tmath::sparse::MUMPS Class Reference

solver for sparse matrices using MUMPS. More...

#include <sparsesolver_mumps.hpp>

Inheritance diagram for tmath::sparse::MUMPS:

[legend]

Collaboration diagram for tmath::sparse::MUMPS:

[legend]

List of all members.

Public Member Functions
virtual bool	Compute (SymSparseMatrix &)
	computes the factorization, returns true if succeeded
virtual bool	Compute (SparseMatrix &)
	computes the factorization, returns true if succeeded
	MUMPS (SymSparseMatrix &)
	MUMPS (SparseMatrix &)
	constructor which already factorizes the matrix
	MUMPS (const bool use_matrix_buffer_as_MUMPS_input_data=false)
	default constructor (inits m_use_matrix_buffer_as_MUMPS_input_data)
void	SetPositive (bool pos)
	Tell MUMPS if the matrix is positive definite (default: false).
virtual bool	Solve (const Matrix &b, Matrix &x)
virtual bool	SolveInPlace (Matrix &b)
virtual bool	SolveTransposed (const Matrix &b, Matrix &x)
virtual std::string	typeIdent () const
	returns the name of the datatype
virtual	~MUMPS ()
	destructor
Private Types
typedef TSolverInterface::MUMPS_STRUC_C	TSolverData
typedef TMumpsInterface< TScalar >	TSolverInterface
Private Member Functions
void	clear ()
template<class SparseMatrixType>
bool	Factorize (SparseMatrixType &)
void	FillData (SymSparseMatrix &, double values[], int rows[], int cols[])
void	FillData (SparseMatrix &, double values[], int rows[], int cols[])
void	init (const bool is_symmetric)
const std::string	mumpserror (int id) const
Private Attributes
bool	m_ispositive
	an internal flag (default: false) which tells MUMPS that the matrix is positive definite
TSolverData	m_solver
bool	m_solvercontainsdata
bool	m_use_matrix_buffer_as_MUMPS_input_data
	if this is on then the matrix buffer is directly used by MUMPS (warning: It may happen that MUMPS will change it unexpectedly)

---

Detailed Description

solver for sparse matrices using MUMPS.

---

Member Function Documentation

virtual bool tmath::sparse::MUMPS::Solve	(	const Matrix &	b,
		Matrix &	x
	)			[virtual]

Solves the given vector on return; returns true if succeeded

Reimplemented from tmath::sparse::SparseSolver.

virtual bool tmath::sparse::MUMPS::SolveInPlace

(

Matrix &

b

)

[virtual]

Factorizes this matrix and solves the given vector in place; returns true if succeeded

Reimplemented from tmath::sparse::SparseSolver.

virtual bool tmath::sparse::MUMPS::SolveTransposed	(	const Matrix &	b,
		Matrix &	x
	)			[virtual]

Solves the system (transpose(Matrix)*x=b for x; returns true if succeeded

Reimplemented from tmath::sparse::SparseSolver.

---

The documentation for this class was generated from the following file:

• modules/tmath/tmath/sparse/sparsesolver_mumps.hpp