genericROM.Containers.OperatorCompressionData.OperatorCompressionDataMechanical

class OperatorCompressionDataMechanical(solutionName)[source]

Bases: Mordicus.Containers.OperatorCompressionData.OperatorCompressionDataBase.OperatorCompressionDataBase

Class containing an OperatorCompressionDataMechanical, used in the operator compression step of the POD-ECM method. The implementation uses Lagrange isoparametric finite elements as high-dimension integration model.

gradPhiAtIntegPoint

of length dimensionality of the mesh, of scipy.sparse.coo_matrix of size (numberOfIntegrationPoints, numberOfModes), components of the gradient of the shape functions at the integration points

Type

list

integrationWeights

of size (numberOfIntegrationPoints,), vector containing the integration weights associated to the computation of the internal forces vector with the high-fidelity integration scheme

Type

1D np.ndarray

listOfTags

(of length numberOfIntegrationPoints) at each integration point, containing all the tags of the element containing the integration points

Type

list of lists (of str)

reducedIntegrationPoints

of size (numberOfReducedIntegrationPoints,), vector containing the of the integration points associated to the computation of the internal forces vector with the reduced integration scheme

Type

1D np.ndarray

reducedIntegrationWeights

of size (numberOfReducedIntegrationPoints,), vector containing the integration weights associated to the computation of the internal forces vector with the reduced integration scheme

Type

1D np.ndarray

reducedListOTags

(of length numberOfReducedIntegrationPoints) at each reduced integration point containing all the tags of the element containing the reduced integration points. It is an extraction of listOfTags at the reduced integration points.

Type

list of lists (of str)

reducedEpsilonAtReducedIntegPoints

of size (numberOfSigmaComponents,numberOfReducedIntegrationPoints,numberOfModes), dtype = float, containing \(\epsilon(\Psi)(x_k'')\) where \(\Psi\) is a vector from the reducedOrderBasis associated to tje solution “U” and where \(x_k'\) are the reduced integration points

Type

np.ndarray

dualReconstructionData

dictionary containing data used for reconstructing dual quantities in the online stage, with following key:values

  • “methodDualReconstruction” : str (“GappyPOD” or “MetaModel”)

  • name of dual quantities (e.g. “evrcum”):

    • if “MetaModel” : tuple

      model: sklearn.model_selection._search.GridSearchCV

      scalerX: sklearn.preprocessing._data.StandardScaler

      scalery: sklearn.preprocessing._data.StandardScaler

    • if “GappyPOD” : tuple

      reducedOrderBasisAtReducedIntegrationPoints: np.ndarray of size (numberOfModes, nReducedIntegrationPoints)

Type

dict

GetDualReconstructionData()[source]

Get the dualReconstructionData attribute

Returns

dictionary containing data used for reconstructing dual quantities in the online stage

Return type

dict

GetGradPhiAtIntegPoint()[source]

Get the gradPhiAtIntegPoint attribute

Returns

of length dimensionality of the mesh, of scipy.sparse.coo_matrix of size (numberOfIntegrationPoints, numberOfModes)

Return type

list

GetIntegrationWeights()[source]

Get the integrationWeights attribute

Returns

of size (numberOfIntegrationPoints,)

Return type

1D np.ndarray

GetListOfTags()[source]

Get the listOfTags attribute

Returns

(of length numberOfIntegrationPoints)

Return type

list of lists (of str)

GetNumberOfIntegrationPoints()[source]

Get the number of integration points

Returns

number of integration points

Return type

int

GetNumberOfModes()[source]

Get the number of reducedOrderBasis modes

Returns

number of reducedOrderBasis modes

Return type

int

GetNumberOfReducedIntegrationPoints()[source]

Get the number of reduced iontegration points

Returns

number of reduced iontegration points

Return type

int

GetNumberOfSigmaComponents()[source]

Get the number of second order tensor components

Returns

number number of second order tensor components

Return type

int

GetReducedEpsilonAtReducedIntegPoints()[source]
Parameters

reducedEpsilonAtReducedIntegPoints (np.ndarray of size) – (numberOfSigmaComponents,numberOfReducedIntegrationPoints,numberOfModes), dtype = float contains \(\epsilon(\Psi)(x_k)\) where \(\Psi\) is a POD mode and \(x_k\) are the reduced integration points

GetReducedIntegrationPoints()[source]

Get the reducedIntegrationPoints attribute

Returns

of size (numberOfReducedIntegrationPoints,)

Return type

1D np.ndarray

GetReducedIntegrationWeights()[source]

Get the reducedIntegrationWeights attribute

Returns

of size (numberOfReducedIntegrationPoints,)

Return type

1D np.ndarray

GetReducedListOTags()[source]

Get the reducedListOTags attribute

Returns

(of length numberOfReducedIntegrationPoints)

Return type

list of lists (of str)

SetDualReconstructionData(dualReconstructionData)[source]

Sets the dualReconstructionData attribute

Parameters

dualReconstructionData (dict) – dictionary containing data used for reconstructing dual quantities in the online stage

SetOperatorPreCompressionData(operatorPreCompressionDataMechanical)[source]

Sets the gradPhiAtIntegPoint, integrationWeights and listOfTags attributes from an OperatorPreCompressionDataMechanical

Parameters

operatorPreCompressionDataMechanical (OperatorPreCompressionDataMechanical) – data structure used in a precomputation of the operator compression step of the POD-ECM method

SetReducedEpsilonAtReducedIntegPoints(reducedEpsilonAtReducedIntegPoints)[source]

Sets the reducedEpsilonAtReducedIntegPoints attribute

Parameters

reducedEpsilonAtReducedIntegPoints (np.ndarray) – of size (numberOfSigmaComponents,numberOfReducedIntegrationPoints,numberOfModes)

SetReducedIntegrationPoints(reducedIntegrationPoints)[source]

Sets the reducedIntegrationPoints attribute

Parameters

reducedIntegrationPoints (1D np.ndarray) – of size (numberOfReducedIntegrationPoints,)

SetReducedIntegrationWeights(reducedIntegrationWeights)[source]

Sets the reducedIntegrationWeights attribute

Parameters

reducedIntegrationWeights (1D np.ndarray) – of size (numberOfReducedIntegrationPoints,)

SetReducedListOTags(reducedListOTags)[source]

Sets the reducedListOTags attribute

Parameters

reducedListOTags (list of lists (of str)) – (of length numberOfReducedIntegrationPoints) at each reduced