Eigen::PardisoLDLT< MatrixType, Options > Class Template Reference

A sparse direct Cholesky (LDLT) factorization and solver based on the PARDISO library. More...

Inheritance diagram for Eigen::PardisoLDLT< MatrixType, Options >:

Public Types
enum	{ UpLo }
typedef Base::RealScalar	RealScalar
typedef Base::Scalar	Scalar
typedef Base::StorageIndex	StorageIndex
Public Types inherited from Eigen::PardisoImpl< PardisoLDLT< MatrixType, Options > >
enum
typedef Matrix< StorageIndex, MatrixType::RowsAtCompileTime, 1 >	IntColVectorType
typedef Matrix< StorageIndex, 1, MatrixType::ColsAtCompileTime >	IntRowVectorType
typedef Traits::MatrixType	MatrixType
typedef Array< StorageIndex, 64, 1, DontAlign >	ParameterType
typedef Traits::RealScalar	RealScalar
typedef Traits::Scalar	Scalar
typedef SparseMatrix< Scalar, RowMajor, StorageIndex >	SparseMatrixType
typedef Traits::StorageIndex	StorageIndex
typedef Matrix< Scalar, Dynamic, 1 >	VectorType

Public Member Functions
void	getMatrix (const MatrixType &matrix)
	PardisoLDLT ()
	PardisoLDLT (const MatrixType &matrix)
Public Member Functions inherited from Eigen::PardisoImpl< PardisoLDLT< MatrixType, Options > >
void	_solve_impl (const MatrixBase< BDerived > &b, MatrixBase< XDerived > &x) const
void	_solve_impl (const MatrixBase< Rhs > &b, MatrixBase< Dest > &dest) const
PardisoLDLT< MatrixType, Options > &	analyzePattern (const MatrixType &matrix)
Index	cols () const
PardisoLDLT< MatrixType, Options > &	compute (const MatrixType &matrix)
PardisoLDLT< MatrixType, Options > &	factorize (const MatrixType &matrix)
ComputationInfo	info () const
	Reports whether previous computation was successful. More...
	PardisoImpl ()
ParameterType &	pardisoParameterArray ()
Index	rows () const
	~PardisoImpl ()
Public Member Functions inherited from Eigen::SparseSolverBase< Derived >
Derived &	derived ()
const Derived &	derived () const
template<typename Rhs >
const Solve< Derived, Rhs >	solve (const MatrixBase< Rhs > &b) const
template<typename Rhs >
const Solve< Derived, Rhs >	solve (const SparseMatrixBase< Rhs > &b) const
	SparseSolverBase ()
	SparseSolverBase (SparseSolverBase &&other)
	~SparseSolverBase ()

Protected Types
typedef PardisoImpl< PardisoLDLT< MatrixType, Options > >	Base
Protected Types inherited from Eigen::PardisoImpl< PardisoLDLT< MatrixType, Options > >
typedef SparseSolverBase< PardisoLDLT< MatrixType, Options > >	Base
typedef internal::pardiso_traits< PardisoLDLT< MatrixType, Options > >	Traits

Additional Inherited Members
Protected Member Functions inherited from Eigen::PardisoImpl< PardisoLDLT< MatrixType, Options > >
PardisoLDLT< MatrixType, Options > &	derived ()
const PardisoLDLT< MatrixType, Options > &	derived () const
void	manageErrorCode (Index error) const
void	pardisoInit (int type)
void	pardisoRelease ()
Protected Attributes inherited from Eigen::PardisoImpl< PardisoLDLT< MatrixType, Options > >
bool	m_analysisIsOk
bool	m_factorizationIsOk
ComputationInfo	m_info
ParameterType	m_iparm
bool	m_isInitialized
SparseMatrixType	m_matrix
StorageIndex	m_msglvl
IntColVectorType	m_perm
void *	m_pt [64]
Index	m_size
StorageIndex	m_type
Protected Attributes inherited from Eigen::SparseSolverBase< Derived >
bool	m_isInitialized

Detailed Description

template<typename MatrixType, int Options>
class Eigen::PardisoLDLT< MatrixType, Options >

A sparse direct Cholesky (LDLT) factorization and solver based on the PARDISO library.

This class allows to solve for A.X = B sparse linear problems via a LDL^T Cholesky factorization using the Intel MKL PARDISO library. The sparse matrix A is assumed to be selfajoint and positive definite. For complex matrices, A can also be symmetric only, see the Options template parameter. The vectors or matrices X and B can be either dense or sparse.

By default, it runs in in-core mode. To enable PARDISO's out-of-core feature, set:

solver.pardisoParameterArray()[59] = 1; 

Template Parameters

MatrixType	the type of the sparse matrix A, it must be a SparseMatrix<>
Options	can be any bitwise combination of Upper, Lower, and Symmetric. The default is Upper, meaning only the upper triangular part has to be used. Symmetric can be used for symmetric, non-selfadjoint complex matrices, the default being to assume a selfadjoint matrix. Upper|Lower can be used to tell both triangular parts can be used as input.

This class follows the sparse solver concept .

See also

Sparse solver concept, class SimplicialLDLT

Definition at line 506 of file PardisoSupport.h.

Member Typedef Documentation

Base

template<typename MatrixType , int Options>

typedef PardisoImpl< PardisoLDLT<MatrixType,Options> > Eigen::PardisoLDLT< MatrixType, Options >::Base

protected

Definition at line 509 of file PardisoSupport.h.

RealScalar

template<typename MatrixType , int Options>

typedef Base::RealScalar Eigen::PardisoLDLT< MatrixType, Options >::RealScalar

Definition at line 517 of file PardisoSupport.h.

Scalar

template<typename MatrixType , int Options>

typedef Base::Scalar Eigen::PardisoLDLT< MatrixType, Options >::Scalar

Definition at line 516 of file PardisoSupport.h.

StorageIndex

template<typename MatrixType , int Options>

typedef Base::StorageIndex Eigen::PardisoLDLT< MatrixType, Options >::StorageIndex

Definition at line 518 of file PardisoSupport.h.

Member Enumeration Documentation

anonymous enum

template<typename MatrixType , int Options>

anonymous enum

Enumerator
UpLo

Definition at line 520 of file PardisoSupport.h.

{ UpLo = Options&(Upper|Lower) };

Constructor & Destructor Documentation

PardisoLDLT() [1/2]

template<typename MatrixType , int Options>

Eigen::PardisoLDLT< MatrixType, Options >::PardisoLDLT ( )

inline

Definition at line 522 of file PardisoSupport.h.

      : Base()
    {
      pardisoInit(Base::ScalarIsComplex ? ( bool(Options&Symmetric) ? 6 : -4 ) : -2);
    }

PardisoLDLT() [2/2]

template<typename MatrixType , int Options>

Eigen::PardisoLDLT< MatrixType, Options >::PardisoLDLT ( const MatrixType &  matrix )

inlineexplicit

Definition at line 528 of file PardisoSupport.h.

      : Base()
    {
      pardisoInit(Base::ScalarIsComplex ? ( bool(Options&Symmetric) ? 6 : -4 ) : -2);
      compute(matrix);
    }

Member Function Documentation

getMatrix()

template<typename MatrixType , int Options>

void Eigen::PardisoLDLT< MatrixType, Options >::getMatrix ( const MatrixType &  matrix )

inline

Definition at line 535 of file PardisoSupport.h.

    {
      // PARDISO supports only upper, row-major matrices
      PermutationMatrix<Dynamic,Dynamic,StorageIndex> p_null;
      m_matrix.resize(matrix.rows(), matrix.cols());
      m_matrix.template selfadjointView<Upper>() = matrix.template selfadjointView<UpLo>().twistedBy(p_null);
      m_matrix.makeCompressed();
    }

---

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

• PardisoSupport.h