repo
stringlengths
2
99
file
stringlengths
13
225
code
stringlengths
0
18.3M
file_length
int64
0
18.3M
avg_line_length
float64
0
1.36M
max_line_length
int64
0
4.26M
extension_type
stringclasses
1 value
smt
smt-master/smt/surrogate_models/krg.py
""" Author: Dr. Mohamed A. Bouhlel <mbouhlel@umich.edu> This package is distributed under New BSD license. """ from smt.surrogate_models.krg_based import KrgBased from smt.utils.kriging import componentwise_distance class KRG(KrgBased): name = "Kriging" def _initialize(self): super()._initialize() declare = self.options.declare declare( "corr", "squar_exp", values=("pow_exp", "abs_exp", "squar_exp", "matern52", "matern32"), desc="Correlation function type", types=(str), ) def _componentwise_distance(self, dx, opt=0, theta=None, return_derivative=False): d = componentwise_distance( dx, self.options["corr"], self.nx, self.options["pow_exp_power"], theta=theta, return_derivative=return_derivative, ) return d
921
25.342857
86
py
smt
smt-master/smt/surrogate_models/kplsk.py
""" Author: Dr. Mohamed A. Bouhlel <mbouhlel@umich.edu> This package is distributed under New BSD license. """ from smt.surrogate_models import KPLS from smt.utils.kriging import componentwise_distance_PLS, componentwise_distance class KPLSK(KPLS): name = "KPLSK" def _initialize(self): super()._initialize() declare = self.options.declare # KPLSK used only with "squar_exp" correlations declare( "corr", "squar_exp", values=("squar_exp"), desc="Correlation function type", types=(str), ) def _componentwise_distance(self, dx, opt=0, theta=None, return_derivative=False): if opt == 0: # Kriging step d = componentwise_distance( dx, self.options["corr"], self.nx, power=self.options["pow_exp_power"], theta=theta, return_derivative=return_derivative, ) else: # KPLS step d = componentwise_distance_PLS( dx, self.options["corr"], self.options["n_comp"], self.coeff_pls, power=self.options["pow_exp_power"], theta=theta, return_derivative=return_derivative, ) return d
1,388
27.346939
86
py
smt
smt-master/smt/surrogate_models/rmtb.py
""" Author: Dr. John T. Hwang <hwangjt@umich.edu> This package is distributed under New BSD license. """ import numpy as np import scipy.sparse from numbers import Integral from smt.utils.linear_solvers import get_solver from smt.utils.line_search import get_line_search_class from smt.surrogate_models.rmts import RMTS from smt.surrogate_models.rmtsclib import PyRMTB class RMTB(RMTS): """ Regularized Minimal-energy Tensor-product B-spline (RMTB) interpolant. RMTB builds an approximation from a tensor product of B-spline curves. In 1-D it is a B-spline curve, in 2-D it is a B-spline surface, in 3-D it is a B-spline volume, and so on - it works for any arbitrary number of dimensions. However, the training points should preferably be arranged in a structured fashion. Advantages: - Evaluation time is independent of the number of training points - The smoothness can be tuned by adjusting the B-spline order and the number of B-spline control points (the latter also affects performance) Disadvantages: - Training time scales poorly with the # dimensions - The data should be structured - RMTB does not handle track data well - RMTB approximates, not interpolates - it does not pass through the training points """ name = "RMTB" def _initialize(self): super(RMTB, self)._initialize() declare = self.options.declare declare( "order", 3, types=(Integral, tuple, list, np.ndarray), desc="B-spline order in each dimension - length [nx]", ) declare( "num_ctrl_pts", 15, types=(Integral, tuple, list, np.ndarray), desc="# B-spline control points in each dimension - length [nx]", ) def _setup(self): options = self.options nx = self.training_points[None][0][0].shape[1] for name in ["smoothness", "num_ctrl_pts", "order"]: if isinstance(options[name], (int, float)): options[name] = [options[name]] * nx options[name] = np.atleast_1d(options[name]) self.printer.max_print_depth = options["max_print_depth"] num = {} # number of inputs and outputs num["x"] = self.training_points[None][0][0].shape[1] num["y"] = self.training_points[None][0][1].shape[1] num["order_list"] = np.array(options["order"], int) num["order"] = np.prod(num["order_list"]) num["ctrl_list"] = np.array(options["num_ctrl_pts"], int) num["ctrl"] = np.prod(num["ctrl_list"]) num["elem_list"] = np.array(num["ctrl_list"] - num["order_list"] + 1, int) num["elem"] = np.prod(num["elem_list"]) num["knots_list"] = num["order_list"] + num["ctrl_list"] num["knots"] = np.sum(num["knots_list"]) # total number of training points (function values and derivatives) num["t"] = 0 for kx in self.training_points[None]: num["t"] += self.training_points[None][kx][0].shape[0] # for RMT num["coeff"] = num["ctrl"] num["support"] = num["order"] num["dof"] = num["ctrl"] self.num = num self.rmtsc = PyRMTB() self.rmtsc.setup( num["x"], np.array(self.options["xlimits"][:, 0]), np.array(self.options["xlimits"][:, 1]), np.array(num["order_list"], np.int32), np.array(num["ctrl_list"], np.int32), ) def _compute_jac_raw(self, ix1, ix2, x): xlimits = self.options["xlimits"] t = np.zeros(x.shape) for kx in range(self.num["x"]): t[:, kx] = (x[:, kx] - xlimits[kx, 0]) / (xlimits[kx, 1] - xlimits[kx, 0]) t = np.maximum(t, 0.0 + 1e-15) t = np.minimum(t, 1.0 - 1e-15) n = x.shape[0] nnz = n * self.num["order"] # data, rows, cols = RMTBlib.compute_jac(ix1, ix2, self.num['x'], n, nnz, # self.num['order_list'], self.num['ctrl_list'], t) data = np.empty(nnz) rows = np.empty(nnz, dtype=np.int32) cols = np.empty(nnz, dtype=np.int32) self.rmtsc.compute_jac(ix1 - 1, ix2 - 1, n, t.flatten(), data, rows, cols) if ix1 != 0: data /= xlimits[ix1 - 1, 1] - xlimits[ix1 - 1, 0] if ix2 != 0: data /= xlimits[ix2 - 1, 1] - xlimits[ix2 - 1, 0] return data, rows, cols def _compute_dof2coeff(self): return None
4,507
33.945736
86
py
smt
smt-master/smt/surrogate_models/surrogate_model.py
""" Author: Dr. Mohamed A. Bouhlel <mbouhlel@umich.edu> Dr. John T. Hwang <hwangjt@umich.edu> This package is distributed under New BSD license. Paul Saves : Mixed Integer """ from typing import Optional import numpy as np from collections import defaultdict from abc import ABCMeta, abstractmethod from smt.utils.printer import Printer from smt.utils.options_dictionary import OptionsDictionary from smt.utils.checks import check_support, check_nx, ensure_2d_array class SurrogateModel(metaclass=ABCMeta): """ Base class for all surrogate models. Attributes ---------- options : OptionsDictionary Dictionary of options. Options values can be set on this attribute directly or they can be passed in as keyword arguments during instantiation. supports : dict Dictionary containing information about what this surrogate model supports. Examples -------- >>> from smt.surrogate_models import RBF >>> sm = RBF(print_training=False) >>> sm.options['print_prediction'] = False """ def __init__(self, **kwargs): """ Constructor where values of options can be passed in. For the list of options, see the documentation for the surrogate model being used. Parameters ---------- **kwargs : named arguments Set of options that can be optionally set; each option must have been declared. Examples -------- >>> from smt.surrogate_models import RBF >>> sm = RBF(print_global=False) """ self.options = OptionsDictionary() self.supports = supports = {} supports["training_derivatives"] = False supports["derivatives"] = False supports["output_derivatives"] = False supports["adjoint_api"] = False supports["variances"] = False supports["variance_derivatives"] = False supports["x_hierarchy"] = False declare = self.options.declare declare( "print_global", True, types=bool, desc="Global print toggle. If False, all printing is suppressed", ) declare( "print_training", True, types=bool, desc="Whether to print training information", ) declare( "print_prediction", True, types=bool, desc="Whether to print prediction information", ) declare( "print_problem", True, types=bool, desc="Whether to print problem information", ) declare( "print_solver", True, types=bool, desc="Whether to print solver information" ) self._initialize() self.options.update(kwargs) self.training_points = defaultdict(dict) self.printer = Printer() self._final_initialize() @property @abstractmethod def name(self): pass def set_training_values(self, xt: np.ndarray, yt: np.ndarray, name=None) -> None: """ Set training data (values). Parameters ---------- xt : np.ndarray[nt, nx] or np.ndarray[nt] The input values for the nt training points. yt : np.ndarray[nt, ny] or np.ndarray[nt] The output values for the nt training points. name : str or None An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). """ xt = ensure_2d_array(xt, "xt") yt = ensure_2d_array(yt, "yt") if xt.shape[0] != yt.shape[0]: raise ValueError( "the first dimension of xt and yt must have the same length" ) self.nt = xt.shape[0] self.nx = xt.shape[1] self.ny = yt.shape[1] kx = 0 self.training_points[name][kx] = [np.array(xt), np.array(yt)] def update_training_values( self, yt: np.ndarray, name: Optional[str] = None ) -> None: """ Update the training data (values) at the previously set input values. Parameters ---------- yt : np.ndarray[nt, ny] or np.ndarray[nt] The output values for the nt training points. name : str or None, optional An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). The default is None. Raises ------ ValueError The training points must be set first with set_training_values before calling update_training_values. The number of training points does not agree with the earlier call of set_training_values. """ yt = ensure_2d_array(yt, "yt") kx = 0 if kx not in self.training_points[name]: raise ValueError( "The training points must be set first with set_training_values " + "before calling update_training_values." ) xt = self.training_points[name][kx][0] if xt.shape[0] != yt.shape[0]: raise ValueError( "The number of training points does not agree with the earlier call of " + "set_training_values." ) self.training_points[name][kx][1] = np.array(yt) def set_training_derivatives( self, xt: np.ndarray, dyt_dxt: np.ndarray, kx: int, name: Optional[str] = None ) -> None: """ Set training data (derivatives). Parameters ---------- xt : np.ndarray[nt, nx] or np.ndarray[nt] The input values for the nt training points. dyt_dxt : np.ndarray[nt, ny] or np.ndarray[nt] The derivatives values for the nt training points. kx : int 0-based index of the derivatives being set. name : str or None An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). """ check_support(self, "training_derivatives") xt = ensure_2d_array(xt, "xt") dyt_dxt = ensure_2d_array(dyt_dxt, "dyt_dxt") if xt.shape[0] != dyt_dxt.shape[0]: raise ValueError( "the first dimension of xt and dyt_dxt must have the same length" ) if not isinstance(kx, int): raise ValueError("kx must be an int") self.training_points[name][kx + 1] = [np.array(xt), np.array(dyt_dxt)] def update_training_derivatives( self, dyt_dxt: np.ndarray, kx: int, name: Optional[str] = None ) -> None: """ Update the training data (values) at the previously set input values. Parameters ---------- dyt_dxt : np.ndarray[nt, ny] or np.ndarray[nt] The derivatives values for the nt training points. kx : int 0-based index of the derivatives being set. name :str or None, optional An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). Raises ------ ValueError The training points must be set first with set_training_values before calling update_training_values.. The number of training points does not agree with the earlier call of set_training_values. """ check_support(self, "training_derivatives") dyt_dxt = ensure_2d_array(dyt_dxt, "dyt_dxt") if kx not in self.training_points[name]: raise ValueError( "The training points must be set first with set_training_values " + "before calling update_training_values." ) xt = self.training_points[name][kx][0] if xt.shape[0] != dyt_dxt.shape[0]: raise ValueError( "The number of training points does not agree with the earlier call of " + "set_training_values." ) self.training_points[name][kx + 1][1] = np.array(dyt_dxt) def train(self) -> None: """ Train the model """ n_exact = self.training_points[None][0][0].shape[0] self.printer.active = self.options["print_global"] self.printer._line_break() self.printer._center(self.name) self.printer.active = ( self.options["print_global"] and self.options["print_problem"] ) self.printer._title("Problem size") self.printer(" %-25s : %i" % ("# training points.", n_exact)) self.printer() self.printer.active = ( self.options["print_global"] and self.options["print_training"] ) if self.name == "MixExp": # Mixture of experts model self.printer._title("Training of the Mixture of experts") else: self.printer._title("Training") # Train the model using the specified model-method with self.printer._timed_context("Training", "training"): self._train() def predict_values(self, x: np.ndarray) -> np.ndarray: """ Predict the output values at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. Returns ------- y : np.ndarray[nt, ny] Output values at the prediction points. """ x = ensure_2d_array(x, "x") self._check_xdim(x) n = x.shape[0] x2 = np.copy(x) self.printer.active = ( self.options["print_global"] and self.options["print_prediction"] ) if self.name == "MixExp": # Mixture of experts model self.printer._title("Evaluation of the Mixture of experts") else: self.printer._title("Evaluation") self.printer(" %-12s : %i" % ("# eval points.", n)) self.printer() # Evaluate the unknown points using the specified model-method with self.printer._timed_context("Predicting", key="prediction"): y = self._predict_values(x2) time_pt = self.printer._time("prediction")[-1] / n self.printer() self.printer("Prediction time/pt. (sec) : %10.7f" % time_pt) self.printer() return y.reshape((n, self.ny)) def predict_derivatives(self, x: np.ndarray, kx: int) -> np.ndarray: """ Predict the dy_dx derivatives at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- dy_dx : np.ndarray[nt, ny] Derivatives. """ check_support(self, "derivatives") x = ensure_2d_array(x, "x") self._check_xdim(x) n = x.shape[0] self.printer.active = ( self.options["print_global"] and self.options["print_prediction"] ) if self.name == "MixExp": # Mixture of experts model self.printer._title("Evaluation of the Mixture of experts") else: self.printer._title("Evaluation") self.printer(" %-12s : %i" % ("# eval points.", n)) self.printer() # Evaluate the unknown points using the specified model-method with self.printer._timed_context("Predicting", key="prediction"): y = self._predict_derivatives(x, kx) time_pt = self.printer._time("prediction")[-1] / n self.printer() self.printer("Prediction time/pt. (sec) : %10.7f" % time_pt) self.printer() return y.reshape((n, self.ny)) def predict_output_derivatives(self, x: np.ndarray) -> dict: """ Predict the derivatives dy_dyt at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. Returns ------- dy_dyt : dict of np.ndarray[nt, nt] Dictionary of output derivatives. Key is None for derivatives wrt yt and kx for derivatives wrt dyt_dxt. """ check_support(self, "output_derivatives") x = ensure_2d_array(x, "x") self._check_xdim(x) dy_dyt = self._predict_output_derivatives(x) return dy_dyt def predict_variances(self, x: np.ndarray) -> np.ndarray: """ Predict the variances at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. Returns ------- s2 : np.ndarray[nt, ny] Variances. """ check_support(self, "variances") x = ensure_2d_array(x, "x") self._check_xdim(x) n = x.shape[0] x2 = np.copy(x) s2 = self._predict_variances(x2) return s2.reshape((n, self.ny)) def predict_variance_derivatives(self, x, kx): """ Predict the derivation of the variance at a point Parameters: ----------- x : np.ndarray Input value for the prediction point. Returns: -------- derived_variance: np.ndarray The jacobian of the variance """ x = ensure_2d_array(x, "x") check_support(self, "variance_derivatives") self._check_xdim(x) n = x.shape[0] self.printer.active = ( self.options["print_global"] and self.options["print_prediction"] ) if self.name == "MixExp": # Mixture of experts model self.printer._title("Evaluation of the Mixture of experts") else: self.printer._title("Evaluation") self.printer(" %-12s : %i" % ("# eval points.", n)) self.printer() # Evaluate the unknown points using the specified model-method with self.printer._timed_context("Predicting", key="prediction"): y = self._predict_variance_derivatives(x, kx) time_pt = self.printer._time("prediction")[-1] / n self.printer() self.printer("Prediction time/pt. (sec) : %10.7f" % time_pt) self.printer() return y def _initialize(self): """ Implemented by surrogate models to declare options and declare what they support (optional). Examples -------- self.options.declare('option_name', default_value, types=(bool, int), desc='description') self.supports['derivatives'] = True """ pass def _train(self) -> None: """ Implemented by surrogate models to perform training (optional, but typically implemented). """ pass def _final_initialize(self): """ Implemented by surrogate models to complete the initialization after options are declared and possibly updated by the user. """ pass @abstractmethod def _predict_values(self, x: np.ndarray) -> np.ndarray: """ Implemented by surrogate models to predict the output values. Parameters ---------- x : np.ndarray[nt, nx] Input values for the prediction points. Returns ------- y : np.ndarray[nt, ny] Output values at the prediction points. """ raise Exception("This surrogate model is incorrectly implemented") def _predict_derivatives(self, x: np.ndarray, kx: int) -> np.ndarray: """ Implemented by surrogate models to predict the dy_dx derivatives (optional). If this method is implemented, the surrogate model should have :: self.supports['derivatives'] = True in the _initialize() implementation. Parameters ---------- x : np.ndarray[nt, nx] Input values for the prediction points. kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- dy_dx : np.ndarray[nt, ny] Derivatives. """ check_support(self, "derivatives", fail=True) def _predict_output_derivatives(self, x: np.ndarray) -> dict: """ Implemented by surrogate models to predict the dy_dyt derivatives (optional). If this method is implemented, the surrogate model should have :: self.supports['output_derivatives'] = True in the _initialize() implementation. Parameters ---------- x : np.ndarray[nt, nx] Input values for the prediction points. Returns ------- dy_dyt : dict of np.ndarray[nt, nt] Dictionary of output derivatives. Key is None for derivatives wrt yt and kx for derivatives wrt dyt_dxt. """ check_support(self, "output_derivatives", fail=True) return {} def _predict_variances(self, x: np.ndarray) -> np.ndarray: """ Implemented by surrogate models to predict the variances at a set of points (optional). If this method is implemented, the surrogate model should have :: self.supports['variances'] = True in the _initialize() implementation. Parameters ---------- x : np.ndarray[nt, nx] Input values for the prediction points. is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying for each design variable whether it is acting or not (for hierarchical design spaces) Returns ------- s2 : np.ndarray[nt, ny] Variances. """ check_support(self, "variances", fail=True) def _predict_variance_derivatives(self, x): """ Implemented by surrogate models to predict the derivation of the variance at a point (optional). Parameters: ----------- x : np.ndarray Input value for the prediction point. Returns: -------- derived_variance: np.ndarray The jacobian of the variance """ check_support(self, "variance_derivatives", fail=True) def _check_xdim(self, x): """Raise a ValueError if x dimension is not consistent with surrogate model training data dimension. This method is used as a guard in preamble of predict methods""" check_nx(self.nx, x)
18,858
31.571675
131
py
smt
smt-master/smt/surrogate_models/kpls.py
""" Author: Dr. Mohamed A. Bouhlel <mbouhlel@umich.edu> This package is distributed under New BSD license. """ import numpy as np from sklearn.cross_decomposition import PLSRegression as pls from smt.surrogate_models.krg_based import KrgBased from smt.utils.kriging import componentwise_distance_PLS import warnings import sys class KPLS(KrgBased): name = "KPLS" def _initialize(self): super(KPLS, self)._initialize() declare = self.options.declare declare("n_comp", 1, types=int, desc="Number of principal components") # KPLS used only with "abs_exp" and "squar_exp" correlations declare( "corr", "squar_exp", values=("abs_exp", "squar_exp", "pow_exp"), desc="Correlation function type", types=(str), ) declare( "eval_n_comp", False, types=(bool), values=(True, False), desc="n_comp evaluation flag", ) declare( "eval_comp_treshold", 1.0, types=(float), desc="n_comp evaluation treshold for Wold's R criterion", ) declare( "cat_kernel_comps", None, types=list, desc="Number of components for PLS categorical kernel", ) def _compute_pls(self, X, y): _pls = pls(self.options["n_comp"]) self.coeff_pls = 0 if np.shape(X)[0] < self.options["n_comp"] + 1: raise ValueError( "ValueError: The database should be at least " + str(self.options["n_comp"] + 1) + " points (currently " + str(np.shape(X)[0]) + ")." ) else: if np.shape(X)[1] == 1: self.coeff_pls = np.atleast_2d(np.array([1])) else: self.coeff_pls = abs(_pls.fit(X.copy(), y.copy()).x_rotations_) return X, y def _componentwise_distance(self, dx, opt=0, theta=None, return_derivative=False): d = componentwise_distance_PLS( dx, self.options["corr"], self.options["n_comp"], self.coeff_pls, power=self.options["pow_exp_power"], theta=theta, return_derivative=return_derivative, ) return d def _estimate_number_of_components(self): """ self.options[n_comp] value from user is ignored and replaced by an estimated one wrt Wold's R criterion. """ eval_comp_treshold = self.options["eval_comp_treshold"] X = self.training_points[None][0][0] y = self.training_points[None][0][1] k_fold = 4 nbk = int(self.nt / k_fold) press_m = 0.0 press_m1 = 0.0 self.options["n_comp"] = 0 nextcomp = True while nextcomp: self.options["n_comp"] += 1 press_m = press_m1 press_m1 = 0 self.options["theta0"] = [0.1] for fold in range(k_fold): self.nt = len(X) - nbk todel = np.arange(fold * nbk, (fold + 1) * nbk) Xfold = np.copy(X) Xfold = np.delete(X, todel, axis=0) yfold = np.copy(y) yfold = np.delete(y, todel, axis=0) Xtest = np.copy(X)[fold * nbk : (fold + 1) * nbk, :] ytest = np.copy(y)[fold * nbk : (fold + 1) * nbk, :] self.training_points[None][0][0] = Xfold self.training_points[None][0][1] = yfold try: self._new_train() except ValueError: self.options["n_comp"] -= 1 nextcomp = False break ye = self._predict_values(Xtest) press_m1 = press_m1 + np.sum(np.power((1 / len(X)) * (ye - ytest), 2)) if self.options["n_comp"] > 1 and press_m1 / press_m > eval_comp_treshold: self.options["n_comp"] -= 1 nextcomp = False self.training_points[None][0][0] = X self.training_points[None][0][1] = y self.nt = len(X) self.options["theta0"] = [0.1] def _train(self): """ Train the model """ # outputs['sol'] = self.sol if self.options["eval_n_comp"]: self._estimate_number_of_components() self._new_train()
4,500
31.854015
112
py
smt
smt-master/smt/surrogate_models/__init__.py
from .ls import LS from .qp import QP from .krg import KRG from .kpls import KPLS from .gekpls import GEKPLS from .kplsk import KPLSK from .genn import GENN from .mgp import MGP from .krg_based import MixIntKernelType from smt.utils.design_space import ( DesignSpace, FloatVariable, IntegerVariable, OrdinalVariable, CategoricalVariable, ) from smt.utils.kriging import MixHrcKernelType try: from .idw import IDW from .rbf import RBF from .rmtc import RMTC from .rmtb import RMTB except: pass
535
18.851852
46
py
smt
smt-master/smt/surrogate_models/krg_based.py
""" Author: Dr. Mohamed Amine Bouhlel <mbouhlel@umich.edu> Some functions are copied from gaussian_process submodule (Scikit-learn 0.14) This package is distributed under New BSD license. """ import numpy as np from enum import Enum from scipy import linalg, optimize from copy import deepcopy import warnings from smt.surrogate_models.surrogate_model import SurrogateModel from smt.utils.kriging import ( differences, constant, linear, quadratic, pow_exp, squar_exp, abs_exp, act_exp, cross_distances, matern52, matern32, gower_componentwise_distances, componentwise_distance, componentwise_distance_PLS, compute_X_cont, cross_levels, compute_X_cross, cross_levels_homo_space, MixHrcKernelType, matrix_data_corr_levels_cat_matrix, matrix_data_corr_levels_cat_mod, matrix_data_corr_levels_cat_mod_comps, ) from smt.utils.misc import standardization from smt.utils.checks import ensure_2d_array, check_support from scipy.stats import multivariate_normal as m_norm from smt.sampling_methods import LHS from smt.utils.design_space import ( BaseDesignSpace, ensure_design_space, CategoricalVariable, ) class MixIntKernelType(Enum): EXP_HOMO_HSPHERE = "EXP_HOMO_HSPHERE" HOMO_HSPHERE = "HOMO_HSPHERE" CONT_RELAX = "CONT_RELAX" GOWER = "GOWER" class KrgBased(SurrogateModel): _regression_types = {"constant": constant, "linear": linear, "quadratic": quadratic} _correlation_types = { "pow_exp": pow_exp, "abs_exp": abs_exp, "squar_exp": squar_exp, "act_exp": act_exp, "matern52": matern52, "matern32": matern32, } name = "KrigingBased" def _initialize(self): super(KrgBased, self)._initialize() declare = self.options.declare supports = self.supports declare( "poly", "constant", values=("constant", "linear", "quadratic"), desc="Regression function type", types=(str), ) declare( "corr", "squar_exp", values=( "pow_exp", "abs_exp", "squar_exp", "act_exp", "matern52", "matern32", ), desc="Correlation function type", ) declare( "pow_exp_power", 1.9, types=(float), desc="Power for the pow_exp kernel function (valid values in (0.0, 2.0]), This option is set automatically when corr option is squar, abs, or matern.", ) declare( "categorical_kernel", MixIntKernelType.CONT_RELAX, values=[ MixIntKernelType.CONT_RELAX, MixIntKernelType.GOWER, MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ], desc="The kernel to use for categorical inputs. Only for non continuous Kriging", ) declare( "hierarchical_kernel", MixHrcKernelType.ALG_KERNEL, values=[ MixHrcKernelType.ALG_KERNEL, MixHrcKernelType.ARC_KERNEL, ], desc="The kernel to use for mixed hierarchical inputs. Only for non continuous Kriging", ) declare( "nugget", 100.0 * np.finfo(np.double).eps, types=(float), desc="a jitter for numerical stability", ) declare( "theta0", [1e-2], types=(list, np.ndarray), desc="Initial hyperparameters" ) # In practice, in 1D and for X in [0,1], theta^{-2} in [1e-2,infty), i.e. # theta in (0,1e1], is a good choice to avoid overfitting. By standardising # X in R, X_norm = (X-X_mean)/X_std, then X_norm in [-1,1] if considering # one std intervals. This leads to theta in (0,2e1] declare( "theta_bounds", [1e-6, 2e1], types=(list, np.ndarray), desc="bounds for hyperparameters", ) declare( "hyper_opt", "Cobyla", values=("Cobyla", "TNC"), desc="Optimiser for hyperparameters optimisation", types=str, ) declare( "eval_noise", False, types=bool, values=(True, False), desc="noise evaluation flag", ) declare( "noise0", [0.0], types=(list, np.ndarray), desc="Initial noise hyperparameters", ) declare( "noise_bounds", [100.0 * np.finfo(np.double).eps, 1e10], types=(list, np.ndarray), desc="bounds for noise hyperparameters", ) declare( "use_het_noise", False, types=bool, values=(True, False), desc="heteroscedastic noise evaluation flag", ) declare( "n_start", 10, types=int, desc="number of optimizer runs (multistart method)", ) declare( "xlimits", None, types=(list, np.ndarray), desc="definition of a design space of float (continuous) variables: " "array-like of size nx x 2 (lower, upper bounds)", ) declare( "design_space", None, types=(BaseDesignSpace, list, np.ndarray), desc="definition of the (hierarchical) design space: " "use `smt.utils.design_space.DesignSpace` as the main API. Also accepts list of float variable bounds", ) self.best_iteration_fail = None self.nb_ill_matrix = 5 self.is_acting_points = {} supports["derivatives"] = True supports["variances"] = True supports["variance_derivatives"] = True supports["x_hierarchy"] = True def _final_initialize(self): # initialize default power values if self.options["corr"] == "squar_exp": self.options["pow_exp_power"] = 2.0 elif self.options["corr"] in ["abs_exp", "matern32", "matern52"]: self.options["pow_exp_power"] = 1.0 # Check the pow_exp_power is >0 and <=2 assert ( self.options["pow_exp_power"] > 0 and self.options["pow_exp_power"] <= 2 ), ( "The power value for exponential power function can only be >0 and <=2, but %s was given" % self.options["pow_exp_power"] ) @property def design_space(self) -> BaseDesignSpace: xt = self.training_points.get(None) if xt is not None: xt = xt[0][0] if self.options["design_space"] is None: self.options["design_space"] = ensure_design_space(xt=xt) elif not isinstance(self.options["design_space"], BaseDesignSpace): ds_input = self.options["design_space"] self.options["design_space"] = ensure_design_space( xt=xt, xlimits=ds_input, design_space=ds_input ) return self.options["design_space"] @property def is_continuous(self) -> bool: return self.design_space.is_all_cont def set_training_values( self, xt: np.ndarray, yt: np.ndarray, name=None, is_acting=None ) -> None: """ Set training data (values). Parameters ---------- xt : np.ndarray[nt, nx] or np.ndarray[nt] The input values for the nt training points. yt : np.ndarray[nt, ny] or np.ndarray[nt] The output values for the nt training points. name : str or None An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying which of the design variables is acting in a hierarchical design space """ super().set_training_values(xt, yt, name=name) if is_acting is not None: self.is_acting_points[name] = is_acting def _new_train(self): # Sampling points X and y X = self.training_points[None][0][0] y = self.training_points[None][0][1] # Get is_acting status from design space model if needed (might correct training points) is_acting = self.is_acting_points.get(None) if is_acting is None: X, is_acting = self.design_space.correct_get_acting(X) self.training_points[None][0][0] = X self.is_acting_points[None] = is_acting # Compute PLS-coefficients (attr of self) and modified X and y (if GEKPLS is used) if self.name not in ["Kriging", "MGP"]: if self.is_continuous: X, y = self._compute_pls(X.copy(), y.copy()) self._check_param() self.X_train = X self.is_acting_train = is_acting self._corr_params = None if not (self.is_continuous): D, self.ij, X = gower_componentwise_distances( X=X, x_is_acting=is_acting, design_space=self.design_space, hierarchical_kernel=self.options["hierarchical_kernel"], ) if self.options["categorical_kernel"] == MixIntKernelType.CONT_RELAX: X2, _ = self.design_space.unfold_x(self.training_points[None][0][0]) ( self.X2_norma, _, self.X2_offset, _, self.X2_scale, _, ) = standardization(X2, self.training_points[None][0][1]) D, _ = cross_distances(self.X2_norma) self.Lij, self.n_levels = cross_levels( X=self.X_train, ij=self.ij, design_space=self.design_space ) _, self.cat_features = compute_X_cont(self.X_train, self.design_space) # Center and scale X and y ( self.X_norma, self.y_norma, self.X_offset, self.y_mean, self.X_scale, self.y_std, ) = standardization(X, y) if not self.options["eval_noise"]: self.optimal_noise = np.array(self.options["noise0"]) elif self.options["use_het_noise"]: # hetGP works with unique design variables when noise variance are not given ( self.X_norma, index_unique, nt_reps, ) = np.unique(self.X_norma, return_inverse=True, return_counts=True, axis=0) self.nt = self.X_norma.shape[0] # computing the mean of the output per unique design variable (see Binois et al., 2018) y_norma_unique = [] for i in range(self.nt): y_norma_unique.append(np.mean(self.y_norma[index_unique == i])) # pointwise sensible estimates of the noise variances (see Ankenman et al., 2010) self.optimal_noise = self.options["noise0"] * np.ones(self.nt) for i in range(self.nt): diff = self.y_norma[index_unique == i] - y_norma_unique[i] if np.sum(diff**2) != 0.0: self.optimal_noise[i] = np.std(diff, ddof=1) ** 2 self.optimal_noise = self.optimal_noise / nt_reps self.y_norma = y_norma_unique if self.is_continuous: # Calculate matrix of distances D between samples D, self.ij = cross_distances(self.X_norma) if np.min(np.sum(np.abs(D), axis=1)) == 0.0: warnings.warn( "Warning: multiple x input features have the same value (at least same row twice)." ) #### # Regression matrix and parameters self.F = self._regression_types[self.options["poly"]](self.X_norma) n_samples_F = self.F.shape[0] if self.F.ndim > 1: p = self.F.shape[1] else: p = 1 self._check_F(n_samples_F, p) # Optimization ( self.optimal_rlf_value, self.optimal_par, self.optimal_theta, ) = self._optimize_hyperparam(D) if self.name in ["MGP"]: self._specific_train() else: if self.options["eval_noise"] and not self.options["use_het_noise"]: self.optimal_noise = self.optimal_theta[-1] self.optimal_theta = self.optimal_theta[:-1] # if self.name != "MGP": # del self.y_norma, self.D def _train(self): """ Train the model """ # outputs['sol'] = self.sol self._new_train() def _initialize_theta(self, theta, n_levels, cat_features, cat_kernel): self.n_levels_origin = n_levels if self._corr_params is not None: return self._corr_params nx = self.nx try: cat_kernel_comps = self.options["cat_kernel_comps"] if cat_kernel_comps is not None: n_levels = np.array(cat_kernel_comps) except KeyError: cat_kernel_comps = None try: ncomp = self.options["n_comp"] try: self.pls_coeff_cont except AttributeError: self.pls_coeff_cont = [] except KeyError: cat_kernel_comps = None ncomp = 1e5 theta_cont_features = np.zeros((len(theta), 1), dtype=bool) theta_cat_features = np.zeros((len(theta), len(n_levels)), dtype=bool) i = 0 j = 0 n_theta_cont = 0 for feat in cat_features: if feat: if cat_kernel in [ MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ]: theta_cat_features[ j : j + int(n_levels[i] * (n_levels[i] - 1) / 2), i ] = [True] * int(n_levels[i] * (n_levels[i] - 1) / 2) j += int(n_levels[i] * (n_levels[i] - 1) / 2) i += 1 else: if n_theta_cont < ncomp: theta_cont_features[j] = True j += 1 n_theta_cont += 1 theta_cat_features = ( [ np.where(theta_cat_features[:, i_lvl])[0] for i_lvl in range(len(n_levels)) ], np.any(theta_cat_features, axis=1) if len(n_levels) > 0 else None, ) self._corr_params = params = ( cat_kernel_comps, ncomp, theta_cat_features, theta_cont_features, nx, n_levels, ) return params def _matrix_data_corr( self, corr, design_space, power, theta, theta_bounds, dx, Lij, n_levels, cat_features, cat_kernel, x=None, ): """ matrix kernel correlation model. Parameters ---------- corr: correlation_types - The autocorrelation model design_space: BaseDesignSpace - The design space definition theta : list[small_d * n_comp] Hyperparameters of the correlation model dx: np.ndarray[n_obs * (n_obs - 1) / 2, n_comp] - The gower_componentwise_distances between the samples. Lij: np.ndarray [n_obs * (n_obs - 1) / 2, 2] - The levels corresponding to the indices i and j of the vectors in X. n_levels: np.ndarray - The number of levels for every categorical variable. cat_features: np.ndarray [dim] - Indices of the categorical input dimensions. cat_kernel : string - The kernel to use for categorical inputs. Only for non continuous Kriging", x : np.ndarray[n_obs , n_comp] - The input instead of dx for homo_hs prediction Returns ------- r: np.ndarray[n_obs * (n_obs - 1) / 2,1] An array containing the values of the autocorrelation model. """ _correlation_types = { "pow_exp": pow_exp, "abs_exp": abs_exp, "squar_exp": squar_exp, "act_exp": act_exp, "matern52": matern52, "matern32": matern32, } # Initialize static parameters ( cat_kernel_comps, ncomp, theta_cat_features, theta_cont_features, nx, n_levels, ) = self._initialize_theta(theta, n_levels, cat_features, cat_kernel) # Sampling points X and y X = self.training_points[None][0][0] y = self.training_points[None][0][1] if cat_kernel == MixIntKernelType.CONT_RELAX: X_pls_space, _ = design_space.unfold_x(X) nx = len(theta) elif cat_kernel == MixIntKernelType.GOWER: X_pls_space = np.copy(X) else: X_pls_space, _ = compute_X_cont(X, design_space) if cat_kernel_comps is not None or ncomp < 1e5: ###Modifier la condition : if PLS cont if self.pls_coeff_cont == []: X, y = self._compute_pls(X_pls_space.copy(), y.copy()) self.pls_coeff_cont = self.coeff_pls if cat_kernel in [MixIntKernelType.GOWER, MixIntKernelType.CONT_RELAX]: d = componentwise_distance_PLS( dx, corr, self.options["n_comp"], self.pls_coeff_cont, power, theta=None, return_derivative=False, ) r = _correlation_types[corr](theta, d) return r else: d_cont = componentwise_distance_PLS( dx[:, np.logical_not(cat_features)], corr, self.options["n_comp"], self.pls_coeff_cont, power, theta=None, return_derivative=False, ) else: d = componentwise_distance( dx, corr, nx, power, theta=None, return_derivative=False, ) if cat_kernel in [MixIntKernelType.GOWER, MixIntKernelType.CONT_RELAX]: r = _correlation_types[corr](theta, d) return r else: d_cont = d[:, np.logical_not(cat_features)] theta_cont = theta[theta_cont_features[:, 0]] r_cont = _correlation_types[corr](theta_cont, d_cont) r_cat = np.copy(r_cont) * 0 r = np.copy(r_cont) ##Theta_cat_i loop try: self.coeff_pls_cat except AttributeError: self.coeff_pls_cat = [] theta_cat_kernel = theta if len(n_levels) > 0: theta_cat_kernel = theta.copy() if cat_kernel == MixIntKernelType.EXP_HOMO_HSPHERE: theta_cat_kernel[theta_cat_features[1]] *= 0.5 * np.pi / theta_bounds[1] elif cat_kernel == MixIntKernelType.HOMO_HSPHERE: theta_cat_kernel[theta_cat_features[1]] *= 2.0 * np.pi / theta_bounds[1] for i in range(len(n_levels)): theta_cat = theta_cat_kernel[theta_cat_features[0][i]] T = matrix_data_corr_levels_cat_matrix( i, n_levels, theta_cat, theta_bounds, is_ehh=cat_kernel == MixIntKernelType.EXP_HOMO_HSPHERE, ) if cat_kernel_comps is not None: # Sampling points X and y X = self.training_points[None][0][0] y = self.training_points[None][0][1] X_icat = X[:, cat_features] X_icat = X_icat[:, i] old_n_comp = ( self.options["n_comp"] if "n_comp" in self.options else None ) self.options["n_comp"] = int(n_levels[i] / 2 * (n_levels[i] - 1)) X_full_space = compute_X_cross(X_icat, self.n_levels_origin[i]) try: self.coeff_pls = self.coeff_pls_cat[i] except IndexError: _, _ = self._compute_pls(X_full_space.copy(), y.copy()) self.coeff_pls_cat.append(self.coeff_pls) if x is not None: x_icat = x[:, cat_features] x_icat = x_icat[:, i] x_full_space = compute_X_cross(x_icat, self.n_levels_origin[i]) dx_cat_i = cross_levels_homo_space( x_full_space, self.ij, y=X_full_space ) else: dx_cat_i = cross_levels_homo_space(X_full_space, self.ij) d_cat_i = componentwise_distance_PLS( dx_cat_i, "squar_exp", self.options["n_comp"], self.coeff_pls, power=self.options["pow_exp_power"], theta=None, return_derivative=False, ) matrix_data_corr_levels_cat_mod_comps( i, Lij, r_cat, n_levels, T, d_cat_i, has_cat_kernel=cat_kernel in [ MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ], ) else: matrix_data_corr_levels_cat_mod( i, Lij, r_cat, T, has_cat_kernel=cat_kernel in [ MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ], ) r = np.multiply(r, r_cat) if cat_kernel_comps is not None: if old_n_comp == None: self.options._dict.pop("n_comp", None) else: self.options["n_comp"] = old_n_comp return r def _reduced_likelihood_function(self, theta): """ This function determines the BLUP parameters and evaluates the reduced likelihood function for the given autocorrelation parameters theta. Maximizing this function wrt the autocorrelation parameters theta is equivalent to maximizing the likelihood of the assumed joint Gaussian distribution of the observations y evaluated onto the design of experiments X. Parameters ---------- theta: list(n_comp), optional - An array containing the autocorrelation parameters at which the Gaussian Process model parameters should be determined. Returns ------- reduced_likelihood_function_value: real - The value of the reduced likelihood function associated to the given autocorrelation parameters theta. par: dict() - A dictionary containing the requested Gaussian Process model parameters: sigma2 Gaussian Process variance. beta Generalized least-squares regression weights for Universal Kriging or for Ordinary Kriging. gamma Gaussian Process weights. C Cholesky decomposition of the correlation matrix [R]. Ft Solution of the linear equation system : [R] x Ft = F Q, G QR decomposition of the matrix Ft. """ # Initialize output reduced_likelihood_function_value = -np.inf par = {} # Set up R nugget = self.options["nugget"] if self.options["eval_noise"]: nugget = 0 noise = self.noise0 tmp_var = theta if self.options["use_het_noise"]: noise = self.optimal_noise if self.options["eval_noise"] and not self.options["use_het_noise"]: theta = tmp_var[0:-1] noise = tmp_var[-1] if not (self.is_continuous): dx = self.D r = self._matrix_data_corr( corr=self.options["corr"], design_space=self.design_space, power=self.options["pow_exp_power"], theta=theta, theta_bounds=self.options["theta_bounds"], dx=dx, Lij=self.Lij, n_levels=self.n_levels, cat_features=self.cat_features, cat_kernel=self.options["categorical_kernel"], ).reshape(-1, 1) else: r = self._correlation_types[self.options["corr"]](theta, self.D).reshape( -1, 1 ) R = np.eye(self.nt) * (1.0 + nugget + noise) R[self.ij[:, 0], self.ij[:, 1]] = r[:, 0] R[self.ij[:, 1], self.ij[:, 0]] = r[:, 0] # Cholesky decomposition of R try: C = linalg.cholesky(R, lower=True) except (linalg.LinAlgError, ValueError) as e: print("exception : ", e) print(np.linalg.eig(R)[0]) return reduced_likelihood_function_value, par # Get generalized least squared solution Ft = linalg.solve_triangular(C, self.F, lower=True) Q, G = linalg.qr(Ft, mode="economic") sv = linalg.svd(G, compute_uv=False) rcondG = sv[-1] / sv[0] if rcondG < 1e-10: # Check F sv = linalg.svd(self.F, compute_uv=False) condF = sv[0] / sv[-1] if condF > 1e15: raise Exception( "F is too ill conditioned. Poor combination " "of regression model and observations." ) else: # Ft is too ill conditioned, get out (try different theta) return reduced_likelihood_function_value, par Yt = linalg.solve_triangular(C, self.y_norma, lower=True) beta = linalg.solve_triangular(G, np.dot(Q.T, Yt)) rho = Yt - np.dot(Ft, beta) # The determinant of R is equal to the squared product of the diagonal # elements of its Cholesky decomposition C detR = (np.diag(C) ** (2.0 / self.nt)).prod() # Compute/Organize output p = 0 q = 0 if self.name in ["MFK", "MFKPLS", "MFKPLSK"]: p = self.p q = self.q sigma2 = (rho**2.0).sum(axis=0) / (self.nt - p - q) reduced_likelihood_function_value = -(self.nt - p - q) * np.log10( sigma2.sum() ) - self.nt * np.log10(detR) par["sigma2"] = sigma2 * self.y_std**2.0 par["beta"] = beta par["gamma"] = linalg.solve_triangular(C.T, rho) par["C"] = C par["Ft"] = Ft par["G"] = G par["Q"] = Q if self.name in ["MGP"]: reduced_likelihood_function_value += self._reduced_log_prior(theta) # A particular case when f_min_cobyla fail if (self.best_iteration_fail is not None) and ( not np.isinf(reduced_likelihood_function_value) ): if reduced_likelihood_function_value > self.best_iteration_fail: self.best_iteration_fail = reduced_likelihood_function_value self._thetaMemory = np.array(tmp_var) elif (self.best_iteration_fail is None) and ( not np.isinf(reduced_likelihood_function_value) ): self.best_iteration_fail = reduced_likelihood_function_value self._thetaMemory = np.array(tmp_var) if reduced_likelihood_function_value > 1e15: reduced_likelihood_function_value = 1e15 return reduced_likelihood_function_value, par def _reduced_likelihood_gradient(self, theta): """ Evaluates the reduced_likelihood_gradient at a set of hyperparameters. Parameters --------- theta : list(n_comp), optional - An array containing the autocorrelation parameters at which the Gaussian Process model parameters should be determined. Returns ------- grad_red : np.ndarray (dim,1) Derivative of the reduced_likelihood par: dict() - A dictionary containing the requested Gaussian Process model parameters: sigma2 Gaussian Process variance. beta Generalized least-squares regression weights for Universal Kriging or for Ordinary Kriging. gamma Gaussian Process weights. C Cholesky decomposition of the correlation matrix [R]. Ft Solution of the linear equation system : [R] x Ft = F Q, G QR decomposition of the matrix Ft. dr List of all the correlation matrix derivative tr List of all the trace part in the reduce likelihood derivatives dmu List of all the mean derivatives arg List of all minus_Cinv_dRdomega_gamma dsigma List of all sigma derivatives """ red, par = self._reduced_likelihood_function(theta) C = par["C"] gamma = par["gamma"] Q = par["Q"] G = par["G"] sigma_2 = par["sigma2"] nb_theta = len(theta) grad_red = np.zeros(nb_theta) dr_all = [] tr_all = [] dmu_all = [] arg_all = [] dsigma_all = [] dbeta_all = [] for i_der in range(nb_theta): # Compute R derivatives dr = self._correlation_types[self.options["corr"]]( theta, self.D, grad_ind=i_der ) dr_all.append(dr) dR = np.zeros((self.nt, self.nt)) dR[self.ij[:, 0], self.ij[:, 1]] = dr[:, 0] dR[self.ij[:, 1], self.ij[:, 0]] = dr[:, 0] # Compute beta derivatives Cinv_dR_gamma = linalg.solve_triangular(C, np.dot(dR, gamma), lower=True) dbeta = -linalg.solve_triangular(G, np.dot(Q.T, Cinv_dR_gamma)) arg_all.append(Cinv_dR_gamma) dbeta_all.append(dbeta) # Compute mu derivatives dmu = np.dot(self.F, dbeta) dmu_all.append(dmu) # Compute log(detR) derivatives tr_1 = linalg.solve_triangular(C, dR, lower=True) tr = linalg.solve_triangular(C.T, tr_1) tr_all.append(tr) # Compute Sigma2 Derivatives dsigma_2 = ( (1 / self.nt) * ( -dmu.T.dot(gamma) - gamma.T.dot(dmu) - np.dot(gamma.T, dR.dot(gamma)) ) * self.y_std**2.0 ) dsigma_all.append(dsigma_2) # Compute reduced log likelihood derivatives grad_red[i_der] = ( -self.nt / np.log(10) * (dsigma_2 / sigma_2 + np.trace(tr) / self.nt) ) par["dr"] = dr_all par["tr"] = tr_all par["dmu"] = dmu_all par["arg"] = arg_all par["dsigma"] = dsigma_all par["dbeta_all"] = dbeta_all grad_red = np.atleast_2d(grad_red).T if self.name in ["MGP"]: grad_red += self._reduced_log_prior(theta, grad=True) return grad_red, par def _reduced_likelihood_hessian(self, theta): """ Evaluates the reduced_likelihood_gradient at a set of hyperparameters. Parameters ---------- theta : list(n_comp), optional - An array containing the autocorrelation parameters at which the Gaussian Process model parameters should be determined. Returns ------- hess : np.ndarray Hessian values. hess_ij: np.ndarray [nb_theta * (nb_theta + 1) / 2, 2] - The indices i and j of the vectors in theta associated to the hessian in hess. par: dict() - A dictionary containing the requested Gaussian Process model parameters: sigma2 Gaussian Process variance. beta Generalized least-squared regression weights for Universal Kriging or for Ordinary Kriging. gamma Gaussian Process weights. C Cholesky decomposition of the correlation matrix [R]. Ft Solution of the linear equation system : [R] x Ft = F Q, G QR decomposition of the matrix Ft. dr List of all the correlation matrix derivative tr List of all the trace part in the reduce likelihood derivatives dmu List of all the mean derivatives arg List of all minus_Cinv_dRdomega_gamma dsigma List of all sigma derivatives """ dred, par = self._reduced_likelihood_gradient(theta) C = par["C"] gamma = par["gamma"] Q = par["Q"] G = par["G"] sigma_2 = par["sigma2"] nb_theta = len(theta) dr_all = par["dr"] tr_all = par["tr"] dmu_all = par["dmu"] arg_all = par["arg"] dsigma = par["dsigma"] Rinv_dRdomega_gamma_all = [] Rinv_dmudomega_all = [] n_val_hess = nb_theta * (nb_theta + 1) // 2 hess_ij = np.zeros((n_val_hess, 2), dtype=np.int32) hess = np.zeros((n_val_hess, 1)) ind_1 = 0 if self.name in ["MGP"]: log_prior = self._reduced_log_prior(theta, hessian=True) for omega in range(nb_theta): ind_0 = ind_1 ind_1 = ind_0 + nb_theta - omega hess_ij[ind_0:ind_1, 0] = omega hess_ij[ind_0:ind_1, 1] = np.arange(omega, nb_theta) dRdomega = np.zeros((self.nt, self.nt)) dRdomega[self.ij[:, 0], self.ij[:, 1]] = dr_all[omega][:, 0] dRdomega[self.ij[:, 1], self.ij[:, 0]] = dr_all[omega][:, 0] dmudomega = dmu_all[omega] Cinv_dmudomega = linalg.solve_triangular(C, dmudomega, lower=True) Rinv_dmudomega = linalg.solve_triangular(C.T, Cinv_dmudomega) Rinv_dmudomega_all.append(Rinv_dmudomega) Rinv_dRdomega_gamma = linalg.solve_triangular(C.T, arg_all[omega]) Rinv_dRdomega_gamma_all.append(Rinv_dRdomega_gamma) for i, eta in enumerate(hess_ij[ind_0:ind_1, 1]): dRdeta = np.zeros((self.nt, self.nt)) dRdeta[self.ij[:, 0], self.ij[:, 1]] = dr_all[eta][:, 0] dRdeta[self.ij[:, 1], self.ij[:, 0]] = dr_all[eta][:, 0] dr_eta_omega = self._correlation_types[self.options["corr"]]( theta, self.D, grad_ind=omega, hess_ind=eta ) dRdetadomega = np.zeros((self.nt, self.nt)) dRdetadomega[self.ij[:, 0], self.ij[:, 1]] = dr_eta_omega[:, 0] dRdetadomega[self.ij[:, 1], self.ij[:, 0]] = dr_eta_omega[:, 0] # Compute beta second derivatives dRdeta_Rinv_dmudomega = np.dot(dRdeta, Rinv_dmudomega) dmudeta = dmu_all[eta] Cinv_dmudeta = linalg.solve_triangular(C, dmudeta, lower=True) Rinv_dmudeta = linalg.solve_triangular(C.T, Cinv_dmudeta) dRdomega_Rinv_dmudeta = np.dot(dRdomega, Rinv_dmudeta) dRdeta_Rinv_dRdomega_gamma = np.dot(dRdeta, Rinv_dRdomega_gamma) Rinv_dRdeta_gamma = linalg.solve_triangular(C.T, arg_all[eta]) dRdomega_Rinv_dRdeta_gamma = np.dot(dRdomega, Rinv_dRdeta_gamma) dRdetadomega_gamma = np.dot(dRdetadomega, gamma) beta_sum = ( dRdeta_Rinv_dmudomega + dRdomega_Rinv_dmudeta + dRdeta_Rinv_dRdomega_gamma + dRdomega_Rinv_dRdeta_gamma - dRdetadomega_gamma ) Qt_Cinv_beta_sum = np.dot( Q.T, linalg.solve_triangular(C, beta_sum, lower=True) ) dbetadetadomega = linalg.solve_triangular(G, Qt_Cinv_beta_sum) # Compute mu second derivatives dmudetadomega = np.dot(self.F, dbetadetadomega) # Compute sigma2 second derivatives sigma_arg_1 = ( -np.dot(dmudetadomega.T, gamma) + np.dot(dmudomega.T, Rinv_dRdeta_gamma) + np.dot(dmudeta.T, Rinv_dRdomega_gamma) ) sigma_arg_2 = ( -np.dot(gamma.T, dmudetadomega) + np.dot(gamma.T, dRdeta_Rinv_dmudomega) + np.dot(gamma.T, dRdomega_Rinv_dmudeta) ) sigma_arg_3 = np.dot(dmudeta.T, Rinv_dmudomega) + np.dot( dmudomega.T, Rinv_dmudeta ) sigma_arg_4_in = ( -dRdetadomega_gamma + dRdeta_Rinv_dRdomega_gamma + dRdomega_Rinv_dRdeta_gamma ) sigma_arg_4 = np.dot(gamma.T, sigma_arg_4_in) dsigma2detadomega = ( (1 / self.nt) * (sigma_arg_1 + sigma_arg_2 + sigma_arg_3 + sigma_arg_4) * self.y_std**2.0 ) # Compute Hessian dreddetadomega_tr_1 = np.trace(np.dot(tr_all[eta], tr_all[omega])) dreddetadomega_tr_2 = np.trace( linalg.solve_triangular( C.T, linalg.solve_triangular(C, dRdetadomega, lower=True) ) ) dreddetadomega_arg1 = (self.nt / sigma_2) * ( dsigma2detadomega - (1 / sigma_2) * dsigma[omega] * dsigma[eta] ) dreddetadomega = ( -(dreddetadomega_arg1 - dreddetadomega_tr_1 + dreddetadomega_tr_2) / self.nt ) hess[ind_0 + i, 0] = self.nt / np.log(10) * dreddetadomega if self.name in ["MGP"] and eta == omega: hess[ind_0 + i, 0] += log_prior[eta] par["Rinv_dR_gamma"] = Rinv_dRdomega_gamma_all par["Rinv_dmu"] = Rinv_dmudomega_all return hess, hess_ij, par def predict_values(self, x: np.ndarray, is_acting=None) -> np.ndarray: """ Predict the output values at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying for each design variable whether it is acting or not (for hierarchical design spaces) Returns ------- y : np.ndarray[nt, ny] Output values at the prediction points. """ x = ensure_2d_array(x, "x") self._check_xdim(x) if is_acting is not None: is_acting = ensure_2d_array(is_acting, "is_acting") if is_acting.shape != x.shape: raise ValueError( f"is_acting should have the same dimensions as x: {is_acting.shape} != {x.shape}" ) n = x.shape[0] x2 = np.copy(x) self.printer.active = ( self.options["print_global"] and self.options["print_prediction"] ) if self.name == "MixExp": # Mixture of experts model self.printer._title("Evaluation of the Mixture of experts") else: self.printer._title("Evaluation") self.printer(" %-12s : %i" % ("# eval points.", n)) self.printer() # Evaluate the unknown points using the specified model-method with self.printer._timed_context("Predicting", key="prediction"): y = self._predict_values(x2, is_acting=is_acting) time_pt = self.printer._time("prediction")[-1] / n self.printer() self.printer("Prediction time/pt. (sec) : %10.7f" % time_pt) self.printer() return y.reshape((n, self.ny)) def _predict_values(self, x: np.ndarray, is_acting=None) -> np.ndarray: """ Evaluates the model at a set of points. Parameters ---------- x : np.ndarray [n_evals, dim] Evaluation point input variable values is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying for each design variable whether it is acting or not (for hierarchical design spaces) Returns ------- y : np.ndarray Evaluation point output variable values """ # Initialization if is_acting is None: x, is_acting = self.design_space.correct_get_acting(x) n_eval, n_features_x = x.shape if not (self.is_continuous): dx = gower_componentwise_distances( x, x_is_acting=is_acting, design_space=self.design_space, hierarchical_kernel=self.options["hierarchical_kernel"], y=np.copy(self.X_train), y_is_acting=self.is_acting_train, ) if self.options["categorical_kernel"] == MixIntKernelType.CONT_RELAX: Xpred, _ = self.design_space.unfold_x(x) Xpred_norma = (Xpred - self.X2_offset) / self.X2_scale dx = differences(Xpred_norma, Y=self.X2_norma.copy()) _, ij = cross_distances(x, self.X_train) Lij, _ = cross_levels( X=x, ij=ij, design_space=self.design_space, y=self.X_train ) self.ij = ij r = self._matrix_data_corr( corr=self.options["corr"], design_space=self.design_space, power=self.options["pow_exp_power"], theta=self.optimal_theta, theta_bounds=self.options["theta_bounds"], dx=dx, Lij=Lij, n_levels=self.n_levels, cat_features=self.cat_features, cat_kernel=self.options["categorical_kernel"], x=x, ).reshape(n_eval, self.nt) X_cont, _ = compute_X_cont(x, self.design_space) X_cont = (X_cont - self.X_offset) / self.X_scale else: X_cont = (x - self.X_offset) / self.X_scale # Get pairwise componentwise L1-distances to the input training set dx = differences(X_cont, Y=self.X_norma.copy()) d = self._componentwise_distance(dx) # Compute the correlation function r = self._correlation_types[self.options["corr"]]( self.optimal_theta, d ).reshape(n_eval, self.nt) y = np.zeros(n_eval) # Compute the regression function f = self._regression_types[self.options["poly"]](X_cont) # Scaled predictor y_ = np.dot(f, self.optimal_par["beta"]) + np.dot(r, self.optimal_par["gamma"]) # Predictor y = (self.y_mean + self.y_std * y_).ravel() return y def _predict_derivatives(self, x, kx): """ Evaluates the derivatives at a set of points. Parameters --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- y : np.ndarray Derivative values. """ # Initialization n_eval, n_features_x = x.shape x = (x - self.X_offset) / self.X_scale # Get pairwise componentwise L1-distances to the input training set dx = differences(x, Y=self.X_norma.copy()) d = self._componentwise_distance(dx) dd = self._componentwise_distance( dx, theta=self.optimal_theta, return_derivative=True ) # Compute the correlation function derivative_dic = {"dx": dx, "dd": dd} r, dr = self._correlation_types[self.options["corr"]]( self.optimal_theta, d, derivative_params=derivative_dic ) r = r.reshape(n_eval, self.nt) drx = dr[:, kx].reshape(n_eval, self.nt) if self.options["poly"] == "constant": df = np.zeros((1, self.nx)) elif self.options["poly"] == "linear": df = np.zeros((self.nx + 1, self.nx)) df[1:, :] = np.eye(self.nx) else: raise ValueError( "The derivative is only available for ordinary kriging or " + "universal kriging using a linear trend" ) # Beta and gamma = R^-1(y-FBeta) beta = self.optimal_par["beta"] gamma = self.optimal_par["gamma"] df_dx = np.dot(df.T, beta) y = (df_dx[kx] + np.dot(drx, gamma)) * self.y_std / self.X_scale[kx] return y def predict_variances(self, x: np.ndarray, is_acting=None) -> np.ndarray: """ Predict the variances at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying for each design variable whether it is acting or not (for hierarchical design spaces) Returns ------- s2 : np.ndarray[nt, ny] Variances. """ check_support(self, "variances") x = ensure_2d_array(x, "x") self._check_xdim(x) if is_acting is not None: is_acting = ensure_2d_array(is_acting, "is_acting") if is_acting.shape != x.shape: raise ValueError( f"is_acting should have the same dimensions as x: {is_acting.shape} != {x.shape}" ) n = x.shape[0] x2 = np.copy(x) s2 = self._predict_variances(x2, is_acting=is_acting) return s2.reshape((n, self.ny)) def _predict_variances(self, x: np.ndarray, is_acting=None) -> np.ndarray: """ Provide uncertainty of the model at a set of points Parameters ---------- x : np.ndarray [n_evals, dim] Evaluation point input variable values is_acting : np.ndarray[nt, nx] or np.ndarray[nt] Matrix specifying for each design variable whether it is acting or not (for hierarchical design spaces) Returns ------- MSE : np.ndarray Evaluation point output variable MSE """ # Initialization if is_acting is None: x, is_acting = self.design_space.correct_get_acting(x) n_eval, n_features_x = x.shape X_cont = x if not (self.is_continuous): dx = gower_componentwise_distances( x, x_is_acting=is_acting, design_space=self.design_space, hierarchical_kernel=self.options["hierarchical_kernel"], y=np.copy(self.X_train), y_is_acting=self.is_acting_train, ) if self.options["categorical_kernel"] == MixIntKernelType.CONT_RELAX: Xpred, _ = self.design_space.unfold_x(x) Xpred_norma = (Xpred - self.X2_offset) / self.X2_scale dx = differences(Xpred_norma, Y=self.X2_norma.copy()) _, ij = cross_distances(x, self.X_train) Lij, _ = cross_levels( X=x, ij=ij, design_space=self.design_space, y=self.X_train ) self.ij = ij r = self._matrix_data_corr( corr=self.options["corr"], design_space=self.design_space, power=self.options["pow_exp_power"], theta=self.optimal_theta, theta_bounds=self.options["theta_bounds"], dx=dx, Lij=Lij, n_levels=self.n_levels, cat_features=self.cat_features, cat_kernel=self.options["categorical_kernel"], x=x, ).reshape(n_eval, self.nt) X_cont, _ = compute_X_cont(x, self.design_space) X_cont = (X_cont - self.X_offset) / self.X_scale else: x = (x - self.X_offset) / self.X_scale X_cont = np.copy(x) # Get pairwise componentwise L1-distances to the input training set dx = differences(x, Y=self.X_norma.copy()) d = self._componentwise_distance(dx) # Compute the correlation function r = self._correlation_types[self.options["corr"]]( self.optimal_theta, d ).reshape(n_eval, self.nt) C = self.optimal_par["C"] rt = linalg.solve_triangular(C, r.T, lower=True) u = linalg.solve_triangular( self.optimal_par["G"].T, np.dot(self.optimal_par["Ft"].T, rt) - self._regression_types[self.options["poly"]](X_cont).T, ) A = self.optimal_par["sigma2"] B = 1.0 - (rt**2.0).sum(axis=0) + (u**2.0).sum(axis=0) MSE = np.einsum("i,j -> ji", A, B) # Mean Squared Error might be slightly negative depending on # machine precision: force to zero! MSE[MSE < 0.0] = 0.0 return MSE def _predict_variance_derivatives(self, x, kx): """ Provide the derivative of the variance of the model at a set of points Parameters ----------- x : np.ndarray [n_evals, dim] Evaluation point input variable values kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- derived_variance: np.ndarray The jacobian of the variance of the kriging model (the kx-th derivative) """ # Initialization n_eval, n_features_x = x.shape x = (x - self.X_offset) / self.X_scale theta = self.optimal_theta # Get pairwise componentwise L1-distances to the input training set dx = differences(x, Y=self.X_norma.copy()) d = self._componentwise_distance(dx) dd = self._componentwise_distance( dx, theta=self.optimal_theta, return_derivative=True ) derivative_dic = {"dx": dx, "dd": dd} sigma2 = self.optimal_par["sigma2"] C = self.optimal_par["C"] # p1 : derivative of (rt**2.0).sum(axis=0) r, dr = self._correlation_types[self.options["corr"]]( theta, d, derivative_params=derivative_dic ) r = r.reshape(n_eval, self.nt) drx = dr[:, kx] drx = drx.reshape(n_eval, self.nt) rt = linalg.solve_triangular(C, r.T, lower=True) invKr = linalg.solve_triangular(C.T, rt) p1 = 2 * np.dot(drx, invKr).T # p2 : derivative of (u**2.0).sum(axis=0) f_x = self._regression_types[self.options["poly"]](x).T F = self.F rho2 = linalg.solve_triangular(C, F, lower=True) invKF = linalg.solve_triangular(C.T, rho2) A = f_x.T - np.dot(r, invKF) B = np.dot(F.T, invKF) rho3 = linalg.cholesky(B, lower=True) invBAt = linalg.solve_triangular(rho3, A.T, lower=True) D = linalg.solve_triangular(rho3.T, invBAt) if self.options["poly"] == "constant": df = np.zeros((1, self.nx)) elif self.options["poly"] == "linear": df = np.zeros((self.nx + 1, self.nx)) df[1:, :] = np.eye(self.nx) else: raise ValueError( "The derivative is only available for ordinary kriging or " + "universal kriging using a linear trend" ) dA = df[:, kx].T - np.dot(drx, invKF) p3 = 2 * np.dot(dA, D).T # prime : derivative of MSE # MSE ~1.0 - (rt**2.0).sum(axis=0) + (u**2.0).sum(axis=0) prime = 0 - p1 + p3 ## scaling factors x_std = self.X_scale[kx] derived_variance = np.array((np.outer(sigma2, np.diag(prime.T)) / x_std)) return np.atleast_2d(derived_variance.T) def _optimize_hyperparam(self, D): """ This function evaluates the Gaussian Process model at x. Parameters ---------- D: np.ndarray [n_obs * (n_obs - 1) / 2, dim] - The componentwise cross-spatial-correlation-distance between the vectors in X. Returns ------- best_optimal_rlf_value: real - The value of the reduced likelihood function associated to the best autocorrelation parameters theta. best_optimal_par: dict() - A dictionary containing the requested Gaussian Process model parameters. best_optimal_theta: list(n_comp) or list(dim) - The best hyperparameters found by the optimization. """ # reinitialize optimization best values self.best_iteration_fail = None self._thetaMemory = None # Initialize the hyperparameter-optimization if self.name in ["MGP"]: def minus_reduced_likelihood_function(theta): res = -self._reduced_likelihood_function(theta)[0] return res def grad_minus_reduced_likelihood_function(theta): grad = -self._reduced_likelihood_gradient(theta)[0] return grad else: def minus_reduced_likelihood_function(log10t): return -self._reduced_likelihood_function(theta=10.0**log10t)[0] def grad_minus_reduced_likelihood_function(log10t): log10t_2d = np.atleast_2d(log10t).T res = ( -np.log(10.0) * (10.0**log10t_2d) * (self._reduced_likelihood_gradient(10.0**log10t_2d)[0]) ) return res limit, _rhobeg = max(10 * len(self.options["theta0"]), 25), 0.5 exit_function = False if "KPLSK" in self.name: n_iter = 1 else: n_iter = 0 for ii in range(n_iter, -1, -1): ( best_optimal_theta, best_optimal_rlf_value, best_optimal_par, constraints, ) = ( [], [], [], [], ) bounds_hyp = [] self.theta0 = deepcopy(self.options["theta0"]) for i in range(len(self.theta0)): # In practice, in 1D and for X in [0,1], theta^{-2} in [1e-2,infty), # i.e. theta in (0,1e1], is a good choice to avoid overfitting. # By standardising X in R, X_norm = (X-X_mean)/X_std, then # X_norm in [-1,1] if considering one std intervals. This leads # to theta in (0,2e1] theta_bounds = self.options["theta_bounds"] if self.theta0[i] < theta_bounds[0] or self.theta0[i] > theta_bounds[1]: self.theta0[i] = np.random.rand() self.theta0[i] = ( self.theta0[i] * (theta_bounds[1] - theta_bounds[0]) + theta_bounds[0] ) warnings.warn( "Warning: theta0 is out the feasible bounds. A random initialisation is used instead." ) if self.name in ["MGP"]: # to be discussed with R. Priem constraints.append(lambda theta, i=i: theta[i] + theta_bounds[1]) constraints.append(lambda theta, i=i: theta_bounds[1] - theta[i]) bounds_hyp.append((-theta_bounds[1], theta_bounds[1])) else: log10t_bounds = np.log10(theta_bounds) constraints.append(lambda log10t, i=i: log10t[i] - log10t_bounds[0]) constraints.append(lambda log10t, i=i: log10t_bounds[1] - log10t[i]) bounds_hyp.append(log10t_bounds) if self.name in ["MGP"]: theta0_rand = m_norm.rvs( self.options["prior"]["mean"] * len(self.theta0), self.options["prior"]["var"], 1, ) theta0 = self.theta0 else: theta_bounds = self.options["theta_bounds"] log10t_bounds = np.log10(theta_bounds) theta0_rand = np.random.rand(len(self.theta0)) theta0_rand = ( theta0_rand * (log10t_bounds[1] - log10t_bounds[0]) + log10t_bounds[0] ) theta0 = np.log10(self.theta0) if not (self.is_continuous): self.D = D else: ##from abs distance to kernel distance self.D = self._componentwise_distance(D, opt=ii) # Initialization k, incr, stop, best_optimal_rlf_value, max_retry = 0, 0, 1, -1e20, 10 while k < stop: # Use specified starting point as first guess self.noise0 = np.array(self.options["noise0"]) noise_bounds = self.options["noise_bounds"] if self.options["eval_noise"] and not self.options["use_het_noise"]: self.noise0[self.noise0 == 0.0] = noise_bounds[0] for i in range(len(self.noise0)): if ( self.noise0[i] < noise_bounds[0] or self.noise0[i] > noise_bounds[1] ): self.noise0[i] = noise_bounds[0] warnings.warn( "Warning: noise0 is out the feasible bounds. The lowest possible value is used instead." ) theta0 = np.concatenate( [theta0, np.log10(np.array([self.noise0]).flatten())] ) theta0_rand = np.concatenate( [ theta0_rand, np.log10(np.array([self.noise0]).flatten()), ] ) for i in range(len(self.noise0)): noise_bounds = np.log10(noise_bounds) constraints.append( lambda log10t: log10t[i + len(self.theta0)] - noise_bounds[0] ) constraints.append( lambda log10t: noise_bounds[1] - log10t[i + len(self.theta0)] ) bounds_hyp.append(noise_bounds) theta_limits = np.repeat( np.log10([theta_bounds]), repeats=len(theta0), axis=0 ) theta_all_loops = np.vstack((theta0, theta0_rand)) if self.options["n_start"] > 1: sampling = LHS( xlimits=theta_limits, criterion="maximin", random_state=41 ) theta_lhs_loops = sampling(self.options["n_start"]) theta_all_loops = np.vstack((theta_all_loops, theta_lhs_loops)) optimal_theta_res = {"fun": float("inf")} optimal_theta_res_loop = None try: if self.options["hyper_opt"] == "Cobyla": for theta0_loop in theta_all_loops: optimal_theta_res_loop = optimize.minimize( minus_reduced_likelihood_function, theta0_loop, constraints=[ {"fun": con, "type": "ineq"} for con in constraints ], method="COBYLA", options={ "rhobeg": _rhobeg, "tol": 1e-4, "maxiter": limit, }, ) if optimal_theta_res_loop["fun"] < optimal_theta_res["fun"]: optimal_theta_res = optimal_theta_res_loop elif self.options["hyper_opt"] == "TNC": theta_all_loops = 10**theta_all_loops for theta0_loop in theta_all_loops: optimal_theta_res_loop = optimize.minimize( minus_reduced_likelihood_function, theta0_loop, method="TNC", jac=grad_minus_reduced_likelihood_function, bounds=bounds_hyp, options={"maxiter": 100}, ) if optimal_theta_res_loop["fun"] < optimal_theta_res["fun"]: optimal_theta_res = optimal_theta_res_loop if "x" not in optimal_theta_res: raise ValueError( f"Optimizer encountered a problem: {optimal_theta_res_loop!s}" ) optimal_theta = optimal_theta_res["x"] if self.name not in ["MGP"]: optimal_theta = 10**optimal_theta optimal_rlf_value, optimal_par = self._reduced_likelihood_function( theta=optimal_theta ) # Compare the new optimizer to the best previous one if k > 0: if np.isinf(optimal_rlf_value): stop += 1 if incr != 0: return if stop > max_retry: raise ValueError( "%d attempts to train the model failed" % max_retry ) else: if optimal_rlf_value >= self.best_iteration_fail: if optimal_rlf_value > best_optimal_rlf_value: best_optimal_rlf_value = optimal_rlf_value best_optimal_par = optimal_par best_optimal_theta = optimal_theta else: if ( self.best_iteration_fail > best_optimal_rlf_value ): best_optimal_theta = self._thetaMemory ( best_optimal_rlf_value, best_optimal_par, ) = self._reduced_likelihood_function( theta=best_optimal_theta ) else: if np.isinf(optimal_rlf_value): stop += 1 else: best_optimal_rlf_value = optimal_rlf_value best_optimal_par = optimal_par best_optimal_theta = optimal_theta k += 1 except ValueError as ve: # raise ve # If iteration is max when fmin_cobyla fail is not reached if self.nb_ill_matrix > 0: self.nb_ill_matrix -= 1 k += 1 stop += 1 # One evaluation objectif function is done at least if self.best_iteration_fail is not None: if self.best_iteration_fail > best_optimal_rlf_value: best_optimal_theta = self._thetaMemory ( best_optimal_rlf_value, best_optimal_par, ) = self._reduced_likelihood_function( theta=best_optimal_theta ) # Optimization fail elif best_optimal_par == []: print("Optimization failed. Try increasing the ``nugget``") raise ve # Break the while loop else: k = stop + 1 print("fmin_cobyla failed but the best value is retained") if "KPLSK" in self.name: if self.options["eval_noise"]: # best_optimal_theta contains [theta, noise] if eval_noise = True theta = best_optimal_theta[:-1] else: # best_optimal_theta contains [theta] if eval_noise = False theta = best_optimal_theta if exit_function: return best_optimal_rlf_value, best_optimal_par, best_optimal_theta if self.options["corr"] == "squar_exp": self.options["theta0"] = (theta * self.coeff_pls**2).sum(1) else: self.options["theta0"] = (theta * np.abs(self.coeff_pls)).sum(1) self.options["n_comp"] = int(self.nx) limit = 10 * self.options["n_comp"] self.best_iteration_fail = None exit_function = True return best_optimal_rlf_value, best_optimal_par, best_optimal_theta def _check_param(self): """ This function checks some parameters of the model and amend theta0 if possible (see _amend_theta0_option). """ d = self.options["n_comp"] if "n_comp" in self.options else self.nx if self.name in ["KPLS"]: if self.options["corr"] not in ["pow_exp", "squar_exp", "abs_exp"]: raise ValueError( "KPLS only works with a squared exponential, or an absolute exponential kernel with variable power" ) if ( self.options["categorical_kernel"] not in [ MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ] and self.name == "KPLS" ): if self.options["cat_kernel_comps"] is not None: raise ValueError( "cat_kernel_comps option is for homoscedastic kernel." ) mat_dim = ( self.options["cat_kernel_comps"] if "cat_kernel_comps" in self.options else None ) n_comp = self.options["n_comp"] if "n_comp" in self.options else None n_param = compute_n_param( self.design_space, self.options["categorical_kernel"], d, n_comp, mat_dim, ) self.options["theta0"] *= np.ones(n_param) if len(self.options["theta0"]) != d and ( self.options["categorical_kernel"] == MixIntKernelType.GOWER or self.is_continuous ): if len(self.options["theta0"]) == 1: self.options["theta0"] *= np.ones(d) else: raise ValueError( "the length of theta0 (%s) should be equal to the number of dim (%s)." % (len(self.options["theta0"]), d) ) if self.options["use_het_noise"] and not self.options["eval_noise"]: if len(self.options["noise0"]) != self.nt: if len(self.options["noise0"]) == 1: self.options["noise0"] *= np.ones(self.nt) else: raise ValueError( "for the heteroscedastic case, the length of noise0 (%s) should be equal to the number of observations (%s)." % (len(self.options["noise0"]), self.nt) ) if not self.options["use_het_noise"]: if len(self.options["noise0"]) != 1: raise ValueError( "for the homoscedastic noise case, the length of noise0 (%s) should be equal to one." % (len(self.options["noise0"])) ) if self.supports["training_derivatives"]: if not (1 in self.training_points[None]): raise Exception("Derivative values are needed for using the GEK model.") def _check_F(self, n_samples_F, p): """ This function check the F-parameters of the model. """ if n_samples_F != self.nt: raise Exception( "Number of rows in F and X do not match. Most " "likely something is going wrong with the " "regression model." ) if p > n_samples_F: raise Exception( ( "Ordinary least squares problem is undetermined " "n_samples=%d must be greater than the " "regression model size p=%d." ) % (self.nt, p) ) def compute_n_param(design_space, cat_kernel, d, n_comp, mat_dim): """ Returns the he number of parameters needed for an homoscedastic or full group kernel. Parameters ---------- design_space: BaseDesignSpace - design space definition cat_kernel : string -The kernel to use for categorical inputs. Only for non continuous Kriging, d: int - n_comp or nx n_comp : int - if PLS, then it is the number of components else None, mat_dim : int - if PLS, then it is the number of components for matrix kernel (mixed integer) else None, Returns ------- n_param: int - The number of parameters. """ n_param = design_space.n_dv if n_comp is not None: n_param = d if cat_kernel == MixIntKernelType.CONT_RELAX: return n_param if mat_dim is not None: return int(np.sum([l * (l - 1) / 2 for l in mat_dim]) + n_param) if cat_kernel == MixIntKernelType.GOWER: return n_param for i, dv in enumerate(design_space.design_variables): if isinstance(dv, CategoricalVariable): n_values = dv.n_values if design_space.n_dv == d: n_param -= 1 if cat_kernel in [ MixIntKernelType.EXP_HOMO_HSPHERE, MixIntKernelType.HOMO_HSPHERE, ]: n_param += int(n_values * (n_values - 1) / 2) if cat_kernel == MixIntKernelType.CONT_RELAX: n_param += int(n_values) return n_param
72,518
36.987952
163
py
smt
smt-master/smt/surrogate_models/qp.py
""" Author: Dr. Mohamed Amine Bouhlel <mbouhlel@umich.edu> Dr. Nathalie.bartoli <nathalie@onera.fr> This package is distributed under New BSD license. TO DO: - define outputs['sol'] = self.sol """ import numpy as np import scipy from smt.surrogate_models.surrogate_model import SurrogateModel from smt.utils.caching import cached_operation from smt.utils.misc import standardization class QP(SurrogateModel): """ Square polynomial approach """ name = "QP" def _initialize(self): super(QP, self)._initialize() declare = self.options.declare supports = self.supports declare( "data_dir", values=None, types=str, desc="Directory for loading / saving cached data; None means do not save or load", ) supports["derivatives"] = True ############################################################################ # Model functions ############################################################################ def _new_train(self): """ Train the model """ X = self.training_points[None][0][0] y = self.training_points[None][0][1] ( self.X_norma, self.y_norma, self.X_offset, self.y_mean, self.X_scale, self.y_std, ) = standardization(X, y) if X.shape[0] < (self.nx + 1) * (self.nx + 2) / 2.0: raise Exception( "Number of training points should be greater or equal to %d." % ((self.nx + 1) * (self.nx + 2) / 2.0) ) X = self._response_surface(self.X_norma) self.coef = np.dot(np.linalg.inv(np.dot(X.T, X)), (np.dot(X.T, self.y_norma))) def _train(self): """ Train the model """ inputs = {"self": self} with cached_operation(inputs, self.options["data_dir"]) as outputs: if outputs: self.sol = outputs["sol"] else: self._new_train() def _response_surface(self, x): """ Build the response surface of degree 2 argument ----------- x : np.ndarray [nt, nx] Training points Returns ------- M : np.ndarray Matrix of the surface """ dim = self.nx n = x.shape[0] n_app = int(scipy.special.binom(dim + 2, dim)) M = np.zeros((n_app, n)) x = x.T M[0, :] = np.ones((1, n)) for i in range(1, dim + 1): M[i, :] = x[i - 1, :] for i in range(dim + 1, 2 * dim + 1): M[i, :] = x[i - (dim + 1), :] ** 2 for i in range(dim - 1): for j in range(i + 1, dim): k = int(2 * dim + 2 + (i) * dim - ((i + 1) * (i)) / 2 + (j - (i + 2))) M[k, :] = x[i, :] * x[j, :] return M.T def _predict_derivatives(self, x, kx): """ Evaluates the derivatives at a set of points. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- y : np.ndarray Derivative values. """ dim = self.nx x = (x - self.X_offset) / self.X_scale linear_coef = self.coef[1 + kx, :] quad_coef = 2 * self.coef[1 + dim + kx, :] * x[:, kx] neval = np.size(quad_coef, 0) cross_coef = np.zeros(neval) for i in range(dim): if i > kx: k = int( 2 * dim + 2 + (kx) * dim - ((kx + 1) * (kx)) / 2 + (i - (kx + 2)) ) cross_coef += self.coef[k, :] * x[:, i] elif i < kx: k = int(2 * dim + 2 + (i) * dim - ((i + 1) * (i)) / 2 + (kx - (i + 2))) cross_coef += self.coef[k, :] * x[:, i] y = ( (linear_coef + quad_coef + cross_coef).reshape((x.shape[0], self.ny)) * self.y_std / self.X_scale[kx] ) return y def _predict_values(self, x): """ Evaluates the model at a set of points. Arguments --------- x : np.ndarray [n_evals, nx] Evaluation point input variable values Returns ------- y : np.ndarray [n_evals, ny] Evaluation point output variable values """ x = (x - self.X_offset) / self.X_scale M = self._response_surface(x) y_ = np.dot(M, self.coef) y = (self.y_mean + self.y_std * y_).ravel() return y
4,775
28.300613
98
py
smt
smt-master/smt/surrogate_models/tests/test_krg_het_noise.py
""" Author: Andres Lopez-Lopera <<andres.lopez_lopera@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import KRG from smt.utils.sm_test_case import SMTestCase from smt.utils import compute_rms_error class Test(SMTestCase): def test_predict_output(self): xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 1.1, 1.0]) # Adding noisy repetitions np.random.seed(6) yt_std_rand = np.std(yt) * np.random.uniform(size=yt.shape) xt_full = np.array(3 * xt.tolist()) yt_full = np.concatenate((yt, yt + 0.2 * yt_std_rand, yt - 0.2 * yt_std_rand)) sm = KRG(theta0=[1.0], eval_noise=True, use_het_noise=True, n_start=1) sm.set_training_values(xt_full, yt_full) sm.train() yt = yt.reshape(-1, 1) y = sm.predict_values(xt) t_error = np.linalg.norm(y - yt) / np.linalg.norm(yt) self.assert_error(t_error, 0.0, 1e-2) if __name__ == "__main__": unittest.main()
1,072
27.236842
86
py
smt
smt-master/smt/surrogate_models/tests/test_kpls.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import KPLS from smt.problems import Sphere from smt.sampling_methods import FullFactorial, LHS class TestKPLS(unittest.TestCase): def test_predict_output(self): for corr_str in [ "pow_exp", "abs_exp", "squar_exp", ]: d, n = (3, 3) sx = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), d, axis=0), criterion="m", random_state=42, ) x = sx(n) sy = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), 1, axis=0), criterion="m", random_state=42, ) y = sy(n) kriging = KPLS(n_comp=2, corr=corr_str) kriging.set_training_values(x, y) kriging.train() x_fail_1 = np.asarray([[0, 0, 0, 0]]) x_fail_2 = np.asarray([[0]]) self.assertRaises(ValueError, lambda: kriging.predict_values(x_fail_1)) self.assertRaises(ValueError, lambda: kriging.predict_values(x_fail_2)) var = kriging.predict_variances(x) self.assertEqual(y.shape[0], var.shape[0]) kriging = KPLS(n_comp=3) kriging.set_training_values(x, y) self.assertRaises(ValueError, lambda: kriging.train()) def test_kpls_training_with_zeroed_outputs(self): # Test scikit-learn 0.24 regression cf. https://github.com/SMTorg/smt/issues/274 for corr_str in [ "pow_exp", "abs_exp", "squar_exp", ]: x = np.random.rand(50, 3) y = np.zeros(50) kpls = KPLS() kpls.options["corr"] = corr_str kpls.set_training_values(x, y) kpls.train() x_test = np.asarray([[0, 0, 0], [0.5, 0.5, 0.5], [1, 1, 1]]) y_test = kpls.predict_values(x_test) # KPLS training fails anyway but not due to PLS exception StopIteration self.assertEqual(np.linalg.norm(y_test - np.asarray([[0, 0, 0]])), 0) if __name__ == "__main__": unittest.main()
2,294
30.875
88
py
smt
smt-master/smt/surrogate_models/tests/test_mgp.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import MGP from smt.problems import Sphere from smt.sampling_methods import FullFactorial, LHS class TestMGP(unittest.TestCase): def test_predict_output_shapes(self): d, n = (3, 10) sx = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), d, axis=0), criterion="m", random_state=42, ) x = sx(n) sy = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), 1, axis=0), criterion="m", random_state=42, ) y = sy(n) y = y.flatten() mgp = MGP(n_comp=2) mgp.set_training_values(x, y) mgp.train() x_fail_1 = np.asarray([0, 0, 0, 0]) x_fail_2 = np.asarray([0]) self.assertRaises(ValueError, lambda: mgp.predict_values(x_fail_1)) self.assertRaises(ValueError, lambda: mgp.predict_values(x_fail_2)) self.assertRaises(ValueError, lambda: mgp.predict_variances(x_fail_1)) self.assertRaises(ValueError, lambda: mgp.predict_variances(x_fail_2)) xtest = np.array([[0.0, 0.0, 0.0], [1.0, 2.0, 3.0]]) n_samples = xtest.shape[0] ypred = mgp.predict_values(xtest) self.assertEqual(ypred.shape, (n_samples, 1)) var = mgp.predict_variances(xtest) self.assertEqual(var.shape, (n_samples, 1)) var_no_uq = mgp.predict_variances_no_uq(xtest) self.assertEqual(var_no_uq.shape, (n_samples, 1)) if __name__ == "__main__": unittest.main()
1,659
28.122807
78
py
smt
smt-master/smt/surrogate_models/tests/test_krg_based.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models.krg_based import KrgBased class TestKrgBased(unittest.TestCase): def test_theta0_default_init(self): krg = KrgBased() krg.set_training_values(np.array([[1, 2, 3]]), np.array([[1]])) krg._check_param() self.assertTrue(np.array_equal(krg.options["theta0"], [1e-2, 1e-2, 1e-2])) def test_theta0_one_dim_init(self): krg = KrgBased(theta0=[2e-2]) krg.set_training_values(np.array([[1, 2, 3]]), np.array([[1]])) krg._check_param() self.assertTrue(np.array_equal(krg.options["theta0"], [2e-2, 2e-2, 2e-2])) def test_theta0_erroneous_init(self): krg = KrgBased(theta0=[2e-2, 1e-2]) krg.set_training_values(np.array([[1, 2]]), np.array([[1]])) # correct krg._check_param() krg.set_training_values(np.array([[1, 2, 3]]), np.array([[1]])) # erroneous self.assertRaises(ValueError, krg._check_param) if __name__ == "__main__": unittest.main()
1,127
30.333333
84
py
smt
smt-master/smt/surrogate_models/tests/test_rmts.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np import matplotlib.pyplot as plt from smt.utils.silence import Silence from smt.utils.sm_test_case import SMTestCase from smt.utils import compute_rms_error from smt.surrogate_models import RMTB, RMTC def function_test_1d(x): # function xsinx x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y = (x - 3.5) * np.sin((x - 3.5) / (np.pi)) return y.reshape((-1, 1)) class TestRMTS(SMTestCase): def setUp(self): # xt = np.random.rand(5, 1) * 10.0 self.xt = np.array( [[3.6566495, 4.64266046, 7.23645433, 6.04862594, 8.85571712]] ).T self.yt = function_test_1d(self.xt) self.xlimits = np.array([[0.0, 25.0]]) smref = RMTB(xlimits=self.xlimits, print_global=False) smref.set_training_values(self.xt, self.yt) with Silence(): smref.train() self.xref = np.array([[0.0, 6.25, 12.5, 18.75, 25.0]]).T self.yref = smref.predict_values(self.xref) self.sms = {} def test_linear_search(self): for ls in ["bracketed", "cubic", "quadratic", "null"]: self.sms[ls] = RMTB( xlimits=self.xlimits, line_search=ls, print_global=False ) self.sms[ls].set_training_values(self.xt, self.yt) with Silence(): self.sms[ls].train() error = compute_rms_error(self.sms[ls], self.xref, self.yref) self.assert_error(error, 0.0, 1e-1) def test_linear_solver(self): for ls in [ "krylov-dense", "dense-chol", "lu", "ilu", "krylov", "krylov-lu", "krylov-mg", "gs", "jacobi", "mg", "null", ]: self.sms[ls] = RMTB(xlimits=self.xlimits, solver=ls, print_global=False) self.sms[ls].set_training_values(self.xt, self.yt) with Silence(): self.sms[ls].train() error = compute_rms_error(self.sms[ls], self.xref, self.yref) self.assert_error(error, 0.0, 1.1) if __name__ == "__main__": unittest.main()
2,291
26.614458
84
py
smt
smt-master/smt/surrogate_models/tests/test_krg_training.py
#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ Created on Mon Mar 23 15:20:29 2020 @author: ninamoello """ from __future__ import print_function, division import numpy as np import unittest from smt.utils.sm_test_case import SMTestCase from smt.utils.kriging import ( pow_exp, abs_exp, squar_exp, act_exp, cross_distances, componentwise_distance, matern52, matern32, ) from smt.utils.misc import standardization from smt.sampling_methods.lhs import LHS from smt.surrogate_models import KRG, MGP print_output = False class Test(SMTestCase): def setUp(self): eps = 1e-8 xlimits = np.asarray([[0, 1], [0, 1]]) self.random = np.random.RandomState(42) lhs = LHS(xlimits=xlimits, random_state=self.random) X = lhs(8) y = LHS(xlimits=np.asarray([[0, 1]]), random_state=self.random)(8) X_norma, y_norma, X_offset, y_mean, X_scale, y_std = standardization(X, y) D, ij = cross_distances(X_norma) theta = self.random.rand(2) corr_str = [ "pow_exp", "abs_exp", "squar_exp", "act_exp", "matern32", "matern52", ] corr_def = [pow_exp, abs_exp, squar_exp, act_exp, matern32, matern52] power_val = { "pow_exp": 1.9, "abs_exp": 1.0, "squar_exp": 2.0, "act_exp": 1.0, "matern32": 1.0, "matern52": 1.0, } self.eps = eps self.X = X self.y = y ( self.X_norma, self.y_norma, self.X_offset, self.y_mean, self.X_scale, self.y_std, ) = ( X_norma, y_norma, X_offset, y_mean, X_scale, y_std, ) self.D, self.ij = D, ij self.theta = theta self.corr_str = corr_str self.corr_def = corr_def self.power_val = power_val def test_noise_estimation(self): xt = np.array([[0.0], [1.0], [2.0], [3.0], [4.0]]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = KRG(hyper_opt="Cobyla", eval_noise=True, noise0=[1e-4]) sm.set_training_values(xt, yt) sm.train() self.assert_error(np.array(sm.optimal_theta), np.array([1.6]), 1e-1, 1e-1) def test_corr_derivatives(self): for ind, corr in enumerate(self.corr_def): # For every kernel # self.corr_str[ind] = self.corr_def[ind] D = componentwise_distance( self.D, self.corr_str[ind], self.X.shape[1], self.power_val[self.corr_str[ind]], ) k = corr(self.theta, D) K = np.eye(self.X.shape[0]) K[self.ij[:, 0], self.ij[:, 1]] = k[:, 0] K[self.ij[:, 1], self.ij[:, 0]] = k[:, 0] grad_norm_all = [] diff_norm_all = [] ind_theta = [] for i, theta_i in enumerate(self.theta): eps_theta = np.zeros(self.theta.shape) eps_theta[i] = self.eps k_dk = corr(self.theta + eps_theta, D) K_dk = np.eye(self.X.shape[0]) K_dk[self.ij[:, 0], self.ij[:, 1]] = k_dk[:, 0] K_dk[self.ij[:, 1], self.ij[:, 0]] = k_dk[:, 0] grad_eps = (K_dk - K) / self.eps dk = corr(self.theta, D, grad_ind=i) dK = np.zeros((self.X.shape[0], self.X.shape[0])) dK[self.ij[:, 0], self.ij[:, 1]] = dk[:, 0] dK[self.ij[:, 1], self.ij[:, 0]] = dk[:, 0] grad_norm_all.append(np.linalg.norm(dK)) diff_norm_all.append(np.linalg.norm(grad_eps)) ind_theta.append(r"$x_%d$" % i) self.assert_error( np.array(grad_norm_all), np.array(diff_norm_all), 1e-5, 1e-5 ) # from utils/smt_test_case.py def test_corr_hessian(self): for ind, corr in enumerate(self.corr_def): # For every kernel # self.corr_str[ind] = self.corr_def[ind] D = componentwise_distance( self.D, self.corr_str[ind], self.X.shape[1], self.power_val[self.corr_str[ind]], ) grad_norm_all = [] diff_norm_all = [] for i, theta_i in enumerate(self.theta): k = corr(self.theta, D, grad_ind=i) K = np.eye(self.X.shape[0]) K[self.ij[:, 0], self.ij[:, 1]] = k[:, 0] K[self.ij[:, 1], self.ij[:, 0]] = k[:, 0] for j, omega_j in enumerate(self.theta): eps_omega = np.zeros(self.theta.shape) eps_omega[j] = self.eps k_dk = corr(self.theta + eps_omega, D, grad_ind=i) K_dk = np.eye(self.X.shape[0]) K_dk[self.ij[:, 0], self.ij[:, 1]] = k_dk[:, 0] K_dk[self.ij[:, 1], self.ij[:, 0]] = k_dk[:, 0] grad_eps = (K_dk - K) / self.eps dk = corr(self.theta, D, grad_ind=i, hess_ind=j) dK = np.zeros((self.X.shape[0], self.X.shape[0])) dK[self.ij[:, 0], self.ij[:, 1]] = dk[:, 0] dK[self.ij[:, 1], self.ij[:, 0]] = dk[:, 0] grad_norm_all.append(np.linalg.norm(dK)) diff_norm_all.append(np.linalg.norm(grad_eps)) self.assert_error( np.array(grad_norm_all), np.array(diff_norm_all), 1e-5, 1e-5 ) # from utils/smt_test_case.py def test_likelihood_derivatives(self): for corr_str in [ "pow_exp", "abs_exp", "squar_exp", "act_exp", "matern32", "matern52", ]: # For every kernel for poly_str in ["constant", "linear", "quadratic"]: # For every method if corr_str == "act_exp": kr = MGP(print_global=False) theta = self.random.rand(4) else: kr = KRG(print_global=False) theta = self.theta kr.options["poly"] = poly_str kr.options["corr"] = corr_str kr.options["pow_exp_power"] = self.power_val[corr_str] kr.set_training_values(self.X, self.y) kr.train() grad_red, dpar = kr._reduced_likelihood_gradient(theta) red, par = kr._reduced_likelihood_function(theta) grad_norm_all = [] diff_norm_all = [] ind_theta = [] for i, theta_i in enumerate(theta): eps_theta = theta.copy() eps_theta[i] = eps_theta[i] + self.eps red_dk, par_dk = kr._reduced_likelihood_function(eps_theta) dred_dk = (red_dk - red) / self.eps grad_norm_all.append(grad_red[i]) diff_norm_all.append(float(dred_dk)) ind_theta.append(r"$x_%d$" % i) grad_norm_all = np.atleast_2d(grad_norm_all) diff_norm_all = np.atleast_2d(diff_norm_all).T self.assert_error( grad_norm_all, diff_norm_all, atol=1e-5, rtol=1e-3 ) # from utils/smt_test_case.py def test_likelihood_hessian(self): for corr_str in [ "pow_exp", "abs_exp", "squar_exp", "act_exp", "matern32", "matern52", ]: # For every kernel for poly_str in ["constant", "linear", "quadratic"]: # For every method if corr_str == "act_exp": kr = MGP(print_global=False) theta = self.random.rand(4) else: kr = KRG(print_global=False) theta = self.theta kr.options["poly"] = poly_str kr.options["corr"] = corr_str kr.options["pow_exp_power"] = self.power_val[corr_str] kr.set_training_values(self.X, self.y) kr.train() grad_red, dpar = kr._reduced_likelihood_gradient(theta) hess, hess_ij, _ = kr._reduced_likelihood_hessian(theta) Hess = np.zeros((theta.shape[0], theta.shape[0])) Hess[hess_ij[:, 0], hess_ij[:, 1]] = hess[:, 0] Hess[hess_ij[:, 1], hess_ij[:, 0]] = hess[:, 0] grad_norm_all = [] diff_norm_all = [] ind_theta = [] for j, omega_j in enumerate(theta): eps_omega = theta.copy() eps_omega[j] += self.eps grad_red_eps, _ = kr._reduced_likelihood_gradient(eps_omega) for i, theta_i in enumerate(theta): hess_eps = (grad_red_eps[i] - grad_red[i]) / self.eps grad_norm_all.append( np.linalg.norm(Hess[i, j]) / np.linalg.norm(Hess) ) diff_norm_all.append( np.linalg.norm(hess_eps) / np.linalg.norm(Hess) ) ind_theta.append(r"$x_%d,x_%d$" % (j, i)) self.assert_error( np.array(grad_norm_all), np.array(diff_norm_all), atol=1e-5, rtol=1e-3, ) # from utils/smt_test_case.py def test_variance_derivatives(self): for corr_str in [ "abs_exp", "squar_exp", "matern32", "matern52", "pow_exp", ]: kr = KRG(print_global=False) kr.options["poly"] = "constant" kr.options["corr"] = corr_str kr.options["pow_exp_power"] = self.power_val[corr_str] kr.set_training_values(self.X, self.y) kr.train() e = 1e-6 xa = self.random.random() xb = self.random.random() x_valid = [[xa, xb], [xa + e, xb], [xa - e, xb], [xa, xb + e], [xa, xb - e]] y_predicted = kr.predict_variances(np.array(x_valid)) y_jacob = np.zeros((2, 5)) for i in range(np.shape(x_valid)[0]): l0 = kr.predict_variance_derivatives(np.atleast_2d(x_valid[i]), 0)[0] l1 = kr.predict_variance_derivatives(np.atleast_2d(x_valid[i]), 1)[0] y_jacob[0, i] = l0 y_jacob[1, i] = l1 diff_g = (y_predicted[1] - y_predicted[2]) / (2 * e) diff_d = (y_predicted[3] - y_predicted[4]) / (2 * e) jac_rel_error1 = abs((y_jacob[0][0] - diff_g) / y_jacob[0][0]) self.assert_error(jac_rel_error1, 1e-3, atol=0.01, rtol=0.01) jac_rel_error2 = abs((y_jacob[1][0] - diff_d) / y_jacob[1][1]) self.assert_error(jac_rel_error2, 1e-3, atol=0.01, rtol=0.01) if __name__ == "__main__": print_output = True unittest.main()
11,319
35.050955
88
py
smt
smt-master/smt/surrogate_models/tests/test_surrogate_model_examples.py
""" Author: John Hwang <<hwangjt@umich.edu>> This package is distributed under New BSD license. """ import unittest import matplotlib matplotlib.use("Agg") try: from smt.surrogate_models import IDW, RBF, RMTB, RMTC compiled_available = True except: compiled_available = False class Test(unittest.TestCase): @unittest.skipIf(not compiled_available, "C compilation failed") def test_idw(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import IDW xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = IDW(p=2) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() @unittest.skipIf(not compiled_available, "C compilation failed") def test_rbf(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import RBF xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = RBF(d0=5) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() @unittest.skipIf(not compiled_available, "C compilation failed") def test_rmtb(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import RMTB xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) xlimits = np.array([[0.0, 4.0]]) sm = RMTB( xlimits=xlimits, order=4, num_ctrl_pts=20, energy_weight=1e-15, regularization_weight=0.0, ) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() @unittest.skipIf(not compiled_available, "C compilation failed") def test_rmtc(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import RMTC xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) xlimits = np.array([[0.0, 4.0]]) sm = RMTC( xlimits=xlimits, num_elements=20, energy_weight=1e-15, regularization_weight=0.0, ) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() def test_ls(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import LS xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = LS() sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() def test_qp(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import QP xt = np.array([[0.0, 1.0, 2.0, 3.0, 4.0]]).T yt = np.array([[0.2, 1.4, 1.5, 0.9, 1.0], [0.0, 1.0, 2.0, 4, 3]]).T sm = QP() sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) t1, _ = plt.plot(xt, yt[:, 0], "o", "C0") p1 = plt.plot(x, y[:, 0], "C0", label="Prediction 1") t2, _ = plt.plot(xt, yt[:, 1], "o", "C1") p2 = plt.plot(x, y[:, 1], "C1", label="Prediction 2") plt.xlabel("x") plt.ylabel("y") plt.legend() plt.show() def test_krg(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KRG xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = KRG(theta0=[1e-2]) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) # estimated variance s2 = sm.predict_variances(x) # derivative according to the first variable dydx = sm.predict_derivatives(xt, 0) fig, axs = plt.subplots(1) # add a plot with variance axs.plot(xt, yt, "o") axs.plot(x, y) axs.fill_between( np.ravel(x), np.ravel(y - 3 * np.sqrt(s2)), np.ravel(y + 3 * np.sqrt(s2)), color="lightgrey", ) axs.set_xlabel("x") axs.set_ylabel("y") axs.legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="lower right", ) plt.show() def test_mixed_int_krg(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KRG from smt.applications.mixed_integer import MixedIntegerKrigingModel from smt.utils.design_space import DesignSpace, IntegerVariable xt = np.array([0.0, 2.0, 3.0]) yt = np.array([0.0, 1.5, 0.9]) design_space = DesignSpace( [ IntegerVariable(0, 4), ] ) sm = MixedIntegerKrigingModel( surrogate=KRG(design_space=design_space, theta0=[1e-2]) ) sm.set_training_values(xt, yt) sm.train() num = 500 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) # estimated variance s2 = sm.predict_variances(x) fig, axs = plt.subplots(1) axs.plot(xt, yt, "o") axs.plot(x, y) axs.fill_between( np.ravel(x), np.ravel(y - 3 * np.sqrt(s2)), np.ravel(y + 3 * np.sqrt(s2)), color="lightgrey", ) axs.set_xlabel("x") axs.set_ylabel("y") axs.legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="lower right", ) plt.show() def test_mixed_gower_krg(self): from smt.surrogate_models import ( MixIntKernelType, KRG, DesignSpace, CategoricalVariable, ) from smt.applications.mixed_integer import ( MixedIntegerKrigingModel, ) import matplotlib.pyplot as plt import numpy as np xt = np.array([0, 3, 4]) yt = np.array([0.0, 1.0, 1.5]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-2], categorical_kernel=MixIntKernelType.GOWER, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 5, 5) y = sm.predict_values(x) plt.plot(xt, yt, "o", label="data") plt.plot(x, y, "d", color="red", markersize=3, label="pred") plt.xlabel("x") plt.ylabel("y") plt.legend() plt.show() def test_kpls_auto(self): import numpy as np from smt.surrogate_models import KPLS from smt.problems import TensorProduct from smt.sampling_methods import LHS # The problem is the exponential problem with dimension 10 ndim = 10 prob = TensorProduct(ndim=ndim, func="exp") sm = KPLS(eval_n_comp=True) samp = LHS(xlimits=prob.xlimits, random_state=42) np.random.seed(0) xt = samp(50) yt = prob(xt) np.random.seed(1) sm.set_training_values(xt, yt) sm.train() ## The model automatically choose a dimension of 3 l = sm.options["n_comp"] print("\n The model automatically choose " + str(l) + " components.") ## You can predict a 10-dimension point from the 3-dimensional model print(sm.predict_values(np.array([[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]]))) print(sm.predict_variances(np.array([[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]]))) def test_kpls(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KPLS xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = KPLS(theta0=[1e-2]) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) # estimated variance # add a plot with variance s2 = sm.predict_variances(x) # to compute the derivative according to the first variable dydx = sm.predict_derivatives(xt, 0) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() plt.plot(xt, yt, "o") plt.plot(x, y) plt.fill_between( np.ravel(x), np.ravel(y - 3 * np.sqrt(s2)), np.ravel(y + 3 * np.sqrt(s2)), color="lightgrey", ) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction", "Confidence Interval 99%"]) plt.show() def test_kplsk(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KPLSK xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.9, 1.0]) sm = KPLSK(theta0=[1e-2]) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) # estimated variance s2 = sm.predict_variances(x) # derivative according to the first variable dydx = sm.predict_derivatives(xt, 0) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() # add a plot with variance plt.plot(xt, yt, "o") plt.plot(x, y) plt.fill_between( np.ravel(x), np.ravel(y - 3 * np.sqrt(s2)), np.ravel(y + 3 * np.sqrt(s2)), color="lightgrey", ) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction", "Confidence Interval 99%"]) plt.show() def test_gekpls(self): import numpy as np from mpl_toolkits.mplot3d import Axes3D import matplotlib.pyplot as plt from smt.surrogate_models import GEKPLS, DesignSpace from smt.problems import Sphere from smt.sampling_methods import LHS # Construction of the DOE fun = Sphere(ndim=2) sampling = LHS(xlimits=fun.xlimits, criterion="m") xt = sampling(20) yt = fun(xt) # Compute the gradient for i in range(2): yd = fun(xt, kx=i) yt = np.concatenate((yt, yd), axis=1) design_space = DesignSpace(fun.xlimits) # Build the GEKPLS model n_comp = 2 sm = GEKPLS( design_space=design_space, theta0=[1e-2] * n_comp, extra_points=1, print_prediction=False, n_comp=n_comp, ) sm.set_training_values(xt, yt[:, 0]) for i in range(2): sm.set_training_derivatives(xt, yt[:, 1 + i].reshape((yt.shape[0], 1)), i) sm.train() # Test the model X = np.arange(fun.xlimits[0, 0], fun.xlimits[0, 1], 0.25) Y = np.arange(fun.xlimits[1, 0], fun.xlimits[1, 1], 0.25) X, Y = np.meshgrid(X, Y) Z = np.zeros((X.shape[0], X.shape[1])) for i in range(X.shape[0]): for j in range(X.shape[1]): Z[i, j] = sm.predict_values( np.hstack((X[i, j], Y[i, j])).reshape((1, 2)) ) fig = plt.figure() ax = fig.add_subplot(projection="3d") surf = ax.plot_surface(X, Y, Z) plt.show() def test_genn(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models.genn import GENN, load_smt_data # Training data lower_bound = -np.pi upper_bound = np.pi number_of_training_points = 4 xt = np.linspace(lower_bound, upper_bound, number_of_training_points) yt = xt * np.sin(xt) dyt_dxt = np.sin(xt) + xt * np.cos(xt) # Validation data number_of_validation_points = 30 xv = np.linspace(lower_bound, upper_bound, number_of_validation_points) yv = xv * np.sin(xv) dyv_dxv = np.sin(xv) + xv * np.cos(xv) # Truth model x = np.arange(lower_bound, upper_bound, 0.01) y = x * np.sin(x) # GENN genn = GENN() genn.options["alpha"] = 0.1 # learning rate that controls optimizer step size genn.options["beta1"] = 0.9 # tuning parameter to control ADAM optimization genn.options["beta2"] = 0.99 # tuning parameter to control ADAM optimization genn.options[ "lambd" ] = 0.1 # lambd = 0. = no regularization, lambd > 0 = regularization genn.options[ "gamma" ] = 1.0 # gamma = 0. = no grad-enhancement, gamma > 0 = grad-enhancement genn.options["deep"] = 2 # number of hidden layers genn.options["wide"] = 6 # number of nodes per hidden layer genn.options[ "mini_batch_size" ] = 64 # used to divide data into training batches (use for large data sets) genn.options["num_epochs"] = 20 # number of passes through data genn.options[ "num_iterations" ] = 100 # number of optimizer iterations per mini-batch genn.options["is_print"] = True # print output (or not) load_smt_data( genn, xt, yt, dyt_dxt ) # convenience function to read in data that is in SMT format genn.train() # API function to train model genn.plot_training_history() # non-API function to plot training history (to check convergence) genn.goodness_of_fit( xv, yv, dyv_dxv ) # non-API function to check accuracy of regression y_pred = genn.predict_values( x ) # API function to predict values at new (unseen) points # Plot fig, ax = plt.subplots() ax.plot(x, y_pred) ax.plot(x, y, "k--") ax.plot(xv, yv, "ro") ax.plot(xt, yt, "k+", mew=3, ms=10) ax.set(xlabel="x", ylabel="y", title="GENN") ax.legend(["Predicted", "True", "Test", "Train"]) plt.show() def test_mgp(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import MGP from smt.sampling_methods import LHS # Construction of the DOE dim = 3 def fun(x): import numpy as np res = ( np.sum(x, axis=1) ** 2 - np.sum(x, axis=1) + 0.2 * (np.sum(x, axis=1) * 1.2) ** 3 ) return res sampling = LHS( xlimits=np.asarray([(-1, 1)] * dim), criterion="m", random_state=42 ) xt = sampling(8) yt = np.atleast_2d(fun(xt)).T # Build the MGP model sm = MGP( theta0=[1e-2], print_prediction=False, n_comp=1, ) sm.set_training_values(xt, yt[:, 0]) sm.train() # Get the transfert matrix A emb = sm.embedding["C"] # Compute the smallest box containing all points of A upper = np.sum(np.abs(emb), axis=0) lower = -upper # Test the model u_plot = np.atleast_2d(np.arange(lower, upper, 0.01)).T x_plot = sm.get_x_from_u(u_plot) # Get corresponding points in Omega y_plot_true = fun(x_plot) y_plot_pred = sm.predict_values(u_plot) sigma_MGP = sm.predict_variances(u_plot) sigma_KRG = sm.predict_variances_no_uq(u_plot) u_train = sm.get_u_from_x(xt) # Get corresponding points in A # Plots fig, ax = plt.subplots() ax.plot(u_plot, y_plot_pred, label="Predicted") ax.plot(u_plot, y_plot_true, "k--", label="True") ax.plot(u_train, yt, "k+", mew=3, ms=10, label="Train") ax.fill_between( u_plot[:, 0], y_plot_pred[:, 0] - 3 * sigma_MGP[:, 0], y_plot_pred[:, 0] + 3 * sigma_MGP[:, 0], color="r", alpha=0.5, label="Variance with hyperparameters uncertainty", ) ax.fill_between( u_plot[:, 0], y_plot_pred[:, 0] - 3 * sigma_KRG[:, 0], y_plot_pred[:, 0] + 3 * sigma_KRG[:, 0], color="b", alpha=0.5, label="Variance without hyperparameters uncertainty", ) ax.set(xlabel="x", ylabel="y", title="MGP") fig.legend(loc="upper center", ncol=2) fig.tight_layout() fig.subplots_adjust(top=0.74) plt.show() if __name__ == "__main__": unittest.main()
18,375
28.354633
104
py
smt
smt-master/smt/surrogate_models/tests/test_qp.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>>, Frederic Zahle This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import QP, KRG from smt.examples.rans_crm_wing.rans_crm_wing import ( get_rans_crm_wing, plot_rans_crm_wing, ) class TestQP(unittest.TestCase): def test_ny(self): xt, yt, _ = get_rans_crm_wing() interp = QP() interp.set_training_values(xt, yt) interp.train() v0 = np.zeros((4, 2)) for ix, i in enumerate([10, 11, 12, 13]): v0[ix, :] = interp.predict_values(np.atleast_2d(xt[i, :])) v1 = interp.predict_values(np.atleast_2d(xt[10:14, :])) expected_diff = np.zeros((4, 2)) np.testing.assert_allclose(v1 - v0, expected_diff, atol=1e-15)
818
25.419355
70
py
smt
smt-master/smt/surrogate_models/tests/__init__.py
0
0
0
py
smt
smt-master/smt/surrogate_models/tests/test_krg_predictions.py
""" Authors: Nathalie Bartoli, Paul Saves This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import KRG from smt.sampling_methods import LHS from smt.utils.sm_test_case import SMTestCase class Test(SMTestCase): def setUp(self): def pb(x): # sin + linear trend y = ( np.atleast_2d(np.sin(x[:, 0])).T + np.atleast_2d(2 * x[:, 0] + 5 * x[:, 1]).T + 10 ) # + linear trend return y xlimits = np.array([[-5, 10], [-5, 10]]) sampling = LHS(xlimits=xlimits, random_state=42) self.xt = sampling(12) self.yt = pb(self.xt) def test_predictions(self): trends = ["constant", "linear"] kernels = ["pow_exp", "squar_exp", "abs_exp", "matern32", "matern52"] powers = [1.0, 1.5, 2.0] for trend in trends: for kernel in kernels: if kernel == "pow_exp": for power in powers: sm = KRG( theta0=[0.01], print_global=False, poly=trend, corr=kernel, pow_exp_power=power, ) # ,eval_noise=True) sm.set_training_values(self.xt, self.yt) sm.train() print(f"\n*** TREND = {trend} & kernel = {kernel} ***\n") # quality of the surrogate on validation points Test._check_prediction_variances(self, sm) Test._check_prediction_derivatives(self, sm) else: sm = KRG( theta0=[0.01], print_global=False, poly=trend, corr=kernel ) # ,eval_noise=True) sm.set_training_values(self.xt, self.yt) sm.train() print(f"\n*** TREND = {trend} & kernel = {kernel} ***\n") # quality of the surrogate on validation points Test._check_prediction_variances(self, sm) Test._check_prediction_derivatives(self, sm) @staticmethod def _check_prediction_variances(self, sm): y_predicted = sm.predict_variances(self.xt) variance_at_training_inputs = np.sum(y_predicted**2) np.testing.assert_allclose(variance_at_training_inputs, 0, atol=1e-9) @staticmethod def _check_prediction_derivatives(self, sm): e = 5e-6 xa = -1.3 xb = 2.5 x_valid = np.array( [[xa, xb], [xa + e, xb], [xa - e, xb], [xa, xb + e], [xa, xb - e]] ) y_predicted = sm.predict_variances(x_valid) x = np.atleast_2d(x_valid[0]) diff_g = (y_predicted[1, 0] - y_predicted[2, 0]) / (2 * e) diff_d = (y_predicted[3, 0] - y_predicted[4, 0]) / (2 * e) deriv = np.array( [ sm.predict_variance_derivatives(x, 0)[0], sm.predict_variance_derivatives(x, 1)[0], ] ).T pred_errors = np.array( [ np.abs((diff_g - deriv[0][0]) / diff_g), np.abs((diff_d - deriv[0][1]) / diff_d), ] ) total_error = np.sum(pred_errors**2) np.testing.assert_allclose(total_error, 0, atol=5e-3) y_predicted = sm.predict_values(x_valid) x = np.atleast_2d(x_valid[0]) diff_g = (y_predicted[1, 0] - y_predicted[2, 0]) / (2 * e) diff_d = (y_predicted[3, 0] - y_predicted[4, 0]) / (2 * e) deriv = np.array( [sm.predict_derivatives(x, 0)[0], sm.predict_derivatives(x, 1)[0]] ).T pred_errors = np.array( [ np.abs((diff_g - deriv[0][0]) / diff_g), np.abs((diff_d - deriv[0][1]) / diff_d), ] ) total_error = np.sum(pred_errors**2) np.testing.assert_allclose(total_error, 0, atol=1e-9) ### VECTORIZATION TESTS x_valid = np.concatenate( ( x_valid, np.atleast_2d(np.array([x_valid[0][0] + 1.0, x_valid[0][1] + 1.0])), ) ) # test predict values & variances vectorization all_vals1 = np.zeros((6, 2)) for i, x in enumerate(x_valid): all_vals1[i, 0] = sm.predict_values(np.atleast_2d(x)) all_vals1[i, 1] = sm.predict_variances(np.atleast_2d(x)) all_vals2x = sm.predict_values(np.atleast_2d(x_valid)).flatten() all_vals2y = sm.predict_variances(np.atleast_2d(x_valid)).flatten() total_error = np.sum( [ np.power(all_vals1[:, 0] - all_vals2x, 2), np.power(all_vals1[:, 1] - all_vals2y, 2), ] ) np.testing.assert_allclose(total_error, 0, atol=1e-9) # test predict_derivatives vectorization all_vals1 = np.zeros((6, 2)) for i, x in enumerate(x_valid): all_vals1[i, 0] = sm.predict_derivatives(np.atleast_2d(x), 0) all_vals1[i, 1] = sm.predict_derivatives(np.atleast_2d(x), 1) all_vals2x = sm.predict_derivatives(np.atleast_2d(x_valid), 0).flatten() all_vals2y = sm.predict_derivatives(np.atleast_2d(x_valid), 1).flatten() total_error = np.sum( [ np.power(all_vals1[:, 0] - all_vals2x, 2), np.power(all_vals1[:, 1] - all_vals2y, 2), ] ) np.testing.assert_allclose(total_error, 0, atol=1e-9) # test predict_variance_derivatives vectorization all_vals1 = np.zeros((6, 2)) for i, x in enumerate(x_valid): all_vals1[i, 0] = sm.predict_variance_derivatives(np.atleast_2d(x), 0) all_vals1[i, 1] = sm.predict_variance_derivatives(np.atleast_2d(x), 1) all_vals2x = sm.predict_variance_derivatives( np.atleast_2d(x_valid), 0 ).flatten() all_vals2y = sm.predict_variance_derivatives( np.atleast_2d(x_valid), 1 ).flatten() total_error = np.sum( [ np.power(all_vals1[:, 0] - all_vals2x, 2), np.power(all_vals1[:, 1] - all_vals2y, 2), ] ) np.testing.assert_allclose(total_error, 0, atol=1e-9) if __name__ == "__main__": unittest.main()
6,514
34.026882
84
py
smt
smt-master/smt/surrogate_models/tests/test_krg_outputs.py
""" Author: Remi Lafage <<remi.lafage@onera.fr>> This package is distributed under New BSD license. """ import unittest import numpy as np from smt.surrogate_models import KRG from smt.problems import Sphere from smt.sampling_methods import FullFactorial, LHS class TestKRG(unittest.TestCase): def test_predict_output_shape(self): d, n = (3, 10) sx = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), d, axis=0), criterion="m", random_state=42, ) x = sx(n) # 2-dimensional output n_s = 2 sy = LHS( xlimits=np.repeat(np.atleast_2d([0.0, 1.0]), n_s, axis=0), criterion="m", random_state=42, ) y = sy(n) kriging = KRG(poly="linear") kriging.set_training_values(x, y) kriging.train() val = kriging.predict_values(x) self.assertEqual(y.shape, val.shape) var = kriging.predict_variances(x) self.assertEqual(y.shape, var.shape) for kx in range(d): val_deriv = kriging.predict_derivatives(x, kx) self.assertEqual(y.shape, val_deriv.shape) var_deriv = kriging.predict_variance_derivatives(x, 0) self.assertEqual((n, n_s), var_deriv.shape) if __name__ == "__main__": unittest.main()
1,347
25.431373
70
py
smt
smt-master/smt/applications/mfk.py
# -*- coding: utf-8 -*- """ Created on Fri May 04 10:26:49 2018 @author: Mostafa Meliani <melimostafa@gmail.com> Multi-Fidelity co-Kriging: recursive formulation with autoregressive model of order 1 (AR1) Adapted on January 2021 by Andres Lopez-Lopera to the new SMT version """ from copy import deepcopy import numpy as np from scipy.linalg import solve_triangular from scipy import linalg from scipy.spatial.distance import cdist from sklearn.cross_decomposition import PLSRegression as pls from smt.surrogate_models.krg_based import KrgBased from smt.sampling_methods import LHS from smt.utils.kriging import ( cross_distances, componentwise_distance, differences, ) from smt.utils.misc import standardization class NestedLHS(object): def __init__(self, nlevel, xlimits, random_state=None): """ Constructor where values of options can be passed in. Parameters ---------- nlevel : integer. The number of design of experiments to be built xlimits : ndarray The interval of the domain in each dimension with shape (nx, 2) random_state : Numpy RandomState object or seed number which controls random draws """ self.nlevel = nlevel self.xlimits = xlimits self.random_state = random_state def __call__(self, nb_samples_hifi): """ Builds nlevel nested design of experiments of dimension dim and size n_samples. Each doe sis built with the optmized lhs procedure. Builds the highest level first; nested properties are ensured by deleting the nearest neighbours in lower levels of fidelity. Parameters ---------- nb_samples_hifi: The number of samples of the highest fidelity model. nb_samples_fi(n-1) = 2 * nb_samples_fi(n) Returns ------ list of length nlevel of design of experiemnts from low to high fidelity level. """ nt = [] for i in range(self.nlevel, 0, -1): nt.append(pow(2, i - 1) * nb_samples_hifi) if len(nt) != self.nlevel: raise ValueError("nt must be a list of nlevel elements") if np.allclose(np.sort(nt)[::-1], nt) == False: raise ValueError("nt must be a list of decreasing integers") doe = [] p0 = LHS(xlimits=self.xlimits, criterion="ese", random_state=self.random_state) doe.append(p0(nt[0])) for i in range(1, self.nlevel): p = LHS( xlimits=self.xlimits, criterion="ese", random_state=self.random_state ) doe.append(p(nt[i])) for i in range(1, self.nlevel)[::-1]: ind = [] d = cdist(doe[i], doe[i - 1], "euclidean") for j in range(doe[i].shape[0]): dj = np.sort(d[j, :]) k = dj[0] l = (np.where(d[j, :] == k))[0][0] m = 0 while l in ind: m = m + 1 k = dj[m] l = (np.where(d[j, :] == k))[0][0] ind.append(l) doe[i - 1] = np.delete(doe[i - 1], ind, axis=0) doe[i - 1] = np.vstack((doe[i - 1], doe[i])) return doe class MFK(KrgBased): def _initialize(self): super()._initialize() declare = self.options.declare declare( "rho_regr", "constant", values=("constant", "linear", "quadratic"), desc="Regression function type for rho", ) declare( "optim_var", False, types=bool, values=(True, False), desc="If True, the variance at HF samples is forced to zero", ) declare( "propagate_uncertainty", True, types=bool, values=(True, False), desc="If True, the variance cotribution of lower fidelity levels are considered", ) self.name = "MFK" def _differences(self, X, Y): """ Compute the distances """ return differences(X, Y) def _check_list_structure(self, X, y): """ checks if the data structure is compatible with MFK. sets class attributes such as (number of levels of Fidelity, training points in each level, ...) Arguments : X : list of arrays, each array corresponds to a fidelity level. starts from lowest to highest y : same as X """ if type(X) is not list: nlevel = 1 X = [X] else: nlevel = len(X) if type(y) is not list: y = [y] if len(X) != len(y): raise ValueError("X and y must have the same length.") n_samples = np.zeros(nlevel, dtype=int) n_features = np.zeros(nlevel, dtype=int) n_samples_y = np.zeros(nlevel, dtype=int) for i in range(nlevel): n_samples[i], n_features[i] = X[i].shape if i > 1 and n_features[i] != n_features[i - 1]: raise ValueError("All X must have the same number of columns.") y[i] = np.asarray(y[i]).ravel()[:, np.newaxis] n_samples_y[i] = y[i].shape[0] if n_samples[i] != n_samples_y[i]: raise ValueError("X and y must have the same number of rows.") self.nx = n_features[0] self.nt_all = n_samples self.nlvl = nlevel self.ny = y[0].shape[1] self.X = X[:] self.y = y[:] def _new_train(self): """ Overrides KrgBased implementation Trains the Multi-Fidelity model """ self._new_train_init() theta0 = self.options["theta0"].copy() noise0 = self.options["noise0"].copy() for lvl in range(self.nlvl): self._new_train_iteration(lvl) self.options["theta0"] = theta0 self.options["noise0"] = noise0 self._reinterpolate(lvl) def _new_train_init(self): if self.name in ["MFKPLS", "MFKPLSK"]: _pls = pls(self.options["n_comp"]) # As of sklearn 0.24.1 PLS with zeroed outputs raises an exception while sklearn 0.23 returns zeroed x_rotations # For now the try/except below is a workaround to restore the 0.23 behaviour try: # PLS is done on the highest fidelity identified by the key None self.m_pls = _pls.fit( self.training_points[None][0][0].copy(), self.training_points[None][0][1].copy(), ) self.coeff_pls = self.m_pls.x_rotations_ except StopIteration: self.coeff_pls = np.zeros( self.training_points[None][0][0].shape[1], self.options["n_comp"] ) xt = [] yt = [] i = 0 while self.training_points.get(i, None) is not None: xt.append(self.training_points[i][0][0]) yt.append(self.training_points[i][0][1]) i = i + 1 xt.append(self.training_points[None][0][0]) yt.append(self.training_points[None][0][1]) self._check_list_structure(xt, yt) self._check_param() X = self.X y = self.y _, _, self.X_offset, self.y_mean, self.X_scale, self.y_std = standardization( np.concatenate(xt, axis=0), np.concatenate(yt, axis=0) ) nlevel = self.nlvl # initialize lists self.optimal_noise_all = nlevel * [0] self.D_all = nlevel * [0] self.F_all = nlevel * [0] self.p_all = nlevel * [0] self.q_all = nlevel * [0] self.optimal_rlf_value = nlevel * [0] self.optimal_par = nlevel * [{}] self.optimal_theta = nlevel * [0] self.X_norma_all = [(x - self.X_offset) / self.X_scale for x in X] self.y_norma_all = [(f - self.y_mean) / self.y_std for f in y] def _new_train_iteration(self, lvl): n_samples = self.nt_all self.options["noise0"] = np.array([self.options["noise0"][lvl]]).flatten() self.options["theta0"] = self.options["theta0"][lvl, :] self.X_norma = self.X_norma_all[lvl] self.y_norma = self.y_norma_all[lvl] if self.options["eval_noise"]: if self.options["use_het_noise"]: # hetGP works with unique design variables ( self.X_norma, self.index_unique, # do we need to store it? self.nt_reps, # do we need to store it? ) = np.unique( self.X_norma, return_inverse=True, return_counts=True, axis=0 ) self.nt_all[lvl] = self.X_norma.shape[0] # computing the mean of the output per unique design variable (see Binois et al., 2018) y_norma_unique = [] for i in range(self.nt_all[lvl]): y_norma_unique.append(np.mean(self.y_norma[self.index_unique == i])) y_norma_unique = np.array(y_norma_unique).reshape(-1, 1) # pointwise sensible estimates of the noise variances (see Ankenman et al., 2010) self.optimal_noise = self.options["noise0"] * np.ones(self.nt_all[lvl]) for i in range(self.nt_all[lvl]): diff = self.y_norma[self.index_unique == i] - y_norma_unique[i] if np.sum(diff**2) != 0.0: self.optimal_noise[i] = np.std(diff, ddof=1) ** 2 self.optimal_noise = self.optimal_noise / self.nt_reps self.optimal_noise_all[lvl] = self.optimal_noise self.y_norma = y_norma_unique self.X_norma_all[lvl] = self.X_norma self.y_norma_all[lvl] = self.y_norma else: self.optimal_noise = self.options["noise0"] / self.y_std**2 self.optimal_noise_all[lvl] = self.optimal_noise # Calculate matrix of distances D between samples self.D_all[lvl] = cross_distances(self.X_norma) # Regression matrix and parameters self.F_all[lvl] = self._regression_types[self.options["poly"]](self.X_norma) self.p_all[lvl] = self.F_all[lvl].shape[1] # Concatenate the autoregressive part for levels > 0 if lvl > 0: F_rho = self._regression_types[self.options["rho_regr"]](self.X_norma) self.q_all[lvl] = F_rho.shape[1] self.F_all[lvl] = np.hstack( ( F_rho * np.dot( self._predict_intermediate_values( self.X_norma, lvl, descale=False ), np.ones((1, self.q_all[lvl])), ), self.F_all[lvl], ) ) else: self.q_all[lvl] = 0 n_samples_F_i = self.F_all[lvl].shape[0] if n_samples_F_i != n_samples[lvl]: raise Exception( "Number of rows in F and X do not match. Most " "likely something is going wrong with the " "regression model." ) if int(self.p_all[lvl] + self.q_all[lvl]) >= n_samples_F_i: raise Exception( ( "Ordinary least squares problem is undetermined " "n_samples=%d must be greater than the regression" " model size p+q=%d." ) % (n_samples_F_i, self.p_all[lvl] + self.q_all[lvl]) ) # Determine Gaussian Process model parameters self.F = self.F_all[lvl] D, self.ij = self.D_all[lvl] self._lvl = lvl self.nt = self.nt_all[lvl] self.q = self.q_all[lvl] self.p = self.p_all[lvl] ( self.optimal_rlf_value[lvl], self.optimal_par[lvl], self.optimal_theta[lvl], ) = self._optimize_hyperparam(D) if self.options["eval_noise"] and not self.options["use_het_noise"]: tmp_list = self.optimal_theta[lvl] self.optimal_theta[lvl] = tmp_list[:-1] self.optimal_noise = tmp_list[-1] self.optimal_noise_all[lvl] = self.optimal_noise del self.y_norma, self.D, self.optimal_noise def _reinterpolate(self, lvl): if self.options["eval_noise"] and self.options["optim_var"]: X = self.X for lvl in range(self.nlvl - 1): self.set_training_values( X[lvl], self._predict_intermediate_values(X[lvl], lvl + 1), name=lvl ) self.set_training_values( X[-1], self._predict_intermediate_values(X[-1], self.nlvl) ) self.options["eval_noise"] = False self._new_train() self.options["eval_noise"] = True def _componentwise_distance(self, dx, opt=0): d = componentwise_distance( dx, self.options["corr"], self.nx, power=self.options["pow_exp_power"] ) return d def _predict_intermediate_values(self, X, lvl, descale=True): """ Evaluates the model at a set of points. Used for training the model at level lvl. Allows to relax the order problem. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values lvl : level at which the prediction is made Returns ------- y : np.ndarray Evaluation point output variable values """ n_eval, _ = X.shape # if n_features_X != self.n_features: # raise ValueError("Design must be an array of n_features columns.") # Calculate kriging mean and variance at level 0 mu = np.zeros((n_eval, lvl)) if descale: X = (X - self.X_offset) / self.X_scale f = self._regression_types[self.options["poly"]](X) f0 = self._regression_types[self.options["poly"]](X) dx = self._differences(X, Y=self.X_norma_all[0]) d = self._componentwise_distance(dx) beta = self.optimal_par[0]["beta"] r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[0], d ).reshape(n_eval, self.nt_all[0]) gamma = self.optimal_par[0]["gamma"] # Scaled predictor mu[:, 0] = (np.dot(f, beta) + np.dot(r_, gamma)).ravel() # Calculate recursively kriging mean and variance at level i for i in range(1, lvl): g = self._regression_types[self.options["rho_regr"]](X) dx = self._differences(X, Y=self.X_norma_all[i]) d = self._componentwise_distance(dx) r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[i], d ).reshape(n_eval, self.nt_all[i]) f = np.vstack((g.T * mu[:, i - 1], f0.T)) beta = self.optimal_par[i]["beta"] gamma = self.optimal_par[i]["gamma"] # scaled predictor mu[:, i] = (np.dot(f.T, beta) + np.dot(r_, gamma)).ravel() # scaled predictor if descale: mu = mu * self.y_std + self.y_mean return mu[:, -1].reshape((n_eval, 1)) def _predict_values(self, X, is_acting=None): """ Evaluates the model at a set of points. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values Returns ------- y : np.ndarray Evaluation point output variable values """ return self._predict_intermediate_values(X, self.nlvl) def _predict_variances(self, X: np.ndarray, is_acting=None) -> np.ndarray: """ Evaluates the model at a set of points. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values Returns ------- y : np.ndarray Evaluation point output variable values """ return self.predict_variances_all_levels(X)[0][:, -1] def predict_variances_all_levels(self, X): """ Evaluates the model at a set of points. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values Returns ------- y : np.ndarray Evaluation point output variable values """ # Initialization X = atleast_2d(X) nlevel = self.nlvl sigma2_rhos = [] n_eval, n_features_X = X.shape # if n_features_X != self.n_features: # raise ValueError("Design must be an array of n_features columns.") X = (X - self.X_offset) / self.X_scale # Calculate kriging mean and variance at level 0 mu = np.zeros((n_eval, nlevel)) f = self._regression_types[self.options["poly"]](X) f0 = self._regression_types[self.options["poly"]](X) dx = self._differences(X, Y=self.X_norma_all[0]) d = self._componentwise_distance(dx) # Get regression function and correlation F = self.F_all[0] C = self.optimal_par[0]["C"] beta = self.optimal_par[0]["beta"] Ft = solve_triangular(C, F, lower=True) # yt = solve_triangular(C, self.y_norma_all[0], lower=True) r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[0], d ).reshape(n_eval, self.nt_all[0]) gamma = self.optimal_par[0]["gamma"] # Scaled predictor mu[:, 0] = (np.dot(f, beta) + np.dot(r_, gamma)).ravel() self.sigma2_rho = nlevel * [None] MSE = np.zeros((n_eval, nlevel)) r_t = solve_triangular(C, r_.T, lower=True) G = self.optimal_par[0]["G"] u_ = solve_triangular(G.T, f.T - np.dot(Ft.T, r_t), lower=True) sigma2 = self.optimal_par[0]["sigma2"] / self.y_std**2 MSE[:, 0] = sigma2 * ( # 1 + self.optimal_noise_all[0] - (r_t ** 2).sum(axis=0) + (u_ ** 2).sum(axis=0) 1 - (r_t**2).sum(axis=0) + (u_**2).sum(axis=0) ) # Calculate recursively kriging variance at level i for i in range(1, nlevel): F = self.F_all[i] C = self.optimal_par[i]["C"] g = self._regression_types[self.options["rho_regr"]](X) dx = self._differences(X, Y=self.X_norma_all[i]) d = self._componentwise_distance(dx) r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[i], d ).reshape(n_eval, self.nt_all[i]) f = np.vstack((g.T * mu[:, i - 1], f0.T)) Ft = solve_triangular(C, F, lower=True) yt = solve_triangular(C, self.y_norma_all[i], lower=True) r_t = solve_triangular(C, r_.T, lower=True) G = self.optimal_par[i]["G"] beta = self.optimal_par[i]["beta"] # scaled predictor sigma2 = self.optimal_par[i]["sigma2"] / self.y_std**2 q = self.q_all[i] u_ = solve_triangular(G.T, f - np.dot(Ft.T, r_t), lower=True) sigma2_rho = np.dot( g, sigma2 * linalg.inv(np.dot(G.T, G))[:q, :q] + np.dot(beta[:q], beta[:q].T), ) sigma2_rho = (sigma2_rho * g).sum(axis=1) sigma2_rhos.append(sigma2_rho) if self.name in ["MFKPLS", "MFKPLSK"]: p = self.p_all[i] Q_ = (np.dot((yt - np.dot(Ft, beta)).T, yt - np.dot(Ft, beta)))[0, 0] MSE[:, i] = ( # sigma2_rho * MSE[:, i - 1] +Q_ / (2 * (self.nt_all[i] - p - q)) # * (1 + self.optimal_noise_all[i] - (r_t ** 2).sum(axis=0)) * (1 - (r_t**2).sum(axis=0)) + sigma2 * (u_**2).sum(axis=0) ) else: MSE[:, i] = sigma2 * ( # 1 + self.optimal_noise_all[i] - (r_t ** 2).sum(axis=0) + (u_ ** 2).sum(axis=0) 1 - (r_t**2).sum(axis=0) + (u_**2).sum(axis=0) ) # + sigma2_rho * MSE[:, i - 1] if self.options["propagate_uncertainty"]: MSE[:, i] = MSE[:, i] + sigma2_rho * MSE[:, i - 1] # scaled predictor MSE *= self.y_std**2 return MSE, sigma2_rhos def _predict_derivatives(self, x, kx): """ Evaluates the derivatives at a set of points. Arguments --------- x : np.ndarray [n_evals, dim] Evaluation point input variable values kx : int The 0-based index of the input variable with respect to which derivatives are desired. Returns ------- y : np.ndarray*self.y_std/self.X_scale[kx]) Derivative values. """ lvl = self.nlvl # Initialization n_eval, n_features_x = x.shape x = (x - self.X_offset) / self.X_scale dy_dx = np.zeros((n_eval, lvl)) if self.options["corr"] != "squar_exp": raise ValueError( "The derivative is only available for square exponential kernel" ) if self.options["poly"] == "constant": df = np.zeros([n_eval, 1]) elif self.options["poly"] == "linear": df = np.zeros((n_eval, self.nx + 1)) df[:, 1:] = 1 else: raise ValueError( "The derivative is only available for ordinary kriging or " + "universal kriging using a linear trend" ) df0 = deepcopy(df) if self.options["rho_regr"] != "constant": raise ValueError( "The derivative is only available for regression rho constant" ) # Get pairwise componentwise L1-distances to the input training set dx = self._differences(x, Y=self.X_norma_all[0]) d = self._componentwise_distance(dx) # Compute the correlation function r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[0], d ).reshape(n_eval, self.nt_all[0]) # Beta and gamma = R^-1(y-FBeta) beta = self.optimal_par[0]["beta"] gamma = self.optimal_par[0]["gamma"] df_dx = np.dot(df, beta) d_dx = x[:, kx].reshape((n_eval, 1)) - self.X_norma_all[0][:, kx].reshape( (1, self.nt_all[0]) ) theta = self._get_theta(0) dy_dx[:, 0] = np.ravel((df_dx - 2 * theta[kx] * np.dot(d_dx * r_, gamma))) # Calculate recursively derivative at level i for i in range(1, lvl): g = self._regression_types[self.options["rho_regr"]](x) dx = self._differences(x, Y=self.X_norma_all[i]) d = self._componentwise_distance(dx) r_ = self._correlation_types[self.options["corr"]]( self.optimal_theta[i], d ).reshape(n_eval, self.nt_all[i]) df = np.vstack((g.T * dy_dx[:, i - 1], df0.T)) beta = self.optimal_par[i]["beta"] gamma = self.optimal_par[i]["gamma"] df_dx = np.dot(df.T, beta) d_dx = x[:, kx].reshape((n_eval, 1)) - self.X_norma_all[i][:, kx].reshape( (1, self.nt_all[i]) ) theta = self._get_theta(i) # scaled predictor dy_dx[:, i] = np.ravel(df_dx - 2 * theta[kx] * np.dot(d_dx * r_, gamma)) return dy_dx[:, -1] * self.y_std / self.X_scale[kx] def _get_theta(self, i): return self.optimal_theta[i] def _check_param(self): """ Overrides KrgBased implementation This function checks some parameters of the model. """ if self.name in ["MFKPLS", "MFKPLSK"]: d = self.options["n_comp"] else: d = self.nx if self.options["corr"] == "act_exp": raise ValueError("act_exp correlation function must be used with MGP") if self.name in ["MFKPLS"]: if self.options["corr"] not in ["squar_exp", "abs_exp"]: raise ValueError( "MFKPLS only works with a squared exponential or an absolute exponential kernel" ) elif self.name in ["MFKPLSK"]: if self.options["corr"] not in ["squar_exp"]: raise ValueError( "MFKPLSK only works with a squared exponential kernel (until we prove the contrary)" ) if isinstance(self.options["theta0"], np.ndarray): if self.options["theta0"].shape != (self.nlvl, d): raise ValueError( "the dimensions of theta0 %s should coincide to the number of dim %s" % (self.options["theta0"].shape, (self.nlvl, d)) ) else: if len(self.options["theta0"]) != d: if len(self.options["theta0"]) == 1: self.options["theta0"] *= np.ones((self.nlvl, d)) elif len(self.options["theta0"]) == self.nlvl: self.options["theta0"] = np.array(self.options["theta0"]).reshape( -1, 1 ) self.options["theta0"] *= np.ones((1, d)) else: raise ValueError( "the length of theta0 (%s) should be equal to the number of dim (%s) or levels of fidelity (%s)." % (len(self.options["theta0"]), d, self.nlvl) ) else: self.options["theta0"] *= np.ones((self.nlvl, 1)) if len(self.options["noise0"]) != self.nlvl: if len(self.options["noise0"]) == 1: self.options["noise0"] = self.nlvl * [self.options["noise0"]] else: raise ValueError( "the length of noise0 (%s) should be equal to the number of levels of fidelity (%s)." % (len(self.options["noise0"]), self.nlvl) ) for i in range(self.nlvl): if self.options["use_het_noise"]: if len(self.X[i]) == len(np.unique(self.X[i])): if len(self.options["noise0"][i]) != self.nt_all[i]: if len(self.options["noise0"][i]) == 1: self.options["noise0"][i] *= np.ones(self.nt_all[i]) else: raise ValueError( "for the level of fidelity %s, the length of noise0 (%s) should be equal to the number of observations (%s)." % (i, len(self.options["noise0"][i]), self.nt_all[i]) ) else: if np.size(self.options["noise0"][i]) != 1: raise ValueError( "for the level of fidelity %s, the length of noise0 (%s) should be equal to one." % (i, len(self.options["noise0"][i])) )
27,471
35.776439
141
py
smt
smt-master/smt/applications/application.py
""" Author: Dr. Mohamed Amine Bouhlel <mbouhlel@umich.edu>, R. Lafage This package is distributed under New BSD license. """ from smt.utils.options_dictionary import OptionsDictionary from smt.surrogate_models import LS, QP, KPLS, KRG, KPLSK, GEKPLS, MGP try: from smt.surrogate_models import IDW, RBF, RMTC, RMTB compiled_available = True except: compiled_available = False class SurrogateBasedApplication: if compiled_available: _surrogate_type = { "KRG": KRG, "LS": LS, "QP": QP, "KPLS": KPLS, "KPLSK": KPLSK, "GEKPLS": GEKPLS, "RBF": RBF, "RMTC": RMTC, "RMTB": RMTB, "IDW": IDW, "MGP": MGP, } else: _surrogate_type = { "KRG": KRG, "LS": LS, "QP": QP, "KPLS": KPLS, "KPLSK": KPLSK, "GEKPLS": GEKPLS, "MGP": MGP, } def __init__(self, **kwargs): """ Constructor where values of options can be passed in. For the list of options, see the documentation for the surrogate model being used. Parameters ---------- **kwargs : named arguments Set of options that can be optionally set; each option must have been declared. Examples -------- >>> from smt.applications import VFM >>> extension = VFM(type_bridge = 'Additive', name_model_LF = QP, name_model_bridge = LS, X_LF = xLF, y_LF = yLF, X_HF = xHF, y_HF = yHF, options_LF = dictOptionLFModel, options_bridge = dictOptionBridgeModel) """ self.options = OptionsDictionary() self._initialize() self.options.update(kwargs) def _initialize(self): """ Implemented by the application to declare options and declare what they support (optional). Examples -------- self.options.declare('option_name', default_value, types=(bool, int), desc='description') """ pass
2,127
27
99
py
smt
smt-master/smt/applications/vfm.py
""" Author: Dr. Mohamed Amine Bouhlel <mbouhlel@umich.edu>, R. Lafage This package is distributed under New BSD license. Variable-fidelity modeling: two types of bridge functions are available; i.e., additive and multiplicative """ import numpy as np from smt.utils.options_dictionary import OptionsDictionary from types import FunctionType from smt.applications.application import SurrogateBasedApplication class VFM(SurrogateBasedApplication): def _initialize(self): super(VFM, self)._initialize() declare = self.options.declare declare( "name_model_LF", types=object, values=( "KRG", "LS", "QP", "KPLS", "KPLSK", "GEKPLS", "RBF", "RMTC", "RMTB", "IDW", ), desc="Name of the low-fidelity model", ) declare("options_LF", {}, types=dict, desc="Options for the low-fidelity model") declare( "name_model_bridge", types=object, values=( "KRG", "LS", "QP", "KPLS", "KPLSK", "GEKPLS", "RBF", "RMTC", "RMTB", "IDW", ), desc="Name of the bridge model", ) declare("options_bridge", {}, types=dict, desc="Options for the bridge model") declare( "type_bridge", "Additive", types=str, values=("Additive", "Multiplicative"), desc="Bridge function type", ) declare("X_LF", None, types=np.ndarray, desc="Low-fidelity inputs") declare("y_LF", None, types=np.ndarray, desc="Low-fidelity output") declare("X_HF", None, types=np.ndarray, desc="High-fidelity inputs") declare("y_HF", None, types=np.ndarray, desc="High-fidelity output") declare("dy_LF", None, types=np.ndarray, desc="Low-fidelity derivatives") declare("dy_HF", None, types=np.ndarray, desc="High-fidelity derivatives") self.nx = None self.ny = None self.sm_HF = None self._trained = False def predict_values(self, x): """ Predict the output values at a set of points x. Parameters ---------- x: np.ndarray[n, nx] or np.ndarray[n] Input values for the prediction result analysis. kx : int The 0-based index of the input variable with respect to which derivatives are desired. return ------ y: np.ndarray Output values at the prediction points. """ if not self._trained: self._apply() y = self.sm_HF["predict_values"](x) return y def predict_derivatives(self, x, kx): """ Predict the dy_dx derivatives at a set of points. Parameters ---------- x: np.ndarray[n, nx] or np.ndarray[n] Input values for the prediction result analysis. kx : int The 0-based index of the input variable with respect to which derivatives are desired. return ------ y: np.ndarray Derivatives at the prediction points. """ if not self._trained: self._apply() if kx is None: y = np.zeros(x.shape) for i in range(x.shape[1]): y[:, i] = self.sm_HF["predict_derivatives"][i](x).reshape((x.shape[0])) else: y = self.sm_HF["predict_derivatives"][kx](x).reshape((x.shape[0], self.ny)) return y def _apply(self): """ Algorithm of the VFM method """ # For seek of readability if ( self.options["X_LF"] is not None and self.options["y_LF"] is not None and self.options["X_HF"] is not None and self.options["y_HF"] is not None ): X_LF = self.options["X_LF"] y_LF = self.options["y_LF"] X_HF = self.options["X_HF"] y_HF = self.options["y_HF"] else: raise ValueError("Check X_LF, y_LF, X_HF, and y_FH") self.nx = X_LF.shape[1] self.ny = y_LF.shape[1] if self.options["dy_LF"] is not None: dy_LF = self.options["dy_LF"] if self.options["dy_HF"] is not None: dy_HF = self.options["dy_HF"] # Check parameters self._check_param() # Train the low fidelity model self.LF_deriv = self.options["options_LF"]["deriv"] del self.options["options_LF"]["deriv"] sm_LF = self.options["name_model_LF"](**self.options["options_LF"]) sm_LF.options["print_global"] = False sm_LF.set_training_values(X_LF, y_LF) if self.LF_deriv: for i in range(sm_LF.nx): sm_LF.set_training_derivatives(X_LF, dy_LF[:, i], i) sm_LF.train() # compute the bridge data if self.options["type_bridge"] == "Multiplicative": y_bridge = y_HF / sm_LF.predict_values(X_HF) self.B_deriv = self.options["options_bridge"]["deriv"] del self.options["options_bridge"]["deriv"] if self.B_deriv: dy_bridge = np.zeros(dy_HF.shape) for i in range(sm_LF.nx): dy_bridge[:, i] = ( ( ( dy_HF[:, i].reshape((y_HF.shape[0], 1)) * sm_LF.predict_values(X_HF) ) - (y_HF * sm_LF.predict_derivatives(X_HF, kx=i)) ) / sm_LF.predict_values(X_HF) ** 2 ).reshape(y_HF.shape[0]) elif self.options["type_bridge"] == "Additive": y_bridge = y_HF - sm_LF.predict_values(X_HF) self.B_deriv = self.options["options_bridge"]["deriv"] del self.options["options_bridge"]["deriv"] if self.B_deriv: dy_bridge = np.zeros(dy_HF.shape) for i in range(sm_LF.nx): dy_bridge[:, i] = ( dy_HF[:, i].reshape((y_HF.shape[0], 1)) - sm_LF.predict_derivatives(X_HF, kx=i) ).reshape((y_HF.shape[0])) else: raise ValueError("Only Additive and Multiplicative bridges are available") # Construct of the bridge function sm_bridge = self.options["name_model_bridge"](**self.options["options_bridge"]) sm_bridge.set_training_values(X_HF, y_bridge) if self.B_deriv and self.options["dy_HF"] is not None: for i in range(sm_bridge.nx): sm_bridge.set_training_derivatives(X_HF, dy_bridge[:, i], i) sm_bridge.train() # Construct the final model sm_HF = {} if self.options["type_bridge"] == "Multiplicative": sm_HF["predict_values"] = lambda x: sm_bridge.predict_values( x ) * sm_LF.predict_values(x) if sm_bridge.supports["derivatives"] and sm_LF.supports["derivatives"]: sm_HF["predict_derivatives"] = [] for i in range(sm_LF.nx): sm_HF["predict_derivatives"].append( lambda x, i=i: sm_bridge.predict_derivatives(x, i) * sm_LF.predict_values(x) + sm_bridge.predict_values(x) * sm_LF.predict_derivatives(x, i) ) else: sm_HF["predict_values"] = lambda x: sm_bridge.predict_values( x ) + sm_LF.predict_values(x) if sm_bridge.supports["derivatives"] and sm_LF.supports["derivatives"]: sm_HF["predict_derivatives"] = [] for i in range(sm_LF.nx): sm_HF["predict_derivatives"].append( lambda x, i=i: sm_bridge.predict_derivatives(x, i) + sm_LF.predict_derivatives(x, i) ) self._trained = True self.sm_HF = sm_HF def _check_param(self): """ This function check some parameters of the model. """ # Check surrogates if not callable(self.options["name_model_LF"]): if self.options["name_model_LF"] in self._surrogate_type: self.options["name_model_LF"] = self._surrogate_type[ self.options["name_model_LF"] ] else: raise ValueError( "The LF surrogate should be one of %s, " "%s was given." % (self._surrogate_type.keys(), self.options["name_model_LF"]) ) if not callable(self.options["name_model_bridge"]): if self.options["name_model_bridge"] in self._surrogate_type: self.options["name_model_bridge"] = self._surrogate_type[ self.options["name_model_bridge"] ] else: raise ValueError( "The bridge surrogate should be one of %s, " "%s was given." % (self._surrogate_type.keys(), self.options["name_model_bridge"]) ) # Initialize the parameter deriv if "deriv" not in self.options["options_LF"].keys(): self.options["options_LF"]["deriv"] = False if "deriv" not in self.options["options_bridge"].keys(): self.options["options_bridge"]["deriv"] = False
9,813
34.557971
97
py
smt
smt-master/smt/applications/moe.py
""" Author: Remi Lafage <remi.lafage@onera.fr> This package is distributed under New BSD license. Mixture of Experts """ # TODO : support for best number of clusters # TODO : implement verbosity 'print_global' # TODO : documentation import numpy as np import warnings OLD_SKLEARN = False try: # scikit-learn < 0.20.0 from sklearn.mixture import GMM as GaussianMixture OLD_SKLEARN = True except: from sklearn.mixture import GaussianMixture from scipy.stats import multivariate_normal from smt.utils.options_dictionary import OptionsDictionary from smt.applications.application import SurrogateBasedApplication from smt.utils.misc import compute_rms_error from smt.surrogate_models.surrogate_model import SurrogateModel warnings.filterwarnings("ignore", category=DeprecationWarning) MOE_EXPERT_NAMES = [ "KRG", "KPLS", "KPLSK", "LS", "QP", "RBF", "IDW", "RMTB", "RMTC", ] class MOESurrogateModel(SurrogateModel): """Wrapper class exposing MOE features as a SurrogateModel subclass.""" name = "MOE" def _initialize(self): super(MOESurrogateModel, self)._initialize() # Copy over options from MOE object self.moe = moe = MOE() for key, data in moe.options._declared_entries.items(): self.options._declared_entries[key] = data value = moe.options[key] if value is not None: self.options[key] = value def _setup(self): for key in self.moe.options._declared_entries: if key in self.options: self.moe.options[key] = self.options[key] # self.supports['derivatives'] = self.options['derivatives_support'] # Interface not yet implemented self.supports["variances"] = self.options["variances_support"] def train(self): if len(self.training_points) == 0: xt = self.options["xt"] yt = self.options["yt"] self.set_training_values(xt, yt) super(MOESurrogateModel, self).train() def _train(self): self._setup() for name in self.training_points: xt, yt = self.training_points[name][0] self.moe.set_training_values(xt, yt, name=name) self.moe.train() def _predict_values(self, x: np.ndarray, is_acting=None) -> np.ndarray: return self.moe.predict_values(x) def _predict_variances(self, x: np.ndarray, is_acting=None) -> np.ndarray: return self.moe.predict_variances(x) class MOE(SurrogateBasedApplication): # Names of experts available to be part of the mixture AVAILABLE_EXPERTS = [ name for name in MOE_EXPERT_NAMES if name in SurrogateBasedApplication._surrogate_type ] def _initialize(self): super(MOE, self)._initialize() declare = self.options.declare declare("xt", None, types=np.ndarray, desc="Training inputs") declare("yt", None, types=np.ndarray, desc="Training outputs") declare( "ct", None, types=np.ndarray, desc="Training derivative outputs used for clustering", ) declare("xtest", None, types=np.ndarray, desc="Test inputs") declare("ytest", None, types=np.ndarray, desc="Test outputs") declare( "ctest", None, types=np.ndarray, desc="Derivatives test outputs used for clustering", ) declare("n_clusters", 2, types=int, desc="Number of clusters") declare( "smooth_recombination", True, types=bool, desc="Continuous cluster transition", ) declare( "heaviside_optimization", False, types=bool, desc="Optimize Heaviside scaling factor when smooth recombination is used", ) declare( "derivatives_support", False, types=bool, desc="Use only experts that support derivatives prediction", ) declare( "variances_support", False, types=bool, desc="Use only experts that support variance prediction", ) declare( "allow", [], desc="Names of allowed experts to be possibly part of the mixture. " "Empty list corresponds to all surrogates allowed.", ) declare( "deny", [], desc="Names of forbidden experts", ) self.x = None self.y = None self.c = None self.n_clusters = None self.smooth_recombination = None self.heaviside_optimization = None self.heaviside_factor = 1.0 # dictionary {name: class} of possible experts wrt to options self._enabled_expert_types = self._get_enabled_expert_types() # list of experts after MOE training self._experts = [] self.xt = None self.yt = None @property def enabled_experts(self): """ Returns the names of enabled experts after taking into account MOE options """ self._enabled_expert_types = self._get_enabled_expert_types() return list(self._enabled_expert_types.keys()) def set_training_values(self, xt, yt, name=None): """ Set training data (values). Parameters ---------- xt : np.ndarray[nt, nx] or np.ndarray[nt] The input values for the nt training points. yt : np.ndarray[nt, ny] or np.ndarray[nt] The output values for the nt training points. name : str or None An optional label for the group of training points being set. This is only used in special situations (e.g., multi-fidelity applications). """ self.xt = xt self.yt = yt def train(self): """ Supports for surrogate model API. Build and train the mixture of experts surrogate. """ if self.xt is not None and self.yt is not None: # set_training_values has been called self.x = x = self.xt self.y = y = self.yt else: self.x = x = self.options["xt"] self.y = y = self.options["yt"] self.c = c = self.options["ct"] if not self.c: self.c = c = y self.n_clusters = self.options["n_clusters"] self.smooth_recombination = self.options["smooth_recombination"] self.heaviside_optimization = ( self.options["smooth_recombination"] and self.options["heaviside_optimization"] ) self.heaviside_factor = 1.0 self._check_inputs() self._enabled_expert_types = self._get_enabled_expert_types() self._experts = [] # Set test values and trained values xtest = self.options["xtest"] ytest = self.options["ytest"] ctest = self.options["ctest"] if not ctest: ctest = ytest values = np.c_[x, y, c] self.test_data_present = xtest is not None and ytest is not None if self.test_data_present: self._test_values = np.c_[xtest, ytest, ctest] self._training_values = values else: self._test_values, self._training_values = self._extract_part(values, 10) self.ndim = nx = x.shape[1] xt = self._training_values[:, 0:nx] yt = self._training_values[:, nx : nx + 1] ct = self._training_values[:, nx + 1 :] # Clustering self.cluster = GaussianMixture( n_components=self.n_clusters, covariance_type="full", n_init=20 ) self.cluster.fit(np.c_[xt, ct]) if not self.cluster.converged_: raise Exception("Clustering not converged") # Choice of the experts and training self._fit() xtest = self._test_values[:, 0:nx] ytest = self._test_values[:, nx : nx + 1] # Heaviside factor if self.heaviside_optimization and self.n_clusters > 1: self.heaviside_factor = self._find_best_heaviside_factor(xtest, ytest) print("Best Heaviside factor = {}".format(self.heaviside_factor)) self.distribs = self._create_clusters_distributions(self.heaviside_factor) if not self.test_data_present: # if we have used part of data to validate, fit on overall data self._training_values = values self._fit(new_model=False) def predict_values(self, x): """ Predict the output values at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. Returns ------- y : np.ndarray[nt, ny] Output values at the prediction points. """ if self.smooth_recombination: y = self._predict_smooth_output(x) else: y = self._predict_hard_output(x) return y def predict_variances(self, x): """ Predict the output variances at a set of points. Parameters ---------- x : np.ndarray[nt, nx] or np.ndarray[nt] Input values for the prediction points. Returns ------- y : np.ndarray[nt, ny] Output variances at the prediction points. """ if not self.options["variances_support"]: raise RuntimeError( "Experts not selected taking variance support into account: use variances_support=True " "when creating MOE" ) if self.smooth_recombination: y = self._predict_smooth_output(x, output_variances=True) else: y = self._predict_hard_output(x, output_variances=True) return y def _check_inputs(self): """ Check the input data given by the client is correct. raise Value error with relevant message """ if self.x is None or self.y is None: raise ValueError("check x and y values") if self.x.shape[0] != self.y.shape[0]: raise ValueError( "The number of input points %d doesn t match with the number of output points %d." % (self.x.shape[0], self.y.shape[0]) ) if self.y.shape[0] != self.c.shape[0]: raise ValueError( "The number of output points %d doesn t match with the number of criterion weights %d." % (self.y.shape[0], self.c.shape[0]) ) # choice of number of cluster max_n_clusters = int(len(self.x) / 10) + 1 if self.n_clusters > max_n_clusters: print("Number of clusters should be inferior to {0}".format(max_n_clusters)) raise ValueError( "The number of clusters is too high considering the number of points" ) def _get_enabled_expert_types(self): """ Select relevant surrogate models (experts) regarding MOE feature options """ prototypes = { name: smclass() for name, smclass in self._surrogate_type.items() if name in MOE_EXPERT_NAMES } if self.options["derivatives_support"]: prototypes = { name: proto for name, proto in prototypes.items() if proto.supports["derivatives"] } if self.options["variances_support"]: prototypes = { name: proto for name, proto in prototypes.items() if proto.supports["variances"] } if self.options["allow"]: prototypes = { name: proto for name, proto in prototypes.items() if name in self.options["allow"] } if self.options["deny"]: prototypes = { name: proto for name, proto in prototypes.items() if name not in self.options["deny"] } if not prototypes: raise ValueError( f"List of experts is empty: check support, allow and deny options wrt " f"possible experts: {MOE_EXPERT_NAMES}" ) return {name: self._surrogate_type[name] for name in prototypes} def _fit(self, new_model=True): """ Find the best model for each cluster (clustering already done) and train it if new_model is True otherwise train the points given (choice of best models by cluster already done) Arguments --------- - new_model : bool (optional) Set true to search the best local model """ self.distribs = self._create_clusters_distributions(self.heaviside_factor) nx = self.ndim xt = self._training_values[:, 0:nx] yt = self._training_values[:, nx : nx + 1] ct = self._training_values[:, nx + 1 :] xtest = self._test_values[:, 0:nx] ytest = self._test_values[:, nx : nx + 1] ctest = self._test_values[:, nx + 1 :] # sort trained_values for each cluster cluster_classifier = self.cluster.predict(np.c_[xt, ct]) clustered_values = self._cluster_values(np.c_[xt, yt], cluster_classifier) # sort test_values for each cluster only used in case of new model if new_model: test_cluster_classifier = self.cluster.predict(np.c_[xtest, ctest]) clustered_test_values = self._cluster_values( np.c_[xtest, ytest], test_cluster_classifier ) # find model for each cluster for i in range(self.n_clusters): if new_model: model = self._find_best_model( clustered_values[i], clustered_test_values[i] ) self._experts.append(model) else: # retrain the experts # used when self._training_values changed with expert best models already found training_values = np.array(clustered_values[i]) xtrain = training_values[:, 0 : self.ndim] ytrain = training_values[:, self.ndim] self._experts[i].set_training_values(xtrain, ytrain) self._experts[i].train() def _predict_hard_output(self, x, output_variances=False): """ This method predicts the output of a x samples for a discontinuous recombination. Arguments --------- - x : array_like x samples Return ------ - predicted_values : array_like predicted output """ predicted_values = [] probs = self._proba_cluster(x) sort_cluster = np.apply_along_axis(np.argmax, 1, probs) for i in range(len(sort_cluster)): model = self._experts[sort_cluster[i]] if output_variances: predicted_values.append(model.predict_variances(np.atleast_2d(x[i]))[0]) else: predicted_values.append(model.predict_values(np.atleast_2d(x[i]))[0]) predicted_values = np.array(predicted_values) return predicted_values def _predict_smooth_output(self, x, distribs=None, output_variances=False): """ This method predicts the output of x with a smooth recombination. Arguments: ---------- - x: np.ndarray x samples - distribs: distribution list (optional) array of membership distributions (use self ones if None) Returns ------- - predicted_values : array_like predicted output """ predicted_values = [] if distribs is None: distribs = self.distribs sort_proba = self._proba_cluster(x, distribs) for i in range(len(sort_proba)): recombined_value = 0 for j in range(len(self._experts)): if output_variances: expert_value = ( self._experts[j].predict_variances(np.atleast_2d(x[i]))[0] * sort_proba[i][j] ** 2 ) else: expert_value = ( self._experts[j].predict_values(np.atleast_2d(x[i]))[0] * sort_proba[i][j] ) recombined_value += expert_value predicted_values.append(recombined_value) predicted_values = np.array(predicted_values) return predicted_values @staticmethod def _extract_part(values, quantile): """ Divide the values list in quantile parts to return one part of (num/quantile) values out of num values. Arguments ---------- - values : np.ndarray[num, -1] the values list to extract from - quantile : int the quantile Returns ------- - extracted, remaining : np.ndarray, np.ndarray the extracted values part, the remaining values """ num = values.shape[0] indices = np.arange(0, num, quantile) # uniformly distributed mask = np.zeros(num, dtype=bool) mask[indices] = True return values[mask], values[~mask] def _find_best_model(self, clustered_values, clustered_test_values): """ Find the best model which minimizes the errors. Arguments : ------------ - clustered_values: array_like training samples [[X1,X2, ..., Xn, Y], ... ] Returns : --------- - model : surrogate model best trained surrogate model """ dim = self.ndim scores = {} sms = {} training_values = np.array(clustered_values) test_values = np.array(clustered_test_values) for name, sm_class in self._enabled_expert_types.items(): kwargs = {} if name in ["RMTB", "RMTC"]: # Note: RMTS checks for xlimits, # we take limits on all x (not just the trained_values ones) as # the surrogate is finally re-trained on the whole x set. xlimits = np.zeros((dim, 2)) for i in range(dim): xlimits[i][0] = np.amin(self.x[:, i]) xlimits[i][1] = np.amax(self.x[:, i]) kwargs = {"xlimits": xlimits} sm = sm_class(**kwargs) sm.options["print_global"] = False sm.set_training_values(training_values[:, 0:dim], training_values[:, dim]) sm.train() expected = test_values[:, dim] actual = sm.predict_values(test_values[:, 0:dim]).reshape(-1) l_two = np.linalg.norm(expected - actual, 2) # l_two_rel = l_two / np.linalg.norm(expected, 2) # mse = (l_two**2) / len(expected) # rmse = mse ** 0.5 scores[sm.name] = l_two print(sm.name, l_two) sms[sm.name] = sm best_name = None best_score = None for name, rmse in scores.items(): if best_score is None or rmse < best_score: best_name, best_score = name, rmse print("Best expert = {}".format(best_name)) return sms[best_name] def _find_best_heaviside_factor(self, x, y): """ Find the best heaviside factor to smooth approximated values. Arguments --------- - x: array_like input training samples - y: array_like output training samples Returns ------- hfactor : float best heaviside factor wrt given samples """ heaviside_factor = 1.0 if self.n_clusters > 1: hfactors = np.linspace(0.1, 2.1, num=21) errors = [] for hfactor in hfactors: distribs = self._create_clusters_distributions(hfactor) ypred = self._predict_smooth_output(x, distribs) err_rel = np.linalg.norm(y - ypred, 2) / np.linalg.norm(y, 2) errors.append(err_rel) if max(errors) < 1e-6: heaviside_factor = 1.0 else: min_error_index = errors.index(min(errors)) heaviside_factor = hfactors[min_error_index] return heaviside_factor """ Functions related to clustering """ def _create_clusters_distributions(self, heaviside_factor=1.0): """ Create an array of frozen multivariate normal distributions (distribs). Arguments --------- - heaviside_factor: float Heaviside factor used to scale covariance matrices Returns: -------- - distribs: array_like Array of frozen multivariate normal distributions with clusters means and covariances """ distribs = [] dim = self.ndim means = self.cluster.means_ if OLD_SKLEARN: cov = heaviside_factor * self.cluster.covars_ else: cov = heaviside_factor * self.cluster.covariances_ for k in range(self.n_clusters): meansk = means[k][0:dim] covk = cov[k][0:dim, 0:dim] mvn = multivariate_normal(meansk, covk, allow_singular=True) distribs.append(mvn) return distribs def _cluster_values(self, values, classifier): """ Classify values regarding the given classifier info. Arguments --------- - values: array_like values to cluster - classifier: array_like Cluster corresponding to each point of value in the same order Returns ------- - clustered: array_like Samples sort by cluster Example: --------- values: [[ 1.67016597e-01 5.42927264e-01 9.25779645e+00] [ 5.20618344e-01 9.88223010e-01 1.51596837e+02] [ 6.09979830e-02 2.66824984e-01 1.17890707e+02] [ 9.62783472e-01 7.36979149e-01 7.37641826e+01] [ 3.01194132e-01 8.58084068e-02 4.88696602e+01] [ 6.40398203e-01 6.91090937e-01 8.91963162e+01] [ 7.90710374e-01 1.40464471e-01 1.89390766e+01] [ 4.64498124e-01 3.61009635e-01 1.04779656e+01]] cluster_classifier: [1 0 0 2 1 2 1 1] clustered: [[array([ 0.52061834, 0.98822301, 151.59683723]), array([ 6.09979830e-02, 2.66824984e-01, 1.17890707e+02])] [array([ 0.1670166 , 0.54292726, 9.25779645]), array([ 0.30119413, 0.08580841, 48.86966023]), array([ 0.79071037, 0.14046447, 18.93907662]), array([ 0.46449812, 0.36100964, 10.47796563])] [array([ 0.96278347, 0.73697915, 73.76418261]), array([ 0.6403982 , 0.69109094, 89.19631619])]] """ num = len(classifier) assert values.shape[0] == num clusters = [[] for n in range(self.n_clusters)] for i in range(num): clusters[classifier[i]].append(values[i]) return clusters def _proba_cluster_one_sample(self, x, distribs): """ Compute membership probabilities to each cluster for one sample. Arguments --------- - x: array_like a sample for which probabilities must be calculated - distribs: multivariate_normal objects list array of normal distributions Returns ------- - prob: array_like x membership probability for each cluster """ weights = np.array(self.cluster.weights_) rvs = np.array([distribs[k].pdf(x) for k in range(len(weights))]) probs = weights * rvs rad = np.sum(probs) if rad > 0: probs = probs / rad return probs def _proba_cluster(self, x, distribs=None): """ Calculate membership probabilities to each cluster for each sample Arguments --------- - x: array_like samples where probabilities must be calculated - distribs : multivariate_normal objects list (optional) array of membership distributions. If None, use self ones. Returns ------- - probs: array_like x membership probabilities to each cluster. Examples : ---------- x: [[ 0. 0.] [ 0. 1.] [ 1. 0.] [ 1. 1.]] prob: [[ 1.49050563e-02 9.85094944e-01] [ 9.90381299e-01 9.61870088e-03] [ 9.99208990e-01 7.91009759e-04] [ 1.48949963e-03 9.98510500e-01]] """ if distribs is None: distribs = self.distribs if self.n_clusters == 1: probs = np.ones((x.shape[0], 1)) else: probs = np.array( [self._proba_cluster_one_sample(x[i], distribs) for i in range(len(x))] ) return probs
25,371
31.695876
109
py
smt
smt-master/smt/applications/mfkplsk.py
# -*- coding: utf-8 -*- """ Created on Fri May 04 10:26:49 2018 @author: Mostafa Meliani <melimostafa@gmail.com> Multi-Fidelity co-Kriging: recursive formulation with autoregressive model of order 1 (AR1) Partial Least Square decomposition added on highest fidelity level KPLSK model combined PLS followed by a Krging model in the initial dimension Adapted on March 2020 by Nathalie Bartoli to the new SMT version Adapted on January 2021 by Andres Lopez-Lopera to the new SMT version """ from smt.utils.kriging import componentwise_distance from smt.applications import MFKPLS class MFKPLSK(MFKPLS): def _initialize(self): super(MFKPLSK, self)._initialize() declare = self.options.declare # Like KPLSK, MFKPLSK used only with "squar_exp" correlations declare( "corr", "squar_exp", values=("squar_exp"), desc="Correlation function type", types=(str), ) self.name = "MFKPLSK" def _componentwise_distance(self, dx, opt=0): # Modif for KPLSK model if opt == 0: # Kriging step d = componentwise_distance( dx, self.options["corr"], self.nx, power=self.options["pow_exp_power"] ) else: # KPLS step d = super(MFKPLSK, self)._componentwise_distance(dx, opt) return d def _new_train(self): """ Overrides KrgBased implementation Trains the Multi-Fidelity model + PLS (done on the highest fidelity level) + Kriging (MFKPLSK) """ self._new_train_init() self.n_comp = self.options["n_comp"] theta0 = self.options["theta0"].copy() noise0 = self.options["noise0"].copy() for lvl in range(self.nlvl): self._new_train_iteration(lvl) self.options["n_comp"] = self.n_comp self.options["theta0"] = theta0 self.options["noise0"] = noise0 self._reinterpolate(lvl) def _get_theta(self, i): return self.optimal_theta[i]
2,074
31.421875
103
py
smt
smt-master/smt/applications/__init__.py
from .vfm import VFM from .moe import MOE, MOESurrogateModel from .mfk import MFK, NestedLHS from .mfkpls import MFKPLS from .mfkplsk import MFKPLSK from .ego import EGO, Evaluator
181
25
39
py
smt
smt-master/smt/applications/mixed_integer.py
""" Author: Remi Lafage <remi.lafage@onera.fr> This package is distributed under New BSD license. """ import numpy as np from smt.surrogate_models.surrogate_model import SurrogateModel from smt.sampling_methods.sampling_method import SamplingMethod from smt.utils.checks import ensure_2d_array from smt.surrogate_models.krg_based import KrgBased, MixIntKernelType from smt.utils.design_space import ( BaseDesignSpace, CategoricalVariable, ensure_design_space, ) import warnings class MixedIntegerSamplingMethod(SamplingMethod): """ Sampling method decorator that takes an SMT continuous sampling method and cast values according x types specification to implement a sampling method handling integer (ORD) or categorical (ENUM) features """ def __init__(self, sampling_method_class, design_space, **kwargs): """ Parameters ---------- sampling_method_class: class name SMT sampling method class design_space: BaseDesignSpace design space definition kwargs: options of the given sampling method options used to instanciate the SMT sampling method with the additional 'output_in_folded_space' boolean option specifying if doe output should be in folded space (enum indexes) or not (enum masks) """ warnings.warn( "MixedIntegerSamplingMethod has been deprecated, use DesignSpace.sample_valid_x instead!", category=DeprecationWarning, ) self._design_space = design_space self._unfolded_xlimits = design_space.get_unfolded_num_bounds() self._output_in_folded_space = kwargs.get("output_in_folded_space", True) kwargs.pop("output_in_folded_space", None) self._sampling_method = sampling_method_class( xlimits=self._unfolded_xlimits, **kwargs ) super().__init__() def _compute(self, nt): doe = self._sampling_method(nt) x_doe, _ = self._design_space.correct_get_acting(doe) if self._output_in_folded_space: x_doe, _ = self._design_space.fold_x(x_doe) return x_doe def __call__(self, nt): return self._compute(nt) def expand_lhs(self, x, nt, method="basic"): doe = self._sampling_method(nt) x_doe, _ = self._design_space.correct_get_acting(doe) if self._output_in_folded_space: x_doe, _ = self._design_space.fold_x(x_doe) return x_doe class MixedIntegerSurrogateModel(SurrogateModel): """ Surrogate model (not Kriging) decorator that takes an SMT continuous surrogate model and cast values according x types specification to implement a surrogate model handling integer (ORD) or categorical (ENUM) features """ def __init__( self, design_space, surrogate, input_in_folded_space=True, ): """ Parameters ---------- design_space: BaseDesignSpace design space definition surrogate: SMT surrogate model (not Kriging) instance of a SMT surrogate model input_in_folded_space: bool whether x data are in given in folded space (enum indexes) or not (enum masks) categorical_kernel: string the kernel to use for categorical inputs. Only for non continuous Kriging. """ super().__init__() self._surrogate = surrogate if isinstance(self._surrogate, KrgBased): raise ValueError( "Using MixedIntegerSurrogateModel integer model with " + str(self._surrogate.name) + " is not supported. Please use MixedIntegerKrigingModel instead." ) self.design_space = ensure_design_space(design_space=design_space) self._input_in_folded_space = input_in_folded_space self.supports = self._surrogate.supports self.options["print_global"] = False if "poly" in self._surrogate.options: if self._surrogate.options["poly"] != "constant": raise ValueError("constant regression must be used with mixed integer") @property def name(self): return "MixedInteger" + self._surrogate.name def _initialize(self): self.supports["derivatives"] = False def set_training_values(self, xt, yt, name=None) -> None: xt = ensure_2d_array(xt, "xt") # Round inputs design_space = self.design_space xt, _ = design_space.correct_get_acting(xt) if self._input_in_folded_space: xt_apply, _ = design_space.unfold_x(xt) else: xt_apply = xt super().set_training_values(xt_apply, yt) self._surrogate.set_training_values(xt_apply, yt, name) def update_training_values(self, yt, name=None): super().update_training_values(yt, name) self._surrogate.update_training_values(yt, name) def _train(self): self._surrogate._train() def predict_values(self, x: np.ndarray) -> np.ndarray: x_corr, is_acting = self._get_x_for_surrogate_model(x) return self._surrogate.predict_values(x_corr) def predict_variances(self, x: np.ndarray) -> np.ndarray: x_corr, is_acting = self._get_x_for_surrogate_model(x) return self._surrogate.predict_variances(x_corr) def _get_x_for_surrogate_model(self, x): xp = ensure_2d_array(x, "xp") x_corr, is_acting = self.design_space.correct_get_acting(xp) if self._input_in_folded_space: x_corr, is_acting = self.design_space.unfold_x(x_corr, is_acting=is_acting) return x_corr, is_acting def _predict_values(self, x: np.ndarray) -> np.ndarray: pass class MixedIntegerKrigingModel(KrgBased): """ Kriging model decorator that takes an SMT continuous surrogate model and cast values according x types specification to implement a surrogate model handling integer (ORD) or categorical (ENUM) features """ def __init__( self, surrogate, input_in_folded_space=True, ): """ Parameters ---------- surrogate: SMT Kriging surrogate model instance of a SMT Kriging surrogate model """ super().__init__() self._surrogate = surrogate if not (isinstance(self._surrogate, KrgBased)): raise ValueError( "Using MixedIntegerKrigingModel integer model with " + str(self._surrogate.name) + " is not supported. Please use MixedIntegerSurrogateModel instead." ) self.options["design_space"] = self._surrogate.design_space self._input_in_folded_space = input_in_folded_space self.supports = self._surrogate.supports self.options["print_global"] = False if "poly" in self._surrogate.options: if self._surrogate.options["poly"] != "constant": raise ValueError("constant regression must be used with mixed integer") design_space = self.design_space if ( any( isinstance(dv, CategoricalVariable) for dv in design_space.design_variables ) and self._surrogate.options["categorical_kernel"] is None ): self._surrogate.options[ "categorical_kernel" ] = MixIntKernelType.HOMO_HSPHERE warnings.warn( "Using MixedIntegerSurrogateModel integer model with Continuous Relaxation is not supported. Switched to homoscedastic hypersphere kernel instead." ) if self._surrogate.options["categorical_kernel"] is not None: self._input_in_folded_space = False @property def name(self): return "MixedInteger" + self._surrogate.name def _initialize(self): super()._initialize() self.supports["derivatives"] = False def set_training_values(self, xt, yt, name=None, is_acting=None): xt = ensure_2d_array(xt, "xt") # If the is_acting matrix is not given, assume input is not corrected (rounding, imputation, etc.) yet design_space = self.design_space if is_acting is None: xt, is_acting = design_space.correct_get_acting(xt) if self._input_in_folded_space: xt_apply, is_acting_apply = design_space.unfold_x(xt, is_acting) else: xt_apply, is_acting_apply = xt, is_acting super().set_training_values(xt_apply, yt, is_acting=is_acting_apply) self._surrogate.set_training_values( xt_apply, yt, name, is_acting=is_acting_apply ) def update_training_values(self, yt, name=None): super().update_training_values(yt, name) self._surrogate.update_training_values(yt, name) def _train(self): self._surrogate._train() def predict_values(self, x: np.ndarray, is_acting=None) -> np.ndarray: x_corr, is_acting = self._get_x_for_surrogate_model(x) return self._surrogate.predict_values(x_corr, is_acting=is_acting) def predict_variances(self, x: np.ndarray, is_acting=None) -> np.ndarray: x_corr, is_acting = self._get_x_for_surrogate_model(x) return self._surrogate.predict_variances(x_corr, is_acting=is_acting) def _get_x_for_surrogate_model(self, x): xp = ensure_2d_array(x, "xp") x_corr, is_acting = self.design_space.correct_get_acting(xp) if self._input_in_folded_space: x_corr, is_acting = self.design_space.unfold_x(x_corr, is_acting=is_acting) return x_corr, is_acting def _predict_values(self, x: np.ndarray, is_acting=None) -> np.ndarray: pass class MixedIntegerContext(object): """ Class which acts as sampling method and surrogate model factory to handle integer and categorical variables consistently. """ def __init__(self, design_space, work_in_folded_space=True): """ Parameters ---------- design_space: BaseDesignSpace the design space definition (includes mixed-discrete and/or hierarchical specifications) work_in_folded_space: bool whether x data are in given in folded space (enum indexes) or not (enum masks) """ self._design_space = ensure_design_space(design_space=design_space) self._unfold_space = not work_in_folded_space self._unfolded_xlimits = self._design_space.get_unfolded_num_bounds() self._work_in_folded_space = work_in_folded_space @property def design_space(self) -> BaseDesignSpace: return self._design_space def build_sampling_method(self, *_, **__): """ Build MixedIntegerSamplingMethod from given SMT sampling method. """ return_folded = self._work_in_folded_space def sample(n): x, _ = self._design_space.sample_valid_x(n, unfolded=not return_folded) return x return sample def build_kriging_model(self, surrogate): """ Build MixedIntegerKrigingModel from given SMT surrogate model. """ surrogate.options["design_space"] = self._design_space return MixedIntegerKrigingModel( surrogate=surrogate, input_in_folded_space=self._work_in_folded_space, ) def build_surrogate_model(self, surrogate): """ Build MixedIntegerKrigingModel from given SMT surrogate model. """ return MixedIntegerSurrogateModel( self._design_space, surrogate=surrogate, input_in_folded_space=self._work_in_folded_space, ) def get_unfolded_xlimits(self): """ Returns relaxed xlimits Each level of an enumerate gives a new continuous dimension in [0, 1]. Each integer dimensions are relaxed continuously. """ return self._unfolded_xlimits def get_unfolded_dimension(self): """ Returns x dimension (int) taking into account unfolded categorical features """ return len(self._unfolded_xlimits)
12,226
34.44058
163
py
smt
smt-master/smt/applications/mfkpls.py
# -*- coding: utf-8 -*- """ Created on Fri May 04 10:26:49 2018 @author: Mostafa Meliani <melimostafa@gmail.com> Multi-Fidelity co-Kriging: recursive formulation with autoregressive model of order 1 (AR1) Partial Least Square decomposition added on highest fidelity level Adapted on March 2020 by Nathalie Bartoli to the new SMT version Adapted on January 2021 by Andres Lopez-Lopera to the new SMT version """ import numpy as np from sklearn.cross_decomposition import PLSRegression as pls from sklearn.metrics.pairwise import manhattan_distances from smt.applications import MFK from smt.utils.kriging import componentwise_distance_PLS class MFKPLS(MFK): """ Multi-Fidelity model + PLS (done on the highest fidelity level) """ def _initialize(self): super()._initialize() declare = self.options.declare # Like KPLS, MFKPLS used only with "abs_exp" and "squar_exp" correlations declare( "corr", "squar_exp", values=("abs_exp", "squar_exp"), desc="Correlation function type", types=(str), ) declare("n_comp", 1, types=int, desc="Number of principal components") self.name = "MFKPLS" def _differences(self, X, Y): """ Overrides differences function for MFK Compute the manhattan_distances """ return manhattan_distances(X, Y, sum_over_features=False) def _componentwise_distance(self, dx, opt=0): d = componentwise_distance_PLS( dx, self.options["corr"], self.options["n_comp"], self.coeff_pls, power=self.options["pow_exp_power"], ) return d def _compute_pls(self, X, y): _pls = pls(self.options["n_comp"]) # As of sklearn 0.24.1 PLS with zeroed outputs raises an exception while sklearn 0.23 returns zeroed x_rotations # For now the try/except below is a workaround to restore the 0.23 behaviour try: self.coeff_pls = _pls.fit(X.copy(), y.copy()).x_rotations_ except StopIteration: self.coeff_pls = np.zeros((X.shape[1], self.options["n_comp"])) return X, y def _get_theta(self, i): return np.sum(self.optimal_theta[i] * self.coeff_pls**2, axis=1)
2,313
32.057143
120
py
smt
smt-master/smt/applications/ego.py
""" Authors: Nathalie Bartoli, Remy Priem, Remi Lafage, Emile Roux <emile.roux@univ-smb.fr> This package is distributed under New BSD license. """ import numpy as np from types import FunctionType from scipy.stats import norm from scipy.optimize import minimize from smt.surrogate_models import KPLS, KRG, KPLSK, MGP, GEKPLS from smt.applications.application import SurrogateBasedApplication from smt.applications.mixed_integer import MixedIntegerContext from smt.utils.design_space import ( BaseDesignSpace, DesignSpace, FloatVariable, CategoricalVariable, ) class Evaluator(object): """ An interface for evaluation of a function at x points (nsamples of dimension nx). User can derive this interface and override the run() method to implement custom multiprocessing. """ def run(self, fun, x): """ Evaluates fun at x. Parameters --------- fun : function to evaluate: (nsamples, nx) -> (nsample, 1) x : np.ndarray[nsamples, nx] nsamples points of nx dimensions. Returns ------- np.ndarray[nsample, 1] fun evaluations at the nsamples points. """ return fun(x) class EGO(SurrogateBasedApplication): def _initialize(self): super(EGO, self)._initialize() declare = self.options.declare declare("fun", None, types=FunctionType, desc="Function to minimize") declare( "criterion", "EI", types=str, values=["EI", "SBO", "LCB"], desc="criterion for next evaluation point determination: Expected Improvement, \ Surrogate-Based Optimization or Lower Confidence Bound", ) declare("n_iter", None, types=int, desc="Number of optimizer steps") declare( "n_max_optim", 20, types=int, desc="Maximum number of internal optimizations", ) declare("n_start", 20, types=int, desc="Number of optimization start points") declare( "n_parallel", 1, types=int, desc="Number of parallel samples to compute using qEI criterion", ) declare( "qEI", "KBLB", types=str, values=["KB", "KBLB", "KBUB", "KBRand", "CLmin"], desc="Approximated q-EI maximization strategy", ) declare( "evaluator", default=Evaluator(), types=Evaluator, desc="Object used to run function fun to optimize at x points (nsamples, nxdim)", ) declare( "n_doe", None, types=int, desc="Number of points of the initial LHS doe, only used if xdoe is not given", ) declare("xdoe", None, types=np.ndarray, desc="Initial doe inputs") declare("ydoe", None, types=np.ndarray, desc="Initial doe outputs") declare("verbose", False, types=bool, desc="Print computation information") declare( "enable_tunneling", False, types=bool, desc="Enable the penalization of points that have been already evaluated in EI criterion", ) declare( "surrogate", KRG(print_global=False), types=(KRG, KPLS, KPLSK, GEKPLS, MGP), desc="SMT kriging-based surrogate model used internaly", ) self.options.declare( "random_state", types=(type(None), int, np.random.RandomState), desc="Numpy RandomState object or seed number which controls random draws", ) def optimize(self, fun): """ Optimizes fun Parameters ---------- fun: function to optimize: ndarray[n, nx] or ndarray[n] -> ndarray[n, 1] Returns ------- [nx, 1]: x optimum [1, 1]: y optimum int: index of optimum in data arrays [ndoe + n_iter, nx]: coord-x data [ndoe + n_iter, 1]: coord-y data """ x_data, y_data = self._setup_optimizer(fun) n_iter = self.options["n_iter"] n_parallel = self.options["n_parallel"] for k in range(n_iter): # Virtual enrichement loop for p in range(n_parallel): # find next best x-coord point to evaluate x_et_k, success = self._find_best_point( x_data, y_data, self.options["enable_tunneling"] ) if not success: self.log( "Internal optimization failed at EGO iter = {}.{}".format(k, p) ) break elif success: self.log( "Internal optimization succeeded at EGO iter = {}.{}".format( k, p ) ) # Set temporaly the y-coord point based on the kriging prediction x_et_k = np.atleast_2d(x_et_k) if self.mixint: x_et_k, _ = self.design_space.correct_get_acting(x_et_k) y_et_k = self._get_virtual_point(x_et_k, y_data) # Update y_data with predicted value y_data = y_data.reshape(y_data.shape[0], self.gpr.ny) y_data = np.vstack((y_data, y_et_k)) x_data = np.atleast_2d(np.append(x_data, x_et_k, axis=0)) # Compute the real values of y_data x_to_compute = np.atleast_2d(x_data[-n_parallel:]) y = self._evaluator.run(fun, x_to_compute) y_data[-n_parallel:] = y # Find the optimal point ind_best = np.argmin(y_data if y_data.ndim == 1 else y_data[:, 0]) x_opt = x_data[ind_best] y_opt = y_data[ind_best] return x_opt, y_opt, ind_best, x_data, y_data def log(self, msg): if self.options["verbose"]: print(msg) def EI(self, points, enable_tunneling=False, x_data=None): """Expected improvement""" y_data = np.atleast_2d(self.gpr.training_points[None][0][1]) f_min = y_data[np.argmin(y_data[:, 0])] pred = self.gpr.predict_values(points) sig = np.sqrt(self.gpr.predict_variances(points)) args0 = (f_min - pred) / sig args1 = (f_min - pred) * norm.cdf(args0) args2 = sig * norm.pdf(args0) if sig.size == 1 and sig == 0.0: # can be use only if one point is computed return 0.0 ei = args1 + args2 # penalize the points already evaluated with tunneling if enable_tunneling: for i in range(len(points)): p = np.atleast_2d(points[i]) EIp = self.EI(p, enable_tunneling=False) for x in x_data: x = np.atleast_2d(x) # if np.abs(p-x)<1: # ei[i]=ei[i]*np.reciprocal(1+100*np.exp(-np.reciprocal(1-np.square(p-x)))) pena = (EIp - self.EI(x, enable_tunneling=False)) / ( 1e-9 + np.power(np.linalg.norm(p - x), 4) ) if pena > 0: ei[i] = ei[i] - pena ei[i] = max(ei[i], 0) return ei def SBO(self, point): """Surrogate based optimization: min the surrogate model by suing the mean mu""" res = self.gpr.predict_values(point) return res def LCB(self, point): """Lower confidence bound optimization: minimize by using mu - 3*sigma""" pred = self.gpr.predict_values(point) var = self.gpr.predict_variances(point) res = pred - 3.0 * np.sqrt(var) return res def _setup_optimizer(self, fun): """ Instanciate internal surrogate used for optimization and setup function evaluator wrt options Parameters ---------- fun: function to optimize: ndarray[n, nx] or ndarray[n] -> ndarray[n, 1] Returns ------- ndarray: initial coord-x doe ndarray: initial coord-y doe = fun(xdoe) """ # Set the model self.gpr = self.options["surrogate"] self.design_space: BaseDesignSpace = self.gpr.design_space if isinstance(self.design_space, DesignSpace): self.design_space.seed = self.options["random_state"] # Handle mixed integer optimization is_continuous = self.design_space.is_all_cont if not is_continuous: self.categorical_kernel = self.gpr.options["categorical_kernel"] self.mixint = MixedIntegerContext( self.design_space, work_in_folded_space=True, ) underlying_gpr = self.gpr self.gpr = self.mixint.build_kriging_model(self.gpr) self.categorical_kernel = underlying_gpr.options["categorical_kernel"] self.mixint = MixedIntegerContext( self.design_space, work_in_folded_space=True, ) self._sampling = self.mixint.build_sampling_method() else: self.mixint = None self._sampling = lambda n: self.design_space.sample_valid_x(n)[0] self.categorical_kernel = None # Build DOE self._evaluator = self.options["evaluator"] xdoe = self.options["xdoe"] if xdoe is None: self.log("Build initial DOE with LHS") n_doe = self.options["n_doe"] x_doe = self._sampling(n_doe) else: self.log("Initial DOE given") x_doe = np.atleast_2d(xdoe) ydoe = self.options["ydoe"] if ydoe is None: y_doe = self._evaluator.run(fun, x_doe) else: # to save time if y_doe is already given to EGO y_doe = ydoe return x_doe, y_doe def _train_gpr(self, x_data, y_data): self.gpr.set_training_values(x_data, y_data) if self.gpr.supports["training_derivatives"]: for kx in range(self.gpr.nx): self.gpr.set_training_derivatives( x_data, y_data[:, 1 + kx].reshape((y_data.shape[0], 1)), kx ) self.gpr.train() def _find_best_point(self, x_data=None, y_data=None, enable_tunneling=False): """ Function that analyse a set of x_data and y_data and give back the more interesting point to evaluates according to the selected criterion Parameters ---------- x_data: ndarray(n_points, nx) y_data: ndarray(n_points, 1) Returns ------- ndarray(nx, 1): the next best point to evaluate boolean: success flag """ self._train_gpr(x_data, y_data) criterion = self.options["criterion"] n_start = self.options["n_start"] n_max_optim = self.options["n_max_optim"] method = "SLSQP" options = {"maxiter": 200} if self.mixint: bounds = self.design_space.get_num_bounds() cons = [] for j in range(len(bounds)): lower, upper = bounds[j] l = {"type": "ineq", "fun": lambda x, lb=lower, i=j: x[i] - lb} u = {"type": "ineq", "fun": lambda x, ub=upper, i=j: ub - x[i]} cons.append(l) cons.append(u) bounds = None options = {"maxiter": 300} else: bounds = self.design_space.get_num_bounds() cons = () if criterion == "EI": self.obj_k = lambda x: -self.EI(np.atleast_2d(x), enable_tunneling, x_data) elif criterion == "SBO": self.obj_k = lambda x: self.SBO(np.atleast_2d(x)) elif criterion == "LCB": self.obj_k = lambda x: self.LCB(np.atleast_2d(x)) success = False n_optim = 1 # in order to have some success optimizations with SLSQP while not success and n_optim <= n_max_optim: opt_all = [] x_start = self._sampling(n_start) for ii in range(n_start): try: opt_all.append( minimize( lambda x: float(np.array(self.obj_k(x)).flat[0]), x_start[ii, :], method=method, bounds=bounds, constraints=cons, options=options, ) ) except ValueError: # in case "x0 violates bound constraints" error print("warning: `x0` violates bound constraints") print("x0={}".format(x_start[ii, :])) print("bounds={}".format(bounds)) opt_all.append({"success": False}) opt_all = np.asarray(opt_all) for opt_i in opt_all: if ( opt_i["message"] == "Maximum number of function evaluations has been exceeded." ): opt_i["success"] = True opt_success = opt_all[[opt_i["success"] for opt_i in opt_all]] obj_success = np.array([opt_i["fun"] for opt_i in opt_success]) success = obj_success.size != 0 if not success: self.log("New start point for the internal optimization") n_optim += 1 if n_optim >= n_max_optim: # self.log("Internal optimization failed at EGO iter = {}".format(k)) return np.atleast_2d(0), False ind_min = np.argmin(obj_success) opt = opt_success[ind_min] x_et_k = np.atleast_2d(opt["x"]) return x_et_k, True def _get_virtual_point(self, x, y_data): """ Depending on the qEI attribute return a predicted value at given point x Parameters ---------- x: ndarray(1, 1) the x-coord point where to forecast the y-coord virtual point y_data: current y evaluation list only used when qEI is CLmin Returns ------- ndarray(1, 1): the so-called virtual y-coord point """ qEI = self.options["qEI"] if qEI == "CLmin": return np.min(y_data) if qEI == "KB": return self.gpr.predict_values(x) if qEI == "KBUB": conf = 3.0 if qEI == "KBLB": conf = -3.0 if qEI == "KBRand": conf = np.random.randn() pred = self.gpr.predict_values(x) var = self.gpr.predict_variances(x) return pred + conf * np.sqrt(var)
14,811
33.446512
102
py
smt
smt-master/smt/applications/tests/test_mfkpls.py
# -*- coding: utf-8 -*- """ Created on Mon May 07 14:20:11 2018 @author: m.meliani Adapted to new SMT version in march 2020 by Nathalie Bartoli """ import matplotlib matplotlib.use("Agg") import unittest import numpy as np import unittest import inspect from collections import OrderedDict from smt.problems import Sphere, TensorProduct from smt.sampling_methods import LHS, FullFactorial from smt.utils.sm_test_case import SMTestCase from smt.utils.silence import Silence from smt.utils import compute_rms_error from smt.surrogate_models import LS, QP, KPLS, KRG, KPLSK, GEKPLS, GENN from smt.applications.mfk import MFK, NestedLHS from smt.applications.mfkpls import MFKPLS from copy import deepcopy print_output = False class TestMFKPLS(SMTestCase): def setUp(self): self.nt = 100 self.ne = 100 self.ndim = 3 self.n_comp = 2 def test_mfkpls(self): self.problems = ["exp", "tanh", "cos"] for fname in self.problems: prob = TensorProduct(ndim=self.ndim, func=fname) sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) for i in range(self.ndim): yt = np.concatenate((yt, prob(xt, kx=i)), axis=1) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) # Modif MM sm = MFKPLS() if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False # to test some options sm.options["eval_noise"] = True sm.options["optim_var"] = True # modif MM sm.options["n_comp"] = self.n_comp sm.options["theta0"] = [1e-2] * self.n_comp sm.set_training_values(xt, yt[:, 0]) sm.set_training_values(x_lf, y_lf[:, 0], name=0) sm.set_training_values(xt, yt[:, 0]) sm.set_training_values(x_lf, y_lf[:, 0], name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) self.assert_error(t_error, 0.0, 1.5) self.assert_error(e_error, 0.0, 1.5) def test_mfkpls_derivs(self): if self.ndim < 2: print("To try test_mfkpls_derivs the dimension must be greater than 1") prob = Sphere(ndim=self.ndim) sampling = LHS(xlimits=prob.xlimits) # Modif MM nt = 100 np.random.seed(0) xt = sampling(nt) yt = prob(xt) dyt = {} for kx in range(prob.xlimits.shape[0]): dyt[kx] = prob(xt, kx=kx) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) dye = {} for kx in range(prob.xlimits.shape[0]): dye[kx] = prob(xe, kx=kx) print("n_comp mfkpls_deriv", self.n_comp) # modif MM sm = MFKPLS() if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False # modif MM sm.options["n_comp"] = self.n_comp sm.options["theta0"] = [1e-2] * self.n_comp sm.set_training_values(xt, yt) sm.set_training_values(x_lf, y_lf, name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) e_error0 = compute_rms_error(sm, xe, dye[0], 0) e_error1 = compute_rms_error(sm, xe, dye[1], 1) if print_output: print( "%8s %6s %18.9e %18.9e %18.9e %18.9e" % (pname[:6], sname, t_error, e_error, e_error0, e_error1) ) self.assert_error(e_error0, 0.0, 1e-1) self.assert_error(e_error1, 0.0, 1e-1) @staticmethod def run_mfkpls_example(): import numpy as np import matplotlib.pyplot as plt from smt.applications.mfk import MFK, NestedLHS from smt.applications.mfkpls import MFKPLS # low fidelity model def lf_function(x): import numpy as np return ( 0.5 * ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) + (x - 0.5) * 10.0 - 5 ) # high fidelity model def hf_function(x): import numpy as np return ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) # Problem set up xlimits = np.array([[0.0, 1.0]]) xdoes = NestedLHS(nlevel=2, xlimits=xlimits, random_state=0) xt_c, xt_e = xdoes(7) # Evaluate the HF and LF functions yt_e = hf_function(xt_e) yt_c = lf_function(xt_c) # choice of number of PLS components ncomp = 1 sm = MFKPLS(n_comp=ncomp, theta0=ncomp * [1.0]) # low-fidelity dataset names being integers from 0 to level-1 sm.set_training_values(xt_c, yt_c, name=0) # high-fidelity dataset without name sm.set_training_values(xt_e, yt_e) # train the model sm.train() x = np.linspace(0, 1, 101, endpoint=True).reshape(-1, 1) # query the outputs y = sm.predict_values(x) mse = sm.predict_variances(x) derivs = sm.predict_derivatives(x, kx=0) plt.figure() plt.plot(x, hf_function(x), label="reference") plt.plot(x, y, linestyle="-.", label="mean_gp") plt.scatter(xt_e, yt_e, marker="o", color="k", label="HF doe") plt.scatter(xt_c, yt_c, marker="*", color="g", label="LF doe") plt.legend(loc=0) plt.ylim(-10, 17) plt.xlim(-0.1, 1.1) plt.xlabel(r"$x$") plt.ylabel(r"$y$") plt.show() if __name__ == "__main__": unittest.main()
6,025
27.027907
83
py
smt
smt-master/smt/applications/tests/test_mfkplsk.py
# -*- coding: utf-8 -*- """ Created on Mon May 07 14:20:11 2018 @author: m.meliani Adapted to new SMT version in march 2020 by Nathalie Bartoli """ import matplotlib matplotlib.use("Agg") import unittest import numpy as np import unittest import inspect from collections import OrderedDict from smt.problems import Sphere, TensorProduct from smt.sampling_methods import LHS, FullFactorial from smt.utils.sm_test_case import SMTestCase from smt.utils.silence import Silence from smt.utils import compute_rms_error from smt.surrogate_models import LS, QP, KPLS, KRG, KPLSK, GEKPLS, GENN from smt.applications.mfk import MFK, NestedLHS from smt.applications.mfkplsk import MFKPLSK from copy import deepcopy print_output = False class TestMFKPLSK(SMTestCase): def setUp(self): self.nt = 100 self.ne = 100 self.ndim = 3 self.n_comp = 2 def test_mfkplsk(self): self.problems = ["exp", "tanh", "cos"] for fname in self.problems: prob = TensorProduct(ndim=self.ndim, func=fname) sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) for i in range(self.ndim): yt = np.concatenate((yt, prob(xt, kx=i)), axis=1) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) # Modif MM sm = MFKPLSK() if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False # to test some options sm.options["eval_noise"] = True sm.options["optim_var"] = True # modif MM sm.options["n_comp"] = self.n_comp sm.options["theta0"] = [1e-2] * self.n_comp sm.set_training_values(xt, yt[:, 0]) sm.set_training_values(x_lf, y_lf[:, 0], name=0) sm.set_training_values(xt, yt[:, 0]) sm.set_training_values(x_lf, y_lf[:, 0], name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) self.assert_error(t_error, 0.0, 1.5) self.assert_error(e_error, 0.0, 1.5) def test_mfkplsk_derivs(self): if self.ndim < 2: print("To try test_mfkplsk_derivs the dimension must be greater than 1") prob = Sphere(ndim=self.ndim) sampling = LHS(xlimits=prob.xlimits) # Modif MM nt = 100 np.random.seed(0) xt = sampling(nt) yt = prob(xt) dyt = {} for kx in range(prob.xlimits.shape[0]): dyt[kx] = prob(xt, kx=kx) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) dye = {} for kx in range(prob.xlimits.shape[0]): dye[kx] = prob(xe, kx=kx) # modif MM sm = MFKPLSK() if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False # to test some options sm.options["eval_noise"] = False # modif MM sm.options["n_comp"] = self.n_comp sm.options["theta0"] = [1e-2] * self.n_comp sm.set_training_values(xt, yt) sm.set_training_values(x_lf, y_lf, name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) e_error0 = compute_rms_error(sm, xe, dye[0], 0) e_error1 = compute_rms_error(sm, xe, dye[1], 1) if print_output: print( "%8s %6s %18.9e %18.9e %18.9e %18.9e" % (pname[:6], sname, t_error, e_error, e_error0, e_error1) ) self.assert_error(e_error0, 0.0, 1e-1) self.assert_error(e_error1, 0.0, 1e-1) @staticmethod def run_mfkplsk_example(): import numpy as np import matplotlib.pyplot as plt from smt.applications.mfk import MFK, NestedLHS from smt.applications.mfkplsk import MFKPLSK # low fidelity modelk def lf_function(x): import numpy as np return ( 0.5 * ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) + (x - 0.5) * 10.0 - 5 ) # high fidelity model def hf_function(x): import numpy as np return ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) # Problem set up xlimits = np.array([[0.0, 1.0]]) xdoes = NestedLHS(nlevel=2, xlimits=xlimits, random_state=0) xt_c, xt_e = xdoes(7) # Evaluate the HF and LF functions yt_e = hf_function(xt_e) yt_c = lf_function(xt_c) # choice of number of PLS components ncomp = 1 sm = MFKPLSK(n_comp=ncomp, theta0=ncomp * [1.0]) # low-fidelity dataset names being integers from 0 to level-1 sm.set_training_values(xt_c, yt_c, name=0) # high-fidelity dataset without name sm.set_training_values(xt_e, yt_e) # train the model sm.train() x = np.linspace(0, 1, 101, endpoint=True).reshape(-1, 1) # query the outputs y = sm.predict_values(x) mse = sm.predict_variances(x) derivs = sm.predict_derivatives(x, kx=0) plt.figure() plt.plot(x, hf_function(x), label="reference") plt.plot(x, y, linestyle="-.", label="mean_gp") plt.scatter(xt_e, yt_e, marker="o", color="k", label="HF doe") plt.scatter(xt_c, yt_c, marker="*", color="g", label="LF doe") plt.legend(loc=0) plt.ylim(-10, 17) plt.xlim(-0.1, 1.1) plt.xlabel(r"$x$") plt.ylabel(r"$y$") plt.show() if __name__ == "__main__": unittest.main()
6,062
26.811927
84
py
smt
smt-master/smt/applications/tests/test_ego.py
# coding: utf-8 """ Author: Remi Lafage <remi.lafage@onera.fr> and Nathalie Bartoli This package is distributed under New BSD license. """ import warnings warnings.filterwarnings("ignore") import os import unittest import numpy as np from sys import argv import matplotlib matplotlib.use("Agg") from smt.applications import EGO from smt.applications.ego import Evaluator from smt.utils.sm_test_case import SMTestCase from smt.problems import Branin, Rosenbrock from smt.sampling_methods import FullFactorial from multiprocessing import Pool from smt.sampling_methods import LHS from smt.surrogate_models import ( KRG, GEKPLS, KPLS, MixIntKernelType, DesignSpace, OrdinalVariable, FloatVariable, CategoricalVariable, IntegerVariable, ) from smt.applications.mixed_integer import ( MixedIntegerContext, MixedIntegerSamplingMethod, ) # This implementation only works with Python > 3.3 class ParallelEvaluator(Evaluator): def run(self, fun, x): with Pool(3) as p: return np.array( [y[0] for y in p.map(fun, [np.atleast_2d(x[i]) for i in range(len(x))])] ) class TestEGO(SMTestCase): plot = None @staticmethod def function_test_1d(x): # function xsinx x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y = (x - 3.5) * np.sin((x - 3.5) / (np.pi)) return y.reshape((-1, 1)) def test_evaluator(self): x = [[1], [2], [3]] expected = TestEGO.function_test_1d(x) actual = ParallelEvaluator().run(TestEGO.function_test_1d, x) for i in range(len(x)): self.assertAlmostEqual(expected[i, 0], actual[i, 0]) def test_function_test_1d(self): n_iter = 15 xlimits = np.array([[0.0, 25.0]]) criterion = "EI" design_space = DesignSpace(xlimits) ego = EGO( n_iter=n_iter, criterion=criterion, n_doe=3, surrogate=KRG(design_space=design_space, print_global=False), random_state=42, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_1d) self.assertAlmostEqual(18.9, float(x_opt), delta=1) self.assertAlmostEqual(-15.1, float(y_opt), delta=1) def test_function_test_1d_parallel(self): n_iter = 3 xlimits = np.array([[0.0, 25.0]]) design_space = DesignSpace(xlimits) criterion = "EI" n_parallel = 3 ego = EGO( n_iter=n_iter, criterion=criterion, n_doe=3, surrogate=KRG(design_space=design_space, print_global=False), n_parallel=n_parallel, evaluator=ParallelEvaluator(), random_state=42, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_1d) self.assertAlmostEqual(18.9, float(x_opt), delta=1) self.assertAlmostEqual(-15.1, float(y_opt), delta=1) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_rosenbrock_2D(self): n_iter = 50 fun = Rosenbrock(ndim=2) xlimits = fun.xlimits criterion = "LCB" #'EI' or 'SBO' or 'LCB' random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = FullFactorial(xlimits=xlimits)(10) ego = EGO( n_start=30, xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=KRG(design_space=design_space, n_start=25, print_global=False), random_state=random_state, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) self.assertTrue(np.allclose([[1, 1]], x_opt, rtol=0.55)) self.assertAlmostEqual(0.0, float(y_opt), delta=1) def test_rosenbrock_2D_SBO(self): n_iter = 10 fun = Rosenbrock(ndim=2) xlimits = fun.xlimits criterion = "SBO" #'EI' or 'SBO' or 'LCB' design_space = DesignSpace(xlimits) xdoe = FullFactorial(xlimits=xlimits)(50) ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=KRG(design_space=design_space, print_global=False), random_state=42, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) self.assertTrue(np.allclose([[1, 1]], x_opt, atol=1)) self.assertAlmostEqual(0.0, float(y_opt), delta=1) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_rosenbrock_2D_parallel(self): n_iter = 20 n_parallel = 5 fun = Rosenbrock(ndim=2) xlimits = fun.xlimits criterion = "LCB" #'EI' or 'SBO' or 'LCB' random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = FullFactorial(xlimits=xlimits)(10) qEI = "KB" ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=KRG(design_space=design_space, print_global=False), n_parallel=n_parallel, qEI=qEI, evaluator=ParallelEvaluator(), random_state=random_state, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) print("Rosenbrock: ", x_opt) self.assertTrue(np.allclose([[1, 1]], x_opt, rtol=0.5)) self.assertAlmostEqual(0.0, float(y_opt), delta=1) def test_branin_2D(self): n_iter = 20 fun = Branin(ndim=2) criterion = "LCB" #'EI' or 'SBO' or 'LCB' design_space = fun.design_space ego = EGO( surrogate=KRG(design_space=design_space, print_global=False), n_iter=n_iter, criterion=criterion, n_doe=10, random_state=42, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) # 3 optimal points possible: [-pi, 12.275], [pi, 2.275], [9.42478, 2.475] self.assertTrue( np.allclose([[-3.14, 12.275]], x_opt, rtol=0.25) or np.allclose([[3.14, 2.275]], x_opt, rtol=0.25) or np.allclose([[9.42, 2.475]], x_opt, rtol=0.25) ) self.assertAlmostEqual(0.39, float(y_opt), delta=0.8) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_branin_2D_parallel(self): n_iter = 10 fun = Branin(ndim=2) n_parallel = 5 xlimits = fun.xlimits criterion = "EI" #'EI' or 'SBO' or 'LCB' design_space = DesignSpace(xlimits) xdoe = FullFactorial(xlimits=xlimits)(10) ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=KRG(design_space=design_space, print_global=False), n_parallel=n_parallel, random_state=42, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) # 3 optimal points possible: [-pi, 12.275], [pi, 2.275], [9.42478, 2.475] self.assertTrue( np.allclose([[-3.14, 12.275]], x_opt, rtol=0.5) or np.allclose([[3.14, 2.275]], x_opt, rtol=0.5) or np.allclose([[9.42, 2.475]], x_opt, rtol=0.5) ) print("Branin=", x_opt) self.assertAlmostEqual(0.39, float(y_opt), delta=1) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_branin_2D_mixed_parallel(self): n_parallel = 5 n_iter = 20 fun = Branin(ndim=2) xlimits = fun.xlimits criterion = "EI" #'EI' or 'SBO' or 'LCB' qEI = "CLmin" random_state = 42 design_space = DesignSpace( [ IntegerVariable(*xlimits[0]), FloatVariable(*xlimits[1]), ], seed=random_state, ) sm = KRG(design_space=design_space, print_global=False, n_start=25) mixint = MixedIntegerContext(design_space) sampling = mixint.build_sampling_method(FullFactorial) xdoe = sampling(10) ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, n_parallel=n_parallel, qEI=qEI, n_start=30, evaluator=ParallelEvaluator(), surrogate=sm, random_state=random_state, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) # 3 optimal points possible: [-pi, 12.275], [pi, 2.275], [9.42478, 2.475] self.assertTrue( np.allclose([[-3, 12.275]], x_opt, rtol=0.2) or np.allclose([[3, 2.275]], x_opt, rtol=0.2) or np.allclose([[9, 2.475]], x_opt, rtol=0.2) ) self.assertAlmostEqual(0.494, float(y_opt), delta=1) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_branin_2D_mixed(self): n_iter = 20 fun = Branin(ndim=2) xlimits = fun.xlimits random_state = 42 design_space = DesignSpace( [ IntegerVariable(*xlimits[0]), FloatVariable(*xlimits[1]), ], seed=random_state, ) criterion = "EI" #'EI' or 'SBO' or 'LCB' sm = KRG(design_space=design_space, print_global=False) sampling = MixedIntegerSamplingMethod(FullFactorial, design_space) xdoe = sampling(10) ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=sm, enable_tunneling=False, random_state=random_state, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) # 3 optimal points possible: [-pi, 12.275], [pi, 2.275], [9.42478, 2.475] self.assertTrue( np.allclose([[-3, 12.275]], x_opt, rtol=0.2) or np.allclose([[3, 2.275]], x_opt, rtol=0.2) or np.allclose([[9, 2.475]], x_opt, rtol=0.2) ) self.assertAlmostEqual(0.494, float(y_opt), delta=1) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_branin_2D_mixed_tunnel(self): n_iter = 20 fun = Branin(ndim=2) xlimits = fun.xlimits random_state = 42 design_space = DesignSpace( [ IntegerVariable(*xlimits[0]), FloatVariable(*xlimits[1]), ], seed=random_state, ) criterion = "EI" #'EI' or 'SBO' or 'LCB' sm = KRG(design_space=design_space, print_global=False) sampling = MixedIntegerSamplingMethod(FullFactorial, design_space) xdoe = sampling(30) ego = EGO( xdoe=xdoe, n_iter=n_iter, criterion=criterion, surrogate=sm, enable_tunneling=True, random_state=random_state, ) x_opt, y_opt, _, _, _ = ego.optimize(fun=fun) # 3 optimal points possible: [-pi, 12.275], [pi, 2.275], [9.42478, 2.475] self.assertTrue( np.allclose([[-3, 12.275]], x_opt, rtol=2) or np.allclose([[3, 2.275]], x_opt, rtol=2) or np.allclose([[9, 2.475]], x_opt, rtol=2) ) self.assertAlmostEqual(0.494, float(y_opt), delta=2) @staticmethod def function_test_mixed_integer(X): import numpy as np # float x1 = X[:, 0] # XType.ENUM 1 c1 = X[:, 1] x2 = c1 == 0 x3 = c1 == 1 x4 = c1 == 2 # XType.ENUM 2 c2 = X[:, 2] x5 = c2 == 0 x6 = c2 == 1 # int i = X[:, 3] y = ( (x2 + 2 * x3 + 3 * x4) * x5 * x1 + (x2 + 2 * x3 + 3 * x4) * x6 * 0.95 * x1 + i ) return y @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_ego_mixed_integer(self): n_iter = 15 n_doe = 5 random_state = 42 design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red", "green"]), CategoricalVariable(["large", "small"]), OrdinalVariable([0, 2, 3]), ], seed=random_state, ) xdoe, _ = design_space.sample_valid_x(n_doe) criterion = "EI" #'EI' or 'SBO' or 'LCB' ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=KRG(design_space=design_space, print_global=False), enable_tunneling=False, random_state=random_state, ) _, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_mixed_integer) self.assertAlmostEqual(-15, float(y_opt), delta=5) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_ego_mixed_integer_gower_distance(self): n_iter = 15 n_doe = 5 random_state = 42 design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red", "green"]), CategoricalVariable(["large", "small"]), IntegerVariable(0, 2), ], seed=random_state, ) xdoe, _ = design_space.sample_valid_x(n_doe) criterion = "EI" #'EI' or 'SBO' or 'LCB' ego = EGO( n_start=30, n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=KRG( n_start=25, design_space=design_space, categorical_kernel=MixIntKernelType.GOWER, print_global=False, ), enable_tunneling=False, random_state=random_state, ) _, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_mixed_integer) self.assertAlmostEqual(-15, float(y_opt), delta=5) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_ego_mixed_integer_hierarchical_NN(self): random_state = 42 def f_neu(x1, x2, x3, x4): if x4 == 0: return 2 * x1 + x2 - 0.5 * x3 elif x4 == 1: return -x1 + 2 * x2 - 0.5 * x3 elif x4 == 2: return -x1 + x2 + 0.5 * x3 else: raise ValueError(f"Unexpected x4: {x4}") def f1(x1, x2, x3, x4, x5): return f_neu(x1, x2, x3, x4) + x5**2 def f2(x1, x2, x3, x4, x5, x6): return f_neu(x1, x2, x3, x4) + (x5**2) + 0.3 * x6 def f3(x1, x2, x3, x4, x5, x6, x7): return f_neu(x1, x2, x3, x4) + (x5**2) + 0.3 * x6 - 0.1 * x7**3 def f_hv(X): y = [] for x in X: x3_decoded = design_space.decode_values(x, i_dv=3)[0] if x[0] == 0: y.append(f1(x[1], x[2], x3_decoded, x[4], x[5])) elif x[0] == 1: y.append(f2(x[1], x[2], x3_decoded, x[4], x[5], x[6])) elif x[0] == 2: y.append(f3(x[1], x[2], x3_decoded, x[4], x[5], x[6], x[7])) else: raise ValueError(f"Unexpected x0: {x[0]}") return np.array(y) random_state = 42 design_space = DesignSpace( [ OrdinalVariable(values=[1, 2, 3]), # x0 FloatVariable(-5, 2), FloatVariable(-5, 2), OrdinalVariable(values=[8, 16, 32, 64, 128, 256]), # x3 CategoricalVariable(values=["ReLU", "SELU", "ISRLU"]), # x4 IntegerVariable(0, 5), # x5 IntegerVariable(0, 5), # x6 IntegerVariable(0, 5), # x7 ], seed=random_state, ) # x6 is active when x0 >= 2 design_space.declare_decreed_var(decreed_var=6, meta_var=0, meta_value=[2, 3]) # x7 is active when x0 >= 3 design_space.declare_decreed_var(decreed_var=7, meta_var=0, meta_value=3) n_doe = 4 neutral_var_ds = DesignSpace(design_space.design_variables[1:]) sampling = MixedIntegerSamplingMethod( LHS, neutral_var_ds, criterion="ese", random_state=random_state ) x_cont = sampling(3 * n_doe) xdoe1 = np.zeros((n_doe, 8)) x_cont2 = x_cont[:n_doe, :5] xdoe1[:, 0] = np.zeros(n_doe) xdoe1[:, 1:6] = x_cont2 # ydoe1 = f_hv(xdoe1) xdoe1 = np.zeros((n_doe, 8)) xdoe1[:, 0] = np.zeros(n_doe) xdoe1[:, 1:6] = x_cont2 xdoe2 = np.zeros((n_doe, 8)) x_cont2 = x_cont[n_doe : 2 * n_doe, :6] xdoe2[:, 0] = np.ones(n_doe) xdoe2[:, 1:7] = x_cont2 # ydoe2 = f_hv(xdoe2) xdoe2 = np.zeros((n_doe, 8)) xdoe2[:, 0] = np.ones(n_doe) xdoe2[:, 1:7] = x_cont2 xdoe3 = np.zeros((n_doe, 8)) xdoe3[:, 0] = 2 * np.ones(n_doe) xdoe3[:, 1:] = x_cont[2 * n_doe :, :] # ydoe3 = f_hv(xdoe3) Xt = np.concatenate((xdoe1, xdoe2, xdoe3), axis=0) # Yt = np.concatenate((ydoe1, ydoe2, ydoe3), axis=0) n_iter = 6 criterion = "EI" ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=Xt, surrogate=KRG( design_space=design_space, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, theta0=[1e-2], n_start=5, corr="abs_exp", print_global=False, ), enable_tunneling=False, random_state=random_state, ) x_opt, y_opt, dnk, x_data, y_data = ego.optimize(fun=f_hv) self.assertAlmostEqual( f_hv(np.atleast_2d([2, -5, -5, 5, 0, 0, 0, 5])), float(y_opt), delta=15, ) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_ego_mixed_integer_hierarchical_Goldstein(self): def H(x1, x2, x3, x4, z3, z4, x5, cos_term): h = ( 53.3108 + 0.184901 * x1 - 5.02914 * x1**3 * 10 ** (-6) + 7.72522 * x1**z3 * 10 ** (-8) - 0.0870775 * x2 - 0.106959 * x3 + 7.98772 * x3**z4 * 10 ** (-6) + 0.00242482 * x4 + 1.32851 * x4**3 * 10 ** (-6) - 0.00146393 * x1 * x2 - 0.00301588 * x1 * x3 - 0.00272291 * x1 * x4 + 0.0017004 * x2 * x3 + 0.0038428 * x2 * x4 - 0.000198969 * x3 * x4 + 1.86025 * x1 * x2 * x3 * 10 ** (-5) - 1.88719 * x1 * x2 * x4 * 10 ** (-6) + 2.50923 * x1 * x3 * x4 * 10 ** (-5) - 5.62199 * x2 * x3 * x4 * 10 ** (-5) ) if cos_term: h += 5.0 * np.cos(2.0 * np.pi * (x5 / 100.0)) - 2.0 return h def f1(x1, x2, z1, z2, z3, z4, x5, cos_term): c1 = z2 == 0 c2 = z2 == 1 c3 = z2 == 2 c4 = z3 == 0 c5 = z3 == 1 c6 = z3 == 2 y = ( c4 * ( c1 * H(x1, x2, 20, 20, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 20, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 20, z3, z4, x5, cos_term) ) + c5 * ( c1 * H(x1, x2, 20, 50, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 50, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 50, z3, z4, x5, cos_term) ) + c6 * ( c1 * H(x1, x2, 20, 80, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 80, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 80, z3, z4, x5, cos_term) ) ) return y def f2(x1, x2, x3, z2, z3, z4, x5, cos_term): c1 = z2 == 0 c2 = z2 == 1 c3 = z2 == 2 y = ( c1 * H(x1, x2, x3, 20, z3, z4, x5, cos_term) + c2 * H(x1, x2, x3, 50, z3, z4, x5, cos_term) + c3 * H(x1, x2, x3, 80, z3, z4, x5, cos_term) ) return y def f3(x1, x2, x4, z1, z3, z4, x5, cos_term): c1 = z1 == 0 c2 = z1 == 1 c3 = z1 == 2 y = ( c1 * H(x1, x2, 20, x4, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, x4, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, x4, z3, z4, x5, cos_term) ) return y def f_hv(X): y = [] for x in X: if x[0] == 0: y.append( f1(x[2], x[3], x[7], x[8], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 1: y.append( f2(x[2], x[3], x[4], x[8], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 2: y.append( f3(x[2], x[3], x[5], x[7], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 3: y.append( H(x[2], x[3], x[4], x[5], x[9], x[10], x[6], cos_term=x[1]) ) else: raise ValueError return np.array(y) random_state = 0 ds = DesignSpace( [ CategoricalVariable(values=[0, 1, 2, 3]), # meta OrdinalVariable(values=[0, 1]), # x1 FloatVariable(0, 100), # x2 FloatVariable(0, 100), FloatVariable(0, 100), FloatVariable(0, 100), FloatVariable(0, 100), IntegerVariable(0, 2), # x7 IntegerVariable(0, 2), IntegerVariable(0, 2), IntegerVariable(0, 2), ], seed=random_state, ) # x4 is acting if meta == 1, 3 ds.declare_decreed_var(decreed_var=4, meta_var=0, meta_value=[1, 3]) # x5 is acting if meta == 2, 3 ds.declare_decreed_var(decreed_var=5, meta_var=0, meta_value=[2, 3]) # x7 is acting if meta == 0, 2 ds.declare_decreed_var(decreed_var=7, meta_var=0, meta_value=[0, 2]) # x8 is acting if meta == 0, 1 ds.declare_decreed_var(decreed_var=8, meta_var=0, meta_value=[0, 1]) n_doe = 25 ds.seed = random_state Xt, x_is_active = ds.sample_valid_x(n_doe) n_iter = 10 criterion = "EI" ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=Xt, surrogate=KRG( design_space=ds, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, theta0=[1e-2], n_start=10, corr="squar_exp", print_global=False, ), verbose=True, enable_tunneling=False, random_state=random_state, n_start=25, ) x_opt, y_opt, dnk, x_data, y_data = ego.optimize(fun=f_hv) self.assertAlmostEqual( 9.022, float(y_opt), delta=25, ) def test_ego_mixed_integer_homo_gaussian(self): n_iter = 15 random_state = 42 design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red", "green"]), CategoricalVariable(["large", "small"]), IntegerVariable(0, 2), ], seed=random_state, ) n_doe = 5 sampling = MixedIntegerSamplingMethod( LHS, design_space, criterion="ese", random_state=random_state, output_in_folded_space=True, ) xdoe = sampling(n_doe) criterion = "EI" #'EI' or 'SBO' or 'LCB' ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=KRG( design_space=design_space, categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, print_global=False, ), enable_tunneling=False, random_state=random_state, ) _, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_mixed_integer) self.assertAlmostEqual(-15, float(y_opt), delta=5) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 1, "too slow") def test_ego_mixed_integer_homo_gaussian_pls(self): n_iter = 15 random_state = 42 design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red", "green"]), CategoricalVariable(["large", "small"]), IntegerVariable(0, 2), ], seed=random_state, ) sampling = MixedIntegerSamplingMethod( LHS, design_space, criterion="ese", random_state=random_state, output_in_folded_space=True, ) n_doe = 5 xdoe = sampling(n_doe) criterion = "EI" #'EI' or 'SBO' or 'LCB' sm = KPLS( print_global=False, design_space=design_space, categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, n_comp=1, cat_kernel_comps=[2, 2], ) ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=sm, enable_tunneling=False, random_state=random_state, ) _, y_opt, _, _, _ = ego.optimize(fun=TestEGO.function_test_mixed_integer) self.assertAlmostEqual(-15, float(y_opt), delta=5) def test_ydoe_option(self): n_iter = 15 fun = Branin(ndim=2) xlimits = fun.xlimits criterion = "LCB" #'EI' or 'SBO' or 'LCB' random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = FullFactorial(xlimits=xlimits)(10) ydoe = fun(xdoe) ego = EGO( xdoe=xdoe, ydoe=ydoe, n_iter=n_iter, criterion=criterion, surrogate=KRG(design_space=design_space, print_global=False), random_state=random_state, ) _, y_opt, _, _, _ = ego.optimize(fun=fun) self.assertAlmostEqual(0.39, float(y_opt), delta=1) def test_find_best_point(self): fun = TestEGO.function_test_1d xlimits = np.array([[0.0, 25.0]]) random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = FullFactorial(xlimits=xlimits)(3) ydoe = fun(xdoe) ego = EGO( xdoe=xdoe, ydoe=ydoe, n_iter=1, criterion="LCB", surrogate=KRG(design_space=design_space, print_global=False), n_start=30, enable_tunneling=False, random_state=random_state, ) _, _, _, _, _ = ego.optimize(fun=fun) x, _ = ego._find_best_point(xdoe, ydoe, enable_tunneling=False) self.assertAlmostEqual(6.5, float(x), delta=1) @staticmethod def initialize_ego_gek(func="exp", criterion="LCB"): from smt.problems import TensorProduct class TensorProductIndirect(TensorProduct): def __init__(self, **kwargs): super().__init__(**kwargs) self.super = super() def _evaluate(self, x, kx): assert kx is None response = self.super._evaluate(x, kx) sens = np.hstack( self.super._evaluate(x, ki) for ki in range(x.shape[1]) ) return np.hstack((response, sens)) fun = TensorProductIndirect(ndim=2, func=func) random_state = 42 design_space = DesignSpace(fun.xlimits, seed=42) # Construction of the DOE sampling = LHS(xlimits=fun.xlimits, criterion="m", random_state=random_state) xdoe = sampling(20) ydoe = fun(xdoe) # Build the GEKPLS surrogate model n_comp = 2 sm = GEKPLS( theta0=[1e-2] * n_comp, design_space=design_space, extra_points=1, eval_comp_treshold=0.8, print_prediction=False, n_comp=n_comp, ) # Build the EGO optimizer and optimize ego = EGO( xdoe=xdoe, ydoe=ydoe, n_iter=5, criterion=criterion, surrogate=sm, n_start=30, enable_tunneling=False, random_state=random_state, ) return ego, fun def test_ego_gek(self): ego, fun = self.initialize_ego_gek() x_opt, _, _, _, _ = ego.optimize(fun=fun) self.assertAlmostEqual(-1.0, float(x_opt[0]), delta=1e-4) self.assertAlmostEqual(-1.0, float(x_opt[1]), delta=1e-4) def test_ei_gek(self): ego, fun = self.initialize_ego_gek(func="cos", criterion="EI") x_data, y_data = ego._setup_optimizer(fun) ego._train_gpr(x_data, y_data) # Test the EI value at the following point ei = ego.EI(np.array([[0.8398599985874058, -0.3240337426231973]])) self.assertTrue(np.allclose(ei, [6.87642e-12, 1.47804e-10, 2.76223], atol=1e-1)) def test_qei_criterion_default(self): fun = TestEGO.function_test_1d xlimits = np.array([[0.0, 25.0]]) random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = FullFactorial(xlimits=xlimits)(3) ydoe = fun(xdoe) ego = EGO( xdoe=xdoe, ydoe=ydoe, n_iter=1, n_parallel=2, criterion="SBO", surrogate=KRG(design_space=design_space, print_global=False), n_start=30, random_state=random_state, ) ego._setup_optimizer(fun) ego.gpr.set_training_values(xdoe, ydoe) ego.gpr.train() xtest = np.array([[10.0]]) # test that default virtual point should be equal to 3sigma lower bound kriging interval expected = float( ego.gpr.predict_values(xtest) - 3 * np.sqrt(ego.gpr.predict_variances(xtest)) ) actual = float(ego._get_virtual_point(xtest, fun(xtest))[0]) self.assertAlmostEqual(expected, actual) @unittest.skipIf(int(os.getenv("RUN_SLOW", 0)) < 2, "too slow") def test_examples(self): self.run_ego_example() self.run_ego_parallel_example() self.run_ego_mixed_integer_example() @staticmethod def run_ego_example(): import numpy as np from smt.applications import EGO from smt.surrogate_models import KRG from smt.utils.design_space import DesignSpace import matplotlib.pyplot as plt def function_test_1d(x): # function xsinx import numpy as np x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y = (x - 3.5) * np.sin((x - 3.5) / (np.pi)) return y.reshape((-1, 1)) n_iter = 6 xlimits = np.array([[0.0, 25.0]]) random_state = 42 # for reproducibility design_space = DesignSpace(xlimits, seed=random_state) xdoe = np.atleast_2d([0, 7, 25]).T n_doe = xdoe.size criterion = "EI" #'EI' or 'SBO' or 'LCB' ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=KRG(design_space=design_space, print_global=False), random_state=random_state, ) x_opt, y_opt, _, x_data, y_data = ego.optimize(fun=function_test_1d) print("Minimum in x={:.1f} with f(x)={:.1f}".format(float(x_opt), float(y_opt))) x_plot = np.atleast_2d(np.linspace(0, 25, 100)).T y_plot = function_test_1d(x_plot) fig = plt.figure(figsize=[10, 10]) for i in range(n_iter): k = n_doe + i x_data_k = x_data[0:k] y_data_k = y_data[0:k] ego.gpr.set_training_values(x_data_k, y_data_k) ego.gpr.train() y_gp_plot = ego.gpr.predict_values(x_plot) y_gp_plot_var = ego.gpr.predict_variances(x_plot) y_ei_plot = -ego.EI(x_plot) ax = fig.add_subplot((n_iter + 1) // 2, 2, i + 1) ax1 = ax.twinx() (ei,) = ax1.plot(x_plot, y_ei_plot, color="red") (true_fun,) = ax.plot(x_plot, y_plot) (data,) = ax.plot( x_data_k, y_data_k, linestyle="", marker="o", color="orange" ) if i < n_iter - 1: (opt,) = ax.plot( x_data[k], y_data[k], linestyle="", marker="*", color="r" ) (gp,) = ax.plot(x_plot, y_gp_plot, linestyle="--", color="g") sig_plus = y_gp_plot + 3 * np.sqrt(y_gp_plot_var) sig_moins = y_gp_plot - 3 * np.sqrt(y_gp_plot_var) un_gp = ax.fill_between( x_plot.T[0], sig_plus.T[0], sig_moins.T[0], alpha=0.3, color="g" ) lines = [true_fun, data, gp, un_gp, opt, ei] fig.suptitle("EGO optimization of $f(x) = x \sin{x}$") fig.subplots_adjust(hspace=0.4, wspace=0.4, top=0.8) ax.set_title("iteration {}".format(i + 1)) fig.legend( lines, [ "f(x)=xsin(x)", "Given data points", "Kriging prediction", "Kriging 99% confidence interval", "Next point to evaluate", "Expected improvment function", ], ) plt.show() # Check the optimal point is x_opt=18.9, y_opt =-15.1 @staticmethod def run_ego_mixed_integer_example(): import numpy as np from smt.applications import EGO from smt.applications.mixed_integer import MixedIntegerContext from smt.surrogate_models import MixIntKernelType from smt.utils.design_space import ( DesignSpace, CategoricalVariable, FloatVariable, IntegerVariable, ) import matplotlib.pyplot as plt from smt.surrogate_models import KRG from smt.sampling_methods import LHS # Regarding the interface, the function to be optimized should handle # categorical values as index values in the enumeration type specification. # For instance, here "blue" will be passed to the function as the index value 2. # This allows to keep the numpy ndarray X handling numerical values. def function_test_mixed_integer(X): # float x1 = X[:, 0] # enum 1 c1 = X[:, 1] x2 = c1 == 0 x3 = c1 == 1 x4 = c1 == 2 # enum 2 c2 = X[:, 2] x5 = c2 == 0 x6 = c2 == 1 # int i = X[:, 3] y = ( (x2 + 2 * x3 + 3 * x4) * x5 * x1 + (x2 + 2 * x3 + 3 * x4) * x6 * 0.95 * x1 + i ) return y.reshape((-1, 1)) n_iter = 15 random_state = 42 design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red", "green"]), CategoricalVariable(["square", "circle"]), IntegerVariable(0, 2), ], seed=random_state, ) criterion = "EI" #'EI' or 'SBO' or 'LCB' qEI = "KBRand" sm = KRG( design_space=design_space, categorical_kernel=MixIntKernelType.GOWER, print_global=False, ) mixint = MixedIntegerContext(design_space) n_doe = 3 sampling = mixint.build_sampling_method( LHS, criterion="ese", random_state=random_state ) xdoe = sampling(n_doe) ydoe = function_test_mixed_integer(xdoe) ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, ydoe=ydoe, surrogate=sm, qEI=qEI, n_parallel=2, random_state=random_state, ) x_opt, y_opt, _, _, y_data = ego.optimize(fun=function_test_mixed_integer) print("Minimum in x={} with f(x)={:.1f}".format(x_opt, float(y_opt))) # print("Minimum in typed x={}".format(ego.mixint.cast_to_mixed_integer(x_opt))) min_ref = -15 mini = np.zeros(n_iter) for k in range(n_iter): mini[k] = np.log(np.abs(np.min(y_data[0 : k + n_doe - 1]) - min_ref)) x_plot = np.linspace(1, n_iter + 0.5, n_iter) u = max(np.floor(max(mini)) + 1, -100) l = max(np.floor(min(mini)) - 0.2, -10) fig = plt.figure() axes = fig.add_axes([0.1, 0.1, 0.8, 0.8]) axes.plot(x_plot, mini, color="r") axes.set_ylim([l, u]) plt.title("minimum convergence plot", loc="center") plt.xlabel("number of iterations") plt.ylabel("log of the difference w.r.t the best") plt.show() @staticmethod def run_ego_parallel_example(): import numpy as np from smt.applications import EGO from smt.applications.ego import Evaluator from smt.surrogate_models import KRG, DesignSpace import matplotlib.pyplot as plt def function_test_1d(x): # function xsinx import numpy as np x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y = (x - 3.5) * np.sin((x - 3.5) / (np.pi)) return y.reshape((-1, 1)) n_iter = 3 n_parallel = 3 n_start = 50 xlimits = np.array([[0.0, 25.0]]) random_state = 42 design_space = DesignSpace(xlimits, seed=random_state) xdoe = np.atleast_2d([0, 7, 25]).T n_doe = xdoe.size class ParallelEvaluator(Evaluator): """ Implement Evaluator interface using multiprocessing ThreadPool object (Python 3 only). """ def run(self, fun, x): n_thread = 5 # Caveat: import are made here due to SMT documentation building process import numpy as np from sys import version_info from multiprocessing.pool import ThreadPool if version_info.major == 2: return fun(x) # Python 3 only with ThreadPool(n_thread) as p: return np.array( [ y[0] for y in p.map( fun, [np.atleast_2d(x[i]) for i in range(len(x))] ) ] ) criterion = "EI" #'EI' or 'SBO' or 'LCB' qEI = "KBUB" # "KB", "KBLB", "KBUB", "KBRand" ego = EGO( n_iter=n_iter, criterion=criterion, xdoe=xdoe, surrogate=KRG(design_space=design_space, print_global=False), n_parallel=n_parallel, qEI=qEI, n_start=n_start, evaluator=ParallelEvaluator(), random_state=random_state, ) x_opt, y_opt, _, x_data, y_data = ego.optimize(fun=function_test_1d) print("Minimum in x={:.1f} with f(x)={:.1f}".format(float(x_opt), float(y_opt))) x_plot = np.atleast_2d(np.linspace(0, 25, 100)).T y_plot = function_test_1d(x_plot) fig = plt.figure(figsize=[10, 10]) for i in range(n_iter): k = n_doe + (i) * (n_parallel) x_data_k = x_data[0:k] y_data_k = y_data[0:k] x_data_sub = x_data_k.copy() y_data_sub = y_data_k.copy() for p in range(n_parallel): ego.gpr.set_training_values(x_data_sub, y_data_sub) ego.gpr.train() y_ei_plot = -ego.EI(x_plot) y_gp_plot = ego.gpr.predict_values(x_plot) y_gp_plot_var = ego.gpr.predict_variances(x_plot) x_data_sub = np.append(x_data_sub, x_data[k + p]) y_KB = ego._get_virtual_point(np.atleast_2d(x_data[k + p]), y_data_sub) y_data_sub = np.append(y_data_sub, y_KB) ax = fig.add_subplot(n_iter, n_parallel, i * (n_parallel) + p + 1) ax1 = ax.twinx() (ei,) = ax1.plot(x_plot, y_ei_plot, color="red") (true_fun,) = ax.plot(x_plot, y_plot) (data,) = ax.plot( x_data_sub[: -1 - p], y_data_sub[: -1 - p], linestyle="", marker="o", color="orange", ) (virt_data,) = ax.plot( x_data_sub[-p - 1 : -1], y_data_sub[-p - 1 : -1], linestyle="", marker="o", color="g", ) (opt,) = ax.plot( x_data_sub[-1], y_data_sub[-1], linestyle="", marker="*", color="r" ) (gp,) = ax.plot(x_plot, y_gp_plot, linestyle="--", color="g") sig_plus = y_gp_plot + 3.0 * np.sqrt(y_gp_plot_var) sig_moins = y_gp_plot - 3.0 * np.sqrt(y_gp_plot_var) un_gp = ax.fill_between( x_plot.T[0], sig_plus.T[0], sig_moins.T[0], alpha=0.3, color="g" ) lines = [true_fun, data, gp, un_gp, opt, ei, virt_data] fig.suptitle("EGOp optimization of $f(x) = x \sin{x}$") fig.subplots_adjust(hspace=0.4, wspace=0.4, top=0.8) ax.set_title("iteration {}.{}".format(i, p)) fig.legend( lines, [ "f(x)=xsin(x)", "Given data points", "Kriging prediction", "Kriging 99% confidence interval", "Next point to evaluate", "Expected improvment function", "Virtula data points", ], ) plt.show() if __name__ == "__main__": if "--plot" in argv: TestEGO.plot = True argv.remove("--plot") if "--example" in argv: TestEGO.run_ego_mixed_integer_example() exit() unittest.main()
43,079
32.525292
98
py
smt
smt-master/smt/applications/tests/test_mfk_variance.py
# -*- coding: utf-8 -*- """ Created on Mon Apr 27 15:36:13 2020 @author: Vincent Drouet and Nathalie Bartoli in order to validate the variance formula for multifidelity on the branin 2D function Comparisons are based on the paper Le Gratiet et Cannamela 2015 : Le Gratiet, L., & Cannamela, C. (2015). Cokriging-based sequential design strategies using fast cross-validation techniques for multi-fidelity computer codes. Technometrics, 57(3), 418-427. https://doi.org/10.1080/00401706.2014.928233 """ import numpy as np from smt.applications.mfk import MFK, NestedLHS from smt.sampling_methods import LHS import unittest from smt.utils.sm_test_case import SMTestCase print_output = True # %% # Define the low and high fidelity models # Example on a 2D problem: Branin Function A_corr = 0 # high fidelity model def HF(point): # Expensive Forretal function res = ( ( point[:, 1] - (5.1 / (4 * np.pi**2)) * point[:, 0] ** 2 + (5 / np.pi) * point[:, 0] - 6 ) ** 2 + 10 * (1 - (1 / 8 / np.pi)) * np.cos(point[:, 0]) + 10 + 5 * point[:, 0] ) return res # low fidelity model def LF(point): # Cheap Forretal function res = ( HF(point) - (0.5 * A_corr**2 + A_corr + 0.2) * ( point[:, 1] - 5.1 / 4 / np.pi**2 * point[:, 0] ** 2 + 5.0 / np.pi * point[:, 0] - 6 ) ** 2 ) return res class TestMFK_variance(SMTestCase): @staticmethod def corr(x1, x2, sm, lvl): # Normalization of data with X_scale and X_offset from sm # lvl start to 1 X_scale = sm.X_scale X_offset = sm.X_offset X1 = (x1 - X_offset) / X_scale X2 = (x2 - X_offset) / X_scale thetas = sm.optimal_theta[lvl - 1] if sm.options["corr"] == "abs_exp": p = 1 else: p = 2 return np.prod(np.exp(-thetas * np.abs(X1 - X2) ** p)) @staticmethod def hyperparam_LG(sm): # computation of the hyperparameters based on the formula from LeGratiet 2015 y_t_lf = sm.training_points[0][0][1] n_LF = y_t_lf.shape[0] C = sm.optimal_par[0]["C"] R = C @ C.T H = np.ones((n_LF, 1)) R_inv_H = np.linalg.solve(R, H) M = H.T @ R_inv_H M_inv = np.linalg.inv(M) R_inv_y = np.linalg.solve(R, y_t_lf) beta_LG_1 = M_inv @ H.T @ R_inv_y sigma2_LG_1 = ( (y_t_lf - H @ beta_LG_1).T @ np.linalg.solve(R, y_t_lf - H @ beta_LG_1) / (n_LF - 1) )[0, 0] y_t_hf = sm.training_points[None][0][1] n_HF = y_t_hf.shape[0] y_D_l = y_t_lf[-n_HF:, :].reshape( -1, 1 ) # to get y^l-1(D^l) we need n_HF last points of y_t_lf H = np.hstack((y_D_l, np.ones((n_HF, 1)))) C = sm.optimal_par[1]["C"] R = C @ C.T R_inv_H = np.linalg.solve(R, H) M = H.T @ R_inv_H M_inv = np.linalg.inv(M) R_inv_y = np.linalg.solve(R, y_t_hf) rho_beta = M_inv @ H.T @ R_inv_y rho_LG = rho_beta[0, 0] beta_LG_2 = rho_beta[1, 0] sigma2_LG_2 = ( (y_t_hf - H @ rho_beta).T @ np.linalg.solve(R, y_t_hf - H @ rho_beta) / (n_HF - 2) )[0, 0] sigma2_rho_LG = rho_LG**2 + sigma2_LG_2 * M_inv[0, 0] return beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG @staticmethod def Mu_LG_LG(x, sm): # computation of the mean based on the formula from LeGratiet 2015 # and using beta from LeGratiet 2015 ( beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG, ) = TestMFK_variance.hyperparam_LG(sm) C = sm.optimal_par[0]["C"] R = C @ C.T x_t_lf = sm.training_points[0][0][0] y_t_lf = sm.training_points[0][0][1] n_LF = x_t_lf.shape[0] r_x = np.empty((n_LF, 1)) for i in range(n_LF): r_x[i, 0] = TestMFK_variance.corr(x, x_t_lf[i], sm, 1) M = np.linalg.solve(R, y_t_lf - beta_LG_1 * np.ones((n_LF, 1))) mu_0 = beta_LG_1 + r_x.T @ M x_t_hf = sm.training_points[None][0][0] y_t_hf = sm.training_points[None][0][1] n_HF = x_t_hf.shape[0] C = sm.optimal_par[1]["C"] R = C @ C.T r_x = np.empty((n_HF, 1)) for i in range(n_HF): r_x[i, 0] = TestMFK_variance.corr(x, x_t_hf[i], sm, 2) M = np.linalg.solve( R, y_t_hf - beta_LG_2 * np.ones((n_HF, 1)) - rho_LG * y_t_lf[-n_HF:, 0, np.newaxis], ) return rho_LG * mu_0 + beta_LG_2 + r_x.T @ M @staticmethod def Mu_LG_sm(x, sm): # computation of the mean based on the formula from LeGratiet 2015 # and using beta from sm beta = sm.optimal_par[0]["beta"] C = sm.optimal_par[0]["C"] R = C @ C.T x_t_lf = sm.training_points[0][0][0] y_t_lf = sm.training_points[0][0][1] n_LF = x_t_lf.shape[0] r_x = np.empty((n_LF, 1)) for i in range(n_LF): r_x[i, 0] = TestMFK_variance.corr(x, x_t_lf[i], sm, 1) M = np.linalg.solve(R, y_t_lf - beta * np.ones((n_LF, 1))) mu_0 = beta + r_x.T @ M x_t_hf = sm.training_points[None][0][0] y_t_hf = sm.training_points[None][0][1] n_HF = x_t_hf.shape[0] beta = sm.optimal_par[1]["beta"] rho = beta[0] C = sm.optimal_par[1]["C"] R = C @ C.T r_x = np.empty((n_HF, 1)) for i in range(n_HF): r_x[i, 0] = TestMFK_variance.corr(x, x_t_hf[i], sm, 2) M = np.linalg.solve( R, y_t_hf - beta[1] * np.ones((n_HF, 1)) - rho * y_t_lf[-n_HF:, 0, np.newaxis], ) return rho * mu_0 + beta[1] + r_x.T @ M @staticmethod def Cov_LG_sm(x1, x2, sm): # computation of the covariance based on the formula from LeGratiet 2015 # Using the Sigma2 output from sm x_t_lf = sm.training_points[0][0][0] y_t_lf = sm.training_points[0][0][1] n_LF = x_t_lf.shape[0] sigma2 = sm.optimal_par[0]["sigma2"] C = sm.optimal_par[0]["C"] R = C @ C.T r_x1 = np.empty((n_LF, 1)) r_x2 = np.empty((n_LF, 1)) for i in range(n_LF): r_x1[i, 0] = TestMFK_variance.corr(x1, x_t_lf[i], sm, 1) r_x2[i, 0] = TestMFK_variance.corr(x2, x_t_lf[i], sm, 1) h_x1 = np.ones((1, 1)) h_x2 = np.ones((1, 1)) H = np.ones((n_LF, 1)) hr_1 = np.vstack((h_x1, r_x1)) hr_2 = np.vstack((h_x2, r_x2)) M1 = np.hstack((np.zeros((1, 1)), H.T)) M2 = np.hstack((H, R)) M3 = np.vstack((M1, M2)) M4 = np.linalg.solve(M3, hr_2) # M4 = M3^-1 @ hr_2 k_0 = sigma2 * ( TestMFK_variance.corr(x1, x2, sm, 1) - hr_1.T @ M4 ) # covariance of level 0 x_t_hf = sm.training_points[None][0][0] n_HF = x_t_hf.shape[0] sigma2 = sm.optimal_par[1]["sigma2"] (var_all_pred, sigma2_rho) = sm.predict_variances_all_levels(x1) sigma2_rho = sigma2_rho[0] C = sm.optimal_par[1]["C"] R = C @ C.T r_x1 = np.empty((n_HF, 1)) r_x2 = np.empty((n_HF, 1)) for i in range(n_HF): r_x1[i, 0] = TestMFK_variance.corr(x1, x_t_hf[i], sm, 2) r_x2[i, 0] = TestMFK_variance.corr(x2, x_t_hf[i], sm, 2) mu_x1 = sm._predict_intermediate_values(x1, 1).reshape(-1, 1) mu_x2 = sm._predict_intermediate_values(x2, 1).reshape(-1, 1) h_x1 = np.vstack((mu_x1, np.ones((1, 1)))) h_x2 = np.vstack((mu_x2, np.ones((1, 1)))) y_D_l = y_t_lf[-n_HF:, :].reshape( -1, 1 ) # to get y^l-1(D^l) we need n_HF last points of y_t_lf H = np.hstack((y_D_l, np.ones((n_HF, 1)))) hr_1 = np.vstack((h_x1, r_x1)) hr_2 = np.vstack((h_x2, r_x2)) M1 = np.hstack((np.zeros((2, 2)), H.T)) M2 = np.hstack((H, R)) M3 = np.vstack((M1, M2)) M4 = np.linalg.solve(M3, hr_2) # M4 = M3^-1 @ hr_2 k_1 = sigma2_rho * k_0 + sigma2 * ( TestMFK_variance.corr(x1, x2, sm, 2) - hr_1.T @ M4 ) return k_0, k_1 @staticmethod def Cov_LG_LG(x1, x2, sm): # computation of the covariance based on the formula from LeGratiet 2015 # Using the Sigma2 output from Le Gratiet paper x_t_lf = sm.training_points[0][0][0] y_t_lf = sm.training_points[0][0][1] n_LF = x_t_lf.shape[0] ( beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG, ) = TestMFK_variance.hyperparam_LG(sm) sigma2 = sigma2_LG_1 C = sm.optimal_par[0]["C"] R = C @ C.T r_x1 = np.empty((n_LF, 1)) r_x2 = np.empty((n_LF, 1)) for i in range(n_LF): r_x1[i, 0] = TestMFK_variance.corr(x1, x_t_lf[i], sm, 1) r_x2[i, 0] = TestMFK_variance.corr(x2, x_t_lf[i], sm, 1) h_x1 = np.ones((1, 1)) h_x2 = np.ones((1, 1)) H = np.ones((n_LF, 1)) hr_1 = np.vstack((h_x1, r_x1)) hr_2 = np.vstack((h_x2, r_x2)) M1 = np.hstack((np.zeros((1, 1)), H.T)) M2 = np.hstack((H, R)) M3 = np.vstack((M1, M2)) M4 = np.linalg.solve(M3, hr_2) # M4 = M3^-1 @ hr_2 k_0 = sigma2 * ( TestMFK_variance.corr(x1, x2, sm, 1) - hr_1.T @ M4 ) # covariance of level 0 x_t_hf = sm.training_points[None][0][0] n_HF = x_t_hf.shape[0] sigma2 = sigma2_LG_2 C = sm.optimal_par[1]["C"] R = C @ C.T r_x1 = np.empty((n_HF, 1)) r_x2 = np.empty((n_HF, 1)) for i in range(n_HF): r_x1[i, 0] = TestMFK_variance.corr(x1, x_t_hf[i], sm, 2) r_x2[i, 0] = TestMFK_variance.corr(x2, x_t_hf[i], sm, 2) mu_x1 = sm._predict_intermediate_values(x1, 1).reshape(-1, 1) mu_x2 = sm._predict_intermediate_values(x2, 1).reshape(-1, 1) h_x1 = np.vstack((mu_x1, np.ones((1, 1)))) h_x2 = np.vstack((mu_x2, np.ones((1, 1)))) y_D_l = y_t_lf[-n_HF:, :].reshape( -1, 1 ) # to get y^l-1(D^l) we need n_HF last points of y_t_lf H = np.hstack((y_D_l, np.ones((n_HF, 1)))) hr_1 = np.vstack((h_x1, r_x1)) hr_2 = np.vstack((h_x2, r_x2)) M1 = np.hstack((np.zeros((2, 2)), H.T)) M2 = np.hstack((H, R)) M3 = np.vstack((M1, M2)) M4 = np.linalg.solve(M3, hr_2) # M4 = M3^-1 @ hr_2 k_1 = sigma2_rho_LG * k_0 + sigma2 * ( TestMFK_variance.corr(x1, x2, sm, 2) - hr_1.T @ M4 ) return k_0, k_1 @staticmethod def verif_hyperparam(sm, x_test_LHS): # get the hyperparameters from sm beta_sm_1 = sm.optimal_par[0]["beta"][0, 0] sigma2_sm_1 = sm.optimal_par[0]["sigma2"][0] rho_beta_sm = sm.optimal_par[1]["beta"] beta_sm_2 = rho_beta_sm[1, 0] sigma2_sm_2 = sm.optimal_par[1]["sigma2"][0] rho_sm = rho_beta_sm[0, 0] (var_all_pred, sigma2_rho) = sm.predict_variances_all_levels(x_test_LHS) sigma2_rho_sm = sigma2_rho[0] ( beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG, ) = TestMFK_variance.hyperparam_LG(sm) return ( beta_sm_1, sigma2_sm_1, beta_sm_2, sigma2_sm_2, rho_sm, sigma2_rho_sm, beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG, ) def test_mfk_variance(self): # To create the doe # dim = 2 nlevel = 2 ub0 = 10.0 ub1 = 15.0 lb0 = -5.0 lb1 = 0.0 xlimits = np.array([[lb0, ub0], [lb1, ub1]]) # Constants n_HF = 5 # number of high fidelity points (number of low fi is twice) xdoes = NestedLHS(nlevel=nlevel, xlimits=xlimits) x_t_lf, x_t_hf = xdoes(n_HF) # Evaluate the HF and LF functions y_t_lf = LF(x_t_lf) y_t_hf = HF(x_t_hf) sm = MFK( theta0=x_t_hf.shape[1] * [1e-2], print_global=False, rho_regr="constant", ) # low-fidelity dataset names being integers from 0 to level-1 sm.set_training_values(x_t_lf, y_t_lf, name=0) # high-fidelity dataset without name sm.set_training_values(x_t_hf, y_t_hf) # train the model sm.train() # Validation set # for validation with LHS ntest = 1 sampling = LHS(xlimits=xlimits) x_test_LHS = sampling(ntest) # y_test_LHS = HF(x_test_LHS) # compare the mean value between different formula if print_output: print("Mu sm : {}".format(sm.predict_values(x_test_LHS)[0, 0])) print( "Mu LG_sm : {}".format(TestMFK_variance.Mu_LG_sm(x_test_LHS, sm)[0, 0]) ) print( "Mu LG_LG : {}".format(TestMFK_variance.Mu_LG_LG(x_test_LHS, sm)[0, 0]) ) # self.assertAlmostEqual( # TestMFK_variance.Mu_LG_sm(x_test_LHS, sm)[0, 0], # TestMFK_variance.Mu_LG_LG(x_test_LHS, sm)[0, 0], # delta=1, # ) self.assertAlmostEqual( sm.predict_values(x_test_LHS)[0, 0], TestMFK_variance.Mu_LG_LG(x_test_LHS, sm)[0, 0], delta=1, ) # compare the variance value between different formula (k_0_LG_sm, k_1_LG_sm) = TestMFK_variance.Cov_LG_sm(x_test_LHS, x_test_LHS, sm) (k_0_LG_LG, k_1_LG_LG) = TestMFK_variance.Cov_LG_LG(x_test_LHS, x_test_LHS, sm) k_0_sm = sm.predict_variances_all_levels(x_test_LHS)[0][0, 0] k_1_sm = sm.predict_variances_all_levels(x_test_LHS)[0][0, 1] if print_output: print("Level 0") print("Var sm : {}".format(k_0_sm)) print("Var LG_sm : {}".format(k_0_LG_sm[0, 0])) print("Var LG_LG : {}".format(k_0_LG_LG[0, 0])) print("Level 1") print("Var sm : {}".format(k_1_sm)) print("Var LG_sm : {}".format(k_1_LG_sm[0, 0])) print("Var LG_LG : {}".format(k_1_LG_LG[0, 0])) # for level 0 self.assertAlmostEqual(k_0_sm, k_0_LG_sm[0, 0], delta=1) self.assertAlmostEqual(k_0_LG_sm[0, 0], k_0_LG_LG[0, 0], delta=1) # for level 1 self.assertAlmostEqual(k_1_sm, k_1_LG_sm[0, 0], delta=1) self.assertAlmostEqual(k_1_LG_sm[0, 0], k_1_LG_LG[0, 0], delta=1) ( beta_sm_1, sigma2_sm_1, beta_sm_2, sigma2_sm_2, rho_sm, sigma2_rho_sm, beta_LG_1, sigma2_LG_1, beta_LG_2, sigma2_LG_2, rho_LG, sigma2_rho_LG, ) = TestMFK_variance.verif_hyperparam(sm, x_test_LHS) if print_output: print("Hyperparameters") print("rho_sm : {}".format(rho_sm)) print("rho_LG : {}".format(rho_LG)) print("sigma2_rho_sm : {}".format(sigma2_rho_sm[0])) print("sigma2_rho_LG : {}".format(sigma2_rho_LG)) print("beta_sm_1 : {}".format(beta_sm_1)) print("beta_LG_1 : {}".format(beta_LG_1[0, 0])) print("beta_sm_2 : {}".format(beta_sm_2)) print("beta_LG_2 : {}".format(beta_LG_2)) print("sigma2_sm_1 : {}".format(sigma2_sm_1)) print("sigma2_LG_1 : {}".format(sigma2_LG_1)) print("sigma2_sm_2 : {}".format(sigma2_sm_2)) print("sigma2_LG_2 : {}".format(sigma2_LG_2)) if __name__ == "__main__": unittest.main()
16,153
33.080169
88
py
smt
smt-master/smt/applications/tests/test_vfm.py
""" Author: Mohamed Amine Bouhlel <mbouhlel@umich.edu> This package is distributed under New BSD license. """ import unittest import matplotlib matplotlib.use("Agg") import numpy as np from scipy import linalg from smt.utils.sm_test_case import SMTestCase from smt.utils import compute_rms_error from smt.utils.silence import Silence from smt.problems import WaterFlowLFidelity, WaterFlow from smt.sampling_methods import LHS from smt.applications import VFM from smt.utils.misc import compute_rms_error from smt.examples.rans_crm_wing.rans_crm_wing import ( get_rans_crm_wing, plot_rans_crm_wing, ) def setupCRM(LF_candidate="QP", Bridge_candidate="KRG", type_bridge="Additive"): xt, yt, _ = get_rans_crm_wing() optionsLF = {} optionsB = {} M = VFM( type_bridge=type_bridge, name_model_LF=LF_candidate, name_model_bridge=Bridge_candidate, X_LF=xt, y_LF=yt, X_HF=xt, y_HF=yt, options_LF=optionsLF, options_bridge=optionsB, ) return M, xt class TestVFM(SMTestCase): def test_vfm(self): # Problem set up ndim = 8 ntest = 500 ndoeLF = int(10 * ndim) ndoeHF = int(3) funLF = WaterFlowLFidelity(ndim=ndim) funHF = WaterFlow(ndim=ndim) deriv1 = True deriv2 = True LF_candidate = "QP" Bridge_candidate = "KRG" type_bridge = "Multiplicative" optionsLF = {} optionsB = {"theta0": [1e-2] * ndim, "print_prediction": False, "deriv": False} # Construct low/high fidelity data and validation points sampling = LHS(xlimits=funLF.xlimits, criterion="m", random_state=42) xLF = sampling(ndoeLF) yLF = funLF(xLF) if deriv1: dy_LF = np.zeros((ndoeLF, 1)) for i in range(ndim): yd = funLF(xLF, kx=i) dy_LF = np.concatenate((dy_LF, yd), axis=1) sampling = LHS(xlimits=funHF.xlimits, criterion="m", random_state=43) xHF = sampling(ndoeHF) yHF = funHF(xHF) if deriv2: dy_HF = np.zeros((ndoeHF, 1)) for i in range(ndim): yd = funHF(xHF, kx=i) dy_HF = np.concatenate((dy_HF, yd), axis=1) xtest = sampling(ntest) ytest = funHF(xtest) dytest = np.zeros((ntest, ndim)) for i in range(ndim): dytest[:, i] = funHF(xtest, kx=i).T # Initialize VFM vfm = VFM( type_bridge=type_bridge, name_model_LF=LF_candidate, name_model_bridge=Bridge_candidate, X_LF=xLF, y_LF=yLF, X_HF=xHF, y_HF=yHF, options_LF=optionsLF, options_bridge=optionsB, dy_LF=dy_LF, dy_HF=dy_HF, ) # Prediction of the validation points rms_error = compute_rms_error(vfm, xtest, ytest) self.assert_error(rms_error, 0.0, 3e-1) @staticmethod def run_vfm_example(self): import matplotlib.pyplot as plt import numpy as np from scipy import linalg from smt.utils import compute_rms_error from smt.problems import WaterFlowLFidelity, WaterFlow from smt.sampling_methods import LHS from smt.applications import VFM # Problem set up ndim = 8 ntest = 500 ndoeLF = int(10 * ndim) ndoeHF = int(3) funLF = WaterFlowLFidelity(ndim=ndim) funHF = WaterFlow(ndim=ndim) deriv1 = True deriv2 = True LF_candidate = "QP" Bridge_candidate = "KRG" type_bridge = "Multiplicative" optionsLF = {} optionsB = {"theta0": [1e-2] * ndim, "print_prediction": False, "deriv": False} # Construct low/high fidelity data and validation points sampling = LHS(xlimits=funLF.xlimits, criterion="m") xLF = sampling(ndoeLF) yLF = funLF(xLF) if deriv1: dy_LF = np.zeros((ndoeLF, 1)) for i in range(ndim): yd = funLF(xLF, kx=i) dy_LF = np.concatenate((dy_LF, yd), axis=1) sampling = LHS(xlimits=funHF.xlimits, criterion="m") xHF = sampling(ndoeHF) yHF = funHF(xHF) if deriv2: dy_HF = np.zeros((ndoeHF, 1)) for i in range(ndim): yd = funHF(xHF, kx=i) dy_HF = np.concatenate((dy_HF, yd), axis=1) xtest = sampling(ntest) ytest = funHF(xtest) dytest = np.zeros((ntest, ndim)) for i in range(ndim): dytest[:, i] = funHF(xtest, kx=i).T # Initialize the extension VFM M = VFM( type_bridge=type_bridge, name_model_LF=LF_candidate, name_model_bridge=Bridge_candidate, X_LF=xLF, y_LF=yLF, X_HF=xHF, y_HF=yHF, options_LF=optionsLF, options_bridge=optionsB, dy_LF=dy_LF, dy_HF=dy_HF, ) # Prediction of the validation points y = M.predict_values(x=xtest) plt.figure() plt.plot(ytest, ytest, "-.") plt.plot(ytest, y, ".") plt.xlabel(r"$y$ True") plt.ylabel(r"$y$ prediction") plt.show() def test_KRG_KRG_additive(self): with Silence(): M, xt = setupCRM( LF_candidate="KRG", Bridge_candidate="KRG", type_bridge="Additive" ) with Silence(): yp = M.predict_values(np.atleast_2d(xt[0])) dyp = M.predict_derivatives(np.atleast_2d(xt[0]), kx=0) self.assert_error(yp, np.array([[0.015368, 0.367424]]), atol=2e-2, rtol=3e-2) self.assert_error(dyp, np.array([[0.07007729, 3.619421]]), atol=3e-1, rtol=1e-2) def test_QP_KRG_additive(self): with Silence(): M, xt = setupCRM( LF_candidate="QP", Bridge_candidate="KRG", type_bridge="Additive" ) with Silence(): yp = M.predict_values(np.atleast_2d(xt[0])) dyp = M.predict_derivatives(np.atleast_2d(xt[0]), kx=0) self.assert_error(yp, np.array([[0.015368, 0.367424]]), atol=1e-2, rtol=1e-2) self.assert_error( dyp, np.array([[1.16130832e-03, 4.36712162e00]]), atol=3e-1, rtol=1e-2 ) def test_KRG_KRG_mult(self): with Silence(): M, xt = setupCRM( LF_candidate="KRG", Bridge_candidate="KRG", type_bridge="Multiplicative" ) with Silence(): yp = M.predict_values(np.atleast_2d(xt[0])) dyp = M.predict_derivatives(np.atleast_2d(xt[0]), kx=0) self.assert_error(yp, np.array([[0.015368, 0.367424]]), atol=2e-2, rtol=3e-2) self.assert_error(dyp, np.array([[0.07007729, 3.619421]]), atol=3e-1, rtol=1e-2) def test_QP_KRG_mult(self): with Silence(): M, xt = setupCRM( LF_candidate="QP", Bridge_candidate="KRG", type_bridge="Multiplicative" ) with Silence(): yp = M.predict_values(np.atleast_2d(xt[0])) dyp = M.predict_derivatives(np.atleast_2d(xt[0]), kx=0) self.assert_error( yp, np.array([[0.01537882, 0.36681699]]), atol=3e-1, rtol=1e-2 ) self.assert_error( dyp, np.array([[0.21520949, 4.50217261]]), atol=3e-1, rtol=1e-2 ) if __name__ == "__main__": unittest.main()
7,530
29.366935
88
py
smt
smt-master/smt/applications/tests/test_mfk_1fidelity.py
import matplotlib matplotlib.use("Agg") import unittest import numpy as np import unittest from smt.problems import TensorProduct from smt.sampling_methods import LHS from smt.utils.sm_test_case import SMTestCase from smt.utils.silence import Silence from smt.utils import compute_rms_error from smt.applications.mfk import MFK print_output = False class TestMFKOneFidelity(SMTestCase): def setUp(self): self.nt = 20 self.ne = 50 self.ndim = 1 def test_mfk_1fidelity(self): self.problems = ["exp", "tanh", "cos"] for fname in self.problems: prob = TensorProduct(ndim=self.ndim, func=fname) sampling = LHS(xlimits=prob.xlimits, random_state=0) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) for i in range(self.ndim): yt = np.concatenate((yt, prob(xt, kx=i)), axis=1) sampling = LHS(xlimits=prob.xlimits, random_state=1) xv = sampling(self.ne) yv = prob(xv) sm = MFK( theta0=[1e-2] * self.ndim, print_global=False, ) sm.set_training_values(xt, yt[:, 0]) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xv, yv) self.assert_error(t_error, 0.0, 1e-6) self.assert_error(e_error, 0.0, 1e-6) @staticmethod def run_mfk_example_1fidelity(): import numpy as np import matplotlib.pyplot as plt from smt.applications.mfk import MFK, NestedLHS # Consider only 1 fidelity level # high fidelity model def hf_function(x): import numpy as np return ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) # Problem set up xlimits = np.array([[0.0, 1.0]]) xdoes = NestedLHS(nlevel=1, xlimits=xlimits, random_state=0) xt_e = xdoes(7)[0] # Evaluate the HF function yt_e = hf_function(xt_e) sm = MFK(theta0=xt_e.shape[1] * [1.0]) # High-fidelity dataset without name sm.set_training_values(xt_e, yt_e) # Train the model sm.train() x = np.linspace(0, 1, 101, endpoint=True).reshape(-1, 1) # Query the outputs y = sm.predict_values(x) mse = sm.predict_variances(x) derivs = sm.predict_derivatives(x, kx=0) plt.figure() plt.plot(x, hf_function(x), label="reference") plt.plot(x, y, linestyle="-.", label="mean_gp") plt.scatter(xt_e, yt_e, marker="o", color="k", label="HF doe") plt.legend(loc=0) plt.ylim(-10, 17) plt.xlim(-0.1, 1.1) plt.xlabel(r"$x$") plt.ylabel(r"$y$") plt.show() if __name__ == "__main__": unittest.main()
2,882
24.741071
70
py
smt
smt-master/smt/applications/tests/test_mfk.py
# -*- coding: utf-8 -*- """ Created on Mon May 07 14:20:11 2018 @author: m.meliani """ import matplotlib matplotlib.use("Agg") import unittest import numpy as np import unittest import inspect from collections import OrderedDict from smt.problems import Sphere, TensorProduct from smt.sampling_methods import LHS, FullFactorial from smt.utils.sm_test_case import SMTestCase from smt.utils.silence import Silence from smt.utils import compute_rms_error from smt.surrogate_models import LS, QP, KPLS, KRG, KPLSK, GEKPLS, GENN from smt.applications.mfk import MFK, NestedLHS from copy import deepcopy print_output = False class TestMFK(SMTestCase): def setUp(self): self.nt = 100 self.ne = 100 self.ndim = 3 def test_nested_lhs(self): xlimits = np.array([[0.0, 1.0], [0.0, 1.0]]) xnorm = NestedLHS(nlevel=3, xlimits=xlimits, random_state=0) xlow, xmedium, xhigh = xnorm(15) for items1 in xmedium: found = False for items0 in xlow: if items1.all() == items0.all(): found = True self.assertTrue(found) for items1 in xhigh: found = False for items0 in xmedium: if items1.all() == items0.all(): found = True self.assertTrue(found) def test_mfk(self): self.problems = ["exp", "tanh", "cos"] for fname in self.problems: prob = TensorProduct(ndim=self.ndim, func=fname) sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) for i in range(self.ndim): yt = np.concatenate((yt, prob(xt, kx=i)), axis=1) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) sm = MFK(theta0=[1e-2] * self.ndim) if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False sm.set_training_values(xt, yt[:, 0]) sm.set_training_values(x_lf, y_lf[:, 0], name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) self.assert_error(t_error, 0.0, 1) self.assert_error(e_error, 0.0, 1) def test_mfk_derivs(self): prob = Sphere(ndim=self.ndim) sampling = LHS(xlimits=prob.xlimits) nt = 500 np.random.seed(0) xt = sampling(nt) yt = prob(xt) dyt = {} for kx in range(prob.xlimits.shape[0]): dyt[kx] = prob(xt, kx=kx) y_lf = 2 * prob(xt) + 2 x_lf = deepcopy(xt) np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) dye = {} for kx in range(prob.xlimits.shape[0]): dye[kx] = prob(xe, kx=kx) sm = MFK(theta0=[1e-2] * self.ndim) if sm.options.is_declared("xlimits"): sm.options["xlimits"] = prob.xlimits sm.options["print_global"] = False sm.set_training_values(xt, yt) sm.set_training_values(x_lf, y_lf, name=0) with Silence(): sm.train() t_error = compute_rms_error(sm) e_error = compute_rms_error(sm, xe, ye) e_error0 = compute_rms_error(sm, xe, dye[0], 0) e_error1 = compute_rms_error(sm, xe, dye[1], 1) if print_output: print( "%8s %6s %18.9e %18.9e %18.9e %18.9e" % (pname[:6], sname, t_error, e_error, e_error0, e_error1) ) self.assert_error(e_error0, 0.0, 1e-1) self.assert_error(e_error1, 0.0, 1e-1) @staticmethod def run_mfk_example(): import numpy as np import matplotlib.pyplot as plt from smt.applications.mfk import MFK, NestedLHS # low fidelity model def lf_function(x): import numpy as np return ( 0.5 * ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) + (x - 0.5) * 10.0 - 5 ) # high fidelity model def hf_function(x): import numpy as np return ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) # Problem set up xlimits = np.array([[0.0, 1.0]]) xdoes = NestedLHS(nlevel=2, xlimits=xlimits, random_state=0) xt_c, xt_e = xdoes(7) # Evaluate the HF and LF functions yt_e = hf_function(xt_e) yt_c = lf_function(xt_c) sm = MFK(theta0=xt_e.shape[1] * [1.0]) # low-fidelity dataset names being integers from 0 to level-1 sm.set_training_values(xt_c, yt_c, name=0) # high-fidelity dataset without name sm.set_training_values(xt_e, yt_e) # train the model sm.train() x = np.linspace(0, 1, 101, endpoint=True).reshape(-1, 1) # query the outputs y = sm.predict_values(x) mse = sm.predict_variances(x) derivs = sm.predict_derivatives(x, kx=0) plt.figure() plt.plot(x, hf_function(x), label="reference") plt.plot(x, y, linestyle="-.", label="mean_gp") plt.scatter(xt_e, yt_e, marker="o", color="k", label="HF doe") plt.scatter(xt_c, yt_c, marker="*", color="g", label="LF doe") plt.legend(loc=0) plt.ylim(-10, 17) plt.xlim(-0.1, 1.1) plt.xlabel(r"$x$") plt.ylabel(r"$y$") plt.show() if __name__ == "__main__": unittest.main()
5,714
26.742718
74
py
smt
smt-master/smt/applications/tests/__init__.py
0
0
0
py
smt
smt-master/smt/applications/tests/test_mixed_integer.py
""" Created on Tue Oct 12 10:48:01 2021 @author: psaves """ import unittest import numpy as np import matplotlib import itertools matplotlib.use("Agg") from smt.applications.mixed_integer import ( MixedIntegerContext, MixedIntegerSamplingMethod, MixedIntegerKrigingModel, ) from smt.problems import Sphere, HierarchicalGoldstein, HierarchicalNeuralNetwork from smt.utils.design_space import ( DesignSpace, FloatVariable, IntegerVariable, OrdinalVariable, CategoricalVariable, ) from smt.sampling_methods import LHS from smt.surrogate_models import ( KRG, KPLS, QP, MixIntKernelType, MixHrcKernelType, ) class TestMixedInteger(unittest.TestCase): def test_krg_mixed_3D(self): design_space = DesignSpace( [ FloatVariable(-10, 10), CategoricalVariable(["blue", "red", "green"]), IntegerVariable(-10, 10), ] ) mixint = MixedIntegerContext(design_space) sm = mixint.build_kriging_model(KRG(print_prediction=False)) sampling = mixint.build_sampling_method(LHS, criterion="m") fun = Sphere(ndim=3) xt = sampling(20) yt = fun(xt) sm.set_training_values(xt, yt) sm.train() eq_check = True for i in range(xt.shape[0]): if abs(float(xt[i, :][2]) - int(float(xt[i, :][2]))) > 10e-8: eq_check = False if not (xt[i, :][1] == 0 or xt[i, :][1] == 1 or xt[i, :][1] == 2): eq_check = False self.assertTrue(eq_check) def test_krg_mixed_3D_bad_regr(self): design_space = DesignSpace( [ FloatVariable(-10, 10), CategoricalVariable(["blue", "red", "green"]), IntegerVariable(-10, 10), ] ) mixint = MixedIntegerContext(design_space) with self.assertRaises(ValueError): sm = mixint.build_kriging_model(KRG(print_prediction=False, poly="linear")) def test_qp_mixed_2D_INT(self): design_space = DesignSpace( [ FloatVariable(-10, 10), IntegerVariable(-10, 10), ] ) mixint = MixedIntegerContext(design_space) sm = mixint.build_surrogate_model(QP(print_prediction=False)) sampling = mixint.build_sampling_method(LHS, criterion="m") fun = Sphere(ndim=2) xt = sampling(10) yt = fun(xt) sm.set_training_values(xt, yt) sm.train() eq_check = True for i in range(xt.shape[0]): if abs(float(xt[i, :][1]) - int(float(xt[i, :][1]))) > 10e-8: eq_check = False self.assertTrue(eq_check) def test_compute_unfolded_dimension(self): design_space = DesignSpace( [ FloatVariable(0, 1), CategoricalVariable(["A", "B"]), ] ) assert design_space.get_unfolded_num_bounds().shape[0] == 3 def test_unfold_with_enum_mask(self): x = np.array([[1.5, 1], [1.5, 0], [1.5, 1]]) expected = [[1.5, 0, 1], [1.5, 1, 0], [1.5, 0, 1]] design_space = DesignSpace( [ FloatVariable(1, 2), CategoricalVariable(["A", "B"]), ] ) x_unfolded, _ = design_space.unfold_x(x) self.assertListEqual(expected, x_unfolded.tolist()) def test_unfold_with_enum_mask_with_enum_first(self): x = np.array([[1, 1.5], [0, 1.5], [1, 1.5]]) expected = [[0, 1, 1.5], [1, 0, 1.5], [0, 1, 1.5]] design_space = DesignSpace( [ CategoricalVariable(["A", "B"]), FloatVariable(1, 2), ] ) x_unfolded, _ = design_space.unfold_x(x) self.assertListEqual(expected, x_unfolded.tolist()) def test_fold_with_enum_index(self): x = np.array([[1.5, 0, 1], [1.5, 1, 0], [1.5, 0, 1]]) expected = [[1.5, 1], [1.5, 0], [1.5, 1]] design_space = DesignSpace( [ FloatVariable(1, 2), CategoricalVariable(["A", "B"]), ] ) x_folded, _ = design_space.fold_x(x) self.assertListEqual(expected, x_folded.tolist()) def test_fold_with_enum_index_with_list(self): expected = [[1.5, 1]] x = np.array([1.5, 0, 1]) design_space = DesignSpace( [ FloatVariable(1, 2), CategoricalVariable(["A", "B"]), ] ) x_folded, _ = design_space.fold_x(x) self.assertListEqual(expected, x_folded.tolist()) x = [1.5, 0, 1] x_folded, _ = design_space.fold_x(x) self.assertListEqual(expected, x_folded.tolist()) def test_cast_to_enum_value(self): design_space = DesignSpace( [ FloatVariable(0, 4), CategoricalVariable(["blue", "red"]), ] ) x = np.zeros((5, 2)) x[:, 1] = [1, 1, 0, 1, 0] decoded = design_space.decode_values(x, i_dv=1) expected = ["red", "red", "blue", "red", "blue"] self.assertListEqual(expected, decoded) def test_unfolded_xlimits_type(self): design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["2", "3"]), CategoricalVariable(["4", "5"]), IntegerVariable(0, 2), ] ) sampling = MixedIntegerSamplingMethod(LHS, design_space, criterion="ese") doe = sampling(10) self.assertEqual((10, 4), doe.shape) def test_unfold_xlimits_with_continuous_limits(self): design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), IntegerVariable(0, 2), ] ) unfolded_limits = design_space.get_unfolded_num_bounds() self.assertEqual( np.array_equal( [[-5, 5], [0, 1], [0, 1], [0, 1], [0, 1], [0, 1], [0, 2]], unfolded_limits, ), True, ) def test_unfold_xlimits_with_continuous_limits_and_ordinal_values(self): design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), OrdinalVariable(["0", "3", "4"]), ] ) unfolded_limits = design_space.get_unfolded_num_bounds() self.assertEqual( np.array_equal( [[-5, 5], [0, 1], [0, 1], [0, 1], [0, 1], [0, 1], [0, 2]], unfolded_limits, ), True, ) def test_cast_to_discrete_values(self): design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), IntegerVariable(0, 4), ] ) x = np.array([[2.6, 0.3, 0.5, 0.25, 0.45, 0.85, 3.1]]) self.assertEqual( np.array_equal( np.array([[2.6, 0, 1, 0, 0, 1, 3]]), design_space.correct_get_acting(x)[0], ), True, ) def test_cast_to_discrete_values_with_smooth_rounding_ordinal_values(self): x = np.array([[2.6, 0.3, 0.5, 0.25, 0.45, 0.85, 1.1]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), OrdinalVariable(["0", "2", "4"]), ] ) self.assertEqual( np.array_equal( np.array([[2.6, 0, 1, 0, 0, 1, 1]]), design_space.correct_get_acting(x)[0], ), True, ) def test_cast_to_discrete_values_with_hard_rounding_ordinal_values(self): x = np.array([[2.6, 0.3, 0.5, 0.25, 0.45, 0.85, 0.9]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), OrdinalVariable(["0", "4"]), ] ) self.assertEqual( np.array_equal( np.array([[2.6, 0, 1, 0, 0, 1, 1]]), design_space.correct_get_acting(x)[0], ), True, ) def test_cast_to_discrete_values_with_non_integer_ordinal_values(self): x = np.array([[2.6, 0.3, 0.5, 0.25, 0.45, 0.85, 0.8]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["blue", "red"]), CategoricalVariable(["short", "medium", "long"]), OrdinalVariable(["0", "3.5"]), ] ) self.assertEqual( np.array_equal( np.array([[2.6, 0, 1, 0, 0, 1, 1]]), design_space.correct_get_acting(x)[0], ), True, ) def test_examples(self): self.run_mixed_integer_lhs_example() self.run_mixed_integer_qp_example() self.run_mixed_integer_context_example() self.run_hierarchical_variables_Goldstein() self.run_mixed_discrete_design_space_example() self.run_hierarchical_design_space_example() def run_mixed_integer_lhs_example(self): import numpy as np import matplotlib.pyplot as plt from matplotlib import colors from smt.utils.design_space import ( DesignSpace, FloatVariable, CategoricalVariable, ) float_var = FloatVariable(0, 4) cat_var = CategoricalVariable(["blue", "red"]) design_space = DesignSpace( [ float_var, cat_var, ] ) num = 40 x, x_is_acting = design_space.sample_valid_x(num) cmap = colors.ListedColormap(cat_var.values) plt.scatter(x[:, 0], np.zeros(num), c=x[:, 1], cmap=cmap) plt.show() def run_mixed_integer_qp_example(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import QP from smt.applications.mixed_integer import MixedIntegerSurrogateModel from smt.utils.design_space import DesignSpace, IntegerVariable xt = np.array([0.0, 1.0, 2.0, 3.0, 4.0]) yt = np.array([0.0, 1.0, 1.5, 0.5, 1.0]) # Specify the design space using the DesignSpace # class and various available variable types design_space = DesignSpace( [ IntegerVariable(0, 4), ] ) sm = MixedIntegerSurrogateModel(design_space=design_space, surrogate=QP()) sm.set_training_values(xt, yt) sm.train() num = 100 x = np.linspace(0.0, 4.0, num) y = sm.predict_values(x) plt.plot(xt, yt, "o") plt.plot(x, y) plt.xlabel("x") plt.ylabel("y") plt.legend(["Training data", "Prediction"]) plt.show() def run_mixed_integer_context_example(self): import matplotlib.pyplot as plt from smt.surrogate_models import KRG from smt.applications.mixed_integer import MixedIntegerContext from smt.utils.design_space import ( DesignSpace, FloatVariable, IntegerVariable, CategoricalVariable, ) design_space = DesignSpace( [ IntegerVariable(0, 5), FloatVariable(0.0, 4.0), CategoricalVariable(["blue", "red", "green", "yellow"]), ] ) def ftest(x): return (x[:, 0] * x[:, 0] + x[:, 1] * x[:, 1]) * (x[:, 2] + 1) # Helper class for creating surrogate models mi_context = MixedIntegerContext(design_space) # DOE for training sampler = mi_context.build_sampling_method() num = mi_context.get_unfolded_dimension() * 5 print("DOE point nb = {}".format(num)) xt = sampler(num) yt = ftest(xt) # Surrogate sm = mi_context.build_kriging_model(KRG()) sm.set_training_values(xt, yt) sm.train() # DOE for validation xv = sampler(50) yv = ftest(xv) yp = sm.predict_values(xv) plt.plot(yv, yv) plt.plot(yv, yp, "o") plt.xlabel("actual") plt.ylabel("prediction") plt.show() def test_hierarchical_variables_Goldstein(self): problem = HierarchicalGoldstein() ds = problem.design_space self.assertIsInstance(ds, DesignSpace) self.assertEqual(ds.n_dv, 11) x = np.array( [ [0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], ] ) y = problem(x) self.assertTrue( np.linalg.norm( y - np.array([50.75285716, 56.62074043, 50.97693309, 56.29235443]) ) < 1e-8 ) self.assertTrue( np.linalg.norm( problem.eval_is_acting.astype(int) - np.array( [ [1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1], [1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1], [1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1], ] ) ) < 1e-8 ) self.assertTrue( np.linalg.norm( problem.eval_x - np.array( [ [0, 1, 1, 1, 50, 50, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 50, 1, 0, 1, 1, 1], [2, 1, 1, 1, 50, 1, 1, 1, 0, 1, 1], [3, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1], ] ) ) < 1e-8 ) self.assertTrue(np.linalg.norm(y - problem(problem.eval_x)) < 1e-8) n_doe = 15 ds.seed = 42 Xt, is_acting = ds.sample_valid_x(n_doe) Yt = problem(Xt) sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=ds, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, hierarchical_kernel=MixHrcKernelType.ARC_KERNEL, theta0=[1e-2], corr="abs_exp", n_start=10, ), ) sm.set_training_values(Xt, Yt, is_acting=is_acting) sm.train() y_s = sm.predict_values(Xt)[:, 0] pred_RMSE = np.linalg.norm(y_s - Yt) / len(Yt) y_sv = sm.predict_variances(Xt)[:, 0] var_RMSE = np.linalg.norm(y_sv) / len(Yt) self.assertTrue(pred_RMSE < 1e-7) print("Pred_RMSE", pred_RMSE) self.assertTrue(var_RMSE < 1e-7) self.assertTrue( np.linalg.norm( sm.predict_values( np.array( [ [0.0, 1.0, 64.0, 4.0, 56.0, 37.0, 35.0, 1.0, 2.0, 1.0, 1.0], [1.0, 0.0, 31.0, 92.0, 24.0, 3.0, 17.0, 1.0, 2.0, 1.0, 1.0], [2.0, 1.0, 28.0, 60.0, 77.0, 66.0, 9.0, 0.0, 1.0, 1.0, 1.0], [ 3.0, 1.0, 50.0, 40.0, 99.0, 35.0, 51.0, 2.0, 1.0, 1.0, 2.0, ], ] ) )[:, 0] - sm.predict_values( np.array( [ [0.0, 1.0, 64.0, 4.0, 6.0, 7.0, 35.0, 1.0, 2.0, 1.0, 1.0], [ 1.0, 0.0, 31.0, 92.0, 24.0, 30.0, 17.0, 0.0, 2.0, 1.0, 1.0, ], [2.0, 1.0, 28.0, 60.0, 7.0, 66.0, 9.0, 0.0, 2.0, 1.0, 1.0], [ 3.0, 1.0, 50.0, 40.0, 99.0, 35.0, 51.0, 0.0, 0.0, 1.0, 2.0, ], ] ) )[:, 0] ) < 1e-8 ) self.assertTrue( np.linalg.norm( sm.predict_values( np.array( [[1.0, 0.0, 31.0, 92.0, 24.0, 3.0, 17.0, 1.0, 2.0, 1.0, 1.0]] ) ) - sm.predict_values( np.array( [[1.0, 0.0, 31.0, 92.0, 24.0, 3.0, 17.0, 1.0, 1.0, 1.0, 1.0]] ) ) ) > 1e-8 ) def run_mixed_discrete_design_space_example(self): import numpy as np from smt.utils.design_space import ( DesignSpace, FloatVariable, IntegerVariable, OrdinalVariable, CategoricalVariable, ) ds = DesignSpace( [ CategoricalVariable( ["A", "B"] ), # x0 categorical: A or B; order is not relevant OrdinalVariable( ["C", "D", "E"] ), # x1 ordinal: C, D or E; order is relevant IntegerVariable( 0, 2 ), # x2 integer between 0 and 2 (inclusive): 0, 1, 2 FloatVariable(0, 1), # c3 continuous between 0 and 1 ] ) # Sample the design space # Note: is_acting_sampled specifies for each design variable whether it is acting or not x_sampled, is_acting_sampled = ds.sample_valid_x(100) # Correct design vectors: round discrete variables, correct hierarchical variables x_corr, is_acting = ds.correct_get_acting( np.array( [ [0, 0, 2, 0.25], [0, 2, 1, 0.75], ] ) ) print(is_acting) def run_hierarchical_design_space_example(self): import numpy as np from smt.utils.design_space import ( DesignSpace, FloatVariable, IntegerVariable, OrdinalVariable, CategoricalVariable, ) ds = DesignSpace( [ CategoricalVariable( ["A", "B"] ), # x0 categorical: A or B; order is not relevant OrdinalVariable( ["C", "D", "E"] ), # x1 ordinal: C, D or E; order is relevant IntegerVariable( 0, 2 ), # x2 integer between 0 and 2 (inclusive): 0, 1, 2 FloatVariable(0, 1), # c3 continuous between 0 and 1 ] ) # Declare that x1 is acting if x0 == A ds.declare_decreed_var(decreed_var=1, meta_var=0, meta_value="A") # Sample the design space # Note: is_acting_sampled specifies for each design variable whether it is acting or not x_sampled, is_acting_sampled = ds.sample_valid_x(100) # Correct design vectors: round discrete variables, correct hierarchical variables x_corr, is_acting = ds.correct_get_acting( np.array( [ [0, 0, 2, 0.25], [1, 2, 1, 0.66], ] ) ) # Observe the hierarchical behavior: assert np.all( is_acting == np.array( [ [True, True, True, True], [True, False, True, True], # x1 is not acting if x0 != A ] ) ) assert np.all( x_corr == np.array( [ [0, 0, 2, 0.25], # x1 is not acting, so it is corrected ("imputed") to its non-acting value (0 for discrete vars) [1, 0, 1, 0.66], ] ) ) def run_hierarchical_variables_Goldstein(self): import numpy as np from smt.utils.design_space import ( DesignSpace, CategoricalVariable, IntegerVariable, FloatVariable, ) from smt.applications.mixed_integer import MixedIntegerKrigingModel from smt.surrogate_models import MixIntKernelType, MixHrcKernelType, KRG def f_hv(X): import numpy as np def H(x1, x2, x3, x4, z3, z4, x5, cos_term): import numpy as np h = ( 53.3108 + 0.184901 * x1 - 5.02914 * x1**3 * 10 ** (-6) + 7.72522 * x1**z3 * 10 ** (-8) - 0.0870775 * x2 - 0.106959 * x3 + 7.98772 * x3**z4 * 10 ** (-6) + 0.00242482 * x4 + 1.32851 * x4**3 * 10 ** (-6) - 0.00146393 * x1 * x2 - 0.00301588 * x1 * x3 - 0.00272291 * x1 * x4 + 0.0017004 * x2 * x3 + 0.0038428 * x2 * x4 - 0.000198969 * x3 * x4 + 1.86025 * x1 * x2 * x3 * 10 ** (-5) - 1.88719 * x1 * x2 * x4 * 10 ** (-6) + 2.50923 * x1 * x3 * x4 * 10 ** (-5) - 5.62199 * x2 * x3 * x4 * 10 ** (-5) ) if cos_term: h += 5.0 * np.cos(2.0 * np.pi * (x5 / 100.0)) - 2.0 return h def f1(x1, x2, z1, z2, z3, z4, x5, cos_term): c1 = z2 == 0 c2 = z2 == 1 c3 = z2 == 2 c4 = z3 == 0 c5 = z3 == 1 c6 = z3 == 2 y = ( c4 * ( c1 * H(x1, x2, 20, 20, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 20, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 20, z3, z4, x5, cos_term) ) + c5 * ( c1 * H(x1, x2, 20, 50, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 50, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 50, z3, z4, x5, cos_term) ) + c6 * ( c1 * H(x1, x2, 20, 80, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, 80, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, 80, z3, z4, x5, cos_term) ) ) return y def f2(x1, x2, x3, z2, z3, z4, x5, cos_term): c1 = z2 == 0 c2 = z2 == 1 c3 = z2 == 2 y = ( c1 * H(x1, x2, x3, 20, z3, z4, x5, cos_term) + c2 * H(x1, x2, x3, 50, z3, z4, x5, cos_term) + c3 * H(x1, x2, x3, 80, z3, z4, x5, cos_term) ) return y def f3(x1, x2, x4, z1, z3, z4, x5, cos_term): c1 = z1 == 0 c2 = z1 == 1 c3 = z1 == 2 y = ( c1 * H(x1, x2, 20, x4, z3, z4, x5, cos_term) + c2 * H(x1, x2, 50, x4, z3, z4, x5, cos_term) + c3 * H(x1, x2, 80, x4, z3, z4, x5, cos_term) ) return y y = [] for x in X: if x[0] == 0: y.append( f1(x[2], x[3], x[7], x[8], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 1: y.append( f2(x[2], x[3], x[4], x[8], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 2: y.append( f3(x[2], x[3], x[5], x[7], x[9], x[10], x[6], cos_term=x[1]) ) elif x[0] == 3: y.append( H(x[2], x[3], x[4], x[5], x[9], x[10], x[6], cos_term=x[1]) ) return np.array(y) design_space = DesignSpace( [ CategoricalVariable(values=[0, 1, 2, 3]), # meta IntegerVariable(0, 1), # x1 FloatVariable(0, 100), # x2 FloatVariable(0, 100), FloatVariable(0, 100), FloatVariable(0, 100), FloatVariable(0, 100), IntegerVariable(0, 2), # x7 IntegerVariable(0, 2), IntegerVariable(0, 2), IntegerVariable(0, 2), ] ) # x4 is acting if meta == 1, 3 design_space.declare_decreed_var(decreed_var=4, meta_var=0, meta_value=[1, 3]) # x5 is acting if meta == 2, 3 design_space.declare_decreed_var(decreed_var=5, meta_var=0, meta_value=[2, 3]) # x7 is acting if meta == 0, 2 design_space.declare_decreed_var(decreed_var=7, meta_var=0, meta_value=[0, 2]) # x8 is acting if meta == 0, 1 design_space.declare_decreed_var(decreed_var=8, meta_var=0, meta_value=[0, 1]) # Sample from the design spaces, correctly considering hierarchy n_doe = 15 Xt, Xt_is_acting = design_space.sample_valid_x(n_doe) Yt = f_hv(Xt) sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, hierarchical_kernel=MixHrcKernelType.ALG_KERNEL, # ALG or ARC theta0=[1e-2], corr="abs_exp", n_start=5, ), ) sm.set_training_values(Xt, Yt, is_acting=Xt_is_acting) sm.train() y_s = sm.predict_values(Xt)[:, 0] pred_RMSE = np.linalg.norm(y_s - Yt) / len(Yt) y_sv = sm.predict_variances(Xt)[:, 0] var_RMSE = np.linalg.norm(y_sv) / len(Yt) def test_hierarchical_variables_NN(self): problem = HierarchicalNeuralNetwork() ds = problem.design_space self.assertEqual(ds.n_dv, 8) n_doe = 100 ds_sampling = DesignSpace(ds.design_variables[1:]) sampling = MixedIntegerSamplingMethod( LHS, ds_sampling, criterion="ese", random_state=42 ) x_cont = sampling(3 * n_doe) xdoe1 = np.zeros((n_doe, 8)) x_cont2 = x_cont[:n_doe, :5] xdoe1[:, 0] = np.zeros(n_doe) xdoe1[:, 1:6] = x_cont2 ydoe1 = problem(xdoe1) xdoe1 = np.zeros((n_doe, 8)) xdoe1[:, 0] = np.zeros(n_doe) xdoe1[:, 1:6] = x_cont2 xdoe2 = np.zeros((n_doe, 8)) x_cont2 = x_cont[n_doe : 2 * n_doe, :6] xdoe2[:, 0] = np.ones(n_doe) xdoe2[:, 1:7] = x_cont2 ydoe2 = problem(xdoe2) xdoe2 = np.zeros((n_doe, 8)) xdoe2[:, 0] = np.ones(n_doe) xdoe2[:, 1:7] = x_cont2 xdoe3 = np.zeros((n_doe, 8)) xdoe3[:, 0] = 2 * np.ones(n_doe) xdoe3[:, 1:] = x_cont[2 * n_doe :, :] ydoe3 = problem(xdoe3) Xt = np.concatenate((xdoe1, xdoe2, xdoe3), axis=0) Yt = np.concatenate((ydoe1, ydoe2, ydoe3), axis=0) sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=problem.design_space, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, hierarchical_kernel=MixHrcKernelType.ALG_KERNEL, theta0=[1e-2], corr="abs_exp", n_start=5, ), ) sm.set_training_values(Xt, Yt) sm.train() y_s = sm.predict_values(Xt)[:, 0] pred_RMSE = np.linalg.norm(y_s - Yt) / len(Yt) y_sv = sm.predict_variances(Xt)[:, 0] var_RMSE = np.linalg.norm(y_sv) / len(Yt) self.assertTrue(pred_RMSE < 1e-7) print("Pred_RMSE", pred_RMSE) self.assertTrue(var_RMSE < 1e-7) self.assertTrue( np.linalg.norm( sm.predict_values( np.array( [ [0, -1, -2, 8, 0, 2, 0, 0], [1, -1, -2, 16, 1, 2, 1, 0], [2, -1, -2, 32, 2, 2, 1, -2], ] ) )[:, 0] - sm.predict_values( np.array( [ [0, -1, -2, 8, 0, 2, 10, 10], [1, -1, -2, 16, 1, 2, 1, 10], [2, -1, -2, 32, 2, 2, 1, -2], ] ) )[:, 0] ) < 1e-8 ) self.assertTrue( np.linalg.norm( sm.predict_values(np.array([[0, -1, -2, 8, 0, 2, 0, 0]])) - sm.predict_values(np.array([[0, -1, -2, 8, 0, 12, 10, 10]])) ) > 1e-8 ) def test_mixed_gower_2D(self): xt = np.array([[0, 5], [2, -1], [4, 0.5]]) yt = np.array([[0.0], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-2], corr="abs_exp", categorical_kernel=MixIntKernelType.GOWER, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x, x2): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertAlmostEqual(y[20, 0], 0) self.assertAlmostEqual(y[50, 0], 1) self.assertAlmostEqual(y[95, 0], 1.5) self.assertTrue(np.abs(np.sum(np.array([y[20], y[50], y[95]]) - yt)) < 1e-6) self.assertTrue(np.abs(np.sum(np.array([yvar[20], yvar[50], yvar[95]]))) < 1e-6) self.assertEqual(np.shape(y), (105, 1)) def test_mixed_CR_2D(self): xt = np.array([[0, 5], [2, -1], [4, 0.5]]) yt = np.array([[0.0], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-2], corr="abs_exp", categorical_kernel=MixIntKernelType.GOWER, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x, x2): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertAlmostEqual(y[20, 0], 0) self.assertAlmostEqual(y[50, 0], 1) self.assertAlmostEqual(y[95, 0], 1.5) self.assertTrue(np.abs(np.sum(np.array([y[20], y[50], y[95]]) - yt)) < 1e-6) self.assertTrue(np.abs(np.sum(np.array([yvar[20], yvar[50], yvar[95]]))) < 1e-6) self.assertEqual(np.shape(y), (105, 1)) def test_mixed_CR_PLS_noisy_2D(self): xt = np.array([[0, 5], [2, -1], [4, 0.5]]) yt = np.array([[0.0], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KPLS( n_comp=1, eval_noise=True, design_space=design_space, theta0=[1e-2], categorical_kernel=MixIntKernelType.CONT_RELAX, corr="abs_exp", ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x, x2): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertTrue(np.abs(np.sum(np.array([y[20], y[50], y[95]]) - yt)) < 1e-6) self.assertTrue(np.abs(np.sum(np.array([yvar[20], yvar[50], yvar[95]]))) < 1e-6) self.assertEqual(np.shape(y), (105, 1)) def test_mixed_homo_gaussian_2D(self): xt = np.array([[0, 5], [2, -1], [4, 0.5]]) yt = np.array([[0.0], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-2], corr="abs_exp", categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x, x2): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertTrue(np.abs(np.sum(np.array([y[20], y[50], y[95]]) - yt)) < 1e-6) self.assertTrue(np.abs(np.sum(np.array([yvar[20], yvar[50], yvar[95]]))) < 1e-6) self.assertEqual(np.shape(y), (105, 1)) def test_mixed_homo_hyp_2D(self): xt = np.array([[0, 5], [2, -1], [4, 0.5]]) yt = np.array([[0.0], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-2], categorical_kernel=MixIntKernelType.HOMO_HSPHERE, corr="abs_exp", ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x, x2): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertTrue(np.abs(np.sum(np.array([y[20], y[50], y[95]]) - yt)) < 1e-6) self.assertTrue(np.abs(np.sum(np.array([yvar[20], yvar[50], yvar[95]]))) < 1e-6) self.assertEqual(np.shape(y), (105, 1)) def test_mixed_homo_gaussian_3D_PLS(self): xt = np.array([[0.5, 0, 5], [2, 3, 4], [5, 2, -1], [-2, 4, 0.5]]) yt = np.array([[0.0], [3], [1.0], [1.5]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = surrogate = KPLS( design_space=design_space, theta0=[1e-2], n_comp=1, categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, cat_kernel_comps=[3], corr="squar_exp", ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x2, x, x2): x1.append(np.array(element)) x_pred = np.array(x1) i = 0 i += 1 y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) self.assertTrue((np.abs(np.sum(np.array(sm.predict_values(xt) - yt)))) < 1e-6) self.assertTrue((np.abs(np.sum(np.array(sm.predict_variances(xt) - 0)))) < 1e-6) def test_mixed_homo_gaussian_3D_PLS_cate(self): xt = np.array([[0.5, 0, 5], [2, 3, 4], [5, 2, -1], [-2, 4, 0.5]]) yt = np.array([[0.0], [3], [1.0], [1.5]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = KPLS( design_space=design_space, theta0=[1e-2], n_comp=2, corr="abs_exp", cat_kernel_comps=[3], categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x2, x, x2): x1.append(np.array(element)) x_pred = np.array(x1) i = 0 i += 1 y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) self.assertTrue((np.abs(np.sum(np.array(sm.predict_values(xt) - yt)))) < 1e-6) self.assertTrue((np.abs(np.sum(np.array(sm.predict_variances(xt) - 0)))) < 1e-6) def test_mixed_homo_hyp_3D_PLS_cate(self): xt = np.array([[0.5, 0, 5], [2, 3, 4], [5, 2, -1], [-2, 4, 0.5]]) yt = np.array([[0.0], [3], [1.0], [1.5]]) design_space = DesignSpace( [ FloatVariable(-5, 5), CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KPLS( design_space=design_space, theta0=[1e-2], n_comp=1, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, cat_kernel_comps=[3], corr="squar_exp", ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x1 = [] for element in itertools.product(x2, x, x2): x1.append(np.array(element)) x_pred = np.array(x1) i = 0 i += 1 y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) self.assertTrue((np.abs(np.sum(np.array(sm.predict_values(xt) - yt)))) < 1e-6) self.assertTrue((np.abs(np.sum(np.array(sm.predict_variances(xt) - 0)))) < 1e-6) def test_mixed_homo_gaussian_3D_ord_cate(self): xt = np.array([[0.5, 0, 5], [2, 3, 4], [5, 2, -1], [-2, 4, 0.5]]) yt = np.array([[0.0], [3], [1.0], [1.5]]) design_space = DesignSpace( [ CategoricalVariable(["0.0", "1.0", " 2.0", "3.0", "4.0"]), FloatVariable(-5, 5), CategoricalVariable(["0.0", "1.0", " 2.0", "3.0"]), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KPLS( design_space=design_space, theta0=[1e-2], n_comp=1, categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, cat_kernel_comps=[3, 2], corr="squar_exp", ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation x = np.linspace(0, 4, 5) x2 = np.linspace(-5, 5, 21) x3 = np.linspace(0, 3, 4) x1 = [] for element in itertools.product(x, x2, x3): x1.append(np.array(element)) x_pred = np.array(x1) y = sm.predict_values(x_pred) yvar = sm.predict_variances(x_pred) # prediction are correct on known points self.assertTrue((np.abs(np.sum(np.array(sm.predict_values(xt) - yt)) < 1e-6))) self.assertTrue((np.abs(np.sum(np.array(sm.predict_variances(xt) - 0)) < 1e-6))) def test_mixed_gower_3D(self): design_space = DesignSpace( [ FloatVariable(-10, 10), IntegerVariable(-10, 10), IntegerVariable(-10, 10), ] ) mixint = MixedIntegerContext(design_space) sm = mixint.build_kriging_model( KRG(categorical_kernel=MixIntKernelType.GOWER, print_prediction=False) ) sampling = mixint.build_sampling_method(LHS, criterion="m") fun = Sphere(ndim=3) xt = sampling(10) yt = fun(xt) sm.set_training_values(xt, yt) sm.train() eq_check = True for i in range(xt.shape[0]): if abs(float(xt[i, :][1]) - int(float(xt[i, :][1]))) > 10e-8: eq_check = False self.assertTrue(eq_check) def run_mixed_gower_example(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KRG, MixIntKernelType from smt.applications.mixed_integer import MixedIntegerKrigingModel from smt.utils.design_space import ( DesignSpace, CategoricalVariable, FloatVariable, ) xt1 = np.array([[0, 0.0], [0, 2.0], [0, 4.0]]) xt2 = np.array([[1, 0.0], [1, 2.0], [1, 3.0]]) xt3 = np.array([[2, 1.0], [2, 2.0], [2, 4.0]]) xt = np.concatenate((xt1, xt2, xt3), axis=0) xt[:, 1] = xt[:, 1].astype(np.float64) yt1 = np.array([0.0, 9.0, 16.0]) yt2 = np.array([0.0, -4, -13.0]) yt3 = np.array([-10, 3, 11.0]) yt = np.concatenate((yt1, yt2, yt3), axis=0) design_space = DesignSpace( [ CategoricalVariable(["Blue", "Red", "Green"]), FloatVariable(0, 4), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, categorical_kernel=MixIntKernelType.GOWER, theta0=[1e-1], corr="squar_exp", n_start=20, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation n = 100 x_cat1 = [] x_cat2 = [] x_cat3 = [] for i in range(n): x_cat1.append(0) x_cat2.append(1) x_cat3.append(2) x_cont = np.linspace(0.0, 4.0, n) x1 = np.concatenate( (np.asarray(x_cat1).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x2 = np.concatenate( (np.asarray(x_cat2).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x3 = np.concatenate( (np.asarray(x_cat3).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) y1 = sm.predict_values(x1) y2 = sm.predict_values(x2) y3 = sm.predict_values(x3) # estimated variance s2_1 = sm.predict_variances(x1) s2_2 = sm.predict_variances(x2) s2_3 = sm.predict_variances(x3) fig, axs = plt.subplots(3, figsize=(8, 6)) axs[0].plot(xt1[:, 1].astype(np.float64), yt1, "o", linestyle="None") axs[0].plot(x_cont, y1, color="Blue") axs[0].fill_between( np.ravel(x_cont), np.ravel(y1 - 3 * np.sqrt(s2_1)), np.ravel(y1 + 3 * np.sqrt(s2_1)), color="lightgrey", ) axs[0].set_xlabel("x") axs[0].set_ylabel("y") axs[0].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[1].plot( xt2[:, 1].astype(np.float64), yt2, marker="o", color="r", linestyle="None" ) axs[1].plot(x_cont, y2, color="Red") axs[1].fill_between( np.ravel(x_cont), np.ravel(y2 - 3 * np.sqrt(s2_2)), np.ravel(y2 + 3 * np.sqrt(s2_2)), color="lightgrey", ) axs[1].set_xlabel("x") axs[1].set_ylabel("y") axs[1].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[2].plot( xt3[:, 1].astype(np.float64), yt3, marker="o", color="r", linestyle="None" ) axs[2].plot(x_cont, y3, color="Green") axs[2].fill_between( np.ravel(x_cont), np.ravel(y3 - 3 * np.sqrt(s2_3)), np.ravel(y3 + 3 * np.sqrt(s2_3)), color="lightgrey", ) axs[2].set_xlabel("x") axs[2].set_ylabel("y") axs[2].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) plt.tight_layout() plt.show() def run_mixed_homo_gaussian_example(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KRG, MixIntKernelType from smt.applications.mixed_integer import MixedIntegerKrigingModel from smt.utils.design_space import ( DesignSpace, CategoricalVariable, FloatVariable, ) xt1 = np.array([[0, 0.0], [0, 2.0], [0, 4.0]]) xt2 = np.array([[1, 0.0], [1, 2.0], [1, 3.0]]) xt3 = np.array([[2, 1.0], [2, 2.0], [2, 4.0]]) xt = np.concatenate((xt1, xt2, xt3), axis=0) xt[:, 1] = xt[:, 1].astype(np.float64) yt1 = np.array([0.0, 9.0, 16.0]) yt2 = np.array([0.0, -4, -13.0]) yt3 = np.array([-10, 3, 11.0]) yt = np.concatenate((yt1, yt2, yt3), axis=0) design_space = DesignSpace( [ CategoricalVariable(["Blue", "Red", "Green"]), FloatVariable(0, 4), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, theta0=[1e-1], corr="squar_exp", n_start=20, categorical_kernel=MixIntKernelType.EXP_HOMO_HSPHERE, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation n = 100 x_cat1 = [] x_cat2 = [] x_cat3 = [] for i in range(n): x_cat1.append(0) x_cat2.append(1) x_cat3.append(2) x_cont = np.linspace(0.0, 4.0, n) x1 = np.concatenate( (np.asarray(x_cat1).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x2 = np.concatenate( (np.asarray(x_cat2).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x3 = np.concatenate( (np.asarray(x_cat3).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) y1 = sm.predict_values(x1) y2 = sm.predict_values(x2) y3 = sm.predict_values(x3) # estimated variance s2_1 = sm.predict_variances(x1) s2_2 = sm.predict_variances(x2) s2_3 = sm.predict_variances(x3) fig, axs = plt.subplots(3, figsize=(8, 6)) axs[0].plot(xt1[:, 1].astype(np.float64), yt1, "o", linestyle="None") axs[0].plot(x_cont, y1, color="Blue") axs[0].fill_between( np.ravel(x_cont), np.ravel(y1 - 3 * np.sqrt(s2_1)), np.ravel(y1 + 3 * np.sqrt(s2_1)), color="lightgrey", ) axs[0].set_xlabel("x") axs[0].set_ylabel("y") axs[0].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[1].plot( xt2[:, 1].astype(np.float64), yt2, marker="o", color="r", linestyle="None" ) axs[1].plot(x_cont, y2, color="Red") axs[1].fill_between( np.ravel(x_cont), np.ravel(y2 - 3 * np.sqrt(s2_2)), np.ravel(y2 + 3 * np.sqrt(s2_2)), color="lightgrey", ) axs[1].set_xlabel("x") axs[1].set_ylabel("y") axs[1].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[2].plot( xt3[:, 1].astype(np.float64), yt3, marker="o", color="r", linestyle="None" ) axs[2].plot(x_cont, y3, color="Green") axs[2].fill_between( np.ravel(x_cont), np.ravel(y3 - 3 * np.sqrt(s2_3)), np.ravel(y3 + 3 * np.sqrt(s2_3)), color="lightgrey", ) axs[2].set_xlabel("x") axs[2].set_ylabel("y") axs[2].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) plt.tight_layout() plt.show() def run_mixed_homo_hyp_example(self): import numpy as np import matplotlib.pyplot as plt from smt.surrogate_models import KRG, MixIntKernelType from smt.applications.mixed_integer import MixedIntegerKrigingModel from smt.utils.design_space import ( DesignSpace, CategoricalVariable, FloatVariable, ) xt1 = np.array([[0, 0.0], [0, 2.0], [0, 4.0]]) xt2 = np.array([[1, 0.0], [1, 2.0], [1, 3.0]]) xt3 = np.array([[2, 1.0], [2, 2.0], [2, 4.0]]) xt = np.concatenate((xt1, xt2, xt3), axis=0) xt[:, 1] = xt[:, 1].astype(np.float64) yt1 = np.array([0.0, 9.0, 16.0]) yt2 = np.array([0.0, -4, -13.0]) yt3 = np.array([-10, 3, 11.0]) yt = np.concatenate((yt1, yt2, yt3), axis=0) design_space = DesignSpace( [ CategoricalVariable(["Blue", "Red", "Green"]), FloatVariable(0, 4), ] ) # Surrogate sm = MixedIntegerKrigingModel( surrogate=KRG( design_space=design_space, categorical_kernel=MixIntKernelType.HOMO_HSPHERE, theta0=[1e-1], corr="squar_exp", n_start=20, ), ) sm.set_training_values(xt, yt) sm.train() # DOE for validation n = 100 x_cat1 = [] x_cat2 = [] x_cat3 = [] for i in range(n): x_cat1.append(0) x_cat2.append(1) x_cat3.append(2) x_cont = np.linspace(0.0, 4.0, n) x1 = np.concatenate( (np.asarray(x_cat1).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x2 = np.concatenate( (np.asarray(x_cat2).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) x3 = np.concatenate( (np.asarray(x_cat3).reshape(-1, 1), x_cont.reshape(-1, 1)), axis=1 ) y1 = sm.predict_values(x1) y2 = sm.predict_values(x2) y3 = sm.predict_values(x3) # estimated variance s2_1 = sm.predict_variances(x1) s2_2 = sm.predict_variances(x2) s2_3 = sm.predict_variances(x3) fig, axs = plt.subplots(3, figsize=(8, 6)) axs[0].plot(xt1[:, 1].astype(np.float64), yt1, "o", linestyle="None") axs[0].plot(x_cont, y1, color="Blue") axs[0].fill_between( np.ravel(x_cont), np.ravel(y1 - 3 * np.sqrt(s2_1)), np.ravel(y1 + 3 * np.sqrt(s2_1)), color="lightgrey", ) axs[0].set_xlabel("x") axs[0].set_ylabel("y") axs[0].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[1].plot( xt2[:, 1].astype(np.float64), yt2, marker="o", color="r", linestyle="None" ) axs[1].plot(x_cont, y2, color="Red") axs[1].fill_between( np.ravel(x_cont), np.ravel(y2 - 3 * np.sqrt(s2_2)), np.ravel(y2 + 3 * np.sqrt(s2_2)), color="lightgrey", ) axs[1].set_xlabel("x") axs[1].set_ylabel("y") axs[1].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) axs[2].plot( xt3[:, 1].astype(np.float64), yt3, marker="o", color="r", linestyle="None" ) axs[2].plot(x_cont, y3, color="Green") axs[2].fill_between( np.ravel(x_cont), np.ravel(y3 - 3 * np.sqrt(s2_3)), np.ravel(y3 + 3 * np.sqrt(s2_3)), color="lightgrey", ) axs[2].set_xlabel("x") axs[2].set_ylabel("y") axs[2].legend( ["Training data", "Prediction", "Confidence Interval 99%"], loc="upper left", bbox_to_anchor=[0, 1], ) plt.tight_layout() plt.show() if __name__ == "__main__": TestMixedInteger().run_mixed_integer_context_example() unittest.main()
55,817
31.584939
116
py
smt
smt-master/smt/applications/tests/test_moe.py
""" Author: Remi Lafage <remi.lafage@onera.fr> This package is distributed under New BSD license. """ import matplotlib matplotlib.use("Agg") import unittest import numpy as np from sys import argv from smt.applications import MOE, MOESurrogateModel from smt.utils.sm_test_case import SMTestCase from smt.problems import Branin, LpNorm from smt.sampling_methods import FullFactorial, LHS from smt.utils.misc import compute_rms_error from smt.surrogate_models import RMTB, RMTC class TestMOE(SMTestCase): """ Test class """ plot = None @staticmethod def function_test_1d(x): x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y[x < 0.4] = x[x < 0.4] ** 2 y[(x >= 0.4) & (x < 0.8)] = 3 * x[(x >= 0.4) & (x < 0.8)] + 1 y[x >= 0.8] = np.sin(10 * x[x >= 0.8]) return y.reshape((-1, 1)) # @unittest.skip('disabled') def test_1d_50(self): self.ndim = 1 self.nt = 50 self.ne = 50 np.random.seed(0) xt = np.random.sample(self.nt).reshape((-1, 1)) yt = self.function_test_1d(xt) moe = MOE( smooth_recombination=True, heaviside_optimization=True, n_clusters=3, xt=xt, yt=yt, ) moe.train() # validation data np.random.seed(1) xe = np.random.sample(self.ne) ye = self.function_test_1d(xe) rms_error = compute_rms_error(moe, xe, ye) self.assert_error(rms_error, 0.0, 3e-1) self.assertRaises(RuntimeError, lambda: moe.predict_variances(xe)) if TestMOE.plot: import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D y = moe.predict_values(xe) plt.figure(1) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel(r"$y$ actual") plt.ylabel(r"$y$ prediction") plt.figure(2) xv = np.linspace(0, 1, 100) yv = self.function_test_1d(xv) plt.plot(xv, yv, "-.") plt.plot(xe, y, "o") plt.show() # @unittest.skip('disabled') def test_1d_50_var(self): self.ndim = 1 self.nt = 50 self.ne = 50 np.random.seed(0) xt = np.random.sample(self.nt).reshape((-1, 1)) yt = self.function_test_1d(xt) moe = MOE( smooth_recombination=True, heaviside_optimization=True, n_clusters=3, xt=xt, yt=yt, variances_support=True, ) moe.train() # validation data np.random.seed(1) xe = np.random.sample(self.ne) ye = self.function_test_1d(xe) rms_error = compute_rms_error(moe, xe, ye) self.assert_error(rms_error, 0.0, 3e-1) moe.predict_variances(xe) moe_hard = MOE( smooth_recombination=False, heaviside_optimization=True, n_clusters=3, xt=xt, yt=yt, variances_support=True, ) moe_hard.train() moe_hard.predict_variances(xe) if TestMOE.plot: import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D y = moe.predict_values(xe) plt.figure(1) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel(r"$y$ actual") plt.ylabel(r"$y$ prediction") plt.figure(2) xv = np.linspace(0, 1, 100) yv = self.function_test_1d(xv) y = moe.predict_values(xv) y_std = np.sqrt(moe.predict_variances(xv)) plt.plot(xv, yv, "--k", linewidth=1) plt.plot(xv, y, "-b", linewidth=1) plt.plot(xv, y + y_std, "--b", linewidth=1) plt.plot(xv, y - y_std, "--b", linewidth=1) plt.show() # @unittest.skip('disabled') def test_1d_50_surrogate_model(self): self.ndim = 1 self.nt = 50 self.ne = 50 np.random.seed(0) xt = np.random.sample(self.nt).reshape((-1, 1)) yt = self.function_test_1d(xt) moe = MOESurrogateModel( smooth_recombination=True, heaviside_optimization=True, n_clusters=3, xt=xt, yt=yt, ) self.assertIsInstance(moe.moe, MOE) moe.train() self.assertFalse(moe.supports["variances"]) # validation data np.random.seed(1) xe = np.random.sample(self.ne) ye = self.function_test_1d(xe) rms_error = compute_rms_error(moe, xe, ye) self.assert_error(rms_error, 0.0, 3e-1) self.assertRaises(RuntimeError, lambda: moe.predict_variances(xe)) moe_var = MOESurrogateModel( smooth_recombination=True, heaviside_optimization=True, n_clusters=3, xt=xt, yt=yt, variances_support=True, ) moe_var.train() self.assertTrue(moe_var.supports["variances"]) moe_var.predict_variances(xe) if TestMOE.plot: import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D y = moe.predict_values(xe) plt.figure(1) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel(r"$y$ actual") plt.ylabel(r"$y$ prediction") plt.figure(2) xv = np.linspace(0, 1, 100) yv = self.function_test_1d(xv) plt.plot(xv, yv, "-.") plt.plot(xe, y, "o") plt.show() # @unittest.skip('disabled') def test_norm1_2d_200(self): self.ndim = 2 self.nt = 200 self.ne = 200 prob = LpNorm(ndim=self.ndim) # training data sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) # mixture of experts moe = MOE(smooth_recombination=False, n_clusters=5) moe.set_training_values(xt, yt) moe.train() # validation data np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) rms_error = compute_rms_error(moe, xe, ye) self.assert_error(rms_error, 0.0, 1e-1) if TestMOE.plot: import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D y = moe.predict_values(xe) plt.figure(1) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel(r"$y$ actual") plt.ylabel(r"$y$ prediction") fig = plt.figure(2) ax = fig.add_subplot(111, projection="3d") ax.scatter(xt[:, 0], xt[:, 1], yt) plt.title("L1 Norm") plt.show() # @unittest.skip('disabled for now as it blocks unexpectedly on travis linux') def test_branin_2d_200(self): self.ndim = 2 self.nt = 200 self.ne = 200 prob = Branin(ndim=self.ndim) # training data sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(self.nt) yt = prob(xt) # mixture of experts moe = MOE(n_clusters=5) moe.set_training_values(xt, yt) moe.options["heaviside_optimization"] = True moe.train() # validation data np.random.seed(1) xe = sampling(self.ne) ye = prob(xe) rms_error = compute_rms_error(moe, xe, ye) self.assert_error(rms_error, 0.0, 1e-1) if TestMOE.plot: import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D y = moe.analyse_results(x=xe, operation="predict_values") plt.figure(1) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel(r"$y$ actual") plt.ylabel(r"$y$ prediction") fig = plt.figure(2) ax = fig.add_subplot(111, projection="3d") ax.scatter(xt[:, 0], xt[:, 1], yt) plt.title("Branin function") plt.show() def test_enabled_expert_types(self): moe = MOE(variances_support=True) expected = ["KPLS", "KPLSK", "KRG"] self.assertEqual(expected, sorted(moe.enabled_experts)) moe = MOE(derivatives_support=True) expected = ["IDW", "KPLS", "KPLSK", "KRG", "LS", "QP", "RBF", "RMTB", "RMTC"] self.assertEqual(expected, sorted(moe.enabled_experts)) moe = MOE(deny=["KRG", "RMTB"]) expected = ["IDW", "KPLS", "KPLSK", "LS", "QP", "RBF", "RMTC"] self.assertEqual(expected, sorted(moe.enabled_experts)) moe = MOE(allow=["KRG", "RMTB"]) expected = ["KRG", "RMTB"] self.assertEqual(expected, sorted(moe.enabled_experts)) moe = MOE(variances_support=True, allow=["KRG", "RMTB"]) expected = ["KRG"] self.assertEqual(expected, sorted(moe.enabled_experts)) moe = MOE(derivatives_support=True, deny=["RBF", "IDW", "KPLSK"]) expected = ["KPLS", "KRG", "LS", "QP", "RMTB", "RMTC"] self.assertEqual(expected, sorted(moe.enabled_experts)) def test_fix_moe_rmts_bug(self): def myfunc(x): return -0.5 * ( np.sin(40 * (x - 0.85) ** 4) * np.cos(2.5 * (x - 0.95)) + 0.5 * (x - 0.9) + 1 ) nt1 = 11 nt2 = 15 ne = 101 # Training data X1 = np.linspace(0.001, 0.3, nt1).reshape(nt1, 1) X1 = np.concatenate((X1, np.array([[0.35]])), axis=0) X2 = np.linspace(0.4, 1.0, nt2).reshape(nt2, 1) xt = np.concatenate((X1, X2), axis=0) yt = myfunc(xt) moe = MOE(smooth_recombination=True, n_clusters=2, heaviside_optimization=True) moe._surrogate_type = {"RMTB": RMTB, "RMTC": RMTC} moe.set_training_values(xt, yt) moe.train() def test_fix_test_data_bug(self): # MOE does not interpolate the first training point def myfunc(x): return ((x * 6 - 2) ** 2) * np.sin((x * 6 - 2) * 2) # limits of teh design space xlimits = np.array([[0.0, 1.0]]) # LHS DOE with ndoe points ndoe = 6 # Construction of the DOE sampling = LHS(xlimits=xlimits, criterion="m", random_state=0) x1D = sampling(ndoe) x1D = np.sort(x1D, axis=0) # Compute the output y1D = myfunc(x1D) # test data num = 50 xv1D = sampling(num) xv1D = np.sort(xv1D, axis=0) yv1D = myfunc(xv1D) moe1D = MOE(n_clusters=1, xtest=xv1D, ytest=yv1D, allow=["KRG"]) moe1D.set_training_values(x1D, y1D) moe1D.train() # Check that moe1D is interpolating all training values ypred = moe1D.predict_values(x1D) self.assertTrue(np.allclose(y1D, ypred)) def test_bad_allow_value(self): nt = 35 sampling = FullFactorial(xlimits=np.array([[0, 1]]), clip=True) np.random.seed(0) xt = sampling(nt) yt = self.function_test_1d(xt) moe = MOE(n_clusters=1, allow=["TOTO"]) moe.set_training_values(xt, yt) with self.assertRaises(ValueError) as context: moe.train() self.assertEqual( "List of experts is empty: check support, allow and deny options wrt " "possible experts: ['KRG', 'KPLS', 'KPLSK', 'LS', 'QP', 'RBF', 'IDW', 'RMTB', 'RMTC']", str(context.exception), ) @staticmethod def run_moe_example_1d(): import numpy as np from smt.applications import MOE from smt.sampling_methods import FullFactorial import matplotlib.pyplot as plt nt = 35 def function_test_1d(x): import numpy as np # Note: only required by SMT doc testing toolchain x = np.reshape(x, (-1,)) y = np.zeros(x.shape) y[x < 0.4] = x[x < 0.4] ** 2 y[(x >= 0.4) & (x < 0.8)] = 3 * x[(x >= 0.4) & (x < 0.8)] + 1 y[x >= 0.8] = np.sin(10 * x[x >= 0.8]) return y.reshape((-1, 1)) x = np.linspace(0, 1, 100) ytrue = function_test_1d(x) # Training data sampling = FullFactorial(xlimits=np.array([[0, 1]]), clip=True) np.random.seed(0) xt = sampling(nt) yt = function_test_1d(xt) # Mixture of experts print("MOE Experts: ", MOE.AVAILABLE_EXPERTS) # MOE1: Find the best surrogate model on the whole domain moe1 = MOE(n_clusters=1) print("MOE1 enabled experts: ", moe1.enabled_experts) moe1.set_training_values(xt, yt) moe1.train() y_moe1 = moe1.predict_values(x) # MOE2: Set nb of cluster with just KRG, LS and IDW surrogate models moe2 = MOE(smooth_recombination=False, n_clusters=3, allow=["KRG", "LS", "IDW"]) print("MOE2 enabled experts: ", moe2.enabled_experts) moe2.set_training_values(xt, yt) moe2.train() y_moe2 = moe2.predict_values(x) fig, axs = plt.subplots(1) axs.plot(x, ytrue, ".", color="black") axs.plot(x, y_moe1) axs.plot(x, y_moe2) axs.set_xlabel("x") axs.set_ylabel("y") axs.legend(["Training data", "MOE 1 Prediction", "MOE 2 Prediction"]) plt.show() @staticmethod def run_moe_example_2d(): import numpy as np from smt.applications import MOE from smt.problems import LpNorm from smt.sampling_methods import FullFactorial import sklearn import matplotlib.pyplot as plt from matplotlib import colors from mpl_toolkits.mplot3d import Axes3D ndim = 2 nt = 200 ne = 200 # Problem: L1 norm (dimension 2) prob = LpNorm(ndim=ndim) # Training data sampling = FullFactorial(xlimits=prob.xlimits, clip=True) np.random.seed(0) xt = sampling(nt) yt = prob(xt) # Mixture of experts print("MOE Experts: ", MOE.AVAILABLE_EXPERTS) moe = MOE(smooth_recombination=True, n_clusters=5, deny=["RMTB", "KPLSK"]) print("Enabled Experts: ", moe.enabled_experts) moe.set_training_values(xt, yt) moe.train() # Validation data np.random.seed(1) xe = sampling(ne) ye = prob(xe) # Prediction y = moe.predict_values(xe) fig = plt.figure(1) fig.set_size_inches(12, 11) # Cluster display colors_ = list(colors.cnames.items()) GMM = moe.cluster weight = GMM.weights_ mean = GMM.means_ if sklearn.__version__ < "0.20.0": cov = GMM.covars_ else: cov = GMM.covariances_ prob_ = moe._proba_cluster(xt) sort = np.apply_along_axis(np.argmax, 1, prob_) xlim = prob.xlimits x0 = np.linspace(xlim[0, 0], xlim[0, 1], 20) x1 = np.linspace(xlim[1, 0], xlim[1, 1], 20) xv, yv = np.meshgrid(x0, x1) x = np.array(list(zip(xv.reshape((-1,)), yv.reshape((-1,))))) prob = moe._proba_cluster(x) plt.subplot(221, projection="3d") ax = plt.gca() for i in range(len(sort)): color = colors_[int(((len(colors_) - 1) / sort.max()) * sort[i])][0] ax.scatter(xt[i][0], xt[i][1], yt[i], c=color) plt.title("Clustered Samples") plt.subplot(222, projection="3d") ax = plt.gca() for i in range(len(weight)): color = colors_[int(((len(colors_) - 1) / len(weight)) * i)][0] ax.plot_trisurf( x[:, 0], x[:, 1], prob[:, i], alpha=0.4, linewidth=0, color=color ) plt.title("Membership Probabilities") plt.subplot(223) for i in range(len(weight)): color = colors_[int(((len(colors_) - 1) / len(weight)) * i)][0] plt.tricontour(x[:, 0], x[:, 1], prob[:, i], 1, colors=color, linewidths=3) plt.title("Cluster Map") plt.subplot(224) plt.plot(ye, ye, "-.") plt.plot(ye, y, ".") plt.xlabel("actual") plt.ylabel("prediction") plt.title("Predicted vs Actual") plt.show() if __name__ == "__main__": if "--plot" in argv: TestMOE.plot = True argv.remove("--plot") if "--example" in argv: TestMOE.run_moe_example_1d() exit() unittest.main()
16,658
29.344262
99
py
smt
smt-master/doc/preprocess_test.py
""" Author: Dr. John T. Hwang <hwangjt@umich.edu> This package is distributed under New BSD license. """ import os, sys import inspect import importlib import contextlib try: from StringIO import StringIO except: from io import StringIO import matplotlib matplotlib.use("Agg") import matplotlib.pyplot as plt @contextlib.contextmanager def stdoutIO(stdout=None): old = sys.stdout if stdout is None: stdout = StringIO() sys.stdout = stdout yield stdout sys.stdout = old def process_test(root, file_name, iline, line): file_path = root + "/" + file_name embed_num_indent = line.find(".. embed-test") if line[:embed_num_indent] != " " * embed_num_indent: return line include_print_output = ( "embed-test-print" in line or "embed-test-print-plot" in line or "embed-test-print-plot" in line ) include_plot_output = ( "embed-test-plot" in line or "embed-test-print-plot" in line or "embed-test-print-plot" in line ) split_line = line.replace(" ", "").split(",") if len(split_line) != 3 or len(split_line[0].split("::")) != 2: raise Exception( "Invalid format for embed-test in file {} line {}".format( file_path, iline + 1 ) ) py_file_path = split_line[0].split("::")[1] class_name = split_line[1] method_name = split_line[2][:-1] index = len(py_file_path.split("/")[-1]) py_root = py_file_path[:-index] py_file_name = py_file_path[-index:] sys.path.append(os.path.dirname(os.path.realpath(__file__)) + "/" + py_root) py_module = importlib.import_module(py_file_name[:-3]) obj = getattr(py_module, class_name) method = getattr(obj, method_name) method_lines = inspect.getsource(method).split("\n") for imethod_line, method_line in enumerate(method_lines): if "def" in method_line and method_name in method_line: imethod_line += 1 break method_lines = method_lines[imethod_line:] first_line = method_lines[0] py_num_indent = first_line.find(first_line.strip()) for imethod_line, method_line in enumerate(method_lines): method_lines[imethod_line] = method_line[py_num_indent:] replacement_lines = [] replacement_lines.append(" " * embed_num_indent + ".. code-block:: python\n") replacement_lines.append("\n") replacement_lines.extend( [ " " * embed_num_indent + " " * 2 + method_line + "\n" for method_line in method_lines ] ) if include_print_output: joined_method_lines = "\n".join(method_lines) with stdoutIO() as s: exec(joined_method_lines) output_lines = s.getvalue().split("\n") if len(output_lines) > 1: replacement_lines.append(" " * embed_num_indent + "::\n") replacement_lines.append("\n") replacement_lines.extend( [ " " * embed_num_indent + " " * 2 + output_line + "\n" for output_line in output_lines ] ) if include_plot_output: joined_method_lines = "\n".join(method_lines) plt.clf() with stdoutIO() as s: exec(joined_method_lines) abs_plot_name = file_path[:-5] + ".png" plt.savefig(abs_plot_name) rel_plot_name = file_name[:-5] + ".png" replacement_lines.append( " " * embed_num_indent + ".. figure:: {}\n".format(rel_plot_name) ) replacement_lines.append(" " * embed_num_indent + " :scale: 80 %\n") replacement_lines.append(" " * embed_num_indent + " :align: center\n") return replacement_lines
3,771
27.793893
81
py
smt
smt-master/doc/preprocess_options.py
""" Author: Dr. John T. Hwang <hwangjt@umich.edu> This package is distributed under New BSD license. """ def process_options(root, file_name, iline, line): file_path = root + "/" + file_name embed_num_indent = line.find(".. embed-options-table") if "embed-options-table-surrogate_models" in line: type_ = "surrogate_models" elif "embed-options-table-problems" in line: type_ = "problems" elif "embed-options-table-sampling_methods" in line: type_ = "sampling_methods" else: raise Exception("embed-options-table is an invalid name") if line[:embed_num_indent] != " " * embed_num_indent: return line split_line = line.replace(" ", "").split("::") if len(split_line) != 2: raise Exception( "Invalid format for embed-options-table in file {} line {}".format( file_path, iline + 1 ) ) class_name = split_line[1] exec("from smt.{} import {}".format(type_, class_name), globals()) exec("sm_class = {}".format(class_name), globals()) sm = sm_class() options = sm.options outputs = [] for option_name, option_data in options._declared_entries.items(): name = option_name default = option_data["default"] values = option_data["values"] types = option_data["types"] desc = option_data["desc"] if types is not None: if not isinstance(types, (tuple, list)): types = (types,) types = [type_.__name__ for type_ in types] if values is not None: if not isinstance(values, (tuple, list)): values = (values,) values = [value for value in values] outputs.append([name, default, values, types, desc]) replacement_lines = [] replacement_lines.append( " " * embed_num_indent + ".. list-table:: List of options\n" ) replacement_lines.append(" " * embed_num_indent + " " * 2 + ":header-rows: 1\n") replacement_lines.append( " " * embed_num_indent + " " * 2 + ":widths: 15, 10, 20, 20, 30\n" ) replacement_lines.append(" " * embed_num_indent + " " * 2 + ":stub-columns: 0\n") replacement_lines.append("\n") replacement_lines.append(" " * embed_num_indent + " " * 2 + "* - Option\n") replacement_lines.append(" " * embed_num_indent + " " * 2 + " - Default\n") replacement_lines.append( " " * embed_num_indent + " " * 2 + " - Acceptable values\n" ) replacement_lines.append( " " * embed_num_indent + " " * 2 + " - Acceptable types\n" ) replacement_lines.append(" " * embed_num_indent + " " * 2 + " - Description\n") for output in outputs: for entry in [output[0]]: replacement_lines.append( " " * embed_num_indent + " " * 2 + "* - %s\n" % entry ) for entry in output[1:]: replacement_lines.append( " " * embed_num_indent + " " * 2 + " - %s\n" % entry ) return replacement_lines
3,088
31.861702
86
py
smt
smt-master/doc/conf.py
#!/usr/bin/env python3 # -*- coding: utf-8 -*- # # SMT documentation build configuration file, created by # sphinx-quickstart on Sun Aug 6 19:36:14 2017. # # This file is execfile()d with the current directory set to its # containing dir. # # Note that not all possible configuration values are present in this # autogenerated file. # # All configuration values have a default; values that are commented out # serve to show the default. # If extensions (or modules to document with autodoc) are in another directory, # add these directories to sys.path here. If the directory is relative to the # documentation root, use os.path.abspath to make it absolute, like shown here. # # import os # import sys # sys.path.insert(0, os.path.abspath('.')) from smt import __version__ # -- General configuration ------------------------------------------------ # If your documentation needs a minimal Sphinx version, state it here. # # needs_sphinx = '1.0' # Add any Sphinx extension module names here, as strings. They can be # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom # ones. extensions = [ "sphinx.ext.autodoc", "sphinx.ext.mathjax", "sphinx.ext.viewcode", "matplotlib.sphinxext.plot_directive", "numpydoc", ] # Add any paths that contain templates here, relative to this directory. templates_path = ["_templates"] # The suffix(es) of source filenames. # You can specify multiple suffix as a list of string: # # source_suffix = ['.rst', '.md'] source_suffix = ".rst" # The master toctree document. master_doc = "index" # General information about the project. project = "SMT" copyright = "2017, John Hwang" author = "John Hwang, Mohamed Amine Bouhlel, Remi Lafage" # The version info for the project you're documenting, acts as replacement for # |version| and |release|, also used in various other places throughout the # built documents. # # The short X.Y version. # Take the full version when no need to distinguish version and release version = __version__ # The full version, including alpha/beta/rc tags. release = version # The language for content autogenerated by Sphinx. Refer to documentation # for a list of supported languages. # # This is also used if you do content translation via gettext catalogs. # Usually you set "language" from the command line for these cases. language = None # List of patterns, relative to source directory, that match files and # directories to ignore when looking for source files. # This patterns also effect to html_static_path and html_extra_path exclude_patterns = ["_build", "Thumbs.db", ".DS_Store"] # The name of the Pygments (syntax highlighting) style to use. pygments_style = "sphinx" # If true, `todo` and `todoList` produce output, else they produce nothing. todo_include_todos = False # -- Options for HTML output ---------------------------------------------- # The theme to use for HTML and HTML Help pages. See the documentation for # a list of builtin themes. html_theme = "bizstyle" # Theme options are theme-specific and customize the look and feel of a theme # further. For a list of options available for each theme, see the # documentation. # html_theme_options = { "rightsidebar": False, "sidebarwidth": 250, "body_min_width": 1100, "body_max_width": 1100, } html_logo = "smt_logo.png" # Add any paths that contain custom static files (such as style sheets) here, # relative to this directory. They are copied after the builtin static files, # so a file named "default.css" will overwrite the builtin "default.css". html_static_path = ["_static"] # -- Options for HTMLHelp output ------------------------------------------ # Output file base name for HTML help builder. htmlhelp_basename = "SMTdoc" # -- Options for LaTeX output --------------------------------------------- latex_elements = { # The paper size ('letterpaper' or 'a4paper'). # # 'papersize': 'letterpaper', # The font size ('10pt', '11pt' or '12pt'). # # 'pointsize': '10pt', # Additional stuff for the LaTeX preamble. # # 'preamble': '', # Latex figure (float) alignment # # 'figure_align': 'htbp', } # Grouping the document tree into LaTeX files. List of tuples # (source start file, target name, title, # author, documentclass [howto, manual, or own class]). latex_documents = [ ( master_doc, "SMT.tex", "SMT Documentation", "John Hwang, Mohamed Amine Bouhlel, Remi Lafage", "manual", ) ] # -- Options for manual page output --------------------------------------- # One entry per manual page. List of tuples # (source start file, name, description, authors, manual section). man_pages = [(master_doc, "smt", "SMT Documentation", [author], 1)] # -- Options for Texinfo output ------------------------------------------- # Grouping the document tree into Texinfo files. List of tuples # (source start file, target name, title, author, # dir menu entry, description, category) texinfo_documents = [ ( master_doc, "SMT", "SMT Documentation", author, "SMT", "One line description of project.", "Miscellaneous", ) ]
5,197
28.873563
79
py
smt
smt-master/doc/preprocess.py
""" Author: Dr. John T. Hwang <hwangjt@umich.edu> This package is distributed under New BSD license. """ import os from preprocess_test import process_test from preprocess_options import process_options def get_file_paths(): file_paths_list = [] file_paths_list_rstx = [] file_paths_list_py = [] for root, dirs, files in os.walk("."): for file_name in files: file_paths_list.append((root, file_name)) if file_name[-5:] == ".rstx": file_paths_list_rstx.append((root, file_name)) if file_name[-3:] == ".py": file_paths_list_py.append((root, file_name)) return file_paths_list, file_paths_list_rstx, file_paths_list_py file_paths_list, file_paths_list_rstx, file_paths_list_py = get_file_paths() for root, file_name in file_paths_list_rstx: file_path = root + "/" + file_name with open(file_path, "r") as f: lines = f.readlines() for iline, line in enumerate(lines): if ".. embed-test" in line: lines[iline] = process_test(root, file_name, iline, line) elif ".. embed-options-table" in line: lines[iline] = process_options(root, file_name, iline, line) new_lines = [] for line in lines: if isinstance(line, list): new_lines.extend(line) else: new_lines.append(line) new_file_path = file_path[:-5] + ".rst" with open(new_file_path, "w") as f: f.writelines(new_lines)
1,505
27.415094
76
py
smt
smt-master/doc/embed_directives/directive_embed_options.py
from sphinx_auto_embed.directive import Directive class DirectiveEmbedOptions(Directive): """ Directive for embedding a table from an OptionsDictionary instance. The 3 arguments are the module name, class name, and attribute name. """ NAME = "embed-options-table" NUM_ARGS = 3 def run(self, file_dir, file_name, embed_num_indent, args): module_path, class_name, attribute_name = args exec("from {} import {}".format(module_path, class_name), globals()) exec("obj = {}()".format(class_name), globals()) options = getattr(obj, attribute_name) outputs = [] for option_name, option_data in options._declared_entries.items(): name = option_name default = option_data["default"] values = option_data["values"] types = option_data["types"] desc = option_data["desc"] if types is not None: if not isinstance(types, (tuple, list)): types = (types,) types = [type_.__name__ for type_ in types] if values is not None: if not isinstance(values, (tuple, list)): values = (values,) values = [value for value in values] outputs.append([name, default, values, types, desc]) lines = [] lines.append(" " * embed_num_indent + ".. list-table:: List of options\n") lines.append(" " * embed_num_indent + " " * 2 + ":header-rows: 1\n") lines.append(" " * embed_num_indent + " " * 2 + ":widths: 15, 10, 20, 20, 30\n") lines.append(" " * embed_num_indent + " " * 2 + ":stub-columns: 0\n") lines.append("\n") lines.append(" " * embed_num_indent + " " * 2 + "* - Option\n") lines.append(" " * embed_num_indent + " " * 2 + " - Default\n") lines.append(" " * embed_num_indent + " " * 2 + " - Acceptable values\n") lines.append(" " * embed_num_indent + " " * 2 + " - Acceptable types\n") lines.append(" " * embed_num_indent + " " * 2 + " - Description\n") for output in outputs: for entry in [output[0]]: lines.append(" " * embed_num_indent + " " * 2 + "* - %s\n" % entry) for entry in output[1:]: lines.append(" " * embed_num_indent + " " * 2 + " - %s\n" % entry) return lines
2,419
37.412698
88
py
pyzor
pyzor-master/setup.py
import sys import setuptools import distutils.core import pyzor long_description = """ Pyzor is spam-blocking networked system that uses spam signatures to identify them. """ classifiers = ["Operating System :: POSIX", "Operating System :: Microsoft :: Windows", "Environment :: Console", "Environment :: No Input/Output (Daemon)", "Programming Language :: Python", "Programming Language :: Python :: 2.6", "Programming Language :: Python :: 3", "Intended Audience :: System Administrators", "Topic :: Communications :: Email", "Topic :: Communications :: Email :: Filters", "Development Status :: 5 - Production/Stable", "License :: OSI Approved :: GNU General Public License v2 (" "GPLv2)", ] distutils.core.setup( name='pyzor', version=pyzor.__version__, description='networked spam-signature detection', long_description=long_description, author='Frank J. Tobin', author_email='ftobin@neverending.org', license='GPL', platforms='POSIX', keywords='spam', url='http://www.pyzor.org/', scripts=['scripts/pyzor', 'scripts/pyzord', 'scripts/pyzor-migrate'], packages=['pyzor', 'pyzor.engines', 'pyzor.hacks'], classifiers=classifiers, test_suite="tests.suite", )
1,532
28.480769
75
py
pyzor
pyzor-master/web/application.py
#! /usr/bin/env python import os import email import socket import logging import smtplib import datetime import email.utils import email.mime.base import email.mime.text import email.mime.multipart try: import configparser as ConfigParser except ImportError: import ConfigParser import flask from flask_wtf.form import Form from flask.views import MethodView from wtforms.fields.simple import TextField, SubmitField, TextAreaField from flask_wtf.html5 import EmailField from flask_wtf.file import FileField from flask_wtf.recaptcha.fields import RecaptchaField from wtforms.validators import required, length try: import sentry_sdk from sentry_sdk.integrations.flask import FlaskIntegration except ImportError: pass import pyzor import pyzor.digest import pyzor.client MSG_TEMPLATE_TXT = """ Whitelist request: - Date: %s - Name: %%(name)s - Email: %%(email)s - Digest: %%(digest)s - Request IP: %%(ip)s =============== %%(comment)s =============== Pyzor Version: %s """ % (datetime.datetime.utcnow(), pyzor.__version__) def load_configuration(): """Load server-specific configuration settings.""" conf = ConfigParser.ConfigParser() defaults = { "captcha": { "ssl": "False", "public_key": "", "private_key": "", }, "email": { "host": "localhost", "port": "25", "username": "", "password": "", "recipients": "", "sender": "no-reply@%s" % socket.gethostname(), }, "logging": { "file": "/var/log/pyzor/web.log", "level": "INFO", "sentry": "", "sentry_level": "WARNING", } } # Load in default values. for section, values in defaults.items(): conf.add_section(section) for option, value in values.items(): conf.set(section, option, value) if os.path.exists("/etc/pyzor/web.conf"): # Overwrite with local values. conf.read("/etc/pyzor/web.conf") return conf def setup_logging(): logger = app.logger file_handler = logging.FileHandler(CONF.get("logging", "file")) file_handler.setFormatter(logging.Formatter( '%(asctime)s %(levelname)s %(message)s')) log_level = getattr(logging, CONF.get("logging", "level")) logger.setLevel(log_level) logger.addHandler(file_handler) CONF = load_configuration() SENTRY_DSN = CONF.get("logging", "sentry") if SENTRY_DSN: sentry_sdk.init( dsn="https://b23e6ec6b4e14bdfb44dcfa5e9a8ba16@sentry.io/1207622", integrations=[FlaskIntegration()] ) setup_logging() app = flask.Flask(__name__) app.config.update({ "RECAPTCHA_USE_SSL": CONF.get("captcha", "ssl").lower() == "true", "RECAPTCHA_PUBLIC_KEY": CONF.get("captcha", "public_key"), "RECAPTCHA_PRIVATE_KEY": CONF.get("captcha", "private_key"), }) class MessageForm(Form): digest = TextField("Pyzor digest*", validators=[length(40, 40, "Invalid Digest"), required()]) message = FileField('Raw message*') name = TextField('Name') email = EmailField('Email') comment = TextAreaField('Other details') recaptcha = RecaptchaField() submit = SubmitField() def __init___(self, *args, **kwargs): super(MessageForm, self).__init__(*args, **kwargs) self.msg = None self.raw_message = None self.logger = app.logger def validate(self): if not Form.validate(self): return False self.raw_message = flask.request.files["message"].stream.read() try: digest = pyzor.digest.DataDigester( email.message_from_string(self.raw_message)).value if digest != self.digest.data: self.add_error("digest", "Digest does not match message.") return False client = pyzor.client.Client(timeout=20) try: response = client.check(digest) except pyzor.TimeoutError as e: self.add_error("message", "Temporary error please try again.") self.logger.warn("Timeout: %s", e) return False except pyzor.CommError as e: self.add_error("message", "Temporary error please try again.") self.logger.warn("Error: %s", e) return False if not response.is_ok(): self.add_error("message", "Temporary error please try again.") self.logger.warn("Invalid response from server: %s", response) return False if int(response["Count"]) == 0: self.add_error("message", "Message not reported as spam.") return False if int(response["WL-Count"]) != 0: self.add_error("message", "Message is already whitelisted.") return False except AssertionError: self.add_error("message", "Invalid message.") return False return True def add_error(self, field, message): try: self.errors[field].append(message) except (KeyError, TypeError): self.errors[field] = [message] class WhitelistMessage(MethodView): def __init__(self): self.form = MessageForm(flask.request.form, csrf_enabled=False) self.logger = app.logger def get(self): return flask.render_template('whitelist.html', form=self.form, error=None) def post(self): success = False if self.form.validate(): msg = self.build_notification() self.send_email(msg) success = True return flask.render_template('whitelist.html', form=self.form, success=success) def build_notification(self): data = {"name": self.form.name.data, "email": self.form.email.data, "digest": self.form.digest.data, "comment": self.form.comment.data, "ip": flask.request.remote_addr} msg = email.mime.multipart.MIMEMultipart() msg["Date"] = email.utils.formatdate(localtime=True) msg["Subject"] = "[Pyzor] Whitelist request" msg["From"] = CONF.get("email", "sender") msg["To"] = CONF.get("email", "recipients") msg.preamble = "This is a multi-part message in MIME format." msg.epilogue = "" msg.attach(email.mime.text.MIMEText(MSG_TEMPLATE_TXT % data)) original_attachment = email.mime.base.MIMEBase("message", "rfc822") original_attachment.add_header("Content-Disposition", "attachment") original_attachment.set_payload(self.form.raw_message) msg.attach(original_attachment) return msg def send_email(self, msg): smtp = smtplib.SMTP(host=CONF.get("email", "host"), port=CONF.get("email", "port")) smtp.ehlo() try: code, err = smtp.mail(CONF.get("email", "sender")) if code != 250: raise smtplib.SMTPSenderRefused(code, err, CONF.get("email", "sender")) rcpterrs = {} for rcpt in CONF.get("email", "recipients").split(","): code, err = smtp.rcpt(rcpt) if code not in (250, 251): rcpterrs[rcpt] = (code, err) if rcpterrs: raise smtplib.SMTPRecipientsRefused(rcpterrs) code, err = smtp.data(msg.as_string()) if code != 250: raise smtplib.SMTPDataError(code, err) finally: try: smtp.quit() except smtplib.SMTPServerDisconnected: pass app.add_url_rule("/whitelist/", view_func=WhitelistMessage.as_view("whitelist")) @app.errorhandler(500) def unhandled_exception(error): """Generic error message.""" setup_logging() app.logger.error("Unhandled Exception: %s", error, exc_info=True) return flask.render_template('error.html', error=error) if __name__ == '__main__': app.debug = True app.run()
8,333
31.940711
80
py
pyzor
pyzor-master/pyzor/message.py
"""This modules contains the various messages used in the pyzor client server communication. """ import random import email.message import pyzor class Message(email.message.Message): def __init__(self): email.message.Message.__init__(self) self.setup() def setup(self): pass def init_for_sending(self): self.ensure_complete() def __str__(self): # The parent class adds the unix From header. return self.as_string() def ensure_complete(self): pass class ThreadedMessage(Message): def init_for_sending(self): if 'Thread' not in self: self.set_thread(ThreadId.generate()) assert 'Thread' in self self["PV"] = str(pyzor.proto_version) Message.init_for_sending(self) def ensure_complete(self): if 'PV' not in self or 'Thread' not in self: raise pyzor.IncompleteMessageError("Doesn't have fields for a " "ThreadedMessage.") Message.ensure_complete(self) def get_protocol_version(self): return float(self['PV']) def get_thread(self): return ThreadId(self['Thread']) def set_thread(self, i): self['Thread'] = str(i) class Response(ThreadedMessage): ok_code = 200 def ensure_complete(self): if 'Code' not in self or 'Diag' not in self: raise pyzor.IncompleteMessageError("doesn't have fields for a " "Response") ThreadedMessage.ensure_complete(self) def is_ok(self): return self.get_code() == self.ok_code def get_code(self): return int(self['Code']) def get_diag(self): return self['Diag'] def head_tuple(self): return self.get_code(), self.get_diag() class Request(ThreadedMessage): """This is the class that should be used to read in Requests of any type. Subclasses are responsible for setting 'Op' if they are generating a message,""" def get_op(self): return self['Op'] def ensure_complete(self): if 'Op' not in self: raise pyzor.IncompleteMessageError("doesn't have fields for a " "Request") ThreadedMessage.ensure_complete(self) class ClientSideRequest(Request): op = None def setup(self): Request.setup(self) self["Op"] = self.op class SimpleDigestBasedRequest(ClientSideRequest): def __init__(self, digest=None): ClientSideRequest.__init__(self) self.digest_count = 0 if digest: self.add_digest(digest) def add_digest(self, digest): self.add_header("Op-Digest", digest) self.digest_count += 1 class SimpleDigestSpecBasedRequest(SimpleDigestBasedRequest): def __init__(self, digest=None, spec=None): SimpleDigestBasedRequest.__init__(self, digest) if spec: flat_spec = [item for sublist in spec for item in sublist] self["Op-Spec"] = ",".join(str(part) for part in flat_spec) class PingRequest(ClientSideRequest): op = "ping" class PongRequest(SimpleDigestBasedRequest): op = "pong" class CheckRequest(SimpleDigestBasedRequest): op = "check" class InfoRequest(SimpleDigestBasedRequest): op = "info" class ReportRequest(SimpleDigestSpecBasedRequest): op = "report" class WhitelistRequest(SimpleDigestSpecBasedRequest): op = "whitelist" class ThreadId(int): # (0, 1024) is reserved full_range = (0, 2 ** 16) ok_range = (1024, full_range[1]) error_value = 0 def __new__(cls, i): self = int.__new__(cls, i) if not (cls.full_range[0] <= self < cls.full_range[1]): raise ValueError("value outside of range") return self @classmethod def generate(cls): return cls(random.randrange(*cls.ok_range)) def in_ok_range(self): return self.ok_range[0] <= self < self.ok_range[1]
4,023
23.993789
77
py
pyzor
pyzor-master/pyzor/digest.py
"""Handle digesting the messages.""" from __future__ import print_function import re import sys import codecs import hashlib try: import HTMLParser except ImportError: import html.parser as HTMLParser # Hard-coded for the moment. digest_spec = ([(20, 3), (60, 3)]) HASH = hashlib.sha1 HASH_SIZE = len(HASH(b"").hexdigest()) if sys.version_info[0] == 2: sys.stdout = codecs.getwriter('utf8')(sys.stdout) class HTMLStripper(HTMLParser.HTMLParser): """Strip all tags from the HTML.""" def __init__(self, collector): HTMLParser.HTMLParser.__init__(self) self.reset() self.collector = collector self.collect = True def handle_data(self, data): """Keep track of the data.""" data = data.strip() if data and self.collect: self.collector.append(data) def handle_starttag(self, tag, attrs): HTMLParser.HTMLParser.handle_starttag(self, tag, attrs) if tag.lower() in ("script", "style"): self.collect = False def handle_endtag(self, tag): HTMLParser.HTMLParser.handle_endtag(self, tag) if tag.lower() in ("script", "style"): self.collect = True class DataDigester(object): """The major workhouse class.""" __slots__ = ['value', 'digest'] # Minimum line length for it to be included as part of the digest. min_line_length = 8 # If a message is this many lines or less, then we digest the whole # message. atomic_num_lines = 4 # We're not going to try to match email addresses as per the spec # because it's too difficult. Plus, regular expressions don't work well # for them. (BNF is better at balanced parens and such). email_ptrn = re.compile(r'\S+@\S+') # Same goes for URLs. url_ptrn = re.compile(r'[a-z]+:\S+', re.IGNORECASE) # We also want to remove anything that is so long it looks like possibly # a unique identifier. longstr_ptrn = re.compile(r'\S{10,}') ws_ptrn = re.compile(r'\s') # String that the above patterns will be replaced with. # Note that an empty string will always be used to remove whitespace. unwanted_txt_repl = '' def __init__(self, msg, spec=None): if spec is None: spec = digest_spec self.value = None self.digest = HASH() # Need to know the total number of lines in the content. lines = [] for payload in self.digest_payloads(msg): for line in payload.splitlines(): norm = self.normalize(line) if self.should_handle_line(norm): try: lines.append(norm.encode("utf8", "ignore")) except UnicodeError: continue if len(lines) <= self.atomic_num_lines: self.handle_atomic(lines) else: self.handle_pieced(lines, spec) self.value = self.digest.hexdigest() assert len(self.value) == HASH_SIZE def handle_atomic(self, lines): """We digest everything.""" for line in lines: self.handle_line(line) def handle_pieced(self, lines, spec): """Digest stuff according to the spec.""" for offset, length in spec: for i in range(length): try: line = lines[int(offset * len(lines) // 100) + i] except IndexError: pass else: self.handle_line(line) def handle_line(self, line): self.digest.update(line.rstrip()) @classmethod def normalize(cls, s): s = s.replace("\x00", "") repl = cls.unwanted_txt_repl s = cls.longstr_ptrn.sub(repl, s) s = cls.email_ptrn.sub(repl, s) s = cls.url_ptrn.sub(repl, s) # Make sure we do the whitespace last because some of the previous # patterns rely on whitespace. return cls.ws_ptrn.sub('', s).strip() @staticmethod def normalize_html_part(s): data = [] stripper = HTMLStripper(data) try: stripper.feed(s) except (UnicodeDecodeError, HTMLParser.HTMLParseError): # We can't parse the HTML, so just strip it. This is still # better than including generic HTML/CSS text. pass return " ".join(data) @classmethod def should_handle_line(cls, s): return len(s) and cls.min_line_length <= len(s) @classmethod def digest_payloads(cls, msg): for part in msg.walk(): if part.get_content_maintype() == "text": payload = part.get_payload(decode=True) charset = part.get_content_charset() if charset: charset = charset.replace("\x00", "") errors = "ignore" if not charset: charset = "ascii" elif (charset.lower().replace("_", "-") in ("quopri-codec", "quopri", "quoted-printable", "quotedprintable")): errors = "strict" try: payload = payload.decode(charset, errors) except (LookupError, UnicodeError, AssertionError): try: payload = payload.decode("ascii", "ignore") except UnicodeError: continue if part.get_content_subtype() == "html": yield cls.normalize_html_part(payload) else: yield payload elif part.is_multipart(): # Skip, because walk() will give us the payload next. pass else: # Non-text parts are passed through as-is. yield part.get_payload() class PrintingDataDigester(DataDigester): """Extends DataDigester: prints out what we're digesting.""" def handle_line(self, line): print(line.decode("utf8")) super(PrintingDataDigester, self).handle_line(line)
6,107
30.979058
76
py
pyzor
pyzor-master/pyzor/server.py
"""Networked spam-signature detection server. The server receives the request in the form of a RFC5321 message, and responds with another RFC5321 message. Neither of these messages has a body - all of the data is encapsulated in the headers. The response headers will always include a "Code" header, which is a HTTP-style response code, and a "Diag" header, which is a human-readable message explaining the response code (typically this will be "OK"). Both the request and response headers always include a "PV" header, which indicates the protocol version that is being used (in a major.minor format). Both the requestion and response headers also always include a "Thread", which uniquely identifies the request (this is a requirement of using UDP). Responses to requests may arrive in any order, but the "Thread" header of a response will always match the "Thread" header of the appropriate request. Authenticated requests must also have "User", "Time" (timestamp), and "Sig" (signature) headers. """ import os import sys import time import errno import socket import signal import logging import threading import traceback import email.message try: import SocketServer except ImportError: import socketserver as SocketServer import pyzor.config import pyzor.account import pyzor.engines.common import pyzor.hacks.py26 pyzor.hacks.py26.hack_all() def _eintr_retry(func, *args): """restart a system call interrupted by EINTR""" while True: try: return func(*args) except OSError as e: if e.args[0] != errno.EINTR: raise class Server(SocketServer.UDPServer): """The pyzord server. Handles incoming UDP connections in a single thread and single process.""" max_packet_size = 8192 time_diff_allowance = 180 def __init__(self, address, database, passwd_fn, access_fn, forwarder=None): if ":" in address[0]: Server.address_family = socket.AF_INET6 else: Server.address_family = socket.AF_INET self.log = logging.getLogger("pyzord") self.usage_log = logging.getLogger("pyzord-usage") self.database = database self.one_step = getattr(self.database, "handles_one_step", False) # Handle configuration files self.passwd_fn = passwd_fn self.access_fn = access_fn self.accounts = {} self.acl = {} self.load_config() self.forwarder = forwarder self.log.debug("Listening on %s", address) SocketServer.UDPServer.__init__(self, address, RequestHandler, bind_and_activate=False) try: self.socket.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_V6ONLY, 0) except (AttributeError, socket.error) as e: self.log.debug("Unable to set IPV6_V6ONLY to false %s", e) self.server_bind() self.server_activate() # Finally, set signals signal.signal(signal.SIGUSR1, self.reload_handler) signal.signal(signal.SIGTERM, self.shutdown_handler) def load_config(self): """Reads the configuration files and loads the accounts and ACLs.""" self.accounts = pyzor.config.load_passwd_file(self.passwd_fn) self.acl = pyzor.config.load_access_file(self.access_fn, self.accounts) def shutdown_handler(self, *args, **kwargs): """Handler for the SIGTERM signal. This should be used to kill the daemon and ensure proper clean-up. """ self.log.info("SIGTERM received. Shutting down.") t = threading.Thread(target=self.shutdown) t.start() def reload_handler(self, *args, **kwargs): """Handler for the SIGUSR1 signal. This should be used to reload the configuration files. """ self.log.info("SIGUSR1 received. Reloading configuration.") t = threading.Thread(target=self.load_config) t.start() def handle_error(self, request, client_address): self.log.error("Error while processing request from: %s", client_address, exc_info=True) class PreForkServer(Server): """The same as Server, but prefork itself when starting the self, by forking a number of child-processes. The parent process will then wait for all his child process to complete. """ def __init__(self, address, database, passwd_fn, access_fn, prefork=4): """The same as Server.__init__ but requires a list of databases instead of a single database connection. """ self.pids = None Server.__init__(self, address, database, passwd_fn, access_fn) self._prefork = prefork def serve_forever(self, poll_interval=0.5): """Fork the current process and wait for all children to finish.""" pids = [] for dummy in range(self._prefork): database = next(self.database) pid = os.fork() if not pid: # Create the database in the child process, to prevent issues self.database = database() Server.serve_forever(self, poll_interval=poll_interval) os._exit(0) else: pids.append(pid) self.pids = pids for pid in self.pids: _eintr_retry(os.waitpid, pid, 0) def shutdown(self): """If this is the parent process send the TERM signal to all children, else call the super method. """ for pid in self.pids or (): os.kill(pid, signal.SIGTERM) if self.pids is None: Server.shutdown(self) def load_config(self): """If this is the parent process send the USR1 signal to all children, else call the super method. """ for pid in self.pids or (): os.kill(pid, signal.SIGUSR1) if self.pids is None: Server.load_config(self) class ThreadingServer(SocketServer.ThreadingMixIn, Server): """A threaded version of the pyzord server. Each connection is served in a new thread. This may not be suitable for all database types.""" pass class BoundedThreadingServer(ThreadingServer): """Same as ThreadingServer but this also accepts a limited number of concurrent threads. """ def __init__(self, address, database, passwd_fn, access_fn, max_threads, forwarding_server=None): ThreadingServer.__init__(self, address, database, passwd_fn, access_fn, forwarder=forwarding_server) self.semaphore = threading.Semaphore(max_threads) def process_request(self, request, client_address): self.semaphore.acquire() ThreadingServer.process_request(self, request, client_address) def process_request_thread(self, request, client_address): ThreadingServer.process_request_thread(self, request, client_address) self.semaphore.release() class ProcessServer(SocketServer.ForkingMixIn, Server): """A multi-processing version of the pyzord server. Each connection is served in a new process. This may not be suitable for all database types. """ def __init__(self, address, database, passwd_fn, access_fn, max_children=40, forwarding_server=None): ProcessServer.max_children = max_children Server.__init__(self, address, database, passwd_fn, access_fn, forwarder=forwarding_server) class RequestHandler(SocketServer.DatagramRequestHandler): """Handle a single pyzord request.""" def __init__(self, *args, **kwargs): self.response = email.message.Message() SocketServer.DatagramRequestHandler.__init__(self, *args, **kwargs) def handle(self): """Handle a pyzord operation, cleanly handling any errors.""" self.response["Code"] = "200" self.response["Diag"] = "OK" self.response["PV"] = "%s" % pyzor.proto_version try: self._really_handle() except NotImplementedError as e: self.handle_error(501, "Not implemented: %s" % e) except pyzor.UnsupportedVersionError as e: self.handle_error(505, "Version Not Supported: %s" % e) except pyzor.ProtocolError as e: self.handle_error(400, "Bad request: %s" % e) except pyzor.SignatureError as e: self.handle_error(401, "Unauthorized: Signature Error: %s" % e) except pyzor.AuthorizationError as e: self.handle_error(403, "Forbidden: %s" % e) except Exception as e: self.handle_error(500, "Internal Server Error: %s" % e) self.server.log.error(traceback.format_exc()) self.server.log.debug("Sending: %r", self.response.as_string()) self.wfile.write(self.response.as_string().encode("utf8")) def _really_handle(self): """handle() without the exception handling.""" self.server.log.debug("Received: %r", self.packet) # Read the request. # Old versions of the client sent a double \n after the signature, # which screws up the RFC5321 format. Specifically handle that # here - this could be removed in time. request = email.message_from_bytes( self.rfile.read().replace(b"\n\n", b"\n") + b"\n") # Ensure that the response can be paired with the request. self.response["Thread"] = request["Thread"] # If this is an authenticated request, then check the authentication # details. user = request["User"] or pyzor.anonymous_user if user != pyzor.anonymous_user: try: pyzor.account.verify_signature(request, self.server.accounts[user]) except KeyError: raise pyzor.SignatureError("Unknown user.") if "PV" not in request: raise pyzor.ProtocolError("Protocol Version not specified in " "request") # The protocol version is compatible if the major number is # identical (changes in the minor number are unimportant). try: if int(float(request["PV"])) != int(pyzor.proto_version): raise pyzor.UnsupportedVersionError() except ValueError: self.server.log.warn("Invalid PV: %s", request["PV"]) raise pyzor.ProtocolError("Invalid Protocol Version") # Check that the user has permission to execute the requested # operation. opcode = request["Op"] if opcode not in self.server.acl[user]: raise pyzor.AuthorizationError( "User is not authorized to request the operation.") self.server.log.debug("Got a %s command from %s", opcode, self.client_address[0]) # Get a handle to the appropriate method to execute this operation. try: dispatch = self.dispatches[opcode] except KeyError: raise NotImplementedError("Requested operation is not " "implemented.") # Get the existing record from the database (or a blank one if # there is no matching record). digests = request.get_all("Op-Digest") # Do the requested operation, log what we have done, and return. if dispatch and digests: dispatch(self, digests) self.server.usage_log.info("%s,%s,%s,%r,%s", user, self.client_address[0], opcode, digests, self.response["Code"]) def handle_error(self, code, message): """Create an appropriate response for an error.""" self.server.usage_log.error("%s: %s", code, message) self.response.replace_header("Code", "%d" % code) self.response.replace_header("Diag", message) def handle_pong(self, digests): """Handle the 'pong' command. This command returns maxint for report counts and 0 whitelist. """ self.server.log.debug("Request pong for %s", digests[0]) self.response["Count"] = "%d" % sys.maxsize self.response["WL-Count"] = "%d" % 0 def handle_check(self, digests): """Handle the 'check' command. This command returns the spam/ham counts for the specified digest. """ digest = digests[0] try: record = self.server.database[digest] except KeyError: record = pyzor.engines.common.Record() self.server.log.debug("Request to check digest %s", digest) self.response["Count"] = "%d" % record.r_count self.response["WL-Count"] = "%d" % record.wl_count def handle_report(self, digests): """Handle the 'report' command in a single step. This command increases the spam count for the specified digests.""" self.server.log.debug("Request to report digests %s", digests) if self.server.one_step: self.server.database.report(digests) else: for digest in digests: try: record = self.server.database[digest] except KeyError: record = pyzor.engines.common.Record() record.r_increment() self.server.database[digest] = record if self.server.forwarder: for digest in digests: self.server.forwarder.queue_forward_request(digest) def handle_whitelist(self, digests): """Handle the 'whitelist' command in a single step. This command increases the ham count for the specified digests.""" self.server.log.debug("Request to whitelist digests %s", digests) if self.server.one_step: self.server.database.whitelist(digests) else: for digest in digests: try: record = self.server.database[digest] except KeyError: record = pyzor.engines.common.Record() record.wl_increment() self.server.database[digest] = record if self.server.forwarder: for digest in digests: self.server.forwarder.queue_forward_request(digest, True) def handle_info(self, digests): """Handle the 'info' command. This command returns diagnostic data about a digest (timestamps for when the digest was first/last seen as spam/ham, and spam/ham counts). """ digest = digests[0] try: record = self.server.database[digest] except KeyError: record = pyzor.engines.common.Record() self.server.log.debug("Request for information about digest %s", digest) def time_output(time_obj): """Convert a datetime object to a POSIX timestamp. If the object is None, then return 0. """ if not time_obj: return 0 return time.mktime(time_obj.timetuple()) self.response["Entered"] = "%d" % time_output(record.r_entered) self.response["Updated"] = "%d" % time_output(record.r_updated) self.response["WL-Entered"] = "%d" % time_output(record.wl_entered) self.response["WL-Updated"] = "%d" % time_output(record.wl_updated) self.response["Count"] = "%d" % record.r_count self.response["WL-Count"] = "%d" % record.wl_count dispatches = { 'ping': None, 'pong': handle_pong, 'info': handle_info, 'check': handle_check, 'report': handle_report, 'whitelist': handle_whitelist, }
15,776
37.574572
79
py
pyzor
pyzor-master/pyzor/account.py
"""A collection of utilities that facilitate working with Pyzor accounts. Note that accounts are not necessary (on the client or server), as an "anonymous" account always exists.""" import time import hashlib import pyzor def sign_msg(hashed_key, timestamp, msg, hash_=hashlib.sha1): """Converts the key, timestamp (epoch seconds), and msg into a digest. lower(H(H(M) + ':' T + ':' + K)) M is message T is integer epoch timestamp K is hashed_key H is the hash function (currently SHA1) """ msg = msg.as_string().strip().encode("utf8") digest = hash_() digest.update(hash_(msg).digest()) digest.update((":%d:%s" % (timestamp, hashed_key)).encode("utf8")) return digest.hexdigest().lower() def hash_key(key, user, hash_=hashlib.sha1): """Returns the hash key for this username and password. lower(H(U + ':' + lower(K))) K is key (hex string) U is username H is the hash function (currently SHA1) """ result = ("%s:%s" % (user, key.lower())).encode("utf8") return hash_(result).hexdigest().lower() def verify_signature(msg, user_key): """Verify that the provided message is correctly signed. The message must have "User", "Time", and "Sig" headers. If the signature is valid, then the function returns normally. If the signature is not valid, then a pyzor.SignatureError() exception is raised.""" timestamp = int(msg["Time"]) user = msg["User"] provided_signature = msg["Sig"] # Check that this signature is not too old. if abs(time.time() - timestamp) > pyzor.MAX_TIMESTAMP_DIFFERENCE: raise pyzor.SignatureError("Timestamp not within allowed range.") # Calculate what the correct signature is. hashed_user_key = hash_key(user_key, user) # The signature is not part of the message that is signed. del msg["Sig"] correct_signature = sign_msg(hashed_user_key, timestamp, msg) if correct_signature != provided_signature: raise pyzor.SignatureError("Invalid signature.") class Account(object): def __init__(self, username, salt, key): self.username = username self.salt = salt self.key = key def key_from_hexstr(s): try: salt, key = s.split(",") except ValueError: raise ValueError("Invalid number of parts for key; perhaps you " "forgot the comma at the beginning for the " "salt divider?") return salt, key AnonymousAccount = Account(pyzor.anonymous_user, None, "")
2,546
30.8375
74
py
pyzor
pyzor-master/pyzor/config.py
"""Functions that handle parsing pyzor configuration files.""" import os import re import logging import collections try: import sentry_sdk _has_sentry = True except ImportError: _has_sentry = False import pyzor.account _COMMENT_P = re.compile(r"((?<=[^\\])#.*)") # Configuration files for the Pyzor Server def load_access_file(access_fn, accounts): """Load the ACL from the specified file, if it exists, and return an ACL dictionary, where each key is a username and each value is a set of allowed permissions (if the permission is not in the set, then it is not allowed). 'accounts' is a dictionary of accounts that exist on the server - only the keys are used, which must be the usernames (these are the users that are granted permission when the 'all' keyword is used, as described below). Each line of the file should be in the following format: operation : user : allow|deny where 'operation' is a space-separated list of pyzor commands or the keyword 'all' (meaning all commands), 'username' is a space-separated list of usernames or the keyword 'all' (meaning all users) - the anonymous user is called "anonymous", and "allow|deny" indicates whether or not the specified user(s) may execute the specified operations. The file is processed from top to bottom, with the final match for user/operation being the value taken. Every file has the following implicit final rule: all : all : deny If the file does not exist, then the following default is used: check report ping info : anonymous : allow """ log = logging.getLogger("pyzord") # A defaultdict is safe, because if we get a non-existant user, we get # the empty set, which is the same as a deny, which is the final # implicit rule. acl = collections.defaultdict(set) if not os.path.exists(access_fn): log.info("Using default ACL: the anonymous user may use the check, " "report, ping and info commands.") acl[pyzor.anonymous_user] = set(("check", "report", "ping", "pong", "info")) return acl accessf = open(access_fn) for line in accessf: if not line.strip() or line[0] == "#": continue try: operations, users, allowed = [part.lower().strip() for part in line.split(":")] except ValueError: log.warn("Invalid ACL line: %r", line) continue try: allowed = {"allow": True, "deny": False}[allowed] except KeyError: log.warn("Invalid ACL line: %r", line) continue if operations == "all": operations = ("check", "report", "ping", "pong", "info", "whitelist") else: operations = [operation.strip() for operation in operations.split()] if users == "all": users = accounts else: users = [user.strip() for user in users.split()] for user in users: if allowed: log.debug("Granting %s to %s.", ",".join(operations), user) # If these operations are already allowed, this will have # no effect. acl[user].update(operations) else: log.debug("Revoking %s from %s.", ",".join(operations), user) # If these operations are not allowed yet, this will have # no effect. acl[user].difference_update(operations) accessf.close() log.info("ACL: %r", acl) return acl def load_passwd_file(passwd_fn): """Load the accounts from the specified file. Each line of the file should be in the format: username : key If the file does not exist, then an empty dictionary is returned; otherwise, a dictionary of (username, key) items is returned. """ log = logging.getLogger("pyzord") accounts = {} if not os.path.exists(passwd_fn): log.info("Accounts file does not exist - only the anonymous user " "will be available.") return accounts passwdf = open(passwd_fn) for line in passwdf: if not line.strip() or line[0] == "#": continue try: user, key = line.split(":") except ValueError: log.warn("Invalid accounts line: %r", line) continue user = user.strip() key = key.strip() log.debug("Creating an account for %s with key %s.", user, key) accounts[user] = key passwdf.close() # Don't log the keys at 'info' level, just ther usernames. log.info("Accounts: %s", ",".join(accounts)) return accounts # Configuration files for the Pyzor Client def load_accounts(filepath): """Layout of file is: host : port : username : salt,key""" accounts = {} log = logging.getLogger("pyzor") if os.path.exists(filepath): accountsf = open(filepath) for lineno, orig_line in enumerate(accountsf): line = orig_line.strip() if not line or line.startswith('#'): continue try: host, port, username, key = [x.strip() for x in line.split(":")] except ValueError: log.warn("account file: invalid line %d: wrong number of " "parts", lineno) continue try: port = int(port) except ValueError as ex: log.warn("account file: invalid line %d: %s", lineno, ex) continue address = (host, port) try: salt, key = pyzor.account.key_from_hexstr(key) except ValueError as ex: log.warn("account file: invalid line %d: %s", lineno, ex) continue if not salt and not key: log.warn("account file: invalid line %d: keystuff can't be " "all None's", lineno) continue accounts[address] = pyzor.account.Account(username, salt, key) accountsf.close() else: log.warn("No accounts are setup. All commands will be executed by " "the anonymous user.") return accounts def load_servers(filepath): """Load the servers file.""" logger = logging.getLogger("pyzor") if not os.path.exists(filepath): servers = [] else: servers = [] with open(filepath) as serverf: for line in serverf: line = line.strip() if re.match("[^#][a-zA-Z0-9.-]+:[0-9]+", line): address, port = line.rsplit(":", 1) servers.append((address, int(port))) if not servers: logger.info("No servers specified, defaulting to public.pyzor.org.") servers = [("public.pyzor.org", 24441)] return servers def load_local_whitelist(filepath): """Load the local digest skip file.""" if not os.path.exists(filepath): return set() whitelist = set() with open(filepath) as serverf: for line in serverf: # Remove any comments line = _COMMENT_P.sub("", line).strip() if line: whitelist.add(line) return whitelist # Common configurations def setup_logging(log_name, filepath, debug, sentry_dsn=None, sentry_lvl="WARN"): """Setup logging according to the specified options. Return the Logger object. """ fmt = logging.Formatter('%(asctime)s (%(process)d) %(levelname)s ' '%(message)s') stream_handler = logging.StreamHandler() if debug: stream_log_level = logging.DEBUG file_log_level = logging.DEBUG else: stream_log_level = logging.CRITICAL file_log_level = logging.INFO logger = logging.getLogger(log_name) logger.setLevel(file_log_level) stream_handler.setLevel(stream_log_level) stream_handler.setFormatter(fmt) logger.addHandler(stream_handler) if filepath: file_handler = logging.FileHandler(filepath) file_handler.setLevel(file_log_level) file_handler.setFormatter(fmt) logger.addHandler(file_handler) if sentry_dsn and _has_sentry: sentry_sdk.init("https://8cb95c088d04414e885879898a952d05@sentry.io/1443213") return logger def expand_homefiles(homefiles, category, homedir, config): """Set the full file path for these configuration files.""" for filename in homefiles: filepath = config.get(category, filename) if not filepath: continue filepath = os.path.expanduser(filepath) if not os.path.isabs(filepath): filepath = os.path.join(homedir, filepath) config.set(category, filename, filepath)
9,034
34.155642
85
py
pyzor
pyzor-master/pyzor/client.py
"""Networked spam-signature detection client. >>> import pyzor >>> import pyzor.client >>> import pyzor.digest >>> import pyzor.config To load the accounts file: >>> accounts = pyzor.config.load_accounts(filename) To create a client (to then issue commands): >>> client = pyzor.client.Client(accounts) To create a client, using the anonymous user: >>> client = pyzor.client.Client() To get a digest (of an email.message.Message object, or similar): >>> digest = pyzor.digest.DataDigester(msg).value To query a server (where address is a (host, port) pair): >>> client.ping(address) >>> client.info(digest, address) >>> client.report(digest, address) >>> client.whitelist(digest, address) >>> client.check(digest, address) To query the default server (public.pyzor.org): >>> client.ping() >>> client.info(digest) >>> client.report(digest) >>> client.whitelist(digest) >>> client.check(digest) Response will contain, depending on the type of request, some of the following keys (e.g. client.ping()['Code']): All responses will have: - 'Diag' 'OK' or error message - 'Code' '200' if OK - 'PV' Protocol Version - 'Thread' `info` and `check` responses will also contain: - '[WL-]Count' Whitelist/Blacklist count `info` responses will also have: - '[WL-]Entered' timestamp when message was first whitelisted/blacklisted - '[WL-]Updated' timestamp when message was last whitelisted/blacklisted """ import time import email import socket import logging import functools import collections import pyzor.digest import pyzor.account import pyzor.message import pyzor.hacks.py26 pyzor.hacks.py26.hack_email() class Client(object): timeout = 5 max_packet_size = 8192 def __init__(self, accounts=None, timeout=None, spec=None): if accounts is None: accounts = {} self.accounts = dict(((host, int(port)), account) for (host, port), account in accounts.items()) if spec is None: spec = pyzor.digest.digest_spec self.spec = spec if timeout is not None: self.timeout = timeout self.log = logging.getLogger("pyzor") def ping(self, address=("public.pyzor.org", 24441)): msg = pyzor.message.PingRequest() sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def pong(self, digest, address=("public.pyzor.org", 24441)): msg = pyzor.message.PongRequest(digest) sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def info(self, digest, address=("public.pyzor.org", 24441)): msg = pyzor.message.InfoRequest(digest) sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def report(self, digest, address=("public.pyzor.org", 24441)): msg = pyzor.message.ReportRequest(digest, self.spec) sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def whitelist(self, digest, address=("public.pyzor.org", 24441)): msg = pyzor.message.WhitelistRequest(digest, self.spec) sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def check(self, digest, address=("public.pyzor.org", 24441)): msg = pyzor.message.CheckRequest(digest) sock = self.send(msg, address) return self.read_response(sock, msg.get_thread()) def _mock_check(self, digests, address=None): msg = (u"Code: %s\nDiag: OK\nPV: %s\nThread: 1024\nCount: 0\n" u"WL-Count: 0" % (pyzor.message.Response.ok_code, pyzor.proto_version)).encode('ascii') return email.message_from_bytes(msg, _class=pyzor.message.Response) def send(self, msg, address=("public.pyzor.org", 24441)): address = (address[0], int(address[1])) msg.init_for_sending() try: account = self.accounts[address] except KeyError: account = pyzor.account.AnonymousAccount timestamp = int(time.time()) msg["User"] = account.username msg["Time"] = str(timestamp) msg["Sig"] = pyzor.account.sign_msg(pyzor.account.hash_key( account.key, account.username), timestamp, msg) self.log.debug("sending: %r", msg.as_string()) return self._send(msg, address) @staticmethod def _send(msg, addr): sock = None for res in socket.getaddrinfo(addr[0], addr[1], 0, socket.SOCK_DGRAM, socket.IPPROTO_UDP): af, socktype, proto, _, sa = res try: sock = socket.socket(af, socktype, proto) except socket.error: sock = None continue try: sock.sendto(msg.as_string().encode("utf8"), 0, sa) except socket.timeout: sock.close() raise pyzor.TimeoutError("Sending to %s time-outed" % sa) except socket.error: sock.close() sock = None continue break if sock is None: raise pyzor.CommError("Unable to send to %s:%s" % addr) return sock def read_response(self, sock, expected_id): sock.settimeout(self.timeout) try: packet, address = sock.recvfrom(self.max_packet_size) except socket.timeout as ex: sock.close() raise pyzor.TimeoutError("Reading response timed-out.") except socket.error as ex: sock.close() raise pyzor.CommError("Socket error while reading response: %s" % ex) self.log.debug("received: %r/%r", packet, address) msg = email.message_from_bytes(packet, _class=pyzor.message.Response) msg.ensure_complete() try: thread_id = msg.get_thread() if thread_id != expected_id: if thread_id.in_ok_range(): raise pyzor.ProtocolError( "received unexpected thread id %d (expected %d)" % (thread_id, expected_id)) self.log.warn("received error thread id %d (expected %d)", thread_id, expected_id) except KeyError: self.log.warn("no thread id received") return msg class BatchClient(Client): """Like the normal Client but with support for batching reports.""" def __init__(self, accounts=None, timeout=None, spec=None, batch_size=10): Client.__init__(self, accounts=accounts, timeout=timeout, spec=spec) self.batch_size = batch_size self.r_requests = {} self.w_requests = {} self.flush() def report(self, digest, address=("public.pyzor.org", 24441)): self._add_digest(digest, address, self.r_requests) def whitelist(self, digest, address=("public.pyzor.org", 24441)): self._add_digest(digest, address, self.w_requests) def _add_digest(self, digest, address, requests): address = (address[0], int(address[1])) msg = requests[address] msg.add_digest(digest) if msg.digest_count >= self.batch_size: try: return self.send(msg, address) finally: del requests[address] def flush(self): """Deleting any saved digest reports.""" self.r_requests = collections.defaultdict( functools.partial(pyzor.message.ReportRequest, spec=self.spec)) self.w_requests = collections.defaultdict( functools.partial(pyzor.message.WhitelistRequest, spec=self.spec)) def force(self): """Force send any remaining reports.""" for address, msg in self.r_requests.items(): try: self.send(msg, address) except: continue for address, msg in self.w_requests.items(): try: self.send(msg, address) except: continue def __del__(self): self.force() class ClientRunner(object): def __init__(self, routine): self.log = logging.getLogger("pyzor") self.routine = routine self.all_ok = True self.results = [] def run(self, server, args, kwargs=None): if kwargs is None: kwargs = {} message = "%s:%s\t" % server response = None try: response = self.routine(*args, **kwargs) self.handle_response(response, message) except (pyzor.CommError, KeyError, ValueError) as e: self.results.append("%s%s\n" % (message, (e.code, str(e)))) self.log.error("%s\t%s: %s", server, e.__class__.__name__, e) self.all_ok = False def handle_response(self, response, message): """mesaage is a string we've built up so far""" if not response.is_ok(): self.all_ok = False self.results.append("%s%s\n" % (message, response.head_tuple())) class CheckClientRunner(ClientRunner): def __init__(self, routine, r_count=0, wl_count=0): ClientRunner.__init__(self, routine) self.found_hit = False self.whitelisted = False self.hit_count = 0 self.whitelist_count = 0 self.r_count_found = r_count self.wl_count_clears = wl_count def handle_response(self, response, message): message += "%s\t" % str(response.head_tuple()) if response.is_ok(): self.hit_count = int(response['Count']) self.whitelist_count = int(response['WL-Count']) if self.whitelist_count > self.wl_count_clears: self.whitelisted = True elif self.hit_count > self.r_count_found: self.found_hit = True message += "%d\t%d" % (self.hit_count, self.whitelist_count) else: self.all_ok = False self.results.append(message + "\n") class InfoClientRunner(ClientRunner): def handle_response(self, response, message): message += "%s\n" % str(response.head_tuple()) if response.is_ok(): for key in ('Count', 'Entered', 'Updated', 'WL-Count', 'WL-Entered', 'WL-Updated'): if key in response: val = int(response[key]) if 'Count' in key: stringed = str(val) elif val == -1: stringed = 'Never' else: stringed = time.ctime(val) message += ("\t%s: %s\n" % (key, stringed)) else: self.all_ok = False self.results.append(message + "\n")
10,826
33.262658
78
py
pyzor
pyzor-master/pyzor/__init__.py
"""Networked spam-signature detection.""" __author__ = "Frank J. Tobin, ftobin@neverending.org" __credits__ = "Tony Meyer, Dreas von Donselaar, all the Pyzor contributors." __version__ = "1.1.1" import hashlib proto_name = 'pyzor' proto_version = 2.1 anonymous_user = 'anonymous' # We would like to use sha512, but that would mean that all the digests # changed, so for now, we stick with sha1 (which is the same as the old # sha module). sha = hashlib.sha1 # This is the maximum time between a client signing a Pyzor request and the # server checking the signature. MAX_TIMESTAMP_DIFFERENCE = 300 # seconds class CommError(Exception): """Something in general went wrong with the transaction.""" code = 400 class ProtocolError(CommError): """Something is wrong with talking the protocol.""" code = 400 class TimeoutError(CommError): """The connection timed out.""" code = 504 class IncompleteMessageError(ProtocolError): """A complete requested was not received.""" pass class UnsupportedVersionError(ProtocolError): """Client is using an unsupported protocol version.""" pass class SignatureError(CommError): """Unknown user, signature on msg invalid, or not within allowed time range.""" pass class AuthorizationError(CommError): """The signature was valid, but the user is not permitted to do the requested action.""" pass
1,410
23.327586
76
py
pyzor
pyzor-master/pyzor/forwarder.py
"""Manage the forwarder process.""" import logging import threading try: import Queue except ImportError: import queue as Queue class Forwarder(object): """Forwards digest to remote pyzor servers""" def __init__(self, forwarding_client, remote_servers, max_queue_size=10000): """ forward_client: a pyzor.client.Client instance to use as forwarding client remote_servers: a list of (hostname,port) tuples where digests should be forwarded to max_queue_size: max amount of queued digests """ self.log = logging.getLogger("pyzord") self.forwarding_client = forwarding_client self.forward_queue = Queue.Queue(max_queue_size) self.remote_servers = remote_servers def _forward_loop(self): """read forwarding requests from the queue""" while True: try: digest, whitelist = self.forward_queue.get(block=True, timeout=2) except Queue.Empty: # If the forwarding client has been deleted we should # end the thread if self.forwarding_client is None: return else: continue for server in self.remote_servers: try: if whitelist: self.forwarding_client.whitelist(digest, server) else: self.forwarding_client.report(digest, server) except Exception as ex: self.log.warn('Forwarding digest %s to %s failed: %s', digest, server, ex) def queue_forward_request(self, digest, whitelist=False): """If forwarding is enabled, insert a digest into the forwarding queue if whitelist is True, the digest will be forwarded as whitelist request if the queue is full, the digest is dropped """ if self.forwarding_client is None: # forwarding has been disabled return try: self.forward_queue.put_nowait((digest, whitelist),) except Queue.Full: pass def start_forwarding(self): """start the forwarding thread""" threading.Thread(target=self._forward_loop).start() def stop_forwarding(self): """disable forwarding and tell the forwarding thread to end itself""" self.forwarding_client = None
2,568
34.191781
79
py
pyzor
pyzor-master/pyzor/engines/redis_.py
"""Redis database engine.""" import time import logging import datetime import functools try: import redis _has_redis = True except ImportError: redis = None _has_redis = False from pyzor.engines.common import * VERSION = "1" NAMESPACE = "pyzord.digest_v%s" % VERSION def encode_date(date): """Convert the date to Unix Timestamp""" if date is None: return 0 return int(time.mktime(date.timetuple())) def decode_date(stamp): """Return a datetime object from a Unix Timestamp.""" stamp = int(stamp) if stamp == 0: return None return datetime.datetime.fromtimestamp(stamp) def safe_call(f): """Decorator that wraps a method for handling database operations.""" def wrapped_f(self, *args, **kwargs): # This only logs the error and raise the usual Error for consistency, # the redis library takes care of reconnecting and everything else. try: return f(self, *args, **kwargs) except redis.exceptions.RedisError as e: self.log.error("Redis error while calling %s: %s", f.__name__, e) raise DatabaseError("Database temporarily unavailable.") return wrapped_f class RedisDBHandle(BaseEngine): absolute_source = False handles_one_step = True log = logging.getLogger("pyzord") def __init__(self, fn, mode, max_age=None): self.max_age = max_age # The 'fn' is host,port,password,db. We ignore mode. # We store the authentication details so that we can reconnect if # necessary. self._dsn = fn fn = fn.split(",") self.host = fn[0] or "localhost" self.port = fn[1] or "6379" self.passwd = fn[2] or None self.db_name = fn[3] or "0" self.db = self._get_new_connection() self._check_version() @staticmethod def _encode_record(r): return {"r_count": r.r_count, "r_entered": encode_date(r.r_entered), "r_updated": encode_date(r.r_updated), "wl_count": r.wl_count, "wl_entered": encode_date(r.wl_entered), "wl_updated": encode_date(r.wl_updated) } @staticmethod def _decode_record(r): if not r: return Record() return Record(r_count=int(r.get(b"r_count", 0)), r_entered=decode_date(r.get(b"r_entered", 0)), r_updated=decode_date(r.get(b"r_updated", 0)), wl_count=int(r.get(b"wl_count", 0)), wl_entered=decode_date(r.get(b"wl_entered", 0)), wl_updated=decode_date(r.get(b"wl_updated", 0))) def __iter__(self): for key in self.db.keys(self._real_key("*")): yield key.rsplit(".", 1)[-1] def _iteritems(self): for key in self: try: yield key, self[key] except Exception as ex: self.log.warning("Invalid record %s: %s", key, ex) def iteritems(self): return self._iteritems() def items(self): return list(self._iteritems()) @staticmethod def _real_key(key): return "%s.%s" % (NAMESPACE, key) @safe_call def _get_new_connection(self): if "/" in self.host: return redis.StrictRedis(unix_socket_path=self.host, db=int(self.db_name), password=self.passwd) return redis.StrictRedis(host=self.host, port=int(self.port), db=int(self.db_name), password=self.passwd) @safe_call def __getitem__(self, key): return self._decode_record(self.db.hgetall(self._real_key(key))) @safe_call def __setitem__(self, key, value): real_key = self._real_key(key) self.db.hmset(real_key, self._encode_record(value)) if self.max_age is not None: self.db.expire(real_key, self.max_age) @safe_call def __delitem__(self, key): self.db.delete(self._real_key(key)) @safe_call def report(self, keys): now = int(time.time()) for key in keys: real_key = self._real_key(key) self.db.hincrby(real_key, "r_count") self.db.hsetnx(real_key, "r_entered", now) self.db.hset(real_key, "r_updated", now) if self.max_age: self.db.expire(real_key, self.max_age) @safe_call def whitelist(self, keys): now = int(time.time()) for key in keys: real_key = self._real_key(key) self.db.hincrby(real_key, "wl_count") self.db.hsetnx(real_key, "wl_entered", now) self.db.hset(real_key, "wl_updated", now) if self.max_age: self.db.expire(real_key, self.max_age) @classmethod def get_prefork_connections(cls, fn, mode, max_age=None): """Yields a number of database connections suitable for a Pyzor pre-fork server. """ while True: yield functools.partial(cls, fn, mode, max_age=max_age) def _check_version(self): """Check if there are deprecated records and warn the user.""" old_keys = len(self.db.keys("pyzord.digest.*")) if old_keys: cmd = ("pyzor-migrate --delete --se=redis_v0 --sd=%s " "--de=redis --dd=%s" % (self._dsn, self._dsn)) self.log.critical("You have %s records in the deprecated version " "of the redis engine.", old_keys) self.log.critical("Please migrate the records with: %r", cmd) class ThreadedRedisDBHandle(RedisDBHandle): def __init__(self, fn, mode, max_age=None, bound=None): RedisDBHandle.__init__(self, fn, mode, max_age=max_age) if not _has_redis: handle = DBHandle(single_threaded=None, multi_threaded=None, multi_processing=None, prefork=None) else: handle = DBHandle(single_threaded=RedisDBHandle, multi_threaded=ThreadedRedisDBHandle, multi_processing=None, prefork=RedisDBHandle)
6,235
31.310881
80
py
pyzor
pyzor-master/pyzor/engines/gdbm_.py
"""Gdbm database engine.""" try: import gdbm as gdbm _has_gdbm = True except ImportError: try: import dbm.gnu as gdbm _has_gdbm = True except ImportError: _has_gdbm = False import time import logging import datetime import threading from pyzor.engines.common import Record, DBHandle, BaseEngine def _dt_decode(datetime_str): """Decode a string into a datetime object.""" if datetime_str == 'None': return None try: return datetime.datetime.strptime(datetime_str, "%Y-%m-%d %H:%M:%S.%f") except ValueError: return datetime.datetime.strptime(datetime_str, "%Y-%m-%d %H:%M:%S") class GdbmDBHandle(BaseEngine): absolute_source = True handles_one_step = False sync_period = 60 reorganize_period = 3600 * 24 # 1 day fields = ('r_count', 'r_entered', 'r_updated', 'wl_count', 'wl_entered', 'wl_updated') _fields = [('r_count', int), ('r_entered', _dt_decode), ('r_updated', _dt_decode), ('wl_count', int), ('wl_entered', _dt_decode), ('wl_updated', _dt_decode)] this_version = '1' log = logging.getLogger("pyzord") def __init__(self, fn, mode, max_age=None): self.max_age = max_age self.db = gdbm.open(fn, mode) self.reorganize_timer = None self.sync_timer = None self.start_reorganizing() self.start_syncing() def __iter__(self): k = self.db.firstkey() while k is not None: yield k k = self.db.nextkey(k) def _iteritems(self): for k in self: try: yield k, self._really_getitem(k) except Exception as e: self.log.warning("Invalid record %s: %s", k, e) def iteritems(self): return self._iteritems() def items(self): return list(self._iteritems()) def apply_method(self, method, varargs=(), kwargs=None): if kwargs is None: kwargs = {} return method(*varargs, **kwargs) def __getitem__(self, key): return self.apply_method(self._really_getitem, (key,)) def _really_getitem(self, key): return GdbmDBHandle.decode_record(self.db[key]) def __setitem__(self, key, value): self.apply_method(self._really_setitem, (key, value)) def _really_setitem(self, key, value): self.db[key] = GdbmDBHandle.encode_record(value) def __delitem__(self, key): self.apply_method(self._really_delitem, (key,)) def _really_delitem(self, key): del self.db[key] def start_syncing(self): if self.db: self.apply_method(self._really_sync) self.sync_timer = threading.Timer(self.sync_period, self.start_syncing) self.sync_timer.setDaemon(True) self.sync_timer.start() def _really_sync(self): self.db.sync() def start_reorganizing(self): if not self.max_age: return if self.db: self.apply_method(self._really_reorganize) self.reorganize_timer = threading.Timer(self.reorganize_period, self.start_reorganizing) self.reorganize_timer.setDaemon(True) self.reorganize_timer.start() def _really_reorganize(self): self.log.debug("reorganizing the database") key = self.db.firstkey() breakpoint = time.time() - self.max_age while key is not None: rec = self._really_getitem(key) delkey = None if int(time.mktime(rec.r_updated.timetuple())) < breakpoint: self.log.debug("deleting key %s", key) delkey = key key = self.db.nextkey(key) if delkey: self._really_delitem(delkey) self.db.reorganize() @classmethod def encode_record(cls, value): values = [cls.this_version] values.extend(["%s" % getattr(value, x) for x in cls.fields]) return ",".join(values) @classmethod def decode_record(cls, s): try: s = s.decode("utf8") except UnicodeError: raise StandardError("don't know how to handle db value %s" % repr(s)) parts = s.split(',') version = parts[0] if len(parts) == 3: dispatch = cls.decode_record_0 elif version == '1': dispatch = cls.decode_record_1 else: raise StandardError("don't know how to handle db value %s" % repr(s)) return dispatch(s) @staticmethod def decode_record_0(s): r = Record() parts = s.split(',') fields = ('r_count', 'r_entered', 'r_updated') assert len(parts) == len(fields) for i in range(len(parts)): setattr(r, fields[i], int(parts[i])) return r @classmethod def decode_record_1(cls, s): r = Record() parts = s.split(',')[1:] assert len(parts) == len(cls.fields) for part, field in zip(parts, cls._fields): f, decode = field setattr(r, f, decode(part)) return r class ThreadedGdbmDBHandle(GdbmDBHandle): """Like GdbmDBHandle, but handles multi-threaded access.""" def __init__(self, fn, mode, max_age=None, bound=None): self.db_lock = threading.Lock() GdbmDBHandle.__init__(self, fn, mode, max_age=max_age) def apply_method(self, method, varargs=(), kwargs=None): if kwargs is None: kwargs = {} with self.db_lock: return GdbmDBHandle.apply_method(self, method, varargs=varargs, kwargs=kwargs) # This won't work because the gdbm object needs to be in shared memory of the # spawned processes. # class ProcessGdbmDBHandle(ThreadedGdbmDBHandle): # def __init__(self, fn, mode, max_age=None, bound=None): # ThreadedGdbmDBHandle.__init__(self, fn, mode, max_age=max_age, # bound=bound) # self.db_lock = multiprocessing.Lock() if not _has_gdbm: handle = DBHandle(single_threaded=None, multi_threaded=None, multi_processing=None, prefork=None) else: handle = DBHandle(single_threaded=GdbmDBHandle, multi_threaded=ThreadedGdbmDBHandle, multi_processing=None, prefork=None)
6,609
30.327014
79
py
pyzor
pyzor-master/pyzor/engines/mysql.py
"""MySQLdb database engine.""" import time import logging import datetime import itertools import functools import threading try: import Queue except ImportError: import queue as Queue try: import MySQLdb import MySQLdb.cursors _has_mysql = True except ImportError: _has_mysql = False from pyzor.engines.common import * class MySQLDBHandle(BaseEngine): absolute_source = False handles_one_step = True # The table must already exist, and have this schema: # CREATE TABLE `public` ( # `digest` char(40) default NULL, # `r_count` int(11) default NULL, # `wl_count` int(11) default NULL, # `r_entered` datetime default NULL, # `wl_entered` datetime default NULL, # `r_updated` datetime default NULL, # `wl_updated` datetime default NULL, # PRIMARY KEY (`digest`) # ) # XXX Re-organising might be faster with a r_updated index. However, # XXX the re-organisation time isn't that important, and that would # XXX (slightly) slow down all inserts, so we leave it for now. reorganize_period = 3600 * 24 # 1 day reconnect_period = 60 # seconds log = logging.getLogger("pyzord") def __init__(self, fn, mode, max_age=None): self.max_age = max_age self.db = None # The 'fn' is host,user,password,db,table. We ignore mode. # We store the authentication details so that we can reconnect if # necessary. self.host, self.user, self.passwd, self.db_name, \ self.table_name = fn.split(",") self.last_connect_attempt = 0 # We have never connected. self.reorganize_timer = None self.reconnect() self.start_reorganizing() def _get_new_connection(self): """Returns a new db connection.""" db = MySQLdb.connect(host=self.host, user=self.user, db=self.db_name, passwd=self.passwd) db.autocommit(True) return db def _check_reconnect_time(self): if time.time() - self.last_connect_attempt < self.reconnect_period: # Too soon to reconnect. self.log.debug("Can't reconnect until %s", (time.ctime(self.last_connect_attempt + self.reconnect_period))) return False return True def reconnect(self): if not self._check_reconnect_time(): return if self.db: try: self.db.close() except MySQLdb.Error: pass try: self.db = self._get_new_connection() except MySQLdb.Error as e: self.log.error("Unable to connect to database: %s", e) self.db = None # Keep track of when we connected, so that we don't retry too often. self.last_connect_attempt = time.time() def _iter(self, db): c = db.cursor(cursorclass=MySQLdb.cursors.SSCursor) c.execute("SELECT digest FROM %s" % self.table_name) while True: row = c.fetchone() if not row: break yield row[0] c.close() def __iter__(self): return self._safe_call("iter", self._iter, ()) def _iteritems(self, db): c = db.cursor(cursorclass=MySQLdb.cursors.SSCursor) c.execute("SELECT digest, r_count, wl_count, r_entered, r_updated, " "wl_entered, wl_updated FROM %s" % self.table_name) while True: row = c.fetchone() if not row: break yield row[0], Record(*row[1:]) c.close() def iteritems(self): return self._safe_call("iteritems", self._iteritems, ()) def items(self): return list(self._safe_call("iteritems", self._iteritems, ())) def __del__(self): """Close the database when the object is no longer needed.""" try: if self.db: self.db.close() except MySQLdb.Error: pass def _safe_call(self, name, method, args): try: return method(*args, db=self.db) except (MySQLdb.Error, AttributeError) as ex: self.log.error("%s failed: %s", name, ex) self.reconnect() # Retrying just complicates the logic - we don't really care if # a single query fails (and it's possible that it would fail) # on the second attempt anyway. Any exceptions are caught by # the server, and a 'nice' message provided to the caller. raise DatabaseError("Database temporarily unavailable.") def report(self, keys): return self._safe_call("report", self._report, (keys,)) def whitelist(self, keys): return self._safe_call("whitelist", self._whitelist, (keys,)) def __getitem__(self, key): return self._safe_call("getitem", self._really__getitem__, (key,)) def __setitem__(self, key, value): return self._safe_call("setitem", self._really__setitem__, (key, value)) def __delitem__(self, key): return self._safe_call("delitem", self._really__delitem__, (key,)) def _report(self, keys, db=None): c = db.cursor() try: c.executemany("INSERT INTO %s (digest, r_count, wl_count, " "r_entered, r_updated, wl_entered, wl_updated) " "VALUES (%%s, 1, 0, NOW(), NOW(), NOW(), NOW()) ON " "DUPLICATE KEY UPDATE r_count=r_count+1, " "r_updated=NOW()" % self.table_name, map(lambda key: (key,), keys)) finally: c.close() def _whitelist(self, keys, db=None): c = db.cursor() try: c.executemany("INSERT INTO %s (digest, r_count, wl_count, " "r_entered, r_updated, wl_entered, wl_updated) " "VALUES (%%s, 0, 1, NOW(), NOW(), NOW(), NOW()) ON " "DUPLICATE KEY UPDATE wl_count=wl_count+1, " "wl_updated=NOW()" % self.table_name, map(lambda key: (key,), keys)) finally: c.close() def _really__getitem__(self, key, db=None): """__getitem__ without the exception handling.""" c = db.cursor() # The order here must match the order of the arguments to the # Record constructor. c.execute("SELECT r_count, wl_count, r_entered, r_updated, " "wl_entered, wl_updated FROM %s WHERE digest=%%s" % self.table_name, (key,)) try: try: return Record(*c.fetchone()) except TypeError: # fetchone() returned None, i.e. there is no such record raise KeyError() finally: c.close() def _really__setitem__(self, key, value, db=None): """__setitem__ without the exception handling.""" c = db.cursor() try: c.execute("INSERT INTO %s (digest, r_count, wl_count, " "r_entered, r_updated, wl_entered, wl_updated) " "VALUES (%%s, %%s, %%s, %%s, %%s, %%s, %%s) ON " "DUPLICATE KEY UPDATE r_count=%%s, wl_count=%%s, " "r_entered=%%s, r_updated=%%s, wl_entered=%%s, " "wl_updated=%%s" % self.table_name, (key, value.r_count, value.wl_count, value.r_entered, value.r_updated, value.wl_entered, value.wl_updated, value.r_count, value.wl_count, value.r_entered, value.r_updated, value.wl_entered, value.wl_updated)) finally: c.close() def _really__delitem__(self, key, db=None): """__delitem__ without the exception handling.""" c = db.cursor() try: c.execute("DELETE FROM %s WHERE digest=%%s" % self.table_name, (key,)) finally: c.close() def start_reorganizing(self): if not self.max_age: return self.log.debug("reorganizing the database") breakpoint = (datetime.datetime.now() - datetime.timedelta(seconds=self.max_age)) db = self._get_new_connection() c = db.cursor() try: c.execute("DELETE FROM %s WHERE r_updated<%%s" % self.table_name, (breakpoint,)) except (MySQLdb.Error, AttributeError) as e: self.log.warn("Unable to reorganise: %s", e) finally: c.close() db.close() self.reorganize_timer = threading.Timer(self.reorganize_period, self.start_reorganizing) self.reorganize_timer.setDaemon(True) self.reorganize_timer.start() @classmethod def get_prefork_connections(cls, fn, mode, max_age=None): """Yields a number of database connections suitable for a Pyzor pre-fork server. """ # Only run the reorganize timer in the first child process. yield functools.partial(cls, fn, mode, max_age=max_age) while True: yield functools.partial(cls, fn, mode, max_age=None) class ThreadedMySQLDBHandle(MySQLDBHandle): def __init__(self, fn, mode, max_age=None, bound=None): self.bound = bound if self.bound: self.db_queue = Queue.Queue() MySQLDBHandle.__init__(self, fn, mode, max_age=max_age) def _get_connection(self): if self.bound: return self.db_queue.get() else: return self._get_new_connection() def _release_connection(self, db): if self.bound: self.db_queue.put(db) else: db.close() def _safe_call(self, name, method, args): db = self._get_connection() try: return method(*args, db=db) except (MySQLdb.Error, AttributeError) as ex: self.log.error("%s failed: %s", name, ex) if not self.bound: raise DatabaseError("Database temporarily unavailable.") try: # Connection might be timeout, ping and retry db.ping(True) return method(*args, db=db) except (MySQLdb.Error, AttributeError) as ex: # attempt a new connection, if we can retry db = self._reconnect(db) raise DatabaseError("Database temporarily unavailable.") finally: self._release_connection(db) def reconnect(self): if not self.bound: return for _ in range(self.bound): self.db_queue.put(self._get_new_connection()) def _reconnect(self, db): if not self._check_reconnect_time(): return db else: self.last_connect_attempt = time.time() return self._get_new_connection() def __del__(self): if not self.bound: return for db in iter(self.db_queue.get_nowait): try: db.close() except MySQLdb.Error: continue except Queue.Empty: break class ProcessMySQLDBHandle(MySQLDBHandle): def __init__(self, fn, mode, max_age=None): MySQLDBHandle.__init__(self, fn, mode, max_age=max_age) def reconnect(self): pass def __del__(self): pass def _safe_call(self, name, method, args): db = None try: db = self._get_new_connection() return method(*args, db=db) except (MySQLdb.Error, AttributeError) as ex: self.log.error("%s failed: %s", name, ex) raise DatabaseError("Database temporarily unavailable.") finally: if db is not None: db.close() if not _has_mysql: handle = DBHandle(single_threaded=None, multi_threaded=None, multi_processing=None, prefork=None) else: handle = DBHandle(single_threaded=MySQLDBHandle, multi_threaded=ThreadedMySQLDBHandle, multi_processing=ProcessMySQLDBHandle, prefork=MySQLDBHandle)
12,438
34.438746
78
py
pyzor
pyzor-master/pyzor/engines/redis_v0.py
"""Redis database engine. XXX Deprecated version. """ import logging import datetime import functools try: import redis _has_redis = True except ImportError: redis = None _has_redis = False from pyzor.engines.common import * NAMESPACE = "pyzord.digest" encode_date = lambda d: "" if d is None else d.strftime("%Y-%m-%d %H:%M:%S") decode_date = lambda x: None if x == "" else datetime.datetime.strptime( x, "%Y-%m-%d %H:%M:%S") def safe_call(f): """Decorator that wraps a method for handling database operations.""" def wrapped_f(self, *args, **kwargs): # This only logs the error and raise the usual Error for consistency, # the redis library takes care of reconnecting and everything else. try: return f(self, *args, **kwargs) except redis.exceptions.RedisError as e: self.log.error("Redis error while calling %s: %s", f.__name__, e) raise DatabaseError("Database temporarily unavailable.") return wrapped_f class RedisDBHandle(BaseEngine): absolute_source = False handles_one_step = False log = logging.getLogger("pyzord") def __init__(self, fn, mode, max_age=None): self.max_age = max_age # The 'fn' is host,port,password,db. We ignore mode. # We store the authentication details so that we can reconnect if # necessary. fn = fn.split(",") self.host = fn[0] or "localhost" self.port = fn[1] or "6379" self.passwd = fn[2] or None self.db_name = fn[3] or "0" self.db = self._get_new_connection() @staticmethod def _encode_record(r): return ("%s,%s,%s,%s,%s,%s" % (r.r_count, encode_date(r.r_entered), encode_date(r.r_updated), r.wl_count, encode_date(r.wl_entered), encode_date(r.wl_updated))).encode() @staticmethod def _decode_record(r): if r is None: return Record() fields = r.decode().split(",") return Record(r_count=int(fields[0]), r_entered=decode_date(fields[1]), r_updated=decode_date(fields[2]), wl_count=int(fields[3]), wl_entered=decode_date(fields[4]), wl_updated=decode_date(fields[5])) def __iter__(self): for key in self.db.keys(self._real_key("*")): yield key.rsplit(".", 1)[-1] def _iteritems(self): for key in self: try: yield key, self[key] except Exception as ex: self.log.warning("Invalid record %s: %s", key, ex) def iteritems(self): return self._iteritems() def items(self): return list(self._iteritems()) @staticmethod def _real_key(key): return "%s.%s" % (NAMESPACE, key) @safe_call def _get_new_connection(self): if "/" in self.host: return redis.StrictRedis(unix_socket_path=self.host, db=int(self.db_name), password=self.passwd) return redis.StrictRedis(host=self.host, port=int(self.port), db=int(self.db_name), password=self.passwd) @safe_call def __getitem__(self, key): return self._decode_record(self.db.get(self._real_key(key))) @safe_call def __setitem__(self, key, value): if self.max_age is None: self.db.set(self._real_key(key), self._encode_record(value)) else: self.db.setex(self._real_key(key), self.max_age, self._encode_record(value)) @safe_call def __delitem__(self, key): self.db.delete(self._real_key(key)) @classmethod def get_prefork_connections(cls, fn, mode, max_age=None): """Yields a number of database connections suitable for a Pyzor pre-fork server. """ while True: yield functools.partial(cls, fn, mode, max_age=max_age) class ThreadedRedisDBHandle(RedisDBHandle): def __init__(self, fn, mode, max_age=None, bound=None): RedisDBHandle.__init__(self, fn, mode, max_age=max_age) if not _has_redis: handle = DBHandle(single_threaded=None, multi_threaded=None, multi_processing=None, prefork=None) else: handle = DBHandle(single_threaded=RedisDBHandle, multi_threaded=ThreadedRedisDBHandle, multi_processing=None, prefork=RedisDBHandle)
4,666
29.703947
80
py
pyzor
pyzor-master/pyzor/engines/common.py
"""Common library shared by different engines.""" import sys import datetime from collections import namedtuple __all__ = ["DBHandle", "DatabaseError", "Record", "BaseEngine"] DBHandle = namedtuple("DBHandle", ["single_threaded", "multi_threaded", "multi_processing", "prefork"]) class DatabaseError(Exception): pass class Record(object): """Prefix conventions used in this class: r = report (spam) wl = whitelist """ def __init__(self, r_count=0, wl_count=0, r_entered=None, r_updated=None, wl_entered=None, wl_updated=None): self.r_count = r_count self.wl_count = wl_count self.r_entered = r_entered self.r_updated = r_updated self.wl_entered = wl_entered self.wl_updated = wl_updated def wl_increment(self): # overflow prevention if self.wl_count < sys.maxsize: self.wl_count += 1 if self.wl_entered is None: self.wl_entered = datetime.datetime.now() self.wl_update() def r_increment(self): # overflow prevention if self.r_count < sys.maxsize: self.r_count += 1 if self.r_entered is None: self.r_entered = datetime.datetime.now() self.r_update() def r_update(self): self.r_updated = datetime.datetime.now() def wl_update(self): self.wl_updated = datetime.datetime.now() class BaseEngine(object): """Base class for Pyzor engines.""" absolute_source = True handles_one_step = False def __iter__(self): """Iterate over all keys""" raise NotImplementedError() def iteritems(self): """Iterate over pairs of (key, record).""" raise NotImplementedError() def items(self): """Return a list of (key, record).""" raise NotImplementedError() def __getitem__(self, key): """Get the record for this corresponding key.""" raise NotImplementedError() def __setitem__(self, key, value): """Set the record for this corresponding key. 'value' should be a instance of the ``Record`` class. """ raise NotImplementedError() def __delitem__(self, key): """Remove the corresponding record from the database.""" raise NotImplementedError() def report(self, keys): """Report the corresponding key as spam, incrementing the report count. Engines that implement don't implement this method should have handles_one_step set to False. """ raise NotImplementedError() def whitelist(self, keys): """Report the corresponding key as ham, incrementing the whitelist count. Engines that implement don't implement this method should have handles_one_step set to False. """ raise NotImplementedError() @classmethod def get_prefork_connections(cls, fn, mode, max_age=None): """Yields an unlimited number of partial functions that return a new engine instance, suitable for using toghether with the Pre-Fork server. """ raise NotImplementedError()
3,190
28.009091
79
py
pyzor
pyzor-master/pyzor/engines/__init__.py
"""Database backends for pyzord. The database class must expose a dictionary-like interface, allowing access via __getitem__, __setitem__, and __delitem__. The key will be a forty character string, and the value should be an instance of the Record class. If the database backend cannot store the Record objects natively, then it must transparently take care of translating to/from Record objects in __setitem__ and __getitem__. The database class should take care of expiring old values at the appropriate interval. """ from pyzor.engines import gdbm_ from pyzor.engines import mysql from pyzor.engines import redis_ from pyzor.engines import redis_v0 __all__ = ["database_classes"] database_classes = {"gdbm": gdbm_.handle, "mysql": mysql.handle, "redis_v0": redis_v0.handle, "redis": redis_.handle, }
891
30.857143
75
py
pyzor
pyzor-master/pyzor/hacks/py3.py
"""Hacks for python2-3 compatibility.""" import sys def reload(module): """Reload the modeule. This is handled differently according to the python version. This even varies across Python3 versions """ if sys.version_info[0] == 2: # Built-in method return reload(module) elif sys.version_info[0] == 3 and sys.version_info[1] <= 3: import imp return imp.reload(module) else: import importlib return importlib.reload(module)
506
22.045455
63
py
pyzor
pyzor-master/pyzor/hacks/__init__.py
"""Various hack to make pyzor compatible with different Python versions."""
76
37.5
75
py
pyzor
pyzor-master/pyzor/hacks/py26.py
"""Hacks for Python 2.6""" __all__ = ["hack_all", "hack_email", "hack_select"] def hack_all(email=True, select=True): """Apply all Python 2.6 patches.""" if email: hack_email() if select: hack_select() def hack_email(): """The python2.6 version of email.message_from_string, doesn't work with unicode strings. And in python3 it will only work with a decoded. So switch to using only message_from_bytes. """ import email if not hasattr(email, "message_from_bytes"): email.message_from_bytes = email.message_from_string def hack_select(): """The python2.6 version of SocketServer does not handle interrupt calls from signals. Patch the select call if necessary. """ import sys if sys.version_info[0] == 2 and sys.version_info[1] == 6: import select import errno real_select = select.select def _eintr_retry(*args): """restart a system call interrupted by EINTR""" while True: try: return real_select(*args) except (OSError, select.error) as ex: if ex.args[0] != errno.EINTR: raise select.select = _eintr_retry
1,258
26.977778
76
py
pyzor
pyzor-master/scripts/summarise.py
#! /usr/bin/env python # -*- coding: utf-8 -*- """Summarise Pyzor database. Generate a summary of the current state of a Pyzor database. This currently only works with a MySQL (or compatible) database. This can currently only output to a Slack channel. There are extra requirements for this script: * click * requests """ import os import datetime import ConfigParser import MySQLdb import requests import click @click.command() @click.option("--config", default=None) @click.argument("hook") def summarise(config, hook): """Generate a summary of a Pyzor database.""" if config is None: config = os.path.expanduser("~/.pyzor/config") conf = ConfigParser.ConfigParser() conf.read(config) (host, user, password, db_name, table) = conf.get("server", "DigestDB").split(",") db = MySQLdb.connect( host=host, user=user, db=db_name, passwd=password, ) c = db.cursor() # TODO: With a newer Python, this could use f-strings. data = {} c.execute( "SELECT COUNT(*) FROM `%s`" % table ) data["total"] = c.fetchone()[0] c.execute( "SELECT MIN(wl_entered), MIN(wl_updated), " "MIN(r_entered), MIN(r_updated), MAX(wl_entered), MAX(wl_updated), " "MAX(r_entered), MAX(r_updated) from `%s`" % table ) (data["oldest_white"], data["oldest_white_update"], data["oldest_spam"], data["oldest_spam_update"], data["newest_white"], data["newest_white_update"], data["newest_spam"], data["newest_spam_update"] ) = c.fetchone() c.execute( "SELECT MAX(r_count), MAX(wl_count) FROM `%s`" % table ) data["max_spam"], data["max_white"] = c.fetchone() # Frequency table for counts. for column in ("r_count", "wl_count"): buckets = [] for bucket in range(10): low = bucket * 100 high = (bucket + 1) * 100 c.execute( "SELECT COUNT(*) FROM `%s` WHERE %s BETWEEN %%s AND %%s" % (table, column), (low, high) ) buckets.append(c.fetchone()[0]) data[column] = buckets # Frequency table for age. for column in ("r_updated", "wl_updated"): buckets = [] for bucket in range(10): now = datetime.datetime.now() low = now - datetime.timedelta(days=(bucket + 1) * 7) high = now - datetime.timedelta(days=bucket * 7) c.execute( "SELECT COUNT(*) FROM `%s` WHERE %s BETWEEN %%s AND %%s" % (table, column), (low, high) ) buckets.append(c.fetchone()[0]) data[column] = buckets data["table"] = table notify_slack(hook, data) c.close() db.close() # Borrowed from https://raw.githubusercontent.com/kennethreitz/spark.py/master/spark.py def spark_string(ints, fit_min=False): """Returns a spark string from given iterable of ints. Keyword Arguments: fit_min: Matches the range of the sparkline to the input integers rather than the default of zero. Useful for large numbers with relatively small differences between the positions """ ticks = u" ▁▂▃▄▅▆▇█" min_range = min(ints) if fit_min else 0 step_range = max(ints) - min_range step = (step_range / float(len(ticks) - 1)) or 1 return u''.join(ticks[int(round((i - min_range) / step))] for i in ints) def notify_slack(hook, data): """Send a notification containing a summary of a Pyzor database to a Slack channel.""" text = "Pyzor summary for _%(table)s_ (%(total)s digests)" % data format = "%d %b %Y" if data["max_spam"] < 100: spam_colour = "danger" else: spam_colour = "good" if data["max_white"] < 100: white_colour = "danger" else: white_colour = "good" if (datetime.datetime.now() - data["newest_spam_update"]).days > 2: spam_age_colour = "danger" else: spam_age_colour = "good" if (datetime.datetime.now() - data["newest_white_update"]).days > 2: white_age_colour = "danger" else: white_age_colour = "good" attachments = [ { "title": "Spam Reports", "text": spark_string(data["r_count"], fit_min=True), "fields": [ { "title": "Most common count", "value": data["max_spam"], "short": True, }, ], "color": spam_colour, }, { "title": "Whitelist Reports", "text": spark_string(data["wl_count"], fit_min=True), "fields": [ { "title": "Most common count", "value": data["max_white"], "short": True, }, ], "color": white_colour, }, { "title": "Spam Age", "text": spark_string(data["r_updated"], fit_min=True), "fields": [ { "title": "Oldest", "value": data["oldest_spam"].strftime(format), "short": True, }, { "title": "Oldest Update", "value": data["oldest_spam_update"].strftime(format), "short": True, }, { "title": "Latest", "value": data["newest_spam"].strftime(format), "short": True, }, { "title": "Latest Update", "value": data["newest_spam_update"].strftime(format), "short": True, }, ], "color": spam_age_colour, }, { "title": "Whitelist Age", "text": spark_string(data["wl_updated"], fit_min=True), "fields": [ { "title": "Oldest", "value": data["oldest_white"].strftime(format), "short": True, }, { "title": "Oldest Update", "value": data["oldest_white_update"].strftime(format), "short": True, }, { "title": "Latest", "value": data["newest_white"].strftime(format), "short": True, }, { "title": "Latest Update", "value": data["newest_white_update"].strftime(format), "short": True, }, ], "color": white_age_colour, }, ] response = requests.post( hook, json={"text": text, "attachments": attachments} ) if __name__ == "__main__": summarise()
6,995
29.819383
87
py
pyzor
pyzor-master/tests/__init__.py
"""Package reserved for tests and test utilities.""" import unittest def suite(): """Gather all the tests from this package in a test suite.""" import tests.unit as unit import tests.functional as functional test_suite = unittest.TestSuite() test_suite.addTest(unit.suite()) test_suite.addTest(functional.suite()) return test_suite if __name__ == '__main__': unittest.main(defaultTest='suite')
433
20.7
65
py
pyzor
pyzor-master/tests/unit/test_forwarder.py
"""Test the pyzor.forwarder module """ import time import unittest import threading try: from unittest.mock import call, Mock except ImportError: from mock import call, Mock import pyzor.forwarder class ForwarderTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) def tearDown(self): unittest.TestCase.tearDown(self) def test_queue(self): client = Mock() servlist = [] max_qsize = 10 forwarder = pyzor.forwarder.Forwarder(client, servlist, max_queue_size=max_qsize) for _ in range(max_qsize * 2): forwarder.queue_forward_request('975422c090e7a43ab7c9bf0065d5b661259e6d74') self.assertGreater(forwarder.forward_queue.qsize(), 0, 'queue insert failed') self.assertLessEqual(forwarder.forward_queue.qsize(), max_qsize, 'queue overload') self.assertEqual(forwarder.forward_queue.qsize(), max_qsize, 'queue should be full at this point') t = threading.Thread(target=forwarder._forward_loop) t.start() time.sleep(1) self.assertEqual(forwarder.forward_queue.qsize(), 0, 'queue should be empty') forwarder.stop_forwarding() t.join(5) self.assertFalse(t.is_alive(), 'forward thread did not end') def test_remote_servers(self): client = Mock() digest = '975422c090e7a43ab7c9bf0065d5b661259e6d74' servlist = [("test1.example.com", 24441), ("test2.example.com", 24442)] forwarder = pyzor.forwarder.Forwarder(client, servlist) forwarder.queue_forward_request(digest) forwarder.queue_forward_request(digest, whitelist=True) forwarder.start_forwarding() time.sleep(2) forwarder.stop_forwarding() client.report.assert_has_calls([call(digest, servlist[0]), call(digest, servlist[1])]) client.whitelist.assert_has_calls([call(digest, servlist[0]), call(digest, servlist[1])]) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(ForwarderTest)) return test_suite if __name__ == '__main__': unittest.main()
2,352
32.614286
106
py
pyzor
pyzor-master/tests/unit/test_account.py
"""Test the pyzor.account module""" import io import os import time import email import hashlib import unittest import pyzor import pyzor.config import pyzor.account class AccountTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.timestamp = 1381219396 self.msg = email.message_from_string("") self.msg["Op"] = "ping" self.msg["Thread"] = "14941" self.msg["PV"] = "2.1" self.msg["User"] = "anonymous" self.msg["Time"] = str(self.timestamp) def tearDown(self): unittest.TestCase.tearDown(self) def test_sign_msg(self): """Test the sign message function""" hashed_key = hashlib.sha1(b"test_key").hexdigest() expected = "2ab1bad2aae6fd80c656a896c82eef0ec1ec38a0" result = pyzor.account.sign_msg(hashed_key, self.timestamp, self.msg) self.assertEqual(result, expected) def test_hash_key(self): """Test the hash key function""" user = "testuser" key = "testkey" expected = "0957bd79b58263657127a39762879098286d8477" result = pyzor.account.hash_key(key, user) self.assertEqual(result, expected) def test_verify_signature(self): """Test the verify signature function""" def mock_sm(h, t, m): return "testsig" real_sm = pyzor.account.sign_msg pyzor.account.sign_msg = mock_sm try: self.msg["Sig"] = "testsig" del self.msg["Time"] self.msg["Time"] = str(int(time.time())) pyzor.account.verify_signature(self.msg, "testkey") finally: pyzor.account.sign_msg = real_sm def test_verify_signature_old_timestamp(self): """Test the verify signature with old timestamp""" def mock_sm(h, t, m): return "testsig" real_sm = pyzor.account.sign_msg pyzor.account.sign_msg = mock_sm try: self.msg["Sig"] = "testsig" self.assertRaises(pyzor.SignatureError, pyzor.account.verify_signature, self.msg, "testkey") finally: pyzor.account.sign_msg = real_sm def test_verify_signature_bad_signature(self): """Test the verify signature with invalid signature""" def mock_sm(h, t, m): return "testsig" real_sm = pyzor.account.sign_msg pyzor.account.sign_msg = mock_sm try: self.msg["Sig"] = "testsig-bad" del self.msg["Time"] self.msg["Time"] = str(int(time.time())) self.assertRaises(pyzor.SignatureError, pyzor.account.verify_signature, self.msg, "testkey") finally: pyzor.account.sign_msg = real_sm class LoadAccountTest(unittest.TestCase): """Tests for the load_accounts function""" filepath = "test_file" def setUp(self): unittest.TestCase.setUp(self) self.real_exists = os.path.exists os.path.exists = lambda p: True if p == self.filepath else \ self.real_exists(p) self.mock_file = io.StringIO() try: self.real_open = pyzor.account.__builtins__.open except AttributeError: self.real_open = pyzor.account.__builtins__["open"] def mock_open(path, mode="r", buffering=-1): if path == self.filepath: self.mock_file.seek(0) return self.mock_file else: return self.real_open(path, mode, buffering) try: pyzor.account.__builtins__.open = mock_open except AttributeError: pyzor.account.__builtins__["open"] = mock_open def tearDown(self): unittest.TestCase.tearDown(self) os.path.exists = self.real_exists try: pyzor.account.__builtins__.open = self.real_open except AttributeError: pyzor.account.__builtins__["open"] = self.real_open def test_load_accounts_nothing(self): result = pyzor.config.load_accounts("foobar") self.assertEqual(result, {}) def test_load_accounts(self): """Test loading the account file""" self.mock_file.write(u"public.pyzor.org : 24441 : test : 123abc,cba321\n" u"public2.pyzor.org : 24441 : test2 : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertIn(("public.pyzor.org", 24441), result) self.assertIn(("public2.pyzor.org", 24441), result) account = result[("public.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test", "123abc", "cba321")) account = result[("public2.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test2", "123abc", "cba321")) def test_load_accounts_invalid_line(self): """Test loading the account file""" self.mock_file.write(u"public.pyzor.org : 24441 ; test : 123abc,cba321\n" u"public2.pyzor.org : 24441 : test2 : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertNotIn(("public.pyzor.org", 24441), result) self.assertEqual(len(result), 1) self.assertIn(("public2.pyzor.org", 24441), result) account = result[("public2.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test2", "123abc", "cba321")) def test_load_accounts_invalid_port(self): """Test loading the account file""" self.mock_file.write(u"public.pyzor.org : a4441 : test : 123abc,cba321\n" u"public2.pyzor.org : 24441 : test2 : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertNotIn(("public.pyzor.org", 24441), result) self.assertEqual(len(result), 1) self.assertIn(("public2.pyzor.org", 24441), result) account = result[("public2.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test2", "123abc", "cba321")) def test_load_accounts_invalid_key(self): """Test loading the account file""" self.mock_file.write(u"public.pyzor.org : 24441 : test : ,\n" u"public2.pyzor.org : 24441 : test2 : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertNotIn(("public.pyzor.org", 24441), result) self.assertEqual(len(result), 1) self.assertIn(("public2.pyzor.org", 24441), result) account = result[("public2.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test2", "123abc", "cba321")) def test_load_accounts_invalid_missing_comma(self): """Test loading the account file""" self.mock_file.write(u"public.pyzor.org : 24441 : test : 123abccba321\n" u"public2.pyzor.org : 24441 : test2 : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertNotIn(("public.pyzor.org", 24441), result) self.assertEqual(len(result), 1) self.assertIn(("public2.pyzor.org", 24441), result) account = result[("public2.pyzor.org", 24441)] self.assertEqual((account.username, account.salt, account.key), ("test2", "123abc", "cba321")) def test_load_accounts_comment(self): """Test skipping commented lines""" self.mock_file.write(u"#public1.pyzor.org : 24441 : test : 123abc,cba321") result = pyzor.config.load_accounts(self.filepath) self.assertNotIn(("public.pyzor.org", 24441), result) self.assertFalse(result) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(AccountTest)) test_suite.addTest(unittest.makeSuite(LoadAccountTest)) return test_suite if __name__ == '__main__': unittest.main()
8,254
39.665025
104
py
pyzor
pyzor-master/tests/unit/test_config.py
import os import logging import unittest try: import configparser as ConfigParser except ImportError: import ConfigParser try: from unittest.mock import patch, Mock except ImportError: from mock import patch, Mock import pyzor.config from tests.util import mock_open class MockData(list): def close(self): pass class TestPasswdLoad(unittest.TestCase): fp = "pyzord.passwd" alice_key = "alice_key" bob_key = "bob_key" def setUp(self): super(TestPasswdLoad, self).setUp() self.data = MockData() self.exists = True real_exists = os.path.exists patch("pyzor.config.open", return_value=self.data, create=True).start() _exists = lambda fp: True if fp == self.fp else real_exists(fp) patch("pyzor.config.os.path.exists", side_effect=_exists).start() def get_passwd(self, fp=None): if not fp: fp = self.fp return pyzor.config.load_passwd_file(fp) def tearDown(self): super(TestPasswdLoad, self).tearDown() patch.stopall() def test_nothing(self): result = self.get_passwd() self.assertEqual(result, {}) def test_default(self): result = self.get_passwd("foobar") self.assertEqual(result, {}) def test_passwd(self): self.data.append("alice : %s\n" % self.alice_key) self.data.append("bob : %s\n" % self.bob_key) result = self.get_passwd() self.assertEqual(result, {"alice": self.alice_key, "bob": self.bob_key}) def test_invalid_line(self): self.data.append("alice ; %s\n" % self.alice_key) self.data.append("bob : %s\n" % self.bob_key) result = self.get_passwd() self.assertEqual(result, {"bob": self.bob_key}) def test_ignore_comment(self): self.data.append("alice : %s\n" % self.alice_key) self.data.append("# bob : %s\n" % self.bob_key) result = self.get_passwd() self.assertEqual(result, {"alice": self.alice_key}) class TestAccessLoad(unittest.TestCase): fp = "pyzord.access" accounts = ["alice", "bob"] all = {'report', 'info', 'pong', 'ping', 'check', 'whitelist'} anonymous_privileges = {'report', 'info', 'pong', 'ping', 'check'} def setUp(self): super(TestAccessLoad, self).setUp() self.data = MockData() self.exists = True real_exists = os.path.exists patch("pyzor.config.open", return_value=self.data, create=True).start() _exists = lambda fp: True if fp == self.fp else real_exists(fp) patch("pyzor.config.os.path.exists", side_effect=_exists).start() def get_access(self, fp=None, accounts=None): if not fp: fp = self.fp if not accounts: accounts = self.accounts return pyzor.config.load_access_file(fp, accounts) def tearDown(self): super(TestAccessLoad, self).tearDown() patch.stopall() def test_nothing(self): result = self.get_access() self.assertEqual(result, {}) def test_default(self): result = self.get_access(fp="foobar") self.assertEqual(result, {'anonymous': self.anonymous_privileges}) def test_invalid_line(self): self.data.append("all : allice ; allow\n") self.data.append("ping : bob : allow\n") result = self.get_access() self.assertEqual(result, {'bob': {'ping'}}) def test_invalid_action(self): self.data.append("all : allice : don't allow\n") self.data.append("ping : bob : allow\n") result = self.get_access() self.assertEqual(result, {'bob': {'ping'}}) def test_all_privilege(self): self.data.append("all : bob : allow\n") result = self.get_access() self.assertEqual(result, {'bob': self.all}) def test_all_accounts(self): self.data.append("all : all : allow\n") result = self.get_access() self.assertEqual(result, {'alice': self.all, 'bob': self.all}) def test_deny_action(self): self.data.append("all : all : allow\n") self.data.append("ping : bob : deny\n") result = self.get_access() self.assertEqual(result, {'alice': self.all, 'bob': self.all - {'ping'}}) def test_multiple_users(self): self.data.append("all : alice bob: allow\n") result = self.get_access() self.assertEqual(result, {'alice': self.all, 'bob': self.all}) def test_multiple_privileges(self): self.data.append("ping pong : alice: allow\n") result = self.get_access() self.assertEqual(result, {'alice': {'ping', 'pong'}}) def test_ignore_comments(self): self.data.append("all: alice: allow\n") self.data.append("# all: bob : allow\n") result = self.get_access() self.assertEqual(result, {'alice': self.all}) class TestServersLoad(unittest.TestCase): fp = "servers" public_server = ("public.pyzor.org", 24441) random_server1 = ("random.pyzor.org", 33544) random_server2 = ("127.1.2.45", 13587) def setUp(self): super(TestServersLoad, self).setUp() self.data = [] self.exists = True real_exists = os.path.exists _exists = lambda fp: True if fp == self.fp else real_exists(fp) patch("pyzor.config.os.path.exists", side_effect=_exists).start() def get_servers(self, fp=None): if not fp: fp = self.fp name = "pyzor.config.open" with patch(name, mock_open(read_data=''.join(self.data)), create=True) as m: return pyzor.config.load_servers(fp) def tearDown(self): super(TestServersLoad, self).tearDown() patch.stopall() def test_nothing(self): result = self.get_servers() self.assertEqual(result, [self.public_server]) def test_default(self): result = self.get_servers("foobar") self.assertEqual(result, [self.public_server]) def test_servers(self): self.data.append("%s:%s\n" % self.random_server1) self.data.append("%s:%s\n" % self.random_server2) result = self.get_servers() self.assertEqual(result, [self.random_server1, self.random_server2]) def test_ignore_comment(self): self.data.append("#%s:%s\n" % self.random_server1) self.data.append("%s:%s\n" % self.random_server2) result = self.get_servers() self.assertEqual(result, [self.random_server2]) class TestLogSetup(unittest.TestCase): log_file = "this_is_a_test_log_file" def setUp(self): super(TestLogSetup, self).setUp() def tearDown(self): super(TestLogSetup, self).tearDown() try: os.remove(self.log_file) except OSError: pass def test_logging(self): pyzor.config.setup_logging("pyzor.test1", None, False) log = logging.getLogger("pyzor.test1") self.assertEqual(log.getEffectiveLevel(), logging.INFO) self.assertEqual(log.handlers[0].level, logging.CRITICAL) def test_logging_debug(self): pyzor.config.setup_logging("pyzor.test2", None, True) log = logging.getLogger("pyzor.test2") self.assertEqual(log.getEffectiveLevel(), logging.DEBUG) self.assertEqual(log.handlers[0].level, logging.DEBUG) def test_logging_file(self): pyzor.config.setup_logging("pyzor.test3", self.log_file, False) log = logging.getLogger("pyzor.test3") self.assertEqual(log.getEffectiveLevel(), logging.INFO) self.assertEqual(log.handlers[0].level, logging.CRITICAL) self.assertEqual(log.handlers[1].level, logging.INFO) def test_logging_file_debug(self): pyzor.config.setup_logging("pyzor.test4", self.log_file, True) log = logging.getLogger("pyzor.test4") self.assertEqual(log.getEffectiveLevel(), logging.DEBUG) self.assertEqual(log.handlers[0].level, logging.DEBUG) self.assertEqual(log.handlers[1].level, logging.DEBUG) class TestExpandHomeFiles(unittest.TestCase): home = "/home/user/pyzor" def setUp(self): super(TestExpandHomeFiles, self).setUp() def tearDown(self): super(TestExpandHomeFiles, self).tearDown() def check_expand(self, homefiles, homedir, config, expected): section = "test" conf = ConfigParser.ConfigParser() conf.add_section(section) for key, value in config.items(): conf.set(section, key, value) pyzor.config.expand_homefiles(homefiles, section, homedir, conf) result = dict(conf.items(section)) self.assertEqual(result, expected) def test_homedir(self): self.check_expand( ["testfile"], self.home, {"testfile": "my.file"}, {"testfile": "%s/my.file" % self.home}, ) def test_homedir_none(self): self.check_expand( ["testfile"], self.home, {"testfile": ""}, {"testfile": ""}, ) def test_homedir_abs(self): self.check_expand( ["testfile"], self.home, {"testfile": "/home/user2/pyzor"}, {"testfile": "/home/user2/pyzor"}, ) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(TestLogSetup)) test_suite.addTest(unittest.makeSuite(TestAccessLoad)) test_suite.addTest(unittest.makeSuite(TestPasswdLoad)) test_suite.addTest(unittest.makeSuite(TestServersLoad)) return test_suite if __name__ == '__main__': unittest.main()
9,903
30.948387
74
py
pyzor
pyzor-master/tests/unit/test_server.py
"""Test the pyzor.server module """ import io import sys import time import logging import unittest try: import socketserver as SocketServer except ImportError: import SocketServer from datetime import datetime, timedelta try: from unittest.mock import patch except ImportError: from mock import patch import pyzor.server import pyzor.engines.common class MockServer(): """Mocks the pyzor.server.Server class""" def __init__(self): self.log = logging.getLogger("pyzord") self.usage_log = logging.getLogger("pyzord-usage") self.log.addHandler(logging.NullHandler()) self.usage_log.addHandler(logging.NullHandler()) self.forwarder = None self.one_step = False class MockDatagramRequestHandler(): """ Mock the SocketServer.DatagramRequestHand.""" def __init__(self, headers, database=None, acl=None, accounts=None): """Initiates an request handler and set's the data in `headers` as the request. Also set's the database, acl and accounts for the MockServer. This will be set as base class for RequestHandler. """ self.rfile = io.BytesIO() self.wfile = io.BytesIO() for i, j in headers.items(): self.rfile.write(("%s: %s\n" % (i, j)).encode("utf8")) self.rfile.seek(0) self.packet = None self.client_address = ["127.0.0.1"] # Setup MockServer data self.server = MockServer() self.server.database = database if acl: self.server.acl = acl else: self.server.acl = {pyzor.anonymous_user: ("check", "report", "ping", "pong", "info", "whitelist",)} self.server.accounts = accounts self.handle() def handle(self): pass class RequestHandlerTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.real_drh = SocketServer.DatagramRequestHandler SocketServer.DatagramRequestHandler = MockDatagramRequestHandler pyzor.server.RequestHandler.__bases__ = (MockDatagramRequestHandler,) # setup the basic values for request and response self.request = {"User": pyzor.anonymous_user, "Time": str(int(time.time())), "PV": str(pyzor.proto_version), "Thread": "3597"} self.expected_response = {"Code": "200", "Diag": "OK", "PV": str(pyzor.proto_version), "Thread": "3597"} def tearDown(self): unittest.TestCase.tearDown(self) SocketServer.DatagramRequestHandler = self.real_drh pyzor.server.RequestHandler.__bases__ = (self.real_drh,) patch.stopall() def check_response(self, handler): """Checks if the response from the handler is equal to the expected response. """ handler.wfile.seek(0) response = handler.wfile.read() response = response.decode("utf8").replace("\n\n", "\n") result = {} try: for line in response.splitlines(): key = line.split(":", 1)[0].strip() value = line.split(":")[1].strip() result[key] = value except (IndexError, TypeError) as e: self.fail("Error parsing %r: %s" % (response, e)) self.assertEqual(result, self.expected_response) def timestamp(self, time_obj): if not time_obj: return 0 else: return str(int(time.mktime(time_obj.timetuple()))) def test_ping(self): """Tests the ping command handler""" self.request["Op"] = "ping" handler = pyzor.server.RequestHandler(self.request) self.check_response(handler) def test_pong(self): """Tests the pong command handler""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {digest: pyzor.engines.common.Record(24, 42)} self.request["Op"] = "pong" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.expected_response["Count"] = str(sys.maxsize) self.expected_response["WL-Count"] = "0" self.check_response(handler) def test_check(self): """Tests the check command handler""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {digest: pyzor.engines.common.Record(24, 42)} self.request["Op"] = "check" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.expected_response["Count"] = "24" self.expected_response["WL-Count"] = "42" self.check_response(handler) def test_check_new(self): """Tests the check command handler with a new record""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {} self.request["Op"] = "check" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.expected_response["Count"] = "0" self.expected_response["WL-Count"] = "0" self.check_response(handler) def test_info(self): """Tests the info command handler""" entered = datetime.now() - timedelta(days=10) updated = datetime.now() wl_entered = datetime.now() - timedelta(days=20) wl_updated = datetime.now() - timedelta(days=2) digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {digest: pyzor.engines.common.Record(24, 42, entered, updated, wl_entered, wl_updated)} self.request["Op"] = "info" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.expected_response["Count"] = "24" self.expected_response["WL-Count"] = "42" self.expected_response["Entered"] = self.timestamp(entered) self.expected_response["Updated"] = self.timestamp(updated) self.expected_response["WL-Entered"] = self.timestamp(wl_entered) self.expected_response["WL-Updated"] = self.timestamp(wl_updated) self.check_response(handler) def test_info_new(self): """Tests the info command handler with a new record""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {} self.request["Op"] = "info" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.expected_response["Count"] = "0" self.expected_response["WL-Count"] = "0" self.expected_response["Entered"] = "0" self.expected_response["Updated"] = "0" self.expected_response["WL-Entered"] = "0" self.expected_response["WL-Updated"] = "0" self.check_response(handler) def test_report(self): """Tests the report command handler""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {digest: pyzor.engines.common.Record(24, 42)} self.request["Op"] = "report" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.check_response(handler) self.assertEqual(database[digest].r_count, 25) def test_report_new(self): """Tests the report command handler with a new record""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {} self.request["Op"] = "report" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.check_response(handler) self.assertEqual(database[digest].r_count, 1) def test_whitelist(self): """Tests the whitelist command handler""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {digest: pyzor.engines.common.Record(24, 42)} self.request["Op"] = "whitelist" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.check_response(handler) self.assertEqual(database[digest].wl_count, 43) def test_whitelist_new(self): """Tests the whitelist command handler with a new record""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" database = {} self.request["Op"] = "whitelist" self.request["Op-Digest"] = digest handler = pyzor.server.RequestHandler(self.request, database) self.check_response(handler) self.assertEqual(database[digest].wl_count, 1) def test_handle_no_version(self): """Tests handling an request with no version specified""" self.request["Op"] = "ping" del self.request["PV"] handler = pyzor.server.RequestHandler(self.request) self.expected_response["Code"] = "400" self.expected_response["Diag"] = "Bad request" self.check_response(handler) def test_handle_unsupported_version(self): """Tests handling an request with an unsupported version specified""" self.request["Op"] = "ping" self.request["PV"] = str(pyzor.proto_version + 2) handler = pyzor.server.RequestHandler(self.request) self.expected_response["Code"] = "505" self.expected_response["Diag"] = "Version Not Supported" self.check_response(handler) def test_handle_not_implemented(self): """Tests handling an request with an unimplemented command""" self.request["Op"] = "notimplemented" acl = {pyzor.anonymous_user: "notimplemented"} handler = pyzor.server.RequestHandler(self.request, acl=acl) self.expected_response["Code"] = "501" self.expected_response["Diag"] = "Not implemented" self.check_response(handler) def test_handle_unauthorized(self): """Tests handling an request with an unauthorized command""" self.request["Op"] = "report" acl = {pyzor.anonymous_user: ("ping", "check")} handler = pyzor.server.RequestHandler(self.request, acl=acl) self.expected_response["Code"] = "403" self.expected_response["Diag"] = "Forbidden" self.check_response(handler) def test_handle_account(self): """Tests handling an request where user is not anonymous""" self.request["Op"] = "ping" self.request["User"] = "testuser" acl = {"testuser": ("ping", "check")} accounts = {"testuser": "testkey"} mock_vs = lambda x, y: None real_vs = pyzor.account.verify_signature pyzor.account.verify_signature = mock_vs try: handler = pyzor.server.RequestHandler(self.request, acl=acl, accounts=accounts) self.check_response(handler) finally: pyzor.account.verify_signature = real_vs def test_handle_unknown_account(self): """Tests handling an request where user is unkwown""" self.request["Op"] = "ping" self.request["User"] = "testuser" acl = {"testuser": ("ping", "check")} accounts = {} self.expected_response["Code"] = "401" self.expected_response["Diag"] = "Unauthorized" def mock_vs(x, y): pass real_vs = pyzor.account.verify_signature pyzor.account.verify_signature = mock_vs try: handler = pyzor.server.RequestHandler(self.request, acl=acl, accounts=accounts) self.check_response(handler) finally: pyzor.account.verify_signature = real_vs def test_handle_invalid_signature(self): """Tests handling an request where user key is invalid""" self.request["Op"] = "ping" self.request["User"] = "testuser" acl = {"testuser": ("ping", "check")} accounts = {"testuser": ("ping", "check")} self.expected_response["Code"] = "401" self.expected_response["Diag"] = "Unauthorized" def mock_vs(x, y): raise pyzor.SignatureError("Invalid signature.") real_vs = pyzor.account.verify_signature pyzor.account.verify_signature = mock_vs try: handler = pyzor.server.RequestHandler(self.request, acl=acl, accounts=accounts) self.check_response(handler) finally: pyzor.account.verify_signature = real_vs def test_invalid_pv(self): self.request["Op"] = "ping" self.request["PV"] = "ab2.13" handler = pyzor.server.RequestHandler(self.request) self.expected_response["Code"] = "400" self.expected_response["Diag"] = "Bad request" self.check_response(handler) def test_uncaught_exception(self): patch("pyzor.server.RequestHandler._really_handle", side_effect=Exception("test")).start() self.request["Op"] = "ping" handler = pyzor.server.RequestHandler(self.request) self.expected_response["Code"] = "500" self.expected_response["Diag"] = "Internal Server Error" del self.expected_response["Thread"] self.check_response(handler) class ServerTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.mock_config = patch("pyzor.config").start() def tearDown(self): unittest.TestCase.tearDown(self) patch.stopall() def test_server(self): pyzor.server.Server(("127.0.0.1", 24441), {}, "passwd_fn", "access_fn", None) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(RequestHandlerTest)) test_suite.addTest(unittest.makeSuite(ServerTest)) return test_suite if __name__ == '__main__': unittest.main()
14,225
34.654135
81
py
pyzor
pyzor-master/tests/unit/test_client.py
import time import email import unittest try: from unittest.mock import Mock, patch, call except ImportError: from mock import Mock, patch, call import pyzor.client import pyzor.account class TestBase(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.thread = 33715 self.timeout = None self.time = str(int(time.time())) patch("pyzor.account.sign_msg", return_value="TestSig").start() patch("pyzor.account.hash_key").start() # the response the mock socket will send self.response = {"Code": "200", "Diag": "OK", "PV": "2.1", "Thread": "33715", "Time": self.time } self.mresponse = None self.mock_socket = None # the expected request that the client should send self.expected = {"Thread": str(self.thread), "PV": str(pyzor.proto_version), "User": "anonymous", "Time": self.time, "Sig": "TestSig" } def tearDown(self): unittest.TestCase.tearDown(self) patch.stopall() def get_requests(self): for mock_call in self.mock_socket.mock_calls: name, args, kwargs = mock_call if name == "socket().sendto": yield args, kwargs def check_request(self): """Check if the request sent by the client is equal to the expected one. """ req = {} for args, _ in self.get_requests(): self.assertEqual(args[2], ('127.0.0.1', 24441)) req = dict(email.message_from_string(args[0].decode())) self.assertEqual(req, self.expected) def patch_all(self, conf=None): if conf is None: conf = {} patch("pyzor.message.ThreadId.generate", return_value=self.thread).start() if self.response: response = "\n".join("%s: %s" % (key, value) for key, value in self.response.items()) + "\n\n" self.mresponse = response.encode(), ("127.0.0.1", 24441) else: self.mresponse = None addrinfo = [(2, 2, 17, '', ('127.0.0.1', 24441))] config = {"socket.return_value": Mock(), "socket.return_value.recvfrom.return_value": self.mresponse, "getaddrinfo.return_value": addrinfo} config.update(conf) self.mock_socket = patch("pyzor.client.socket", **config).start() def check_client(self, accounts, method, *args, **kwargs): """Tests if the request and response are sent and read correctly by the client. """ client = pyzor.client.Client(accounts=accounts, timeout=self.timeout) got_response = getattr(client, method)(*args, **kwargs) self.assertEqual(str(got_response), self.mresponse[0].decode()) if self.expected is not None: self.check_request() return client class ClientTest(TestBase): def test_ping(self): """Test the client ping request""" self.expected["Op"] = "ping" self.patch_all() self.check_client(None, "ping") def test_pong(self): """Test the client pong request""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op"] = "pong" self.expected["Op-Digest"] = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() self.check_client(None, "pong", digest) def test_check(self): """Test the client check request""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op"] = "check" self.expected["Op-Digest"] = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() self.check_client(None, "check", digest) def test_info(self): """Test the client info request""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op"] = "info" self.expected["Op-Digest"] = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() self.check_client(None, "info", digest) def test_report(self): """Test the client report request""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op"] = "report" self.expected["Op-Digest"] = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op-Spec"] = "20,3,60,3" self.patch_all() self.check_client(None, "report", digest) def test_whitelist(self): """Test the client whitelist request""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op"] = "whitelist" self.expected["Op-Digest"] = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.expected["Op-Spec"] = "20,3,60,3" self.patch_all() self.check_client(None, "whitelist", digest) def test_handle_account(self): """Test client handling accounts""" test_account = pyzor.account.Account("TestUser", "TestKey", "TestSalt") self.expected["Op"] = "ping" self.expected["User"] = "TestUser" self.patch_all() self.check_client({("public.pyzor.org", 24441): test_account}, "ping") def test_handle_invalid_thread(self): """Test invalid thread id""" self.thread += 20 self.expected["Op"] = "ping" self.patch_all() self.assertRaises(pyzor.ProtocolError, self.check_client, None, "ping") def test_set_timeout(self): self.expected = None self.patch_all() self.timeout = 10 self.check_client(None, "ping") calls = [call.socket().settimeout(10), ] self.mock_socket.assert_has_calls(calls) class BatchClientTest(TestBase): def test_report(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(10): client.report(digest) args, kwargs = list(self.get_requests())[0] msg = email.message_from_string(args[0].decode()) self.assertEqual(len(msg.get_all("Op-Digest")), 10) def test_report_to_few(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.report(digest) self.assertEqual(list(self.get_requests()), []) def test_whitelist(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(10): client.whitelist(digest) args, kwargs = list(self.get_requests())[0] msg = email.message_from_string(args[0].decode()) self.assertEqual(len(msg.get_all("Op-Digest")), 10) def test_whitelist_to_few(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.whitelist(digest) self.assertEqual(list(self.get_requests()), []) def test_force_report(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.report(digest) client.force() args, kwargs = list(self.get_requests())[0] msg = email.message_from_string(args[0].decode()) self.assertEqual(len(msg.get_all("Op-Digest")), 9) def test_force_whitelist(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.whitelist(digest) client.force() args, kwargs = list(self.get_requests())[0] msg = email.message_from_string(args[0].decode()) self.assertEqual(len(msg.get_all("Op-Digest")), 9) def test_flush_report(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.report(digest) client.flush() client.report(digest) self.assertEqual(list(self.get_requests()), []) def test_flush_whitelist(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.patch_all() client = pyzor.client.BatchClient() for i in range(9): client.whitelist(digest) client.flush() client.whitelist(digest) self.assertEqual(list(self.get_requests()), []) class ClientRunnerTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.server = "test.example.com", 24441 def tearDown(self): unittest.TestCase.tearDown(self) def check_runner(self, test_class, response, results, kwargs=None): if kwargs is None: kwargs = {} kwargs["return_value"] = response mock_routine = Mock(**kwargs) runner = test_class(mock_routine) runner.run(self.server, ()) self.assertEqual(runner.results, results) return runner def test_normal(self): response = pyzor.message.Response() response["Diag"] = "OK" response["Code"] = "200" server = "%s:%s\t" % self.server results = ["%s%s\n" % (server, response.head_tuple()), ] self.check_runner(pyzor.client.ClientRunner, response, results) def test_check(self): response = pyzor.message.Response() response["Diag"] = "OK" response["Code"] = "200" response["Count"] = "2" response["WL-Count"] = "1" server = "%s:%s\t" % self.server results = ["%s%s\t%s\t%s\n" % (server, response.head_tuple(), "2", "1")] self.check_runner(pyzor.client.CheckClientRunner, response, results) def test_info(self): response = """Code: 200 Diag: OK PV: 2.1 Thread: 8521 Entered: 1400221786 Updated: 1400221794 WL-Entered: 0 WL-Updated: 0 Count: 4 WL-Count: 0 """ response = email.message_from_string(response, _class=pyzor.message.Response) server = "%s:%s" % self.server result = ("%s\t(200, 'OK')\n" "\tCount: 4\n" "\tEntered: %s\n" "\tUpdated: %s\n" "\tWL-Count: 0\n" "\tWL-Entered: %s\n" "\tWL-Updated: %s\n\n" % (server, time.ctime(1400221786), time.ctime(1400221794), time.ctime(0), time.ctime(0))) self.maxDiff = None self.check_runner(pyzor.client.InfoClientRunner, response, [result]) def test_info_never(self): response = """Code: 200 Diag: OK PV: 2.1 Thread: 8521 Entered: 1400221786 Updated: 1400221794 WL-Entered: -1 WL-Updated: -1 Count: 4 WL-Count: 0 """ response = email.message_from_string(response, _class=pyzor.message.Response) server = "%s:%s" % self.server result = ("%s\t(200, 'OK')\n" "\tCount: 4\n" "\tEntered: %s\n" "\tUpdated: %s\n" "\tWL-Count: 0\n" "\tWL-Entered: Never\n" "\tWL-Updated: Never\n\n" % (server, time.ctime(1400221786), time.ctime(1400221794))) self.maxDiff = None self.check_runner(pyzor.client.InfoClientRunner, response, [result]) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(ClientTest)) test_suite.addTest(unittest.makeSuite(BatchClientTest)) test_suite.addTest(unittest.makeSuite(ClientRunnerTest)) return test_suite if __name__ == '__main__': unittest.main()
12,149
30.889764
82
py
pyzor
pyzor-master/tests/unit/__init__.py
"""A suite of unit tests that verifies the correct behaviour of various functions/methods in the pyzord code. Note these tests the source of pyzor, not the version currently installed. """ import unittest def suite(): """Gather all the tests from this package in a test suite.""" import test_client import test_config import test_digest import test_server import test_account import test_forwarder import test_engines test_suite = unittest.TestSuite() test_suite.addTest(test_engines.suite()) test_suite.addTest(test_client.suite()) test_suite.addTest(test_config.suite()) test_suite.addTest(test_digest.suite()) test_suite.addTest(test_server.suite()) test_suite.addTest(test_account.suite()) test_suite.addTest(test_forwarder.suite()) return test_suite if __name__ == '__main__': unittest.main(defaultTest='suite')
897
26.212121
74
py
pyzor
pyzor-master/tests/unit/test_digest.py
"""The the pyzor.digest module """ import unittest import pyzor.digest from pyzor.digest import * try: from unittest.mock import patch, Mock, call except ImportError: from mock import patch, Mock, call HTML_TEXT = """<html><head><title>Email spam</title></head><body> <p><b>Email spam</b>, also known as <b>junk email</b> or <b>unsolicited bulk email</b> (<i>UBE</i>), is a subset of <a href="/wiki/Spam_(electronic)" title="Spam (electronic)">electronic spam</a> involving nearly identical messages sent to numerous recipients by <a href="/wiki/Email" title="Email"> email</a>. Clicking on <a href="/wiki/Html_email#Security_vulnerabilities" title="Html email" class="mw-redirect"> links in spam email</a> may send users to <a href="/wiki/Phishing" title="Phishing">phishing</a> web sites or sites that are hosting <a href="/wiki/Malware" title="Malware">malware</a>.</body></html>""" HTML_TEXT_STRIPED = 'Email spam Email spam , also known as junk email or unsolicited bulk email ( UBE ),' \ ' is a subset of electronic spam involving nearly identical messages sent to numerous recipients by email' \ ' . Clicking on links in spam email may send users to phishing web sites or sites that are hosting malware .' HTML_STYLE = """<html><head></head><sTyle>Some random style</stylE> <body>This is a test.</body></html> """ HTML_SCRIPT = """<html><head></head><SCRIPT>Some random script</SCRIPT> <body>This is a test.</body></html> """ class HTMLStripperTests(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.data = [] def tearDown(self): unittest.TestCase.tearDown(self) def test_HTMLStripper(self): stripper = HTMLStripper(self.data) stripper.feed(HTML_TEXT) res = " ".join(self.data) self.assertEqual(res, HTML_TEXT_STRIPED) def test_strip_style(self): stripper = HTMLStripper(self.data) stripper.feed(HTML_STYLE) res = " ".join(self.data) self.assertEqual(res, "This is a test.") def test_strip_script(self): stripper = HTMLStripper(self.data) stripper.feed(HTML_SCRIPT) res = " ".join(self.data) self.assertEqual(res, "This is a test.") class PreDigestTests(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.lines = [] def mock_digest_paylods(c, message): yield message.decode("utf8") def mock_handle_line(s, line): self.lines.append(line.decode("utf8")) self.real_digest_payloads = DataDigester.digest_payloads self.real_handle_line = DataDigester.handle_line DataDigester.digest_payloads = mock_digest_paylods DataDigester.handle_line = mock_handle_line def tearDown(self): unittest.TestCase.tearDown(self) DataDigester.digest_payloads = self.real_digest_payloads DataDigester.handle_line = self.real_handle_line def test_predigest_emails(self): """Test email removal in the predigest process""" real_longstr = DataDigester.longstr_ptrn DataDigester.longstr_ptrn = re.compile(r'\S{100,}') emails = ["test@example.com", "test123@example.com", "test+abc@example.com", "test.test2@example.com", "test.test2+abc@example.com", ] message = "Test %s Test2" expected = "TestTest2" try: for email in emails: self.lines = [] DataDigester((message % email).encode("utf8")) self.assertEqual(self.lines[0], expected) finally: DataDigester.longstr_ptrn = real_longstr def test_predigest_urls(self): """Test url removal in the predigest process""" real_longstr = DataDigester.longstr_ptrn DataDigester.longstr_ptrn = re.compile(r'\S{100,}') urls = ["http://www.example.com", # "www.example.com", # XXX This also fail "http://example.com", # "example.com", # XXX This also fails "http://www.example.com/test/" "http://www.example.com/test/test2", ] message = "Test %s Test2" expected = "TestTest2" try: for url in urls: self.lines = [] DataDigester((message % url).encode("utf8")) self.assertEqual(self.lines[0], expected) finally: DataDigester.longstr_ptrn = real_longstr def test_predigest_long(self): """Test long "words" removal in the predigest process""" strings = ["0A2D3f%a#S", "3sddkf9jdkd9", "@@#@@@@@@@@@"] message = "Test %s Test2" expected = "TestTest2" for string in strings: self.lines = [] DataDigester((message % string).encode("utf8")) self.assertEqual(self.lines[0], expected) def test_predigest_min_line_lenght(self): """Test small lines removal in the predigest process""" message = "This line is included\n" \ "not this\n" \ "This also" expected = ["Thislineisincluded", "Thisalso"] DataDigester(message.encode("utf8")) self.assertEqual(self.lines, expected) def test_predigest_atomic(self): """Test atomic messages (lines <= 4) in the predigest process""" message = "All this message\nShould be included\nIn the predigest" expected = ["Allthismessage", "Shouldbeincluded", "Inthepredigest"] DataDigester(message.encode("utf8")) self.assertEqual(self.lines, expected) def test_predigest_pieced(self): """Test pieced messages (lines > 4) in the predigest process""" message = "" for i in range(100): message += "Line%d test test test\n" % i expected = [] for i in [20, 21, 22, 60, 61, 62]: expected.append("Line%dtesttesttest" % i) DataDigester(message.encode("utf8")) self.assertEqual(self.lines, expected) class DigestTests(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) self.lines = [] def mock_digest_paylods(c, message): yield message.decode("utf8") self.real_digest_payloads = DataDigester.digest_payloads DataDigester.digest_payloads = mock_digest_paylods def tearDown(self): unittest.TestCase.tearDown(self) DataDigester.digest_payloads = self.real_digest_payloads def test_digest(self): message = b"That's some good ham right there" predigested = b"That'ssomegoodhamrightthere" digest = hashlib.sha1() digest.update(predigested) expected = digest.hexdigest() result = DataDigester(message).value self.assertEqual(result, expected) def test_digest_w_null(self): message = b"That's some good ham rig\x00ht there" predigested = b"That'ssomegoodhamrightthere" digest = hashlib.sha1() digest.update(predigested) expected = digest.hexdigest() result = DataDigester(message).value self.assertEqual(result, expected) class MessageDigestTest(unittest.TestCase): def setUp(self): unittest.TestCase.setUp(self) patch("pyzor.digest.DataDigester.normalize_html_part", return_value="normalized").start() self.config = { "get_content_maintype.return_value": "text", "get_content_charset.return_value": "utf8", "get_payload.return_value": Mock(), "get_payload.return_value.decode.return_value": "decoded" } def tearDown(self): unittest.TestCase.tearDown(self) patch.stopall() def check_msg(self): mock_part = Mock(**self.config) conf = {"walk.return_value": [mock_part]} mock_msg = Mock(**conf) return mock_part, mock_msg, list(DataDigester.digest_payloads(mock_msg)) def test_text(self): mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, ["decoded"]) expected = [call.decode('utf8', 'ignore')] payload = mock_part.get_payload.return_value payload.assert_has_calls(expected, True) def test_text_no_charset(self): self.config["get_content_charset.return_value"] = None mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, ["decoded"]) expected = [call.decode('ascii', 'ignore')] payload = mock_part.get_payload.return_value payload.assert_has_calls(expected) def test_text_quopri(self): self.config["get_content_charset.return_value"] = "quopri" mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, ["decoded"]) expected = [call.decode('quopri', 'strict')] payload = mock_part.get_payload.return_value payload.assert_has_calls(expected) def test_text_lookuperror(self): def _decode(encoding, errors): if encoding not in ("ascii",): raise LookupError() return "decoded" self.config["get_payload.return_value.decode.side_effect"] = _decode mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, ["decoded"]) expected = [call.decode('utf8', 'ignore'), call.decode('ascii', 'ignore')] payload = mock_part.get_payload.return_value payload.assert_has_calls(expected) def test_text_unicodeerror(self): self.config["get_payload.return_value.decode.side_effect"] = UnicodeError mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, []) expected = [call.decode('utf8', 'ignore'), call.decode('ascii', 'ignore')] payload = mock_part.get_payload.return_value payload.assert_has_calls(expected) def test_html(self): self.config["get_content_subtype.return_value"] = "html" mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, ["normalized"]) def test_multipart(self): self.config["get_content_maintype.return_value"] = "nottext" self.config["is_multipart.return_value"] = True mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, []) def test_nontext(self): self.config["get_content_maintype.return_value"] = "nottext" self.config["is_multipart.return_value"] = False mock_part, mock_msg, result = self.check_msg() self.assertEqual(result, [mock_part.get_payload.return_value]) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(HTMLStripperTests)) test_suite.addTest(unittest.makeSuite(PreDigestTests)) test_suite.addTest(unittest.makeSuite(DigestTests)) test_suite.addTest(unittest.makeSuite(MessageDigestTest)) return test_suite if __name__ == '__main__': unittest.main()
11,210
35.399351
129
py
pyzor
pyzor-master/tests/unit/test_engines/test_redis.py
"""Test the pyzor.engines.gdbm_ module.""" import time import logging import unittest from datetime import datetime try: from unittest.mock import Mock, patch, call except ImportError: from mock import Mock, patch, call import pyzor.engines.redis_ import pyzor.engines.common class EncodingRedisTest(unittest.TestCase): """Test the RedisDBHandle class""" r_count = 24 wl_count = 42 entered = datetime(2014, 4, 23, 15, 41, 30) updated = datetime(2014, 4, 25, 17, 22, 25) wl_entered = datetime(2014, 2, 12, 11, 10, 55) wl_updated = datetime(2014, 3, 25, 5, 1, 50) def setUp(self): unittest.TestCase.setUp(self) self.record = pyzor.engines.common.Record(self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated) self.entered_st = int(time.mktime(self.entered.timetuple())) self.updated_st = int(time.mktime(self.updated.timetuple())) self.wl_entered_st = int(time.mktime(self.wl_entered.timetuple())) self.wl_updated_st = int(time.mktime(self.wl_updated.timetuple())) def compare_records(self, r1, r2): attrs = ("r_count", "r_entered", "r_updated", "wl_count", "wl_entered", "wl_updated") self.assertTrue(all(getattr(r1, attr) == getattr(r2, attr) for attr in attrs)) def tearDown(self): unittest.TestCase.tearDown(self) def test_encode_record(self): expected = { "r_count": 24, "r_entered": self.entered_st, "r_updated": self.updated_st, "wl_count": 42, "wl_entered": self.wl_entered_st, "wl_updated": self.wl_updated_st } result = pyzor.engines.redis_.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_encode_record_no_date(self): expected = { "r_count": 24, "r_entered": self.entered_st, "r_updated": 0, "wl_count": 42, "wl_entered": self.wl_entered_st, "wl_updated": self.wl_updated_st } self.record.r_updated = None result = pyzor.engines.redis_.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_encode_record_no_white(self): expected = { "r_count": 24, "r_entered": self.entered_st, "r_updated": self.updated_st, "wl_count": 0, "wl_entered": 0, "wl_updated": 0 } self.record.wl_count = 0 self.record.wl_entered = None self.record.wl_updated = None result = pyzor.engines.redis_.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_decode_record(self): encoded = { b"r_count": 24, b"r_entered": self.entered_st, b"r_updated": self.updated_st, b"wl_count": 42, b"wl_entered": self.wl_entered_st, b"wl_updated": self.wl_updated_st } result = pyzor.engines.redis_.RedisDBHandle._decode_record(encoded) self.compare_records(result, self.record) def test_decode_record_no_date(self): encoded = { b"r_count": 24, b"r_entered": self.entered_st, b"r_updated": 0, b"wl_count": 42, b"wl_entered": self.wl_entered_st, b"wl_updated": self.wl_updated_st } result = pyzor.engines.redis_.RedisDBHandle._decode_record(encoded) self.record.r_updated = None self.compare_records(result, self.record) def test_decode_record_no_white(self): encoded = { b"r_count": 24, b"r_entered": self.entered_st, b"r_updated": self.updated_st, b"wl_count": 0, b"wl_entered": 0, b"wl_updated": 0 } result = pyzor.engines.redis_.RedisDBHandle._decode_record(encoded) self.record.wl_count = 0 self.record.wl_entered = None self.record.wl_updated = None self.compare_records(result, self.record) class RedisTest(unittest.TestCase): max_age = 60 * 60 def setUp(self): unittest.TestCase.setUp(self) logger = logging.getLogger("pyzord") logger.addHandler(logging.NullHandler()) self.mredis = patch("pyzor.engines.redis_.redis", create=True).start() patch("pyzor.engines.redis_.RedisDBHandle._encode_record", side_effect=lambda x: x).start() patch("pyzor.engines.redis_.RedisDBHandle._decode_record", side_effect=lambda x: x).start() def tearDown(self): unittest.TestCase.tearDown(self) patch.stopall() def test_init(self): expected = {"host": "example.com", "port": 6387, "password": "passwd", "db": 5, } db = pyzor.engines.redis_.RedisDBHandle("example.com,6387,passwd,5", None) self.mredis.StrictRedis.assert_called_with(**expected) def test_init_defaults(self): expected = {"host": "localhost", "port": 6379, "password": None, "db": 0, } db = pyzor.engines.redis_.RedisDBHandle(",,,", None) self.mredis.StrictRedis.assert_called_with(**expected) def test_set(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" value = "record test" db = pyzor.engines.redis_.RedisDBHandle(",,,", None) db[digest] = value expected = ("pyzord.digest_v1.%s" % digest, value) self.mredis.StrictRedis.return_value.hmset.assert_called_with(*expected) def test_set_max_age(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" value = "record test" db = pyzor.engines.redis_.RedisDBHandle(",,,", None, max_age=self.max_age) db[digest] = value expected1 = ("pyzord.digest_v1.%s" % digest, value) expected2 = ("pyzord.digest_v1.%s" % digest, self.max_age) self.mredis.StrictRedis.return_value.hmset.assert_called_with(*expected1) self.mredis.StrictRedis.return_value.expire.assert_called_with(*expected2) def test_get(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_.RedisDBHandle(",,,", None) result = db[digest] expected = ("pyzord.digest_v1.%s" % digest,) self.mredis.StrictRedis.return_value.hgetall.assert_called_with(*expected) def test_items(self): patch("pyzor.engines.redis_.redis.StrictRedis.return_value.keys", return_value=["2aedaac999d71421c9ee49b9d81f627a7bc570aa"]).start() digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_.RedisDBHandle(",,,", None) db.items()[0] expected = ("pyzord.digest_v1.%s" % digest,) self.mredis.StrictRedis.return_value.keys.assert_called_with("pyzord.digest_v1.*") self.mredis.StrictRedis.return_value.hgetall.assert_called_with(*expected) def test_delete(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_.RedisDBHandle(",,,", None) del db[digest] expected = ("pyzord.digest_v1.%s" % digest,) self.mredis.StrictRedis.return_value.delete.assert_called_with(*expected) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(EncodingRedisTest)) test_suite.addTest(unittest.makeSuite(RedisTest)) return test_suite if __name__ == '__main__': unittest.main()
8,064
34.528634
90
py
pyzor
pyzor-master/tests/unit/test_engines/test_mysql.py
"""Test the pyzor.engines.mysql module.""" import unittest import threading from datetime import datetime, timedelta import pyzor.engines import pyzor.engines.mysql import pyzor.engines.common class MockTimer(): def __init__(self, *args, **kwargs): pass def start(self): pass def setDaemon(self, daemon): pass def make_MockMySQL(result, queries): class MockCursor(): def __init__(self): self.done = False def fetchone(self): if not self.done: self.done = True return result else: return None def fetchall(self): return [result] def execute(self, query, args=None): queries.append((query, args)) def close(self): pass class MockDB(): def cursor(self, *args, **kwargs): return MockCursor() def close(self): pass def commit(self): pass def autocommit(self, value): pass class MockMysql(): @staticmethod def connect(*args, **kwargs): return MockDB() class Error(Exception): pass class cursors: class SSCursor: pass return MockMysql class MySQLTest(unittest.TestCase): """Test the GdbmDBHandle class""" max_age = 60 * 60 * 24 * 30 * 4 r_count = 24 wl_count = 42 entered = datetime.now() - timedelta(days=10) updated = datetime.now() - timedelta(days=2) wl_entered = datetime.now() - timedelta(days=20) wl_updated = datetime.now() - timedelta(days=3) handler = pyzor.engines.mysql.MySQLDBHandle def setUp(self): unittest.TestCase.setUp(self) self.real_timer = threading.Timer threading.Timer = MockTimer self.record = pyzor.engines.common.Record(self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated) self.response = self.record_unpack() self.queries = [] mock_MySQL = make_MockMySQL(self.response, self.queries) try: self.real_mysql = pyzor.engines.mysql.MySQLdb except AttributeError: self.real_mysql = None setattr(pyzor.engines.mysql, "MySQLdb", mock_MySQL) def tearDown(self): unittest.TestCase.tearDown(self) threading.Timer = self.real_timer pyzor.engines.mysql.MySQLdb = self.real_mysql def record_unpack(self, record=None): if not record: record = self.record return (record.r_count, record.wl_count, record.r_entered, record.r_updated, record.wl_entered, record.wl_updated) def test_reconnect(self): """Test MySQLDBHandle.__init__""" expected = "DELETE FROM testtable WHERE r_updated<%s" self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) self.assertEqual(self.queries[0][0], expected) def test_no_reorganize(self): self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=None) self.assertFalse(self.queries) def test_set_item(self): """Test MySQLDBHandle.__setitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" expected = ("INSERT INTO testtable (digest, r_count, wl_count, " "r_entered, r_updated, wl_entered, wl_updated) " "VALUES (%s, %s, %s, %s, %s, %s, %s) ON " "DUPLICATE KEY UPDATE r_count=%s, wl_count=%s, " "r_entered=%s, r_updated=%s, wl_entered=%s, " "wl_updated=%s", (digest, self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated, self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated)) handle = self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) handle[digest] = self.record self.assertEqual(self.queries[1], expected) def test_get_item(self): """Test MySQLDBHandle.__getitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" expected = ("SELECT r_count, wl_count, r_entered, r_updated, " "wl_entered, wl_updated FROM testtable WHERE digest=%s", (digest,)) handle = self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) result = handle[digest] self.assertEqual(self.queries[1], expected) self.assertEqual(self.record_unpack(result), self.record_unpack()) def test_del_item(self): """Test MySQLDBHandle.__detitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" expected = ("DELETE FROM testtable WHERE digest=%s", (digest,)) handle = self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) del handle[digest] self.assertEqual(self.queries[1], expected) def test_items(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" expected = ("SELECT digest, r_count, wl_count, r_entered, r_updated, " "wl_entered, wl_updated FROM testtable", None) self.response = (digest, self.response) handle = self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) handle.items() self.assertEqual(self.queries[1], expected) def test_iter(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" expected = ("SELECT digest FROM testtable", None) self.response = (digest,) handle = self.handler("testhost,testuser,testpass,testdb,testtable", None, max_age=self.max_age) for d in handle: pass self.assertEqual(self.queries[1], expected) class ThreadedMySQLTest(MySQLTest): """Test the GdbmDBHandle class""" handler = pyzor.engines.mysql.ThreadedMySQLDBHandle class ProcessesMySQLTest(MySQLTest): """Test the GdbmDBHandle class""" handler = pyzor.engines.mysql.ProcessMySQLDBHandle def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(MySQLTest)) test_suite.addTest(unittest.makeSuite(ThreadedMySQLTest)) test_suite.addTest(unittest.makeSuite(ProcessesMySQLTest)) return test_suite if __name__ == '__main__': unittest.main()
6,907
33.368159
83
py
pyzor
pyzor-master/tests/unit/test_engines/test_gdbm.py
"""Test the pyzor.engines.gdbm_ module.""" import sys import time import unittest import threading from datetime import datetime, timedelta import pyzor.engines.gdbm_ import pyzor.engines.common class MockTimer(): def __init__(self, *args, **kwargs): pass def start(self): pass def setDaemon(self, daemon): pass class MockGdbmDB(dict): """Mock a gdbm database""" def firstkey(self): if not self.keys(): return None self.key_index = 1 return list(self.keys())[0] def nextkey(self, key): if len(self.keys()) <= self.key_index: return None else: self.key_index += 1 return self.keys()[self.key_index] def sync(self): pass def reorganize(self): pass class GdbmTest(unittest.TestCase): """Test the GdbmDBHandle class""" handler = pyzor.engines.gdbm_.GdbmDBHandle max_age = 60 * 60 * 24 * 30 * 4 r_count = 24 wl_count = 42 entered = datetime.now() - timedelta(days=10) updated = datetime.now() - timedelta(days=2) wl_entered = datetime.now() - timedelta(days=20) wl_updated = datetime.now() - timedelta(days=3) def setUp(self): unittest.TestCase.setUp(self) self.real_timer = threading.Timer threading.Timer = MockTimer self.db = MockGdbmDB() class MockGdbm(): @staticmethod def open(fn, mode): return self.db try: self.real_gdbm = pyzor.engines.gdbm_.gdbm except AttributeError: self.real_gdbm = None setattr(pyzor.engines.gdbm_, "gdbm", MockGdbm()) self.record = pyzor.engines.common.Record(self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated) def tearDown(self): unittest.TestCase.tearDown(self) threading.Timer = self.real_timer pyzor.engines.gdbm_.gdbm = self.real_gdbm def record_as_str(self, record=None): if not record: record = self.record return ("1,%s,%s,%s,%s,%s,%s" % (record.r_count, record.r_entered, record.r_updated, record.wl_count, record.wl_entered, record.wl_updated)).encode("utf8") def test_set_item(self): """Test GdbmDBHandle.__setitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" handle = self.handler(None, None, max_age=self.max_age) handle[digest] = self.record self.assertEqual(self.db[digest], self.record_as_str().decode("utf8")) def test_get_item(self): """Test GdbmDBHandle.__getitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" handle = self.handler(None, None, max_age=self.max_age) self.db[digest] = self.record_as_str() result = handle[digest] self.assertEqual(self.record_as_str(result), self.record_as_str()) def test_items(self): """Test GdbmDBHandle.items()""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" handle = self.handler(None, None, max_age=self.max_age) self.db[digest] = self.record_as_str() key, result = handle.items()[0] self.assertEqual(key, digest) self.assertEqual(self.record_as_str(result), self.record_as_str()) def test_del_item(self): """Test GdbmDBHandle.__delitem__""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" handle = self.handler(None, None, max_age=self.max_age) self.db[digest] = self.record_as_str() del handle[digest] self.assertFalse(self.db.get(digest)) def test_reorganize_older(self): """Test GdbmDBHandle.start_reorganizing with older records""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.db[digest] = self.record_as_str() handle = self.handler(None, None, max_age=3600 * 24) self.assertFalse(self.db.get(digest)) def test_reorganize_older_no_max_age(self): """Test GdbmDBHandle.start_reorganizing with older records, but no max_age set. """ digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.db[digest] = self.record_as_str() handle = self.handler(None, None, max_age=None) self.assertEqual(self.db[digest], self.record_as_str()) def test_reorganize_fresh(self): """Test GdbmDBHandle.start_reorganizing with newer records""" digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" self.db[digest] = self.record_as_str() handle = self.handler(None, None, max_age=3600 * 24 * 3) self.assertEqual(self.db[digest], self.record_as_str()) class ThreadingGdbmTest(GdbmTest): """Test the GdbmDBHandle class""" handler = pyzor.engines.gdbm_.ThreadedGdbmDBHandle def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(GdbmTest)) test_suite.addTest(unittest.makeSuite(ThreadingGdbmTest)) return test_suite if __name__ == '__main__': unittest.main()
5,304
29.843023
94
py
pyzor
pyzor-master/tests/unit/test_engines/test_redis_v0.py
"""Test the pyzor.engines.gdbm_ module.""" import unittest from datetime import datetime import pyzor.engines.redis_v0 import pyzor.engines.common class EncodingRedisTest(unittest.TestCase): """Test the RedisDBHandle class""" r_count = 24 wl_count = 42 entered = datetime(2014, 4, 23, 15, 41, 30) updated = datetime(2014, 4, 25, 17, 22, 25) wl_entered = datetime(2014, 2, 12, 11, 10, 55) wl_updated = datetime(2014, 3, 25, 5, 1, 50) def setUp(self): unittest.TestCase.setUp(self) self.record = pyzor.engines.common.Record(self.r_count, self.wl_count, self.entered, self.updated, self.wl_entered, self.wl_updated) def compare_records(self, r1, r2): attrs = ("r_count", "r_entered", "r_updated", "wl_count", "wl_entered", "wl_updated") self.assertTrue(all(getattr(r1, attr) == getattr(r2, attr) for attr in attrs)) def tearDown(self): unittest.TestCase.tearDown(self) def test_encode_record(self): expected = ("24,2014-04-23 15:41:30,2014-04-25 17:22:25," "42,2014-02-12 11:10:55,2014-03-25 05:01:50").encode() result = pyzor.engines.redis_v0.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_encode_record_no_date(self): expected = ("24,2014-04-23 15:41:30,," "42,2014-02-12 11:10:55,2014-03-25 05:01:50").encode() self.record.r_updated = None result = pyzor.engines.redis_v0.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_encode_record_no_white(self): expected = ("24,2014-04-23 15:41:30,2014-04-25 17:22:25," "0,,").encode() self.record.wl_count = 0 self.record.wl_entered = None self.record.wl_updated = None result = pyzor.engines.redis_v0.RedisDBHandle._encode_record(self.record) self.assertEqual(result, expected) def test_decode_record(self): encoded = ("24,2014-04-23 15:41:30,2014-04-25 17:22:25," "42,2014-02-12 11:10:55,2014-03-25 05:01:50").encode() result = pyzor.engines.redis_v0.RedisDBHandle._decode_record(encoded) self.compare_records(result, self.record) def test_decode_record_no_date(self): encoded = ("24,2014-04-23 15:41:30,," "42,2014-02-12 11:10:55,2014-03-25 05:01:50").encode() result = pyzor.engines.redis_v0.RedisDBHandle._decode_record(encoded) self.record.r_updated = None self.compare_records(result, self.record) def test_decode_record_no_white(self): encoded = ("24,2014-04-23 15:41:30,2014-04-25 17:22:25," "0,,").encode() result = pyzor.engines.redis_v0.RedisDBHandle._decode_record(encoded) self.record.wl_count = 0 self.record.wl_entered = None self.record.wl_updated = None self.compare_records(result, self.record) def make_MockRedis(commands): class MockStrictRedis(): def __init__(self, *args, **kwargs): commands.append(("init", args, kwargs)) def set(self, *args, **kwargs): commands.append(("set", args, kwargs)) def setex(self, *args, **kwargs): commands.append(("setex", args, kwargs)) def get(self, *args, **kwargs): commands.append(("get", args, kwargs)) def delete(self, *args, **kwargs): commands.append(("delete", args, kwargs)) def keys(self, *args, **kwargs): commands.append(("keys", args, kwargs)) yield "pyzord.digest.2aedaac999d71421c9ee49b9d81f627a7bc570aa" class MockError(Exception): pass class exceptions(): def __init__(self): self.RedisError = MockError class MockRedis(): def __init__(self): self.StrictRedis = MockStrictRedis self.exceptions = exceptions() return MockRedis() mock_encode_record = lambda s, x: x mock_decode_record = lambda s, x: x class RedisTest(unittest.TestCase): max_age = 60 * 60 def setUp(self): unittest.TestCase.setUp(self) self.commands = [] try: self.real_redis = pyzor.engines.redis_v0.redis except AttributeError: self.real_redis = None self.real_encode = pyzor.engines.redis_v0.RedisDBHandle._encode_record self.real_decode = pyzor.engines.redis_v0.RedisDBHandle._decode_record setattr(pyzor.engines.redis_v0, "redis", make_MockRedis(self.commands)) pyzor.engines.redis_v0.RedisDBHandle._encode_record = mock_encode_record pyzor.engines.redis_v0.RedisDBHandle._decode_record = mock_decode_record def tearDown(self): unittest.TestCase.tearDown(self) pyzor.engines.redis_v0.redis = self.real_redis pyzor.engines.redis_v0.RedisDBHandle._encode_record = self.real_encode pyzor.engines.redis_v0.RedisDBHandle._decode_record = self.real_decode def test_init(self): expected = {"host": "example.com", "port": 6387, "password": "passwd", "db": 5, } db = pyzor.engines.redis_v0.RedisDBHandle("example.com,6387,passwd,5", None) self.assertEqual(self.commands[0], ("init", (), expected)) def test_init_defaults(self): expected = {"host": "localhost", "port": 6379, "password": None, "db": 0, } db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None) self.assertEqual(self.commands[0], ("init", (), expected)) def test_set(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" value = "record test" db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None) db[digest] = value expected = ("pyzord.digest.%s" % digest, value) self.assertEqual(self.commands[1], ("set", expected, {})) def test_set_max_age(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" value = "record test" db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None, max_age=self.max_age) db[digest] = value expected = ("pyzord.digest.%s" % digest, self.max_age, value) self.assertEqual(self.commands[1], ("setex", expected, {})) def test_get(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None) result = db[digest] expected = ("pyzord.digest.%s" % digest,) self.assertEqual(self.commands[1], ("get", expected, {})) def test_items(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None) db.items()[0] expected = ("pyzord.digest.%s" % digest,) self.assertEqual(self.commands[1], ("keys", ("pyzord.digest.*",), {})) self.assertEqual(self.commands[2], ("get", expected, {})) def test_delete(self): digest = "2aedaac999d71421c9ee49b9d81f627a7bc570aa" db = pyzor.engines.redis_v0.RedisDBHandle(",,,", None) del db[digest] expected = ("pyzord.digest.%s" % digest,) self.assertEqual(self.commands[1], ("delete", expected, {})) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(EncodingRedisTest)) test_suite.addTest(unittest.makeSuite(RedisTest)) return test_suite if __name__ == '__main__': unittest.main()
7,898
36.084507
83
py
pyzor
pyzor-master/tests/unit/test_engines/__init__.py
"""A suite of unit tests that verifies the correct behaviour of various functions/methods in the pyzord code. Note these tests the source of pyzor, not the version currently installed. """ import unittest def suite(): """Gather all the tests from this package in a test suite.""" import test_gdbm import test_mysql import test_redis import test_redis_v0 test_suite = unittest.TestSuite() test_suite.addTest(test_gdbm.suite()) test_suite.addTest(test_mysql.suite()) test_suite.addTest(test_redis.suite()) test_suite.addTest(test_redis_v0.suite()) return test_suite if __name__ == '__main__': unittest.main(defaultTest='suite')
682
23.392857
74
py
pyzor
pyzor-master/tests/util/__init__.py
"""This package contains various utilities use in the pyzor tests.""" import os import sys import time import redis import shutil import unittest import subprocess from datetime import datetime try: from unittest.mock import mock_open as _mock_open except ImportError: from mock import mock_open as _mock_open import pyzor.client def mock_open(mock=None, read_data=""): mock = _mock_open(mock, read_data) mock.return_value.__iter__ = lambda x: iter(read_data.splitlines()) return mock msg = """Newsgroups: Date: Wed, 10 Apr 2002 22:23:51 -0400 (EDT) From: Frank Tobin <ftobin@neverending.org> Fcc: sent-mail Message-ID: <20020410222350.E16178@palanthas.neverending.org> X-Our-Headers: X-Bogus,Anon-To X-Bogus: aaron7@neverending.org MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Test Email """ digest = "7421216f915a87e02da034cc483f5c876e1a1338" _dt_decode = lambda x: None if x == 'None' else datetime.strptime(x, "%a %b %d %H:%M:%S %Y") class PyzorTestBase(unittest.TestCase): """Test base that starts the pyzord daemon in setUpClass with specified arguments. The daemon is killed in tearDownClass. This also create the necessary files and the homedir. """ pyzord = None _args = {"homedir": "--homedir", "engine": "-e", "dsn": "--dsn", "address": "-a", "port": "-p", "threads": "--threads", "max_threads": "--max-threads", "processes": "--processes", "max_processes": "--max-processes", "db_connections": "--db-connections", "password_file": "--password-file", "access_file": "--access-file", "cleanup_age": "--cleanup-age", "log_file": "--log-file", "detach": "--detach", "prefork": "--pre-fork", } homedir = "./pyzor-test/" threads = "False" access_file = "pyzord.access" password_file = "pyzord.passwd" log_file = "pyzord-test.log" dsn = "localhost,,,10" engine = "redis" access = """check report ping pong info whitelist : alice : deny check report ping pong info whitelist : bob : allow ALL : dan : allow pong info whitelist : dan : deny """ passwd = """alice : fc7f1cad729b5f3862b2ef192e2d9e0d0d4bd515 bob : cf88277c5d4abdc0a3f56f416011966d04a3f462 dan : c1a50281fc43e860fe78c16c73b9618ada59f959 """ servers = """127.0.0.1:9999 """ accounts_alice = """127.0.0.1 : 9999 : alice : d28f86151e80a9accba4a4eba81c460532384cd6,fc7f1cad729b5f3862b2ef192e2d9e0d0d4bd515 """ accounts_bob = """127.0.0.1 : 9999 : bob : de6ef568787256bf5f55909dc0c398e49b5c9808,cf88277c5d4abdc0a3f56f416011966d04a3f462 """ accounts_chuck = """127.0.0.1 : 9999 : bob : de6ef568787256bf5f55909dc0c398e49b5c9808,af88277c5d4abdc0a3f56f416011966d04a3f462 """ accounts_dan = """127.0.0.1 : 9999 : dan : 1cc2efa77d8833d83556e0cc4fa617c64eebc7fb,c1a50281fc43e860fe78c16c73b9618ada59f959 """ @classmethod def write_homedir_file(cls, name, content): if not name or not content: return with open(os.path.join(cls.homedir, name), "w") as f: f.write(content) @classmethod def setUpClass(cls): super(PyzorTestBase, cls).setUpClass() try: os.mkdir(cls.homedir) except OSError: pass cls.write_homedir_file(cls.access_file, cls.access) cls.write_homedir_file(cls.password_file, cls.passwd) cls.write_homedir_file(cls.password_file, cls.passwd) cls.write_homedir_file("servers", cls.servers) cls.write_homedir_file("alice", cls.accounts_alice) cls.write_homedir_file("bob", cls.accounts_bob) cls.write_homedir_file("chuck", cls.accounts_chuck) cls.write_homedir_file("dan", cls.accounts_dan) args = ["pyzord"] for key, value in cls._args.items(): option = getattr(cls, key, None) if option: args.append(value) args.append(option) cls.pyzord = [] for line in cls.servers.splitlines(): line = line.strip() if not line: continue addr, port = line.rsplit(":", 1) cls.pyzord.append(subprocess.Popen(args + ["-a", addr, "-p", port])) time.sleep(1) # allow time to initialize server def setUp(self): unittest.TestCase.setUp(self) self.client_args = {"--homedir": self.homedir, "--servers-file": "servers", "-t": None, # timeout "-r": None, # report threshold "-w": None, # whitelist threshold "-s": None, # style } def tearDown(self): unittest.TestCase.tearDown(self) @classmethod def tearDownClass(cls): super(PyzorTestBase, cls).tearDownClass() for pyzord in cls.pyzord: pyzord.terminate() pyzord.wait() shutil.rmtree(cls.homedir, True) redis.StrictRedis(db=10).flushdb() def check_pyzor(self, cmd, user, input=None, code=None, exit_code=None, counts=()): """Call the pyzor client with the specified args from self.client_args and verifies the response. """ args = ["pyzor"] if user: args.append("--accounts-file") args.append(user) for key, value in self.client_args.items(): if value: args.append(key) args.append(value) args.append(cmd) pyzor = subprocess.Popen(args, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE) if input: stdout, stderr = pyzor.communicate(input.encode("utf8")) else: stdout, stderr = pyzor.communicate() if stderr: self.fail(stderr) if code is not None: try: stdout = stdout.decode("utf8") results = stdout.strip().split("\t") status = eval(results[1]) except Exception as e: self.fail("Parsing error: %s of %r" % (e, stdout)) self.assertEqual(status[0], code, status) if counts: self.assertEqual(counts, (int(results[2]), int(results[3]))) if exit_code is not None: self.assertEqual(exit_code, pyzor.returncode) return stdout def check_pyzor_multiple(self, cmd, user, input=None, code=None, exit_code=None, counts=()): """Call the pyzor client with the specified args from self.client_args and verifies the response. """ args = ["pyzor"] if user: args.append("--accounts-file") args.append(user) for key, value in self.client_args.items(): if value: args.append(key) args.append(value) args.append(cmd) pyzor = subprocess.Popen(args, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE) if input: stdout, stderr = pyzor.communicate(input.encode("utf8")) else: stdout, stderr = pyzor.communicate() if stderr: self.fail(stderr) stdout = stdout.decode("utf8") for i, line in enumerate(stdout.splitlines()): try: line = line.strip() if not line: continue results = line.strip().split("\t") except Exception as e: self.fail("Parsing error: %s of %r" % (e, stdout)) if code is not None: try: status = eval(results[1]) except Exception as e: self.fail("Parsing error: %s of %r" % (e, stdout)) self.assertEqual(status[0], code[i], status) if counts: self.assertEqual((int(results[2]), int(results[3])), counts[i]) if exit_code is not None: self.assertEqual(exit_code, pyzor.returncode) return stdout def check_digest(self, digest, address, counts=(0, 0)): result = self.client.check(digest, address) self.assertEqual((int(result["Count"]), int(result["WL-Count"])), counts) return result def get_record(self, input, user="bob"): """Uses `pyzor info` to get the record data.""" stdout = self.check_pyzor("info", user, input, code=200, exit_code=0) info = stdout.splitlines()[1:] record = {} try: for line in info: line = line.strip() if not line: continue key, value = line.split(":", 1) record[key.strip()] = value.strip() except Exception as e: self.fail("Error parsing %r: %s" % (info, e)) return record def check_fuzzy_date(self, date1, date2=None, seconds=10): """Check if the given date is almost equal to now.""" date1 = _dt_decode(date1) if not date2: date2 = datetime.now() delta = abs((date2 - date1).total_seconds()) if delta > seconds: self.fail("Delta %s is too big: %s, %s" % (delta , date1, date2)) class PyzorTest(object): """MixIn class for PyzorTestBase that performs a series of basic tests.""" def test_ping(self): self.check_pyzor("ping", "bob") def test_pong(self): input = "Test1 pong1 Test2" self.check_pyzor("pong", "bob", input=input, code=200, exit_code=0, counts=(sys.maxsize, 0)) def test_check(self): input = "Test1 check1 Test2" self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(0, 0)) r = self.get_record(input) self.assertEqual(r["Count"], "0") def test_report(self): input = "Test1 report1 Test2" self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=0, counts=(1, 0)) r = self.get_record(input) self.assertEqual(r["Count"], "1") self.check_fuzzy_date(r["Entered"]) def test_report_update(self): input = "Test1 report update1 Test2" self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=0, counts=(1, 0)) time.sleep(1) self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=0, counts=(2, 0)) r = self.get_record(input) self.assertEqual(r["Count"], "2") self.assertNotEqual(r["Entered"], r["Updated"]) self.check_fuzzy_date(r["Updated"]) def test_whitelist(self): input = "Test1 white list1 Test2" self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(0, 1)) r = self.get_record(input) self.assertEqual(r["WL-Count"], "1") self.check_fuzzy_date(r["WL-Entered"]) def test_whitelist_update(self): input = "Test1 white list update1 Test2" self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(0, 1)) time.sleep(1) self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(0, 2)) r = self.get_record(input) self.assertEqual(r["WL-Count"], "2") self.assertNotEqual(r["WL-Entered"], r["WL-Updated"]) self.check_fuzzy_date(r["WL-Updated"]) def test_report_whitelist(self): input = "Test1 white list report1 Test2" self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(1, 1)) r = self.get_record(input) self.assertEqual(r["Count"], "1") self.check_fuzzy_date(r["Entered"]) self.assertEqual(r["WL-Count"], "1") self.check_fuzzy_date(r["WL-Entered"]) def test_report_whitelist_update(self): input = "Test1 white list report update1 Test2" self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(1, 1)) time.sleep(1) self.check_pyzor("whitelist", "bob", input=input, code=200, exit_code=0) self.check_pyzor("report", "bob", input=input, code=200, exit_code=0) self.check_pyzor("check", "bob", input=input, code=200, exit_code=1, counts=(2, 2)) r = self.get_record(input) self.assertEqual(r["Count"], "2") self.assertNotEqual(r["Entered"], r["Updated"]) self.check_fuzzy_date(r["Updated"]) self.assertEqual(r["WL-Count"], "2") self.assertNotEqual(r["WL-Entered"], r["WL-Updated"]) self.check_fuzzy_date(r["WL-Updated"])
14,111
36.73262
132
py
pyzor
pyzor-master/tests/benchmark/__init__.py
0
0
0
py
pyzor
pyzor-master/tests/benchmark/measure_server_response.py
from __future__ import division import json import Queue import timeit import optparse import threading import collections DIGEST = "da39a3ee5e6b4b0d3255bfef95601890afd80709" SETUP = """ import pyzor import string import random import hashlib import pyzor.client digest = "".join(random.choice(string.letters) for _ in range(50)) digest = hashlib.sha1(digest).hexdigest() client = pyzor.client.Client(timeout=%f) """ CMD = """ try: client.%s(digest, address=(%r, %s)) except pyzor.TimeoutError: pass """ ALL_METHODS = ("pong", "check", "report", "info", "whitelist") def measure_method(method, repeats, timeout, server, queue): setup = SETUP % timeout cmd = CMD % ((method,) + server) results = timeit.repeat(stmt=cmd, setup=setup, repeat=repeats, number=1) timeouts = sum(1 for result in results if result >= timeout) queue.put((method, results, timeouts)) def measure_methods(methods, repeats, timeout, server, queue): if methods == "all": methods = ALL_METHODS else: methods = methods.split(",") threads = [] for method in methods: thread = threading.Thread(target=measure_method, args=(method, repeats, timeout, server, queue)) threads.append(thread) thread.start() for thread in threads: thread.join() def json_handler(res): fres = {} for method, result in res.items(): fres[method] = {"runs": [], "timeouts": result["timeouts"], "totals": {}} total = 0 best_all = [] for i, results in enumerate(result["results"]): results.sort() average = sum(results) / len(results) best = results[:3] total += average best_all.extend(best) fres[method]["runs"].append({"average": average, "best": best}) fres[method]["totals"]["average"] = total / len(result["results"]) fres[method]["totals"]["best"] = best_all[:3] print json.dumps(fres, indent=4) def print_handler(res): for method, result in res.items(): print "=" * 80 print "Method: %s" % method print "Timeouts: %s" % result["timeouts"] print "=" * 80 total = 0 best_all = [] for i, results in enumerate(result["results"]): results.sort() average = sum(results) / len(results) best = results[:3] total += average best_all.extend(best) print "\t(%s) %s %s" % (i, average, best) print "=" * 80 print "Total: %s %s" % (total / len(result["results"]), best_all[:3]) print "\n" def main(): opt = optparse.OptionParser() opt.add_option("-n", "--nice", dest="nice", type="int", help="'nice' level", default=0) opt.add_option("-s", "--server", dest="server", default="127.0.0.1:24441") opt.add_option("-m", "--method", dest="method", default="all") opt.add_option("-f", "--format", dest="format", default="print") opt.add_option("-t", "--threads", dest="threads", type="int", default=1) opt.add_option("--timeout", dest="timeout", type="float", default=5.0) opt.add_option("-r", "--repeats", dest="repeats", type="int", default=1000) options, args = opt.parse_args() server = tuple(options.server.rsplit(":", 1)) queue = Queue.Queue() threads = [] for dummy in range(options.threads): thread = threading.Thread(target=measure_methods, args=(options.method, options.repeats, options.timeout, server, queue)) threads.append(thread) thread.start() for thread in threads: thread.join() def get_new_info(): return {"results": [], "timeouts": 0} res = collections.defaultdict(get_new_info) while True: try: method, results, timeouts = queue.get_nowait() res[method]["results"].append(results) res[method]["timeouts"] += timeouts except Queue.Empty: break globals()["%s_handler" % options.format](res) if __name__ == '__main__': main()
4,339
27.552632
79
py
pyzor
pyzor-master/tests/functional/test_forwarder.py
import os import time import shutil import unittest import subprocess import redis class ForwardSetup(object): """Setup forwarding client and 'remote' pyzord""" def write_homedir_file(self, name, content): if not name or not content: return with open(os.path.join(self.homedir, name), "w") as f: f.write(content) def __init__(self, homedir): self.homedir = homedir try: os.mkdir(homedir) except OSError: pass @unittest.skip("This fails randomly on PyPy.") class ForwarderTest(unittest.TestCase): def setUp(self): self.localserver = ForwardSetup('./pyzor-test-forwardserver') # we also use this dir for the local client self.localserver.write_homedir_file('servers', '127.0.0.1:9999\n') self.fwdclient = ForwardSetup('./pyzor-test-forwardingclient') self.fwdclient.write_homedir_file('servers', '127.0.0.1:9998\n') args = ["pyzord", "--homedir", self.localserver.homedir, '-e', 'redis', '--dsn', 'localhost,,,10', '-a', '127.0.0.1', '-p', '9999', '--forward-client-homedir', self.fwdclient.homedir] self.local_pyzord_proc = subprocess.Popen(args) self.remoteserver = ForwardSetup('./pyzor-test-remoteserver') args = ["pyzord", "--homedir", self.remoteserver.homedir, '-e', 'redis', '--dsn', 'localhost,,,9', '-a', '127.0.0.1', '-p', '9998'] self.remote_pyzord_proc = subprocess.Popen(args) time.sleep(0.3) def test_forward_report(self): # submit hash to local server for i in range(10): self.check_pyzor("report", self.localserver.homedir) # make sure the local submission worked self.check_pyzor("check", self.localserver.homedir, counts=(10, 0)) # now use the forwarding client's config to check forwarded submission time.sleep(1) self.check_pyzor("check", self.fwdclient.homedir, counts=(10, 0)) # submit the hash to the remote system, the count should go up self.check_pyzor("report", self.fwdclient.homedir) self.check_pyzor("check", self.fwdclient.homedir, counts=(11, 0)) # switch back to our local server, the count should still be the old value self.check_pyzor("check", self.localserver.homedir, counts=(10, 0)) def tearDown(self): if self.remote_pyzord_proc is not None: self.remote_pyzord_proc.kill() if self.local_pyzord_proc is not None: self.local_pyzord_proc.kill() shutil.rmtree(self.localserver.homedir, True) shutil.rmtree(self.fwdclient.homedir, True) shutil.rmtree(self.remoteserver.homedir, True) redis.StrictRedis(db=9).flushdb() redis.StrictRedis(db=10).flushdb() def check_pyzor(self, cmd, homedir, counts=None, msg=None): """simplified check_pyzor version from PyzorTestBase""" msg = "This is a test message for the forwading feature" args = ["pyzor", '--homedir', homedir, cmd] pyzor = subprocess.Popen(args, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE) stdout, stderr = pyzor.communicate(msg.encode("utf8")) if stderr: self.fail(stderr) try: stdout = stdout.decode("utf8") results = stdout.strip().split("\t") status = eval(results[1]) except Exception as e: self.fail("Parsing error: %s of %r" % (e, stdout)) self.assertEqual(status[0], 200, status) if counts: self.assertEqual(counts, (int(results[2]), int(results[3]))) return stdout def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(ForwarderTest)) return test_suite if __name__ == '__main__': unittest.main()
4,002
35.390909
191
py
pyzor
pyzor-master/tests/functional/test_account.py
import unittest from tests.util import * class AccountPyzorTest(PyzorTestBase): # test bob which has access to everything def test_ping(self): self.check_pyzor("ping", "bob", code=200, exit_code=0) def test_pong(self): self.check_pyzor("pong", "bob", input=msg, code=200, exit_code=0) def test_check(self): self.check_pyzor("check", "bob", input=msg, code=200) def test_report(self): self.check_pyzor("report", "bob", input=msg, code=200, exit_code=0) def test_whitelist(self): self.check_pyzor("whitelist", "bob", input=msg, code=200, exit_code=0) def test_info(self): self.check_pyzor("info", "bob", input=msg, code=200, exit_code=0) # test alice which does not has access to anything # Error should be 403 Forbidden def test_ping_forbidden(self): self.check_pyzor("ping", "alice", code=403, exit_code=1) def test_pong_forbidden(self): self.check_pyzor("pong", "alice", input=msg, code=403, exit_code=1) def test_check_forbidden(self): self.check_pyzor("check", "alice", input=msg, code=403, exit_code=1) def test_report_forbidden(self): self.check_pyzor("report", "alice", input=msg, code=403, exit_code=1) def test_whitelist_forbidden(self): self.check_pyzor("whitelist", "alice", input=msg, code=403, exit_code=1) def test_info_forbidden(self): self.check_pyzor("info", "alice", input=msg, code=403, exit_code=1) # test chuck which does tries to steal bob's account but has the wrong key # Error should be 401 Unauthorized def test_ping_unauthorized(self): self.check_pyzor("ping", "chuck", code=401, exit_code=1) def test_pong_unauthorized(self): self.check_pyzor("pong", "chuck", input=msg, code=401, exit_code=1) def test_check_unauthorized(self): self.check_pyzor("check", "chuck", input=msg, code=401, exit_code=1) def test_report_unauthorized(self): self.check_pyzor("report", "chuck", input=msg, code=401, exit_code=1) def test_whitelist_unauthorized(self): self.check_pyzor("whitelist", "chuck", input=msg, code=401, exit_code=1) def test_info_unauthorized(self): self.check_pyzor("info", "chuck", input=msg, code=401, exit_code=1) # test dan account, which has some access def test_ping_combo(self): self.check_pyzor("ping", "dan", code=200, exit_code=0) def test_pong_combo(self): self.check_pyzor("pong", "dan", input=msg, code=403, exit_code=1) def test_check_combo(self): self.check_pyzor("check", "dan", input=msg, code=200) def test_report_combo(self): self.check_pyzor("report", "dan", input=msg, code=200, exit_code=0) def test_whitelist_combo(self): self.check_pyzor("whitelist", "dan", input=msg, code=403, exit_code=1) def test_info_combo(self): self.check_pyzor("info", "dan", input=msg, code=403, exit_code=1) # test anonymous account, which should is not currently set up in the server def test_ping_anonymous(self): self.check_pyzor("ping", None, code=403, exit_code=1) def test_pong_anonymous(self): self.check_pyzor("pong", None, input=msg, code=403, exit_code=1) def test_check_anonymous(self): self.check_pyzor("check", None, input=msg, code=403, exit_code=1) def test_report_anonymous(self): self.check_pyzor("report", None, input=msg, code=403, exit_code=1) def test_whitelist_anonymous(self): self.check_pyzor("whitelist", None, input=msg, code=403, exit_code=1) def test_info_anonymous(self): self.check_pyzor("info", None, input=msg, code=403, exit_code=1) class AnonymousPyzorTest(PyzorTestBase): """Test accounts with no access or password file set-up. And test anonymous default access. """ access_file = None password_file = None def test_ping(self): self.check_pyzor("ping", None, code=200, exit_code=0) def test_pong(self): self.check_pyzor("pong", None, input=msg, code=200, exit_code=0) def test_check(self): self.check_pyzor("check", None, input=msg, code=200) def test_report(self): self.check_pyzor("report", None, input=msg, code=200, exit_code=0) def test_whitelist(self): # anonymous account are not allowed to whitelist by default self.check_pyzor("whitelist", None, input=msg, code=403, exit_code=1) def test_info(self): self.check_pyzor("info", None, input=msg, code=200, exit_code=0) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(AccountPyzorTest)) test_suite.addTest(unittest.makeSuite(AnonymousPyzorTest)) return test_suite if __name__ == '__main__': unittest.main()
4,999
34.971223
84
py
pyzor
pyzor-master/tests/functional/test_pyzor.py
import io import sys import redis import unittest from tests.util import * MBOX_FILE_PATH = os.path.join(os.path.dirname(os.path.abspath(__file__)), "test_data", "test.mbx") class PyzorScriptTest(PyzorTestBase): password_file = None access = """ALL : anonymous : allow """ def test_report_threshold(self): input = "Test1 report threshold 1 Test2" self.client_args["-r"] = "2" self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 0)) self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(2, 0)) # Exit code will be success now, since the report count exceeds the # threshold self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(3, 0)) def test_whitelist_threshold(self): input = "Test1 white list threshold 1 Test2" self.client_args["-w"] = "2" self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 0)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 1)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 2)) # Exit code will be failure now, since the whitelist count exceeds the # threshold self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 3)) def test_report_whitelist_threshold(self): input = "Test1 report white list threshold 1 Test2" self.client_args["-w"] = "2" self.client_args["-r"] = "1" self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 0)) # Exit code will be success now, since the report count exceeds the # thresholdRedisPyzorTest self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(2, 0)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(2, 1)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(2, 2)) # Exit code will be failure now, since the whitelist count exceeds the # threshold self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(2, 3)) def test_digest_style(self): input = "da39a3ee5e6b4b0d3255bfef95601890afd80700" self.client_args["-s"] = "digests" self.check_pyzor("pong", None, input=input, code=200, exit_code=0, counts=(sys.maxsize, 0)) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(0, 0)) self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 0)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 1)) r = self.get_record(input, None) self.assertEqual(r["Count"], "1") self.assertEqual(r["WL-Count"], "1") def test_digest_style_multiple(self): input2 = "da39a3ee5e6b4b0d3255bfef95601890afd80705\n"\ "da39a3ee5e6b4b0d3255bfef95601890afd80706\n" input3 = "da39a3ee5e6b4b0d3255bfef95601890afd80705\n"\ "da39a3ee5e6b4b0d3255bfef95601890afd80706\n"\ "da39a3ee5e6b4b0d3255bfef95601890afd80707\n" self.client_args["-s"] = "digests" self.check_pyzor_multiple("pong", None, input=input3, exit_code=0, code=[200, 200, 200], counts=[(sys.maxsize, 0), (sys.maxsize, 0), (sys.maxsize, 0)]) self.check_pyzor_multiple("check", None, input=input3, exit_code=1, code=[200, 200, 200], counts=[(0, 0), (0, 0), (0, 0)]) self.check_pyzor_multiple("report", None, input=input2, exit_code=0) self.check_pyzor_multiple("check", None, input=input3, exit_code=0, code=[200, 200, 200], counts=[(1, 0), (1, 0), (0, 0)]) self.check_pyzor_multiple("whitelist", None, input=input3, exit_code=0) self.check_pyzor_multiple("check", None, input=input3, exit_code=1, code=[200, 200, 200], counts=[(1, 1), (1, 1), (0, 1)]) def test_mbox_style(self): input = "From MAILER-DAEMON Mon Jan 6 15:13:33 2014\n\nTest1 message 0 Test2\n\n" self.client_args["-s"] = "mbox" self.check_pyzor("pong", None, input=input, code=200, exit_code=0, counts=(sys.maxsize, 0)) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(0, 0)) self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 0)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 1)) r = self.get_record(input, None) self.assertEqual(r["Count"], "1") self.assertEqual(r["WL-Count"], "1") def test_mbox_real(self): with io.open(MBOX_FILE_PATH, 'rt', encoding='latin-1') as mbox_file: input = mbox_file.read() self.client_args["-s"] = "mbox" self.check_pyzor("pong", None, input=input, code=200, exit_code=0, counts=(sys.maxsize, 0)) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(0, 0)) self.check_pyzor("report", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=0, counts=(1, 0)) self.check_pyzor("whitelist", None, input=input, code=200, exit_code=0) self.check_pyzor("check", None, input=input, code=200, exit_code=1, counts=(1, 1)) r = self.get_record(input, None) self.assertEqual(r["Count"], "1") self.assertEqual(r["WL-Count"], "1") def test_mbox_style_multiple(self): input2 = "From MAILER-DAEMON Mon Jan 6 15:08:02 2014\n\nTest1 message 1 Test2\n\n"\ "From MAILER-DAEMON Mon Jan 6 15:08:05 2014\n\nTest1 message 2 Test2\n\n" input3 = "From MAILER-DAEMON Mon Jan 6 15:08:02 2014\n\nTest1 message 1 Test2\n\n"\ "From MAILER-DAEMON Mon Jan 6 15:08:05 2014\n\nTest1 message 2 Test2\n\n"\ "From MAILER-DAEMON Mon Jan 6 15:08:08 2014\n\nTest1 message 3 Test2\n\n" self.client_args["-s"] = "mbox" self.check_pyzor_multiple("pong", None, input=input3, exit_code=0, code=[200, 200, 200], counts=[(sys.maxsize, 0), (sys.maxsize, 0), (sys.maxsize, 0)]) self.check_pyzor_multiple("check", None, input=input3, exit_code=1, code=[200, 200, 200], counts=[(0, 0), (0, 0), (0, 0)]) self.check_pyzor_multiple("report", None, input=input2, exit_code=0) self.check_pyzor_multiple("check", None, input=input3, exit_code=0, code=[200, 200, 200], counts=[(1, 0), (1, 0), (0, 0)]) self.check_pyzor_multiple("whitelist", None, input=input3, exit_code=0) self.check_pyzor_multiple("check", None, input=input3, exit_code=1, code=[200, 200, 200], counts=[(1, 1), (1, 1), (0, 1)]) def test_predigest(self): out = self.check_pyzor("predigest", None, input=msg).strip() self.assertEqual(out.decode("utf8"), "TestEmail") def test_digest(self): out = self.check_pyzor("digest", None, input=msg).strip() self.assertEqual(out.decode("utf8"), digest) class MultipleServerPyzorScriptTest(PyzorTestBase): password_file = None access = """ALL : anonymous : allow """ servers = """127.0.0.1:9999 127.0.0.1:9998 127.0.0.1:9997 """ def test_ping(self): self.check_pyzor_multiple("ping", None, exit_code=0, code=[200, 200, 200]) def test_pong(self): input = "Test1 multiple pong Test2" self.check_pyzor_multiple("pong", None, input=input, exit_code=0, code=[200, 200, 200], counts=[(sys.maxsize, 0), (sys.maxsize, 0), (sys.maxsize, 0)]) def test_check(self): input = "Test1 multiple check Test2" self.check_pyzor_multiple("check", None, input=input, exit_code=1, code=[200, 200, 200], counts=[(0, 0), (0, 0), (0, 0)]) def test_report(self): input = "Test1 multiple report Test2" self.check_pyzor_multiple("report", None, input=input, exit_code=0, code=[200, 200, 200]) def test_whitelist(self): input = "Test1 multiple whitelist Test2" self.check_pyzor_multiple("whitelist", None, input=input, exit_code=0, code=[200, 200, 200]) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(PyzorScriptTest)) test_suite.addTest(unittest.makeSuite(MultipleServerPyzorScriptTest)) return test_suite if __name__ == '__main__': unittest.main()
11,584
49.58952
112
py
pyzor
pyzor-master/tests/functional/test_server.py
import sys import time import errno import unittest try: import configparser as ConfigParser except ImportError: import ConfigParser import pyzor.client from tests.util import * try: import MySQLdb has_mysql = True except ImportError: has_mysql = False try: import redis has_redis = True except ImportError: has_redis = False try: import gdbm has_gdbm = True except ImportError: has_gdbm = False class BatchedDigestsTest(object): def setUp(self): PyzorTestBase.setUp(self) self.client = pyzor.client.BatchClient() def test_batched_report(self): digest = "da39a3ee5e6b4b0d3255bfef95601890afd80709" for i in range(9): self.client.report(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999)) self.client.report(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999), (10, 0)) def test_batched_whitelist(self): digest = "da39a3ee5e6b4b0d3255bfef95601890afd80708" for i in range(9): self.client.whitelist(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999)) self.client.whitelist(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999), (0, 10)) def test_batched_combined(self): digest = "da39a3ee5e6b4b0d3255bfef95601890afd80707" for i in range(9): self.client.report(digest, ("127.0.0.1", 9999)) self.client.whitelist(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999)) self.client.report(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999), (10, 0)) self.client.whitelist(digest, ("127.0.0.1", 9999)) self.check_digest(digest, ("127.0.0.1", 9999), (10, 10)) def test_batched_multiple_report(self): digest = "%sa39a3ee5e6b4b0d3255bfef95601890afd80706" for i in range(10): self.client.report(digest % i, ("127.0.0.1", 9999)) for i in range(10): self.check_digest(digest % i, ("127.0.0.1", 9999), (1, 0)) def test_batched_multiple_whitelist(self): digest = "%sa39a3ee5e6b4b0d3255bfef95601890afd80705" for i in range(10): self.client.whitelist(digest % i, ("127.0.0.1", 9999)) for i in range(10): self.check_digest(digest % i, ("127.0.0.1", 9999), (0, 1)) def test_multiple_addresses_report(self): digest1 = "da39a3ee5e6b4b0d3255bfef95601890afd80704" digest2 = "da39a3ee5e6b4b0d3255bfef95601890afd80703" for i in range(9): self.client.report(digest1, ("127.0.0.1", 9999)) self.client.report(digest2, ("127.0.0.1", 9998)) self.check_digest(digest1, ("127.0.0.1", 9999)) self.check_digest(digest2, ("127.0.0.1", 9998)) self.client.report(digest1, ("127.0.0.1", 9999)) self.check_digest(digest1, ("127.0.0.1", 9999), (10, 0)) self.client.report(digest2, ("127.0.0.1", 9998)) self.check_digest(digest2, ("127.0.0.1", 9998), (10, 0)) def test_multiple_addresses_whitelist(self): digest1 = "da39a3ee5e6b4b0d3255bfef95601890afd80702" digest2 = "da39a3ee5e6b4b0d3255bfef95601890afd80701" for i in range(9): self.client.whitelist(digest1, ("127.0.0.1", 9999)) self.client.whitelist(digest2, ("127.0.0.1", 9998)) self.check_digest(digest1, ("127.0.0.1", 9999)) self.check_digest(digest2, ("127.0.0.1", 9998)) self.client.whitelist(digest1, ("127.0.0.1", 9999)) self.check_digest(digest1, ("127.0.0.1", 9999), (0, 10)) self.client.whitelist(digest2, ("127.0.0.1", 9998)) self.check_digest(digest2, ("127.0.0.1", 9998), (0, 10)) schema = """ CREATE TABLE IF NOT EXISTS `%s` ( `digest` char(40) default NULL, `r_count` int(11) default NULL, `wl_count` int(11) default NULL, `r_entered` datetime default NULL, `wl_entered` datetime default NULL, `r_updated` datetime default NULL, `wl_updated` datetime default NULL, PRIMARY KEY (`digest`) ) """ @unittest.skipIf(not os.path.exists("./test.conf"), "test.conf is not available") @unittest.skipIf(not has_mysql, "MySQLdb library not available") class MySQLdbBatchedPyzorTest(BatchedDigestsTest, PyzorTestBase): """Test the mysql engine.""" dsn = None engine = "mysql" password_file = None access = """ALL : anonymous : allow """ servers = """127.0.0.1:9999 127.0.0.1:9998 """ @classmethod def setUpClass(cls): conf = ConfigParser.ConfigParser() conf.read("./test.conf") table = conf.get("test", "table") db = MySQLdb.Connect(host=conf.get("test", "host"), user=conf.get("test", "user"), passwd=conf.get("test", "passwd"), db=conf.get("test", "db")) c = db.cursor() c.execute(schema % table) c.close() db.close() cls.dsn = "%s,%s,%s,%s,%s" % (conf.get("test", "host"), conf.get("test", "user"), conf.get("test", "passwd"), conf.get("test", "db"), conf.get("test", "table")) super(MySQLdbBatchedPyzorTest, cls).setUpClass() @classmethod def tearDownClass(cls): super(MySQLdbBatchedPyzorTest, cls).tearDownClass() try: conf = ConfigParser.ConfigParser() conf.read("./test.conf") table = conf.get("test", "table") db = MySQLdb.Connect(host=conf.get("test", "host"), user=conf.get("test", "user"), passwd=conf.get("test", "passwd"), db=conf.get("test", "db")) c = db.cursor() c.execute("DROP TABLE %s" % table) c.close() db.close() except: pass @unittest.skipIf(not has_redis, "redis library not available") class RedisBatchedPyzorTest(BatchedDigestsTest, PyzorTestBase): """Test the redis engine""" dsn = "localhost,,,10" engine = "redis" password_file = None access = """ALL : anonymous : allow """ servers = """127.0.0.1:9999 127.0.0.1:9998 """ @classmethod def tearDownClass(cls): super(RedisBatchedPyzorTest, cls).tearDownClass() redis.StrictRedis(db=10).flushdb() class DetachPyzorTest(PyzorTestBase): detach = "/dev/null" homedir = os.path.join(os.getcwd(), "pyzor-test") def test_pid(self): self.assertTrue(os.path.exists(os.path.join(self.homedir, "pyzord.pid"))) @staticmethod def is_running(pid): try: os.kill(pid, 0) except OSError as err: if err.errno == errno.ESRCH: return False return True @classmethod def tearDownClass(cls): with open(os.path.join(cls.homedir, "pyzord.pid")) as pidf: pid = int(pidf.read().strip()) os.kill(pid, 15) while cls.is_running(pid): time.sleep(0.25) super(DetachPyzorTest, cls).tearDownClass() def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(MySQLdbBatchedPyzorTest)) # test_suite.addTest(unittest.makeSuite(GdbmBatchedPyzorTest)) test_suite.addTest(unittest.makeSuite(RedisBatchedPyzorTest)) test_suite.addTest(unittest.makeSuite(DetachPyzorTest)) return test_suite if __name__ == '__main__': unittest.main()
7,853
31.188525
70
py
pyzor
pyzor-master/tests/functional/__init__.py
"""A suite of functional tests that verifies the correct behaviour of the pyzor client and server as a whole. Functional test should not touch real data and are usually safe, but it's not recommended to run theses on production servers. Note these tests the installed version of pyzor, not the version from the source. """ import unittest def suite(): """Gather all the tests from this package in a test suite.""" import test_pyzor import test_server import test_digest import test_engines import test_account import test_forwarder test_suite = unittest.TestSuite() test_suite.addTest(test_pyzor.suite()) test_suite.addTest(test_digest.suite()) test_suite.addTest(test_server.suite()) test_suite.addTest(test_engines.suite()) test_suite.addTest(test_account.suite()) test_suite.addTest(test_forwarder.suite()) return test_suite if __name__ == '__main__': unittest.main(defaultTest='suite')
964
26.571429
77
py
pyzor
pyzor-master/tests/functional/test_digest.py
# -*- coding: utf-8 -*- import sys import hashlib import unittest from tests.util import * TEXT = """MIME-Version: 1.0 Sender: chirila@spamexperts.com Received: by 10.216.90.129 with HTTP; Fri, 23 Aug 2013 01:59:03 -0700 (PDT) Date: Fri, 23 Aug 2013 11:59:03 +0300 Delivered-To: chirila@spamexperts.com X-Google-Sender-Auth: p6ay4c-tEtdFpavndA9KBmP0CVs Message-ID: <CAK-mJS9aV6Kb7Z5XCRJ_z_UOKEaQjRY8gMzsuxUQcN5iqxNWUg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila@spamexperts.com> To: Alexandru Chirila <chirila@spamexperts.com> Content-Type: multipart/alternative; boundary=001a11c2893246a9e604e4999ea3 --001a11c2893246a9e604e4999ea3 Content-Type: text/plain; charset=ISO-8859-1 %s --001a11c2893246a9e604e4999ea3 """ HTML_TEXT = """MIME-Version: 1.0 Sender: chirila@gapps.spamexperts.com Received: by 10.216.157.70 with HTTP; Thu, 16 Jan 2014 00:43:31 -0800 (PST) Date: Thu, 16 Jan 2014 10:43:31 +0200 Delivered-To: chirila@gapps.spamexperts.com X-Google-Sender-Auth: ybCmONS9U9D6ZUfjx-9_tY-hF2Q Message-ID: <CAK-mJS8sE-V6qtspzzZ+bZ1eSUE_FNMt3K-5kBOG-z3NMgU_Rg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila@spamexperts.com> To: Alexandru Chirila <chirila@gapps.spamexperts.com> Content-Type: multipart/alternative; boundary=001a11c25ff293069304f0126bfd --001a11c25ff293069304f0126bfd Content-Type: text/plain; charset=ISO-8859-1 Email spam. Email spam, also known as junk email or unsolicited bulk email, is a subset of electronic spam involving nearly identical messages sent to numerous recipients by email. Clicking on links in spam email may send users to phishing web sites or sites that are hosting malware. --001a11c25ff293069304f0126bfd Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"><div>Email spam.</div><div><br></div><div>Email spam, also= known as junk email or unsolicited bulk email, is a subset of electronic s= pam involving nearly identical messages sent to numerous recipients by emai= l. Clicking on links in spam email may send users to phishing web sites or = sites that are hosting malware.</div> </div> --001a11c25ff293069304f0126bfd-- """ HTML_TEXT_STYLE_SCRIPT = """MIME-Version: 1.0 Sender: chirila@gapps.spamexperts.com Received: by 10.216.157.70 with HTTP; Thu, 16 Jan 2014 00:43:31 -0800 (PST) Date: Thu, 16 Jan 2014 10:43:31 +0200 Delivered-To: chirila@gapps.spamexperts.com X-Google-Sender-Auth: ybCmONS9U9D6ZUfjx-9_tY-hF2Q Message-ID: <CAK-mJS8sE-V6qtspzzZ+bZ1eSUE_FNMt3K-5kBOG-z3NMgU_Rg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila@spamexperts.com> To: Alexandru Chirila <chirila@gapps.spamexperts.com> Content-Type: multipart/alternative; boundary=001a11c25ff293069304f0126bfd --001a11c25ff293069304f0126bfd Content-Type: text/plain; charset=ISO-8859-1 This is a test. --001a11c25ff293069304f0126bfd Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"> <style> This is my style.</style> <script> This is my script.</script> <div>This is a test.</div> </div> --001a11c25ff293069304f0126bfd-- """ TEXT_ATTACHMENT = """MIME-Version: 1.0 Received: by 10.76.127.40 with HTTP; Fri, 17 Jan 2014 02:21:43 -0800 (PST) Date: Fri, 17 Jan 2014 12:21:43 +0200 Delivered-To: chirila.s.alexandru@gmail.com Message-ID: <CALTHOsuHFaaatiXJKU=LdDCo4NmD_h49yvG2RDsWw17D0-NXJg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila.s.alexandru@gmail.com> To: Alexandru Chirila <chirila.s.alexandru@gmail.com> Content-Type: multipart/mixed; boundary=f46d040a62c49bb1c804f027e8cc --f46d040a62c49bb1c804f027e8cc Content-Type: multipart/alternative; boundary=f46d040a62c49bb1c404f027e8ca --f46d040a62c49bb1c404f027e8ca Content-Type: text/plain; charset=ISO-8859-1 This is a test mailing --f46d040a62c49bb1c404f027e8ca-- --f46d040a62c49bb1c804f027e8cc Content-Type: image/png; name="tar.png" Content-Disposition: attachment; filename="tar.png" Content-Transfer-Encoding: base64 X-Attachment-Id: f_hqjas5ad0 iVBORw0KGgoAAAANSUhEUgAAAskAAADlCAAAAACErzVVAAAACXBIWXMAAAsTAAALEwEAmpwYAAAD GGlDQ1BQaG90b3Nob3AgSUNDIHByb2ZpbGUAAHjaY2BgnuDo4uTKJMDAUFBUUuQe5BgZERmlwH6e gY2BmYGBgYGBITG5uMAxIMCHgYGBIS8/L5UBFTAyMHy7xsDIwMDAcFnX0cXJlYE0wJpcUFTCwMBw gIGBwSgltTiZgYHhCwMDQ3p5SUEJAwNjDAMDg0hSdkEJAwNjAQMDg0h2SJAzAwNjCwMDE09JakUJ AwMDg3N+QWVRZnpGiYKhpaWlgmNKflKqQnBlcUlqbrGCZ15yflFBflFiSWoKAwMD1A4GBgYGXpf8 EgX3xMw8BSMDVQYqg4jIKAUICxE+CDEESC4tKoMHJQODAIMCgwGDA0MAQyJDPcMChqMMbxjFGV0Y SxlXMN5jEmMKYprAdIFZmDmSeSHzGxZLlg6WW6x6rK2s99gs2aaxfWMPZ9/NocTRxfGFM5HzApcj 1xZuTe4FPFI8U3mFeCfxCfNN45fhXyygI7BD0FXwilCq0A/hXhEVkb2i4aJfxCaJG4lfkaiQlJM8 JpUvLS19QqZMVl32llyfvIv8H4WtioVKekpvldeqFKiaqP5UO6jepRGqqaT5QeuA9iSdVF0rPUG9 V/pHDBYY1hrFGNuayJsym740u2C+02KJ5QSrOutcmzjbQDtXe2sHY0cdJzVnJRcFV3k3BXdlD3VP XS8Tbxsfd99gvwT//ID6wIlBS4N3hVwMfRnOFCEXaRUVEV0RMzN2T9yDBLZE3aSw5IaUNak30zky LDIzs+ZmX8xlz7PPryjYVPiuWLskq3RV2ZsK/cqSql01jLVedVPrHzbqNdU0n22VaytsP9op3VXU fbpXta+x/+5Em0mzJ/+dGj/t8AyNmf2zvs9JmHt6vvmCpYtEFrcu+bYsc/m9lSGrTq9xWbtvveWG bZtMNm/ZarJt+w6rnft3u+45uy9s/4ODOYd+Hmk/Jn58xUnrU+fOJJ/9dX7SRe1LR68kXv13fc5N m1t379TfU75/4mHeY7En+59lvhB5efB1/lv5dxc+NH0y/fzq64Lv4T8Ffp360/rP8f9/AA0ADzT6 lvFdAAAAIGNIUk0AAHolAACAgwAA+f8AAIDpAAB1MAAA6mAAADqYAAAXb5JfxUYAAGrVSURBVHja 7J13YE3nG8c/Nzc7BLFX7F01ajcENVrU3qpVtalds2oWNWtTe1Zqj1IzLaVG0VJq7y1myM79/v44 92ZI/FDNoHn/4Nxz3vOcc28+5z3P+7zPoBivT+sko9V6je65nvWeP32N7rm89Z77m16fe34b0tV8 XZqLl/UX9nyNqMhuveeiLq/N75w6hfWea9q/NvecCaip16V5vtYke742v3PFCJJTvDb33DyJ5CSS k0hORCSn/PDpPXaZ8nokkZxE8mtCcrpKn45Ze+yeJJWNSk+1dmseSjr1uW3G4tX5k7RRO6RqPaRv FfuIj3mbZk8iOYnkBCP5o5uSpCcXjh7dN8otCjzOku78uHRnkL6iaxko+4eknYC5Qds8QJWldyXp WM6Ut1eZgU9CdTzy5DKdxyzZuGzmrA+hWiHK9Ju6YtPBP5aXsB6uNXjh4iVfN47xVijewBGwy+Dp mkRyEskvSXKFKZP7t6mrqUCZbXcO1HCNJPlgLoBs14PzaDSZ74Uv/qRrObBbKoV8BD5aX7tw7e+1 30saB++HXT8vZ+PUCtuDbFdbiJvW5QiWJP8wzbLK3mscDKmKqUbnombodaMkeG6RRpuqrrkjhf/e PInkJJJfXk9OoflQIzjotPRwZgpjn4N2GBtD1EU+DFJf42MH/dnhYWBu1qoQwDb1ljQ+z92Qsnts JPd+fPwHrcqeMnv2zA6gvdk1J4+HG3jZxvwzux1T5O4ZoKHmxZK2ubBXp1yd/9Yhbamvez9Omb41 XF2SSE4i+aVJzq7vyXrvbvF0jebe1NFkhvqstcbBpvoqfC9LVZDcH6SF46F5aa8ZjFAFsnecd1Ff aM453dTnnAqxybMnh+bTcrsT4KjddZ5e0Qj5gYprw7TLY7Sml9+mBeyWuvbUUg/tLqsVAO21O4nk JJJfmuTimsV3agGQbJmaAVBU842D3dT10V766MTPUnD1KvIBN/875r7qfE5S4IQK+jrfBY3AMeRc pERvTWRTIEBabe+uKcOnZo08mEmTW0saayoRvsKE41+Wwr+HBf55KTiTo/bk0lJMVdeHW1olkZxE 8j8geYZ70EXDDPG5PjdMFxphHFwsL//fcFwi/TZZGyboxJX79wNVubsu6maPova8r8G45IA82hQp sYEGcOVvnEq2maoVAyQpyoJ+RQ0o/KekcdtDMwKfatTNs4ukJWTStg80qfIx6a/3k/TkJJJfnuT8 +qGGJhrbPqps1YfbAZDsvp+LfIH875jwu3BIunns+GlN76bRAdpZDOqqFwA1NDlSYm996hZ+b88T SZrVTz82zB+Fy5ZqC2k/PiWtByigjUH7Cwf6l6CwlrdWzyuyjHdOsl0kPMmWtc1yOznl+/i8JIVK kq58UdQzp3evvVK49+hESHI2rRmu6gDkCr7kCMBY1QBguL7NpO+tHe9cDTntDrgGnO6l+nk2KrQZ LdXZyme3SImTVC6vFHp4Rqe+Gt1PjaNxOVQfWl1sPgbIpZ+1nBxpoaYm9lfjTrekOz1MSSQnNMnT wSlfHleS7dLdRuaShyW9Rfp8acHUTTdonAhJLqylK5QLIM1xtTb2rVYegKrBd9N76dt3CwIU1BWt AmC/uqk11HryOHsrGRazFuofKXGO8qXXFjegokZ1iMo4MEclALpLtY3Rf56mA/CZ+g1XTey8vw1Q +ySSE5rkAQwJk0LHUuVJUbyh2oUAB5ZL11Y5my2XqJsISS6qKUuVBUwNzmmhdTA8EeoI5h4BIVX4 RN1O/QQk36ZFGmfVQdrpS6CHhtpIbmhTrAFmqaynlhr69oQGGgEOmTLbVIb1ygFk8rNovhvJugc8 qaIZVh29zZdqCFBZ3yWRnNAkL2W6pIvV6NiftuETK5LhxnoG3v97bx2q6TAdEyHJxfRtb61rN+ys QobY3uqPzuFS6y/5VYEv1Wh6eAHHcoe1tYzGALDSUk2LAI+wAy0MowfemhIpsYfaptM6Q3fYVkG/ LD8VKoXuywfArypcpdv82xpwVgHXpXsV8+trANqqVxu1BZL/qk+SSE5okr9nYphvV0eq3PFIfl9S FwYe57OKQLEb+pk+iZDkIprvsk9S6KrIiZlfwIlQaWMGYIq8K0kh0mKnytbpXaYmTnemAYwbVtuq YWe2XIh0NyqhOXYPtgJwOLi4pJBj69f8eiq7oZVLkgK/sss46cSVvSMzkvJSPQBKak9dDSmX5dPT WulxsHUSyQlL8kaaZYMMMyx7DZ34Au8cpeOYLJB+nzYyPBGS7DykHHbvd6qTKsow2N8/5PRcbwC+ uZmPtod/X9rQhH2dbLYODjYXurbWpbtxDwtGnG1qk5eMRo/CTWjRtnikpxHmGUfXT2z3rktMRyHT z2rb8RdJltkuH0Y1hSSRnDAkQ4v1wdK3TJSkMLPnUdpKFweaCmgbIxIhyc9p9uYX6mZK8+o+bxvz UWD5yq8LkPyUvJNITmiSU5L+qwfSeFZJ0jGq7qCzJBXjzs0SPq8fyQnQPPZFuBslkZxQJG9geDMX 0qzVHIZLstRl8hoGSFIZbib83b4eJDe2zHFMIjmBSd7CQN2f4Gb+6aI540n9XYdSoQuYoNA/2lFU D8aGvN4kN+oeLyTbpUha40twkn+jnaTf7DM+7os5GVS4q+9IW9AFMpxSZUYlQpI/XLVkZLWo5Dja Q54Je87/Nb18dMIyPvS3i5i55Snf+JNSUQ5mqPnl9Jkdnklnnur2MXc2O7C5EoBjUsxIoiP5bxpI UlcWWxYVyVttdpjkg8k1e52ZD6Rh5pmJj+TmFkkag0dZsAP40O9q6nmhYVf+uClNp9YYaF0AAJe9 GmCc8+7swwGS9MgRINXhud/9fF2SdAMgU3bnZvM2bVg0xCvK8stj7c1k3X7rLQB+nPKxRQprCyVD PgTsK6YE76+m9i7rkERywpP8xFRTks64T4zYFTThqG0zMPHpyWnv+hXOVu7Hrxn8yH7PrlTwUZhU R3fTASUPq95ykTrwURnAvEY/WO0YV+S/b/GYnt0HlQf4QJIlxH9401L5kmf5cMQey52Tkn+opDkR lo1j2q4LVvPGXwcBclr+ehBStOKN0JL0UQPAW11cf5Qk3WqTRHKCk6x5Pxv4Js67jUlyIw0xNr6Q 5xMdz1It+KHlrsN62RsZi/rvt0CNsNvZYLx22LSAhSoCwC41B/qoSybXGZas9AiQJP3lo1+yQsYG Fy0201xl7WGCFhkfVoS5A/10TLOgorawUsWAypqUSz8XSllqSoB6JJGc4CQn7haT5A5qa2x0VGP5 Wa49DGir5cxXFrBr/igk39VbQFcdcqtlORexitdHnwGwSvcywTDVhSGq5fDkypGtExtmwSPcFyDT 7Se2iKeZqofbtaD0AEzWu2A+F7ZXFcB0OtTtXIA9UE1TyhuedXmfHE4iOZGRfDXRk1xFJycdvXJ2 hGNrfar5TUJD6rVUd2ZoVOsx5/WwtosOAqYF2uwXWjqKK/3MblM72jFWWgQzVQm6qmUaTc74Vi4T cOcUUO6s+tlOOPvEGXpYw/P6qTW8r1VXgxyB5SoT/hdAdY2qYT3jmF8SyYmL5JVudxM7yXwnhV3x 15yO6qUl5MpHN31mrE+GrvbkLS0BcPpNGhjJVi1JUh76yWIpwXQVgubqkEeSNBk4GTx/6XlZ/Y2A FJZrk5f9ckErAaivYTBDLRTw593wx0H6TqsBmqt3LQ0GSBF6IInkxEXyZZYlepKpXSs56R486KyP QvcDfKyBTNaYHnXTAy30JQDp74VFcZX4RL99Uak8dNHs8KPO01UUPtYnBXVu4/LN3YCTkkK31Yzo X0GSFBB2ycjyoo3YXfF/Vwq8cPTYfR02+G2pjtU0FaCdRiaRnMi0i6z9Ez/JABxVbzW/cBmgrBbz jaoY+4epvrGxTMUje3ezKtef6fMJmj1dReEzNc9udZmHw7qlsOWR12ithRWyubFLyQFMN54kq6Bl dY25Zj9dNaJKmqtzSS0DMvo/Tp9EciIjuVb1xE5yjqwAyZ6c76jmu0IB3J7csusha3TzChvAtgAp AAYbHvF8ou5Ox3ROxaGPGuQyvOuBo8FOH5/Sg4a2/q3UEeBblQNgguqPVwvr49BPfmoK0ExfZtMW sN9mfQ8kkZyISO6XNbGTfOOgEzBQYzqq6XoBsFRVG9hsc9uV3v2jaSva0cUIuzPaN1ase6sJJZ9I FaG7ambRGmwhJ+DY/XFIhYgxuYPhgd/K6om88UJgqsGqCTBKUgWAUlqeTL+Td5t22CeRnNhInmd6 nMhJHqdfmndZrlsebVVvdjAAdTQml29V4/AGrfaXdM++VtTA0rkqZ1/owxYdh2/PDU3D9QF8oQZO utj/61GTF256Z588gFo6ZUWyur4wArBHGZ/3SasZojoAC3RRt+763bjd7GJIZv8Hq4O1LnnSGl+i I/ligfuJnGS3jZL0ZyEyj8/oYey1/65OBEUfSbdHls2Yn3RfZYxka6aOGosgXgBf3csHDYLrcc2Q HfbJCpU34lo7Gv2T+w0y5pITjM+VpDZ436gIUP7rWmcuXLhw6Ub5kl91lfR73SS/i0RIcmJrsc34 6g3oXObZ3vQVKzjE3NlF4We+H9qtYx/Dochs/cf1g5pVvMoWTkPuTi4A7kNtuYfM1lQCtpimnnNd YrvYOzO6FkzyIEoi+VVsFy/ZUjj8o9MyJOVPfqrd27fx5yexHQg/fCA4MZBsORRNP1558uWEXvvx mYf2DAxPBCQn5bT/N0gOnVAJIPXiyF1XN0/+bvkJ6UBOSDfbknAk+/1p/L+QL6L0uW1OfT066Gcl 6fQPP9yTFHjr9N41s8ZaHebOBkpqwQ5b16gPZuhhqR33k0h+Q0h+hFPVLv0+c2aeLnUsuVz6u7kd gKn2dhe7mp+4MzDhSK5u5ytJ9zKZ9xxftewPaXx3i6Q+TLC9NGY/kDTBdER3qwH1FNrZyPNqhEgp 0HWYpA10ubo2UJKqpQ/b/0D+mQdK0hguqrRzcBLJb8qY7JZTklbgtT8FkO9wDz5Z5us7OzdvmxZJ 5z3sgxKM5HXUl6T2VMkO0OW+0/uStI1W1g4/4iNpCaPUEa9JK0/rJrnqftJ9vM/GW5KkX5gkKcCp RH+2SLKkcb7E17pDDUnrTG9JTgX0xpDcKd9/XE/Ol14KX5OOKekdv+1fmBLfsl7SAvvMm76VdNrd PeFIViGn29KvZvfkDm3fKpODdoYvzjW8rcf786ek3bRSMe5I0gn6RRU3jYWSVMalAVMkXaD6X3SW cruc0zkPp726yvv/OsnODXI+k6LSk0d7AThGTvGqzqj0f7BLnzv6ZLBov0IR22UnZyZa1s5ZT1cW HTH969rm/w7JhdPPrO2JecpgxkiqRm8m63x7UvwsKXxVWsYl4Izva7rqbh5TWQboyaNVOLBOkkRh 6/Gy5jBJl6mm+uy6tGDZkz18GVVcf36SpCbkJH+ItIrel6grLef9gKKmBdIR2v3rJJfxafQsLtuH SapKunlnbreBNuMdYYy0M+2MguD++bdFMfX60gXACaDo7AybLArY1igy1WaXMAV1wm72AMN3412A rIZPR8bbAXlT5ajkbnR8rxjmOZKkY9n+MyRnAdI13q2sTvcljaElPb5zoOTfksJr4TItIW0Xj9Ob t33G50Xtbko6DvwuSfcoZhwOdC4kSUHk0lKSA2mm41n0vcsR57fkD0NqdpgojWXhXSpJqkZhOkr6 /dVSy8VKcjuf0tGTsLjbhsUyIfcrl95Yyu0Pi2RpwWXNNY3WXa3qp02mj69Ix+kr/ZUdKgcPAr5T nTu75p6TVtmsxe9ZTmy+YymZU5a3gev3TACrLQ0A0zaNnBYiXc1jX9gOx6BTTos0MmfKwlMse/4z JLsC7uN1lxKStIRmNGsNrp8HSGsNoBPQCrfV7GGX+6FrXknaCxyRpD+pbRw9Zd3wSKnQrm5eI1pS FHDZH4XkS5JUiYIkcz2u3myWOb+kq6koHiTpHO3/bZId589zAs8vvlswOCfm+p7t5/rMMfKBcyCs CsBsdf+iY+DtZJKO6UJ3jR0lHZNllY56+D9crEtpmaPgzPBH6M5jQOXd2mAN9z8WWpg6WlNMmg0e 2gVAqUD/AvC5jp2xHNd6fd9enrBWjY0iI+xViv8Iyf54rW2XijFXqC5JC+hP1UczP8pAuSBNpXtC r4z0xW7vTePW5mAyzGmTGWQc/JOPJUnpUhmfD5GOoVHPbsI5SaEp0r/HNAoHdcNXqTJLuprScPK1 fLzl3yW5ULOOPpM79HFqs2TkJJ+leUr5+PjM/mrau9YEnCsBSlh+AXzkrYOndDLrMDWeqmBtf7uS Fg9Wb8ZoEdukRTgHHf3lFoDjQav8DzUVTMdDSkuPklNIy2yuccfNBQIfnVB7j5mOoRfX3jFBOw3T WMC11l1/+/8IyedoLJ1O4XzN0TNUUj9WOBaRFFqP6XqYpUFCkxyWyxx6h3clqQE1WCBJpUwnjIMn jTE5yJTX+LyF/NFzYHRhv6SRdG7EhdZ078jvypZSUk2y8LMkyw39qySbFvj4+Pj4LE/lmPXdFNV8 2uTz8alnFxlOUt1wxqwLjFAfrUn/kTublXe1SlU18YW6XAzywO10eO4dfpdDixfRcl85APygAlaf 0GLARHWQNJgKEf7Lq9XtkOYZecXvPXi0GfhAG7R98R83wxTW5Y3VLgIDJeluqBT45cRw7aWbpIF8 15jR0tnMznc9U0vSanpLJ+4lNMmqzi15pAqV/rTLvpFukhbZlAuFuKcNluRL0x4fGcpEN76KevIk pkl/uiW/2IF9QQVNhTmifIRrIh/utysuhdUwbfp3x+TCFZaPzJ/NhVKLfZZ09Wnytk/vSITOBJsB Dod7AIM1woiE+iPMfrs8gDVqpYfn/CX1/eVqNR2rqTGblBrgaLgbgJ3fOYBeGqo9AQ/TVbQmEoe8 CtOsNSoJOFmCNBWoogOSQo//OK7MG2u7OOiU4qF03jnTfc0Af01kqqRTVL+elqI13fhGXiy4oeBK /KgjTu7bEprk5vyl5szTzUKsfujkGaqtLslO2w42Zbp0pATLGyS7oYBO5N9LP0lh1/f9eFSSrtiV 1yVPFmoQP2qfGY6oDKH3PZwuqhG/aRf/uhWuik9twGX+wkbj5vnkL+9TLxKhuzeN/24BLNJww5Xt 7H3+CAY4oIFS4MlNG/X92mA2aKd6/KjsQDVtN9wsDIflbpqgWQM1vaLNqRM26GCyKw+MCJUnGgQ0 1SmdVdhc9zfYnrwb9kp/eb4jzaartPOdy5JU3O7mX+9B9gXSzBSYsnpQzaIlsDShSe6Kr444Onkl p7P0KfVamZ23Rhw845aso5eJhuE/UrxTNvJfO0DzuW2LOwJvS5KqUSMtQ6RR7JA6wEmVJHQmXaS1 dNJ+bBr3v0ZyK598QFGfZqRZNtnuo6h2jD/DkwGcC3IE0+mwVkZs9JkwhxPXAS74L1MOM2TS1pni bYvUbLmyQwG/oCLW+pQ7DB/QURrvdDpocCTJLTWUx9cM7/s96gV8pT/Uvt55Xaz4Bq+MnNxnkRQW JMk/cu8PLW5KfpfCJOnsgHKp0ra8L2np7ATXLrrjK23Pa8oz2SLdLg4FfKMc9XHDruZWydIOUra5 aYRgOhdr1HvkZknStQK4T5fkWzdYupc+xWN920kbqlyRnrhmk3x+sPzLJHf3yQRU8amb/FOf9/nS J0oI3Th1gre+uqXaMFjrmqgRwPcqceYcwPmA368C2GvXFHnwk/T+fGWn2s0I582rAVkg5S2/TupN BUtIZHh2A43idGhyYJP6qTeY/7JMV3OcB4YEN0/y6kyQFpPk7z+4LtlKmCl8/97ozmt3d1m9ic4e DpZ0Y2jXSQfDohwPvxhlifJyVJvi2u2vfLexkFzf55ve3Umz2MfHx6eDadIyuyjrdU/C9/4t3Qp7 PONnXcvQ1Kja1EE9rv0BsEdPtgMk009DVIhGUun5en+2wiIyerbV76Wz79TgPvoU5ioy/0V79WC8 +kEVy4EC2mZisNa1V3eg6pOAzEkkJw6SE3OLhWSXIct9FkLhIcNaT5ifvHGbqIvNXscVvqttslb3 pcOFyPyFM0CGzbWuLAaYLsOs5vT44ICQ1Ji2nXKbI+nvCpFrLCslaadrf5WHZAetpYSBnmpDxgeP 27R+GFaVtfrrkO7kKhY6zkgRMDyJ5CSS/8nKCK6prWt6bulj6KhpnABSeeeJVusxZQaAT+43SwlA 7typ37bmf1s+t1FUvwm7RrMXtjaToaUZcC0TIaTAlNzQLFQKagkpF1vCNueAVCYAp365kkhOIvkf kZxgLf/QgUa5KTenpPzJSSS/xiT/1/2TT/d5nERyEsmvP8nh+dwfJpGcRPLrT/JvNNBrRPLJqa8D ycYszcn++Vg5Fcj0/ztUb/l0pb6qU1ctrpZEcozWy4iqeF1InlggMZNcbPLx07v3n/YLvZvb1PCg RRdHuUG2z1q6QZqCZas16vbN8sVWA0e1ftNX/3LyrrTduWHfRhG2ulI7rDEhKa2BJ75Khlv1XFls doqSByXJ0hLytstuuHoM750GoMjEXybFjH/K1hegaJNPy5nfbJLDczs/fJ1IHpAmEZOcL0ChjxRy /9yBza4TdXPD5pvab+4eKl2pstF2yt+GXeGcJCnwyt7Ve25LWpE2tbF4ctlS1uBtSdCHANx5QOnb klYbsFcO0KpKaYof6k7bUAX8VgnzMknX80OXMEn+BU3DbMla3LIbFR+y0OysJP2e840m+bAR8/na kNzd/CDxklxUvxkeOykoYpnkCG4HNdByd+h0Hdk07zdN6/tJZU/rmt/7dStvuJgKSp2ybG9ezNdy 5Uo+oxyUNRUiI/SkGOBiOZT+TsDYXceMbIbZ7gc1sWbQCt//41k9fmuUZl1bp4Om6pY79bK2vF0n h7ESMqWF3aGwhlBZm2kQptPfLzyrvxzfZJLHMvu1IrkHfydeku38npgAaqtIQzUwdIOQ0LLQviH0 1nvR3/vfP4Asl0+XB+ZoTOiJZNhvkI9Ni+gsnU0DJbRwqtpCZS03KpFYU8PZnzzrCk30S/BanNgp 7+PWag+NjR5Bk72ky3YM1KfufpYWgPNWa0bQN5TkDzj1WpH8BX8mYj15mzIaNDX2Uk+gikK1PsJf rVV0kpcHQk1VAIdeYX7ugzWGSdrtEukUdEMbTHyswX7+TpBOvwN5w0852FIlNgC4KlUFumiLLUJq qioCTlrxuZ6oDjNU/VNtM8pXN7F/k0nOmzz0tSK5P7sTMcmLNWDEzgG0UNdyGpG99DcBIVO0fvja 3Td6Q131jE7yYqXCW8t3H/XXoxrYHwweoT9TRxytpolb1ZseGqk1gCn0PNBKQ8F+9LQFC/+85wiw X4/NwIcKUzuYej1o1EV/Z8BdG0dpsPYxU17r1fA/YLsIcCil14rkIUawdCIleaKkkJHU1lAvhUm6 +l4VSQp7FF6cmnqqSvUMvUV5BUqPp+QASlp0NYrbWnEty3QjpNAQTTSCnG49ANpqILj6SdJaAM7r pJEMXyoJA/7UVS02nPKXj1LNnSo2XZWPqzB4bzj6c8c3meTzicua/HyShyfqMXmQ1lRxAS+N9pak H1yhzRc10jg10+CYJA/Sh1TSuJnBmmQCOKXCUVO2aA8ttHmIBmgVYPfED/DSZsAudVYZXvbJLfIF WKkrKg7Ut/6CFNCiUarZWFvHqeZv+gDqWyQ/uzeY5G10e93G5N8SMckdjRKRXhpSSxM2SutSkNkO aKZBMUlurzaU1Gjy/KFeRiKWHFEPh5zH/KsOqvbNwMyQQ3sB8x9GKvymMlIAfCn9ATS2nPpCw8Hl L/krA0ARTR+lqnaH9JOaDdJhM5A55NabPCavZMzrRnJiHpNbGeEd5TSxtrrjvVMH0j/pB3ytj2mk p4Kcm6gf2TQb0voF5wVWKy1EVlK4dh8K+0s1e2hnMqyhTgUfhk31JM35sPMh3XO8M11Xz4Z5JhsQ 4lcw/eOwaa13aNaVy8YSjWaOUm3qK1x93Y5rODBQO99kkqcnMiPcc0nuy8FETPKnqmeUsJn2gQaD eaGG/2RZXKbU/dBcNH/adlFJ3+Kq1UA3rQU2WuxIdW287fC5+0BPqaZprW5tDgs0Bux3Livs8F2N fvuuJF0o3FWP7up+eXj3jqSfHK9ftEvZdO6WPPpxqcrBPmkyRYLC61I1PKRSsjeY5CFsfL1I7szh RExyM1UDyKgfK2gGkPru4/yHJKkLdIkIx7O2gpoEPw0G7LZuBb5/BL100+assekGYP5FNXEc7a9z tjSI7p8f0pPxZrJ+s3zpp86Y+54+NTUzQPLG/SrZsUyPQ6VrmXbN7HfQETpJk6GrQg+GamifoNpv Lsk92Pp6kdyEC4mYZI/dRnaJ3lXtO+QCqDkMuwZz1tUBcszPGp1k+/bvRngc2QGVOkDO5RG+QenT AyRvYAacMkWNMXH+P+5Baacd3TG4REScidOK+1WArk8U2Mf0x6P6bzLJP79eJJd0CknEJCfalqxA Mkib7g3WLtoYyS9fH5Kz5FQSyUn+yTFbC/a9ViRfSVxu+P8+ySmLlS9RysvLK3fmt3J7ej6tR5iT JZH8jPbRa0byz/Fb9STeSS4WPQ9f+P371y6cOXn+wl+HfXceOnH9YUi5JJJjb58lKvPs80n+hmWJ n2THjzz+KWV9n3fFiTbkx//wRVpbFEkBt2eIM+cp4PAfIbkjO18rkutxOfGS/P4GY13jc30dtbTI 4qhcO7Z+OtopHak/HTDo887DPwdaS1LoqZ1Tun1Up8mEWK5oLSXSIETSrVT2zces3HE+NCI5i13U IiMlTf3vS4988v0nSO7GlteJZEu6bEq8JM9QIYC8AY/eGbRlZg7wrgL00WCAglvzANRUNzIsP7+3 Ldhbs3s3vGPIuACUkqZHOHaWUcjlo9eCdPOhJPmPCpaszLrdDmyUcdj35m6S5P/HktLtD6xMj3la sO+YNNBkmQtw73x9/Tnxu8u6m5Oiw3674PNmkzyaFa8TyQeiVZhMbCS31TozOB5Q18OSHpZk189A lrBLAOU1AqCr2tr9psfh+rTRrUyA6VfLfG2uW+eT9o3yAM7SnooVvGm889jJoW1LOwMN5FQ0XFJP ZknWFe+2GoZR56FdrhRgWqjg8FUMV7h07S3WqjG46sgueYL9MC1ZaVHYtUVvNsmTmfE6kfwlvyZi ks2+6g5TtWCWZuS8rF+5uwlgnzyAtzQFYKAavqdtdnnD95XWMKC2Vs2M6gPnL0m7GKVHOcDx7Uat mi96ACsk5XB/LKkHACNlJJMdLW+jgvu+dL025Ak9mrVOT8tph9/1I3jId2u4C1BA0vaGad907WIe 419c3qMrCUyyJUfqkERMMnkCH9VspGP5Qi67YL75xNGIWPJVciCXFhghHZW+1qfwm7L7XzXDT3rv O207ecGWMPymJPkySuepcyjgifEiIvMDhdi/I0nfAjBWTUr2HlmSIVp14GjhFA/upAMmqRWwRE2u KywHWeT7vTIC70na3i/bm07ywpepttcoa2jCkryP3krMJPPRY/+7AQU+Vy9gnYpqHcCNABOQWisz lGkx+LTybVAB1wzj9MliVcQj/JTpKynobm0rZn9J0k+M0lHSpIEOkjaQiXqWU9SSZC0tUl0WSfp4 iCS/oh30JcClB65AI40JDdYyPLV2tooDX2neH1JI+zec5LXPt/so8MhPc4Z2H3Tme1slA8veLRH5 s3cM7TpyR3jMkx7O/O7CC9yiX+npL0Fyc+MP/bwWmmAkM0jqzxKVBObLSxuB5DoCkMN2TmprtZ/V NTWDTzWCTzQ00ubwuyT5MEG/kr64d6URktYwoBLfrqWPJI21xqsentWunZZNV+vksE5vARm01fDn XKGDO9TcUwtHqArYHQ9JR7aON0Lzvtkk//K8GqFXG2e0OmfN9Eh1VZJ0+12w8dcZgK8ky5Si/a9G 5odfnRbsvo4Us7umZ5nu352QfpndOH/2mtOtqehC3nv6nfD/SD5vX/kFvvWV2g75YsRf3558XpLk v2d0i3mRu//aNn+X9TEcWvuZ9r3Zh16Y5OI64chP8gA2qYS2A8WMQKWSOr9mepdqf8ruir+v7/qJ 9/xTPLnjNFfVqK6vIl/9W4yBd4HWMMkqexv9rqd3+oI5irBdGLqM1vygQsB+OQGlNR+guZbJN8/D wGZaNUgfQhsjVLWPOr3ZJJ97XvTTUYeM7k7DZ6w/cORYcp+TacZLIe+Qzuafv538Ow6saP2H1Aeg wCXrWUvNLsMmZaspaWaOQcct2uCUq3pWoNJZcMqeGYosC5cU0pDKoS9Ochd+ef6XvpydvOYCYdLt EZGm8kdvk9nv+NSmuU1AQ9veOx8AFDcYr045XRrnJ0n7mudrbatbcSBEwc9+2mOQnE7bYZPSg/OD AHftB9oY9FXUSKPOY+q7xw1kc81R273KSPnI3PTgK0n92azVRoluST/TUcsws12y/s7kzQtU0qw1 ygH8qNxAXq0BXA6oijZRKyxAP3RXS8o9elIAYI4avdkk36fic4SEqT5+hlKhMaQ4pt406GTzoGvM 98bGQVNqJ6dM1DE+7XVw3yMFhlkN+UWvvY+/dGJOXvJS86F0611YJ9304t1HL64nX3Wu8vzvHPgO wzWIU7qUFVPvH4MfFal1VWpLKlaVgNTeny/6O9jaNaikqVHbos4UeLL/3atSBQ6vY7lk6WVHFtr+ 4OYvhXxhGqRQ+yovTLKHtsN4jctWYb1mcOeBmeR/GfaF9wxe16jA8esAC1S8QPCtJzehovol827d b8rwNGAsuQ5mtzaTr/ZnvQeMGrVxHZ0VntY6GRxpFBzxrwAr1WSLMoJTD/l+kKdytauBuSjmqxVo G7QM05zPNKFfQFj99ivMxaZazpWbnOFNJjmI8s8V05O91q3ekHyGg+fDEmarduDFnYc/zVwepHps DDv9OIfJYP491hrHb7oXbulaj7dLZwySHo93SN6OEZL2JXeijOVGfhoHvMSMr53pBUL4JvGZdMb9 kqWkUbJ0JxR++LNd0Rb4pjZFT5Uxl8+ka+tbMnMgN6UhzN3CaGkYOQ/o48LjOKK73uQ+ozsvMSbj fwnSXZOk35MxSzMHXtAPRnUzY4FknBpskyswXYWZIm2EarppPFz1gQOSNIRfdNPKHUUWMlmqjHGp pdbYlPBNB3TSfpl2bLlmMaZ/mqegW9I6R+4vBfpaqo2U9LAeSyTpZLH7evdNJjnUMe9zxQxjnXWr Hj0cYbI8bK6VDUgPUPK+c/pwSfWtIfzpkk1bubD9h6s1mJ8UKG94V1sbOpF6y2bG6e5wF9NSbxYU oaPlJWwXR80vkPcr1NP1lvFaoFTlz4o0Ii/2vJOfzXnNj7Ok2ulzPErfqsb88Sh1K7hLmsuYnfTS YXP2m9LhQYPYdi4Ptf2k888ueRyT5HOBdpBprM+C5s6Q+bwUNsFwey+rSQBVNXtsaHpglkqQ9o76 QOZAy/F5XRt4f2wH3JGkIwPuSsGnjx3evWnj6uPX5vlL25sZuv9ya8DgSVk25sH7ifTg90Of1Z+1 dEIlU+fTV3d+aoYyeY18ockHzeueDgot3TqvkVOK3X6ub3ReuEyZnytmCj7WrXzcW5K+hUVp3rLu OF4ka6me0yvyCc0kqSlHJUktrH/WoUUyhEtaDJ7rIVPvG/qOL73scV+u404mWltewgpnKet09vlf 2ZfGtqnoIknhpeGTbiZq3DUVlStgH0WGRwZJUniy/OnzSFrGAF/6qYpRGlsT6ZveNMIi6Vd6vjjJ m0Oi1nJ0/6i57YhzsxwA5i5vpc4CkL6hHRTv7QKkily2sPv/X8+i3baemdMDuHumfnFv+zc7V6dn 8ueKmcwa64BnzmJsFHMLitphN46MlKSK1tCk0JVLxo3/+YiTg1GW9wy4DaSTJI2hEXZf3jMUvnUv Y09eaVihntO+sllVSnFLkr6B7qpud+AAbYNx/2JslEjycEobG8lzOL0taS4jVzPwBgWNvQsAw4Vn RfRS2P+f5IypXsnj8rlxBOeSvDqf0UoQ9jwx39hIfkhxY6MTRg3gsPtWyNIySdJjt0xRhQWaoKsk /QmcS0/vMOkrNr1PseuSgkqw5yVIDsmZ8e4LfOVWtuXsjB6yLol11+0jWsXgULyiKyKUl7Ty86t4 4yXpCxbOY8wKm2fHGiDzIUmab9i/XozkV2z2S7Yv27hxy+7de44ePXr0zCW/q38fPXr08M6Ny9du 3b1548/7hieR/IzmbUyJ/18bxQZj44HN0PGzdY1kaIonklSbdvSQNMo6bIZvDzDGsreMvDC9MHPy VF7aS19wKLApua9L6h5rCOGzSJ7Byhf5yp+zy7r+YORZbg+tJGkW40Jtz6GtOeSQVMnlS3qbC0jy 5uhMpk+wYbuGTm1NaY5KOpF8RbyRnBT99I9JbsSJ54mZzdfrxn8z6TvdtJFseYcFxrt8bKCC+5Lj bvKUfwfOc0hzW5J0gRmSLmc3rTC9ZVHA1+Z0vdmgO/nxUT+OyNKT4mFSdzbc8v7+BUkOzl7shcry fsNYY6MIDyTdS2OiriQtZqicc0Tv+7a9nzQWTEfzON/VclMhzWD6RCYbR1cxSQtMnnckBSuJ5MRP csvnZEI5NrZNEWv8gR68Y1NV/07mekvSBXecPd3J8bdGY7bHdYfVtJei9JHT89IyRHWo3DQTaQ5s YYp03FzA0o9fJTVhk/QtSwdS7wVJnmOz6z1PjbTL4SfJN0dGvpWCa/MxJY11so7K4xT9YejPt1Jo NTI/mkzOSuZke7SYkVv4SJJ0Zg6zpCF4B0hXNgQmkZzoSe7wnHWzduDwdrPRSzcujZbiZd8SSdIf nbw983a4J8nn/bJtT9qO9gVwmSz5eUPKjtcUOOaSpMZsn8AiScuZLB0qdvjaUL8XI9mSL98LVkrv TbbhE5s6OC5xduk/siglniTLIkm3KasPWfZd789bdbDFT12xT/u71riYeDegg5nC+6TdNAxM47Q6 NOxEa1N+lkph9fA6c8rtmSnHkkhOPCR3N9ZHn9meHDns/9IXDl/cpsFgwwX01sVI96Kddt/fHvJI 0o0eZ551auwk74hYun1eC+tjD+TYovUpgHoPlS2/JKmC3fGImjMRRhkTKXDZ2JSJenTdIik0Z2qt ccDZEYp+zUJJwS0wDcfupySSEz3JXeIzdcuj53eJneQa7o9e+CIXVi3eFy7pjs+S45J2HZMkXVwY 2qN6+7m+B44evR/R9ccahT49oZAZEc/q+RPSXx28322xNPRYYcMK8tNg/3kHkvTk14HkxBWFESvJ J019E+cvnERy4iG5p+FilbhJbu98I4nkJJL/P8ndXsRT8v+2oDgn+a5rWyWRnETy/yf581dN3bK5 SpyTPClR5UxOIjlxktzLtij2T9vHDeKc5JKFX03m/GVJJP8X7MmvWEupccu4JvkCg15NZr26CUdy 3szglSaJ5LgnueGrZtauXjeuSZ7AsVeTWapWwpF84gpp1TuJ5LgnuTzhr3aREs3imuQPs7+izEz1 E4zk9JZ9jp9oaa4kkuOc5KJur3iRXB/HMcmW1K1fTaSfuUsCkVxl1zmFS9KYJJLjnORcGV/1j9kq jkk+82z34BdrS1gUryRn7//DqJTgtrtpo4thunxM13umSSI5zkn2yPuKF/GIa5KXPd/v9P+32m4P 4pFkh2Ehkg5ntxus0djfu+vQJVoO2iSS44jkIMq94kVSNYljkoc4hL2SxCNxp1zERvI8netYbL7C /XQ9NxV0dslNPTieRHKck3w6Mo/JPx2T49p20TT/KwkMLeV6OR5Jdrp3PAWYOx0+vzS3+wI/SQq6 6mtKIjmuSV7Hq/rmZKoTxySXebULdOYrxeeYnCF9xGaah+FB7derSdLKSDyQPJK5r3iRbBXjmOSc 7V5BnH9b6gTHK8lRW4awrVx/6JhEcjyQ3ORVp1PyjGuSM/f5p7Ku+g5KQ4O4zNz5HJIHqENyHYjt iJtDEsn/LslvO7zqiOVZLo5J9hz+jwRdHp4f7EqvtCjhSE7d0sE8+aMoO+qOGjb862Fjpi72KZ5E 8r9K8h1771e9iId3XJM84h+IudLczlx1yOqrcfwLv6wHUScfW0si+d8leSVDX/EawdSPa5K/fHkp OzycvrwaD7/wy5Kcv2aVUrlyFyxYqGDqJJL/VZIHxpoH6GXaDdrFMclFX97N/leXAmfi5RdO8upM LCSXzhH2itf4g35xbU9+72Vl3PfMdUtJJP+XSH5ifuWwopXMimOSh2V9WRmdXE4rieT/FMl7Wf+q 1xjCT3FM8s+cfTkR+80TlETyf4vk/UWDXp28S3FMckjKqS8lIbxEsdAkkv97vnCv2nJkVxyTrCYf vpSEtf/yWyKJ5P8Eyb/TNs5JXpX8pcbYGuWVACQXvPQvtJOuEa5GhS/FVcufRHJsrf+/nfglFpLD yl9/CQE3HZYlBMlF/xV5jSMGzrh7ixZ+EZKDL/y2cdaMG68byeFn/qliGZ7RIzTOSX65tt3JP5GQ fO/33478tnHH8ulfD/98+ovJ+zvS2+huvJMcdP3X2YP6fPRh9RI5rLV1Pk/UJH9bpETFiu81b9Ww WauPPqhUJKtn9uxOvP0Pr7DpX14X+Yckh++NpPeqt1/Cknx60ajOzWsUSWMXTTl9wQyNXSJOOBiv JPt+UbOAXQyNum+iJjl9bJMAh4B/doXqT5W4SyCSP8Wl8eagBPiFYyH5jlPEr2qXu0qTFh0/Llkq O24vGMF+3t528sL4JNnf+i5IVfy9hh069h40bfnPf94uSuvETHKgHTX7dfy8Y/O6DTp2/GLsvJXb dh9bBCf/0QW20kiJgOTwzAD25RdfSwwkQ41eo+buuq8HjyPegzi+qMT6kR6h8UiyxY3Oc3efjV5y piZVEjPJT+xjhirvhkP/6AJejicSA8lbcF/bvihgV3na6YQmWalZ9XS3RRD4wvqatX0Qr9pF+Vje AV3Il6i1i4yMiIXko/9E/ko6Kz5JfparcVWaStr0DgDlht9KWJLLGLUJo7ZfrbVPX6CFZbKSnC9e Se5B7xgdB5MlUZOcz6iRJ0l6fPnsod27Dy+GaRs33ntZ8dfTJL8ZbyQ/mt84nbt3l+/2hETff+io 5RjGonv4vj65AUyVZl5OQJKb0FrS+YO+vr6+O7evXbNmz5EJ4LN+3vgnLyJymJVkl3gleUmU+uW3 jvmunT9/zjAvnD//uGVwYiU5sCiFWjWsW6NUvuzpnKPN+rq9rPj6TFU8kbyimqvtNot+Ni8qzIVx yopHhGfGgR7FzAC5ZyYQyeFHK1GwU+M8sc2rX8jQvdfW2y8+Sb7v5Pj7Vp8JAz+qWvSpIq8nEifJ H7s+exGz+0tKn0KV8DgjOehhlH3BLQFyRNxptSiWFt+KzoC57vjfbIOH/5axnyTDI2FI9o3xA0cY I9zyxFxECl7bIpuT8zt9LkbuCs1i7X8svki+Nm+gd62n7tslZdpsGSGf90d3EifJVQHSl6/ZoGXv LweOXLhqx+69J67d/MmZ2g8exiriyeLPvIvk+6DLqhg2rrnmHHHxJa0k/5EnytStNxQvmHZwHjCX cgH4NOodpsfTDOBS6esDNm/r+U/XQ40vkn8w20OZ7mOXHrx+4fr9R5IUcD9Qo2FHSMzTl2azZcno Y3sOf+lVwbpvTXyRHLmw6FiweqshMzcdvRAoSQEusSjPEe3y+mlzfwtIMJJ1+4Aplpfc15hi9Wiz TPKwfUmPgdG5nWTKcFxxR7JW5opYot5vwj5PPt6+WwJch0+rZMYUpbTqUtxuhfzav5IbgEfTudcl qTrDE4ZkhU4nWczZQ1AGYiaCDGoFFB04ffHwvFDUOgTXjsBqfHyRPCiNZxHou+vK0/aVBlR+hoy7 w403pFvtVWEJRLIWEsv68l2IrTTznepg59158Nc9qjhCsolRbAdzTZn+VlySrNFFbFPQD8jTDzOc flgUM+3Dzr9F6shMb3WtjIQeGVvBEXDoekon/6lZ8V+Y8ZWmTSwdOxMzk2RLeMuoZB861hXHxdap V1wblGPTk/vhHMsi5AhSxy5iRSqANM4A+X9JIJL7xWrudqDftPwpyn62IOpjGfAO1LfGw90cnh4a R8xjZ5syxVGgXOSMr2tZYwn6iol5f9sB9S13itplMNcLPmgXmVcoyIXlkXe8vjqA2QOX0AQi+bIp 1tDIufDthO92RFXhxkLNCHxOemM2BpOsERkF4pHkslSPpeeC2L0/wrqAudX2e7KcmV/FhLlbSIKQ 3DqakmmbuUZmru7pF6Uro6KYwTqaIszHP9vH0YgcleSw1mX8JOk7kvmrE0Bv+RXF067yoyJ4BERO 9aPZj0+Negsgy5zbCUPyIlLGNg9eavUKqLE5wkZholaUN3N4LzzuS9IE258idzySnDWmDVzSNvhT D3du3nsy6leyNINSEU4Kh8tAeb+EIPn9KObkiLY9yqzVY4F15zqeUvh3Jjc1uCBJt7Kk+EtxTrIs HXL/LqkNdaRzDgB17jz8AGiQB1Zb+8+L6f94tQ6Ae3iCkNyVqrF1HBuZHfy8sacG7tH06fDS1Fh8 XbprBuwKJ4eg+CPZTGzOekfgo+L2AOnHRY67o+HDKLcWNtqRFglBcs3YSB4UmwdUIYo99drY6EAX SWpi3qi4Jrlt91X3tLbEI6kynSV1BCDj1Gv5geQwzNp/AAVjyJgFVZy9E2ZMrh67K2SURETpVt6S dNmOp/J4HAIc2oSoNcDOx6MGhccfyY58G0vPw1GgeM+mGB1z5L3oXOx0oI8l/kluE5vTT+loJJuW S9JPsO/pfl3wlKSpqxXnJJ/uX8Dx/XnhkvIxSdKjTBEeZm0LQERuglbEDBHZAOdvP0kYkovE+p62 ZI/4cTHhvlcah/3T5qLudsAgHXvazhgPJOeO9flbY73nrJXcIIfV7d6LDDdjYMG8hJjxvRuj30kT rlGdU8tKUtNYcoRvhCNxzEYU7eJQvzwPJHkyTZK2mG2qfE4AW/h0U2IWdloP8RTtEJPkwlHnFhHt N8g5d34jp2o5cS4J2QP1HjEzQ/b79AMKSh8AqYLjleRGlIil55RoA1z5UEnaEouZK6BA3Pk7/T8r XLoY/QaCJ2B6p8si34WA/V0pPAUxY+0DnBgTfyRbWzYmSpLmmKP+rt9HmLJikjwGh6CEIrkgsaWx a41rx7Tlek1tb00H98sDx1g1005kkX4F/uUsIs8jeRrmWKxwnxqrNu/17F/O3rbY/iFFY2oSHckb /yQfgacdHx+mBMgz5Kyk+x/jAFuk/cRWcrIUjeOd5AK29/VPeY1ajV07RnEHaBxLXrrqvKeEIrk6 TWN2O2ZPq4EtDA0jZ2Zgxmq4GIu8bnhKqgDkssQnyVdNsT06OQD7Pvf9d4xpmRtoIOmGObaOQ/CM f5KVnbFPdRuOnXePLX8e2f3juGYpAZgmTcc5LLantEK8k1yZ9rbp/S/jv111Tgr2InPkaFf6aRFn 7ZmcYCR3j214qm5T8UuNmJgPB1f6diWXYiU5j81kZ0Q/HQuJF5JViYIxJpgHwBUKlLG9C1Mb+uXl xEJyd7JH/3VuuVN9R/vI93aawgyVesTqIFuLSvFFsu+MqX9bX11Pz+n8k0XaB78ig24viqa9tyHl /Whz65XxSPIGTDFSJm01QGi97Pyi8pBsdTHavRt7itNO5NbNFn1TAs2lkwOyMiR+SN4OM2NXLqK0 +9IoUsYism9EDo0bE+7GG8mHTdEnmmE1ot9uxQst6SJ1jGVmKOUmso7S5cmP45Lkv07sMPLGLrGu pEYGnU6INCcrPYxJwWdRJJx2jLbQu/kd7PbHH8mBKWKoFzczAaT8rLW3G5jqnFVueuYl1oJUVfHW PuPvYL4pdyBjaLyQrPdxe0rrPOEA0esInpQ6Els6hI+tS4RXuzrEjfoZe5aABqS5Fn26lyNttrfz pnMBU6HeB6QmdJJ6USYWfQ82WDcvtDHHQcRIbNrFRZgn/V3fzjZFCs0De63bD1KBa7RQIUtV3KKs +m0GqGKJN5I1AlP0ZHuh7+FYzhb5W+83KcieKelidXF64MRg6UOsKt5MM7Ayfkg+60rOaD5i4d7G fM+WSyYXHJOa4xWLyDzWEPscgN2xeCP5hCulIx1w5tnjZSjEgdfuGjP+snwjjYjNw7cTyWwjowfg eD5OSfY3tKCeUCgsNAsRqQy+JNISNNBqt48EuXfkqokk3asDsDb+SA7Ki8vWqH0+g1E6N+MT70qf TPaTpD9gl0esasMPsE+67AJu4CXNioun8Bn5LmZD0SgjnKU3eE9Z9lvA4bHtPny/cd/Nl+CM1DK2 AW4f1tDFWSkgVg+OuCFZ35l4yxpLHTrCDtPTKrwHq6Tvn/JnkKSLzpEG9AluEBdvkkiSf8000Gpn g2GngSJGj+9NUcIvBjzlN3a8ItSIPg8YD6S9EW8k63hyXGZF4BfUBZpHh3EyTgHZY1tq1UdkM+Z9 dL2wYL8UkhHYEj8kazjkPBfxqTcQzcttHfhLw2Mx4qoOGaxWzysZ4sYd9Vk5iCaYcOnxe6Dl2vel IcZq73U4Il2JJVinDskiiTiTEuKgIHQEyT+4Vjcu1ggwtQCrMrPDBYgoNjkOikARKyoBY5yg3lNu tpZixvwpvkjWL2mhqtVV0PctKPOUP3olqqoKDWJKO+9IT0m64WazbIyBWF/ncUGyxkJ668vrYTMg T7Sjg0gnaRvEcLk5YGK0bXsW4Hoyvkh+0KE0gMkal/P0wugy3IIlvU2zpw5shnFRPo4B0t2MK5J9 zP2tVsBvwQmg1FVJtyc6AZGFXX+GM2Wg5z3J8mfPNJBqWoy38X4z8EP8kay+gEOreb7bvykBNH9q lea0HfP0BVnDYzHWOZ+z8mA6ZShYqYEd8UPywzYNnaHy1L1/7RqaCTCeqohWghqS/F3p9ZS8sHdJ FzH3DysPlLHEE8lN4K227hEz0lYxfs/akjQKl+jqhX82ikf99YPfBmrEFcmNZkfscmaMN1BleO+m Bazzj44RdgE4NApwKlgqLWDXMrZkiD2B9DfjjeQpkDtd5Ix/+VPntMP9kf62i5kVYwn0t75GPLEW IZwElA+PF5IbwKoSUU0VP0c9eslsmOkaku0pa8qgaF/xtLuxHhEfJP8IvKegAwsXb0gJkDb62/h7 60rwLbenbBPNsNsbbcdvDsCSOLddhDswKnRoZIRnqqhx4Bfg6ChSGMZ7jx6xv9eCCoLxeMYHyeeS wRD/iUXswb0Y0OGpIdmJXpKqUvgp14qHGchu00MG2+IQHycHNsYHyb7AivBlVR2B9GmBDNGIHYzz HUna8zSna0zRTeOzAKfT8UFyeF7AyfjJhpbh6SfovAd5ja/QC9O2KAdGEWOSMhyias5xRLI8GCLZ k9HTFexoXjhq6r2LsGsg3tend3Hh2fGT201gWhs/JFu8rVPhwEvXw+oAyaLNqANKkuqmpFOpn7LD WZpChBv+kywmq6LdC/COD5KrAyMkhVw6f+2eG2B/KspRP3dbkrj3SRXVsX69G9miWe9CiwCVQuOB 5FUA/GpV8h2AdFHu7EIeHK3OnA+ykTwikMcy0ozX0245QbmAJnFOcgqmKQgWbzFDJhZ3zB1Fgw82 s/hrikgjgQX/V2rmgHgheTmAs+HNuwGAUlHS4AY3gKWSpJHYb4oKck+iBgB+Y3ssr7gBB+Ke5KNA RFb3yfCU5tgZlysR06OGkfuXu+D+VMjkLtO/7wAVG8nhhQCYKylQCn8b4N2Ixd0z2aMkiZwOLmMN 1cO/JeSPmaVqI0SulcQhySN1H2amA5atvifLqShMZmPIZPJop/n/Jmq9kzxufH5jWeMz4vB8JEln 2yUHKB4xaIXUhh5WAxIk2xn5Z+kM+aJ8rTspc0X5QerHPcmdo+aiq2MocpEOOqttlAc3jmr0DOlp InOM2N+OQMpLcU6yD+QD+kvD7T84dd8lOUAhq6vAMndMg209H5UCyDnx1yPbBqSHIrGpEQ2ADLfi mORM9NN9yAJw4en+Zei3lmQPc0EsBvDINgJg8D++q49ylWzayLtwUS+v97y8vN9Z82ySp+OcMVI5 fmIo8Dmtz9jR0vCBYe3eYQ84zbBO5s54Q6Gob3PNtkWHHzCD6R/PRrYWzv9BixrFC5f2qlTBy7ts vdBnkBzgDpDT+un39Ia7egfrI/hjcnIZo10rIAPUuyIp3PddKHQuxjVveQD5/rH/xZpCBWu2eL9Y 4TJelct7eZdp9iySy1G2HlBDs93+1953h0V1dV+vKQxVKYoFFcXEGjHYG2qMxhqV2BWxxl7e2Gvs FRvGHkti7LHGJPZYYnujRmPBjmhEbFgQBCkz+/vj9joX8Ps9r8/D+Qtm7ty758665+yz99prA16H 1jK7Vbfhp2JOLw4B8gjJ0W5A09Z8ocYg1RqMG24AQrLNv1hbPrhVeOOQ4FqhDUND69cYoI7kQEyj hywZQcE7roKx14CGAOCnE/zJDAYAtko1OfrQj8vXbjp42WjV6k0ZlaajJpIzAjHQH96omBjW/G+i N+3yugKAtcehk6cWN7MAQ5jY4t38gFcIUHPHY6K48W5AI6meSOq3bLdiewkAPsz3djw8u2v58o0/ n7tjlCLXVmb5PQ0kb4PFC7BwP/JvYWyPiJGnH1zaFWaC+xluIYHr01aAW4exnf0AdFFrIPAjAExg zX9wesfy5Zt2nI8x6jo3kpn8Qh3JV4Afy8KEQnesAFCDLo0Xf6qGoMay1wR8R6c6+wHI318rmR4J gONQZsb++fPylZt2XYg1anRZqc3mDFUkl8BseiBtWyQOcw5PZ2WsP9e7VCHAYzJR8onZrcq78Fd0 /ajl+D23nYY/V8pu7hpNJG+C+Q+gLVwXAAUyiejZRQvMhfjS9U3s8Q0AVEzpCgDe3gBsi7VCbecA uI4gurt+YA0htmf2q9VrxRmnQvn2wlLDy2jNyfXQsChEKiEzpZ8ryXjqR6wA6pF9JqsoWHyb6lUX AqgbS45ba/tVE/pzm/OH9vn+L6ez3jtv6aWr2tWRPAZ5UgujItANANCSaAuA4aN9AHgIJexEiYUA /EJEjrhYnfDEWACNH5P92sqvqwgmmAt+1u/HC053WI9l/QAaa83JMyiWOUIZSmuAPlQKMKsExsUj Bi5lol+ta+mqpogX2HPaGd1qo6qyD5zUQrI9GG2vA/OBuoBHOhHZawLVE8YVAGCrvThRvC1sQnTw MzMAWNpo6zmNRNGuj8+PKKHaiqDRiI263YR+lX2ghYaffBRY7ydW8Iqe3BVAgaZuAFByPjNXX83L KWE+jmpVs9HowxozVgd4TM84M7SImsmuTUdu1XVHN8g+0EXdu8gohJ7khpFsdWRILFF9AN0oLebM JcnKPRFwXrR3zh5q8Vye8Ue/QmpGe7Qas1NXCSFSRWNRBckfYQid02rhEoH6jjIoHgTgW62J6TXN 3IvpmTN15B3h/uVuzYqp8/KDT2gheQ8sN08DBwCAYXeuAlCb6N0/F+4KU/+7kgAY/tiTHYuXbNOe Kvq+aOCRfreyXnu9UtO0o7dNZcc200ByTQQ9NgmrKxElNgNgTUyPPn6GA979ILjwm1nt8duTykXS r1bQM7ncnFjNj9eUHdtJHck7gAvJQCXA1KwUmhLRQwuAporzxecBYHqlZ/HeW1Tqj6DyGf8tpWOz OWRhnOYZSssOHqqO5GAMcHwJPuQiHZNRegNQB9Cs731Ko2jwHbcyz79w0nExXy8N2eVp8iN/00Jy U4TRFmA9AOAAET3JBzC0IPFg1u6ZTvc+l6jsytlY8l8nhptqbVA/wVub7MhQdSQfBtafEzm3RM+Z ZUgsp5TyKcxdACe9l0/R0si+2HjQicnmehps1WcmlWVEBcktUIX+BQD0o2KYTESToEpSGQQA+q3P e0dQx++6mvdvdGK07Yvf1U9wV35kD3UkV8aAFewRyo3yPPj4AYAncEz9MuNP7t6w+v5SS4nyTruH un8Ta2Sa0ERyggX7aA5gAWoxK1pvAIBst3yZWRw2OfFxKaap/dtar4Jd6jvve1p5u5qzvwfGkNwW pTJ2AiV5WlNKdbibpW170prBtLqlymMpGifSqPXtuDnxJSyNTE5NrqXaM309DCH5phUraDUAKzq9 sGADUVoAAFRRgMwNgF7DlPuN3h7L+2bV4DsBbo2c3+cGh9TO8Z1BJFdFfQ8AMMNd+XMtZz/bBFBd QVbfS+uQeIro4U/duud1bqlHV2WfimcW+VF7NZC8Dj7p1IPx6IELRBeZH1S6LDmqMWc5o4/ka99S +6XxexyHp9TyM9DDN2SJ0jvqKz+otiqSn9iwkJYDg1GMiQSm1IX7QkDM9HZEAEMSbOCLKtXG1DG0 YwJR5r4pbWoZMLnaaqWj3REqDpESyX2Q/98IAMu6ov424DbRT+oSZMxiPlDb5rSgydRztj09fe/Y tlUMGF17Q6ZTJw7d1JEcAsACoLpa28C1AGoCrt01mmKk9Hr02kGvhnvVftXHUFdncwe5duNqxTF/ aSC5MzrQBleY+jREZeAoUSv4lQRwSxmkAsy6O+ITD2joxJQndLAGjhwtZMjy4qtk9zhD0bHuK1Uk T4L7K5qMAgcBtIgnoqGw7F8HFBKYWjQVaJm5AlASn/jxw683vK+Rg/ZUMp/b52fI5FLrZXNTSl61 6U2B5FduaBgIABdnIagzgomoHkoBKCSz6SRzktUaJo9rmU5r8zxNeeHYWMZ6a0seQ0aXlytbv1DE EUarI7khYBoNVGmiJl/5O9B9NhD+pboe22NKekLJ4/I23eGV32iHco/h0iBpPcURSepIdgRg3lQT 0JdJxByjI8AsC2RiTklMWZ8TBcPfCk5KO0EnavrM6ulvMmp58CnJOfYpDpigiuSP0JVoLAIdA6yA 39fnD5kQSQPxaTPhl9hrQeUUqqPq4HHjjH/yczv9EeK/oEMBw+3gq16QnGOL4oBINSRvBABLBBCz GhYfTCW6CXQEFKt2UxS2ArigYfLvpgGOtLabaXf5wBXNDQME9aUc7cWKAzaoI7kF0PsgEFlGQY0l opTKXnEj4XorWM45Y8b0+ZS6uECFA9uLIgujwDJRdDdB4Vz4a2SrTwLFAGAO7QWA/2YE49MzCmLn eCAQ+arqqsq+uJ8Yf4QuNrONutXdnAXDTWHRsrSxdHyvhuS/gFYHbw9DINHTJV6AxQ+eR+lT9PuG XwRjvFDhOcWb9NJPV648IDrzhdukmx1MWTDZ3DFGluKVjm1qSG4LoPCBk8CDA2BmiknwnCtKfrEj 1oTuAKwaiSSHY6VlFdGx2p6zrzTNCkAsvcXhps8U759SR3Iv+Cf9APNNq3pdTSZ1wCDyV9VjI8rc UrTY5pg2yOIovkvuDIjnEQ0kDwSA1sBWJiUZtwbYtQxWD0lh4U1XfBWEvhV1+yR3gmX/o86WnnGr /bJouLU7v12wBynePaCG5KMcpoKqtBs240tfAHBP9cDcWfDl7cHfRDsAWIdrpaGX5tkSG2btE7c4 bxZNtvXn17g05WR+Vg3JBYCOL+gkEPsIgL+d0gqg52goyFbz4DpXNaHGjP71r9+8f7OJbdjTue5Z NNptGE+GfqZMUjxUR/KDvofpUaP+MZo+WsLGZ3azRhBujc+ypFluyPpo8JdW5EJknxTJFQHLunvA 30S+AJJLoB71QLmKPMeM2dJ4nQR2N8JnehnnZ/GOAi2vXq6WDcPzT2OzPAeU711UQ3LGxCpiR9wG AC73gD2/A2zNcRPAO+LgdKYjwzcaaZzMZJ8Oty5WyobJBSMzRBkj6YhRQ3I/cx8H0Z/AFSoIdCfa Cfy3MwJE8hFERI4gtBsMAF3V6yz+bV885qlnj9jjpbNhdPGV7MK9SPleOul1ljyvIh7Kj5caNYVJ i722lkO2hiXsHyKif5VLZVt1JL80wX0b/Q3EEtUCApYCx6k8InqLpUPirQgaBLxcBYTq8cXSUj37 DbVkz/ISUanMWqYY0VpcuEdnGiEoclDLsmxJXPU9QHQyL50V38EKwIJq2762AqZmh9R8jHevXUb0 MWfP5FIr04lEvY6FEacau7ATEZ0F/qR6wFai5gih6mhnlTbWe2dF3gIoUxLIO0EjV/A23jypqyl7 RpffaFffR1lIF8nH1ZuhMCMOOKzy8kX3Pdm8tQBg6p9ENEf5eg91JDvCCpwjugicI5oOlCuI5pRs wqJZcBOiTXctAOC6P2m4CV7j7mt9n+SI7h2zbziCTrLlJrLxQJPVSaOYKtj0k0UAoNMK2N5RFbTm F8YpQQBga7p8eH4AIT8oAmiJYUOb5sDkMueJntuUr7/W4sIR3QZ20DjgPj0zYwkFo/fHsl58kQBQ 8+VwG2CbohalfVl/Zq0cGF35KtF9JcK89ZG8Sq+U+5pa7unZIlspE3Iyih5R2YOIYlMqTPunwBFi RJDNF+kYcEpaj7zjm7omAC6fFARgqr1ClfBxsXT5nBluGpB4WeUMz7SRPJbVdrtoQkS9ejHDUIRo BmzCsmzfz/56o5v6AShcq7nE2btYKIcmW0Ykn1T5Iu+0kfwU2EQ3qwwlWgXzEyqLAfXQUjb9NbIA nv85Fu4J+C+W3+vMMz5lc2a0dXr6b8pXC+kj+UugqyaSL/AuHRERPUwicixtlQ85Hp1iJrrIX5ui h+Q0pk1caaYCfDG80l9ZZFyFGvDl+00iYIZ4Yk6LIaJrMy05tzxg++kQBSxIG8ljWCR3QplMIgpB S6JYs1gb9xmwelE18S+/kYgo9T4R/T3F7Jdjk4N+O6LwBJk9pzqSX/N50s9RjygQoyNQTkn7BeDS 43hHG1Caj0m9iSOi08Mtvjk2utwfuwPlr1XURbLDF+r6luzvABH34KA1756kyfkKak9ZRQxbmnf5 KXkMYJoeklPdMIWIdvnDdIFoACoSleLqRFg6hRvmJO8bG8YFBs3h/LqXGYqwt5s76wC5gKthyxtf l4fhzDpIbsroW8S7MjfSBdOIKFSszfoYOE0U/11NBsz+Ie53iehdCMLT1rVx9da0o5DVsMlhN7rJ XvHTQXIiE1YkemzGGqIATJkNmyw7tBfoMtwVsLRb1xhAaaZmOfkj9MuMauGnbVkx43N19ztyFvin uki+qEdpsgeKM5IvAgCXGWZtW9rdtotIGK3HuOhaWn1fT5siKquJ5JOMKuG7QviKiBqgOVEHTj6J GcO5IMzdKUEMY+TjJ0ImDfWaaRsesC5NREzwXPqpruGe0zd+LP24DpKLI286EQ1B0XQiessohKyB SYj2ngbq24mILk8pAwA17rHfBs3qa5tcYku6iJbgs0I/TJB3/tri0l2VDpKvMkRgohWwJhCZEbkL eKTgW3rFPR2XBzDV+LwQ4LmBWDpMRDlto0v/Yp8u/Fd4WaB+tGhpVEEFPVkbyRPgYoIG54sOAcjD e0HTAUDbTPNcCQ22ZCKd1w9vWAadkvDoNukhOQrAz0RbmdBmJfQiWgibODPdRuBZ27fUrZG00Ks4 G5t87alveeVYonQBnKsoY4T+hrb03kjx+llaB8muwDGiODemei+Wqc9MchfF8M8LzNnEM3srMMVk 8S76Jtd5xGU1weSI3vXXn+0qHJwq3qpW00HyfsDykIioMWoSEbAiRlSfzowZDBf4dWRJAJ9OsCGI iG6Z9Y1umkD0wof/+Y/Ri3D9abnmn2PFAekwXSQ3R8NgFfopM/qJg+IZxfRnqq1ERPaPODtPEVFi PScuxuzNZYT/DuohuR+A6kT1mG4tRTCY6L9SHaGeACTN2G9d10x7SsbnSZJMTTMHEW12EptptK+P VUIg0kByBoBJRDPg/ZqI6AzwJ5NMExQNrwCw8VP0k9PqjFfpaJkq+V6diIhWOlm4vzwQIXyp5jpI /p5lVCS6YiERWbDJ4Stv8zIeQN63RJT+Uzkg8nL73zmmp/bo4SAiGsf9O0yNSy8f7feFCV+rtx6S k1yxaDw8Nbokl0ETQURmp/41l0npg4zsz7OCTizNM2El77xe1kNyU3gADy8wfnuaK2YQZeSXFMOO ADQE1T/RDxOzmqpszZg/w4Cf4syNC93ZkvuzjTaSHwNoQRnFWV3nHWy8ZSdMvBrOLQBy2We7/rpb gd1fsXNLIKMRO8yZyY128QTLXjpIHg0gnMmnxBBRcaygqvKyosEoYGU75jhmshq1b7x0r85G/+NY z78cw8kNd5Z1D9vFt1Abr4fk48C5jfK0Op+rAdYH8tuqZrpXLMkGQlnyfrlUzeySnIN4cjjrLsfq Ibk4wk34sT0C0ojoGBPoaCVB7mhAtYkL/aVvAKc4ecIMEZkjw2nix232vrKiQLg6kmMB+GZGM8kT ojWwpBIRpecXIrRXgErII2VXOWHUc8x0JlJlYlf+1OLOTPZcvJvdZn+jg+QIMDTOHqhMRBkBmEfD ECS9p6NRtTP/m3OkPf2Ln5IcZmH/TXBKlPNZt5HdxM/WQ/JCuKWfA46qInkXcOtrToAqQX+95UNo Vz0B2LgCr7QAp1B2WXSvFQBxM2clku1WRIWgtSsjXzmXmUkXwlUEgLkI8OZ03yRjqP6qkC7h8EZw /y1wvruu/nC1L8CKSKkj+QYAXKIZcznueFFuTsvH8Z7vAQuFRojM6KYfCcyUMHj5+XySAbpZ/Hde YDweTSR/CQvwhKgsxjDJsSr0HSzSmtfpKH5Grs2nP9XxcTx7HQCCUlRf50a3ejLDBoBJNGohuSMq 01sXtqGZfHyNUrQBpnjee9IZfDXpOjOAgjzfb4iBeMtnB6I8IW7Dq0TyNeDgZJhgfsg498D3RNdE tZ5Ei+A5CAFKknmGPg2ZvysvygPAQC5hHGcgXuQ/5c9aYFGojuQzgE2U8ongkHzNlO8hl+YDDnaG i5jR+NZb97K81m98SUD0EFw3Eg+fdawSwOrZqiO5HMJM2E7pHtjNnvT4Gbl+zxQEUm1pn6cEfRoO vwbdLQJAQNwpA0YHRh0pD7YOTwvJZdGHqJKGDnIdRNAVrpFBmH5ShqMwPfHIG7m5qbl4smQBdBq8 rwFYr2VqI3kLcPccUH8M67+bUIGIgsQCSHNh/YepW5cOJ+V604RnrtmWafkF9niwEcNdWrkAg55r Inkz0BtuPFexEl8e8BdfuvUQOPTYWyLG4MS5WCxM3W22TvQVpH8DjZjs2soCjHmpjeSCGB2AcUTR i+wsG3tM2sziUuX02ShAG4Abopd+1r8sb2Qj9Ng2Ko/pAJfQ8DFitEdLEzA3SRvJbz3xA1FbdVns DHfMIYcvuxL5617JN4VfLHYT0RRewSxG7eDAwMDAwKDg4ODg4ODiAp+nnjaSJ8GHyJcVms6wIQy4 RDSKCeVyEQpXqq/SMtCJm8A1+rzrXiuT6FEA7xB2UAFBQGBgYGBgueDg4ODgcsLEaT2uheT1wA0X oW9oPpWOFv8Ch2mphG7mxE3gCoj/cWlMRDF+/GaiiYo3z5hcPjg4ODi4rOCUuv6theRMK5bUFZTG VgMIVtzU7wFK8pL0UPqPvtFcGO+gqT0RXXSpT5qkSLgXCQwMDAz8JDg4ODi4jKfg59/URPIx4DzR UJnOurCpPkLUlOmWGuds/1Omab+RY7+NKuFNRPQE7e7GP46NvX93nQoepJGSzKT4e6c2L5u3qoiX NpLboybRZ6zC2hngqBemEMX3FpVfLoGZDqgwniKc8BL8a3QdNmrCjEFMBrkXTsc9+ff+vXsxKls+ WSfbtPgnN7ZsWrCkL5ZoIXkBrNQGH7HZ3HdQKbiPBg6QvRJChDKEVk5MLliz2/BRE2b2YkLwbS2X 4p4+uB8Te1tlyyfTRHz36PH1zT8tWBqBn7SQfB3YOwmenHzNRI4iJd/dEXWU4MZJga9LQJ2eI0ZN nNmeKTyu6XEj7un9BzGxN1RqSmSaiKlxj6I3rVu4rBV+0UTyfNjeEs2Fq5pUzwy4JhOtYBQydxlN MnqzWNMZV2937DsrMmrRopnjevXo0CmsVdsBQzp0aeIToI3k0hhMNAGuaQxmXd61U7SE3gC8cdRR doouY9Ty5WxNmPYId8zuPDpyUdSiOTMG9YgI69jiq/Ax3Tq3qIb1Wkgeh0J0gJ+Q7gBwkUc8rwAn iXaICnKpkFGTNxEpBXAkY2Dm5M7jIhdFLZozfUCPrmEdv2zdbUxE52aVsEcLyYeBy9eFkE47lIQy dfYDbESbxBwuu6dRo38jpYqTdExNG9l1wryoRYvmTOvXI/yrji1bfD0mPLxJeRzTRHJbNCCiPVKe EDc+RnsiemLFOpK3DdcbcURsIbTWiG6mG3FU04WzYTnRKVayYAhCaAXcZBSsrUAC/SJrSENEXkYN H0xEt2x6R0Sc0HjjTy0kd0QtygziqNd/MJEMhR9/gcheWvCUMw3f6wlEdEGXbj14r8Ybl7WQvALW VPqE36bWRkSwUqTqB1iI3niIWk0kGDY60un2afoGjTfuayHZUQCTiOi5ShsfoutslOULtCa+r5kR FtPPx5xwgNsqnfwvzOKFV4HkGOAwUYYXE7ypil50QVEP+SNMmUSKQr4k4/zHUcdW6WdySilzlgWr sOhUR3JldCSazTbxpCjYzIqWyYeZh/97AeNxhk22TTy+RJ93Vl7phharADDED1UkD0NJor58ZW8x TPoO1tdq3gUNRr63IraGweEx8+R8/Qm8ilJX4OOSACzPtZB8jl33CqsVpa4CHhMRTUZ+O0u6MGRn fr3UuwaG+tYAqm58ronkHUwNV318QURJFiykFBu/V2PHciCVaCNcpHIEGYatKGbKsuH4ZLgFmHxO M3aRD9OJYk0sp2QIqn0kqdli9/xviCjVny/+NT69BWb9XqPKYBMQeUkzdtEWjYn+YEq6iN4A2y4r 2wtFwUJEd82CfNWd/69G1+8DuEZFa8YuhiNfGkMVUelb2IHtQHsEuEY03+g1Z2Qc2Xl2QZbM9FlW ChDvgxVIngJvOxGNQwE70V/Ar0Sfy1PTK2CyEyX5YpR03bEZtKLA2/u79t4pkiXL239jAtzeacaT 45mwYG12uWmMjl0VPSSXAxkMlSEvG7p8axgTmXd2/n4ja6Trnn3B0ZPVkRyCgUT2oqzjcAK4leEv 5c8S0SJmP9RM2JY8MmpARce1nfsv2LJk9NBO4Apg1ZFciW1iOVJUoc9vzPOwDXRS3fCdU9aFEBlO ISJ6F5AFMz+dZpNS4ZRI7sGwdE4BZ4kWwe0N0XIhhs3dXF+GfuEjXQqNlkWuZcuGjQ/rxJrgY1Sq SP4NeE5Ey+DygogyvTA9Et6yGz0NTMzmoQvfza6gQQN2CNEFg8N1WgjA67ypITnJBWuIaDAKZTJZ yTwOGgJfWZ3vbKYV5m7BJco0uOMzn2azbsaH56xSQg5LFcnJLizFaSugUP88yRS/E1EjtCZOn9bp +J2yCoiGoy38llYLyVWYNdmeD98ShaGBxD6en1qUiOiJTcbb6mTMjNpMdjDReI2G50yGdxGqjeQZ DG4fMM1abgK/HwdkGqBDuERuOxR8ZyTvy48mTD7yqYfx1W92SXFiUw3JZxgFgePs1nk4ShEdVVQt D2SorGkFBFWG2sZMYKs1Y4xPyoXnFRGlNlWRvJ/b6F9U7vhpFqwsAf9beGaS3ZjcAgfBNKPKM6YR X4uCMxpIzvBk81rt8Dkle2I6EaXmkXV5GoBPiIioD0cVE1FpDYxj/Pc2OMqu94MzJLdHDfZJ7EJE vwJxyTamAZ843l2LZzr9xJWxGprd/hYqe4yNkHV54AzJ7KyWWZSJXjREK6Jki7wXYAs20DQSnonG CRQAXLi60P5Gja65zAZnSJ7EeJ9ESWYl86Ilr8T4CxCjmvJSGbsEXBnbfC/kA7hntZF8m8P5JPg4 DrHNGavK+nuHsYg4JmPEXTRkSHUSaBCGRp0lHM+gtTaSA9lg1nh4pRJNRAGiFvKEaigvj1AbFRzc VzAwvhAznI2MxvO4ibCbNpJnsCJw3Znez95YRETVZPf6pTvL1Yox8aU+ewzZ0E605hvbikznooxD tJFcgwfrx+JWX5zbEylwbHcwSjnOKTV82P+EsXDBkhL831e0kXyAK/H+Dbi0HLZkJjv6kYyKxXyb jMIyZfOyRiwRcmH1DFnev6+M0KOG5BfACq4u5ABROKoQrYVZWmxWkg/K7uFIiXZD2wyBiVbR0Oo3 TMh2DtBGcigLtpUwvyaKZfzFwTKNw2UwsY1FPuP3fO8Mrdp87YmjmKGo1kQh3zlWE8nvvDCDn8/q Kuv7zgt7pn5EyUYcyNI80TnZiI5S7SjRUdHaSF4IC7PjSHTH/AFs+d6Psq5UX3DdwkbCP0XqJxkY K42SHhgXOVKUC4zQRPJ3wE2u7ncyURM0JboJ7JOxW7j1MDOQaxo02ojJWwzyVpmRd1EtaRJIHckJ VvahvgNsJ/oRuM/4ofektQuNBeYv5+T0MGL0viztX/IvCZamgdSRfFoQVx+r0BWNgkuqUFcUTEQj DeWqwxYeenw/euegIAPP28gJ4rjiTW0kt+Yf/EaoH8qujY+kkHDwnTFvmqVsg6dGdtWmcuM3/xMf eyqysYGtSOlNJRV7GBUkOwL5vF0b1CaqgR5c4l2SgOLnqXlwYSo/77kY8ZMrTtx67dG9E3PrG6iy rrhRLEb5H00kL4U5nvOM+hH1ZBoUJVslqmpXBDptRjG+ZvmsoUm22rTtN+Nijk6rY0D7J3SDWNBu kiaSl8OFowSvBeLl0eTPRXR8awrRqyC811F0t3SLru1dpHnyZU6z4OXJZcmKShoxHBXaeYShvENW CPw+h2nABkkGXNO7OCv84IvhlvGvC5YS0UCpfv182BJ5b8Sda7g84f2abB75vYQ9rO1dVOebIPdF acrMzz52wUwRHXE7Us9k4W+eB9n7/RptnTxPkojX9i76CDz+a7IW90RBIpX184xQ/Fn392impXuU LLyhjeTdgqfzN8Cr1rfCV5LSan/O/zxngqRFgaP5+7zBHy1rJp1ONJE8CvnShVt4dgEsD5gNtKh5 nf1jcXp9IHyZ9G9GvfdpcrmVDaRZNU0kRwM/CHWbj45yMZ2vUUbklvqJlpVoE++wp1Z5n0ZXXSOL 62kjWSR2mSinOz0T74HSPJgag72292Zm8JnJ8pdiNJHcGR9zc2ymD+D2lstQivT2b1vY1YdxmWtJ vk1K3fdmuG1CtHxt6qOF5E8QznvDeTG3DkOeTvEQqPJEhyU+0lUTJ9KX+P5Q4T7jknwHOVILyWOR j5thn5qxpQOKZLKxOdNjMcNF1E6gpsAvel7mvRmdZ+Gf8h3kdC0kvzQhSlQnMEjy5hGJ6FptNpu9 7j2Z6TUn/Y3ixXtaSM7wFmWym/CVxbRLvEvsBlehrHaTvHPDm3LvyfK60TRc/lpfDSTfEVew1EM1 brqrKTSDJQpHMbsk68pR3BNKvCeTG99RyouO0kCyvZgoiFUBbdy4LFO0OFQSLKlniIQbn1SN9X1P RreLp5YKHoQWkg+KCYZtZbPYeOTJFBtre8MWIr6P8dl1ogzFd9b0Lg6KqgRpCfh7fV+0fY82ix/F N25ydfP7Rd6H4T7fZarUoGh5Fwvhlkrieka2ufAIFBZWC2+hxRlzgz25Ittb+d6Hyfm/dxB9K39V y7s4Kq7XGwTA8i/vbQ4V1aqL+2vFmNhQIxHR33neh9EBG9UyLZrexQTks4siVR6SN7+UxO//4bLQ me1ybqb/WvbZMYrkofDJFEcHuY2dw0sQ3O8Bd3EnqM5SoS0iOuaZY8NNYXGq9EUtJNcQ14es590Q OiRaS7bLsqsxItbnr245Ntnc+alqFkILyb3EbVo2A3ywjdoJEYA6KC5hYTQWV6ZusebYaEvfN2yy 0SCSGzFlTXzGRdKzpbxkr5rhxuk9ZeQUyuY+r/msqHRc1UJyqHA/idaKSovr8Nmx++7SpsRHJSlD BsreObT848MalKRwdSQ/hlg/+zJ4zL7z5eP41BalpOrfrVCaf+E3jxyaXJ4VlVBUaAxUR3JKHrHM wi2ItPpmwTONL4ZeKytxMItU49bnFMpcs583ctHJcRpITrCICb6pVsmWLwFiXVSi+vz04ojyyomZ LXm8vrIYnJOfmsWmfSXqR9KP1xNpzeecWCuDhM0WN27WydEqvZwvURlsbE7+UVL+lmFBEQ6i3fkV 44mLaJ/KlZUIkmCXc7TtC1iXoVXLqDEnHxD7cUSBsNwXuRSnOacjQEqMS3bHXNG/ZyrkxOgS2/gH uaHBOfmAqMU9EVUS/CBmM3VRGpF15b36B22yrfJc4RdJOYV0aOX4doolkl66Aq4pwoqdwH2XIfLE jsdLRYXM6oLZNdw26JmEGi8ZGjm+MIno8Fs3VouNOcE9zkxXWasOeymOFEFElLk42/LJbiNEN0De +1AjxzcYhcXbzxCR8lCmG7fClJJLf1F7VvqTe2inZ9uTyzvljXCembI3tXJ8U+AlVjkJ57SG2Gyp h6RKLlbSVudc9pT3fZZK6l7Hyd6+rYHkIeJQ5mqYBSnEO2wtg70y/N/IAjOuWKMsHkienL2QeFNJ qlbe7KCXKpIfWiQz1W4IigYP+KDt58p2h1PgLe7K+npM9kKfbSQNZB/IJp9hqkh2FOKSf1wS1SQ8 DXXRnoiILilKn2mnvBLt+dBsuRimLs9JujuTjCkaSG4n0QqhObClidnpMjn+UmKNFHKc7J1lSPjO keoxKeqvNbyLzHziqEQTNCjJrx7prsz2aD2kHV0Yf0NV7TAhKutrXyN5/+D6RryLYbDGS5h6AP9P Cdarv2biyfViDoFUEuNpZNajtF9elJ01xIh3cU566Z8gDiOOQEl2zQ+Qd9dK81eUdD2YmuWMsKnT TZJn5wx5F+VEIk9MZlXouXZXLMtNTDGfp7R/c8zQLO2grH0VIorpboZ2fH+LUwfPbFgsahVQH+2J KLEoSitEtH6B6YoalCljQ40s3eDSyo7mEwzs+OwBYi+BHpkhCnr2Q4CdIdH4JSrOXk5RyZy+JiRr aSdll6PBRnZ8XZE/VeJwittp72dWzXNM0YB0DEMBRe+41GXls2R0tZOK00YY2/GFShnfKa7Chpp2 wyJz324qtKApZeeQT4zi+Gu1Zlcyj/6IOpKnwiNFFMsy/bsOJs6V+AYlGF66QtiH0n1luR7RuBrZ xmiAq+x2FSWQw7I5Ww3Jf0o1ZObDwypwwHcxqL5rlsjm8nsSF2WXyX9mtzLqZXz6q0ovNBknt60a kp+7Yqo42AMEssULRESvrFhOREOQX9mM6LTaL0COv6Y2N+plVD+k8jutVfPiVKJwT6S/USURmXo8 TyMRkXzDVa51oqsB5z5Io2nbculh69WRXFsgZxM1Rx2KEW7bargRXXIRmoJJWC6+qaQ5XkYZmDFc Wu9RFZbOkCqLlVND8hj4iaap1NL4qqbQmSg1L+YQ0Ri4q8j9xplVZj0iehZZysA+r/1+NREeSpb+ TjXVkDwRnk8kHLJKa2B6Kkr09iJK9VZtQhuiig4iejjVQKrSs8sfqs2OE6TPQRNtpr2UkDw9VuC7 mpfL3/5WovAq8pEuzGugN8H59z9u17jifWmg+aQqkmPNPKWF6K0NC4mK87PtH8Bzagi/Rypnj5YI eaqMx7v76CkCWkJXPtX6aHdpZFENyZXRXfSBBcCBUSKaSBfUI3rpJ2riIhp1UVWjz2vczz30OPjW BusStExuLTmyqxqSAyU9yh97YNstk2iuHYWCdvpVEmPix7w8EzXv871NXfU0BV2bbHqj9VFp3HSQ QSRLxivFM3LR4paodXTGte3jWlRRlgzka7Hgit5VpvXr2Duia69Bozccv3dbXLEkQvJkuAvX3Q7E EA1EEPvSQ+DPJya1KAUR1UWI0+/59OjSbnU+VjBlTZUH707S+djlfl3De3fu0W/o3F+jY0+lqiD5 irTLYSmE0X6BDk7bi/YiWgDbVbWz/4RiKdrXjj+yOLxWSUUg1Fz9m191PkVn+4V37d25R7//zP/9 +r1T6WpInuwjjndcdqmcRh+JJBe2AX/RPyXUUZSmf5//PTS/Y03lvGGrPeKQ3icP9O/SrVfnHv2/ Wbz/ZszJzOwgWWUkOj0i4crBeX1b163auEVoaJMOk/ZEZ2b3WiIkfyZWxJ2Ohkz0mIvDff1NgmPV SNWlif7xCE03drl3MX+tH9OpUZ0aLRqGftay76qTL7NhsxTJi/h6XiKitCC/G5SUX1bWeb28+qLh 2HLX6dVS755dN6pDw9q1Wnwe2qDVgNWnE7Nzn+VROOmilkDUQeQ1JE3YmUGUkZF9CKXcOv39yHaf 1wpt0SC0YashP5xNysZJco7k/8shQnJdcc21fesDInJ829fQWdL/T22WIvlGo3HiN5MeEdHtI/9r 91ldP1nszpxJ+V+z+YNF8vNDmR+GzRpdzP6nh1Mk/w+ODxbJH4zNuUjORXIuknORnIvkXCTnIjkX yblIzkVyLpJzkZyL5Fwk5yI5F8m5SM5Fci6Sc5Gci+RcJOci+X8LyQFdPpThySPZ94Ox2YdHsucH Y3NBHskuH4zNJYCK+HAGV6rQ/AOymav06PYB2cx1QBxt+nBsrvD/ANIP5RKCEJXDAAAAAElFTkSu QmCC --f46d040a62c49bb1c804f027e8cc--""" TEXT_ATTACHMENT_W_NULL = """MIME-Version: 1.0 Received: by 10.76.127.40 with HTTP; Fri, 17 Jan 2014 02:21:43 -0800 (PST) Date: Fri, 17 Jan 2014 12:21:43 +0200 Delivered-To: chirila.s.alexandru@gmail.com Message-ID: <CALTHOsuHFaaatiXJKU=LdDCo4NmD_h49yvG2RDsWw17D0-NXJg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila.s.alexandru@gmail.com> To: Alexandru Chirila <chirila.s.alexandru@gmail.com> Content-Type: multipart/mixed; boundary=f46d040a62c49bb1c804f027e8cc --f46d040a62c49bb1c804f027e8cc Content-Type: multipart/alternative; boundary=f46d040a62c49bb1c404f027e8ca --f46d040a62c49bb1c404f027e8ca Content-Type: text/plain; charset=ISO-8859-1 This is a test ma\x00iling --f46d040a62c49bb1c804f027e8cc--""" TEXT_ATTACHMENT_W_MULTIPLE_NULLS = """MIME-Version: 1.0 Received: by 10.76.127.40 with HTTP; Fri, 17 Jan 2014 02:21:43 -0800 (PST) Date: Fri, 17 Jan 2014 12:21:43 +0200 Delivered-To: chirila.s.alexandru@gmail.com Message-ID: <CALTHOsuHFaaatiXJKU=LdDCo4NmD_h49yvG2RDsWw17D0-NXJg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila.s.alexandru@gmail.com> To: Alexandru Chirila <chirila.s.alexandru@gmail.com> Content-Type: multipart/mixed; boundary=f46d040a62c49bb1c804f027e8cc --f46d040a62c49bb1c804f027e8cc Content-Type: multipart/alternative; boundary=f46d040a62c49bb1c404f027e8ca --f46d040a62c49bb1c404f027e8ca Content-Type: text/plain; charset=ISO-8859-1 This is a test ma\x00\x00\x00iling --f46d040a62c49bb1c804f027e8cc--""" TEXT_ATTACHMENT_W_SUBJECT_NULL = """MIME-Version: 1.0 Received: by 10.76.127.40 with HTTP; Fri, 17 Jan 2014 02:21:43 -0800 (PST) Date: Fri, 17 Jan 2014 12:21:43 +0200 Delivered-To: chirila.s.alexandru@gmail.com Message-ID: <CALTHOsuHFaaatiXJKU=LdDCo4NmD_h49yvG2RDsWw17D0-NXJg@mail.gmail.com> Subject: Te\x00\x00\x00st From: Alexandru Chirila <chirila.s.alexandru@gmail.com> To: Alexandru Chirila <chirila.s.alexandru@gmail.com> Content-Type: multipart/mixed; boundary=f46d040a62c49bb1c804f027e8cc --f46d040a62c49bb1c804f027e8cc Content-Type: multipart/alternative; boundary=f46d040a62c49bb1c404f027e8ca --f46d040a62c49bb1c404f027e8ca Content-Type: text/plain; charset=ISO-8859-1 This is a test mailing --f46d040a62c49bb1c804f027e8cc--""" TEXT_ATTACHMENT_W_CONTENTTYPE_NULL = """MIME-Version: 1.0 Received: by 10.76.127.40 with HTTP; Fri, 17 Jan 2014 02:21:43 -0800 (PST) Date: Fri, 17 Jan 2014 12:21:43 +0200 Delivered-To: chirila.s.alexandru@gmail.com Message-ID: <CALTHOsuHFaaatiXJKU=LdDCo4NmD_h49yvG2RDsWw17D0-NXJg@mail.gmail.com> Subject: Test From: Alexandru Chirila <chirila.s.alexandru@gmail.com> To: Alexandru Chirila <chirila.s.alexandru@gmail.com> Content-Type: multipart/mixed; boundary=f46d040a62c49bb1c804f027e8cc --f46d040a62c49bb1c804f027e8cc Content-Type: multipart/alternative; boundary=f46d040a62c49bb1c404f027e8ca --f46d040a62c49bb1c404f027e8ca Content-Type: text/plain; charset="iso-8859-1\x00\x00\x00" This is a test mailing --f46d040a62c49bb1c804f027e8cc--""" class PyzorPreDigestTest(PyzorTestBase): # we don't need the pyzord server to test this @classmethod def setUpClass(cls): pass @classmethod def tearDownClass(cls): pass def setUp(self): # no argument necessary self.client_args = {} def test_predigest_email(self): """Test email removal in the predigest process""" emails = ["t@abc.ro", "t1@abc.ro", "t+@abc.ro", "t.@abc.ro", ] message = "Test %s Test2" expected = b"TestTest2\n" for email in emails: msg = message % email res = self.check_pyzor("predigest", None, input=TEXT % msg) self.assertEqual(res, expected) def test_predigest_long(self): """Test long "words" removal in the predigest process""" strings = ["0A2D3f%a#S", "3sddkf9jdkd9", "@@#@@@@@@@@@"] message = "Test %s Test2" expected = b"TestTest2\n" for s in strings: msg = message % s res = self.check_pyzor("predigest", None, input=TEXT % msg) self.assertEqual(res, expected) def test_predigest_line_length(self): """Test small lines removal in the predigest process""" msg = "This line is included\n"\ "not this\n"\ "This also" expected = b"Thislineisincluded\nThisalso\n" res = self.check_pyzor("predigest", None, input=TEXT % msg) self.assertEqual(res, expected) def test_predigest_atomic(self): """Test atomic messages (lines <= 4) in the predigest process""" msg = "All this message\nShould be included\nIn the predigest" expected = b"Allthismessage\nShouldbeincluded\nInthepredigest\n" res = self.check_pyzor("predigest", None, input=TEXT % msg) self.assertEqual(res, expected) def test_predigest_pieced(self): """Test pieced messages (lines > 4) in the predigest process""" msg = "" for i in range(100): msg += "Line%d test test test\n" % i expected = b"" for i in [20, 21, 22, 60, 61, 62]: expected += ("Line%dtesttesttest\n" % i).encode("utf8") res = self.check_pyzor("predigest", None, input=TEXT % msg) self.assertEqual(res, expected) def test_predigest_html(self): expected = """Emailspam,alsoknownasjunkemailorbulkemail,isasubset ofspaminvolvingnearlyidenticalmessagessenttonumerous byemail.Clickingonlinksinspamemailmaysendusersto byemail.Clickingonlinksinspamemailmaysendusersto phishingwebsitesorsitesthatarehostingmalware. Emailspam.Emailspam,alsoknownasjunkemailorbulkemail,isasubsetofspaminvolvingnearlyidenticalmessagessenttonumerousbyemail.Clickingonlinksinspamemailmaysenduserstophishingwebsitesorsitesthatarehostingmalware. """.encode("utf8") res = self.check_pyzor("predigest", None, input=HTML_TEXT) self.assertEqual(res, expected) def test_predigest_html_style_script(self): expected = """Thisisatest. Thisisatest. """.encode("utf8") res = self.check_pyzor("predigest", None, input=HTML_TEXT_STYLE_SCRIPT) self.assertEqual(res, expected) def test_predigest_attachemnt(self): expected = b"Thisisatestmailing\n" res = self.check_pyzor("predigest", None, input=TEXT_ATTACHMENT) self.assertEqual(res, expected) class PyzorDigestTest(PyzorTestBase): # we don't need the pyzord server to test this @classmethod def setUpClass(cls): pass @classmethod def tearDownClass(cls): pass def setUp(self): # no argument necessary self.client_args = {} def test_digest_email(self): """Test email removal in the digest process""" emails = ["t@abc.ro", "t1@abc.ro", "t+@abc.ro", "t.@abc.ro", ] message = "Test %s Test2" expected = b"TestTest2" for email in emails: msg = message % email res = self.check_pyzor("digest", None, input=TEXT % msg) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_long(self): """Test long "words" removal in the digest process""" strings = ["0A2D3f%a#S", "3sddkf9jdkd9", "@@#@@@@@@@@@"] message = "Test %s Test2" expected = b"TestTest2" for s in strings: msg = message % s res = self.check_pyzor("digest", None, input=TEXT % msg) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_line_length(self): """Test small lines removal in the digest process""" msg = "This line is included\n"\ "not this\n"\ "This also" expected = b"ThislineisincludedThisalso" res = self.check_pyzor("digest", None, input=TEXT % msg) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_atomic(self): """Test atomic messages (lines <= 4) in the digest process""" msg = "All this message\nShould be included\nIn the digest" expected = b"AllthismessageShouldbeincludedInthedigest" res = self.check_pyzor("digest", None, input=TEXT % msg) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_pieced(self): """Test pieced messages (lines > 4) in the digest process""" msg = "" for i in range(100): msg += "Line%d test test test\n" % i expected = b"" for i in [20, 21, 22, 60, 61, 62]: expected += ("Line%dtesttesttest" % i).encode("utf8") res = self.check_pyzor("digest", None, input=TEXT % msg) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_html(self): expected = """Emailspam,alsoknownasjunkemailorbulkemail,isasubset ofspaminvolvingnearlyidenticalmessagessenttonumerous byemail.Clickingonlinksinspamemailmaysendusersto byemail.Clickingonlinksinspamemailmaysendusersto phishingwebsitesorsitesthatarehostingmalware. Emailspam.Emailspam,alsoknownasjunkemailorbulkemail,isasubsetofspaminvolvingnearlyidenticalmessagessenttonumerousbyemail.Clickingonlinksinspamemailmaysenduserstophishingwebsitesorsitesthatarehostingmalware. """.replace("\n", "").encode("utf8") res = self.check_pyzor("digest", None, input=HTML_TEXT) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_html_style_script(self): expected = """Thisisatest.Thisisatest.""".encode("utf8") res = self.check_pyzor("digest", None, input=HTML_TEXT_STYLE_SCRIPT) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_attachment(self): expected = b"Thisisatestmailing" res = self.check_pyzor("digest", None, input=TEXT_ATTACHMENT) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_attachment_w_null(self): expected = b"Thisisatestmailing" res = self.check_pyzor("digest", None, input=TEXT_ATTACHMENT_W_NULL) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_attachment_w_multiple_nulls(self): expected = b"Thisisatestmailing" res = self.check_pyzor("digest", None, input=TEXT_ATTACHMENT_W_MULTIPLE_NULLS) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_attachment_w_subject_null(self): expected = b"Thisisatestmailing" res = self.check_pyzor("digest", None, input=TEXT_ATTACHMENT_W_SUBJECT_NULL) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") def test_digest_attachment_w_contenttype_null(self): expected = b"Thisisatestmailing" res = self.check_pyzor("digest", None, input=TEXT_ATTACHMENT_W_CONTENTTYPE_NULL) self.assertEqual(res.decode("utf8"), hashlib.sha1(expected).hexdigest().lower() + "\n") ENCODING_TEST_EMAIL = """From nobody Tue Apr 1 13:18:54 2014 Content-Type: multipart/related; boundary="===============0632694142025794937==" MIME-Version: 1.0 This is a multi-part message in MIME format. --===============0632694142025794937== Content-Type: text/plain; charset="iso-8859-1" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Thist is a t=E9st --===============0632694142025794937== MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 VGhpcyBpcyBhIHRlc3Qg5r+A5YWJ6YCZ --===============0632694142025794937== MIME-Version: 1.0 Content-Type: text/plain; charset="cp1258" Content-Transfer-Encoding: base64 VGhpcyBpcyBhIHTpc3Qg4qXG --===============0632694142025794937==-- """ BAD_ENCODING = """From nobody Tue Apr 1 13:18:54 2014 Content-Type: multipart/related; boundary="===============0632694142025794937==" MIME-Version: 1.0\x00\x00\x00 This is a multi-part message in MIME format. --===============0632694142025794937== Content-Type: text/plain; charset=ISO-8859-1\x00\x00\x00Content-Transfer-Encoding: quoted-printable This is a test --===============0632694142025794937\x00\x00\x00== Content-Type: text/plain; charset=us-asciia Content-Transfer-Encoding: quoted-printable This is a test --===============0632694142025794937== """ class PyzorEncodingTest(PyzorTestBase): # we don't need the pyzord server to test this @classmethod def setUpClass(cls): pass @classmethod def tearDownClass(cls): pass def setUp(self): # no argument necessary self.client_args = {} def test_encodings(self): expected = "47a83cd0e5cc9bd2c64c06c00e3853f79e63014f\n" res = self.check_pyzor("digest", None, input=ENCODING_TEST_EMAIL) self.assertEqual(res.decode("utf8"), expected) def test_bad_encoding(self): expected = "2b4dbf2fb521edd21d997f3f04b1c7155ba91fff\n" res = self.check_pyzor("digest", None, input=BAD_ENCODING) self.assertEqual(res.decode("utf8"), expected) def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(PyzorDigestTest)) test_suite.addTest(unittest.makeSuite(PyzorPreDigestTest)) test_suite.addTest(unittest.makeSuite(PyzorEncodingTest)) return test_suite if __name__ == '__main__': unittest.main()
56,344
56.029352
206
py
pyzor
pyzor-master/tests/functional/test_engines/test_redis.py
import unittest try: import redis has_redis = True except ImportError: has_redis = False from tests.util import * @unittest.skipIf(not has_redis, "redis library not available") class RedisPyzorTest(PyzorTest, PyzorTestBase): """Test the redis engine""" dsn = "localhost,,,10" engine = "redis" @classmethod def tearDownClass(cls): super(RedisPyzorTest, cls).tearDownClass() redis.StrictRedis(db=10).flushdb() class ThreadsRedisPyzorTest(RedisPyzorTest): """Test the redis engine with threads activated.""" threads = "True" class MaxThreadsRedisPyzorTest(RedisPyzorTest): """Test the gdbm engine with with maximum threads.""" threads = "True" max_threads = "10" class PreForkRedisPyzorTest(RedisPyzorTest): """Test the redis engine with threads activated.""" prefork = "4" def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(RedisPyzorTest)) test_suite.addTest(unittest.makeSuite(ThreadsRedisPyzorTest)) test_suite.addTest(unittest.makeSuite(MaxThreadsRedisPyzorTest)) test_suite.addTest(unittest.makeSuite(PreForkRedisPyzorTest)) return test_suite if __name__ == '__main__': unittest.main()
1,306
24.627451
68
py
pyzor
pyzor-master/tests/functional/test_engines/test_mysql.py
import unittest try: import configparser as ConfigParser except ImportError: import ConfigParser from tests.util import * try: import MySQLdb has_mysql = True except ImportError: has_mysql = False schema = """ CREATE TABLE IF NOT EXISTS `%s` ( `digest` char(40) default NULL, `r_count` int(11) default NULL, `wl_count` int(11) default NULL, `r_entered` datetime default NULL, `wl_entered` datetime default NULL, `r_updated` datetime default NULL, `wl_updated` datetime default NULL, PRIMARY KEY (`digest`) ) """ @unittest.skipIf(not os.path.exists("./test.conf"), "test.conf is not available") @unittest.skipIf(not has_mysql, "MySQLdb library not available") class MySQLdbPyzorTest(PyzorTest, PyzorTestBase): """Test the mysql engine.""" dsn = None engine = "mysql" @classmethod def setUpClass(cls): conf = ConfigParser.ConfigParser() conf.read("./test.conf") table = conf.get("test", "table") db = MySQLdb.Connect(host=conf.get("test", "host"), user=conf.get("test", "user"), passwd=conf.get("test", "passwd"), db=conf.get("test", "db")) c = db.cursor() c.execute(schema % table) c.close() db.close() cls.dsn = "%s,%s,%s,%s,%s" % (conf.get("test", "host"), conf.get("test", "user"), conf.get("test", "passwd"), conf.get("test", "db"), conf.get("test", "table")) super(MySQLdbPyzorTest, cls).setUpClass() @classmethod def tearDownClass(cls): super(MySQLdbPyzorTest, cls).tearDownClass() try: conf = ConfigParser.ConfigParser() conf.read("./test.conf") table = conf.get("test", "table") db = MySQLdb.Connect(host=conf.get("test", "host"), user=conf.get("test", "user"), passwd=conf.get("test", "passwd"), db=conf.get("test", "db")) c = db.cursor() c.execute("DROP TABLE %s" % table) c.close() db.close() except: pass class ThreadsMySQLdbPyzorTest(MySQLdbPyzorTest): """Test the mysql engine with threads activated.""" threads = "True" max_threads = "0" class BoundedThreadsMySQLdbPyzorTest(MySQLdbPyzorTest): """Test the mysql engine with threads and DBConnections set.""" threads = "True" max_threads = "0" db_connections = "10" class MaxThreadsMySQLdbPyzorTest(MySQLdbPyzorTest): """Test the mysql engine with threads and MaxThreads set.""" threads = "True" max_threads = "10" class BoundedMaxThreadsMySQLdbPyzorTest(MySQLdbPyzorTest): """Test the mysql engine with threads, MaxThreads and DBConnections set.""" threads = "True" max_threads = "10" db_connections = "10" class ProcessesMySQLdbPyzorTest(MySQLdbPyzorTest): processes = "True" max_processes = "10" class PreForkMySQLdbPyzorTest(MySQLdbPyzorTest): prefork = "4" def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(MySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(ThreadsMySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(BoundedThreadsMySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(MaxThreadsMySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(BoundedMaxThreadsMySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(ProcessesMySQLdbPyzorTest)) test_suite.addTest(unittest.makeSuite(PreForkMySQLdbPyzorTest)) return test_suite if __name__ == '__main__': unittest.main()
3,987
31.16129
79
py
pyzor
pyzor-master/tests/functional/test_engines/test_gdbm.py
import unittest from tests.util import * try: import gdbm has_gdbm = True except ImportError: has_gdbm = False @unittest.skipIf(not has_gdbm, "gdbm library not available") class GdbmPyzorTest(PyzorTest, PyzorTestBase): """Test the gdbm engine""" dsn = "pyzord.db" engine = "gdbm" class ThreadsGdbmPyzorTest(GdbmPyzorTest): """Test the gdbm engine with threads activated.""" threads = "True" max_threads = "0" class MaxThreadsGdbmPyzorTest(GdbmPyzorTest): """Test the gdbm engine with with maximum threads.""" threads = "True" max_threads = "10" def suite(): """Gather all the tests from this module in a test suite.""" test_suite = unittest.TestSuite() test_suite.addTest(unittest.makeSuite(GdbmPyzorTest)) test_suite.addTest(unittest.makeSuite(ThreadsGdbmPyzorTest)) test_suite.addTest(unittest.makeSuite(MaxThreadsGdbmPyzorTest)) return test_suite if __name__ == '__main__': unittest.main()
1,009
25.578947
67
py
pyzor
pyzor-master/tests/functional/test_engines/__init__.py
"""A suite of functional tests that verifies the correct behaviour of the pyzor client and server as a whole. Functional test should not touch real data and are usually safe, but it's not recommended to run theses on production servers. Note these tests the installed version of pyzor, not the version from the source. """ import unittest def suite(): """Gather all the tests from this package in a test suite.""" import test_gdbm import test_mysql import test_redis test_suite = unittest.TestSuite() test_suite.addTest(test_gdbm.suite()) test_suite.addTest(test_mysql.suite()) test_suite.addTest(test_redis.suite()) return test_suite if __name__ == '__main__': unittest.main(defaultTest='suite')
745
24.724138
77
py
pyzor
pyzor-master/docs/conf.py
# -*- coding: utf-8 -*- # # Pyzor documentation build configuration file, created by # sphinx-quickstart on Sat Jun 7 15:20:07 2014. # # This file is execfile()d with the current directory set to its # containing dir. # # Note that not all possible configuration values are present in this # autogenerated file. # # All configuration values have a default; values that are commented out # serve to show the default. import sys import os # If extensions (or modules to document with autodoc) are in another directory, # add these directories to sys.path here. If the directory is relative to the # documentation root, use os.path.abspath to make it absolute, like shown here. # sys.path.insert(0, os.path.abspath('.')) # -- General configuration ------------------------------------------------ # If your documentation needs a minimal Sphinx version, state it here. # needs_sphinx = '1.0' # Add any Sphinx extension module names here, as strings. They can be # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom # ones. extensions = [ 'sphinx.ext.autodoc', 'sphinx.ext.todo', 'sphinx.ext.coverage', 'sphinx.ext.mathjax', 'sphinx.ext.ifconfig', 'sphinx.ext.viewcode', ] # Add any paths that contain templates here, relative to this directory. templates_path = ['.templates'] # The suffix of source filenames. source_suffix = '.rst' # The encoding of source files. # source_encoding = 'utf-8-sig' # The master toctree document. master_doc = 'index' # General information about the project. project = u'Pyzor' copyright = u'2014, Frank Tobin' # The version info for the project you're documenting, acts as replacement for # |version| and |release|, also used in various other places throughout the # built documents. # # The short X.Y version. version = '1.0' # The full version, including alpha/beta/rc tags. release = '1.0' # The language for content autogenerated by Sphinx. Refer to documentation # for a list of supported languages. # language = None # There are two options for replacing |today|: either, you set today to some # non-false value, then it is used: #today = '' # Else, today_fmt is used as the format for a strftime call. #today_fmt = '%B %d, %Y' # List of patterns, relative to source directory, that match files and # directories to ignore when looking for source files. exclude_patterns = ['.build'] # The reST default role (used for this markup: `text`) to use for all # documents. #default_role = None # If true, '()' will be appended to :func: etc. cross-reference text. #add_function_parentheses = True # If true, the current module name will be prepended to all description # unit titles (such as .. function::). #add_module_names = True # If true, sectionauthor and moduleauthor directives will be shown in the # output. They are ignored by default. #show_authors = False # The name of the Pygments (syntax highlighting) style to use. pygments_style = 'sphinx' # A list of ignored prefixes for module index sorting. #modindex_common_prefix = [] # If true, keep warnings as "system message" paragraphs in the built documents. #keep_warnings = False # -- Options for HTML output ---------------------------------------------- # The theme to use for HTML and HTML Help pages. See the documentation for # a list of builtin themes. html_theme = 'default' # Theme options are theme-specific and customize the look and feel of a theme # further. For a list of options available for each theme, see the # documentation. #html_theme_options = {} # Add any paths that contain custom themes here, relative to this directory. #html_theme_path = [] # The name for this set of Sphinx documents. If None, it defaults to # "<project> v<release> documentation". #html_title = None # A shorter title for the navigation bar. Default is the same as html_title. #html_short_title = None # The name of an image file (relative to this directory) to place at the top # of the sidebar. #html_logo = None # The name of an image file (within the static path) to use as favicon of the # docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32 # pixels large. #html_favicon = None # Add any paths that contain custom static files (such as style sheets) here, # relative to this directory. They are copied after the builtin static files, # so a file named "default.css" will overwrite the builtin "default.css". html_static_path = ['.static'] # Add any extra paths that contain custom files (such as robots.txt or # .htaccess) here, relative to this directory. These files are copied # directly to the root of the documentation. #html_extra_path = [] # If not '', a 'Last updated on:' timestamp is inserted at every page bottom, # using the given strftime format. #html_last_updated_fmt = '%b %d, %Y' # If true, SmartyPants will be used to convert quotes and dashes to # typographically correct entities. #html_use_smartypants = True # Custom sidebar templates, maps document names to template names. #html_sidebars = {} # Additional templates that should be rendered to pages, maps page names to # template names. #html_additional_pages = {} # If false, no module index is generated. #html_domain_indices = True # If false, no index is generated. #html_use_index = True # If true, the index is split into individual pages for each letter. #html_split_index = False # If true, links to the reST sources are added to the pages. #html_show_sourcelink = True # If true, "Created using Sphinx" is shown in the HTML footer. Default is True. #html_show_sphinx = True # If true, "(C) Copyright ..." is shown in the HTML footer. Default is True. html_show_copyright = False # If true, an OpenSearch description file will be output, and all pages will # contain a <link> tag referring to it. The value of this option must be the # base URL from which the finished HTML is served. #html_use_opensearch = '' # This is the file name suffix for HTML files (e.g. ".xhtml"). #html_file_suffix = None # Output file base name for HTML help builder. htmlhelp_basename = 'Pyzordoc' # -- Options for LaTeX output --------------------------------------------- latex_elements = { # The paper size ('letterpaper' or 'a4paper'). #'papersize': 'letterpaper', # The font size ('10pt', '11pt' or '12pt'). #'pointsize': '10pt', # Additional stuff for the LaTeX preamble. #'preamble': '', } # Grouping the document tree into LaTeX files. List of tuples # (source start file, target name, title, # author, documentclass [howto, manual, or own class]). latex_documents = [ ('index', 'Pyzor.tex', u'Pyzor Documentation', u'Frank Tobin', 'manual'), ] # The name of an image file (relative to this directory) to place at the top of # the title page. #latex_logo = None # For "manual" documents, if this is true, then toplevel headings are parts, # not chapters. #latex_use_parts = False # If true, show page references after internal links. #latex_show_pagerefs = False # If true, show URL addresses after external links. #latex_show_urls = False # Documents to append as an appendix to all manuals. #latex_appendices = [] # If false, no module index is generated. #latex_domain_indices = True # -- Options for manual page output --------------------------------------- # One entry per manual page. List of tuples # (source start file, name, description, authors, manual section). man_pages = [ ('index', 'pyzor', u'Pyzor Documentation', [u'Frank Tobin'], 1) ] # If true, show URL addresses after external links. #man_show_urls = False # -- Options for Texinfo output ------------------------------------------- # Grouping the document tree into Texinfo files. List of tuples # (source start file, target name, title, author, # dir menu entry, description, category) texinfo_documents = [ ('index', 'Pyzor', u'Pyzor Documentation', u'Frank Tobin', 'Pyzor', 'One line description of project.', 'Miscellaneous'), ] # Documents to append as an appendix to all manuals. #texinfo_appendices = [] # If false, no module index is generated. #texinfo_domain_indices = True # How to display URL addresses: 'footnote', 'no', or 'inline'. #texinfo_show_urls = 'footnote' # If true, do not generate a @detailmenu in the "Top" node's menu. #texinfo_no_detailmenu = False
8,310
30.244361
79
py
MINDER
MINDER-main/setup.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from distutils.core import setup, Extension import os extra_compile_args = ["-std=c++11", "-DNDEBUG", "-O3"] extension = Extension( "seal.cpp_modules._fm_index", include_dirs=["seal/cpp_modules", os.path.expanduser("~/include")], libraries=["stdc++", "sdsl", "divsufsort", "divsufsort64"], library_dirs=[os.path.expanduser("~/lib")], sources=["seal/cpp_modules/fm_index.cpp", "seal/cpp_modules/fm_index.i"], swig_opts=["-I../include", "-c++"], language="c++11", extra_compile_args=extra_compile_args, ) setup( name="SEAL", version="1.0", ext_modules=[extension], )
810
26.965517
77
py
MINDER
MINDER-main/scripts/build_fm_index.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import logging import multiprocessing import re import ftfy import torch import tqdm import pickle from seal.index import FMIndex from datasets import load_dataset import json logger = logging.getLogger() logger.setLevel(logging.ERROR) def process(line): tokens = tokenize(line) return tokens def preprocess_file(input_path, labels, format="kilt", lowercase=False, tokenize=False, pid2query=None): if pid2query: with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) if args.id2code: with open(args.id2code, 'rb') as f: id2code = pickle.load(f) if format =="msmarco": f = load_dataset(input_path, split="train") pieces_it = ((pp['docid'], pp["title"], pp["text"]) for pp in f) pieces_it = tqdm.tqdm(pieces_it) for idx, title, text in pieces_it: idx = idx.strip() title = title.strip() text = re.sub(r"\s+", " ", text) text = ftfy.fix_text(text) text = text.replace("BULLET::::", "") text = text.replace("SECTION::::", "") text = text.strip() if args.include_query: if args.query_format == 'free': queries = pid2query[idx] query = '' for s in queries: query =query + " " + s if args.query_format == 'stable': queries = pid2query[idx] query = '' for s in queries: query =query + " || " + s + ' @@' if args.include_code: code = id2code[idx] code = " || " + ' '.join([*code.strip()]) + ' @@' if not text: continue if tokenize: title = " ".join(word_tokenize(title)) text = " ".join(word_tokenize(text)) title = f"{title} {args.delim}" if args.include_title and title: text = f"{title} {text}" if args.include_query and query: text = f"{text} {query}" if lowercase: text = text.lower() labels.append(idx) yield text else: with open(input_path, "r", 2**16) as f: if format == "dpr": next(f) pieces_it = csv.reader(f, delimiter="\t", quotechar='"') pieces_it = ((pp[0], pp[2], pp[1]) for pp in pieces_it if len(pp) == 3) elif format == "kilt": pieces_it = (line.strip() for line in f) pieces_it = (line.split("\t", 2) for line in pieces_it) pieces_it = ((pp[0], pp[1], pp[2]) for pp in pieces_it if len(pp) == 3) pieces_it = tqdm.tqdm(pieces_it) for idx, title, text in pieces_it: idx = idx.strip() title = title.strip() text = re.sub(r"\s+", " ", text) text = ftfy.fix_text(text) text = text.replace("BULLET::::", "") text = text.replace("SECTION::::", "") text = text.strip() if args.include_query: if args.query_format == 'free': queries = pid2query[idx] query = '' for s in queries: query =query + " " + s if args.query_format == 'stable': queries = pid2query[idx] query = '' for s in queries: query =query + " || " + s + ' @@' if args.include_code: code = id2code[idx] code = " || " + ' '.join([*code.strip()]) + ' @@' if not text: continue if tokenize: title = " ".join(word_tokenize(title)) text = " ".join(word_tokenize(text)) title = f"{title} {args.delim}" if args.include_title and title: text = f"{title} {text}" if args.include_query and query: text = f"{text} {query}" if args.include_code and code: text = f"{text} {code}" if lowercase: text = text.lower() labels.append(idx) yield text def build_index(input_path): labels = [] index = FMIndex() lines = preprocess_file(input_path, labels, args.format, lowercase=args.lowercase, tokenize=args.tokenize, pid2query=args.pid2query) print('start build index') with multiprocessing.Pool(args.jobs) as p: sequences = p.imap(process, lines) index.initialize(sequences) print('start build index 2') index.labels = labels return index def parse_args(): parser = argparse.ArgumentParser() parser.add_argument("input") parser.add_argument("output") parser.add_argument("--jobs", type=int, default=1) parser.add_argument("--include_title", action="store_true") parser.add_argument("--delim", default="@@") parser.add_argument("--format", choices=["kilt", "dpr", "msmarco"], default="kilt") parser.add_argument("--hf_model", default=None, type=str) parser.add_argument("--lowercase", action="store_true") parser.add_argument("--tokenize", action="store_true") parser.add_argument('--pid2query', default=None, type=str) parser.add_argument('--id2code', default=None, type=str) parser.add_argument("--include_query", action="store_true") parser.add_argument("--include_code", action="store_true") parser.add_argument("--query_format", choices=["free", "stable"], default="free") return parser.parse_args() if __name__ == "__main__": args = parse_args() print(args) if args.tokenize: from spacy.lang.en import English nlp = English() _tokenizer = nlp.tokenizer def word_tokenize(text): return [t.text.strip() for t in _tokenizer(text)] if args.hf_model is not None: from transformers import AutoTokenizer tokenizer = AutoTokenizer.from_pretrained(args.hf_model, use_fast=False) is_bart = "bart" in args.hf_model def tokenize(text): text = text.strip() if is_bart: text = " " + text with tokenizer.as_target_tokenizer(): return tokenizer(text, add_special_tokens=False)["input_ids"] + [tokenizer.eos_token_id] else: bart = torch.hub.load("pytorch/fairseq", "bart.large").eval() def tokenize(text): return bart.encode(" " + text.strip()).tolist()[1:] delim = tokenize(args.delim)[:-1] index = build_index(args.input) print('start build index 3') index.save(args.output)
7,216
29.974249
136
py
MINDER
MINDER-main/scripts/training/make_supervised_msmarco_dataset2.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from argparse import ArgumentParser from collections import defaultdict import json import multiprocessing import random import re import tqdm import math import ftfy from fuzzywuzzy import fuzz from nltk.corpus import stopwords import nltk import pickle from datasets import load_dataset nltk.download('stopwords') banned = set(stopwords.words('english')) def parse_args(): parser = ArgumentParser() parser.add_argument('input') parser.add_argument('output') parser.add_argument('--min_score', default=999.0, type=float) parser.add_argument('--min_score_gold', default=999.0, type=float) parser.add_argument('--max_rank', default=1, type=int) parser.add_argument( '--target', default = "span", choices = [ "chunk", "span", "title", "code", "query", ]) parser.add_argument('--min_length', default=10, type=int) parser.add_argument('--max_length', default=10, type=int) parser.add_argument('--temperature', default=1.0, type=float) parser.add_argument('--jobs', default=30, type=int) parser.add_argument('--mark_target', action="store_true") parser.add_argument('--mark_silver', action="store_true") parser.add_argument('--n_samples', default=1, type=int) parser.add_argument('--id2code', default=None, type=str) parser.add_argument('--pid2query', default=None, type=str) parser.add_argument('--mode', choices=["w", "a"], default="w") return parser.parse_args() def read_id2code(id2code_path): id2code = {} with open(id2code_path) as fin: for line in tqdm.tqdm(fin): line = line.strip() if not line: continue idx, code = line.split("\t") id2code[idx] = code return id2code def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) # for j in range(i+1, min(i+1+ngrams, len(tokens) + 1)): # tok_orig = tokens[i:j] # tok = [t for t in tok_orig if t not in banned] # if not tok: # break # yield (i, j) def extract_spans(text, source, n_samples, min_length, max_length, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) yield span def extract_spans_wrapper(args): return args[1], list(extract_spans(*args)) def clean(text): text = re.sub(r'\s+', ' ', text) text = ftfy.fix_text(text) text = text.replace('BULLET::::', '') text = text.replace('SECTION::::', '') text = text.strip() return text def _iterator_span_get_arguments(data, min_score, max_rank, mark_target, mark_silver, min_score_gold): for sample in tqdm.tqdm(data): source = sample['query'].strip() if mark_target: source += " || body" for i, ctx in enumerate(sample['positive_passages'], start=1): ctx['score']=1000.0 if i > max_rank: continue if float(ctx['score']) < min_score: continue text = ctx['text'].strip() if mark_silver and float(ctx['score']) < min_score_gold: yield text, source + " || ?" elif mark_silver: yield text, source + " || +" else: yield text, source def iterator_span(args): if "train" in args.output: data = load_dataset(args.input, split="train") if "dev" in args.output: qid2query={} for s in load_dataset("Tevatron/msmarco-passage", split="dev"): qid2query[s['query_id']] = s['query'] passage_corpus = load_dataset("Tevatron/msmarco-passage-corpus", split="train") data = [] with open("/home/v-yongqili/project/GGR/data/MSMARCO/qrels.msmarco-passage.dev-subset.txt",'r') as f: for l in f: entry = {} l = re.split('[\t\s]', l.strip()) qid = str(l[0]) entry['query_id'] = qid entry['query'] = qid2query[qid] entry['positive_passages'] = [passage_corpus[int(l[2])]] data.append(entry) arg_it = _iterator_span_get_arguments(data, args.min_score, args.max_rank, args.mark_target, args.mark_silver, args.min_score_gold) arg_it = ((text, source, args.n_samples, args.min_length, args.max_length, args.temperature) for text, source in arg_it) with multiprocessing.Pool(args.jobs) as pool: for source, spans in pool.imap(extract_spans_wrapper, arg_it): for target in spans: yield source, target def iterator(args): if args.pid2query and args.target == "query": with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) if args.target == "code" and args.id2code: id2code = read_id2code(args.id2code) if "train" in args.output: data = load_dataset(args.input, split="train") if "dev" in args.output: qid2query={} for s in load_dataset("Tevatron/msmarco-passage", split="dev"): qid2query[s['query_id']] = s['query'] passage_corpus = load_dataset("Tevatron/msmarco-passage-corpus", split="train") data = [] with open("/home/v-yongqili/project/GGR/data/MSMARCO/qrels.msmarco-passage.dev-subset.txt",'r') as f: for l in f: entry = {} l = re.split('[\t\s]', l.strip()) qid = str(l[0]) entry['query_id'] = qid entry['query'] = qid2query[qid] entry['positive_passages'] = [passage_corpus[int(l[2])]] data.append(entry) for sample in tqdm.tqdm(data): source = sample['query'].strip() if args.target == "chunk" and args.mark_target: source += " || body" elif args.target == "title" and args.mark_target: source += " || title" elif args.target == "query" and args.mark_target: source += " || query" elif args.target == "code" and args.mark_target: source += " || code" else: raise ValueError("Wrong target") for i, ctx in enumerate(sample['positive_passages'], start=1): ctx['score']=1000.0 if i > args.max_rank: continue if float(ctx['score']) < args.min_score: continue if args.target == "chunk": target = ctx['text'].strip() for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "title": if ctx['title'].strip()=="-": continue target = ctx['title'].strip() + " @@" for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "query": if "trivia" in args.input: querys = pid2query[str(ctx['psg_id'])] if "nq" in args.input: querys = pid2query[str(ctx['passage_id'])] if "msmarco" in args.input: querys = pid2query[str(ctx['docid'])] for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", querys[min(_, len(querys)-1)] elif args.mark_silver: yield source + " || +", querys[min(_, len(querys)-1)] else: yield source, querys[min(_, len(querys)-1)] elif args.target == "code": idx = ctx['docid'] code = id2code.get(idx) if not code: continue target = code.strip() + " ||" for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target else: raise ValueError("Wrong target") def main(): args = parse_args() with open(args.output + '.source', mode=args.mode) as src, open(args.output + '.target', mode=args.mode) as tgt: for source, target in iterator_span(args) if args.target == "span" else iterator(args): source = " " + source.strip() target = " " + target.strip() src.write(source + "\n") tgt.write(target + "\n") if __name__ == '__main__': main()
10,873
33.302839
135
py
MINDER
MINDER-main/scripts/training/make_supervised_msmarco_dataset3.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from argparse import ArgumentParser from collections import defaultdict import json import multiprocessing import random import re import tqdm import math import ftfy from fuzzywuzzy import fuzz from nltk.corpus import stopwords import nltk import pickle from datasets import load_dataset nltk.download('stopwords') banned = set(stopwords.words('english')) def parse_args(): parser = ArgumentParser() parser.add_argument('input') parser.add_argument('output') parser.add_argument('--min_score', default=999.0, type=float) parser.add_argument('--min_score_gold', default=999.0, type=float) parser.add_argument('--max_rank', default=1, type=int) parser.add_argument( '--target', default = "span", choices = [ "chunk", "span", "title", "code", "query", ]) parser.add_argument('--min_length', default=10, type=int) parser.add_argument('--max_length', default=10, type=int) parser.add_argument('--temperature', default=1.0, type=float) parser.add_argument('--jobs', default=30, type=int) parser.add_argument('--mark_target', action="store_true") parser.add_argument('--mark_silver', action="store_true") parser.add_argument('--n_samples', default=1, type=int) parser.add_argument('--id2code', default=None, type=str) parser.add_argument('--pid2query', default=None, type=str) parser.add_argument('--mode', choices=["w", "a"], default="w") return parser.parse_args() def read_id2code(id2code_path): id2code = {} with open(id2code_path) as fin: for line in tqdm.tqdm(fin): line = line.strip() if not line: continue idx, code = line.split("\t") id2code[idx] = code return id2code def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) # for j in range(i+1, min(i+1+ngrams, len(tokens) + 1)): # tok_orig = tokens[i:j] # tok = [t for t in tok_orig if t not in banned] # if not tok: # break # yield (i, j) def extract_spans(text, source, n_samples, min_length, max_length, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) yield span def extract_spans_wrapper(args): return args[1], list(extract_spans(*args)) def clean(text): text = re.sub(r'\s+', ' ', text) text = ftfy.fix_text(text) text = text.replace('BULLET::::', '') text = text.replace('SECTION::::', '') text = text.strip() return text def _iterator_span_get_arguments(data, min_score, max_rank, mark_target, mark_silver, min_score_gold): for sample in tqdm.tqdm(data): source = sample['query'].strip() if mark_target: source += " || body" for i, ctx in enumerate(sample['positive_passages'], start=1): ctx['score']=1000.0 if i > max_rank: continue if float(ctx['score']) < min_score: continue text = ctx['text'].strip() if mark_silver and float(ctx['score']) < min_score_gold: yield text, source + " || ?" elif mark_silver: yield text, source + " || +" else: yield text, source def iterator_span(args): if "train" in args.output: data = load_dataset(args.input, split="train") if "dev" in args.output: qid2query={} for s in load_dataset("Tevatron/msmarco-passage", split="dev"): qid2query[s['query_id']] = s['query'] passage_corpus = load_dataset("Tevatron/msmarco-passage-corpus", split="train") data = [] with open("/home/v-yongqili/project/GGR/data/MSMARCO/qrels.msmarco-passage.dev-subset.txt",'r') as f: for l in f: entry = {} l = re.split('[\t\s]', l.strip()) qid = str(l[0]) entry['query_id'] = qid entry['query'] = qid2query[qid] entry['positive_passages'] = [passage_corpus[int(l[2])]] data.append(entry) arg_it = _iterator_span_get_arguments(data, args.min_score, args.max_rank, args.mark_target, args.mark_silver, args.min_score_gold) arg_it = ((text, source, args.n_samples, args.min_length, args.max_length, args.temperature) for text, source in arg_it) with multiprocessing.Pool(args.jobs) as pool: for source, spans in pool.imap(extract_spans_wrapper, arg_it): for target in spans: yield source, target def iterator(args): if args.pid2query and args.target == "query": with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) if args.target == "code" and args.id2code: id2code = read_id2code(args.id2code) if "train" in args.output: data = load_dataset(args.input, split="train") if "dev" in args.output: qid2query={} for s in load_dataset("Tevatron/msmarco-passage", split="dev"): qid2query[s['query_id']] = s['query'] passage_corpus = load_dataset("Tevatron/msmarco-passage-corpus", split="train") data = [] with open("/home/v-yongqili/project/GGR/data/MSMARCO/qrels.msmarco-passage.dev-subset.txt",'r') as f: for l in f: entry = {} l = re.split('[\t\s]', l.strip()) qid = str(l[0]) entry['query_id'] = qid entry['query'] = qid2query[qid] entry['positive_passages'] = [passage_corpus[int(l[2])]] data.append(entry) for sample in tqdm.tqdm(data): source = sample['query'].strip() if args.target == "chunk" and args.mark_target: source += " || body" elif args.target == "title" and args.mark_target: source += " || title" elif args.target == "query" and args.mark_target: source += " || query" elif args.target == "code" and args.mark_target: source += " || code" else: raise ValueError("Wrong target") for i, ctx in enumerate(sample['positive_passages'], start=1): ctx['score']=1000.0 if i > args.max_rank: continue if float(ctx['score']) < args.min_score: continue if args.target == "chunk": target = ctx['text'].strip() for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "title": if ctx['title'].strip()=="-": continue target = ctx['title'].strip() + " @@" for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "query": if "trivia" in args.input: querys = pid2query[str(ctx['psg_id'])] if "nq" in args.input: querys = pid2query[str(ctx['passage_id'])] if "msmarco" in args.input: querys = pid2query[str(ctx['docid'])] for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", querys[min(_, len(querys)-1)] elif args.mark_silver: yield source + " || +", querys[min(_, len(querys)-1)] else: yield source, querys[min(_, len(querys)-1)] elif args.target == "code": idx = ctx['docid'] code = id2code.get(idx) if not code: continue target = code.strip() + " ||" for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target else: raise ValueError("Wrong target") def main(): args = parse_args() with open(args.output + '.source', mode=args.mode) as src, open(args.output + '.target', mode=args.mode) as tgt: for source, target in iterator_span(args) if args.target == "span" else iterator(args): source = " " + source.strip() target = " " + target.strip() src.write(source + "\n") tgt.write(target + "\n") if __name__ == '__main__': main()
10,873
33.302839
135
py
MINDER
MINDER-main/scripts/training/make_generated_dataset.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import random import tqdm import pickle from nltk.corpus import stopwords from random import sample from fuzzywuzzy import fuzz from collections import defaultdict import math banned = { "the", "The", "to", "a", "A", "an", "An", "he", "He", "his", "His", "him", "He's", "she", "She", "her", "Her", "she's", "She's", "it", "It", "its", "Its", "it's", "It's", "and", "And", "or", "Or", "this", "This", "that", "That", "those", "Those", "these", "These", '"', '""', "'", "''", } def is_good(token): if token in banned: return False elif token[-1] in '?.!': return False elif token[0] in '([': return False return True def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) def extract_spans(text, source, n_samples, min_length=10, max_length=10, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) spans = [] for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) spans.append(span) return spans def preprocess_file( input_path, num_samples=1, num_title_samples=1, num_query_samples=1, format="dpr", delimiter='@@', min_length_input=1, max_length_input=15, min_length_output=10, max_length_output=10, full_doc_n=0, mark_pretraining=False, pid2query=None ): if format == 'kilt': raise NotImplementedError elif format == 'dpr': if pid2query: with open(pid2query, 'rb') as f: pid2query = pickle.load(f) with open(input_path, 'r', 2 ** 20) as f: next(f) f = csv.reader(f, delimiter='\t', quotechar='"') f = (l for l in f if len(l) == 3) for pid, text, title in tqdm.tqdm(f): try: text = text title = title querys = pid2query[str(pid)] if text == title: continue sampled = 0 select_query = sample(querys, 1)[0] a = select_query+ " || body" if mark_pretraining: a += " || p" spans = extract_spans(text, a, num_samples, min_length=10, max_length=10, temperature=1.0) while sampled < min(num_samples, len(spans)): b = spans[sampled] yield a, b sampled += 1 sampled = 0 select_querys = sample(querys*num_title_samples, num_title_samples) while sampled < num_title_samples: a=select_querys[sampled]+ " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter yield a, b sampled += 1 if len(querys)>1: sampled = 0 select_querys = sample(querys*num_query_samples, num_query_samples) while sampled < num_query_samples: a = select_querys[sampled]+ " || query" if mark_pretraining: a += " || p" num=0 while select_querys[sampled]==querys[num]: num+=1 if(num>=len(querys)): num-=1 break b = querys[num].strip() yield a, b sampled += 1 except Exception as e: print(e) continue else: raise ValueError def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('input') parser.add_argument('source') parser.add_argument('target') parser.add_argument('--delim', default="@@") parser.add_argument('--format', choices=['kilt', 'dpr'], default='dpr') parser.add_argument('--min_length_input', type=int, default=10) parser.add_argument('--max_length_input', type=int, default=10) parser.add_argument('--min_length_output', type=int, default=10) parser.add_argument('--max_length_output', type=int, default=10) parser.add_argument('--num_samples', type=int, default=10) parser.add_argument('--num_title_samples', type=int, default=3) parser.add_argument('--num_query_samples', type=int, default=5) parser.add_argument('--full_doc_n', type=int, default=1) parser.add_argument('--mark_pretraining', action="store_true") parser.add_argument('--pid2query', default=None, type=str) return parser.parse_args() def main(): args = parse_args() with open(args.source, 'w', 2 ** 20) as src, open(args.target, 'w', 2 ** 20) as tgt: for i, (s, t) in enumerate(preprocess_file( args.input, format=args.format, num_samples=args.num_samples, num_title_samples=args.num_title_samples, num_query_samples=args.num_query_samples, full_doc_n=args.full_doc_n, delimiter=args.delim, min_length_input=args.min_length_input, max_length_input=args.max_length_input, min_length_output=args.min_length_output, max_length_output=args.max_length_output, mark_pretraining=args.mark_pretraining, pid2query =args.pid2query, )): if random.random() < 0.1: s = s.lower() s = " " + s t = " " + t src.write(s + '\n') tgt.write(t + '\n') if __name__ == '__main__': main()
7,389
32.139013
110
py
MINDER
MINDER-main/scripts/training/make_generated_dataset3.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import random import tqdm import pickle from nltk.corpus import stopwords from random import sample from fuzzywuzzy import fuzz from collections import defaultdict import math import multiprocessing banned = { "the", "The", "to", "a", "A", "an", "An", "he", "He", "his", "His", "him", "He's", "she", "She", "her", "Her", "she's", "She's", "it", "It", "its", "Its", "it's", "It's", "and", "And", "or", "Or", "this", "This", "that", "That", "those", "Those", "these", "These", '"', '""', "'", "''", } def is_good(token): if token in banned: return False elif token[-1] in '?.!': return False elif token[0] in '([': return False return True def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) def extract_spans(text, source, n_samples, min_length=10, max_length=10, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) spans = [] for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) spans.append(span) return spans def preprocess_file( pid, text, title, querys, num_samples, num_title_samples, num_query_samples, format, delimiter, min_length_input, max_length_input, min_length_output, max_length_output, full_doc_n, mark_pretraining ): if format == 'kilt': raise NotImplementedError elif format == 'dpr': try: text = text title = title tokens = text.split() for _ in range(full_doc_n): a = text.strip() + " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter yield a, b sampled = 0 failures = 0 while sampled < num_title_samples and failures < 10: if random.random() > 0.5: len_a = random.randint(min_length_input, max_length_input) idx_a = random.randint(0, max(0, len(tokens)-len_a)) a = ' '.join(tokens[idx_a:idx_a+len_a]).strip() + " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter else: len_b = random.randint(min_length_output, max_length_output) idx_b = random.randint(0, max(0, len(tokens)-len_b)) if not is_good(tokens[idx_b]): failures += 1 continue b = ' '.join(tokens[idx_b:idx_b+len_b]).strip() a = title.strip() + ' || body' if mark_pretraining: a += " || p" yield a, b sampled += 1 sampled = 0 failures = 0 while sampled < num_samples and failures < 10: len_a = random.randint(min_length_input, max_length_input) len_b = random.randint(min_length_output, max_length_output) idx_a = random.randint(0, max(0, len(tokens)-len_a)) idx_b = random.randint(0, max(0, len(tokens)-len_b)) if idx_a == idx_b or (not is_good(tokens[idx_b])): failures += 1 continue a = ' '.join(tokens[idx_a:idx_a+len_a]).strip() + ' || body' if mark_pretraining: a += " || p" b = ' '.join(tokens[idx_b:idx_b+len_b]).strip() yield a, b sampled += 1 sampled = 0 failures = 0 while sampled < num_query_samples and failures < 10: if random.random() > 0.5: len_a = random.randint(min_length_input, max_length_input) idx_a = random.randint(0, max(0, len(tokens)-len_a)) a = ' '.join(tokens[idx_a:idx_a+len_a]).strip() + " || query" if mark_pretraining: a += " || p" b = sample(querys, 1)[0] else: len_b = random.randint(min_length_output, max_length_output) idx_b = random.randint(0, max(0, len(tokens)-len_b)) if not is_good(tokens[idx_b]): failures += 1 continue b = ' '.join(tokens[idx_b:idx_b+len_b]).strip() a = title.strip() + ' || body' if mark_pretraining: a += " || p" yield a, b sampled += 1 except Exception as e: print(e) else: raise ValueError def preprocess_file_wrapper(args): return list(preprocess_file(*args)) def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('input') parser.add_argument('source') parser.add_argument('target') parser.add_argument('--delim', default="@@") parser.add_argument('--format', choices=['kilt', 'dpr'], default='dpr') parser.add_argument('--min_length_input', type=int, default=10) parser.add_argument('--max_length_input', type=int, default=10) parser.add_argument('--min_length_output', type=int, default=10) parser.add_argument('--max_length_output', type=int, default=10) parser.add_argument('--num_samples', type=int, default=10) parser.add_argument('--num_title_samples', type=int, default=3) parser.add_argument('--num_query_samples', type=int, default=5) parser.add_argument('--full_doc_n', type=int, default=1) parser.add_argument('--mark_pretraining', action="store_true") parser.add_argument('--pid2query', default=None, type=str) parser.add_argument("--jobs", type=int, default=20) return parser.parse_args() def data_read(f, pid2query): for pid, text, title in tqdm.tqdm(f): querys = pid2query[str(pid)] if text == title: continue if len(querys)<2: continue yield pid, text, title, querys def main(): args = parse_args() with open(args.source, 'w', 2 ** 20) as src, open(args.target, 'w', 2 ** 20) as tgt: if args.pid2query: with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) with open(args.input, 'r', 2 ** 20) as f: next(f) f = csv.reader(f, delimiter='\t', quotechar='"') data = data_read(f, pid2query) arg_it = ((pid, text, title, querys, args.num_samples, args.num_title_samples, args.num_query_samples, args.format, args.delim, args.min_length_input,args.max_length_input, args.min_length_output, args.max_length_output, args.full_doc_n, args.mark_pretraining ) for pid, text, title, querys in data) with multiprocessing.Pool(args.jobs) as pool: for sts in pool.imap(preprocess_file_wrapper, arg_it): for s, t in sts: if random.random() < 0.1: s = s.lower() s = " " + s t = " " + t src.write(s + '\n') tgt.write(t + '\n') if __name__ == '__main__': main()
8,962
31.711679
140
py
MINDER
MINDER-main/scripts/training/make_supervised_dpr_dataset.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. from argparse import ArgumentParser from collections import defaultdict import json import multiprocessing import random import re import tqdm import math import ftfy from fuzzywuzzy import fuzz from nltk.corpus import stopwords import nltk import pickle nltk.download('stopwords') banned = set(stopwords.words('english')) def parse_args(): parser = ArgumentParser() parser.add_argument('input') parser.add_argument('output') parser.add_argument('--min_score', default=999.0, type=float) parser.add_argument('--min_score_gold', default=999.0, type=float) parser.add_argument('--max_rank', default=1, type=int) parser.add_argument( '--target', default = "span", choices = [ "chunk", "span", "title", "code", "query", ]) parser.add_argument('--min_length', default=10, type=int) parser.add_argument('--max_length', default=10, type=int) parser.add_argument('--temperature', default=1.0, type=float) parser.add_argument('--jobs', default=30, type=int) parser.add_argument('--mark_target', action="store_true") parser.add_argument('--mark_silver', action="store_true") parser.add_argument('--n_samples', default=1, type=int) parser.add_argument('--id2code', default=None, type=str) parser.add_argument('--pid2query', default=None, type=str) parser.add_argument('--mode', choices=["w", "a"], default="w") return parser.parse_args() def read_id2code(id2code_path): with open(id2code_path, 'rb') as f: id2code = pickle.load(f) return id2code def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) # for j in range(i+1, min(i+1+ngrams, len(tokens) + 1)): # tok_orig = tokens[i:j] # tok = [t for t in tok_orig if t not in banned] # if not tok: # break # yield (i, j) def extract_spans(text, source, n_samples, min_length, max_length, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) yield span def extract_spans_wrapper(args): return args[1], list(extract_spans(*args)) def clean(text): text = re.sub(r'\s+', ' ', text) text = ftfy.fix_text(text) text = text.replace('BULLET::::', '') text = text.replace('SECTION::::', '') text = text.strip() return text def _iterator_span_get_arguments(data, min_score, max_rank, mark_target, mark_silver, min_score_gold): for sample in tqdm.tqdm(data): source = sample['question'].strip() if mark_target: source += " || body" for i, ctx in enumerate(sample['positive_ctxs'], start=1): if i > max_rank: continue if float(ctx['score']) < min_score: continue text = ctx['text'].strip() if mark_silver and float(ctx['score']) < min_score_gold: yield text, source + " || ?" elif mark_silver: yield text, source + " || +" else: yield text, source def iterator_span(args): with open(args.input) as fin: data = json.load(fin) arg_it = _iterator_span_get_arguments(data, args.min_score, args.max_rank, args.mark_target, args.mark_silver, args.min_score_gold) arg_it = ((text, source, args.n_samples, args.min_length, args.max_length, args.temperature) for text, source in arg_it) with multiprocessing.Pool(args.jobs) as pool: for source, spans in pool.imap(extract_spans_wrapper, arg_it): for target in spans: yield source, target def iterator(args): if args.pid2query and args.target == "query": with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) if args.target == "code" and args.id2code: id2code = read_id2code(args.id2code) with open(args.input) as fin: data = json.load(fin) for sample in tqdm.tqdm(data): source = sample['question'].strip() if args.target == "chunk" and args.mark_target: source += " || body" elif args.target == "title" and args.mark_target: source += " || title" elif args.target == "query" and args.mark_target: source += " || query" elif args.target == "code" and args.mark_target: source += " || code" else: raise ValueError("Wrong target") for i, ctx in enumerate(sample['positive_ctxs'], start=1): if i > args.max_rank: continue if float(ctx['score']) < args.min_score: continue if args.target == "chunk": target = ctx['text'].strip() for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "title": target = ctx['title'].strip() + " @@" for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target elif args.target == "query": if "trivia" in args.input: querys = pid2query[str(ctx['psg_id'])] if "nq" in args.input: querys = pid2query[str(ctx['passage_id'])] for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", querys[min(_, len(querys)-1)] elif args.mark_silver: yield source + " || +", querys[min(_, len(querys)-1)] else: yield source, querys[min(_, len(querys)-1)] elif args.target == "code": if "trivia" in args.input: idx = ctx['psg_id'] if "nq" in args.input: idx = ctx['passage_id'] code = id2code.get(idx) if not code: continue target = ' '.join([*code.strip()]) for _ in range(args.n_samples): if args.mark_silver and float(ctx['score']) < args.min_score_gold: yield source + " || ?", target elif args.mark_silver: yield source + " || +", target else: yield source, target else: raise ValueError("Wrong target") def main(): args = parse_args() with open(args.output + '.source', mode=args.mode) as src, open(args.output + '.target', mode=args.mode) as tgt: for source, target in iterator_span(args) if args.target == "span" else iterator(args): source = " " + source.strip() target = " " + target.strip() src.write(source + "\n") tgt.write(target + "\n") if __name__ == '__main__': main()
9,100
31.974638
135
py
MINDER
MINDER-main/scripts/training/make_unsupervised_dataset.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import random import tqdm from nltk.corpus import stopwords banned = { "the", "The", "to", "a", "A", "an", "An", "he", "He", "his", "His", "him", "He's", "she", "She", "her", "Her", "she's", "She's", "it", "It", "its", "Its", "it's", "It's", "and", "And", "or", "Or", "this", "This", "that", "That", "those", "Those", "these", "These", '"', '""', "'", "''", } def is_good(token): if token in banned: return False elif token[-1] in '?.!': return False elif token[0] in '([': return False return True def preprocess_file( input_path, num_samples=1, num_title_samples=1, format="dpr", delimiter='@@', min_length_input=1, max_length_input=15, min_length_output=10, max_length_output=10, full_doc_n=0, mark_pretraining=False, ): if format == 'kilt': raise NotImplementedError elif format == 'dpr': with open(input_path, 'r', 2 ** 20) as f: next(f) f = csv.reader(f, delimiter='\t', quotechar='"') f = (l for l in f if len(l) == 3) for _, text, title in tqdm.tqdm(f): text = text title = title if text == title: continue tokens = text.split() for _ in range(full_doc_n): a = text.strip() + " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter yield a, b sampled = 0 failures = 0 while sampled < num_title_samples and failures < 10: if random.random() > 0.5: len_a = random.randint(min_length_input, max_length_input) idx_a = random.randint(0, max(0, len(tokens)-len_a)) a = ' '.join(tokens[idx_a:idx_a+len_a]).strip() + " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter else: len_b = random.randint(min_length_output, max_length_output) idx_b = random.randint(0, max(0, len(tokens)-len_b)) if not is_good(tokens[idx_b]): failures += 1 continue b = ' '.join(tokens[idx_b:idx_b+len_b]).strip() a = title.strip() + ' || body' if mark_pretraining: a += " || p" yield a, b sampled += 1 sampled = 0 failures = 0 while sampled < num_samples and failures < 10: len_a = random.randint(min_length_input, max_length_input) len_b = random.randint(min_length_output, max_length_output) idx_a = random.randint(0, max(0, len(tokens)-len_a)) idx_b = random.randint(0, max(0, len(tokens)-len_b)) if idx_a == idx_b or (not is_good(tokens[idx_b])): failures += 1 continue a = ' '.join(tokens[idx_a:idx_a+len_a]).strip() + ' || body' if mark_pretraining: a += " || p" b = ' '.join(tokens[idx_b:idx_b+len_b]).strip() yield a, b sampled += 1 else: raise ValueError def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('input') parser.add_argument('source') parser.add_argument('target') parser.add_argument('--delim', default="@@") parser.add_argument('--format', choices=['kilt', 'dpr'], default='dpr') parser.add_argument('--min_length_input', type=int, default=10) parser.add_argument('--max_length_input', type=int, default=10) parser.add_argument('--min_length_output', type=int, default=10) parser.add_argument('--max_length_output', type=int, default=10) parser.add_argument('--num_samples', type=int, default=10) parser.add_argument('--num_title_samples', type=int, default=3) parser.add_argument('--full_doc_n', type=int, default=1) parser.add_argument('--mark_pretraining', action="store_true") return parser.parse_args() def main(): args = parse_args() with open(args.source, 'w', 2 ** 20) as src, open(args.target, 'w', 2 ** 20) as tgt: for i, (s, t) in enumerate(preprocess_file( args.input, format=args.format, num_samples=args.num_samples, num_title_samples=args.num_title_samples, full_doc_n=args.full_doc_n, delimiter=args.delim, min_length_input=args.min_length_input, max_length_input=args.max_length_input, min_length_output=args.min_length_output, max_length_output=args.max_length_output, mark_pretraining=args.mark_pretraining, )): if random.random() < 0.1: s = s.lower() s = " " + s t = " " + t src.write(s + '\n') tgt.write(t + '\n') if __name__ == '__main__': main()
5,756
31.525424
88
py
MINDER
MINDER-main/scripts/training/make_generated_dataset2.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import random import tqdm import pickle from nltk.corpus import stopwords from random import sample from fuzzywuzzy import fuzz from collections import defaultdict import math import multiprocessing banned = { "the", "The", "to", "a", "A", "an", "An", "he", "He", "his", "His", "him", "He's", "she", "She", "her", "Her", "she's", "She's", "it", "It", "its", "Its", "it's", "It's", "and", "And", "or", "Or", "this", "This", "that", "That", "those", "Those", "these", "These", '"', '""', "'", "''", } def is_good(token): if token in banned: return False elif token[-1] in '?.!': return False elif token[0] in '([': return False return True def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) def extract_spans(text, source, n_samples, min_length=10, max_length=10, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) spans = [] for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) spans.append(span) return spans def preprocess_file( pid, text, title, querys, num_samples, num_title_samples, num_query_samples, format, delimiter, min_length_input, max_length_input, min_length_output, max_length_output, full_doc_n, mark_pretraining ): if format == 'kilt': raise NotImplementedError elif format == 'dpr': try: text = text title = title sampled = 0 select_query = sample(querys, 1)[0] a = select_query+ " || body" if mark_pretraining: a += " || p" spans = extract_spans(text, a, num_samples, min_length=10, max_length=10, temperature=1.0) while sampled < min(num_samples, len(spans)): b = spans[sampled] yield a, b sampled += 1 sampled = 0 select_querys = sample(querys*num_title_samples, num_title_samples) while sampled < num_title_samples: a=select_querys[sampled]+ " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter yield a, b sampled += 1 if len(querys)>1: sampled = 0 select_querys = sample(querys*num_query_samples, num_query_samples) while sampled < num_query_samples: a = select_querys[sampled]+ " || query" if mark_pretraining: a += " || p" num=0 while select_querys[sampled]==querys[num]: num+=1 if(num>=len(querys)): num-=1 break b = querys[num].strip() yield a, b sampled += 1 except Exception as e: print(e) else: raise ValueError def preprocess_file_wrapper(args): return list(preprocess_file(*args)) def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('input') parser.add_argument('source') parser.add_argument('target') parser.add_argument('--delim', default="@@") parser.add_argument('--format', choices=['kilt', 'dpr'], default='dpr') parser.add_argument('--min_length_input', type=int, default=10) parser.add_argument('--max_length_input', type=int, default=10) parser.add_argument('--min_length_output', type=int, default=10) parser.add_argument('--max_length_output', type=int, default=10) parser.add_argument('--num_samples', type=int, default=10) parser.add_argument('--num_title_samples', type=int, default=3) parser.add_argument('--num_query_samples', type=int, default=5) parser.add_argument('--full_doc_n', type=int, default=1) parser.add_argument('--mark_pretraining', action="store_true") parser.add_argument('--pid2query', default=None, type=str) parser.add_argument("--jobs", type=int, default=20) return parser.parse_args() def data_read(f, pid2query): for pid, text, title in tqdm.tqdm(f): querys = pid2query[str(pid)] if text == title: continue if len(querys)<2: continue yield pid, text, title, querys def main(): args = parse_args() with open(args.source, 'w', 2 ** 20) as src, open(args.target, 'w', 2 ** 20) as tgt: if args.pid2query: with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) with open(args.input, 'r', 2 ** 20) as f: next(f) f = csv.reader(f, delimiter='\t', quotechar='"') data = data_read(f, pid2query) arg_it = ((pid, text, title, querys, args.num_samples, args.num_title_samples, args.num_query_samples, args.format, args.delim, args.min_length_input,args.max_length_input, args.min_length_output, args.max_length_output, args.full_doc_n, args.mark_pretraining ) for pid, text, title, querys in data) with multiprocessing.Pool(args.jobs) as pool: for sts in pool.imap(preprocess_file_wrapper, arg_it): for s, t in sts: if random.random() < 0.1: s = s.lower() s = " " + s t = " " + t src.write(s + '\n') tgt.write(t + '\n') if __name__ == '__main__': main()
7,171
30.594714
140
py
MINDER
MINDER-main/scripts/training/make_generated_dataset_for_mamarco.py
# Copyright (c) Facebook, Inc. and its affiliates. # All rights reserved. # # This source code is licensed under the license found in the # LICENSE file in the root directory of this source tree. import argparse import csv import random import tqdm import pickle from nltk.corpus import stopwords from random import sample from fuzzywuzzy import fuzz from collections import defaultdict import math from datasets import load_dataset import multiprocessing banned = { "the", "The", "to", "a", "A", "an", "An", "he", "He", "his", "His", "him", "He's", "she", "She", "her", "Her", "she's", "She's", "it", "It", "its", "Its", "it's", "It's", "and", "And", "or", "Or", "this", "This", "that", "That", "those", "Those", "these", "These", '"', '""', "'", "''", } def is_good(token): if token in banned: return False elif token[-1] in '?.!': return False elif token[0] in '([': return False return True def span_iterator(tokens, ngrams=3, banned=banned): for i in range(len(tokens)): if tokens[i] not in banned: yield (i, i+ngrams) def extract_spans(text, source, n_samples, min_length=10, max_length=10, temperature=1.0): source = source.split("||", 1)[0] query_tokens = source.split() query_tokens_lower = [t.lower() for t in query_tokens] passage_tokens = text.split() passage_tokens_lower = [t.lower() for t in passage_tokens] matches = defaultdict(int) for i1, _ in enumerate(query_tokens_lower): j1 = i1+3 str_1 = " ".join(query_tokens_lower[i1:j1]) for (i2, j2) in span_iterator(passage_tokens_lower, 3): str_2 = " ".join(passage_tokens_lower[i2:j2]) ratio = fuzz.ratio(str_1, str_2) / 100.0 matches[i2] += ratio if not matches: indices = [0] else: indices, weights = zip(*sorted(matches.items(), key=lambda x: -(x[1]))) weights = list(weights) sum_weights = float(sum([0] + weights)) if sum_weights == 0.0 or not weights: indices = [0] weights = [1.0] else: weights = [math.exp(float(w) / temperature) for w in weights] Z = sum(weights) weights = [w / Z for w in weights] indices = random.choices(indices, weights=weights, k=n_samples) spans = [] for i in indices: subspan_size = random.randint(min_length, max_length) span = " ".join(passage_tokens[i:i+subspan_size]) spans.append(span) return spans def preprocess_file( pid, text, title, querys, num_samples, num_title_samples, num_query_samples, format, delimiter, min_length_input, max_length_input, min_length_output, max_length_output, full_doc_n, mark_pretraining ): if format == 'kilt': raise NotImplementedError elif format == 'dpr': try: text = text title = title sampled = 0 select_query = sample(querys, 1)[0] a = select_query+ " || body" if mark_pretraining: a += " || p" spans = extract_spans(text, a, num_samples, min_length=10, max_length=10, temperature=1.0) while sampled < min(num_samples, len(spans)): b = spans[sampled] yield a, b sampled += 1 sampled = 0 select_querys = sample(querys*num_title_samples, num_title_samples) if title != "-": while sampled < num_title_samples: a=select_querys[sampled]+ " || title" if mark_pretraining: a += " || p" b = title.strip() + " " + delimiter yield a, b sampled += 1 if len(querys)>1: sampled = 0 select_querys = sample(querys*num_query_samples, num_query_samples) while sampled < num_query_samples: a = select_querys[sampled]+ " || query" if mark_pretraining: a += " || p" num=0 while select_querys[sampled]==querys[num]: num+=1 if(num>=len(querys)): num-=1 break b = querys[num].strip() yield a, b sampled += 1 except Exception as e: print(e) else: raise ValueError def preprocess_file_wrapper(args): return list(preprocess_file(*args)) def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('input') parser.add_argument('source') parser.add_argument('target') parser.add_argument('--delim', default="@@") parser.add_argument('--format', choices=['kilt', 'dpr'], default='dpr') parser.add_argument('--min_length_input', type=int, default=10) parser.add_argument('--max_length_input', type=int, default=10) parser.add_argument('--min_length_output', type=int, default=10) parser.add_argument('--max_length_output', type=int, default=10) parser.add_argument('--num_samples', type=int, default=10) parser.add_argument('--num_title_samples', type=int, default=3) parser.add_argument('--num_query_samples', type=int, default=5) parser.add_argument('--full_doc_n', type=int, default=1) parser.add_argument('--mark_pretraining', action="store_true") parser.add_argument('--pid2query', default=None, type=str) parser.add_argument("--jobs", type=int, default=20) return parser.parse_args() def data_read(f, pid2query): for entry in tqdm.tqdm(f): pid, text, title = entry['docid'], entry['text'], entry['title'] querys = pid2query[str(pid)] if text == title: continue if len(querys)<2: continue yield pid, text, title, querys def main(): args = parse_args() with open(args.source, 'w', 2 ** 20) as src, open(args.target, 'w', 2 ** 20) as tgt: if args.pid2query: with open(args.pid2query, 'rb') as f: pid2query = pickle.load(f) passage_corpus = load_dataset(args.input, split="train") data = data_read(passage_corpus, pid2query) arg_it = ((pid, text, title, querys, args.num_samples, args.num_title_samples, args.num_query_samples, args.format, args.delim, args.min_length_input,args.max_length_input, args.min_length_output, args.max_length_output, args.full_doc_n, args.mark_pretraining ) for pid, text, title, querys in data) with multiprocessing.Pool(args.jobs) as pool: for sts in pool.imap(preprocess_file_wrapper, arg_it): for s, t in sts: if random.random() < 0.1: s = s.lower() s = " " + s t = " " + t src.write(s + '\n') tgt.write(t + '\n') if __name__ == '__main__': main()
7,234
30.872247
136
py