCMDataLoader.batch_slice(tuples, batch_size)))\\n\\n if shuffle:\\n np.random.shuffle(batched_data)\\n for src_sents, trg_sents, fact_sents in batched_data:\\n num_trg_word = sum(len(s[:-1]) for s in trg_sents)\\n src_lengths = [len(s) for s in src_sents]\\n src_seqs_var = to_input_var(src_sents, vocab.src, cuda)\\n trg_seqs_var = to_input_var(trg_sents, vocab.trg, cuda)\\n fact_lengths = [[len (s) for s in fact_sent] for fact_sent in fact_sents]\\n fact_seqs_var = to_input_var_2d(fact_sents, vocab.src, cuda)\\n\\n yield {\\n 'src_seq': src_seqs_var, 'src_lengths': src_lengths,\\n 'fact_seq': fact_seqs_var, 'fact_lengths': fact_lengths,\\n 'trg_seq': trg_seqs_var[:, :-1],\\n 'target': trg_seqs_var[:, 1:],\\n 'num_trg_word': num_trg_word, 'num_trg_seq': len(trg_sents)\\n }\",\n \"def example_generator(data_path, single_pass, device_rank,data_as_tf_example=True):\\n\\t\\t\\n\\trandom.seed(device_rank+1)\\n\\tif data_as_tf_example:\\n\\t\\tepoch = 0\\n\\t\\twhile True:\\n\\t\\t\\tfilelist = glob.glob(data_path) # get the list of datafiles\\n\\t\\t\\tassert filelist, ('Error: Empty filelist at %s' % data_path) # check filelist isn't empty\\n\\t\\t\\tif single_pass:\\n\\t\\t\\t\\tfilelist = sorted(filelist)\\n\\t\\t\\telse:\\n\\t\\t\\t\\trandom.shuffle(filelist)\\n\\t\\t\\t#tf.logging.info(filelist)\\t\\t\\t\\t\\n\\t\\t\\tfor file_no, f in enumerate(filelist):\\n\\t\\t\\t\\treader = open(f, 'rb')\\n\\t\\t\\t\\tall_examples = []\\n\\t\\t\\t\\twhile True:\\n\\t\\t\\t\\t\\tlen_bytes = reader.read(8)\\n\\t\\t\\t\\t\\tif not len_bytes: \\n\\t\\t\\t\\t\\t\\tif not single_pass:\\n\\t\\t\\t\\t\\t\\t\\trandom.shuffle(all_examples)\\n\\t\\t\\t\\t\\t\\tfor k in all_examples:\\n\\t\\t\\t\\t\\t\\t\\tyield example_pb2.Example.FromString(k), epoch\\n\\t\\t\\t\\t\\t\\tbreak # finished reading this file\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\tstr_len = struct.unpack('q', len_bytes)[0]\\n\\t\\t\\t\\t\\texample_str = struct.unpack('%ds' % str_len, reader.read(str_len))[0]\\n\\t\\t\\t\\t\\tall_examples.append(example_str)\\n\\n\\t\\t\\tif single_pass:\\n\\t\\t\\t\\tprint \\\"example_generator completed reading all datafiles. No more data.\\\"\\n\\t\\t\\t\\tbreak\\n\\n\\t\\t\\n\\telse:\\n\\t\\t#pickle format\\n\\t\\twhile True:\\n\\t\\t\\tif single_pass:\\n\\t\\t\\t\\tfor data_ in data_path:\\n\\t\\t\\t\\t\\tfor i in data_:\\n\\t\\t\\t\\t\\t\\tyield i\\n\\t\\t\\telse:\\n\\t\\t\\t\\trandom.shuffle(data_path)\\n\\t\\t\\t\\tfor data_ in data_path:\\n\\t\\t\\t\\t\\tnew_data = data_\\n\\t\\t\\t\\t\\tx = np.arange(len(new_data))\\n\\t\\t\\t\\t\\tnp.random.shuffle(x)\\n\\t\\t\\t\\t\\t# random.shuffle(new_data)\\n\\t\\t\\t\\t\\tfor i in x:\\n\\t\\t\\t\\t\\t\\tyield new_data[i]\\n\\t\\t\\tif single_pass:\\n\\t\\t\\t\\tbreak\",\n \"def __iter__(self):\\n\\n # Open the data reader\\n self.data.open()\\n\\n starts = np.arange(self.start, self.stop, self.chunksize)\\n for a, b in zip_longest(starts, starts[1:], fillvalue=self.stop):\\n yield self.data.read(a, b, **self.kwargs)\",\n \"def batch_iter(X, *tensors, batch_size=256):\\n idxs = torch.randperm(X.size(0))\\n if X.is_cuda:\\n idxs = idxs.cuda()\\n for batch_idxs in idxs.split(batch_size):\\n res = [X[batch_idxs]]\\n for tensor in tensors:\\n res.append(tensor[batch_idxs])\\n yield res\",\n \"def __call__(self, batch_size=20, shuffle=True, augment=True):\\r\\n\\r\\n if batch_size < 1:\\r\\n raise ValueError(\\\"batch_size must be more than 1.\\\")\\r\\n if shuffle:\\r\\n self.shuffle()\\r\\n\\r\\n for start in range(0, self.length, batch_size):\\r\\n batch = self.perm(start, start+batch_size)\\r\\n if augment:\\r\\n assert self._augmenter is not None, \\\"you have to set an augmenter.\\\"\\r\\n yield self._augmenter.augment_dataset(batch, method=[ia.ImageAugmenter.NONE, ia.ImageAugmenter.FLIP])\\r\\n else:\\r\\n yield batch\",\n \"def get_batch(self):\\n result = self.buffer_.get()\\n self._async_next()\\n return result\",\n \"def get_batch(self):\\n result = self.buffer_.get()\\n self._async_next()\\n return result\",\n \"def make_data_iterator(dataset, batch_size):\\n \\n if dataset == '8gaussians':\\n scale = 2.\\n centers = [\\n (1, 0),\\n (-1, 0),\\n (0, 1),\\n (0, -1),\\n (1. / np.sqrt(2), 1. / np.sqrt(2)),\\n (1. / np.sqrt(2), -1. / np.sqrt(2)),\\n (-1. / np.sqrt(2), 1. / np.sqrt(2)),\\n (-1. / np.sqrt(2), -1. / np.sqrt(2))\\n ]\\n centers = [(scale * x, scale * y) for x, y in centers]\\n while True:\\n dataset = []\\n for i in range(batch_size):\\n point = np.random.randn(2) * .2\\n center = random.choice(centers)\\n point[0] += center[0]\\n point[1] += center[1]\\n dataset.append(point)\\n dataset = torch.Tensor(dataset)\\n dataset /= 1.414 # stdev\\n yield dataset\\n \\n elif dataset == 'sine':\\n while True:\\n noise = 0.2\\n x = torch.linspace(-4, 4, batch_size, dtype=torch.float32)\\n y = np.sin(x) + noise*np.random.randn(*x.shape)\\n yield torch.stack([x, y], dim=1)\\n \\n elif dataset == 'heteroscedastic':\\n theta = torch.linspace(0, 2, batch_size)\\n x = np.exp(theta)*np.tan(0.1*theta)\\n while True:\\n b = (0.001 + 0.5 * np.abs(x)) * np.random.normal(1, 1, batch_size)\\n y = np.exp(theta)*np.sin(0.1*theta) + b\\n yield torch.stack([x, y], dim=1)\\n \\n elif dataset == 'moon':\\n noise = 0.1\\n while True:\\n data, _ = sklearn.datasets.make_moons(n_samples=batch_size,\\n noise=noise)\\n yield torch.Tensor(data)\\n \\n elif dataset == 'helix':\\n noise = 0.2\\n while True:\\n t = torch.linspace(0, 20, batch_size)\\n x = np.cos(t)\\n x2 = np.sin(t) + noise * np.random.randn(*x.shape)\\n \\n yield torch.stack([x, x2, t], dim=1)\\n \\n elif dataset == 'circle':\\n while True:\\n t = np.random.random(batch_size) * 2 * np.pi - np.pi\\n length = 1 - np.random.random(batch_size)*0.4\\n x = torch.Tensor(np.multiply(np.cos(t), length))\\n y = torch.Tensor(np.multiply(np.sin(t), length))\\n \\n yield torch.stack([x, y], dim=1)\\n\\n elif dataset == '2spirals':\\n while True:\\n z = torch.randn(batch_size, 2)\\n n = torch.sqrt(torch.rand(batch_size // 2)) * 540 * (2 * math.pi) / 360\\n d1x = - torch.cos(n) * n + torch.rand(batch_size // 2) * 0.5\\n d1y = torch.sin(n) * n + torch.rand(batch_size // 2) * 0.5\\n x = torch.cat([torch.stack([ d1x, d1y], dim=1),\\n torch.stack([-d1x, -d1y], dim=1)], dim=0) / 3\\n yield x + 0.1*z\",\n \"def forward_batch(model, batch, device):\\n inputs, labels = (batch, batch)\\n inputs, labels = inputs.to(device), labels.to(device)\\n\\n outputs = model(inputs, labels=labels)\\n\\n return outputs[:2]\",\n \"def _batch_iter(line_iter, n_objects, batchsize, dtype):\\n new_dt = np.dtype(dtype.descr + [('mask',bool), ('weight', float)])\\n batch = np.zeros((batchsize, n_objects), dtype=new_dt)\\n sample_n = 0\\n for array, wt in line_iter:\\n\\n n_missing = n_objects - array.size\\n\\n array.resize(n_objects, refcheck=False)\\n batch[sample_n, :] = array\\n batch['weight'][sample_n, :] = wt\\n if n_missing > 0:\\n batch['mask'][sample_n, -n_missing:] = True\\n\\n sample_n += 1\\n if sample_n == batchsize:\\n yield batch\\n sample_n = 0\\n batch.fill(0)\\n yield batch[:sample_n, ...]\",\n \"def next(self):\\n if self._curr_batch + 1 > self.num_batches:\\n # no more batches in current iteration through data set so start\\n # new epoch ready for another pass and indicate iteration is at end\\n self.new_epoch()\\n raise StopIteration()\\n # create an index slice corresponding to current batch number\\n batch_slice = slice(self._curr_batch * self.batch_size,\\n (self._curr_batch + 1) * self.batch_size)\\n inputs_batch = self.inputs[batch_slice]\\n targets_batch = self.targets[batch_slice]\\n # target_ids_global = self.target_ids[batch_slice]\\n target_ids_batch = self.target_ids[batch_slice]\\n self._curr_batch += 1\\n\\n batch_inputs, batch_target_ids, batch_targets = \\\\\\n self.transform_batch(inputs_batch, target_ids_batch, targets_batch)\\n\\n return batch_inputs, batch_targets, batch_target_ids\",\n \"def next_batch(self):\\n # Whether an epoch is done.\\n done = False\\n samples = []\\n for _ in range(self.batch_size):\\n # Indeed, `>` will not occur.\\n if self.ptr >= self.dataset_size:\\n done = True\\n break\\n else:\\n self.ptr += 1\\n sample = self.enqueuer.queue.get()\\n samples.append(sample)\\n # print 'queue size: {}'.format(self.enqueuer.queue.qsize())\\n # Indeed, `>` will not occur.\\n if self.ptr >= self.dataset_size:\\n done = True\\n return samples, done\",\n \"def process(self, data_batch: Any, data_samples: Sequence[dict]) -> None:\",\n \"def next(self, batch_size):\\n if self.batch_id == len(self.data):\\n self.batch_id = 0\\n batch_data = (self.data[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n batch_labels = (self.labels[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n batch_seqlen = (self.seqlen[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n self.batch_id = min(self.batch_id + batch_size, len(self.data))\\n\\n\\n return batch_data, batch_labels, batch_seqlen\",\n \"def next(self, batch_size):\\n if self.batch_id == len(self.data):\\n self.batch_id = 0\\n batch_data = (self.data[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n batch_labels = (self.labels[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n batch_seqlen = (self.seqlen[self.batch_id:min(self.batch_id + batch_size, len(self.data))])\\n self.batch_id = min(self.batch_id + batch_size, len(self.data))\\n\\n\\n return batch_data, batch_labels, batch_seqlen\",\n \"def csv_data():\\r\\n for row in data:\\r\\n writer.writerow(row)\\r\\n csv_data = read_and_flush()\\r\\n yield csv_data\",\n \"def data_generator():\\n msg = Message(Message.ADD, queue.uuid, queue)\\n PROVIDER_MQ.put(msg)\\n keep_running = True\\n while keep_running:\\n try:\\n chunk = queue.get()\\n yield chunk\\n except Empty:\\n app.logger.info('Queue empty. Ending stream')\\n keep_running = False\",\n \"def get_batch(self, data, batch_size):\\n minibatch, size_so_far = [], 0\\n for ex in data:\\n minibatch.append(ex)\\n size_so_far = self.simple_batch_size_fn(ex, len(minibatch))\\n if size_so_far == batch_size:\\n yield minibatch\\n minibatch, size_so_far = [], 0\\n elif size_so_far > batch_size:\\n yield minibatch[:-1]\\n minibatch, size_so_far = minibatch[-1:], self.simple_batch_size_fn(ex, 1)\\n if minibatch:\\n yield minibatch\",\n \"def prepare_epoch(dataset):\\n print(\\\"[-] Epoch Start\\\")\\n\\n i = 0\\n for sample in range(len(dataset)):\\n if sample <= i + BATCH_SIZE-1:\\n continue\\n\\n batch = []\\n for i in range(i, i+BATCH_SIZE):\\n batch.append(get_image(dataset[i], OUT_SIZE, CHANNELS))\\n\\n i += BATCH_SIZE + 1\\n\\n batch_images = np.array(batch).astype(np.float32)\\n yield (batch_images, batch_images)\\n print(\\\"i: {}, s: {}\\\".format(i, sample))\\n\\n print(\\\"[+] Epoch complete\\\")\",\n \"def next_batch(memory_dim, batch_length, task):\\n batch_inputs = []\\n batch_targets = []\\n difficulty = 0.0\\n for _ in xrange(batch_length):\\n # Values as integers\\n random_input = task.generate_random_input(difficulty, memory_dim)\\n target_output = task.run(random_input)\\n\\n # Values as MxM memories\\n batch_inputs.append(values_to_distribution_memory(random_input))\\n batch_targets.append(values_to_distribution_memory(target_output))\\n\\n return batch_inputs, batch_targets\"\n]","isConversational":false},"negative_scores":{"kind":"list like","value":["0.80380756","0.681836","0.6770501","0.6756917","0.6754636","0.67446125","0.6652299","0.6626395","0.6621452","0.65598935","0.65336037","0.6526245","0.64669716","0.6410819","0.63767326","0.63346195","0.6315828","0.63127494","0.6302321","0.6286355","0.627287","0.6272642","0.6260652","0.62356985","0.62288654","0.62252927","0.6220692","0.6212761","0.6197703","0.61774296","0.61748695","0.6171532","0.6169881","0.61675245","0.61557204","0.6153821","0.6127181","0.61268765","0.6125732","0.6124463","0.6118403","0.6118403","0.6114333","0.6106637","0.60798174","0.60657865","0.60586697","0.6045184","0.6040677","0.6030159","0.6025972","0.6023496","0.6022595","0.60183406","0.60130954","0.6012649","0.6009346","0.6008823","0.60043305","0.6001723","0.59951913","0.59924954","0.59916186","0.5991304","0.5988138","0.59674877","0.5949666","0.59475535","0.5945445","0.59439564","0.59294826","0.59209627","0.59203225","0.5917931","0.59136105","0.58894026","0.588546","0.5884782","0.5879699","0.5869765","0.58620435","0.58606267","0.58585125","0.58585125","0.5848224","0.5844037","0.58339214","0.58276963","0.5825809","0.5821429","0.5819109","0.5819109","0.5815904","0.5797195","0.5790434","0.5787621","0.578722"],"string":"[\n \"0.80380756\",\n \"0.681836\",\n \"0.6770501\",\n \"0.6756917\",\n \"0.6754636\",\n \"0.67446125\",\n \"0.6652299\",\n \"0.6626395\",\n \"0.6621452\",\n \"0.65598935\",\n \"0.65336037\",\n \"0.6526245\",\n \"0.64669716\",\n \"0.6410819\",\n \"0.63767326\",\n \"0.63346195\",\n \"0.6315828\",\n \"0.63127494\",\n \"0.6302321\",\n \"0.6286355\",\n \"0.627287\",\n \"0.6272642\",\n \"0.6260652\",\n \"0.62356985\",\n \"0.62288654\",\n \"0.62252927\",\n \"0.6220692\",\n \"0.6212761\",\n \"0.6197703\",\n \"0.61774296\",\n \"0.61748695\",\n \"0.6171532\",\n \"0.6169881\",\n \"0.61675245\",\n \"0.61557204\",\n \"0.6153821\",\n \"0.6127181\",\n \"0.61268765\",\n \"0.6125732\",\n \"0.6124463\",\n \"0.6118403\",\n \"0.6118403\",\n \"0.6114333\",\n \"0.6106637\",\n \"0.60798174\",\n \"0.60657865\",\n \"0.60586697\",\n \"0.6045184\",\n \"0.6040677\",\n \"0.6030159\",\n \"0.6025972\",\n \"0.6023496\",\n \"0.6022595\",\n \"0.60183406\",\n \"0.60130954\",\n \"0.6012649\",\n \"0.6009346\",\n \"0.6008823\",\n \"0.60043305\",\n \"0.6001723\",\n \"0.59951913\",\n \"0.59924954\",\n \"0.59916186\",\n \"0.5991304\",\n \"0.5988138\",\n \"0.59674877\",\n \"0.5949666\",\n \"0.59475535\",\n \"0.5945445\",\n \"0.59439564\",\n \"0.59294826\",\n \"0.59209627\",\n \"0.59203225\",\n \"0.5917931\",\n \"0.59136105\",\n \"0.58894026\",\n \"0.588546\",\n \"0.5884782\",\n \"0.5879699\",\n \"0.5869765\",\n \"0.58620435\",\n \"0.58606267\",\n \"0.58585125\",\n \"0.58585125\",\n \"0.5848224\",\n \"0.5844037\",\n \"0.58339214\",\n \"0.58276963\",\n \"0.5825809\",\n \"0.5821429\",\n \"0.5819109\",\n \"0.5819109\",\n \"0.5815904\",\n \"0.5797195\",\n \"0.5790434\",\n \"0.5787621\",\n \"0.578722\"\n]","isConversational":false},"document_score":{"kind":"string","value":"0.68135273"},"document_rank":{"kind":"string","value":"3"}}}],"truncated":false,"partial":true},"paginationData":{"pageIndex":908,"numItemsPerPage":100,"numTotalItems":95772,"offset":90800,"length":100}},"jwt":"eyJhbGciOiJFZERTQSIsImtpZCI6IjVHZDBvd0g5MTM2eDZjc1FvbE1zcktNWUZoRVFoUm5PVVVybEpjOUhaUEEifQ.eyJyZWFkIjp0cnVlLCJwZXJtaXNzaW9ucyI6eyJyZXBvLmNvbnRlbnQucmVhZCI6dHJ1ZX0sImlhdCI6MTc4NTY4MDk3MSwianRpIjoiYTQ1MGVkY2YtZDI3MS00MjNkLTllYWEtMmFmMDg5YzE3MDNkIiwic3ViIjoiL2RhdGFzZXRzL25vbWljLWFpL2Nvcm5zdGFjay1weXRob24tdjEiLCJleHAiOjE3ODU2ODQ1NzEsImlzcyI6Imh0dHBzOi8vaHVnZ2luZ2ZhY2UuY28ifQ.chqdv2txuL7ETHZan7Ec9IOxsLl6g9crf7kOsqydBM_0gwnnMy1ZcL_ZC8IYUdSgDMD4hjZzA25e4C-PwHJmDA","displayUrls":true,"splitSizeSummaries":[{"config":"default","split":"train","numRows":95772,"numBytesParquet":1928375891}]},"discussionsStats":{"closed":1,"open":1,"total":2},"fullWidth":true,"hasGatedAccess":true,"hasFullAccess":true,"isEmbedded":false,"savedQueries":{"community":[],"user":[]}}">
query stringlengths 9 3.4k | document stringlengths 9 87.4k | metadata dict | negatives listlengths 4 101 | negative_scores listlengths 4 101 | document_score stringlengths 3 10 | document_rank stringclasses 102
values |
|---|---|---|---|---|---|---|
Resets all the contents stored | def tile38_setup():
client.flushall
yield None | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def reset(self):\n self._data = []",
"def reset(self):\n self.temp_data.clear()",
"def reset(self):\n # FIXME: this state does not make sense\n self.reset_creation_info()\n self.reset_document()\n self.reset_package()\n self.reset_file_stat()\n self.reset... | [
"0.79640114",
"0.79449594",
"0.7942509",
"0.79245305",
"0.78824085",
"0.78824085",
"0.78824085",
"0.78803",
"0.78803",
"0.78803",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"0.7834564",
"... | 0.0 | -1 |
return the correctness and confidences of predictions | def correct_and_confidence(model, loader, device, with_energy=True):
with torch.no_grad():
model.eval()
confidences = []
corrects = []
for inputs, labels in loader:
inputs, labels = inputs.to(device), labels.to(device)
if with_energy:
logits =... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def prediction():\r\n\r\n\r\n\tpredictVal = []\r\n\taccuracy = 0.0\r\n\r\n\t# Calculate accuracy for each class in testData\r\n\tfor item in testData:\r\n\t\tclass0Prediction = posProb / 100\r\n\t\tclass1Prediction = negProb / 100\r\n\t\t\r\n\t\t# Multiply the prior probablities for negative and positive reviews b... | [
"0.7893121",
"0.7871008",
"0.76539403",
"0.7412509",
"0.73162645",
"0.72805506",
"0.7263658",
"0.7257417",
"0.71888185",
"0.7166795",
"0.71611714",
"0.7130664",
"0.7118593",
"0.7114675",
"0.71143824",
"0.7077957",
"0.7068769",
"0.7055071",
"0.7054077",
"0.6949678",
"0.6939620... | 0.0 | -1 |
compute expected calibration error (ECE) | def expected_calibration_error(data_length, bucket_accs, bucket_confs, bucket_totals):
# TODO: If the docstring is longer than the function logic, usually the docstring should be shorter or it shouldn't
# be a function.
# Remember that the point of a function is to be an abstraction: something that hides co... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def diff_of_errors(self):\n self.e_of_e = self.azimuth_error - self.altitude_error\n return self.e_of_e",
"def calculate_energy_conservation_error(self):\n assert self.data is not None\n # calculate total energy at start and end of simulation\n energy_start = self.measure_total... | [
"0.7079219",
"0.6710523",
"0.6705349",
"0.66143274",
"0.6573631",
"0.64684534",
"0.64388514",
"0.64110225",
"0.6364667",
"0.63354135",
"0.6307564",
"0.62935024",
"0.6287445",
"0.62517846",
"0.6240687",
"0.61839724",
"0.6173702",
"0.6167329",
"0.61644596",
"0.61461425",
"0.613... | 0.6969385 | 1 |
Execute a command and return its value. | def delegator():
import delegator
def run(command):
if "|" in command:
res = delegator.chain(command)
else:
res = delegator.run(command)
if res.return_code == 127:
raise FileNotFoundError(res.err)
if res.return_code != 0:
raise Run... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def execute(self, command, user_context=None):\n return self.parse(command, user_context).execute()",
"def execute_command(self, command):\n return self.ssh.exec_command(command)",
"def _execute_command(self, cmd):\n LOG.info(\"Executing: %s\" % cmd)\n status, stdout, stderr = self.... | [
"0.75513715",
"0.73304343",
"0.73241794",
"0.72709864",
"0.7242012",
"0.71793664",
"0.7133625",
"0.7099978",
"0.70719355",
"0.7040908",
"0.69573665",
"0.6923248",
"0.69084376",
"0.6903973",
"0.69022304",
"0.6891859",
"0.68201387",
"0.6804588",
"0.67375743",
"0.6724226",
"0.66... | 0.0 | -1 |
A delegator to ``git`` setting the appropriate ``gitdir``. | def git_command(delegator, git_directory):
def run(cmd, **kwargs):
cmd = "git --git-dir={} {}".format(quote(git_directory), cmd)
return delegator(cmd, **kwargs)
return run | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def git_directory(delegator):\n return delegator(\"git rev-parse --git-dir\")",
"def mock_git(mocker):\n from demisto_sdk.commands.common.content import Content\n\n # Mock git working directory\n mocker.patch.object(Content, 'git')\n Content.git().working_tree_dir = TEST_CONTENT_REPO\n yield",
... | [
"0.72769815",
"0.6641427",
"0.6349675",
"0.6286185",
"0.6276211",
"0.61569196",
"0.6130058",
"0.61011934",
"0.60696644",
"0.60047036",
"0.59994465",
"0.59891456",
"0.5951636",
"0.5927444",
"0.58606523",
"0.5830278",
"0.5823502",
"0.58090305",
"0.5806452",
"0.579807",
"0.57860... | 0.66274726 | 2 |
The ``.git`` directory for the current project. | def git_directory(delegator):
return delegator("git rev-parse --git-dir") | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def getGitPath() -> osp:\n current_dir = osp.dirname(osp.realpath(__file__))\n git_dir = osp.dirname(osp.dirname(current_dir))\n return git_dir",
"def get_git_dir(tree):\n\n return os.path.join(tree, \".git\")",
"def git_dir():\n return subprocess.check_output([\"git\", \"rev-parse\", \"--git-di... | [
"0.83023816",
"0.8015751",
"0.7522234",
"0.74142706",
"0.70690876",
"0.7059461",
"0.70178574",
"0.6930033",
"0.68814874",
"0.6830864",
"0.67350674",
"0.6637658",
"0.64931214",
"0.64922035",
"0.6490604",
"0.64626306",
"0.646249",
"0.64517623",
"0.6420175",
"0.63963944",
"0.638... | 0.66105205 | 12 |
Calculates shade by converting the time passed since the start of the game to a percentage of the current day/night cycle. | def update(self):
super().update()
time_since_start = self.time_since_start()
curr_mod = time_since_start%self.game.time_cycle_secs
grade = abs(curr_mod - self.game.time_cycle_secs/2) / (self.game.time_cycle_secs/2)
color_value = grade*(255-self.game.max_darkness) + self.game.ma... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_colour_time_based(self):\n\t\treturn self.get_hex_from_rgb(0, 0, int(255.0 * (get_elapsed_seconds_simulated() / SECONDS_PER_DAY)))",
"def report_rest_percentage(self):\n self._logger.info(\"Running report for device {}\".format(self.name))\n tot_time_since_bday = (datetime.utcnow() - self.b... | [
"0.6328449",
"0.5682314",
"0.5677854",
"0.5432159",
"0.5369359",
"0.53085834",
"0.5219234",
"0.5206503",
"0.520525",
"0.5153342",
"0.5086756",
"0.5010156",
"0.5008131",
"0.50076133",
"0.5004993",
"0.5000146",
"0.49980986",
"0.49826407",
"0.49760562",
"0.49475768",
"0.4946041"... | 0.578151 | 1 |
Displays the time in the game relative to the time cycle | def message(self):
if self.display_time:
return "Time: {}".format(int(self.physics_engine.time_since_start()%self.time_cycle_secs)) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def display_time(self, time):\n pygame.draw.rect(self.screen, self.font_fgcolor, self.time_rect)\n self.screen.blit(self.small_font.render(\"Elapsed time: %.0f s\" % time, -1, (0, 0, 0)), (5, 720))\n pygame.display.update(self.time_rect)\n return",
"def ShowTime():\n ClearDisplay()... | [
"0.80268294",
"0.7536709",
"0.7264826",
"0.71524316",
"0.709622",
"0.7039029",
"0.70333403",
"0.6958243",
"0.6939595",
"0.69334245",
"0.693148",
"0.68405247",
"0.6725378",
"0.6715109",
"0.66588914",
"0.66478723",
"0.6636626",
"0.6634973",
"0.66332513",
"0.6627094",
"0.6620326... | 0.7924845 | 1 |
Creates a model, given the type of model and the output method. Adds the output layer with activation corresponding to the output type after the constructed model. This allows for easy switching between binary and multiclass classification, and prevents us from having to implement two types of each model. The model is ... | def create_model(model_type: str, input_size: Tuple[int, ...], output='binary', summary=True) -> keras.Model:
if model_type in MODELS:
inputs, outputs = MODELS[model_type](input_size)
else:
raise ValueError(f'Model type "{model_type}" not supported')
if output == 'binary':
activatio... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def build_model(input_dim, hidden_neurons, output_dim):\n\tmodel = Sequential([\n \tDense(hidden_neurons, input_dim=input_dim),\n Activation('relu'),\n Dropout(0.2),\n Dense(hidden_neurons),\n Activation('relu'),\n Dropout(0.2),\n Dense(output_dim, activation='softmax')... | [
"0.6704527",
"0.66973925",
"0.66615975",
"0.66041964",
"0.6572064",
"0.6570199",
"0.6549063",
"0.647823",
"0.6476947",
"0.6459976",
"0.6454561",
"0.6407263",
"0.64054805",
"0.638561",
"0.63830036",
"0.63781667",
"0.63442785",
"0.6344052",
"0.6315975",
"0.6313631",
"0.63134235... | 0.71340376 | 0 |
process response get from requests | def process_resp(self, msg, operation, status, index):
metric = "%s.%d.%s" % (METRIC_NAME, index, operation)
self.results.append(Event(TIMESTAMP_MILLIS(), "opentsdb", metric, msg, status))
if status == "0":
self.cause.extend(msg)
metric = "%s.%d.%s" % (METRIC_NAME, index,... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _processGETResp(self, output, request):\r\n msg = {'key' : output}\r\n\r\n self._render_GET(request, httplib.OK,\r\n 'application/json; charset=utf-8', json.dumps(msg))",
"def _process_request(self, request, response):\n ...",
"def _process_request(protocol, dom... | [
"0.73667425",
"0.692869",
"0.66959393",
"0.66323155",
"0.66323155",
"0.6591454",
"0.65617144",
"0.65557534",
"0.65041816",
"0.6460191",
"0.643903",
"0.6426434",
"0.6426434",
"0.64241606",
"0.6421691",
"0.6381393",
"0.6347633",
"0.6323247",
"0.63002414",
"0.6279552",
"0.625061... | 0.0 | -1 |
Create UID for metric, tag_key and tag_value | def create_uid(self, host, index):
msg = []
operation = "WRITE"
url = "%s%s%s" % ("http://", host, "/api/uid/assign")
payload = {"metric": [METRIC_NAME], "tagk": [TAGK], "tagv": ["%s.%d" % (TAGV, index)]}
headers = {"content-type": "application/json"}
m_uuid = False
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def create_tag_id():\n return uuid.uuid1().int",
"def create_uid():\n return random_string(5, string.hexdigits.lower())\n # return (\"%x\" % (int(time.time()) * 0x10 % 0x1000000000)\n # + random_string(7, string.hexdigits.lower()))",
"def uid_gen(prefix: str = \"\") -> str:\n uid: str = ... | [
"0.622496",
"0.5914397",
"0.5869701",
"0.5807702",
"0.5769845",
"0.5710975",
"0.56788355",
"0.56183344",
"0.56085813",
"0.5602032",
"0.5578855",
"0.5568348",
"0.555802",
"0.553758",
"0.5533039",
"0.54645866",
"0.54295546",
"0.5413362",
"0.54101545",
"0.53708315",
"0.53593594"... | 0.6909411 | 0 |
Data will be inserted into tsdb table | def write(self, host, index):
msg = []
operation = "WRITE"
if not self.create_uid(host, index):
return False
url = "%s%s%s" % ("http://", host, "/api/put")
payload = {"metric": METRIC_NAME, "timestamp": TIMESTAMP_MILLIS(), \
"value": METRIC_VAL, "tags":{TAGK: ... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def do_insert_data(self, *args):\n print(\"Provide data to insert\")\n self.connection_obj.insert_into_table(**self.__class__.populate_data())\n print(\"Data Insertion Successful\")",
"def insert(self,table,values):\n self.connect.execute(self.insert_disc[table],values)\n self.... | [
"0.72514176",
"0.7214033",
"0.70906794",
"0.70481104",
"0.7026284",
"0.70220715",
"0.7020377",
"0.69829357",
"0.69679713",
"0.6887554",
"0.68298614",
"0.6821573",
"0.6820573",
"0.6811746",
"0.6759082",
"0.6737531",
"0.6731064",
"0.67139596",
"0.67029476",
"0.669869",
"0.66986... | 0.0 | -1 |
Inserted data will be read | def read(self, host, index):
msg = []
operation = "READ"
url = "%s%s%s" % ("http://", host, "/api/query")
payload = {"start": self.test_start_timestamp, "end": TIMESTAMP_MILLIS(), "queries": [{"aggregator": "none", \
"metric": METRIC_NAME, "tags": {TAGK: "%s.%d" % (TAGV, index)}}... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def insert_data(self):\n\n pass",
"def insert_data(self) -> None:\n if self.min_insert_size > self.insert_count:\n LOG.debug(\"Not enough data for insert....\")\n return\n LOG.debug(f'Inserting {self.insert_count} records...')\n self.insert.write(self.copy_traile... | [
"0.75751936",
"0.7198535",
"0.6850029",
"0.6821989",
"0.68143404",
"0.64528984",
"0.64384943",
"0.64332664",
"0.6369282",
"0.6364045",
"0.6362229",
"0.6341396",
"0.6312636",
"0.6263611",
"0.6248877",
"0.623441",
"0.6233147",
"0.62074935",
"0.61562264",
"0.6149473",
"0.6142653... | 0.0 | -1 |
Analyze ok and ko status on hosts | def analyze_results(self, results):
ok_c = 0
ko_c = 0
for row in results:
if "opentsdb.health" not in row[2] and ".health" in row[2]:
if row[4] == "ERROR":
ko_c += 1
else:
ok_c += 1
return ok_c, ko_c | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _status(self, host):\n pass",
"def remote_status():",
"def check_ping(self):\n # Print ping status of all of your hosts, minimum padding of 8 spaces\n padding_size = max(len(max(self.hosts, key=len)) + 4, 8)\n print('{:{padding_size}}{}'.format('Host', 'Status', padding_size=pad... | [
"0.71723616",
"0.7033423",
"0.6631784",
"0.6572632",
"0.64127666",
"0.63266665",
"0.623772",
"0.6207528",
"0.619318",
"0.6136378",
"0.60922295",
"0.6076038",
"0.6061462",
"0.6061255",
"0.60085744",
"0.59968966",
"0.5984662",
"0.59797204",
"0.59797204",
"0.594375",
"0.59303474... | 0.5727689 | 40 |
Copula transformation (empirical CDF). | def ctransform(x):
xr = np.argsort(np.argsort(x)).astype(float)
xr += 1.
xr /= float(xr.shape[-1] + 1)
return xr | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def conToCelc(faren):\n return (faren - 32)*(5/9)",
"def compute_clade_probabilities():\n command = \"phyltr cat -b 10 indoeuropean.nex | phyltr clades > clades.txt\"\n subprocess.call(command, shell=True)",
"def cepstrum(input, nceps):\n # apply the Discrete Cosine Transform\n output = dct(inpu... | [
"0.5663744",
"0.56464165",
"0.55928284",
"0.5529018",
"0.5462121",
"0.5460687",
"0.53793234",
"0.53748983",
"0.5368778",
"0.5362352",
"0.53609055",
"0.5357859",
"0.53566635",
"0.53391796",
"0.53169805",
"0.5315485",
"0.5315349",
"0.5314555",
"0.5309568",
"0.5291648",
"0.52612... | 0.0 | -1 |
Copula normalization for a single vector. | def copnorm_1d(x):
assert isinstance(x, np.ndarray) and (x.ndim == 1)
return ndtri(ctransform(x)) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def normalize(self):\n self.vector /= np.linalg.norm(self.vector)",
"def normalize(v):\n return v / np.linalg.norm(v)",
"def normalize(vector):\n return vector / np.linalg.norm(vector)",
"def normalize(v):\n return np.array(v) / np.linalg.norm(v)",
"def normalize(v):\n\tdim = v.shape \n... | [
"0.7618007",
"0.74454015",
"0.73818594",
"0.7324582",
"0.723282",
"0.7182667",
"0.71597224",
"0.70688415",
"0.70092964",
"0.70035326",
"0.6974502",
"0.6947041",
"0.6946004",
"0.6910571",
"0.6910143",
"0.69073325",
"0.69067377",
"0.6889483",
"0.68736905",
"0.6866754",
"0.68652... | 0.0 | -1 |
Categorical Copula normalization for a single vector. This function apply the copnorm per categories. | def copnorm_cat_1d(x, y):
assert isinstance(x, np.ndarray) and (x.ndim == 1)
assert isinstance(y, np.ndarray) and (x.ndim == 1)
assert y.dtype in CONFIG['INT_DTYPE']
x_cop = np.zeros_like(x)
y_u = np.unique(y)
for yi in y_u:
_idx = y == yi
x_cop[_idx] = copnorm_1d(x[_idx])
re... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def normalize_comp(cvec):\n return cvec / cvec.norm()",
"def copnorm_1d(x):\n assert isinstance(x, np.ndarray) and (x.ndim == 1)\n return ndtri(ctransform(x))",
"def copnorm_1d(x):\n assert isinstance(x, np.ndarray) and (x.ndim == 1)\n return ndtri(ctransform(x))",
"def normalize(vector):\n ... | [
"0.6994723",
"0.61327696",
"0.61327696",
"0.610969",
"0.61016935",
"0.6088704",
"0.60741276",
"0.6062895",
"0.6050322",
"0.6033933",
"0.60321313",
"0.60171866",
"0.60040766",
"0.5986567",
"0.5978743",
"0.5921216",
"0.5875921",
"0.58606005",
"0.5827847",
"0.5823671",
"0.581396... | 0.6042691 | 9 |
Copula normalization for a multidimentional array. | def copnorm_nd(x, axis=-1):
assert isinstance(x, np.ndarray) and (x.ndim >= 1)
return np.apply_along_axis(copnorm_1d, axis, x) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def normalize_array(array):\n\n return array / np.sum(array, axis=1)[:, np.newaxis]",
"def __normalize_after_fft(arr):\n\n n1, n2 = arr.shape[0], arr.shape[1]\n for i in range(n1):\n for j in range(n2):\n arr[i, j] *= n1 * n2\n\n return arr",
"def _normalize(X:... | [
"0.7164983",
"0.6876631",
"0.6776384",
"0.6739342",
"0.6689978",
"0.6635549",
"0.6580354",
"0.65614873",
"0.65260637",
"0.6514676",
"0.6426658",
"0.64239883",
"0.63720423",
"0.63630325",
"0.6362653",
"0.6353261",
"0.63378096",
"0.62933934",
"0.6270285",
"0.62699866",
"0.62672... | 0.0 | -1 |
Categorical Copula normalization for multidimentional array. This function apply the copnorm per categories. | def copnorm_cat_nd(x, y, axis=-1):
assert isinstance(x, np.ndarray) and (x.ndim >= 1)
if y is None:
return copnorm_nd(x, axis=axis)
return np.apply_along_axis(copnorm_cat_1d, axis, x, y) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def copnorm_cat_1d(x, y):\n assert isinstance(x, np.ndarray) and (x.ndim == 1)\n assert isinstance(y, np.ndarray) and (x.ndim == 1)\n assert y.dtype in CONFIG['INT_DTYPE']\n x_cop = np.zeros_like(x)\n y_u = np.unique(y)\n for yi in y_u:\n _idx = y == yi\n x_cop[_idx] = copnorm_1d(x[... | [
"0.63591534",
"0.5967849",
"0.5967849",
"0.5897528",
"0.58845794",
"0.5782783",
"0.5721007",
"0.5717328",
"0.56614244",
"0.56499016",
"0.56471413",
"0.56471413",
"0.5636114",
"0.5636114",
"0.5629407",
"0.560048",
"0.5600006",
"0.55716574",
"0.55567706",
"0.55072486",
"0.54970... | 0.62532276 | 1 |
Gets training data that includes tagged sentences. Returns a dictionary that maps every word in the training set to its most frequent tag. The dictionary should have a default value. | def most_frequent_train(train_data):
### YOUR CODE HERE
tags_counts_for_each_word = {}
# Filling a dictionary from words and tag tags to their counters
# Going over the words and counting their tags appearances
for sentance in train_data:
for word, tag in sentance:
# If first tim... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def MostCommonClassBaseline(training_set, test_set):\n pos_counts_dict = defaultdict(dict)\n max_pos_dict = dict()\n test_common_tags = []\n\n # Dictionary \"pos_counts_dict\" stores a dictionary for each word that stores counts of each pos of the word\n # This loop runs for each sentence (word, pos... | [
"0.62998277",
"0.61069524",
"0.60696596",
"0.6068922",
"0.6001002",
"0.5926774",
"0.5912238",
"0.58869374",
"0.58837795",
"0.5883664",
"0.5861991",
"0.58473736",
"0.5844703",
"0.58288777",
"0.5826551",
"0.58042777",
"0.5799191",
"0.57924163",
"0.5788701",
"0.5764574",
"0.5759... | 0.7100879 | 0 |
Gets test data and tag prediction map. Returns an evaluation of the accuracy of the most frequent tagger. | def most_frequent_eval(test_set, pred_tags):
gold_tag_seqs = []
pred_tag_seqs = []
for sent in test_set:
words, true_tags = zip(*sent)
gold_tag_seqs.append(true_tags)
### YOUR CODE HERE
DEFAULT_TAG = 'O'
pred_tags_list = []
for word in words:
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def evaluate(self, test_X, test_Y):\n correct = 0\n total = 0.0\n\n for i, ((word_indices, word_char_indices), gold_tag_indices) in enumerate(zip(test_X, test_Y)):\n\n output = self.predict(word_indices, word_char_indices)\n predicted_tag_indices = [np.argmax(o.value()) f... | [
"0.68569124",
"0.66788125",
"0.6602641",
"0.647516",
"0.6442005",
"0.64303696",
"0.63846743",
"0.63640356",
"0.6305632",
"0.63004315",
"0.6276655",
"0.6212785",
"0.6211638",
"0.61958474",
"0.61904263",
"0.6182075",
"0.6165822",
"0.61639893",
"0.6145034",
"0.6125559",
"0.61052... | 0.6525517 | 3 |
The function ``is_admin`` detects whether the calling process is running with administrator/superuser privileges. It works crossplatform on either Windows NT systems or Unixbased systems. | def is_admin():
if os.name == 'nt':
try:
# Only Windows users with admin privileges can read
# the C:\windows\temp directory.
os.listdir(os.sep.join([os.environ.get('SystemRoot','C:\\windows'),'temp']))
except:
return False
else:
r... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def is_admin():\n if platform_is(WINDOWS):\n return windll.shell32.IsUserAnAdmin()\n return os.getuid() == 0",
"def is_admin():\n # type: () -> bool\n current_os_name = os.name\n\n # Works with XP SP2 +\n if current_os_name == \"nt\":\n try:\n return IsUserAnAdmin()\n ... | [
"0.8821815",
"0.8732962",
"0.8634657",
"0.80681646",
"0.78383905",
"0.7656703",
"0.75744766",
"0.7562517",
"0.75570256",
"0.74586576",
"0.74525267",
"0.74387234",
"0.74240386",
"0.74240386",
"0.74135554",
"0.73537236",
"0.7341056",
"0.73286927",
"0.7305158",
"0.7260963",
"0.7... | 0.79539204 | 4 |
Create a new tag | def create_a_tag(self, tag_id, contact_id):
data = {"contactTag":{"contact":str(contact_id),"tag":str(tag_id)}}
return self.client._post("/contactTags", json=data) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def createTag(self, authenticationToken, tag):\r\n pass",
"def create(self, name, tag):\n\n\t\turl_json = urllib.urlencode({\"name\": name, \"tag\": tag})\n\t\treturn self._create(\"/tag?json_hash=%s\" % url_json, \"tag\")",
"def make_new_tag():\n return render_template('tags/new_tag.html')",
"def add_... | [
"0.8043615",
"0.77951604",
"0.7368651",
"0.7363008",
"0.7346278",
"0.72604716",
"0.7207542",
"0.7113458",
"0.7074937",
"0.70454365",
"0.7003208",
"0.69938636",
"0.69363683",
"0.6916433",
"0.68520314",
"0.68380255",
"0.67701113",
"0.67701113",
"0.67611784",
"0.6748694",
"0.674... | 0.6495886 | 33 |
Retrieve an existing user | def delete_a_tag(self, contact_tag_id):
return self.client._delete(f"/contactTags/{str(contact_tag_id)}") | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get(self):\n\n user_id = get_jwt_identity()\n user = user_crud.get(user_id)\n if not user:\n abort(404, message=\"User not Found\")\n\n return user",
"def get(self, id):\n\t\ttry:\n\t\t\tflask_app.logger.debug('We are getting the user: %d', id)\n\t\t\treturn user_servic... | [
"0.8212961",
"0.7848171",
"0.7845886",
"0.7653362",
"0.75199753",
"0.7491075",
"0.74858",
"0.74727863",
"0.7465673",
"0.746224",
"0.7382637",
"0.73640716",
"0.73251826",
"0.73248845",
"0.7324632",
"0.73170745",
"0.7291248",
"0.72907436",
"0.7288322",
"0.7282303",
"0.72667825"... | 0.0 | -1 |
This decorator starts the coroutine for us. | def coroutine(f, *a, **kw):
i = f(*a, **kw)
i.next()
return i | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def coroutine(func):\n def start(*args, **kwargs):\n cr = func(*args, **kwargs)\n next(cr)\n return cr\n return start",
"def coroutine(func):\n def start(*args, **kwargs):\n cr = func(*args, **kwargs)\n next(cr)\n return cr\n return start",
"def coroutine(f... | [
"0.73506445",
"0.73506445",
"0.73506445",
"0.73420984",
"0.73402673",
"0.7286279",
"0.70205104",
"0.69449747",
"0.68627673",
"0.6845239",
"0.6845239",
"0.67451215",
"0.6707592",
"0.65871006",
"0.63496244",
"0.6304423",
"0.62721074",
"0.6268496",
"0.6251403",
"0.6251403",
"0.6... | 0.0 | -1 |
converting json_bytes to dict or dict to json_bytes | def convert(data):
result = None
if isinstance(data, bytes):
try:
result = json.loads(data.decode(CODING))
except:
print('wrong data format!')
finally:
return result
elif isinstance(data, dict):
try:
result = json.dumps(data).en... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def json_decode(data):\n if isinstance(data, bytes):\n data = data.decode(\"utf-8\")\n\n return json.loads(data)",
"def _decode(self, data: bytes):\n\n return json.loads(data.decode('utf-8'))",
"def json_loads(s: Union[bytes, str]) -> Dict[str, Any]:\n return json.loads(ensure_text(s, \"... | [
"0.6704753",
"0.6599705",
"0.6597806",
"0.6580214",
"0.6572985",
"0.65350634",
"0.6337056",
"0.6297806",
"0.62290484",
"0.6216683",
"0.62112856",
"0.6185201",
"0.6179623",
"0.6128823",
"0.6126266",
"0.6124865",
"0.6116758",
"0.6101291",
"0.60905",
"0.609006",
"0.6081644",
"... | 0.71246356 | 0 |
Return a list of latest build packages that are tagged with certain tag | def koji_tag2pkgs(kapi, tag):
ret = []
for rpminfo in kapi.listTagged(tag, inherit=True, latest=True):
pkg = _koji_buildinfo2pkg(kapi, rpminfo)
ret.append(pkg)
return ret | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_latest_build(tag, package):\n proc = Popen([\"osg-koji\", \"-q\", \"list-tagged\", \"--latest\", tag, package],\n stdout=PIPE)\n out = proc.communicate()[0] or b''\n ret = proc.returncode\n\n latest_build_line = out.decode(\"latin-1\").strip()\n\n if ret != 0 or not latest_b... | [
"0.6766918",
"0.64562565",
"0.6202034",
"0.61146486",
"0.59529054",
"0.5890708",
"0.58779067",
"0.57897514",
"0.5786915",
"0.5756531",
"0.57446975",
"0.57202804",
"0.57058716",
"0.56544155",
"0.5646688",
"0.5629611",
"0.5629158",
"0.56251156",
"0.557942",
"0.55721",
"0.556434... | 0.6693938 | 1 |
Return a the build pacakges from a package id | def koji_pkgid2pkgs(kapi, pkgid):
ret = []
for binfo in kapi.listBuilds(packageID=pkgid):
pkg = _koji_buildinfo2pkg(kapi, binfo)
ret.append(pkg)
return ret | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_buildrequire_pkgs_from_build(build_id, session, config):\n print(\"Retriewing build metadata from: \", config.koji_host)\n build = session.getBuild(build_id)\n if not build:\n raise Exception(\"Build with id '{id}' has not been found.\".format(id=build_id))\n\n print(\"Build with the ID\... | [
"0.7184097",
"0.6101216",
"0.6087045",
"0.60699105",
"0.60417914",
"0.6040838",
"0.5977515",
"0.5934301",
"0.5913",
"0.58773243",
"0.58617723",
"0.58331734",
"0.5806986",
"0.5787779",
"0.5784125",
"0.5781364",
"0.5760907",
"0.574403",
"0.5714784",
"0.5705453",
"0.56750613",
... | 0.6947734 | 1 |
Return a list of latest packages that are in the given compose | def composed_url2pkgs(baseurl):
import compose
c = compose.Compose(baseurl)
cid = c.data_id()
cstat = c.data_status()
pdata = c.json_rpms()
p = compose.packages_from_compose(pdata)
pb = compose.packages_bin_from_compose(pdata)
return p, pb, cid, cstat | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_packages():\n\n packages = find_packages()\n packages = ['{}.{}'.format('uniq', package) for package in packages]\n packages.append('uniq')\n return packages",
"def get_package_list():\n pip_freeze = subprocess.check_output(('pip', 'freeze')).decode('utf8')\n package_list = [x.strip().s... | [
"0.61419404",
"0.59497553",
"0.58131677",
"0.57777905",
"0.5700319",
"0.5460036",
"0.54240966",
"0.540263",
"0.5388916",
"0.537803",
"0.5358507",
"0.53547126",
"0.53445446",
"0.53266305",
"0.53051114",
"0.5303987",
"0.52810436",
"0.5273311",
"0.52719593",
"0.52524483",
"0.525... | 0.5837896 | 2 |
Return a list of latest modules that are in the given compose | def composed_url2modules(baseurl):
import compose
c = compose.Compose(baseurl)
cid = c.data_id()
cstat = c.data_status()
print('Mod Compose:', cid)
print(' Status:', cstat)
mdata = c.json_modules()
m = compose.modules_from_compose(mdata)
return compose.dedup_modules(m) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def modules(self):\n return sorted([module for module in self._registry.values()],\n key=lambda scomp: (scomp.order, scomp.label))",
"def _list_all(root_pkg, prog):\n res = \"\\n\".join(\n sorted(\n pkinspect.package_module_names(_import(root_pkg)),\n k... | [
"0.5734544",
"0.5620557",
"0.5515039",
"0.54573196",
"0.5440216",
"0.5403242",
"0.54014957",
"0.5378393",
"0.53708357",
"0.5353265",
"0.53318167",
"0.53174704",
"0.5289548",
"0.52598995",
"0.5258763",
"0.52493656",
"0.5242129",
"0.5224267",
"0.5214835",
"0.5213591",
"0.520834... | 0.5680556 | 1 |
Get positive diagonals, going from bottomleft to topright. | def diagonalsPos (matrix, cols, rows):
for diagonal in ([(j, i - j) for j in range(cols)] for i in range(cols + rows -1)):
yield [matrix[i][j] for i, j in diagonal if i >= 0 and j >= 0 and i < cols and j < rows] | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def diagonals(self):\n left_top_shifts = map(lambda i: (-(i + 1), -(i + 1)), range(min(\n self.left_distance, self.top_distance)))\n left_bottom_shifts = map(lambda i: (-(i + 1), +(i + 1)), range(min(\n self.left_distance, self.bottom_distance)))\n right_top_shifts = map(... | [
"0.7523506",
"0.7055278",
"0.70287716",
"0.69674146",
"0.6805341",
"0.67695767",
"0.67601824",
"0.67326593",
"0.6724137",
"0.66729164",
"0.6666689",
"0.66343945",
"0.66061187",
"0.66051644",
"0.6519829",
"0.64941293",
"0.64686644",
"0.646446",
"0.6461399",
"0.64447653",
"0.63... | 0.7020545 | 3 |
Get negative diagonals, going from topleft to bottomright. | def diagonalsNeg (matrix, cols, rows):
for diagonal in ([(j, i - cols + j + 1) for j in range(cols)] for i in range(cols + rows - 1)):
yield [matrix[i][j] for i, j in diagonal if i >= 0 and j >= 0 and i < cols and j < rows] | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def diagonals(self):\n left_top_shifts = map(lambda i: (-(i + 1), -(i + 1)), range(min(\n self.left_distance, self.top_distance)))\n left_bottom_shifts = map(lambda i: (-(i + 1), +(i + 1)), range(min(\n self.left_distance, self.bottom_distance)))\n right_top_shifts = map(... | [
"0.74394697",
"0.70225304",
"0.6999749",
"0.6986131",
"0.69826496",
"0.68484116",
"0.68346524",
"0.6826511",
"0.6775777",
"0.6718986",
"0.66814524",
"0.6667113",
"0.65236944",
"0.65159935",
"0.6481205",
"0.6473899",
"0.6424914",
"0.6395643",
"0.6351728",
"0.6312667",
"0.62737... | 0.74049544 | 1 |
Create a new game. | def __init__ (self, cols = 6, rows = 7, requiredToWin = 4):
self.cols = cols
self.rows = rows
self.win = requiredToWin
self.board = [[NONE] * rows for _ in range(cols)] | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def game_new():\n posted = request.get_json()\n\n # Parse out the settings from the post\n settings = {}\n possible_settings = Game.DEFAULT_SETTINGS.keys()\n for setting in possible_settings:\n if setting in posted:\n settings[setting] = int(posted[setting])\n\n # Create and sta... | [
"0.78133714",
"0.7810131",
"0.7777821",
"0.7729321",
"0.7605243",
"0.75777143",
"0.7482669",
"0.7363624",
"0.7306845",
"0.7306757",
"0.72615993",
"0.7260527",
"0.72555196",
"0.7236042",
"0.7222994",
"0.7206253",
"0.7193306",
"0.71824133",
"0.71792173",
"0.7131112",
"0.7102228... | 0.0 | -1 |
Insert the figure in the given column. | def insert (self, column, figure):
c = self.board[column]
if c[0] != NONE:
raise Exception('Column is full')
i = -1
while c[i] != NONE:
i -= 1
c[i] = figure
self.checkForWin() | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def place_disc(self, column):\n stdout.write(\"place_disc %d\\n\" % column)\n stdout.flush()",
"def addColumn(self, *column):\n self.insertColumn(self._width, *column)",
"def insert_after(self, text, line, col):\n col = self.canonicalize_column_index(line, col)\n col_off = se... | [
"0.663612",
"0.66170853",
"0.63207686",
"0.6298038",
"0.62482417",
"0.6247075",
"0.6229078",
"0.6151331",
"0.6122636",
"0.61149824",
"0.610697",
"0.61038834",
"0.6098983",
"0.60879403",
"0.60723233",
"0.60662925",
"0.6054099",
"0.60376287",
"0.60324526",
"0.59339696",
"0.5914... | 0.8167265 | 0 |
Check the current board for a winner. | def checkForWin (self):
w = self.getWinner()
if w:
self.printBoard()
raise Exception(w + ' won!') | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def check_winner(self):\n pass",
"def check_for_winner(self):\r\n\r\n # Iterate through the rows\r\n for row in range(self.height):\r\n if self.board[row][0] == self.board[row][1] == self.board[row][2] and self.board[row][0] != None:\r\n return Board.WIN if self.boa... | [
"0.7888144",
"0.78121245",
"0.77188283",
"0.7654055",
"0.7651377",
"0.7618898",
"0.759551",
"0.7593995",
"0.75508547",
"0.75505865",
"0.7549655",
"0.75481594",
"0.74787205",
"0.7476814",
"0.7443378",
"0.7436571",
"0.74264586",
"0.73948514",
"0.73858464",
"0.73742837",
"0.7348... | 0.733567 | 22 |
Get the winner on the current board. | def getWinner (self):
lines = (
self.board, # columns
zip(*self.board), # rows
Board.diagonalsPos(self.board, self.cols, self.rows),
Board.diagonalsNeg(self.board, self.cols, self.rows)
)
for line in chain(*lines):
for figure, group in groupby(line):
if figure != NONE and len(list(g... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_winner(self, board):\r\n ids = self.player_ids\r\n board_size = self.board_size\r\n return self.get_winner_c(board, ids, board_size)",
"def get_winner(self):\n \n current_board_state = self.get_board_state()\n current_board_state_winner = self.state_winner_map.ge... | [
"0.84933585",
"0.8486584",
"0.8465523",
"0.83231235",
"0.83130556",
"0.83130556",
"0.83039725",
"0.8224555",
"0.8214615",
"0.8096653",
"0.80511385",
"0.8001037",
"0.7956043",
"0.7904712",
"0.7889448",
"0.78768945",
"0.7862174",
"0.7825047",
"0.7724497",
"0.77161705",
"0.76997... | 0.74699986 | 34 |
Create an organization in the test database | def setUp(self):
#Inheriting the base class functionality
super(CreateTaskAPITestCase, self).setUp()
# Create the org using serializer
create_org_data = {
'name': 'Ecell NITRR Open Source',
'tagline': 'We love open source.'
}
serializer = CreateOrg... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_add_organization(self):\n pass",
"def test_create_org(session): # pylint:disable=unused-argument\n user = factory_user_model()\n org = OrgService.create_org(TestOrgInfo.org1, user_id=user.id)\n assert org\n dictionary = org.as_dict()\n assert dictionary['name'] == TestOrgInfo.org1... | [
"0.81322205",
"0.77501535",
"0.76315355",
"0.76039594",
"0.7451626",
"0.724363",
"0.71316046",
"0.6978583",
"0.69309694",
"0.68478835",
"0.6826861",
"0.67564154",
"0.6742396",
"0.67277336",
"0.6576727",
"0.6567349",
"0.65031064",
"0.64968556",
"0.6489674",
"0.64821607",
"0.64... | 0.6963344 | 8 |
To be run in a directory with LOTAAS classifier out files. Determines which pulsars are in at least 3 of the 5 models provided and lists them in a file | def categorize_classifier_files(out_dir):
#sort all of the classifier files into a dictionary
class_files = glob.glob("feature_extraction_m*")
class_file_dict = {"positive":[], "negative":[]}
class_cand_dict = {"m1":class_file_dict, "m2":class_file_dict, "m3":class_file_dict, "m4":class_file_dict, "m5"... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_example(models, examples_path):\n frame_list = get_lidc_dataframes(examples_path, len(models))\n malignant_files = []\n for frame in frame_list:\n files = frame[\"File\"]\n for model, file in zip(models, files):\n prediction = np.argmax(model.predict(file))\n i... | [
"0.6200996",
"0.58896387",
"0.5857481",
"0.5760066",
"0.57523656",
"0.5741391",
"0.57344854",
"0.5728773",
"0.56916726",
"0.568332",
"0.56797236",
"0.56721956",
"0.56310445",
"0.5614847",
"0.558532",
"0.5576366",
"0.5574889",
"0.557381",
"0.5571876",
"0.55642986",
"0.55564094... | 0.68747187 | 0 |
r"""Return data used for tests in this file. | def test_data():
x = np.array([8, 67, 79, 10, 52, 53, 98, 34, 15, 58], dtype=float)
y = np.array([24, 87, 48, 94, 98, 66, 14, 24, 60, 16], dtype=float)
z = np.array([0.064, 4.489, 6.241, 0.1, 2.704, 2.809, 9.604, 1.156,
0.225, 3.364], dtype=float)
return x, y, z | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_data(self):\n return self._test_data",
"def getTestData(self):\n raise NotImplementedError",
"def GetTestData():\n return os.path.join(GetSrc(), 'chrome', 'test', 'data')",
"def _test_template_data(self):\n chars=string.ascii_uppercase + string.digits\n id = ''.join(rand... | [
"0.8199515",
"0.7520647",
"0.72175485",
"0.71363693",
"0.7092907",
"0.70007837",
"0.69980794",
"0.67899525",
"0.6782763",
"0.6782763",
"0.6779165",
"0.67096573",
"0.67096573",
"0.67096573",
"0.668133",
"0.6639849",
"0.6627223",
"0.6610728",
"0.6577204",
"0.65460837",
"0.65398... | 0.0 | -1 |
r"""Return grid locations used for tests in this file. | def test_grid():
with get_test_data('interpolation_test_grid.npz') as fobj:
data = np.load(fobj)
return data['xg'], data['yg'] | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_grid_position(self):\n tile_size_x = constants.WINDOW_WIDTH / constants.GRID_TILE_LENGTH\n tile_size_y = constants.WINDOW_HEIGHT / constants.GRID_TILE_LENGTH\n grid_x = tile_size_x / self.host.x\n grid_y = tile_size_y / self.host.y\n return grid_x, grid_y",
"def generat... | [
"0.6918649",
"0.63132507",
"0.63131356",
"0.6228707",
"0.615756",
"0.6123201",
"0.6111844",
"0.60771805",
"0.60724473",
"0.60664034",
"0.6015367",
"0.60129166",
"0.6007126",
"0.6001071",
"0.5990227",
"0.59845644",
"0.5919908",
"0.5914542",
"0.5877363",
"0.58720917",
"0.586397... | 0.6534552 | 1 |
r"""Test natural neighbor interpolation function. | def test_natural_neighbor(test_data, test_grid):
xp, yp, z = test_data
xg, yg = test_grid
img = natural_neighbor(xp, yp, z, xg, yg)
with get_test_data('nn_bbox0to100.npz') as fobj:
truth = np.load(fobj)['img']
assert_array_almost_equal(truth, img) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_nn_point(test_data):\n xp, yp, z = test_data\n\n tri = Delaunay(list(zip(xp, yp)))\n\n sim_gridx = [30]\n sim_gridy = [30]\n\n members, tri_info = find_natural_neighbors(tri,\n list(zip(sim_gridx, sim_gridy)))\n\n val = nn_point(xp, yp, z, [s... | [
"0.68133557",
"0.6289575",
"0.6225868",
"0.62149423",
"0.6126334",
"0.60409766",
"0.60360444",
"0.5988397",
"0.59295636",
"0.59061944",
"0.587684",
"0.5838663",
"0.5803554",
"0.5777474",
"0.5770294",
"0.569739",
"0.567572",
"0.5646712",
"0.5637545",
"0.5592249",
"0.5589785",
... | 0.6973079 | 0 |
r"""Test inverse distance interpolation function. | def test_inverse_distance(method, test_data, test_grid):
xp, yp, z = test_data
xg, yg = test_grid
extra_kw = {}
if method == 'cressman':
extra_kw['r'] = 20
extra_kw['min_neighbors'] = 1
test_file = 'cressman_r20_mn1.npz'
elif method == 'barnes':
extra_kw['r'] = 40
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test__inverse_transform_continuous(self):",
"def testInverted(self):\n invertedClass = xyTransformRegistry[\"inverted\"]\n invertedConfig = invertedClass.ConfigClass()\n affineClass = xyTransformRegistry[\"affine\"]\n invertedConfig.transform.retarget(affineClass)\n affineC... | [
"0.64221805",
"0.6114815",
"0.6088234",
"0.60316193",
"0.59683204",
"0.58155304",
"0.58098304",
"0.5610405",
"0.56054175",
"0.56038624",
"0.5599341",
"0.5583385",
"0.55830747",
"0.558048",
"0.55263007",
"0.55263007",
"0.5525398",
"0.55220234",
"0.55166364",
"0.55103624",
"0.5... | 0.71566653 | 0 |
r"""Test find natural neighbors for a point interpolation function. | def test_nn_point(test_data):
xp, yp, z = test_data
tri = Delaunay(list(zip(xp, yp)))
sim_gridx = [30]
sim_gridy = [30]
members, tri_info = find_natural_neighbors(tri,
list(zip(sim_gridx, sim_gridy)))
val = nn_point(xp, yp, z, [sim_gridx[0], sim... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_natural_neighbor(test_data, test_grid):\n xp, yp, z = test_data\n xg, yg = test_grid\n\n img = natural_neighbor(xp, yp, z, xg, yg)\n\n with get_test_data('nn_bbox0to100.npz') as fobj:\n truth = np.load(fobj)['img']\n\n assert_array_almost_equal(truth, img)",
"def problem5(self, s):... | [
"0.6897055",
"0.6649428",
"0.6639052",
"0.6530777",
"0.64982593",
"0.63480467",
"0.6291873",
"0.62584484",
"0.6234219",
"0.62292683",
"0.6225094",
"0.6086198",
"0.60818624",
"0.6075852",
"0.6039569",
"0.60207516",
"0.6012348",
"0.6010177",
"0.6008167",
"0.5977195",
"0.5965110... | 0.73872244 | 0 |
r"""Test Barnes weights function. | def test_barnes_weights():
kappa = 1000000
gamma = 0.5
dist = np.array([1000, 2000, 3000, 4000])**2
weights = barnes_weights(dist, kappa, gamma) * 10000000
truth = [1353352.832366126918939,
3354.626279025118388,
.152299797447126,
.000000126641655]
asse... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_breeding(self):\n Herbivore.set_parameters({\"gamma\": 1})\n self.herb.weight = 50\n newborn_weight = self.herb.breeding(10)\n nt.assert_not_equal(None, newborn_weight)",
"def c_test_weighting_bias(self, function):\r\n return general_weighting_bias(self.verbosity, func... | [
"0.7263572",
"0.72594935",
"0.72594935",
"0.72594935",
"0.69900525",
"0.68501157",
"0.6781213",
"0.6605773",
"0.6576652",
"0.6557013",
"0.6512069",
"0.64957434",
"0.6486371",
"0.6484919",
"0.646704",
"0.6454388",
"0.6386876",
"0.6384625",
"0.6384625",
"0.63698995",
"0.6315728... | 0.7957706 | 1 |
r"""Test Cressman weights function. | def test_cressman_weights():
r = 5000
dist = np.array([1000, 2000, 3000, 4000])**2
weights = cressman_weights(dist, r)
truth = [0.923076923076923,
0.724137931034482,
0.470588235294117,
0.219512195121951]
assert_array_almost_equal(truth, weights) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_basic(self):\n result = NonLinearWeights(0.85)\n self.assertAlmostEqual(result.cval, 0.85)",
"def test_feeding_weight_carn(self):\n original = self.carn.weight\n self.carn.fitness = 1\n herb = [Herbivore(age=90) for _ in range(50)]\n self.carn.feeding(herb)\n ... | [
"0.67748266",
"0.6498026",
"0.6415541",
"0.6391458",
"0.637254",
"0.6250476",
"0.6250476",
"0.6250476",
"0.6240067",
"0.62005264",
"0.61936164",
"0.61265564",
"0.6121124",
"0.6113677",
"0.6018572",
"0.60182",
"0.601679",
"0.599509",
"0.59852964",
"0.59660983",
"0.5918179",
... | 0.8282531 | 1 |
r"""Test Cressman interpolation for a point function. | def test_cressman_point(test_data):
xp, yp, z = test_data
r = 40
obs_tree = cKDTree(list(zip(xp, yp)))
indices = obs_tree.query_ball_point([30, 30], r=r)
dists = dist_2(30, 30, xp[indices], yp[indices])
values = z[indices]
truth = 1.05499444404
value = cressman_point(dists, values,... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_fixed_point(testFunctions, tol, printFlag): \n pass",
"def test_fixed_point_system(testFunctions, tol, printFlag): \n pass",
"def check_interp(self):\n\n points = np.loadtxt(\"skeleton_temp/\" + cell + \"_points.txt\", delimiter=',')\n\n self.initial_scatter = ax.scatter(poin... | [
"0.6262889",
"0.6127593",
"0.60630435",
"0.59690744",
"0.59616745",
"0.59546226",
"0.5874195",
"0.58734214",
"0.58238226",
"0.58026075",
"0.57872707",
"0.5775866",
"0.5738318",
"0.5729117",
"0.5704638",
"0.5668535",
"0.56655055",
"0.5663175",
"0.5661966",
"0.56616485",
"0.566... | 0.72067887 | 0 |
r"""Test Barnes interpolation for a point function. | def test_barnes_point(test_data):
xp, yp, z = test_data
r = 40
obs_tree = cKDTree(list(zip(xp, yp)))
indices = obs_tree.query_ball_point([60, 60], r=r)
dists = dist_2(60, 60, xp[indices], yp[indices])
values = z[indices]
truth = 4.08718241061
ave_spacing = np.mean((cdist(list(zip(x... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_single_point_bspline_radius() -> None:\n control_points = np.array([(0.0, 0.0, 0.0)])\n knot_vector = np.array([0, 0, 1, 1])\n curve = BSplineCurve(control_points, knot_vector)\n for parameter in np.linspace(0.0, 1.0, num=5):\n np.testing.assert_allclose(\n curve.radius_at(pa... | [
"0.65393716",
"0.62299854",
"0.621912",
"0.6113236",
"0.5917149",
"0.5852951",
"0.5850865",
"0.5846754",
"0.5818286",
"0.58133703",
"0.58133703",
"0.57956046",
"0.57805043",
"0.5778642",
"0.575499",
"0.5752062",
"0.57458085",
"0.57200813",
"0.5705935",
"0.56873363",
"0.568733... | 0.7259133 | 0 |
Add metrealstation_no & metforecaststation_no to rl_kpis_df data | def add_met_real_forecast_station_col_to_rlkpis(self):
rl_kpi_siteId_df = pd.DataFrame(self.rlkpi.imputed_rl_kpis_df.site_id.unique(), columns=['site_id'])
# Assign the nearest station from met-real to the site Id
rl_kpi_siteId_df['met-real-station_no'] = \
rl_kpi_siteId_df['si... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def prepare_train_data(self):\r\n ## Impute rlkpis\r\n print(\"Imputing rlKPI df\")\r\n self.rlkpi.add_target_labels(1)\r\n self.rlkpi.impute_rl_kpis()\r\n\r\n print(\"Add 'met-real-station_no' & met-forecast-station_no to rl_kpis_df\")\r\n self.add_met_real_forecast_stati... | [
"0.66093993",
"0.53555703",
"0.53393054",
"0.533251",
"0.53289",
"0.5301145",
"0.5197288",
"0.5096265",
"0.50771284",
"0.5050672",
"0.50454473",
"0.50402033",
"0.5028574",
"0.50062984",
"0.4998292",
"0.49721038",
"0.4968566",
"0.4958544",
"0.49429664",
"0.49366996",
"0.493378... | 0.7614396 | 0 |
1 Add 'metrealstation_no' & metforecaststation_no to rl_kpis_df so that metforecast and metreal data can be merged easily 2 merge metreal resample data 3 merge metforecast data | def prepare_train_data(self):
## Impute rlkpis
print("Imputing rlKPI df")
self.rlkpi.add_target_labels(1)
self.rlkpi.impute_rl_kpis()
print("Add 'met-real-station_no' & met-forecast-station_no to rl_kpis_df")
self.add_met_real_forecast_station_col_to_rlkpis()
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def add_met_real_forecast_station_col_to_rlkpis(self):\r\n rl_kpi_siteId_df = pd.DataFrame(self.rlkpi.imputed_rl_kpis_df.site_id.unique(), columns=['site_id'])\r\n\r\n # Assign the nearest station from met-real to the site Id\r\n rl_kpi_siteId_df['met-real-station_no'] = \\\r\n rl_k... | [
"0.76217365",
"0.5724756",
"0.5574119",
"0.5542644",
"0.5527173",
"0.5513221",
"0.55130845",
"0.5508425",
"0.5384717",
"0.5348086",
"0.5330485",
"0.5295532",
"0.52790254",
"0.52596587",
"0.52467704",
"0.5236038",
"0.520138",
"0.51984364",
"0.5197946",
"0.51753104",
"0.5171422... | 0.67931956 | 1 |
Compute the Mahalanobis Distance between each row of x and the data | def mahalanobis(self, x=None, data=None, cov=None):
x_minus_mu = x - np.mean(data)
if not cov:
cov = np.cov(data.values.T)
inv_covmat = sp.linalg.inv(cov)
left_term = np.dot(x_minus_mu, inv_covmat)
mahal = np.dot(left_term, x_minus_mu.T)
return mahal.di... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def mahalanobis_distance(self, x: np.ndarray) -> float:\n # this method could be vectorized for efficient calls\n error = x - self.mean\n mahalanobis_distance = error.T @ nla.solve(self.cov, error)\n return mahalanobis_distance",
"def mahalanobis_distance(x, y):\r\n diff = x - np.m... | [
"0.8095588",
"0.75557333",
"0.70927775",
"0.7065423",
"0.68101835",
"0.6714455",
"0.66829896",
"0.6671637",
"0.6317888",
"0.63015944",
"0.61623526",
"0.6118895",
"0.61179835",
"0.61020434",
"0.6095306",
"0.5993222",
"0.5977249",
"0.59567136",
"0.59101474",
"0.59001976",
"0.57... | 0.6563914 | 8 |
Apply principle component analysis to the Xarray to 3 dimensions and visualise the resulting matrix | def visualise_data_set(x_arr, y_arr):
# Instantiate a PCA object for the sake of easy visualisation
pca = PCA(n_components=3)
# Fit and transform x to visualise inside a 3D feature space
x_visualisation = pca.fit_transform(x_arr)
figure = plt.figure()
axis = Axes3D(figure)
axis.scatter(x_... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def principal_component_analysis(X, num_components):\n components = []\n for _ in range(num_components):\n component = first_principal_component(X)\n components.append(component)\n X = remove_projection(X, component)\n \n return components",
"def principle_component_analysis(data... | [
"0.6146485",
"0.60670024",
"0.56167203",
"0.559021",
"0.5509418",
"0.54846156",
"0.5475294",
"0.54682386",
"0.54289496",
"0.5405009",
"0.53256935",
"0.53170615",
"0.52998364",
"0.5298442",
"0.5297327",
"0.529588",
"0.527562",
"0.526715",
"0.52493393",
"0.5245872",
"0.52402586... | 0.63614935 | 0 |
Apply principle component analysis to the two Xarray"s to 3 dimensions and visualise the resulting matrices | def visualise_two_data_sets(x_arr, y_arr, x_arr_two, y_arr_two):
# Instantiate a PCA object for the sake of easy visualisation
pca = PCA(n_components=3)
# Fit and transform x to visualise inside a 3D feature space
x_visualisation = pca.fit_transform(x_arr)
figure = plt.figure()
axis = Axes3D(f... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def visualise_data_set(x_arr, y_arr):\n # Instantiate a PCA object for the sake of easy visualisation\n pca = PCA(n_components=3)\n\n # Fit and transform x to visualise inside a 3D feature space\n x_visualisation = pca.fit_transform(x_arr)\n\n figure = plt.figure()\n axis = Axes3D(figure)\n\n ... | [
"0.60191256",
"0.5515485",
"0.541513",
"0.53996515",
"0.5396888",
"0.5376275",
"0.53691566",
"0.5345294",
"0.53282833",
"0.5317449",
"0.5285818",
"0.52789575",
"0.52583647",
"0.52462524",
"0.52080536",
"0.51868224",
"0.5168219",
"0.51670533",
"0.5163496",
"0.5150325",
"0.5136... | 0.6886152 | 0 |
Plots a ROC curve for each classifier | def plot_mean_roc_curve_of_classifiers(classifier_roc_list, data_set_description):
if const.RECORD_RESULTS is True:
fig = plt.figure(figsize=(8, 6.66))
monochrome = (cycler("color", ["k"]) * cycler("marker", [""]) *
cycler("linestyle", ["-", "--", "-."]))
color_arr = ["... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_ROC():\r\n fpr = dict()\r\n tpr = dict()\r\n roc_auc = dict()\r\n threshold = dict()\r\n \r\n for i in range(n_classes):\r\n \r\n fpr[i], tpr[i], threshold[i] = roc_curve(y_test[:, i], y_pred[:, i])\r\n \r\n roc_auc[i] = auc(fpr[i], tpr[i])\r\n\r... | [
"0.7881746",
"0.7821903",
"0.77732265",
"0.7735024",
"0.77113014",
"0.7636098",
"0.7608597",
"0.7540226",
"0.74923337",
"0.74565345",
"0.7433374",
"0.7372081",
"0.7259865",
"0.72583896",
"0.7231202",
"0.72264963",
"0.7219714",
"0.7154107",
"0.71430546",
"0.7135938",
"0.712159... | 0.7588352 | 7 |
Plots a histogram of balancer results for a specific metric to file | def plot_balancer_results_per_classifier(data_balancer_results_per_classifier, parameter=(2, "Balanced Accuracy")):
classifier_arr = []
color = iter(cm.Set1(np.linspace(0, 1, len(data_balancer_results_per_classifier) + 1)))
mean_classifier_arr = [0] * len(data_balancer_results_per_classifier[0][1])
for ... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_histogram(hist, outname, xlabel=\"\", ylabel=\"frequency\"):\n plt.bar(hist[:,0], hist[:,1])\n plt.xlabel(xlabel)\n plt.ylabel(ylabel)\n plt.savefig(outname)\n plt.close()",
"def histogram(histo,nbr_launch,file):\n with open(\"Results/Histogram_{}_{}.txt\".format(nbr_launch,file.strip(... | [
"0.6530286",
"0.65231216",
"0.649234",
"0.64292675",
"0.6372478",
"0.6277188",
"0.6113853",
"0.61108017",
"0.60909116",
"0.6079042",
"0.6041105",
"0.6036437",
"0.59583956",
"0.59488505",
"0.59182066",
"0.5886027",
"0.5879379",
"0.58783364",
"0.5867552",
"0.5864914",
"0.582034... | 0.6124033 | 6 |
Plots a Kaplan Meier graph of time to default to file | def plot_kaplan_meier_graph_of_time_to_default(time_to_default, data_set_description=""):
kmf = KaplanMeierFitter()
kmf.fit(time_to_default, event_observed=[1] * len(time_to_default))
ax = kmf.plot(title="{0} Kaplan Meier analysis of time to default".format(data_set_description).replace("_", " "))
ax.ge... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_time_frames(self):\n\n fig = plt.figure()\n plt.grid(True)\n\n plt.ylim([-1.5,1.5])\n plt.xlim([0,1])\n\n for key in self.timeframes.keys():\n if key == 0:\n plt.plot(self.x, self.timeframes[key], label=\"time: \" + str(round(key*self.dt, 3)), l... | [
"0.6950117",
"0.6604719",
"0.6425373",
"0.6411647",
"0.62945455",
"0.62930465",
"0.6272661",
"0.61956865",
"0.6073431",
"0.60684174",
"0.60633975",
"0.60579836",
"0.6050537",
"0.6033914",
"0.6004738",
"0.6001769",
"0.59888655",
"0.5988027",
"0.5968159",
"0.5948308",
"0.594171... | 0.7070033 | 0 |
A generic function which plots the difference between the first entry and the rest of the entries to file | def plot_percentage_difference_graph(results, datasets, name_suffix="", parameter="BACC", x_label="Feature selection approach", difference_from="no feature selection", figsize=(16, 5), legend_y=None,
label_rotation=0, y_label_pos=None, y_ticks=None, x_label_replacement_dict=None, fe... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_diff(self, ax, d0, d1):\n i1 = 0\n for i0 in range(d0.shape[0]):\n try:\n while d0[i0,0] > d1[i1,0]:\n print((d0[i0,0], d1[i1,0]))\n i1 += 1\n except IndexError:\n break\n if d0[i0,0] == d1[i... | [
"0.6217327",
"0.61763155",
"0.61288965",
"0.6023296",
"0.6011869",
"0.5955028",
"0.5901663",
"0.5797355",
"0.577714",
"0.57608503",
"0.5745395",
"0.5700119",
"0.5661348",
"0.5626618",
"0.56228036",
"0.5606026",
"0.5561598",
"0.55449283",
"0.5531544",
"0.54875207",
"0.545747",... | 0.0 | -1 |
Plots the time to default graph of results to file | def plot_time_to_default_results(time_to_default_results_per_classifier, parameter="Balanced accuracy"):
color = iter(cm.Set1(np.linspace(0, 1, len(time_to_default_results_per_classifier[0][1]) + 1)))
classifier_arr = []
for i in range(len(time_to_default_results_per_classifier[0][1]) + 1):
classifi... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plotPerTimeStamp(options):\n name = options['name'] + '_' + options['scan'] + '_perTime'\n if options['extra']:\n name += '_' + options['extra']\n f = openRootFileR(options['name']+'_perTime')\n histname = plotName(name, timestamp=False)\n filename = plotName(name, timestamp=True)\n fi... | [
"0.70562994",
"0.6904556",
"0.68536395",
"0.6678135",
"0.66616327",
"0.66491324",
"0.6602544",
"0.6593051",
"0.65195054",
"0.6446045",
"0.6383599",
"0.63769704",
"0.63568395",
"0.6311244",
"0.6310854",
"0.6298991",
"0.62926877",
"0.62676895",
"0.6264268",
"0.62527686",
"0.623... | 0.0 | -1 |
Plots difference to median TPR and TNR on an x and y axis for multiple datasets | def visualise_dataset_classifier_results(dataset_results):
current_time = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
file_name = "raw_dump_{0}.txt".format(current_time)
with open(os.path.dirname(os.path.realpath(__file__)) + "/../results/" + file_name, "wb") as output_file:
output_file.write(str(d... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plotTimeDelta(data, type_plot, device):\n mean = data.mean()\n std = data.std()\n max_data = data.max()\n min_data = data.min()\n max_indx = np.argmax(data) # max value index\n min_indx = np.argmin(data) # min value index\n x = np.arange(min_data, max_data, 0.1)\n y = normfun(x, mean,... | [
"0.64062333",
"0.6103581",
"0.60293597",
"0.6005076",
"0.60047275",
"0.59661716",
"0.59610087",
"0.5950974",
"0.59288114",
"0.592674",
"0.5894661",
"0.588506",
"0.5852047",
"0.58322114",
"0.5823186",
"0.5814672",
"0.5804783",
"0.5800278",
"0.57631016",
"0.57512337",
"0.574488... | 0.55366945 | 55 |
Plots difference from no balancer for different classifiers and data balancing schemes | def visualise_dataset_balancer_results(results, range=(-0.5, 0.5),
colors=("#64B3DE", "#1f78b4", "#B9B914", "#FBAC44", "#bc1659", "#33a02c", "grey", "#b15928", "#6a3d9a", "#e31a1c", "#6ABF20", "#ff7f00", "#6a3d9a"),
exclude=("SVM (linear)", "... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_balancer_results_per_classifier(data_balancer_results_per_classifier, parameter=(2, \"Balanced Accuracy\")):\n classifier_arr = []\n color = iter(cm.Set1(np.linspace(0, 1, len(data_balancer_results_per_classifier) + 1)))\n mean_classifier_arr = [0] * len(data_balancer_results_per_classifier[0][1]... | [
"0.69392395",
"0.6658223",
"0.65169877",
"0.64943475",
"0.628095",
"0.62724614",
"0.61827344",
"0.61158895",
"0.60847265",
"0.6029901",
"0.5979186",
"0.5974841",
"0.5973415",
"0.5971678",
"0.59669495",
"0.5962603",
"0.59477496",
"0.5941426",
"0.59301496",
"0.59257036",
"0.591... | 0.6860829 | 1 |
Plots aggregate data balancer results by combining all classifier results for the particular data balancer. | def visualise_dataset_balancer_results_multi_dataset(dataset_results):
current_time = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
file_name = "raw_dump_{0}.txt".format(current_time)
with open(os.path.dirname(os.path.realpath(__file__)) + "/../results/" + file_name, "wb") as output_file:
output_file... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def plot_balancer_results_per_classifier(data_balancer_results_per_classifier, parameter=(2, \"Balanced Accuracy\")):\n classifier_arr = []\n color = iter(cm.Set1(np.linspace(0, 1, len(data_balancer_results_per_classifier) + 1)))\n mean_classifier_arr = [0] * len(data_balancer_results_per_classifier[0][1]... | [
"0.75818884",
"0.671413",
"0.6578102",
"0.6018402",
"0.5996524",
"0.5993794",
"0.5987738",
"0.59610283",
"0.59337884",
"0.5889667",
"0.58879936",
"0.58794236",
"0.5878393",
"0.58589953",
"0.58403677",
"0.58308715",
"0.58304363",
"0.5776435",
"0.5768485",
"0.5756904",
"0.57507... | 0.746169 | 1 |
Simulates green blindness in a color Arguments | def deuteranomaly(c):
cc = np.zeros(3)
cc[0] = (4211 + 0.677*(255*c[1])**2.2 + 0.2802*(255*c[0])**2.2)**(1.0/2.2)
cc[1] = (4211 + 0.677*(255*c[1])**2.2 + 0.2802*(255*c[0])**2.2)**(1.0/2.2)
cc[2] = (4211 + 0.95724*(255*c[2])**2.2 + 0.02138*(255*c[1])**2.2 \
- 0.02138*(255*c[0])**2.2)**(1.0/2.... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def lightness(color):\n\n strongest = max(color.red, color.green, color.blue)\n weakest = min(color.red, color.green, color.blue)\n return 0.5 * (strongest + weakest) / 255",
"def setColorDiffuse(*args):",
"def _color(self, args):",
"def compute_luminosity(red, green, blue):\r\n return (0.299 *... | [
"0.668433",
"0.66246545",
"0.66212654",
"0.6615543",
"0.64932907",
"0.6433904",
"0.6395846",
"0.627033",
"0.62459433",
"0.6244545",
"0.6229325",
"0.622143",
"0.6183974",
"0.61690986",
"0.614985",
"0.6148988",
"0.6143495",
"0.61352485",
"0.60831016",
"0.6057762",
"0.6054294",
... | 0.0 | -1 |
Simulates red blindness in a color Arguments | def protanomaly(c):
cc = np.zeros(3)
cc[0] = (782.7 + 0.8806*(255*cc[1])**2.2 + 0.1115*(255*cc[0])**2.2)**(1.0/2.2)
cc[1] = (782.7 + 0.8806*(255*cc[1])**2.2 + 0.1115*(255*cc[0])**2.2)**(1.0/2.2)
cc[2] = (782.7 + 0.992052*(255*cc[2])**2.2 - 0.003974*(255*cc[1])**2.2 \
+ 0.003974*(255*cc[0])**... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _color(self, args):",
"def lightness(color):\n\n strongest = max(color.red, color.green, color.blue)\n weakest = min(color.red, color.green, color.blue)\n return 0.5 * (strongest + weakest) / 255",
"def compute_luminosity(red, green, blue):\r\n return (0.299 * red) + (0.587 * green) + (0.114 ... | [
"0.6916326",
"0.66381216",
"0.66007376",
"0.65724105",
"0.65425265",
"0.6514243",
"0.6462543",
"0.6461585",
"0.6320996",
"0.63075715",
"0.6278275",
"0.626712",
"0.6259189",
"0.6233183",
"0.62251776",
"0.61820674",
"0.6140597",
"0.6137846",
"0.6115421",
"0.61151356",
"0.609398... | 0.0 | -1 |
Returns a build for compiling typescript files. Using the out directive, typescript will emit a js file with all the included references, so just create a base file that performs all the required includes. A default output file .d.ts is also generated in the output folder. | def compile(base, output, source, bind=True, skip=False):
# Paths
path = lambda *x: r.path(*[base] + list(x))
tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')
output_path = path(*output)
source_path = path(*source)
source_folder = path(*source[:-1])
# Build
build = r.buil... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def Build(self, out_file):\n raise NotImplementedError",
"def compile_files(base, output, source, bind=True, amd=False):\n\n # Paths\n path = lambda *x: r.path(*[base] + list(x))\n tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')\n output_folder = path(*output)\n source_folder = p... | [
"0.6553955",
"0.5779977",
"0.5724019",
"0.5666004",
"0.54626364",
"0.5448146",
"0.53711534",
"0.53711534",
"0.5339161",
"0.5291656",
"0.52618027",
"0.52588505",
"0.5244529",
"0.5209851",
"0.51751953",
"0.5151105",
"0.51451707",
"0.51408684",
"0.514047",
"0.51372755",
"0.50789... | 0.5712727 | 3 |
Compiles a target, then invokes it using node. | def test(base, output, source, bind=True):
# Paths
path = lambda *x: r.path(*[base] + list(x))
runner = 'node'
output_path = path(*output)
source_path = path(*source)
output_folder = path(*output[:-1])
# Build
build = compile(base, output, source, bind=False, skip=True)
build.notic... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def compile(cls, node, filename):\n compiler = cls(filename)\n compiler.visit(node)\n return compiler.stack.pop()",
"def _pyCompile ( target, source, env ) :\n if len(target) != 1 :\n fail ( \"unexpected number of targets for pyCompile: \"+str(target) )\n if len(source) != 1 :\... | [
"0.6068002",
"0.6013169",
"0.5963838",
"0.5867764",
"0.58265406",
"0.559987",
"0.5526865",
"0.552637",
"0.55262053",
"0.54664874",
"0.5444324",
"0.5436896",
"0.5435432",
"0.54198766",
"0.5417057",
"0.54145837",
"0.53992826",
"0.53654575",
"0.53342557",
"0.5288225",
"0.5267427... | 0.5295578 | 19 |
Returns a build for compiling multiple typescript modules This works exactly like compile_files with amd = True. | def compile_modules(base, output, source, bind=True):
return compile_files(base, output, source, bind, amd=True) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def compile_files(base, output, source, bind=True, amd=False):\n\n # Paths\n path = lambda *x: r.path(*[base] + list(x))\n tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')\n output_folder = path(*output)\n source_folder = path(*source)\n\n # Compile each file. Sometimes --module see... | [
"0.7168874",
"0.59924674",
"0.5815421",
"0.5812318",
"0.5812318",
"0.5750447",
"0.5671063",
"0.5620077",
"0.55134124",
"0.5502796",
"0.5469034",
"0.54557586",
"0.5427786",
"0.5426478",
"0.53992474",
"0.5392173",
"0.53844976",
"0.5356187",
"0.53483844",
"0.5321252",
"0.5311751... | 0.71308005 | 1 |
Returns a build for compiling multiple typescript files. Every matching.ts file will be processed separately. If bind is set, the result is bound internally; you typically want to do this, but some ops (eg. rebuild) may want to run the build.execute() manually. | def compile_files(base, output, source, bind=True, amd=False):
# Paths
path = lambda *x: r.path(*[base] + list(x))
tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')
output_folder = path(*output)
source_folder = path(*source)
# Compile each file. Sometimes --module seems to screw u... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def compile_modules(base, output, source, bind=True):\n return compile_files(base, output, source, bind, amd=True)",
"def compile(base, output, source, bind=True, skip=False):\n\n # Paths\n path = lambda *x: r.path(*[base] + list(x))\n tsc_path = path('node_modules', 'typescript', 'bin', 'tsc.js')\n ... | [
"0.6088479",
"0.56501824",
"0.5504514",
"0.5331496",
"0.5290386",
"0.52033865",
"0.5071304",
"0.50649625",
"0.5030966",
"0.49910313",
"0.49463353",
"0.49420282",
"0.49213454",
"0.49104938",
"0.4905638",
"0.49018425",
"0.48995948",
"0.48995948",
"0.48929653",
"0.48861158",
"0.... | 0.6061279 | 1 |
sends an email when lead saved successfully | def send_lead_task(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):
logger.info("in send lead mail task")
return send_lead_email(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())",
"def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has bee... | [
"0.67478585",
"0.64101136",
"0.61686337",
"0.61469436",
"0.61077154",
"0.61048436",
"0.60561985",
"0.6044016",
"0.60433817",
"0.5962845",
"0.59153545",
"0.5879812",
"0.5870707",
"0.5869888",
"0.58417296",
"0.5837118",
"0.5825478",
"0.58196616",
"0.57915473",
"0.5787865",
"0.5... | 0.6403527 | 2 |
sends an email when lead saved successfully | def send_ext_customer_task(email,name,password,phone,shop,address,lead_mail,mem_mail,website):
print("member email",mem_mail)
logger.info("in sending existing customer mail task")
return send_ext_customer_mail(email,name,password,phone,shop,address,lead_mail,mem_mail,website) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())",
"def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has bee... | [
"0.67478585",
"0.64101136",
"0.6403527",
"0.61686337",
"0.61469436",
"0.61077154",
"0.61048436",
"0.60561985",
"0.6044016",
"0.60433817",
"0.5962845",
"0.59153545",
"0.5879812",
"0.5870707",
"0.5869888",
"0.58417296",
"0.5837118",
"0.5825478",
"0.58196616",
"0.57915473",
"0.5... | 0.54872906 | 51 |
sends an email when lead saved successfully | def send_new_credentials(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website):
logger.info("in send lead mail task")
return send_lead_generate(email,name,password,phone,shop,address,lead_mail,fname,mem_mail,website) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def save(self, fail_silently=False):\n send_mail(fail_silently=fail_silently, **self.get_message_dict())",
"def send_created_email(self):\n if settings.NOTIFY_NEW_REG:\n to = settings.NOTIFY_NEW_REG\n message = \"\"\"\\\nGreetings,<br><br>\n\nA new vehicle registration has bee... | [
"0.67476535",
"0.6409794",
"0.64040166",
"0.6167846",
"0.61461926",
"0.61059725",
"0.61046255",
"0.605578",
"0.6043292",
"0.6042979",
"0.5960896",
"0.59149146",
"0.5868797",
"0.5867584",
"0.58403814",
"0.58367157",
"0.58244336",
"0.5819981",
"0.57925284",
"0.57854337",
"0.576... | 0.5880153 | 12 |
Render each drawing by converting the drawing to image ink and then applying broaden & blur filters | def forward(self, drawings, blur_sigma=None, epsilon=None):
# draw the strokes (this part on cpu)
if not isinstance(drawings[0], list):
single = True
drawings = [drawings]
else:
single = False
pimgs = self.painter(drawings)
pimgs = self.broade... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def draw(self):\n self.bufferX = (self.appWidth/2) - self.viewX\n self.bufferY = (self.appHeight/2) - self.viewY\n anwp.sl.engine.clear()\n anwp.sl.engine.drawImage(0, 0, self.appWidth, self.appHeight, self.backgroundImage)\n self.drawWarpLines()\n \n # render engin... | [
"0.62051916",
"0.608807",
"0.60164183",
"0.5994042",
"0.5974199",
"0.5956731",
"0.59398204",
"0.5916923",
"0.5856512",
"0.582398",
"0.5776995",
"0.57500255",
"0.57297295",
"0.57178104",
"0.570983",
"0.5693524",
"0.5684857",
"0.5681157",
"0.5680937",
"0.5669405",
"0.5637934",
... | 0.69175744 | 0 |
In this version, we include all partial canvas renders in addition to the final renders | def forward_partial(self, drawing, blur_sigma=None, epsilon=None, concat=False):
if isinstance(drawing[0], list):
pimgs = [self.painter.forward_partial(d) for d in drawing]
lengths = [len(p) for p in pimgs]
pimgs = torch.cat(pimgs)
pimgs = self.broaden_and_blur(pi... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def on_draw(self):\n\t\tself.render()",
"def on_draw(self):\n\t\tself.render()",
"def dspyRender(self):\n pass",
"def on_draw(self):\n # draw everything",
"def render(self):\n self.axial.Render()\n self.coronal.Render()\n self.sagittal.Render()\n #self.isosurface.R... | [
"0.62798715",
"0.62798715",
"0.6229959",
"0.62175214",
"0.62031156",
"0.6023987",
"0.5880108",
"0.58327454",
"0.58257514",
"0.5803622",
"0.5789395",
"0.57692194",
"0.5759491",
"0.57456064",
"0.57041544",
"0.5677518",
"0.5677518",
"0.5677518",
"0.5677518",
"0.56731313",
"0.567... | 0.0 | -1 |
In this version, we include all partial canvas renders in addition to the final renders | def forward_partial(self, drawing):
assert not self.is_cuda
drawing = drawings_to_cpu(drawing)
ns = len(drawing)
pimgs = torch.zeros(ns+1, *self.imsize)
canvas = torch.zeros(*self.imsize)
for i, stk in enumerate(drawing):
canvas, _ = self.add_stroke(canvas, st... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def on_draw(self):\n\t\tself.render()",
"def on_draw(self):\n\t\tself.render()",
"def dspyRender(self):\n pass",
"def on_draw(self):\n # draw everything",
"def render(self):\n self.axial.Render()\n self.coronal.Render()\n self.sagittal.Render()\n #self.isosurface.R... | [
"0.628021",
"0.628021",
"0.62287074",
"0.6217737",
"0.6203357",
"0.60232645",
"0.5878731",
"0.58328944",
"0.58255154",
"0.5802863",
"0.5789068",
"0.57696724",
"0.5759943",
"0.5747099",
"0.5704188",
"0.56780964",
"0.56780964",
"0.56780964",
"0.56780964",
"0.5672039",
"0.567181... | 0.5282165 | 88 |
Get property value from object. | def get(self, obj):
raise NotImplementedError | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_value_from_object(obj, key):\n if is_dict(obj):\n return obj.get(key)\n return getattr(obj, key, None)",
"def get(self, prop):\r\n prop_parts = prop.split(\".\")\r\n val = None\r\n for part in prop_parts:\r\n if val is None:\r\n val = self.obj.g... | [
"0.7294242",
"0.7260294",
"0.72458637",
"0.7177455",
"0.71202344",
"0.6993076",
"0.69180053",
"0.6906648",
"0.6891664",
"0.68564713",
"0.68375313",
"0.6823808",
"0.6799708",
"0.6789528",
"0.67803353",
"0.67259824",
"0.67239094",
"0.6705578",
"0.6705456",
"0.66965705",
"0.6668... | 0.6502106 | 26 |
Set property value on object. | def set(self, obj, value):
raise NotImplementedError | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def set(self, obj, value):\n pass",
"def set_value ( self, object, value ):\n target, name = self.target_name( object )\n setattr( target, name, value )",
"def __set__(self, obj, value):\r\n pass",
"def set_property(self, name, value):\n self.properties[name] = value",
"def _... | [
"0.80547476",
"0.77362704",
"0.76287323",
"0.76190275",
"0.75847745",
"0.7426509",
"0.73941714",
"0.72769517",
"0.726904",
"0.72673047",
"0.7146816",
"0.7140176",
"0.7101052",
"0.7087965",
"0.7087463",
"0.7041664",
"0.69987345",
"0.69987345",
"0.6997965",
"0.69783264",
"0.695... | 0.79016274 | 1 |
Set `password_hash` only when `password` is not set. | def _create(cls, model_class, *args, **kwargs):
assert 'password' in kwargs
assert 'passwd_hash' in kwargs
if kwargs.get('password'):
kwargs.pop('passwd_hash')
return super()._create(model_class, *args, **kwargs) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def set_password_hash(self, password):\n salt = bcrypt.gensalt()\n self.password_hash = bcrypt.hashpw(password.encode(), salt)",
"def hash_password(self, password):\n self.password_hash = generate_password_hash(password)",
"def set_password(self, password):\n self.password = self.ha... | [
"0.7201298",
"0.7091728",
"0.70867187",
"0.6961345",
"0.69605434",
"0.69136447",
"0.6894837",
"0.6894837",
"0.68855447",
"0.6867788",
"0.68643236",
"0.6862712",
"0.68401206",
"0.68401206",
"0.68401206",
"0.68401206",
"0.67581147",
"0.6730123",
"0.6710391",
"0.6706543",
"0.668... | 0.0 | -1 |
Create a room history entry, which is required for consistency reasons. | def room_history_entries(self, create, extracted, **kwargs):
if create and self.room is not None:
# Set room history entry begin to registration date
rhe = RoomHistoryEntry.q.filter_by(user=self, room=self.room).one()
rhe.active_during = starting_from(self.registered_at)
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_room_history(self, room):\n pass",
"def create_historical_record(self, instance, history_type):\n history_date = getattr(instance, '_history_date', now())\n history_changeset = self.get_history_changeset(instance)\n manager = getattr(instance, self.manager_name)\n attrs... | [
"0.66256154",
"0.62829894",
"0.62376595",
"0.6034511",
"0.6028819",
"0.6023105",
"0.6000676",
"0.5989488",
"0.59509176",
"0.5908362",
"0.582105",
"0.5779776",
"0.5768183",
"0.5764454",
"0.5743833",
"0.57111114",
"0.57085574",
"0.56982267",
"0.56964606",
"0.5662375",
"0.564588... | 0.7202129 | 0 |
If the input URL contains a port number, strip it out | def rewrite_url(line):
url_list = line.split(' ')
old_url = url_list[0]
new_url = '\n'
# http://192.168.1.13/biedronka_logo.jpg 192.168.1.100/192.168.1.100 - GET myip=- myport=3128
if old_url.endswith('jpg') or old_url.endswith('.jpeg'):
filename = str(uuid.uuid4())+".jpg"
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_remove_empty_port():\n assert (normalize_url(\"http://www.example.com:/\") ==\n \"http://www.example.com/\")",
"def test_remove_default_port():\n assert (normalize_url(\"http://www.example.com:80/bar.html\") ==\n \"http://www.example.com/bar.html\")\n assert (normalize_url... | [
"0.7776715",
"0.7397051",
"0.6689824",
"0.6589313",
"0.650951",
"0.650275",
"0.6461937",
"0.6428155",
"0.64266",
"0.6361706",
"0.6282323",
"0.62341475",
"0.61968756",
"0.6188919",
"0.6155203",
"0.6127684",
"0.61132604",
"0.60064256",
"0.59795815",
"0.5977992",
"0.5977227",
... | 0.0 | -1 |
MLP Block in Transformer Encoder | def __init__(self, hidden_layers, dropout=0.1, activation='gelu'):
super(MLPBlock, self).__init__()
layers = []
for num_units in hidden_layers:
layers.extend([
Dense(num_units, activation=activation),
Dropout(dropout)
])
self.mlp ... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):\n super(TransformerEncoder, self).__init__()\n\n # Create num_layers of TransformerBlock\n self.encoder = Sequential(\n [\n TransformerBlock(num_heads=num_heads,\n ... | [
"0.63641286",
"0.61162806",
"0.6006187",
"0.59183264",
"0.56983024",
"0.5687006",
"0.5635837",
"0.56081927",
"0.55380064",
"0.5521205",
"0.54553306",
"0.5424783",
"0.5420473",
"0.54017746",
"0.53908736",
"0.5378384",
"0.5374217",
"0.5369382",
"0.5336188",
"0.5329734",
"0.5307... | 0.5075871 | 54 |
Pass output of multihead attention to mlp block | def call(self, inputs, *args, **kwargs):
outputs = self.mlp(inputs, *args, **kwargs)
return outputs | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def attention_module_multi_head(roi_feat, position_embedding,\n nongt_dim, fc_dim, feat_dim,\n dim=(1024, 1024, 1024),\n group=16, index=1):\n dim_group = (dim[0] / group, dim[1] / group, dim[2] / group)\n nongt_roi_... | [
"0.63806444",
"0.62879974",
"0.61243266",
"0.6001117",
"0.59858584",
"0.58377385",
"0.57497156",
"0.57036",
"0.56529164",
"0.56093395",
"0.5535092",
"0.55335355",
"0.55295086",
"0.54910344",
"0.54799116",
"0.5470056",
"0.5456178",
"0.54472595",
"0.53975135",
"0.53934085",
"0.... | 0.0 | -1 |
Transformer blocks which includes multihead attention layer and mlp block | def __init__(self, num_heads, D, hidden_layers, dropout=0.1, norm_eps=1e-12):
super(TransformerBlock, self).__init__()
# Attention
self.attention = MultiHeadAttention(
num_heads=num_heads, key_dim=D, dropout=dropout
)
self.norm_attention = LayerNormalization(epsilon=n... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def build_model_multihead_attention_multiscaleCNN4_covermore(self,\n dim_attention,headnum,\n embedding_vec,\n load_weights=False, weight_dir=None,\n nb_filter... | [
"0.6663586",
"0.6597379",
"0.64268494",
"0.6360674",
"0.6338788",
"0.6257444",
"0.6235589",
"0.6209039",
"0.60610014",
"0.59759337",
"0.59740186",
"0.59430027",
"0.590821",
"0.5900392",
"0.58956826",
"0.5895034",
"0.58770245",
"0.58639365",
"0.5794944",
"0.5788988",
"0.576384... | 0.6992421 | 0 |
Pass Embedded Patches through the layers | def call(self, inputs):
# Feed attention
norm_attention = self.norm_attention(inputs)
attention = self.attention(query=norm_attention, value=norm_attention)
# Skip Connection
attention += inputs
# Feed MLP
outputs = self.mlp(self.norm_mlp(attention))
#... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def run(layers):",
"def build_layers(self):\n raise NotImplementedError",
"def layers(self): # -> LayerView:\n ...",
"def UpdateLayers(self):\n pass",
"def layers(self, layers):\n self._layers = layers\n prev = None\n for layer in layers:\n if not layer.inputs... | [
"0.64498407",
"0.6189903",
"0.6165279",
"0.5999772",
"0.5932045",
"0.5875921",
"0.5680412",
"0.5599581",
"0.5518509",
"0.5489415",
"0.5385347",
"0.5311613",
"0.53089815",
"0.5294126",
"0.5293941",
"0.52928007",
"0.5287579",
"0.5253471",
"0.52430594",
"0.5234675",
"0.5187667",... | 0.0 | -1 |
Transformer Encoder which comprises several transformer layers | def __init__(self, num_layers, num_heads, D, mlp_dim, dropout=0.1, norm_eps=1e-12):
super(TransformerEncoder, self).__init__()
# Create num_layers of TransformerBlock
self.encoder = Sequential(
[
TransformerBlock(num_heads=num_heads,
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def encoder(enc_input, attn_bias, n_layer, n_head,\n d_key, d_value, d_model, d_inner_hid, pos_enc,\n preporstprocess_dropout, attention_dropout,\n relu_dropout, preprocess_cmd='n',\n postprocess_cmd='da'):\n for i in range(n_layer):\n enc_output = encoder_layer(enc_input, attn_bias, n_head,\... | [
"0.72634786",
"0.7005841",
"0.68940455",
"0.68260175",
"0.68099016",
"0.6762681",
"0.6706989",
"0.66525126",
"0.66428787",
"0.6633862",
"0.6610087",
"0.6552235",
"0.65452635",
"0.6533876",
"0.6485411",
"0.64490217",
"0.6430559",
"0.64192617",
"0.64069134",
"0.638147",
"0.6380... | 0.63730127 | 21 |
Pass Embedded Patches through the layers | def call(self, inputs, *args, **kwargs):
outputs = self.encoder(inputs, *args, **kwargs)
return outputs | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def run(layers):",
"def build_layers(self):\n raise NotImplementedError",
"def layers(self): # -> LayerView:\n ...",
"def UpdateLayers(self):\n pass",
"def layers(self, layers):\n self._layers = layers\n prev = None\n for layer in layers:\n if not layer.inputs... | [
"0.6448038",
"0.6187745",
"0.6163194",
"0.5998497",
"0.5929416",
"0.5873209",
"0.56773156",
"0.5597431",
"0.5518665",
"0.5488085",
"0.5382779",
"0.53090936",
"0.53070503",
"0.52928275",
"0.5291836",
"0.5291313",
"0.5284668",
"0.5253536",
"0.5242199",
"0.523257",
"0.51847124",... | 0.0 | -1 |
Creates an upper confidence tree by simulating N games starting from the given state in order to select the move leading to the best results | def get_uct(self, state):
root = ut.UCTree(state)
for _ in range(self.N):
moves = []
leaf = root.select(moves, state.courant)
win = leaf.simulate(moves)
root.back_propagate(leaf, moves, win)
return root | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def __call__(self, s, n=1000):\n\n root = StateNode(None, s, self.game)\n \n if root.parent is not None:\n raise ValueError(\"Root's parent must be None.\")\n \n for _ in range(n):\n #selection\n node = _get_next_node(root, self.tree_polic... | [
"0.69618815",
"0.65259063",
"0.6470664",
"0.6314135",
"0.62374127",
"0.6210219",
"0.61834335",
"0.6150433",
"0.6120203",
"0.6094854",
"0.6084462",
"0.60547906",
"0.6046755",
"0.6046614",
"0.6031796",
"0.5984574",
"0.5967185",
"0.59669054",
"0.595718",
"0.595556",
"0.5952896",... | 0.6509543 | 2 |
Return the last bytes of a file containing utf8 formatted text rather than the whole contents. The last bytes is determined by size_in_bytes | def read_last_bytes_of_file(file_path: str, size_in_bytes: int) -> str:
with open(file_path, "rb") as file:
# Note the minus sign
file.seek(-size_in_bytes, os.SEEK_END)
content = file.read().decode("utf-8")
return content | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def text_file_last_line(f, encoding=\"utf-8\"):\n f.seek(0, os.SEEK_END) # start at the second to last character, since text files should end in a newline\n if f.tell() < 2: return \"\"\n f.seek(-2, os.SEEK_END) # start at the second to last character, since text files should end in a newline\n while f... | [
"0.61801714",
"0.5925",
"0.59065217",
"0.58524173",
"0.5808898",
"0.56951183",
"0.5602501",
"0.5586138",
"0.55584174",
"0.55470455",
"0.55171245",
"0.5497499",
"0.5476367",
"0.5474634",
"0.54546565",
"0.5451746",
"0.5424187",
"0.541994",
"0.5400577",
"0.53996587",
"0.53982",
... | 0.79039353 | 0 |
Test basic ranking page. | def test_northamerica_population(self):
self.driver.get(self.url_ +
'/ranking/Count_Person/Country/northamerica')
subtitle_present = EC.text_to_be_present_in_element(
(By.TAG_NAME, 'h3'), 'All Countries in North America')
WebDriverWait(self.driver, self.TIME... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_job_rank(self):\n inv_search = 'rank:Postdoc'\n spi_search = 'find rank Postdoc'\n self._compare_searches(inv_search, spi_search)",
"def test_something(self):\n\n allure.dynamic.title(\"Testing compute_ranks\")\n allure.dynamic.severity(allure.severity_leve... | [
"0.71264744",
"0.6681979",
"0.6591885",
"0.6474593",
"0.6324985",
"0.6264448",
"0.6260697",
"0.6223638",
"0.61334574",
"0.61309284",
"0.6111487",
"0.61088467",
"0.60755545",
"0.6048969",
"0.6043513",
"0.5932428",
"0.5879655",
"0.58348274",
"0.58152354",
"0.5808058",
"0.569797... | 0.5600284 | 29 |
Test translations are displayed correctly in hindi, as well as bottom rankings rendered correctly. | def test_energy_consumption_bottom_ranking_hi(self):
self.driver.get(
self.url_ +
'/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FIND&unit=kg&hl=hi&bottom='
)
self.assertEqual(
self.driver.find_elements_by_class_name('navbar-brand')[0].text... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_i18n10(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n10', {'bool': True})\n self.assertEqual(output, 'Ja')",
"def test_i18n27(self):\n with translation.override(\"ru\"):\n output = self.engine.render_to_string(\"i18n27\"... | [
"0.66185194",
"0.65907204",
"0.65775216",
"0.65519655",
"0.65519655",
"0.65463674",
"0.6530294",
"0.6447057",
"0.64026135",
"0.63868695",
"0.6383359",
"0.63773143",
"0.63244116",
"0.6323046",
"0.6316519",
"0.6313638",
"0.63092667",
"0.62892264",
"0.62892264",
"0.6254798",
"0.... | 0.58505183 | 55 |
Test translations are displayed correctly in korean, as well as top rankings rendered correctly. | def test_energy_consumption_top_ranking_ko(self):
self.driver.get(
self.url_ +
'/ranking/Amount_Consumption_Energy_PerCapita/Country/?h=country%2FKOR&unit=kg&hl=ko'
)
self.assertEqual(
self.driver.find_elements_by_class_name('navbar-brand')[0].text,
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def test_i18n10(self):\n with translation.override('de'):\n output = self.engine.render_to_string('i18n10', {'bool': True})\n self.assertEqual(output, 'Ja')",
"def test_i18n27(self):\n with translation.override(\"ru\"):\n output = self.engine.render_to_string(\"i18n27\"... | [
"0.70812863",
"0.6640255",
"0.6615368",
"0.6592845",
"0.6592845",
"0.6576762",
"0.65381277",
"0.64886284",
"0.64697886",
"0.6454866",
"0.642305",
"0.63667893",
"0.636136",
"0.632902",
"0.632902",
"0.6252181",
"0.61971015",
"0.6178152",
"0.61683005",
"0.6137992",
"0.6137528",
... | 0.57219356 | 65 |
Create an instance of the FriCAS interpreter. | def __init__(self, name='fricas', command='fricas -nox -noclef',
script_subdirectory=None, logfile=None,
server=None, server_tmpdir=None,
init_code=[')lisp (si::readline-off)']):
PanAxiom.__init__(self, name, command,
script_subdirectory, logfi... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def __create_execution_engine(self):\n target = self.binding.Target.from_default_triple()\n target_machine = target.create_target_machine()\n backing_mod = binding.parse_assembly(\"\")\n engine = binding.create_mcjit_compiler(backing_mod, target_machine)\n self.engine = engine",
... | [
"0.57146007",
"0.56284124",
"0.56038344",
"0.55532616",
"0.5481942",
"0.54455805",
"0.5445103",
"0.54039806",
"0.5394971",
"0.5394971",
"0.5394971",
"0.5316189",
"0.5302524",
"0.5259621",
"0.5259621",
"0.52579147",
"0.52550256",
"0.52283067",
"0.519386",
"0.5180122",
"0.51798... | 0.6181505 | 0 |
Return the FriCASExpectFunction class. | def _function_class(self):
return FriCASExpectFunction | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _function_element_class(self):\n return FriCASFunctionElement",
"def func(self):\n return self.__class__",
"def choose_botorch_acqf_class() -> Type[AcquisitionFunction]:\n # NOTE: In the future, this dispatch function could leverage any\n # of the attributes of `BoTorchModel` or kwargs ... | [
"0.7460688",
"0.55869156",
"0.5528139",
"0.5359988",
"0.52939034",
"0.52370685",
"0.51732045",
"0.5147477",
"0.5134911",
"0.51322496",
"0.51305723",
"0.5104154",
"0.5016477",
"0.5000046",
"0.49879676",
"0.4959492",
"0.49496135",
"0.49427474",
"0.49407926",
"0.49324685",
"0.49... | 0.8935946 | 0 |
Returns the FriCAS function element class. | def _function_element_class(self):
return FriCASFunctionElement | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _function_class(self):\n return FriCASExpectFunction",
"def getFunctionClass(functionID):\n d = { 1: Linear,\n 2: LinearDrag,\n 11: Gaussian,\n 12: GaussianDrag,\n 21: Lorentzian,\n 22: LorentzianDrag }\n return d[functionID]",
"def cfunc_type(sel... | [
"0.702561",
"0.63299495",
"0.5833264",
"0.5420016",
"0.53800005",
"0.5317835",
"0.52904075",
"0.5280903",
"0.52620566",
"0.5225637",
"0.5214052",
"0.51871824",
"0.5172217",
"0.5167699",
"0.51647925",
"0.5153158",
"0.51406175",
"0.5124004",
"0.51109064",
"0.5101468",
"0.509311... | 0.8683575 | 0 |
Spawn a new FriCAS commandline session. | def console(self):
fricas_console() | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def cli():\n config, auth, execute_now = read_command_line_arguments()\n main(config, auth, execute_now)",
"def main():\n # Parse arguments.\n parser = argparse.ArgumentParser(description='Console for cluster test.')\n parser.add_argument('-c', dest='command_config_file',\n default='command... | [
"0.60068893",
"0.5865383",
"0.5834904",
"0.58237004",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
... | 0.0 | -1 |
A helper function for converting FriCAS domains to the corresponding Sage object. | def _sage_domain(self):
P = self._check_valid()
name = str(self)
if name == 'Integer':
from sage.rings.all import ZZ
return ZZ
elif name == 'DoubleFloat':
from sage.rings.all import RDF
return RDF
elif name.startswith('Fraction('):
... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def domain(self, domain):",
"def convert_to(self, domain):\n if domain == self.domain:\n return self.copy()\n elif domain == QQ and self.domain == ZZ:\n return self._new(flint.fmpq_mat(self.rep), self.shape, domain)\n elif domain == ZZ and self.domain == QQ:\n ... | [
"0.5323861",
"0.5196539",
"0.5191682",
"0.5174157",
"0.5057873",
"0.50186723",
"0.50181675",
"0.50177693",
"0.5009062",
"0.49185956",
"0.49113917",
"0.4876111",
"0.48465654",
"0.48391303",
"0.48309028",
"0.48160523",
"0.48104942",
"0.48038787",
"0.47553554",
"0.47521913",
"0.... | 0.5950558 | 0 |
Return True if x is of type FriCASElement. | def is_FriCASElement(x):
return isinstance(x, FriCASElement) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def is_(t, x):\n return type(x) is t",
"def _isinstance(cls, x):\n return isinstance(x, cls.PYTHON_TYPE_CHECK)",
"def has_x(self):\n return any(map(lambda s: s.is_x, self))",
"def is_my_case(self, type_):\n return (\n isinstance(self.__apply_sequence(type_), self.declaratio... | [
"0.6683116",
"0.6371762",
"0.6274012",
"0.6056215",
"0.5898661",
"0.58613616",
"0.58362883",
"0.57680345",
"0.56762767",
"0.5658424",
"0.56527305",
"0.5597607",
"0.5569221",
"0.55410606",
"0.55354494",
"0.55278945",
"0.552147",
"0.54687995",
"0.54323465",
"0.54302615",
"0.541... | 0.8895811 | 0 |
Returns the FriCAS interface object defined in sage.interfaces.fricas. | def reduce_load_fricas():
return fricas | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def get_fcigraph(self) -> 'FciGraph':\n return self._core",
"def ICRFFile_Instance():\n return _gmat_py.ICRFFile_Instance()",
"def copy(self):\n return FlorisInterface(self.floris.as_dict())",
"def _get_fcoe_intf_port_type(self):\n return self.__fcoe_intf_port_type",
"def get_fsurface(s... | [
"0.57734627",
"0.52523446",
"0.5242085",
"0.52116334",
"0.51110476",
"0.5084472",
"0.50731456",
"0.5066771",
"0.5064217",
"0.5060581",
"0.4997012",
"0.49558657",
"0.4905208",
"0.4905208",
"0.48911813",
"0.4859921",
"0.48593447",
"0.48388115",
"0.4777076",
"0.4754972",
"0.4744... | 0.4545254 | 70 |
Spawn a new FriCAS commandline session. | def fricas_console():
from sage.repl.rich_output.display_manager import get_display_manager
if not get_display_manager().is_in_terminal():
raise RuntimeError('Can use the console only in the terminal. Try %%fricas magics instead.')
os.system('fricas -nox') | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def cli():\n config, auth, execute_now = read_command_line_arguments()\n main(config, auth, execute_now)",
"def main():\n # Parse arguments.\n parser = argparse.ArgumentParser(description='Console for cluster test.')\n parser.add_argument('-c', dest='command_config_file',\n default='command... | [
"0.60068893",
"0.5865383",
"0.5834904",
"0.58237004",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
"0.5808352",
... | 0.0 | -1 |
optimise the order of matrixchain multiplication | def MatMulOrder(D):
num = len(D)-1 # number of matrix in the chain
print(f"There are {num} matrix to multiply")
M = [[0 for _ in range(num)] for _ in range(num)]
P = [[0 for _ in range(num)] for _ in range(num)]
# i要从大到小
# i == j时, M[i][j]=0,所以不用更新
# i-th矩阵到j-th矩阵的乘的最优值初始化为inf
for i in range(num-2, -1... | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def _multi_matmul_chain_order(arrays):\n n = len(arrays)\n # p stores the dimensions of the matrices\n # Example for p: A_{10x100}, B_{100x5}, C_{5x50} --> p = [10, 100, 5, 50]\n # Using -2 to generalize for shapes that are more than 2 dimmensions\n p = [a.shape[-2] for a in arrays] + [arrays[-1].sh... | [
"0.72562563",
"0.66338646",
"0.65753275",
"0.6564074",
"0.65081275",
"0.64550495",
"0.644735",
"0.64060277",
"0.63821524",
"0.6356483",
"0.6325638",
"0.62876886",
"0.6280246",
"0.6212008",
"0.6144578",
"0.6132298",
"0.6067564",
"0.6057658",
"0.5975746",
"0.5975746",
"0.596742... | 0.697833 | 1 |
Pick GPU if available, else CPU | def get_default_device():
if torch.cuda.is_available():
return torch.device('cuda')
else:
return torch.device('cpu') | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def create_gpu_device_if_present():\n d = dpctl.SyclDevice(\"gpu,cpu\")\n print(\"Selected \" + (\"GPU\" if d.is_gpu else \"CPU\") + \" device\")",
"def return_free_GPU():\r\n if torch.cuda.is_available():\r\n gpu_num = torch.cuda.device_count()\r\n device = torch.device('cuda:{}'.format(g... | [
"0.764121",
"0.7275213",
"0.72637",
"0.7186788",
"0.7095117",
"0.70836645",
"0.7030326",
"0.7010705",
"0.6991861",
"0.698923",
"0.698923",
"0.698923",
"0.6968887",
"0.6954786",
"0.6902639",
"0.6902639",
"0.68919474",
"0.6876222",
"0.6868331",
"0.6805084",
"0.67924947",
"0.6... | 0.6680545 | 26 |
Move tensor(s) to chosen device | def to_device(data, device):
if isinstance(data, (list,tuple)):
return [to_device(x, device) for x in data]
return data.to(device, non_blocking=True) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def move_to_device(input_tensors: List[torch.Tensor],\n target_device: Optional[torch.device],\n non_blocking: bool = False) -> Iterable[torch.Tensor]:\n return (tensor if tensor.device == target_device or target_device is None\n else tensor.to(target_device, non_b... | [
"0.6268374",
"0.5992489",
"0.58285815",
"0.58246756",
"0.5760445",
"0.5746895",
"0.56957227",
"0.5695672",
"0.5674205",
"0.5668232",
"0.5656621",
"0.55133724",
"0.5512487",
"0.55118686",
"0.54767984",
"0.5452559",
"0.54362",
"0.5378659",
"0.5353823",
"0.5353823",
"0.53118145"... | 0.0 | -1 |
Yield a batch of data after moving it to device | def __iter__(self):
for b in self.dl:
yield to_device(b, self.device) | {
"objective": {
"self": [],
"paired": [],
"triplet": [
[
"query",
"document",
"negatives"
]
]
}
} | [
"def __iter__(self):\n for b in self.dl: \n yield to_device(b, self.device) # yield pauses the execution, not store values in memory, forgets about them once iterated\n # no need to remove batch of data from device, done automatically",
"def __iter__(self):\n while True:\n ... | [
"0.80380756",
"0.681836",
"0.6770501",
"0.6756917",
"0.6754636",
"0.67446125",
"0.6652299",
"0.6626395",
"0.6621452",
"0.65598935",
"0.65336037",
"0.6526245",
"0.64669716",
"0.6410819",
"0.63767326",
"0.63346195",
"0.6315828",
"0.63127494",
"0.6302321",
"0.6286355",
"0.627287... | 0.68135273 | 3 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.