sentence1
stringlengths
52
3.87M
sentence2
stringlengths
1
47.2k
label
stringclasses
1 value
def subject(self, value): """ An asn1crypto.x509.Name object, or a dict with at least the following keys: - country_name - state_or_province_name - locality_name - organization_name - common_name Less common keys include: - organizational_unit_name - email_address - street_address - postal_code - business_category - incorporation_locality - incorporation_state_or_province - incorporation_country Uncommon keys include: - surname - title - serial_number - name - given_name - initials - generation_qualifier - dn_qualifier - pseudonym - domain_component All values should be unicode strings """ is_dict = isinstance(value, dict) if not isinstance(value, x509.Name) and not is_dict: raise TypeError(_pretty_message( ''' subject must be an instance of asn1crypto.x509.Name or a dict, not %s ''', _type_name(value) )) if is_dict: value = x509.Name.build(value) self._subject = value
An asn1crypto.x509.Name object, or a dict with at least the following keys: - country_name - state_or_province_name - locality_name - organization_name - common_name Less common keys include: - organizational_unit_name - email_address - street_address - postal_code - business_category - incorporation_locality - incorporation_state_or_province - incorporation_country Uncommon keys include: - surname - title - serial_number - name - given_name - initials - generation_qualifier - dn_qualifier - pseudonym - domain_component All values should be unicode strings
entailment
def subject_public_key(self, value): """ An asn1crypto.keys.PublicKeyInfo or oscrypto.asymmetric.PublicKey object of the subject's public key. """ is_oscrypto = isinstance(value, asymmetric.PublicKey) if not isinstance(value, keys.PublicKeyInfo) and not is_oscrypto: raise TypeError(_pretty_message( ''' subject_public_key must be an instance of asn1crypto.keys.PublicKeyInfo or oscrypto.asymmetric.PublicKey, not %s ''', _type_name(value) )) if is_oscrypto: value = value.asn1 self._subject_public_key = value self._key_identifier = self._subject_public_key.sha1 self._authority_key_identifier = None
An asn1crypto.keys.PublicKeyInfo or oscrypto.asymmetric.PublicKey object of the subject's public key.
entailment
def hash_algo(self, value): """ A unicode string of the hash algorithm to use when signing the request - "sha1" (not recommended), "sha256" or "sha512" """ if value not in set(['sha1', 'sha256', 'sha512']): raise ValueError(_pretty_message( ''' hash_algo must be one of "sha1", "sha256", "sha512", not %s ''', repr(value) )) self._hash_algo = value
A unicode string of the hash algorithm to use when signing the request - "sha1" (not recommended), "sha256" or "sha512"
entailment
def _get_subject_alt(self, name): """ Returns the native value for each value in the subject alt name extension reqiest that is an asn1crypto.x509.GeneralName of the type specified by the name param :param name: A unicode string use to filter the x509.GeneralName objects by - is the choice name x509.GeneralName :return: A list of unicode strings. Empty list indicates no subject alt name extension request. """ if self._subject_alt_name is None: return [] output = [] for general_name in self._subject_alt_name: if general_name.name == name: output.append(general_name.native) return output
Returns the native value for each value in the subject alt name extension reqiest that is an asn1crypto.x509.GeneralName of the type specified by the name param :param name: A unicode string use to filter the x509.GeneralName objects by - is the choice name x509.GeneralName :return: A list of unicode strings. Empty list indicates no subject alt name extension request.
entailment
def _set_subject_alt(self, name, values): """ Replaces all existing asn1crypto.x509.GeneralName objects of the choice represented by the name parameter with the values :param name: A unicode string of the choice name of the x509.GeneralName object :param values: A list of unicode strings to use as the values for the new x509.GeneralName objects """ if self._subject_alt_name is not None: filtered_general_names = [] for general_name in self._subject_alt_name: if general_name.name != name: filtered_general_names.append(general_name) self._subject_alt_name = x509.GeneralNames(filtered_general_names) else: self._subject_alt_name = x509.GeneralNames() if values is not None: for value in values: new_general_name = x509.GeneralName(name=name, value=value) self._subject_alt_name.append(new_general_name) if len(self._subject_alt_name) == 0: self._subject_alt_name = None
Replaces all existing asn1crypto.x509.GeneralName objects of the choice represented by the name parameter with the values :param name: A unicode string of the choice name of the x509.GeneralName object :param values: A list of unicode strings to use as the values for the new x509.GeneralName objects
entailment
def set_extension(self, name, value): """ Sets the value for an extension using a fully constructed Asn1Value object from asn1crypto. Normally this should not be needed, and the convenience attributes should be sufficient. See the definition of asn1crypto.x509.Extension to determine the appropriate object type for a given extension. Extensions are marked as critical when RFC5280 or RFC6960 indicate so. If an extension is validly marked as critical or not (such as certificate policies and extended key usage), this class will mark it as non-critical. :param name: A unicode string of an extension id name from asn1crypto.x509.ExtensionId :param value: A value object per the specs defined by asn1crypto.x509.Extension """ extension = x509.Extension({ 'extn_id': name }) # We use native here to convert OIDs to meaningful names name = extension['extn_id'].native spec = extension.spec('extn_value') if not isinstance(value, spec) and value is not None: raise TypeError(_pretty_message( ''' value must be an instance of %s, not %s ''', _type_name(spec), _type_name(value) )) if name in self._special_extensions: setattr(self, '_%s' % name, value) else: if value is None: if name in self._other_extensions: del self._other_extensions[name] else: self._other_extensions[name] = value
Sets the value for an extension using a fully constructed Asn1Value object from asn1crypto. Normally this should not be needed, and the convenience attributes should be sufficient. See the definition of asn1crypto.x509.Extension to determine the appropriate object type for a given extension. Extensions are marked as critical when RFC5280 or RFC6960 indicate so. If an extension is validly marked as critical or not (such as certificate policies and extended key usage), this class will mark it as non-critical. :param name: A unicode string of an extension id name from asn1crypto.x509.ExtensionId :param value: A value object per the specs defined by asn1crypto.x509.Extension
entailment
def _determine_critical(self, name): """ :return: A boolean indicating the correct value of the critical flag for an extension, based on information from RFC5280 and RFC 6960. The correct value is based on the terminology SHOULD or MUST. """ if name == 'subject_alt_name': return len(self._subject) == 0 if name == 'basic_constraints': return self.ca is True return { 'subject_directory_attributes': False, 'key_usage': True, 'issuer_alt_name': False, 'name_constraints': True, # Based on example EV certificates, non-CA certs have this marked # as non-critical, most likely because existing browsers don't # seem to support policies or name constraints 'certificate_policies': False, 'policy_mappings': True, 'policy_constraints': True, 'extended_key_usage': False, 'inhibit_any_policy': True, 'subject_information_access': False, 'tls_feature': False, 'ocsp_no_check': False, }.get(name, False)
:return: A boolean indicating the correct value of the critical flag for an extension, based on information from RFC5280 and RFC 6960. The correct value is based on the terminology SHOULD or MUST.
entailment
def build(self, signing_private_key): """ Validates the certificate information, constructs an X.509 certificate and then signs it :param signing_private_key: An asn1crypto.keys.PrivateKeyInfo or oscrypto.asymmetric.PrivateKey object for the private key to sign the request with. This should be the private key that matches the public key. :return: An asn1crypto.csr.CertificationRequest object of the request """ is_oscrypto = isinstance(signing_private_key, asymmetric.PrivateKey) if not isinstance(signing_private_key, keys.PrivateKeyInfo) and not is_oscrypto: raise TypeError(_pretty_message( ''' signing_private_key must be an instance of asn1crypto.keys.PrivateKeyInfo or oscrypto.asymmetric.PrivateKey, not %s ''', _type_name(signing_private_key) )) signature_algo = signing_private_key.algorithm if signature_algo == 'ec': signature_algo = 'ecdsa' signature_algorithm_id = '%s_%s' % (self._hash_algo, signature_algo) def _make_extension(name, value): return { 'extn_id': name, 'critical': self._determine_critical(name), 'extn_value': value } extensions = [] for name in sorted(self._special_extensions): value = getattr(self, '_%s' % name) if value is not None: extensions.append(_make_extension(name, value)) for name in sorted(self._other_extensions.keys()): extensions.append(_make_extension(name, self._other_extensions[name])) attributes = [] if extensions: attributes.append({ 'type': 'extension_request', 'values': [extensions] }) certification_request_info = csr.CertificationRequestInfo({ 'version': 'v1', 'subject': self._subject, 'subject_pk_info': self._subject_public_key, 'attributes': attributes }) if signing_private_key.algorithm == 'rsa': sign_func = asymmetric.rsa_pkcs1v15_sign elif signing_private_key.algorithm == 'dsa': sign_func = asymmetric.dsa_sign elif signing_private_key.algorithm == 'ec': sign_func = asymmetric.ecdsa_sign if not is_oscrypto: signing_private_key = asymmetric.load_private_key(signing_private_key) signature = sign_func(signing_private_key, certification_request_info.dump(), self._hash_algo) return csr.CertificationRequest({ 'certification_request_info': certification_request_info, 'signature_algorithm': { 'algorithm': signature_algorithm_id, }, 'signature': signature })
Validates the certificate information, constructs an X.509 certificate and then signs it :param signing_private_key: An asn1crypto.keys.PrivateKeyInfo or oscrypto.asymmetric.PrivateKey object for the private key to sign the request with. This should be the private key that matches the public key. :return: An asn1crypto.csr.CertificationRequest object of the request
entailment
def compute_metrics_cv(self, X, y, **kwargs): '''Compute cross-validated metrics. Trains this model on data X with labels y. Returns a list of dict with keys name, scoring_name, value. Args: X (Union[np.array, pd.DataFrame]): data y (Union[np.array, pd.DataFrame, pd.Series]): labels ''' # compute scores results = self.cv_score_mean(X, y) return results
Compute cross-validated metrics. Trains this model on data X with labels y. Returns a list of dict with keys name, scoring_name, value. Args: X (Union[np.array, pd.DataFrame]): data y (Union[np.array, pd.DataFrame, pd.Series]): labels
entailment
def cv_score_mean(self, X, y): '''Compute mean score across cross validation folds. Split data and labels into cross validation folds and fit the model for each fold. Then, for each scoring type in scorings, compute the score. Finally, average the scores across folds. Returns a dictionary mapping scoring to score. Args: X (np.array): data y (np.array): labels scorings (List[str]): scoring types ''' X, y = self._format_inputs(X, y) if self.problem_type.binary_classification: kf = StratifiedKFold( shuffle=True, random_state=RANDOM_STATE + 3) elif self.problem_type.multi_classification: self.target_type_transformer.inverse_transform(y) transformer = self.target_type_transformer kf = StratifiedKFoldMultiClassIndicator( transformer, shuffle=True, n_splits=3, random_state=RANDOM_STATE + 3) elif self.problem_type.regression: kf = KFold(shuffle=True, n_splits=3, random_state=RANDOM_STATE + 4) else: raise NotImplementedError scoring = { scorer_info.name: scorer_info.scorer for scorer_info in self.scorers_info } cv_results = cross_validate( self.estimator, X, y, scoring=scoring, cv=kf, return_train_score=False) # post-processing results = self._process_cv_results(cv_results) return results
Compute mean score across cross validation folds. Split data and labels into cross validation folds and fit the model for each fold. Then, for each scoring type in scorings, compute the score. Finally, average the scores across folds. Returns a dictionary mapping scoring to score. Args: X (np.array): data y (np.array): labels scorings (List[str]): scoring types
entailment
def get_contrib_features(project_root): """Get contributed features for a project at project_root For a project ``foo``, walks modules within the ``foo.features.contrib`` subpackage. A single object that is an instance of ``ballet.Feature`` is imported if present in each module. The resulting ``Feature`` objects are collected. Args: project_root (str, path-like): Path to project root Returns: List[ballet.Feature]: list of Feature objects """ # TODO Project should require ModuleType project = Project(project_root) contrib = project._resolve('.features.contrib') return _get_contrib_features(contrib)
Get contributed features for a project at project_root For a project ``foo``, walks modules within the ``foo.features.contrib`` subpackage. A single object that is an instance of ``ballet.Feature`` is imported if present in each module. The resulting ``Feature`` objects are collected. Args: project_root (str, path-like): Path to project root Returns: List[ballet.Feature]: list of Feature objects
entailment
def _get_contrib_features(module): """Get contributed features from within given module Be very careful with untrusted code. The module/package will be walked, every submodule will be imported, and all the code therein will be executed. But why would you be trying to import from an untrusted package anyway? Args: contrib (module): module (standalone or package) that contains feature definitions Returns: List[Feature]: list of features """ if isinstance(module, types.ModuleType): # any module that has a __path__ attribute is also a package if hasattr(module, '__path__'): yield from _get_contrib_features_from_package(module) else: yield _get_contrib_feature_from_module(module) else: raise ValueError('Input is not a module')
Get contributed features from within given module Be very careful with untrusted code. The module/package will be walked, every submodule will be imported, and all the code therein will be executed. But why would you be trying to import from an untrusted package anyway? Args: contrib (module): module (standalone or package) that contains feature definitions Returns: List[Feature]: list of features
entailment
def quickstart(): """Generate a brand-new ballet project""" import ballet.templating import ballet.util.log ballet.util.log.enable(level='INFO', format=ballet.util.log.SIMPLE_LOG_FORMAT, echo=False) ballet.templating.render_project_template()
Generate a brand-new ballet project
entailment
def update_project_template(push): """Update an existing ballet project from the upstream template""" import ballet.update import ballet.util.log ballet.util.log.enable(level='INFO', format=ballet.util.log.SIMPLE_LOG_FORMAT, echo=False) ballet.update.update_project_template(push=push)
Update an existing ballet project from the upstream template
entailment
def start_new_feature(): """Start working on a new feature from a template""" import ballet.templating import ballet.util.log ballet.util.log.enable(level='INFO', format=ballet.util.log.SIMPLE_LOG_FORMAT, echo=False) ballet.templating.start_new_feature()
Start working on a new feature from a template
entailment
def write_tabular(obj, filepath): """Write tabular object in HDF5 or pickle format Args: obj (array or DataFrame): tabular object to write filepath (path-like): path to write to; must end in '.h5' or '.pkl' """ _, fn, ext = splitext2(filepath) if ext == '.h5': _write_tabular_h5(obj, filepath) elif ext == '.pkl': _write_tabular_pickle(obj, filepath) else: raise NotImplementedError
Write tabular object in HDF5 or pickle format Args: obj (array or DataFrame): tabular object to write filepath (path-like): path to write to; must end in '.h5' or '.pkl'
entailment
def read_tabular(filepath): """Read tabular object in HDF5 or pickle format Args: filepath (path-like): path to read to; must end in '.h5' or '.pkl' """ _, fn, ext = splitext2(filepath) if ext == '.h5': return _read_tabular_h5(filepath) elif ext == '.pkl': return _read_tabular_pickle(filepath) else: raise NotImplementedError
Read tabular object in HDF5 or pickle format Args: filepath (path-like): path to read to; must end in '.h5' or '.pkl'
entailment
def load_table_from_config(input_dir, config): """Load table from table config dict Args: input_dir (path-like): directory containing input files config (dict): mapping with keys 'name', 'path', and 'pd_read_kwargs'. Returns: pd.DataFrame """ path = pathlib.Path(input_dir).joinpath(config['path']) kwargs = config['pd_read_kwargs'] return pd.read_csv(path, **kwargs)
Load table from table config dict Args: input_dir (path-like): directory containing input files config (dict): mapping with keys 'name', 'path', and 'pd_read_kwargs'. Returns: pd.DataFrame
entailment
def validate_feature_api(project, force=False): """Validate feature API""" if not force and not project.on_pr(): raise SkippedValidationTest('Not on PR') validator = FeatureApiValidator(project) result = validator.validate() if not result: raise InvalidFeatureApi
Validate feature API
entailment
def evaluate_feature_performance(project, force=False): """Evaluate feature performance""" if not force and not project.on_pr(): raise SkippedValidationTest('Not on PR') out = project.build() X_df, y, features = out['X_df'], out['y'], out['features'] proposed_feature = get_proposed_feature(project) accepted_features = get_accepted_features(features, proposed_feature) evaluator = GFSSFAcceptanceEvaluator(X_df, y, accepted_features) accepted = evaluator.judge(proposed_feature) if not accepted: raise FeatureRejected
Evaluate feature performance
entailment
def prune_existing_features(project, force=False): """Prune existing features""" if not force and not project.on_master_after_merge(): raise SkippedValidationTest('Not on master') out = project.build() X_df, y, features = out['X_df'], out['y'], out['features'] proposed_feature = get_proposed_feature(project) accepted_features = get_accepted_features(features, proposed_feature) evaluator = GFSSFPruningEvaluator( X_df, y, accepted_features, proposed_feature) redundant_features = evaluator.prune() # propose removal for feature in redundant_features: logger.debug(PRUNER_MESSAGE + feature.source) return redundant_features
Prune existing features
entailment
def validate(package, test_target_type=None): """Entrypoint for ./validate.py script in ballet projects""" project = Project(package) if test_target_type is None: test_target_type = detect_target_type() if test_target_type == BalletTestTypes.PROJECT_STRUCTURE_VALIDATION: check_project_structure(project) elif test_target_type == BalletTestTypes.FEATURE_API_VALIDATION: validate_feature_api(project) elif test_target_type == BalletTestTypes.FEATURE_ACCEPTANCE_EVALUTION: evaluate_feature_performance(project) elif test_target_type == ( BalletTestTypes.FEATURE_PRUNING_EVALUATION): prune_existing_features(project) else: raise NotImplementedError( 'Unsupported test target type: {test_target_type}' .format(test_target_type=test_target_type))
Entrypoint for ./validate.py script in ballet projects
entailment
def spliceext(filepath, s): """Add s into filepath before the extension Args: filepath (str, path): file path s (str): string to splice Returns: str """ root, ext = os.path.splitext(safepath(filepath)) return root + s + ext
Add s into filepath before the extension Args: filepath (str, path): file path s (str): string to splice Returns: str
entailment
def replaceext(filepath, new_ext): """Replace any existing file extension with a new one Example:: >>> replaceext('/foo/bar.txt', 'py') '/foo/bar.py' >>> replaceext('/foo/bar.txt', '.doc') '/foo/bar.doc' Args: filepath (str, path): file path new_ext (str): new file extension; if a leading dot is not included, it will be added. Returns: Tuple[str] """ if new_ext and new_ext[0] != '.': new_ext = '.' + new_ext root, ext = os.path.splitext(safepath(filepath)) return root + new_ext
Replace any existing file extension with a new one Example:: >>> replaceext('/foo/bar.txt', 'py') '/foo/bar.py' >>> replaceext('/foo/bar.txt', '.doc') '/foo/bar.doc' Args: filepath (str, path): file path new_ext (str): new file extension; if a leading dot is not included, it will be added. Returns: Tuple[str]
entailment
def splitext2(filepath): """Split filepath into root, filename, ext Args: filepath (str, path): file path Returns: str """ root, filename = os.path.split(safepath(filepath)) filename, ext = os.path.splitext(safepath(filename)) return root, filename, ext
Split filepath into root, filename, ext Args: filepath (str, path): file path Returns: str
entailment
def isemptyfile(filepath): """Determine if the file both exists and isempty Args: filepath (str, path): file path Returns: bool """ exists = os.path.exists(safepath(filepath)) if exists: filesize = os.path.getsize(safepath(filepath)) return filesize == 0 else: return False
Determine if the file both exists and isempty Args: filepath (str, path): file path Returns: bool
entailment
def synctree(src, dst, onexist=None): """Recursively sync files at directory src to dst This is more or less equivalent to:: cp -n -R ${src}/ ${dst}/ If a file at the same path exists in src and dst, it is NOT overwritten in dst. Pass ``onexist`` in order to raise an error on such conditions. Args: src (path-like): source directory dst (path-like): destination directory, does not need to exist onexist (callable): function to call if file exists at destination, takes the full path to destination file as only argument """ src = pathlib.Path(src).resolve() dst = pathlib.Path(dst).resolve() if not src.is_dir(): raise ValueError if dst.exists() and not dst.is_dir(): raise ValueError if onexist is None: def onexist(): pass _synctree(src, dst, onexist)
Recursively sync files at directory src to dst This is more or less equivalent to:: cp -n -R ${src}/ ${dst}/ If a file at the same path exists in src and dst, it is NOT overwritten in dst. Pass ``onexist`` in order to raise an error on such conditions. Args: src (path-like): source directory dst (path-like): destination directory, does not need to exist onexist (callable): function to call if file exists at destination, takes the full path to destination file as only argument
entailment
def calculate_disc_entropy(X): r"""Calculates the exact Shannon entropy of a discrete dataset, using empirical probabilities according to the equation: $ H(X) = -\sum(c \in X) p(c) \times \log(p(c)) $ Where $ p(c) $ is calculated as the frequency of c in X. If X's columns logically represent continuous features, it is better to use the estimate_cont_entropy function. If you are unsure of which to use, estimate_entropy can take datasets of mixed discrete and continuous functions. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) Returns: float: A floating-point number representing the dataset entropy. """ X = asarray2d(X) n_samples, _ = X.shape _, counts = np.unique(X, axis=0, return_counts=True) empirical_p = counts * 1.0 / n_samples log_p = np.log(empirical_p) entropy = -np.sum(np.multiply(empirical_p, log_p)) return entropy
r"""Calculates the exact Shannon entropy of a discrete dataset, using empirical probabilities according to the equation: $ H(X) = -\sum(c \in X) p(c) \times \log(p(c)) $ Where $ p(c) $ is calculated as the frequency of c in X. If X's columns logically represent continuous features, it is better to use the estimate_cont_entropy function. If you are unsure of which to use, estimate_entropy can take datasets of mixed discrete and continuous functions. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) Returns: float: A floating-point number representing the dataset entropy.
entailment
def estimate_cont_entropy(X, epsilon=None): """Estimate the Shannon entropy of a discrete dataset. Based off the Kraskov Estimator [1] and Kozachenko [2] estimators for a dataset's Shannon entropy. The function relies on nonparametric methods based on entropy estimation from k-nearest neighbors distances as proposed in [1] and augmented in [2] for mutual information estimation. If X's columns logically represent discrete features, it is better to use the calculate_disc_entropy function. If you are unsure of which to use, estimate_entropy can take datasets of mixed discrete and continuous functions. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) epsilon (array-like): An array with shape (n_samples, 1) that is the epsilon used in Kraskov Estimator. Represents the chebyshev distance from an element to its k-th nearest neighbor in the full dataset. Returns: float: A floating-point number. If epsilon is not provided, this will be the Kozacheko Estimator of the dataset's entropy. If epsilon is provided, this is a partial estimation of the Kraskov entropy estimator. The bias is cancelled out when computing mutual information. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. .. [2] L. F. Kozachenko, N. N. Leonenko, "Sample Estimate of the Entropy of a Random Vector:, Probl. Peredachi Inf., 23:2 (1987), 9-16 """ X = asarray2d(X) n_samples, n_features = X.shape if n_samples <= 1: return 0 nn = NearestNeighbors( metric='chebyshev', n_neighbors=NUM_NEIGHBORS, algorithm='kd_tree') nn.fit(X) if epsilon is None: # If epsilon is not provided, revert to the Kozachenko Estimator n_neighbors = NUM_NEIGHBORS radius = 0 # While we have non-zero radii, calculate for a larger k # Potentially expensive while not np.all(radius) and n_neighbors < n_samples: distances, _ = nn.kneighbors( n_neighbors=n_neighbors, return_distance=True) radius = distances[:, -1] n_neighbors += 1 if n_neighbors == n_samples: # This case only happens if all samples are the same # e.g. this isn't a continuous sample... raise ValueError('Should not have discrete column to estimate') return -digamma(n_neighbors) + digamma(n_samples) + \ n_features * np.mean(np.log(2 * radius)) else: ind = nn.radius_neighbors( radius=epsilon.ravel(), return_distance=False) nx = np.array([i.size for i in ind]) return - np.mean(digamma(nx + 1)) + digamma(n_samples)
Estimate the Shannon entropy of a discrete dataset. Based off the Kraskov Estimator [1] and Kozachenko [2] estimators for a dataset's Shannon entropy. The function relies on nonparametric methods based on entropy estimation from k-nearest neighbors distances as proposed in [1] and augmented in [2] for mutual information estimation. If X's columns logically represent discrete features, it is better to use the calculate_disc_entropy function. If you are unsure of which to use, estimate_entropy can take datasets of mixed discrete and continuous functions. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) epsilon (array-like): An array with shape (n_samples, 1) that is the epsilon used in Kraskov Estimator. Represents the chebyshev distance from an element to its k-th nearest neighbor in the full dataset. Returns: float: A floating-point number. If epsilon is not provided, this will be the Kozacheko Estimator of the dataset's entropy. If epsilon is provided, this is a partial estimation of the Kraskov entropy estimator. The bias is cancelled out when computing mutual information. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. .. [2] L. F. Kozachenko, N. N. Leonenko, "Sample Estimate of the Entropy of a Random Vector:, Probl. Peredachi Inf., 23:2 (1987), 9-16
entailment
def estimate_entropy(X, epsilon=None): r"""Estimate a dataset's Shannon entropy. This function can take datasets of mixed discrete and continuous features, and uses a set of heuristics to determine which functions to apply to each. Because this function is a subroutine in a mutual information estimator, we employ the Kozachenko Estimator[1] for continuous features when this function is _not_ used for mutual information and an adaptation of the Kraskov Estimator[2] when it is. Let X be made of continuous features c and discrete features d. To deal with both continuous and discrete features, We use the following reworking of entropy: $ H(X) = H(c,d) = \sum_{x \in d} p(x) \times H(c(x)) + H(d) $ Where c(x) is a dataset that represents the rows of the continuous dataset in the same row as a discrete column with value x in the original dataset. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) epsilon (array-like): An array with shape (n_samples, 1) that is the epsilon used in Kraskov Estimator. Represents the chebyshev distance from an element to its k-th nearest neighbor in the full dataset. Returns: float: A floating-point number representing the entropy in X. If the dataset is fully discrete, an exact calculation is done. If this is not the case and epsilon is not provided, this will be the Kozacheko Estimator of the dataset's entropy. If epsilon is provided, this is a partial estimation of the Kraskov entropy estimator. The bias is cancelled out when computing mutual information. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. .. [2] L. F. Kozachenko, N. N. Leonenko, "Sample Estimate of the Entropy of a Random Vector:, Probl. Peredachi Inf., 23:2 (1987), 9-16. """ X = asarray2d(X) n_samples, n_features = X.shape if n_features < 1: return 0 disc_mask = _get_discrete_columns(X) cont_mask = ~disc_mask # If our dataset is fully discrete/continuous, do something easier if np.all(disc_mask): return calculate_disc_entropy(X) elif np.all(cont_mask): return estimate_cont_entropy(X, epsilon) # Separate the dataset into discrete and continuous datasets d,c disc_features = asarray2d(X[:, disc_mask]) cont_features = asarray2d(X[:, cont_mask]) entropy = 0 uniques, counts = np.unique(disc_features, axis=0, return_counts=True) empirical_p = counts / n_samples # $\sum_{x \in d} p(x) \times H(c(x))$ for i in range(counts.size): unique_mask = np.all(disc_features == uniques[i], axis=1) selected_cont_samples = cont_features[unique_mask, :] if epsilon is None: selected_epsilon = None else: selected_epsilon = epsilon[unique_mask, :] conditional_cont_entropy = estimate_cont_entropy( selected_cont_samples, selected_epsilon) entropy += empirical_p[i] * conditional_cont_entropy # H(d) entropy += calculate_disc_entropy(disc_features) if epsilon is None: entropy = max(0, entropy) return entropy
r"""Estimate a dataset's Shannon entropy. This function can take datasets of mixed discrete and continuous features, and uses a set of heuristics to determine which functions to apply to each. Because this function is a subroutine in a mutual information estimator, we employ the Kozachenko Estimator[1] for continuous features when this function is _not_ used for mutual information and an adaptation of the Kraskov Estimator[2] when it is. Let X be made of continuous features c and discrete features d. To deal with both continuous and discrete features, We use the following reworking of entropy: $ H(X) = H(c,d) = \sum_{x \in d} p(x) \times H(c(x)) + H(d) $ Where c(x) is a dataset that represents the rows of the continuous dataset in the same row as a discrete column with value x in the original dataset. Args: X (array-like): An array-like (np arr, pandas df, etc.) with shape (n_samples, n_features) or (n_samples) epsilon (array-like): An array with shape (n_samples, 1) that is the epsilon used in Kraskov Estimator. Represents the chebyshev distance from an element to its k-th nearest neighbor in the full dataset. Returns: float: A floating-point number representing the entropy in X. If the dataset is fully discrete, an exact calculation is done. If this is not the case and epsilon is not provided, this will be the Kozacheko Estimator of the dataset's entropy. If epsilon is provided, this is a partial estimation of the Kraskov entropy estimator. The bias is cancelled out when computing mutual information. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. .. [2] L. F. Kozachenko, N. N. Leonenko, "Sample Estimate of the Entropy of a Random Vector:, Probl. Peredachi Inf., 23:2 (1987), 9-16.
entailment
def _calculate_epsilon(X): """Calculates epsilon, a subroutine for the Kraskov Estimator [1] Represents the chebyshev distance of each dataset element to its K-th nearest neighbor. Args: X (array-like): An array with shape (n_samples, n_features) Returns: array-like: An array with shape (n_samples, 1) representing epsilon as described above. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. """ disc_mask = _get_discrete_columns(X) if np.all(disc_mask): # if all discrete columns, there's no point getting epsilon return 0 cont_features = X[:, ~disc_mask] nn = NearestNeighbors(metric='chebyshev', n_neighbors=NUM_NEIGHBORS) nn.fit(cont_features) distances, _ = nn.kneighbors() epsilon = np.nextafter(distances[:, -1], 0) return asarray2d(epsilon)
Calculates epsilon, a subroutine for the Kraskov Estimator [1] Represents the chebyshev distance of each dataset element to its K-th nearest neighbor. Args: X (array-like): An array with shape (n_samples, n_features) Returns: array-like: An array with shape (n_samples, 1) representing epsilon as described above. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004.
entailment
def estimate_conditional_information(x, y, z): """ Estimate the conditional mutual information of three datasets. Conditional mutual information is the mutual information of two datasets, given a third: $ I(x;y|z) = H(x,z) + H(y,z) - H(x,y,z) - H(z) $ Where H(x) is the Shannon entropy of x. For continuous datasets, adapts the Kraskov Estimator [1] for mutual information. Equation 8 still holds because the epsilon terms cancel out: Let d_x, represent the dimensionality of the continuous portion of x. Then, we see that: d_xz + d_yz - d_xyz - d_z = (d_x + d_z) + (d_y + d_z) - (d_x + d_y + d_z) - d_z = 0 Args: x (array-like): An array with shape (n_samples, n_features_x) y (array-like): An array with shape (n_samples, n_features_y) z (array-like): An array with shape (n_samples, n_features_z). This is the dataset being conditioned on. Returns: float: A floating point number representing the conditional mutual information of x and y given z. This calculation is *exact* for entirely discrete datasets and *approximate* if there are continuous columns present. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. """ xz = np.concatenate((x, z), axis=1) yz = np.concatenate((y, z), axis=1) xyz = np.concatenate((xz, y), axis=1) epsilon = _calculate_epsilon(xyz) h_xz = estimate_entropy(xz, epsilon) h_yz = estimate_entropy(yz, epsilon) h_xyz = estimate_entropy(xyz, epsilon) h_z = estimate_entropy(z, epsilon) return max(0, h_xz + h_yz - h_xyz - h_z)
Estimate the conditional mutual information of three datasets. Conditional mutual information is the mutual information of two datasets, given a third: $ I(x;y|z) = H(x,z) + H(y,z) - H(x,y,z) - H(z) $ Where H(x) is the Shannon entropy of x. For continuous datasets, adapts the Kraskov Estimator [1] for mutual information. Equation 8 still holds because the epsilon terms cancel out: Let d_x, represent the dimensionality of the continuous portion of x. Then, we see that: d_xz + d_yz - d_xyz - d_z = (d_x + d_z) + (d_y + d_z) - (d_x + d_y + d_z) - d_z = 0 Args: x (array-like): An array with shape (n_samples, n_features_x) y (array-like): An array with shape (n_samples, n_features_y) z (array-like): An array with shape (n_samples, n_features_z). This is the dataset being conditioned on. Returns: float: A floating point number representing the conditional mutual information of x and y given z. This calculation is *exact* for entirely discrete datasets and *approximate* if there are continuous columns present. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004.
entailment
def estimate_mutual_information(x, y): """Estimate the mutual information of two datasets. Mutual information is a measure of dependence between two datasets and is calculated as: $I(x;y) = H(x) + H(y) - H(x,y)$ Where H(x) is the Shannon entropy of x. For continuous datasets, adapts the Kraskov Estimator [1] for mutual information. Args: x (array-like): An array with shape (n_samples, n_features_x) y (array-like): An array with shape (n_samples, n_features_y) Returns: float: A floating point number representing the mutual information of x and y. This calculation is *exact* for entirely discrete datasets and *approximate* if there are continuous columns present. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004. """ xy = np.concatenate((x, y), axis=1) epsilon = _calculate_epsilon(xy) h_x = estimate_entropy(x, epsilon) h_y = estimate_entropy(y, epsilon) h_xy = estimate_entropy(xy, epsilon) return max(0, h_x + h_y - h_xy)
Estimate the mutual information of two datasets. Mutual information is a measure of dependence between two datasets and is calculated as: $I(x;y) = H(x) + H(y) - H(x,y)$ Where H(x) is the Shannon entropy of x. For continuous datasets, adapts the Kraskov Estimator [1] for mutual information. Args: x (array-like): An array with shape (n_samples, n_features_x) y (array-like): An array with shape (n_samples, n_features_y) Returns: float: A floating point number representing the mutual information of x and y. This calculation is *exact* for entirely discrete datasets and *approximate* if there are continuous columns present. References: .. [1] A. Kraskov, H. Stogbauer and P. Grassberger, "Estimating mutual information". Phys. Rev. E 69, 2004.
entailment
def get_diff_endpoints_from_commit_range(repo, commit_range): """Get endpoints of a diff given a commit range The resulting endpoints can be diffed directly:: a, b = get_diff_endpoints_from_commit_range(repo, commit_range) a.diff(b) For details on specifying git diffs, see ``git diff --help``. For details on specifying revisions, see ``git help revisions``. Args: repo (git.Repo): Repo object initialized with project root commit_range (str): commit range as would be interpreted by ``git diff`` command. Unfortunately only patterns of the form ``a..b`` and ``a...b`` are accepted. Note that the latter pattern finds the merge-base of a and b and uses it as the starting point for the diff. Returns: Tuple[git.Commit, git.Commit]: starting commit, ending commit ( inclusive) Raises: ValueError: commit_range is empty or ill-formed See also: <https://stackoverflow.com/q/7251477> """ if not commit_range: raise ValueError('commit_range cannot be empty') result = re_find(COMMIT_RANGE_REGEX, commit_range) if not result: raise ValueError( 'Expected diff str of the form \'a..b\' or \'a...b\' (got {})' .format(commit_range)) a, b = result['a'], result['b'] a, b = repo.rev_parse(a), repo.rev_parse(b) if result['thirddot']: a = one_or_raise(repo.merge_base(a, b)) return a, b
Get endpoints of a diff given a commit range The resulting endpoints can be diffed directly:: a, b = get_diff_endpoints_from_commit_range(repo, commit_range) a.diff(b) For details on specifying git diffs, see ``git diff --help``. For details on specifying revisions, see ``git help revisions``. Args: repo (git.Repo): Repo object initialized with project root commit_range (str): commit range as would be interpreted by ``git diff`` command. Unfortunately only patterns of the form ``a..b`` and ``a...b`` are accepted. Note that the latter pattern finds the merge-base of a and b and uses it as the starting point for the diff. Returns: Tuple[git.Commit, git.Commit]: starting commit, ending commit ( inclusive) Raises: ValueError: commit_range is empty or ill-formed See also: <https://stackoverflow.com/q/7251477>
entailment
def set_config_variables(repo, variables): """Set config variables Args: repo (git.Repo): repo variables (dict): entries of the form 'user.email': 'you@example.com' """ with repo.config_writer() as writer: for k, value in variables.items(): section, option = k.split('.') writer.set_value(section, option, value) writer.release()
Set config variables Args: repo (git.Repo): repo variables (dict): entries of the form 'user.email': 'you@example.com'
entailment
def validate(self): """Collect and validate all new features""" changes = self.change_collector.collect_changes() features = [] imported_okay = True for importer, modname, modpath in changes.new_feature_info: try: mod = importer() features.extend(_get_contrib_features(mod)) except (ImportError, SyntaxError): logger.info( 'Failed to import module at {}' .format(modpath)) logger.exception('Exception details: ') imported_okay = False if not imported_okay: return False # if no features were added at all, reject if not features: logger.info('Failed to collect any new features.') return False return all( validate_feature_api(feature, self.X, self.y, subsample=False) for feature in features )
Collect and validate all new features
entailment
def load_config_at_path(path): """Load config at exact path Args: path (path-like): path to config file Returns: dict: config dict """ if path.exists() and path.is_file(): with path.open('r') as f: return yaml.load(f, Loader=yaml.SafeLoader) else: raise ConfigurationError("Couldn't find ballet.yml config file.")
Load config at exact path Args: path (path-like): path to config file Returns: dict: config dict
entailment
def config_get(config, *path, default=None): """Get a configuration option following a path through the config Example usage: >>> config_get(config, 'problem', 'problem_type_details', 'scorer', default='accuracy') Args: config (dict): config dict *path (list[str]): List of config sections and options to follow. default (default=None): A default value to return in the case that the option does not exist. """ o = object() result = get_in(config, path, default=o) if result is not o: return result else: return default
Get a configuration option following a path through the config Example usage: >>> config_get(config, 'problem', 'problem_type_details', 'scorer', default='accuracy') Args: config (dict): config dict *path (list[str]): List of config sections and options to follow. default (default=None): A default value to return in the case that the option does not exist.
entailment
def make_config_get(conf_path): """Return a function to get configuration options for a specific project Args: conf_path (path-like): path to project's conf file (i.e. foo.conf module) """ project_root = _get_project_root_from_conf_path(conf_path) config = load_config_in_dir(project_root) return partial(config_get, config)
Return a function to get configuration options for a specific project Args: conf_path (path-like): path to project's conf file (i.e. foo.conf module)
entailment
def relative_to_contrib(diff, project): """Compute relative path of changed file to contrib dir Args: diff (git.diff.Diff): file diff project (Project): project Returns: Path """ path = pathlib.Path(diff.b_path) contrib_path = project.contrib_module_path return path.relative_to(contrib_path)
Compute relative path of changed file to contrib dir Args: diff (git.diff.Diff): file diff project (Project): project Returns: Path
entailment
def pr_num(self): """Return the PR number or None if not on a PR""" result = get_pr_num(repo=self.repo) if result is None: result = get_travis_pr_num() return result
Return the PR number or None if not on a PR
entailment
def branch(self): """Return whether the project is on master branch""" result = get_branch(repo=self.repo) if result is None: result = get_travis_branch() return result
Return whether the project is on master branch
entailment
def path(self): """Return the project path (aka project root) If ``package.__file__`` is ``/foo/foo/__init__.py``, then project.path should be ``/foo``. """ return pathlib.Path(self.package.__file__).resolve().parent.parent
Return the project path (aka project root) If ``package.__file__`` is ``/foo/foo/__init__.py``, then project.path should be ``/foo``.
entailment
def asarray2d(a): """Cast to 2d array""" arr = np.asarray(a) if arr.ndim == 1: arr = arr.reshape(-1, 1) return arr
Cast to 2d array
entailment
def get_arr_desc(arr): """Get array description, in the form '<array type> <array shape>'""" type_ = type(arr).__name__ # see also __qualname__ shape = getattr(arr, 'shape', None) if shape is not None: desc = '{type_} {shape}' else: desc = '{type_} <no shape>' return desc.format(type_=type_, shape=shape)
Get array description, in the form '<array type> <array shape>
entailment
def indent(text, n=4): """Indent each line of text by n spaces""" _indent = ' ' * n return '\n'.join(_indent + line for line in text.split('\n'))
Indent each line of text by n spaces
entailment
def has_nans(obj): """Check if obj has any NaNs Compatible with different behavior of np.isnan, which sometimes applies over all axes (py35, py35) and sometimes does not (py34). """ nans = np.isnan(obj) while np.ndim(nans): nans = np.any(nans) return bool(nans)
Check if obj has any NaNs Compatible with different behavior of np.isnan, which sometimes applies over all axes (py35, py35) and sometimes does not (py34).
entailment
def needs_path(f): """Wraps a function that accepts path-like to give it a pathlib.Path""" @wraps(f) def wrapped(pathlike, *args, **kwargs): path = pathlib.Path(pathlike) return f(path, *args, **kwargs) return wrapped
Wraps a function that accepts path-like to give it a pathlib.Path
entailment
def import_module_at_path(modname, modpath): """Import module from path that may not be on system path Args: modname (str): module name from package root, e.g. foo.bar modpath (str): absolute path to module itself, e.g. /home/user/foo/bar.py. In the case of a module that is a package, then the path should be specified as '/home/user/foo' and a file '/home/user/foo/__init__.py' *must be present* or the import will fail. Examples: >>> modname = 'foo.bar.baz' >>> modpath = '/home/user/foo/bar/baz.py' >>> import_module_at_path(modname, modpath) <module 'foo.bar.baz' from '/home/user/foo/bar/baz.py'> >>> modname = 'foo.bar' >>> modpath = '/home/user/foo/bar' >>> import_module_at_path(modname, modpath) <module 'foo.bar' from '/home/user/foo/bar/__init__.py'> """ # TODO just keep navigating up in the source tree until an __init__.py is # not found? modpath = pathlib.Path(modpath).resolve() if modpath.name == '__init__.py': # TODO improve debugging output with recommend change raise ValueError('Don\'t provide the __init__.py!') def is_package(modpath): return modpath.suffix != '.py' def has_init(dir): return dir.joinpath('__init__.py').is_file() def has_package_structure(modname, modpath): modparts = modname.split('.') n = len(modparts) dir = modpath if not is_package(modpath): n = n - 1 dir = dir.parent while n > 0: if not has_init(dir): return False dir = dir.parent n = n - 1 return True if not has_package_structure(modname, modpath): raise ImportError('Module does not have valid package structure.') parentpath = str(pathlib.Path(modpath).parent) finder = pkgutil.get_importer(parentpath) loader = finder.find_module(modname) if loader is None: raise ImportError( 'Failed to find loader for module {} within dir {}' .format(modname, parentpath)) mod = loader.load_module(modname) # TODO figure out what to do about this assert mod.__name__ == modname return mod
Import module from path that may not be on system path Args: modname (str): module name from package root, e.g. foo.bar modpath (str): absolute path to module itself, e.g. /home/user/foo/bar.py. In the case of a module that is a package, then the path should be specified as '/home/user/foo' and a file '/home/user/foo/__init__.py' *must be present* or the import will fail. Examples: >>> modname = 'foo.bar.baz' >>> modpath = '/home/user/foo/bar/baz.py' >>> import_module_at_path(modname, modpath) <module 'foo.bar.baz' from '/home/user/foo/bar/baz.py'> >>> modname = 'foo.bar' >>> modpath = '/home/user/foo/bar' >>> import_module_at_path(modname, modpath) <module 'foo.bar' from '/home/user/foo/bar/__init__.py'>
entailment
def relpath_to_modname(relpath): """Convert relative path to module name Within a project, a path to the source file is uniquely identified with a module name. Relative paths of the form 'foo/bar' are *not* converted to module names 'foo.bar', because (1) they identify directories, not regular files, and (2) already 'foo/bar/__init__.py' would claim that conversion. Args: relpath (str): Relative path from some location on sys.path Example: >>> relpath_to_modname('ballet/util/_util.py') 'ballet.util._util' """ # don't try to resolve! p = pathlib.Path(relpath) if p.name == '__init__.py': p = p.parent elif p.suffix == '.py': p = p.with_suffix('') else: msg = 'Cannot convert a non-python file to a modname' msg_detail = 'The relpath given is: {}'.format(relpath) logger.error(msg + '\n' + msg_detail) raise ValueError(msg) return '.'.join(p.parts)
Convert relative path to module name Within a project, a path to the source file is uniquely identified with a module name. Relative paths of the form 'foo/bar' are *not* converted to module names 'foo.bar', because (1) they identify directories, not regular files, and (2) already 'foo/bar/__init__.py' would claim that conversion. Args: relpath (str): Relative path from some location on sys.path Example: >>> relpath_to_modname('ballet/util/_util.py') 'ballet.util._util'
entailment
def modname_to_relpath(modname, project_root=None, add_init=True): """Convert module name to relative path. The project root is usually needed to detect if the module is a package, in which case the relevant file is the `__init__.py` within the subdirectory. Example: >>> modname_to_relpath('foo.features') 'foo/features.py' >>> modname_to_relpath('foo.features', project_root='/path/to/project') 'foo/features/__init__.py' Args: modname (str): Module name, e.g. `os.path` project_root (str): Path to project root add_init (bool): Whether to add `__init__.py` to the path of modules that are packages. Defaults to True Returns: str """ parts = modname.split('.') relpath = pathlib.Path(*parts) # is the module a package? if so, the relpath identifies a directory # it is easier to check for whether a file is a directory than to try to # import the module dynamically and see whether it is a package if project_root is not None: relpath_resolved = pathlib.Path(project_root).joinpath(relpath) else: relpath_resolved = relpath if relpath_resolved.is_dir(): if add_init: relpath = relpath.joinpath('__init__.py') else: relpath = str(relpath) + '.py' return str(relpath)
Convert module name to relative path. The project root is usually needed to detect if the module is a package, in which case the relevant file is the `__init__.py` within the subdirectory. Example: >>> modname_to_relpath('foo.features') 'foo/features.py' >>> modname_to_relpath('foo.features', project_root='/path/to/project') 'foo/features/__init__.py' Args: modname (str): Module name, e.g. `os.path` project_root (str): Path to project root add_init (bool): Whether to add `__init__.py` to the path of modules that are packages. Defaults to True Returns: str
entailment
def check(self, feature): """Check that the feature's `input` is a str or Iterable[str]""" input = feature.input is_str = isa(str) is_nested_str = all_fn( iterable, lambda x: all(is_str, x)) assert is_str(input) or is_nested_str(input)
Check that the feature's `input` is a str or Iterable[str]
entailment
def check(self, feature): """Check that the feature has a fit/transform/fit_tranform interface""" assert hasattr(feature.transformer, 'fit') assert hasattr(feature.transformer, 'transform') assert hasattr(feature.transformer, 'fit_transform')
Check that the feature has a fit/transform/fit_tranform interface
entailment
def check(self, feature): """Check that fit can be called on reference data""" mapper = feature.as_dataframe_mapper() mapper.fit(self.X, y=self.y)
Check that fit can be called on reference data
entailment
def check(self, feature): """Check that fit_transform can be called on reference data""" mapper = feature.as_dataframe_mapper() mapper.fit_transform(self.X, y=self.y)
Check that fit_transform can be called on reference data
entailment
def check(self, feature): """Check that the dimensions of the transformed data are correct For input X, an n x p array, a n x q array should be produced, where q is the number of features produced by the logical feature. """ mapper = feature.as_dataframe_mapper() X = mapper.fit_transform(self.X, y=self.y) assert self.X.shape[0] == X.shape[0]
Check that the dimensions of the transformed data are correct For input X, an n x p array, a n x q array should be produced, where q is the number of features produced by the logical feature.
entailment
def check(self, feature): """Check that the feature can be pickled This is needed for saving the pipeline to disk """ try: buf = io.BytesIO() pickle.dump(feature, buf, protocol=pickle.HIGHEST_PROTOCOL) buf.seek(0) new_feature = pickle.load(buf) assert new_feature is not None assert isinstance(new_feature, Feature) finally: buf.close()
Check that the feature can be pickled This is needed for saving the pipeline to disk
entailment
def check(self, feature): """Check that the output of the transformer has no missing values""" mapper = feature.as_dataframe_mapper() X = mapper.fit_transform(self.X, y=self.y) assert not np.any(np.isnan(X))
Check that the output of the transformer has no missing values
entailment
def make_multi_lagger(lags, groupby_kwargs=None): """Return a union of transformers that apply different lags Args: lags (Collection[int]): collection of lags to apply groupby_kwargs (dict): keyword arguments to pd.DataFrame.groupby """ laggers = [SingleLagger(l, groupby_kwargs=groupby_kwargs) for l in lags] feature_union = FeatureUnion([ (repr(lagger), lagger) for lagger in laggers ]) return feature_union
Return a union of transformers that apply different lags Args: lags (Collection[int]): collection of lags to apply groupby_kwargs (dict): keyword arguments to pd.DataFrame.groupby
entailment
def start_new_feature(**cc_kwargs): """Start a new feature within a ballet project Renders the feature template into a temporary directory, then copies the feature files into the proper path within the contrib directory. Args: **cc_kwargs: options for the cookiecutter template Raises: ballet.exc.BalletError: the new feature has the same name as an existing one """ project = Project.from_path(pathlib.Path.cwd().resolve()) contrib_dir = project.get('contrib', 'module_path') with tempfile.TemporaryDirectory() as tempdir: # render feature template output_dir = tempdir cc_kwargs['output_dir'] = output_dir rendered_dir = render_feature_template(**cc_kwargs) # copy into contrib dir src = rendered_dir dst = contrib_dir synctree(src, dst, onexist=_fail_if_feature_exists) logger.info('Start new feature successful.')
Start a new feature within a ballet project Renders the feature template into a temporary directory, then copies the feature files into the proper path within the contrib directory. Args: **cc_kwargs: options for the cookiecutter template Raises: ballet.exc.BalletError: the new feature has the same name as an existing one
entailment
def get_proposed_feature(project): """Get the proposed feature The path of the proposed feature is determined by diffing the project against a comparison branch, such as master. The feature is then imported from that path and returned. Args: project (ballet.project.Project): project info Raises: ballet.exc.BalletError: more than one feature collected """ change_collector = ChangeCollector(project) collected_changes = change_collector.collect_changes() try: new_feature_info = one_or_raise(collected_changes.new_feature_info) importer, _, _ = new_feature_info except ValueError: raise BalletError('Too many features collected') module = importer() feature = _get_contrib_feature_from_module(module) return feature
Get the proposed feature The path of the proposed feature is determined by diffing the project against a comparison branch, such as master. The feature is then imported from that path and returned. Args: project (ballet.project.Project): project info Raises: ballet.exc.BalletError: more than one feature collected
entailment
def get_accepted_features(features, proposed_feature): """Deselect candidate features from list of all features Args: features (List[Feature]): collection of all features in the ballet project: both accepted features and candidate ones that have not been accepted proposed_feature (Feature): candidate feature that has not been accepted Returns: List[Feature]: list of features with the proposed feature not in it. Raises: ballet.exc.BalletError: Could not deselect exactly the proposed feature. """ def eq(feature): """Features are equal if they have the same source At least in this implementation... """ return feature.source == proposed_feature.source # deselect features that match the proposed feature result = lfilter(complement(eq), features) if len(features) - len(result) == 1: return result elif len(result) == len(features): raise BalletError( 'Did not find match for proposed feature within \'contrib\'') else: raise BalletError( 'Unexpected condition (n_features={}, n_result={})' .format(len(features), len(result)))
Deselect candidate features from list of all features Args: features (List[Feature]): collection of all features in the ballet project: both accepted features and candidate ones that have not been accepted proposed_feature (Feature): candidate feature that has not been accepted Returns: List[Feature]: list of features with the proposed feature not in it. Raises: ballet.exc.BalletError: Could not deselect exactly the proposed feature.
entailment
def collect_changes(self): """Collect file and feature changes Steps 1. Collects the files that have changed in this pull request as compared to a comparison branch. 2. Categorize these file changes into admissible or inadmissible file changes. Admissible file changes solely contribute python files to the contrib subdirectory. 3. Collect features from admissible new files. Returns: CollectedChanges """ file_diffs = self._collect_file_diffs() candidate_feature_diffs, valid_init_diffs, inadmissible_diffs = \ self._categorize_file_diffs(file_diffs) new_feature_info = self._collect_feature_info(candidate_feature_diffs) return CollectedChanges( file_diffs, candidate_feature_diffs, valid_init_diffs, inadmissible_diffs, new_feature_info)
Collect file and feature changes Steps 1. Collects the files that have changed in this pull request as compared to a comparison branch. 2. Categorize these file changes into admissible or inadmissible file changes. Admissible file changes solely contribute python files to the contrib subdirectory. 3. Collect features from admissible new files. Returns: CollectedChanges
entailment
def _categorize_file_diffs(self, file_diffs): """Partition file changes into admissible and inadmissible changes""" # TODO move this into a new validator candidate_feature_diffs = [] valid_init_diffs = [] inadmissible_files = [] for diff in file_diffs: valid, failures = check_from_class( ProjectStructureCheck, diff, self.project) if valid: if pathlib.Path(diff.b_path).parts[-1] != '__init__.py': candidate_feature_diffs.append(diff) logger.debug( 'Categorized {file} as CANDIDATE FEATURE MODULE' .format(file=diff.b_path)) else: valid_init_diffs.append(diff) logger.debug( 'Categorized {file} as VALID INIT MODULE' .format(file=diff.b_path)) else: inadmissible_files.append(diff) logger.debug( 'Categorized {file} as INADMISSIBLE; ' 'failures were {failures}' .format(file=diff.b_path, failures=failures)) logger.info( 'Admitted {} candidate feature{} ' 'and {} __init__ module{} ' 'and rejected {} file{}' .format(len(candidate_feature_diffs), make_plural_suffix(candidate_feature_diffs), len(valid_init_diffs), make_plural_suffix(valid_init_diffs), len(inadmissible_files), make_plural_suffix(inadmissible_files))) return candidate_feature_diffs, valid_init_diffs, inadmissible_files
Partition file changes into admissible and inadmissible changes
entailment
def _collect_feature_info(self, candidate_feature_diffs): """Collect feature info Args: candidate_feature_diffs (List[git.diff.Diff]): list of Diffs corresponding to admissible file changes compared to comparison ref Returns: List[Tuple]: list of tuple of importer, module name, and module path. The "importer" is a callable that returns a module """ project_root = self.project.path for diff in candidate_feature_diffs: path = diff.b_path modname = relpath_to_modname(path) modpath = project_root.joinpath(path) importer = partial(import_module_at_path, modname, modpath) yield importer, modname, modpath
Collect feature info Args: candidate_feature_diffs (List[git.diff.Diff]): list of Diffs corresponding to admissible file changes compared to comparison ref Returns: List[Tuple]: list of tuple of importer, module name, and module path. The "importer" is a callable that returns a module
entailment
def get_travis_branch(): """Get current branch per Travis environment variables If travis is building a PR, then TRAVIS_PULL_REQUEST is truthy and the name of the branch corresponding to the PR is stored in the TRAVIS_PULL_REQUEST_BRANCH environment variable. Else, the name of the branch is stored in the TRAVIS_BRANCH environment variable. See also: <https://docs.travis-ci.com/user/environment-variables/#default-environment-variables> """ # noqa E501 try: travis_pull_request = get_travis_env_or_fail('TRAVIS_PULL_REQUEST') if truthy(travis_pull_request): travis_pull_request_branch = get_travis_env_or_fail( 'TRAVIS_PULL_REQUEST_BRANCH') return travis_pull_request_branch else: travis_branch = get_travis_env_or_fail('TRAVIS_BRANCH') return travis_branch except UnexpectedTravisEnvironmentError: return None
Get current branch per Travis environment variables If travis is building a PR, then TRAVIS_PULL_REQUEST is truthy and the name of the branch corresponding to the PR is stored in the TRAVIS_PULL_REQUEST_BRANCH environment variable. Else, the name of the branch is stored in the TRAVIS_BRANCH environment variable. See also: <https://docs.travis-ci.com/user/environment-variables/#default-environment-variables>
entailment
def make_mapper(features): """Make a DataFrameMapper from a feature or list of features Args: features (Union[Feature, List[Feature]]): feature or list of features Returns: DataFrameMapper: mapper made from features """ if not features: features = Feature(input=[], transformer=NullTransformer()) if not iterable(features): features = (features, ) return DataFrameMapper( [t.as_input_transformer_tuple() for t in features], input_df=True)
Make a DataFrameMapper from a feature or list of features Args: features (Union[Feature, List[Feature]]): feature or list of features Returns: DataFrameMapper: mapper made from features
entailment
def _name_estimators(estimators): """Generate names for estimators. Adapted from sklearn.pipeline._name_estimators """ def get_name(estimator): if isinstance(estimator, DelegatingRobustTransformer): return get_name(estimator._transformer) return type(estimator).__name__.lower() names = list(map(get_name, estimators)) counter = dict(Counter(names)) counter = select_values(lambda x: x > 1, counter) for i in reversed(range(len(estimators))): name = names[i] if name in counter: names[i] += "-%d" % counter[name] counter[name] -= 1 return list(zip(names, estimators))
Generate names for estimators. Adapted from sklearn.pipeline._name_estimators
entailment
def _push(project): """Push default branch and project template branch to remote With default config (i.e. remote and branch names), equivalent to:: $ git push origin master:master project-template:project-template Raises: ballet.exc.BalletError: Push failed in some way """ repo = project.repo remote_name = project.get('project', 'remote') remote = repo.remote(remote_name) result = _call_remote_push(remote) failures = lfilter(complement(did_git_push_succeed), result) if failures: for push_info in failures: logger.error( 'Failed to push ref {from_ref} to {to_ref}' .format(from_ref=push_info.local_ref.name, to_ref=push_info.remote_ref.name)) raise BalletError('Push failed')
Push default branch and project template branch to remote With default config (i.e. remote and branch names), equivalent to:: $ git push origin master:master project-template:project-template Raises: ballet.exc.BalletError: Push failed in some way
entailment
def build(X_df=None, y_df=None): """Build features and target Args: X_df (DataFrame): raw variables y_df (DataFrame): raw target Returns: dict with keys X_df, features, mapper_X, X, y_df, encoder_y, y """ if X_df is None: X_df, _ = load_data() if y_df is None: _, y_df = load_data() features = get_contrib_features() mapper_X = ballet.feature.make_mapper(features) X = mapper_X.fit_transform(X_df) encoder_y = get_target_encoder() y = encoder_y.fit_transform(y_df) return { 'X_df': X_df, 'features': features, 'mapper_X': mapper_X, 'X': X, 'y_df': y_df, 'encoder_y': encoder_y, 'y': y, }
Build features and target Args: X_df (DataFrame): raw variables y_df (DataFrame): raw target Returns: dict with keys X_df, features, mapper_X, X, y_df, encoder_y, y
entailment
def main(input_dir, output_dir): """Engineer features""" import ballet.util.log ballet.util.log.enable(logger=logger, level='INFO', echo=False) ballet.util.log.enable(logger=ballet.util.log.logger, level='INFO', echo=False) X_df, y_df = load_data(input_dir=input_dir) out = build() mapper_X = out['mapper_X'] encoder_y = out['encoder_y'] X_ft = mapper_X.transform(X_df) y_ft = encoder_y.transform(y_df) save_features(X_ft, output_dir) save_targets(y_ft, output_dir)
Engineer features
entailment
def load_data(input_dir=None): """Load data""" if input_dir is not None: tables = conf.get('tables') entities_table_name = conf.get('data', 'entities_table_name') entities_config = some(where(tables, name=entities_table_name)) X = load_table_from_config(input_dir, entities_config) targets_table_name = conf.get('data', 'targets_table_name') targets_config = some(where(tables, name=targets_table_name)) y = load_table_from_config(input_dir, targets_config) else: raise NotImplementedError return X, y
Load data
entailment
def _write_header(self, epoch_data: EpochData) -> None: """ Write CSV header row with column names. Column names are inferred from the ``epoch_data`` and ``self.variables`` (if specified). Variables and streams expected later on are stored in ``self._variables`` and ``self._streams`` respectively. :param epoch_data: epoch data to be logged """ self._variables = self._variables or list(next(iter(epoch_data.values())).keys()) self._streams = epoch_data.keys() header = ['"epoch_id"'] for stream_name in self._streams: header += [stream_name + '_' + var for var in self._variables] with open(self._file_path, 'a') as file: file.write(self._delimiter.join(header) + '\n') self._header_written = True
Write CSV header row with column names. Column names are inferred from the ``epoch_data`` and ``self.variables`` (if specified). Variables and streams expected later on are stored in ``self._variables`` and ``self._streams`` respectively. :param epoch_data: epoch data to be logged
entailment
def _write_row(self, epoch_id: int, epoch_data: EpochData) -> None: """ Write a single epoch result row to the CSV file. :param epoch_id: epoch number (will be written at the first column) :param epoch_data: epoch data :raise KeyError: if the variable is missing and ``self._on_missing_variable`` is set to ``error`` :raise TypeError: if the variable has wrong type and ``self._on_unknown_type`` is set to ``error`` """ # list of values to be written values = [epoch_id] for stream_name in self._streams: for variable_name in self._variables: column_name = stream_name+'_'+variable_name try: value = epoch_data[stream_name][variable_name] except KeyError as ex: err_message = '`{}` not found in epoch data.'.format(column_name) if self._on_missing_variable == 'error': raise KeyError(err_message) from ex elif self._on_missing_variable == 'warn': logging.warning(err_message) values.append(self._default_value) continue if isinstance(value, dict) and 'mean' in value: value = value['mean'] elif isinstance(value, dict) and 'nanmean' in value: value = value['nanmean'] if np.isscalar(value): values.append(value) else: err_message = 'Variable `{}` value is not scalar.'.format(variable_name) if self._on_unknown_type == 'error': raise TypeError(err_message) elif self._on_unknown_type == 'warn': logging.warning(err_message) values.append(self._default_value) # write the row with open(self._file_path, 'a') as file: row = self._delimiter.join([str(value) for value in values]) file.write(row + '\n')
Write a single epoch result row to the CSV file. :param epoch_id: epoch number (will be written at the first column) :param epoch_data: epoch data :raise KeyError: if the variable is missing and ``self._on_missing_variable`` is set to ``error`` :raise TypeError: if the variable has wrong type and ``self._on_unknown_type`` is set to ``error``
entailment
def after_epoch(self, epoch_id: int, epoch_data: EpochData) -> None: """ Write a new row to the CSV file with the given epoch data. In the case of first invocation, create the CSV header. :param epoch_id: number of the epoch :param epoch_data: epoch data to be logged """ logging.debug('Saving epoch %d data to "%s"', epoch_id, self._file_path) if not self._header_written: self._write_header(epoch_data=epoch_data) self._write_row(epoch_id=epoch_id, epoch_data=epoch_data)
Write a new row to the CSV file with the given epoch data. In the case of first invocation, create the CSV header. :param epoch_id: number of the epoch :param epoch_data: epoch data to be logged
entailment
def get_random_name(sep: str='-'): """ Generate random docker-like name with the given separator. :param sep: adjective-name separator string :return: random docker-like name """ r = random.SystemRandom() return '{}{}{}'.format(r.choice(_left), sep, r.choice(_right))
Generate random docker-like name with the given separator. :param sep: adjective-name separator string :return: random docker-like name
entailment
def _check_train_time(self) -> None: """ Stop the training if the training time exceeded ``self._minutes``. :raise TrainingTerminated: if the training time exceeded ``self._minutes`` """ if self._minutes is not None and (datetime.now() - self._training_start).total_seconds()/60 > self._minutes: raise TrainingTerminated('Training terminated after more than {} minutes'.format(self._minutes))
Stop the training if the training time exceeded ``self._minutes``. :raise TrainingTerminated: if the training time exceeded ``self._minutes``
entailment
def after_batch(self, stream_name: str, batch_data: Batch) -> None: """ If ``stream_name`` equals to :py:attr:`cxflow.constants.TRAIN_STREAM`, increase the iterations counter and possibly stop the training; additionally, call :py:meth:`_check_train_time`. :param stream_name: stream name :param batch_data: ignored :raise TrainingTerminated: if the number of iterations reaches ``self._iters`` """ self._check_train_time() if self._iters is not None and stream_name == self._train_stream_name: self._iters_done += 1 if self._iters_done >= self._iters: raise TrainingTerminated('Training terminated after iteration {}'.format(self._iters_done))
If ``stream_name`` equals to :py:attr:`cxflow.constants.TRAIN_STREAM`, increase the iterations counter and possibly stop the training; additionally, call :py:meth:`_check_train_time`. :param stream_name: stream name :param batch_data: ignored :raise TrainingTerminated: if the number of iterations reaches ``self._iters``
entailment
def after_epoch(self, epoch_id: int, epoch_data: EpochData) -> None: """ Stop the training if the ``epoch_id`` reaches ``self._epochs``; additionally, call :py:meth:`_check_train_time`. :param epoch_id: epoch id :param epoch_data: ignored :raise TrainingTerminated: if the ``epoch_id`` reaches ``self._epochs`` """ self._check_train_time() if self._epochs is not None and epoch_id >= self._epochs: logging.info('EpochStopperHook triggered') raise TrainingTerminated('Training terminated after epoch {}'.format(epoch_id))
Stop the training if the ``epoch_id`` reaches ``self._epochs``; additionally, call :py:meth:`_check_train_time`. :param epoch_id: epoch id :param epoch_data: ignored :raise TrainingTerminated: if the ``epoch_id`` reaches ``self._epochs``
entailment
def sanitize_url(url: str) -> str: """ Sanitize the given url so that it can be used as a valid filename. :param url: url to create filename from :raise ValueError: when the given url can not be sanitized :return: created filename """ for part in reversed(url.split('/')): filename = re.sub(r'[^a-zA-Z0-9_.\-]', '', part) if len(filename) > 0: break else: raise ValueError('Could not create reasonable name for file from url %s', url) return filename
Sanitize the given url so that it can be used as a valid filename. :param url: url to create filename from :raise ValueError: when the given url can not be sanitized :return: created filename
entailment
def maybe_download_and_extract(data_root: str, url: str) -> None: """ Maybe download the specified file to ``data_root`` and try to unpack it with ``shutil.unpack_archive``. :param data_root: data root to download the files to :param url: url to download from """ # make sure data_root exists os.makedirs(data_root, exist_ok=True) # create sanitized filename from url filename = sanitize_url(url) # check whether the archive already exists filepath = os.path.join(data_root, filename) if os.path.exists(filepath): logging.info('\t`%s` already exists; skipping', filepath) return # download with progressbar try: logging.info('\tdownloading %s', filepath) req = requests.get(url, stream=True) req.raise_for_status() except requests.exceptions.RequestException as ex: logging.error('File `%s` could not be downloaded, %s', filepath, ex) return expected_size = int(req.headers.get('content-length')) chunk_size = 1024 with open(filepath, 'wb') as f_out,\ click.progressbar(req.iter_content(chunk_size=chunk_size), length=expected_size/chunk_size) as bar: for chunk in bar: if chunk: f_out.write(chunk) f_out.flush() # extract try: shutil.unpack_archive(filepath, data_root) except (shutil.ReadError, ValueError): logging.info('File `%s` could not be extracted by `shutil.unpack_archive`. Please process it manually.', filepath)
Maybe download the specified file to ``data_root`` and try to unpack it with ``shutil.unpack_archive``. :param data_root: data root to download the files to :param url: url to download from
entailment
def _raise_check_aggregation(aggregation: str): """ Check whether the given aggregation is present in NumPy or it is one of EXTRA_AGGREGATIONS. :param aggregation: the aggregation name :raise ValueError: if the specified aggregation is not supported or found in NumPy """ if aggregation not in ComputeStats.EXTRA_AGGREGATIONS and not hasattr(np, aggregation): raise ValueError('Aggregation `{}` is not a NumPy function or a member ' 'of EXTRA_AGGREGATIONS.'.format(aggregation))
Check whether the given aggregation is present in NumPy or it is one of EXTRA_AGGREGATIONS. :param aggregation: the aggregation name :raise ValueError: if the specified aggregation is not supported or found in NumPy
entailment
def _compute_aggregation(aggregation: str, data: Iterable[Any]): """ Compute the specified aggregation on the given data. :param aggregation: the name of an arbitrary NumPy function (e.g., mean, max, median, nanmean, ...) or one of :py:attr:`EXTRA_AGGREGATIONS`. :param data: data to be aggregated :raise ValueError: if the specified aggregation is not supported or found in NumPy """ ComputeStats._raise_check_aggregation(aggregation) if aggregation == 'nanfraction': return np.sum(np.isnan(data)) / len(data) if aggregation == 'nancount': return int(np.sum(np.isnan(data))) return getattr(np, aggregation)(data)
Compute the specified aggregation on the given data. :param aggregation: the name of an arbitrary NumPy function (e.g., mean, max, median, nanmean, ...) or one of :py:attr:`EXTRA_AGGREGATIONS`. :param data: data to be aggregated :raise ValueError: if the specified aggregation is not supported or found in NumPy
entailment
def _save_stats(self, epoch_data: EpochData) -> None: """ Extend ``epoch_data`` by stream:variable:aggreagation data. :param epoch_data: data source from which the statistics are computed """ for stream_name in epoch_data.keys(): for variable, aggregations in self._variable_aggregations.items(): # variables are already checked in the AccumulatingHook; hence, we do not check them here epoch_data[stream_name][variable] = OrderedDict( {aggr: ComputeStats._compute_aggregation(aggr, self._accumulator[stream_name][variable]) for aggr in aggregations})
Extend ``epoch_data`` by stream:variable:aggreagation data. :param epoch_data: data source from which the statistics are computed
entailment
def after_epoch(self, epoch_data: EpochData, **kwargs) -> None: """ Compute the specified aggregations and save them to the given epoch data. :param epoch_data: epoch data to be processed """ self._save_stats(epoch_data) super().after_epoch(epoch_data=epoch_data, **kwargs)
Compute the specified aggregations and save them to the given epoch data. :param epoch_data: epoch data to be processed
entailment
def save(self) -> None: """ Save the training trace to :py:attr:`CXF_TRACE_FILE` file under the specified directory. :raise ValueError: if no output directory was specified """ if self._output_dir is None: raise ValueError('Can not save TrainingTrace without output dir.') yaml_to_file(self._trace, self._output_dir, CXF_TRACE_FILE)
Save the training trace to :py:attr:`CXF_TRACE_FILE` file under the specified directory. :raise ValueError: if no output directory was specified
entailment
def from_file(filepath: str): """ Load training trace from the given ``filepath``. :param filepath: training trace file path :return: training trace """ trace = TrainingTrace() trace._trace = load_config(filepath) return trace
Load training trace from the given ``filepath``. :param filepath: training trace file path :return: training trace
entailment
def after_epoch(self, epoch_id: int, epoch_data: EpochData): """ Check termination conditions. :param epoch_id: number of the processed epoch :param epoch_data: epoch data to be checked :raise KeyError: if the stream of variable was not found in ``epoch_data`` :raise TypeError: if the monitored variable is not a scalar or scalar ``mean`` aggregation :raise ValueError: if the specified number of epochs exceeded :raise TrainingTerminated: if the monitor variable is above the required level """ if self._stream not in epoch_data: raise KeyError('The hook could not determine whether the threshold was exceeded as the stream `{}`' 'was not found in the epoch data'.format(self._stream)) if self._variable not in epoch_data[self._stream]: raise KeyError('The hook could not determine whether the threshold was exceeded as the variable `{}`' 'was not found in the epoch data stream `{}`'.format(self._variable, self._stream)) value = epoch_data[self._stream][self._variable] if isinstance(value, dict) and 'mean' in value: value = value['mean'] if not np.isscalar(value): raise TypeError('The value to be checked has to be either a scalar or a dict with `mean` key. ' 'Got `{}` instead.'.format(type(value).__name__)) if value > self._required_min_value: raise TrainingTerminated('{} {} level matched (current {} is greater than required {}).' .format(self._stream, self._variable, value, self._required_min_value)) elif epoch_id >= self._max_epoch: raise ValueError('{} {} was only {} in epoch {}, but {} was required. Training failed.' .format(self._stream, self._variable, value, epoch_id, self._required_min_value))
Check termination conditions. :param epoch_id: number of the processed epoch :param epoch_data: epoch data to be checked :raise KeyError: if the stream of variable was not found in ``epoch_data`` :raise TypeError: if the monitored variable is not a scalar or scalar ``mean`` aggregation :raise ValueError: if the specified number of epochs exceeded :raise TrainingTerminated: if the monitor variable is above the required level
entailment
def train(config_path: str, cl_arguments: Iterable[str], output_root: str) -> None: """ Load config and start the training. :param config_path: path to configuration file :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created """ config = None try: config_path = find_config(config_path) config = load_config(config_file=config_path, additional_args=cl_arguments) validate_config(config) logging.debug('\tLoaded config: %s', config) except Exception as ex: # pylint: disable=broad-except fallback('Loading config failed', ex) run(config=config, output_root=output_root)
Load config and start the training. :param config_path: path to configuration file :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created
entailment
def evaluate(model_path: str, stream_name: str, config_path: Optional[str], cl_arguments: Iterable[str], output_root: str) -> None: """ Evaluate the given model on the specified data stream. Configuration is updated by the respective predict.stream_name section, in particular: - hooks section is entirely replaced - model and dataset sections are updated :param model_path: path to the model to be evaluated :param stream_name: data stream name to be evaluated :param config_path: path to the config to be used, if not specified infer the path from ``model_path`` :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created """ config = None try: model_dir = path.dirname(model_path) if not path.isdir(model_path) else model_path config_path = find_config(model_dir if config_path is None else config_path) config = load_config(config_file=config_path, additional_args=cl_arguments) if stream_name == CXF_PREDICT_STREAM and stream_name in config: # old style ``cxflow predict ...`` logging.warning('Old style ``predict`` configuration section is deprecated and will not be supported, ' 'use ``eval.predict`` section instead.') config['eval'] = {'predict': config['predict']} if 'eval' in config and stream_name in config['eval']: update_section = config['eval'][stream_name] for subsection in ['dataset', 'model', 'main_loop']: if subsection in update_section: config[subsection].update(update_section[subsection]) if 'hooks' in update_section: config['hooks'] = update_section['hooks'] else: logging.warning('Config does not contain `eval.%s.hooks` section. ' 'No hook will be employed during the evaluation.', stream_name) config['hooks'] = [] validate_config(config) logging.debug('\tLoaded config: %s', config) except Exception as ex: # pylint: disable=broad-except fallback('Loading config failed', ex) run(config=config, output_root=output_root, restore_from=model_path, eval=stream_name)
Evaluate the given model on the specified data stream. Configuration is updated by the respective predict.stream_name section, in particular: - hooks section is entirely replaced - model and dataset sections are updated :param model_path: path to the model to be evaluated :param stream_name: data stream name to be evaluated :param config_path: path to the config to be used, if not specified infer the path from ``model_path`` :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created
entailment
def predict(config_path: str, restore_from: Optional[str], cl_arguments: Iterable[str], output_root: str) -> None: """ Run prediction from the specified config path. If the config contains a ``predict`` section: - override hooks with ``predict.hooks`` if present - update dataset, model and main loop sections if the respective sections are present :param config_path: path to the config file or the directory in which it is stored :param restore_from: backend-specific path to the already trained model to be restored from. If ``None`` is passed, it is inferred from the configuration file location as the directory it is located in. :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created """ config = None try: config_path = find_config(config_path) restore_from = restore_from or path.dirname(config_path) config = load_config(config_file=config_path, additional_args=cl_arguments) if 'predict' in config: for section in ['dataset', 'model', 'main_loop']: if section in config['predict']: config[section].update(config['predict'][section]) if 'hooks' in config['predict']: config['hooks'] = config['predict']['hooks'] else: logging.warning('Config does not contain `predict.hooks` section. ' 'No hook will be employed during the prediction.') config['hooks'] = [] validate_config(config) logging.debug('\tLoaded config: %s', config) except Exception as ex: # pylint: disable=broad-except fallback('Loading config failed', ex) run(config=config, output_root=output_root, restore_from=restore_from, eval='predict')
Run prediction from the specified config path. If the config contains a ``predict`` section: - override hooks with ``predict.hooks`` if present - update dataset, model and main loop sections if the respective sections are present :param config_path: path to the config file or the directory in which it is stored :param restore_from: backend-specific path to the already trained model to be restored from. If ``None`` is passed, it is inferred from the configuration file location as the directory it is located in. :param cl_arguments: additional command line arguments which will update the configuration :param output_root: output root in which the training directory will be created
entailment
def _create_epoch_data(self, streams: Optional[Iterable[str]]=None) -> EpochData: """Create empty epoch data double dict.""" if streams is None: streams = [self._train_stream_name] + self._extra_streams return OrderedDict([(stream_name, OrderedDict()) for stream_name in streams])
Create empty epoch data double dict.
entailment
def _check_sources(self, batch: Dict[str, object]) -> None: """ Check for unused and missing sources. :param batch: batch to be checked :raise ValueError: if a source is missing or unused and ``self._on_unused_sources`` is set to ``error`` """ unused_sources = [source for source in batch.keys() if source not in self._model.input_names] missing_sources = [source for source in self._model.input_names if source not in batch.keys()] # check stream sources if unused_sources: if self._on_unused_sources == 'warn' and not self._extra_sources_warned: logging.warning('Some sources provided by the stream do not match model placeholders. Set ' '`main_loop.on_unused_sources` to `ignore` in order to suppress this warning. ' 'Extra sources: %s', unused_sources) self._extra_sources_warned = True elif self._on_unused_sources == 'error': raise ValueError('Some sources provided by the stream do not match model placeholders. Set' '`main_loop.on_unused_sources` to `warn` in order to suppress this error.\n' 'Extra sources: {}'.format(unused_sources)) if missing_sources: raise ValueError('Stream does not provide all required sources. Missing sources: {}' .format(missing_sources))
Check for unused and missing sources. :param batch: batch to be checked :raise ValueError: if a source is missing or unused and ``self._on_unused_sources`` is set to ``error``
entailment
def _run_epoch(self, stream: StreamWrapper, train: bool) -> None: """ Iterate through the given stream and evaluate/train the model with the received batches. Calls :py:meth:`cxflow.hooks.AbstractHook.after_batch` events. :param stream: stream to iterate :param train: if set to ``True``, the model will be trained :raise ValueError: in case of empty batch when ``on_empty_batch`` is set to ``error`` :raise ValueError: in case of empty stream when ``on_empty_stream`` is set to ``error`` :raise ValueError: in case of two batch variables having different lengths """ nonempty_batch_count = 0 for i, batch_input in enumerate(stream): self.raise_check_interrupt() batch_sizes = {len(source) for source in batch_input.values()} if len(batch_sizes) == 0 or batch_sizes == {0}: if self._on_empty_batch == 'warn': logging.warning('%i-th batch in stream `%s` appears to be empty (%i-th empty batch in total). Set ' '`main_loop.on_empty_batch` to `ignore` in order to suppress this warning.', i, stream.name, nonempty_batch_count) elif self._on_empty_batch == 'error': raise ValueError('{}-th batch in stream `{}` appears to be empty ({}-th empty batch in total). Set ' '`main_loop.on_empty_batch` to `warn` in order to change this error into warning; ' 'set to `ignore` to remove it.'.format(i, stream.name, nonempty_batch_count)) continue elif self._fixed_batch_size: if batch_sizes != {self._fixed_batch_size}: var, len_ = [(k, len(v)) for k, v in batch_input.items() if len(v) != self._fixed_batch_size][0] logging.debug('%i-th batch in stream `%s` has variable `%s` of length %i inconsistent with ' '`main_loop.fixed_size` = %i', i, stream.name, var, len_, self._fixed_batch_size) continue nonempty_batch_count += 1 self._check_sources(batch_input) with Timer('eval_batch_{}'.format(stream.name), self._epoch_profile): batch_output = self._model.run(batch=batch_input, train=train, stream=stream) assert set(batch_input.keys()).isdisjoint(set(batch_output)), 'Batch inputs and outputs must not overlap.' with Timer('after_batch_hooks_{}'.format(stream.name), self._epoch_profile): batch_data = {**batch_input, **batch_output} for hook in self._hooks: hook.after_batch(stream_name=stream.name, batch_data=batch_data) if nonempty_batch_count == 0: if self._on_empty_stream == 'warn': logging.warning('Stream `%s` appears to be empty. Set `main_loop.on_empty_stream` to `ignore` in order ' 'to suppress this warning.', stream.name) elif self._on_empty_stream == 'error': raise ValueError('Stream `{}` appears to be empty. Set ' '`main_loop.on_empty_stream` to `warn` in order to change this error into warning; ' 'set to `ignore` to remove it.'.format(stream.name))
Iterate through the given stream and evaluate/train the model with the received batches. Calls :py:meth:`cxflow.hooks.AbstractHook.after_batch` events. :param stream: stream to iterate :param train: if set to ``True``, the model will be trained :raise ValueError: in case of empty batch when ``on_empty_batch`` is set to ``error`` :raise ValueError: in case of empty stream when ``on_empty_stream`` is set to ``error`` :raise ValueError: in case of two batch variables having different lengths
entailment
def train_by_stream(self, stream: StreamWrapper) -> None: """ Train the model with the given stream. :param stream: stream to train with """ self._run_epoch(stream=stream, train=True)
Train the model with the given stream. :param stream: stream to train with
entailment
def evaluate_stream(self, stream: StreamWrapper) -> None: """ Evaluate the given stream. :param stream: stream to be evaluated :param stream_name: stream name """ self._run_epoch(stream=stream, train=False)
Evaluate the given stream. :param stream: stream to be evaluated :param stream_name: stream name
entailment
def get_stream(self, stream_name: str) -> StreamWrapper: """ Get a :py:class:`StreamWrapper` with the given name. :param stream_name: stream name :return: dataset function name providing the respective stream :raise AttributeError: if the dataset does not provide the function creating the stream """ if stream_name not in self._streams: stream_fn_name = '{}_stream'.format(stream_name) try: stream_fn = getattr(self._dataset, stream_fn_name) stream_epoch_limit = -1 if self._fixed_epoch_size is not None and stream_name == self._train_stream_name: stream_epoch_limit = self._fixed_epoch_size self._streams[stream_name] = StreamWrapper(stream_fn, buffer_size=self._buffer, epoch_size=stream_epoch_limit, name=stream_name, profile=self._epoch_profile) except AttributeError as ex: raise AttributeError('The dataset does not have a function for creating a stream named `{}`. ' 'The function has to be named `{}`.'.format(stream_name, stream_fn_name)) from ex return self._streams[stream_name]
Get a :py:class:`StreamWrapper` with the given name. :param stream_name: stream name :return: dataset function name providing the respective stream :raise AttributeError: if the dataset does not provide the function creating the stream
entailment
def _run_zeroth_epoch(self, streams: Iterable[str]) -> None: """ Run zeroth epoch on the specified streams. Calls - :py:meth:`cxflow.hooks.AbstractHook.after_epoch` :param streams: stream names to be evaluated """ for stream_name in streams: with self.get_stream(stream_name) as stream: self.evaluate_stream(stream) epoch_data = self._create_epoch_data(streams) for hook in self._hooks: hook.after_epoch(epoch_id=0, epoch_data=epoch_data)
Run zeroth epoch on the specified streams. Calls - :py:meth:`cxflow.hooks.AbstractHook.after_epoch` :param streams: stream names to be evaluated
entailment
def _try_run(self, run_func: Callable[[], None]) -> None: """ Try running the given function (training/prediction). Calls - :py:meth:`cxflow.hooks.AbstractHook.before_training` - :py:meth:`cxflow.hooks.AbstractHook.after_training` :param run_func: function to be run """ # Initialization: before_training for hook in self._hooks: hook.before_training() try: run_func() except TrainingTerminated as ex: logging.info('Training terminated: %s', ex) # After training: after_training for hook in self._hooks: hook.after_training()
Try running the given function (training/prediction). Calls - :py:meth:`cxflow.hooks.AbstractHook.before_training` - :py:meth:`cxflow.hooks.AbstractHook.after_training` :param run_func: function to be run
entailment
def run_training(self, trace: Optional[TrainingTrace]=None) -> None: """ Run the main loop in the training mode. Calls - :py:meth:`cxflow.hooks.AbstractHook.after_epoch` - :py:meth:`cxflow.hooks.AbstractHook.after_epoch_profile` """ for stream_name in [self._train_stream_name] + self._extra_streams: self.get_stream(stream_name) def training(): logging.debug('Training started') self._epochs_done = 0 # Zeroth epoch: after_epoch if not self._skip_zeroth_epoch: logging.info('Evaluating 0th epoch') self._run_zeroth_epoch([self._train_stream_name] + self._extra_streams) logging.info('0th epoch done\n\n') # Training loop: after_epoch, after_epoch_profile while True: epoch_id = self._epochs_done + 1 logging.info('Training epoch %s', epoch_id) self._epoch_profile.clear() epoch_data = self._create_epoch_data() with self.get_stream(self._train_stream_name) as stream: self.train_by_stream(stream) for stream_name in self._extra_streams: with self.get_stream(stream_name) as stream: self.evaluate_stream(stream) with Timer('after_epoch_hooks', self._epoch_profile): for hook in self._hooks: hook.after_epoch(epoch_id=epoch_id, epoch_data=epoch_data) for hook in self._hooks: hook.after_epoch_profile(epoch_id=epoch_id, profile=self._epoch_profile, train_stream_name=self._train_stream_name, extra_streams=self._extra_streams) self._epochs_done = epoch_id if trace is not None: trace[TrainingTraceKeys.EPOCHS_DONE] = self._epochs_done logging.info('Epoch %s done\n\n', epoch_id) self._try_run(training)
Run the main loop in the training mode. Calls - :py:meth:`cxflow.hooks.AbstractHook.after_epoch` - :py:meth:`cxflow.hooks.AbstractHook.after_epoch_profile`
entailment