--- /dev/null
+#-*-coding:iso-8859-1-*-
+#
+# Copyright (C) 2008-2012 EDF R&D
+#
+# This library is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public
+# License as published by the Free Software Foundation; either
+# version 2.1 of the License.
+#
+# This library is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public
+# License along with this library; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+#
+# See http://www.salome-platform.org/ or email : webmaster.salome@opencascade.com
+#
+
+import logging
+from daCore import BasicObjects, PlatformInfo
+m = PlatformInfo.SystemUsage()
+
+import numpy
+
+# ==============================================================================
+class ElementaryAlgorithm(BasicObjects.Algorithm):
+ def __init__(self):
+ BasicObjects.Algorithm.__init__(self, "ADJOINTTEST")
+ self.defineRequiredParameter(
+ name = "ResiduFormula",
+ default = "ScalarProduct",
+ typecast = str,
+ message = "Formule de résidu utilisée",
+ listval = ["ScalarProduct"],
+ )
+ self.defineRequiredParameter(
+ name = "EpsilonMinimumExponent",
+ default = -8,
+ typecast = int,
+ message = "Exposant minimal en puissance de 10 pour le multiplicateur d'incrément",
+ minval = -20,
+ maxval = 0,
+ )
+ self.defineRequiredParameter(
+ name = "InitialDirection",
+ default = [],
+ typecast = list,
+ message = "Direction initiale de la dérivée directionnelle autour du point nominal",
+ )
+ self.defineRequiredParameter(
+ name = "AmplitudeOfInitialDirection",
+ default = 1.,
+ typecast = float,
+ message = "Amplitude de la direction initiale de la dérivée directionnelle autour du point nominal",
+ )
+ self.defineRequiredParameter(
+ name = "SetSeed",
+ typecast = numpy.random.seed,
+ message = "Graine fixée pour le générateur aléatoire",
+ )
+ self.defineRequiredParameter(
+ name = "ResultTitle",
+ default = "",
+ typecast = str,
+ message = "Titre du tableau et de la figure",
+ )
+
+ def run(self, Xb=None, Y=None, H=None, M=None, R=None, B=None, Q=None, Parameters=None):
+ logging.debug("%s Lancement"%self._name)
+ logging.debug("%s Taille mémoire utilisée de %.1f Mo"%(self._name, m.getUsedMemory("Mo")))
+ #
+ # Paramètres de pilotage
+ # ----------------------
+ self.setParameters(Parameters)
+ #
+ # Opérateur d'observation
+ # -----------------------
+ Hm = H["Direct"].appliedTo
+ Ht = H["Tangent"].appliedInXTo
+ Ha = H["Adjoint"].appliedInXTo
+ #
+ # Construction des perturbations
+ # ------------------------------
+ Perturbations = [ 10**i for i in xrange(self._parameters["EpsilonMinimumExponent"],1) ]
+ Perturbations.reverse()
+ #
+ # Calcul du point courant
+ # -----------------------
+ X = numpy.asmatrix(Xb).flatten().T
+ NormeX = numpy.linalg.norm( X )
+ if Y is None:
+ Y = numpy.asmatrix( Hm( X ) ).flatten().T
+ Y = numpy.asmatrix(Y).flatten().T
+ NormeY = numpy.linalg.norm( Y )
+ #
+ # Fabrication de la direction de l'incrément dX
+ # ----------------------------------------------
+ if len(self._parameters["InitialDirection"]) == 0:
+ dX0 = []
+ for v in X.A1:
+ if abs(v) > 1.e-8:
+ dX0.append( numpy.random.normal(0.,abs(v)) )
+ else:
+ dX0.append( numpy.random.normal(0.,X.mean()) )
+ else:
+ dX0 = numpy.asmatrix(self._parameters["InitialDirection"]).flatten()
+ #
+ dX0 = float(self._parameters["AmplitudeOfInitialDirection"]) * numpy.matrix( dX0 ).T
+ #
+ # Utilisation de F(X) si aucune observation n'est donnee
+ # ------------------------------------------------------
+ #
+ # Entete des resultats
+ # --------------------
+ if self._parameters["ResiduFormula"] is "ScalarProduct":
+ __doc__ = """
+ On observe le residu qui est la difference de deux produits scalaires :
+
+ R(Alpha) = | < TangentF_X(dX) , Y > - < dX , AdjointF_X(Y) > |
+
+ qui doit rester constamment egal zero a la precision du calcul.
+ On prend dX0 = Normal(0,X) et dX = Alpha*dX0. F est le code de calcul.
+ Y doit etre dans l'image de F. S'il n'est pas donne, on prend Y = F(X).
+ """
+ else:
+ __doc__ = ""
+ #
+ msgs = " ====" + "="*len(self._parameters["ResultTitle"]) + "====\n"
+ msgs += " " + self._parameters["ResultTitle"] + "\n"
+ msgs += " ====" + "="*len(self._parameters["ResultTitle"]) + "====\n"
+ msgs += __doc__
+ #
+ msg = " i Alpha ||X|| ||Y|| ||dX|| R(Alpha) "
+ nbtirets = len(msg)
+ msgs += "\n" + "-"*nbtirets
+ msgs += "\n" + msg
+ msgs += "\n" + "-"*nbtirets
+ #
+ Normalisation= -1
+ #
+ # Boucle sur les perturbations
+ # ----------------------------
+ for i,amplitude in enumerate(Perturbations):
+ logging.debug("%s Etape de calcul numéro %i, avec la perturbation %8.3e"%(self._name, i, amplitude))
+ #
+ dX = amplitude * dX0
+ NormedX = numpy.linalg.norm( dX )
+ #
+ TangentFXdX = numpy.asmatrix( Ht( (X,dX) ) )
+ AdjointFXY = numpy.asmatrix( Ha( (X,Y) ) )
+ #
+ Residu = abs(float(numpy.dot( TangentFXdX.A1 , Y.A1 ) - numpy.dot( dX.A1 , AdjointFXY.A1 )))
+ #
+ msg = " %2i %5.0e %9.3e %9.3e %9.3e | %9.3e"%(i,amplitude,NormeX,NormeY,NormedX,Residu)
+ msgs += "\n" + msg
+ #
+ self.StoredVariables["CostFunctionJ"].store( Residu )
+ msgs += "\n" + "-"*nbtirets
+ msgs += "\n"
+ #
+ # Sorties eventuelles
+ # -------------------
+ logging.debug("%s Résultats :\n%s"%(self._name, msgs))
+ print
+ print "Results of adjoint stability check:"
+ print msgs
+ #
+ logging.debug("%s Taille mémoire utilisée de %.1f Mo"%(self._name, m.getUsedMemory("Mo")))
+ logging.debug("%s Terminé"%self._name)
+ #
+ return 0
+
+# ==============================================================================
+if __name__ == "__main__":
+ print '\n AUTODIAGNOSTIC \n'
)
self.defineRequiredParameter(
name = "ResultFile",
- default = "gradient_result_file",
+ default = self._name+"_result_file",
typecast = str,
message = "Nom de base (hors extension) des fichiers de sauvegarde des résultats",
)
# --------------------
if self._parameters["ResiduFormula"] is "Taylor":
__doc__ = """
- On observe le residu issu du développement de Taylor de la fonction H :
+ On observe le residu issu du développement de Taylor de la fonction F :
- R(Alpha) = || H(x+Alpha*dx) - H(x) - Alpha * TangentH_x(dx) ||
+ R(Alpha) = || F(X+Alpha*dX) - F(X) - Alpha * TangentF_X(dX) ||
Ce résidu doit décroître en Alpha**2 selon Alpha.
- On prend dX0 = Normal(0,X) et dX = Alpha*dX0. H est le code de calcul.
+ On prend dX0 = Normal(0,X) et dX = Alpha*dX0. F est le code de calcul.
"""
elif self._parameters["ResiduFormula"] is "Norm":
__doc__ = """
On observe le residu, qui est une approximation du gradient :
- || H(X+Alpha*dX) - H(X) ||
+ || F(X+Alpha*dX) - F(X) ||
R(Alpha) = ---------------------------
Alpha
qui doit rester constant jusqu'à ce qu'on atteigne la précision du calcul.
- On prend dX0 = Normal(0,X) et dX = Alpha*dX0. H est le code de calcul.
+ On prend dX0 = Normal(0,X) et dX = Alpha*dX0. F est le code de calcul.
"""
else:
__doc__ = ""
msgs += " ====" + "="*len(self._parameters["ResultTitle"]) + "====\n"
msgs += __doc__
#
- msg = " i Alpha ||X|| ||H(X)|| ||H(X+dX)|| ||dX|| ||H(X+dX)-H(X)|| ||H(X+dX)-H(X)||/||dX|| R(Alpha) "
+ msg = " i Alpha ||X|| ||F(X)|| ||F(X+dX)|| ||dX|| ||F(X+dX)-F(X)|| ||F(X+dX)-F(X)||/||dX|| R(Alpha) "
nbtirets = len(msg)
msgs += "\n" + "-"*nbtirets
msgs += "\n" + msg
Residu = NormedFXsAm
if Normalisation < 0 : Normalisation = Residu
#
- msg = " %2i %5.0e %8.3e %8.3e %8.3e %8.3e %8.3e | %8.3e | %8.3e"%(i,amplitude,NormeX,NormeFX,NormeFXdX,NormedX,NormedFX,NormedFXsdX,Residu)
+ msg = " %2i %5.0e %9.3e %9.3e %9.3e %9.3e %9.3e | %9.3e | %9.3e"%(i,amplitude,NormeX,NormeFX,NormeFXdX,NormedX,NormedFX,NormedFXsdX,Residu)
msgs += "\n" + msg
#
self.StoredVariables["CostFunctionJ"].store( Residu )
