ml_MultiLevelPreconditioner.h File Reference

#include "ml_include.h"
#include "Epetra_SerialDenseMatrix.h"
#include "Epetra_SerialDenseVector.h"
#include "Epetra_SerialDenseSolver.h"
#include "Epetra_Operator.h"
#include "Epetra_RowMatrix.h"
#include "Epetra_MultiVector.h"
#include "Epetra_MsrMatrix.h"
#include "Teuchos_ParameterList.hpp"
#include "EpetraExt_SolverMap_CrsMatrix.h"
#include "ml_epetra_utils.h"

Include dependency graph for ml_MultiLevelPreconditioner.h:

This graph shows which files directly or indirectly include this file:

Classes
class	ML_Epetra::MultiLevelPreconditioner
	MultiLevelPreconditioner: a class to define black-box multilevel preconditioners using aggregation methods. More...
class	MLVec< T >
struct	wrappedCommStruct

Namespaces
	ML_Epetra
	ML_Epetra: default namespace for all Epetra interfaces.

Macros
#define	ML_MEM_SIZE   20
#define	ML_MEM_INITIAL   0
#define	ML_MEM_FINAL   1
#define	ML_MEM_SMOOTHER   2
#define	ML_MEM_COARSE   3
#define	ML_MEM_HIERARCHY   4
#define	ML_MEM_PREC_FIRST   5
#define	ML_MEM_PREC_OTHER   6
#define	ML_MEM_TOT1   7
#define	ML_MEM_TOT2   8
#define	ML_MEM_INITIAL_MALLOC   10
#define	ML_MEM_FINAL_MALLOC   11
#define	ML_MEM_SMOOTHER_MALLOC   12
#define	ML_MEM_COARSE_MALLOC   13
#define	ML_MEM_HIERARCHY_MALLOC   14
#define	ML_MEM_PREC_FIRST_MALLOC   15
#define	ML_MEM_PREC_OTHER_MALLOC   16
#define	ML_MEM_TOT1_MALLOC   17
#define	ML_MEM_TOT2_MALLOC   18

Enumerations
enum	ML_Epetra::AMGType { ML_SA_FAMILY, ML_MAXWELL, ML_COMPOSITE, ML_CLASSICAL_FAMILY }
	Enumerated type indicating the type of AMG solver to be used.
enum	free_s { useFree, useDelete, neither }
enum	epetraOrMLOrMueLu { epetraType, mlType, mueluType }

Functions
int	ML_Epetra::SetDefaults (std::string ProblemType, Teuchos::ParameterList &List, int *options=0, double *params=0, const bool OverWrite=true)
	Sets default parameters for aggregation-based preconditioners. More...
int	ML_Epetra::SetDefaultsDD (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for aggregation-based 2-level domain decomposition preconditioners. More...
int	ML_Epetra::SetDefaultsDD_LU (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for aggregation-based 2-level domain decomposition preconditioners, using LU on each subdomain. More...
int	ML_Epetra::SetDefaultsDD_3Levels (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for aggregation-based 3-level domain decomposition preconditioners. More...
int	ML_Epetra::SetDefaultsDD_3Levels_LU (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for aggregation-based 3-level domain decomposition preconditioners with LU. More...
int	ML_Epetra::SetDefaultsMaxwell (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for the eddy current equations equations. More...
int	ML_Epetra::SetDefaultsSA (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets default parameters for classical smoothed aggregation. More...
int	ML_Epetra::SetDefaultsNSSA (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets defaults for energy minimization preconditioning for nonsymmetric problems. More...
int	ML_Epetra::SetDefaultsClassicalAMG (Teuchos::ParameterList &List, Teuchos::RCP< std::vector< int > > &options, Teuchos::RCP< std::vector< double > > &params, bool Overwrite=true)
	Sets defaults for classical amg.
int	ML_Epetra::ReadXML (const std::string &FileName, Teuchos::ParameterList &List, const Epetra_Comm &Comm)
	Reads in parameter list options from file.
template<class T >
int	nodalComm (MLVec< T > &vector, MLVec< int > &myLocalNodeIds, struct wrappedCommStruct &framework)
template<class T >
int	dofComm (MLVec< T > &vector, struct wrappedCommStruct &framework)
int	dofCommUsingMlNodalMatrix (double *data, void *widget)
int	MLsortCols (const MLVec< int > &ARowPtr, MLVec< int > &ACols, MLVec< double > &AVals)
int	MLnMyGhost (struct wrappedCommStruct &framework)
int	MLfillNodalMaps (MLVec< int > &amalgRowMap, MLVec< int > &amalgColMap, MLVec< int > &myLocalNodeIds, int nLocalDofs, struct wrappedCommStruct &framework, int nLocalNodes, int nLocalPlusGhostNodes)
int	MLcolGlobalIds (struct wrappedCommStruct &framework, MLVec< int > &myGids)
int	MLassignGhostLocalNodeIds (MLVec< int > &myLocalNodeIds, int nLocalDofs, int nLocalPlusGhostDofs, struct wrappedCommStruct &framework, int &nLocalNodes, int &nLocalPlusGhostNodes)
int	MLextractDiag (const MLVec< int > &rowPtr, const MLVec< int > &cols, const MLVec< double > &, MLVec< double > &diagonal, struct wrappedCommStruct &framework)
int	MLfindDirichlets (const MLVec< int > &rowPtr, const MLVec< int > &cols, const MLVec< double > &vals, const MLVec< double > &diagonal, double tol, MLVec< bool > &dirOrNot, struct wrappedCommStruct &framework)
int	MLrmDirichletCols (MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals, const MLVec< double > &diagonal, bool squeeze, MLVec< double > &solution, MLVec< double > &rhs, const MLVec< bool > &dirOrNot, struct wrappedCommStruct &framework)
int	MLsqueezeOutNnzs (MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals, const MLVec< bool > &keep)
int	MLbuildMap (const MLVec< bool > &dofPresent, MLVec< int > &map, int NDof)
int	MLvariableDofAmalg (int nCols, const MLVec< int > &rowPtr, const MLVec< int > &cols, const MLVec< double > &vals, int nNodes, int maxDofPerNode, const MLVec< int > &map, const MLVec< double > &diag, double tol, MLVec< int > &amalgRowPtr, MLVec< int > &amalgCols, struct wrappedCommStruct &framework, MLVec< int > &myLocalNodeIds)
int	MLrmDifferentDofsCrossings (const MLVec< bool > &dofPresent, int maxDofPerNode, MLVec< int > &rowPtr, MLVec< int > &cols, int nCols, struct wrappedCommStruct &framework, MLVec< int > &myLocalNodeIds)
int	MLbuildLaplacian (const MLVec< int > &rowPtr, const MLVec< int > &cols, MLVec< double > &vals, const MLVec< double > &x, const MLVec< double > &y, const MLVec< double > &z)
int	MLunamalgP (const MLVec< int > &amalgRowPtr, const MLVec< int > &amalgCols, const MLVec< double > &amalgVals, int maxDofPerNode, const MLVec< char > &status, bool fineIsPadded, MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals)
	Unamalgamate prolongator so that it is suitable for PDE systems. More...
int	MLfindEmptyRows (const MLVec< int > &rowPtr, const MLVec< int > &cols, const MLVec< double > &vals, MLVec< bool > &rowEmpty)
int	MLreplaceEmptyByDirichlet (MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals, const MLVec< bool > &colEmpty)
int	MLfineStatus (const MLVec< bool > &dofPresent, const MLVec< int > &map, const MLVec< bool > &dirOrNot, MLVec< char > &status)
int	MLcoarseStatus (const MLVec< bool > &rowEmpty, const MLVec< bool > &dirOrNot, MLVec< char > &status)
int	MLShove (ML_Operator *Mat, MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals, int invec_leng, int(*commfunc)(double *vec, void *data), struct wrappedCommStruct &framework, int nGhost)
	Take Crs std::vec's, make arrays and shove them into an ML Operator. More...
int	ZeroDist (MLVec< double > &xxx, MLVec< double > &yyy, MLVec< double > &zzz, MLVec< int > &rowPtr, MLVec< int > &cols, MLVec< double > &vals, MLVec< double > &diagonal, double tol, MLVec< int > &rowZeros, MLVec< int > &colZeros, double disttol)
int	MergeShared (MLVec< int > &cols, MLVec< int > &rowZeros, MLVec< int > &colZeros, MLVec< int > &groupHead, MLVec< int > &groupNext)
int	BuildNonSharedMap (MLVec< int > &newMap, MLVec< int > &groupHead, MLVec< int > &groupNext)
int	buildCompressedA (MLVec< int > &inputRowPtr, MLVec< int > &inputCols, MLVec< double > &inputVals, MLVec< double > &diagonal, MLVec< int > &groupHead, MLVec< int > &groupNext, MLVec< int > &outputRowptr, MLVec< int > &outputCols, MLVec< int > &map, int newN, double tol)

Function Documentation

int MLShove	(	ML_Operator *	Mat,
		MLVec< int > &	rowPtr,
		MLVec< int > &	cols,
		MLVec< double > &	vals,
		int	invec_leng,
		int(*)(double *vec, void *data)	commfunc,
		struct wrappedCommStruct &	framework,
		int	nGhost
	)

Take Crs std::vec's, make arrays and shove them into an ML Operator.

This includes allocating double arrays and a widget to point to them and calling all the appropriate ML functions to clean out what used to be in the operator before inserting the new matrix.

Parameters

[in]	rowPtr,cols,vals	Crs matrix std:vectors
	Mat	On input, a populated ML Operator. On output, old ML Operator is cleaned and a new Crs matrix is filled into the operator

References ML_Operator_Struct::N_nonzeros.

Referenced by ML_Epetra::MultiLevelPreconditioner::MultiLevelPreconditioner().

int MLunamalgP	(	const MLVec< int > &	amalgRowPtr,
		const MLVec< int > &	amalgCols,
		const MLVec< double > &	amalgVals,
		int	maxDofPerNode,
		const MLVec< char > &	status,
		bool	fineIsPadded,
		MLVec< int > &	rowPtr,
		MLVec< int > &	cols,
		MLVec< double > &	vals
	)

Unamalgamate prolongator so that it is suitable for PDE systems.

Unamalgamate Pmat (represented as a CRS matrix via amalgRowPtr, amalgCols, amalgVals for two different situations: Case 1: Fine level matrix is padded In this case, we basically replicate the unamalgamated operator, except that padded dofs and Dirichlet points use injection. Thus, PUnAmalg is equivalent to something like ( Pmat 0 0 ) ( 0 Pmat 0 ) ( 0 0 Pmat ) where entries associated with fine level padded dofs or Dir. BCs are replaced by rows with a single nonzero (whose value equals one) and the entire matrix is instead ordered so that dofs associated with nodes are consecutive.

Case 2: Fine level matrix is not padded. Thus, PUnAmalg is equivalent to ( Pmat 0 0 ) ( 0 Pmat 0 ) ( 0 0 Pmat ) where rows associated with dofs not present on finest level (e.g., an air pressure dof within water region) are removed and rows associated with Dir. BCs are replaced by rows with a single nonzero (whose value equals one) and the entire matrix is instead ordered so dofs associated with nodes are consecutive.

In both of the above cases, the coarse level discretization matrix is assumed to be padded.

amalgNRows local number of rows for amalgamted prolongator amalgRowPtr CRS local row pointer for amalgamted prolongator amalgCols CRS local cols for amalgamted prolongator amalgVals CRS vals for amalgamted prolongator maxDofPerNode maximum number of degrees-of-freedom at any mesh node status status[i*maxDofPerNode+j] refers to the jth dof at the ith node. status()==s ==> standard element: present in fine operator and not a Dirichlet BC. status()==d ==> element corresponds to Dirichlet BC. status()==p ==> element not present or is padded dof. fineIsPadded Indicates whether fine grid matrix includes padded dofs for those not really present in the PDE system rowPtr CRS local row pointer for resulting unamalgamated prolongator cols CRS local cols for resulting unamalgamated prolongator vals CRS vals for resulting unamalgamated prolongator

Referenced by ML_Epetra::MultiLevelPreconditioner::MultiLevelPreconditioner().