1 # -*- coding: utf-8 -*-
3 # Copyright (C) 2008-2020 EDF R&D
5 # This library is free software; you can redistribute it and/or
6 # modify it under the terms of the GNU Lesser General Public
7 # License as published by the Free Software Foundation; either
8 # version 2.1 of the License.
10 # This library is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 # Lesser General Public License for more details.
15 # You should have received a copy of the GNU Lesser General Public
16 # License along with this library; if not, write to the Free Software
17 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19 # See http://www.salome-platform.org/ or email : webmaster.salome@opencascade.com
21 # Author: Jean-Philippe Argaud, jean-philippe.argaud@edf.fr, EDF R&D
24 from daCore import BasicObjects
25 import numpy, scipy.optimize
27 # ==============================================================================
28 class ElementaryAlgorithm(BasicObjects.Algorithm):
30 BasicObjects.Algorithm.__init__(self, "DIFFERENTIALEVOLUTION")
31 self.defineRequiredParameter(
35 message = "Stratégie de minimisation utilisée",
41 # "CURRENTTOBEST1EXP",
45 # "CURRENTTOBEST1BIN",
51 self.defineRequiredParameter(
52 name = "MaximumNumberOfSteps",
55 message = "Nombre maximal de générations",
58 self.defineRequiredParameter(
59 name = "MaximumNumberOfFunctionEvaluations",
62 message = "Nombre maximal d'évaluations de la fonction",
65 self.defineRequiredParameter(
66 name = "PopulationSize",
69 message = "Taille approximative de la population à chaque génération",
72 self.defineRequiredParameter(
73 name = "MutationDifferentialWeight_F",
76 message = "Poids différentiel de mutation, constant ou aléatoire dans l'intervalle, noté F",
80 self.defineRequiredParameter(
81 name = "CrossOverProbability_CR",
84 message = "Probabilité de recombinaison ou de croisement, notée CR",
88 self.defineRequiredParameter(
89 name = "QualityCriterion",
90 default = "AugmentedWeightedLeastSquares",
92 message = "Critère de qualité utilisé",
93 listval = ["AugmentedWeightedLeastSquares","AWLS","DA",
94 "WeightedLeastSquares","WLS",
95 "LeastSquares","LS","L2",
99 self.defineRequiredParameter(
100 name = "StoreInternalVariables",
103 message = "Stockage des variables internes ou intermédiaires du calcul",
105 self.defineRequiredParameter(
106 name = "StoreSupplementaryCalculations",
109 message = "Liste de calculs supplémentaires à stocker et/ou effectuer",
116 "CostFunctionJAtCurrentOptimum",
117 "CostFunctionJbAtCurrentOptimum",
118 "CostFunctionJoAtCurrentOptimum",
123 "InnovationAtCurrentState",
126 "SimulatedObservationAtBackground",
127 "SimulatedObservationAtCurrentOptimum",
128 "SimulatedObservationAtCurrentState",
129 "SimulatedObservationAtOptimum",
132 self.defineRequiredParameter(
134 typecast = numpy.random.seed,
135 message = "Graine fixée pour le générateur aléatoire",
137 self.defineRequiredParameter( # Pas de type
139 message = "Liste des valeurs de bornes",
141 self.requireInputArguments(
142 mandatory= ("Xb", "Y", "HO", "R", "B" ),
145 def run(self, Xb=None, Y=None, U=None, HO=None, EM=None, CM=None, R=None, B=None, Q=None, Parameters=None):
146 self._pre_run(Parameters, Xb, Y, R, B, Q)
148 len_X = numpy.asarray(Xb).size
149 popsize = round(self._parameters["PopulationSize"]/len_X)
150 maxiter = min(self._parameters["MaximumNumberOfSteps"],round(self._parameters["MaximumNumberOfFunctionEvaluations"]/(popsize*len_X) - 1))
151 logging.debug("%s Nombre maximal de générations = %i, taille de la population à chaque génération = %i"%(self._name, maxiter, popsize*len_X))
155 Hm = HO["Direct"].appliedTo
157 # Précalcul des inversions de B et R
158 # ----------------------------------
162 # Définition de la fonction-coût
163 # ------------------------------
164 def CostFunction(x, QualityMeasure="AugmentedWeightedLeastSquares"):
165 _X = numpy.asmatrix(numpy.ravel( x )).T
166 self.StoredVariables["CurrentState"].store( _X )
168 _HX = numpy.asmatrix(numpy.ravel( _HX )).T
169 _Innovation = Y - _HX
170 if self._toStore("SimulatedObservationAtCurrentState") or \
171 self._toStore("SimulatedObservationAtCurrentOptimum"):
172 self.StoredVariables["SimulatedObservationAtCurrentState"].store( _HX )
173 if self._toStore("InnovationAtCurrentState"):
174 self.StoredVariables["InnovationAtCurrentState"].store( _Innovation )
176 if QualityMeasure in ["AugmentedWeightedLeastSquares","AWLS","DA"]:
177 if BI is None or RI is None:
178 raise ValueError("Background and Observation error covariance matrix has to be properly defined!")
179 Jb = 0.5 * (_X - Xb).T * BI * (_X - Xb)
180 Jo = 0.5 * (_Innovation).T * RI * (_Innovation)
181 elif QualityMeasure in ["WeightedLeastSquares","WLS"]:
183 raise ValueError("Observation error covariance matrix has to be properly defined!")
185 Jo = 0.5 * (_Innovation).T * RI * (_Innovation)
186 elif QualityMeasure in ["LeastSquares","LS","L2"]:
188 Jo = 0.5 * (_Innovation).T * (_Innovation)
189 elif QualityMeasure in ["AbsoluteValue","L1"]:
191 Jo = numpy.sum( numpy.abs(_Innovation) )
192 elif QualityMeasure in ["MaximumError","ME"]:
194 Jo = numpy.max( numpy.abs(_Innovation) )
196 J = float( Jb ) + float( Jo )
198 self.StoredVariables["CostFunctionJb"].store( Jb )
199 self.StoredVariables["CostFunctionJo"].store( Jo )
200 self.StoredVariables["CostFunctionJ" ].store( J )
201 if self._toStore("IndexOfOptimum") or \
202 self._toStore("CurrentOptimum") or \
203 self._toStore("CostFunctionJAtCurrentOptimum") or \
204 self._toStore("CostFunctionJbAtCurrentOptimum") or \
205 self._toStore("CostFunctionJoAtCurrentOptimum") or \
206 self._toStore("SimulatedObservationAtCurrentOptimum"):
207 IndexMin = numpy.argmin( self.StoredVariables["CostFunctionJ"][nbPreviousSteps:] ) + nbPreviousSteps
208 if self._toStore("IndexOfOptimum"):
209 self.StoredVariables["IndexOfOptimum"].store( IndexMin )
210 if self._toStore("CurrentOptimum"):
211 self.StoredVariables["CurrentOptimum"].store( self.StoredVariables["CurrentState"][IndexMin] )
212 if self._toStore("SimulatedObservationAtCurrentOptimum"):
213 self.StoredVariables["SimulatedObservationAtCurrentOptimum"].store( self.StoredVariables["SimulatedObservationAtCurrentState"][IndexMin] )
214 if self._toStore("CostFunctionJAtCurrentOptimum"):
215 self.StoredVariables["CostFunctionJAtCurrentOptimum" ].store( self.StoredVariables["CostFunctionJ" ][IndexMin] )
216 if self._toStore("CostFunctionJbAtCurrentOptimum"):
217 self.StoredVariables["CostFunctionJbAtCurrentOptimum"].store( self.StoredVariables["CostFunctionJb"][IndexMin] )
218 if self._toStore("CostFunctionJoAtCurrentOptimum"):
219 self.StoredVariables["CostFunctionJoAtCurrentOptimum"].store( self.StoredVariables["CostFunctionJo"][IndexMin] )
222 # Point de démarrage de l'optimisation : Xini = Xb
223 # ------------------------------------
224 Xini = numpy.ravel(Xb)
226 # Minimisation de la fonctionnelle
227 # --------------------------------
228 nbPreviousSteps = self.StoredVariables["CostFunctionJ"].stepnumber()
230 optResults = scipy.optimize.differential_evolution(
232 self._parameters["Bounds"],
233 strategy = str(self._parameters["Minimizer"]).lower(),
236 mutation = self._parameters["MutationDifferentialWeight_F"],
237 recombination = self._parameters["CrossOverProbability_CR"],
238 disp = self._parameters["optdisp"],
241 IndexMin = numpy.argmin( self.StoredVariables["CostFunctionJ"][nbPreviousSteps:] ) + nbPreviousSteps
242 MinJ = self.StoredVariables["CostFunctionJ"][IndexMin]
243 Minimum = self.StoredVariables["CurrentState"][IndexMin]
245 # Obtention de l'analyse
246 # ----------------------
247 Xa = numpy.ravel( Minimum )
249 self.StoredVariables["Analysis"].store( Xa )
251 # Calculs et/ou stockages supplémentaires
252 # ---------------------------------------
253 if self._toStore("OMA") or self._toStore("SimulatedObservationAtOptimum"):
254 if self._toStore("SimulatedObservationAtCurrentState"):
255 HXa = self.StoredVariables["SimulatedObservationAtCurrentState"][IndexMin]
256 elif self._toStore("SimulatedObservationAtCurrentOptimum"):
257 HXa = self.StoredVariables["SimulatedObservationAtCurrentOptimum"][-1]
260 if self._toStore("Innovation") or \
261 self._toStore("OMB"):
263 if self._toStore("Innovation"):
264 self.StoredVariables["Innovation"].store( numpy.ravel(d) )
265 if self._toStore("OMB"):
266 self.StoredVariables["OMB"].store( numpy.ravel(d) )
267 if self._toStore("BMA"):
268 self.StoredVariables["BMA"].store( numpy.ravel(Xb) - numpy.ravel(Xa) )
269 if self._toStore("OMA"):
270 self.StoredVariables["OMA"].store( numpy.ravel(Y) - numpy.ravel(HXa) )
271 if self._toStore("SimulatedObservationAtBackground"):
272 self.StoredVariables["SimulatedObservationAtBackground"].store( numpy.ravel(Hm(Xb)) )
273 if self._toStore("SimulatedObservationAtOptimum"):
274 self.StoredVariables["SimulatedObservationAtOptimum"].store( numpy.ravel(HXa) )
279 # ==============================================================================
280 if __name__ == "__main__":
281 print('\n AUTODIAGNOSTIC\n')