TrilinosWrappers::SolverDirect Class Reference

#include <deal.II/lac/trilinos_solver.h>

Classes
struct	AdditionalData

Public Member Functions
	SolverDirect (SolverControl &cn, const AdditionalData &data=AdditionalData())
virtual	~SolverDirect ()
void	initialize (const SparseMatrix &A)
void	solve (VectorBase &x, const VectorBase &b)
void	solve (const SparseMatrix &A, VectorBase &x, const VectorBase &b)
void	solve (const SparseMatrix &A, ::Vector< double > &x, const ::Vector< double > &b)
void	solve (const SparseMatrix &A, ::LinearAlgebra::distributed::Vector< double > &x, const ::LinearAlgebra::distributed::Vector< double > &b)
SolverControl &	control () const

Static Public Member Functions
static ::ExceptionBase &	ExcTrilinosError (int arg1)

Private Member Functions
void	do_solve ()

Private Attributes
SolverControl &	solver_control
std_cxx11::shared_ptr< Epetra_LinearProblem >	linear_problem
std_cxx11::shared_ptr< Amesos_BaseSolver >	solver
const AdditionalData	additional_data

Detailed Description

An implementation of Trilinos direct solvers (using the Amesos package). The data field AdditionalData::solver_type can be used to specify the type of solver. It allows the use of built-in solvers Amesos_Klu as well as third-party solvers Amesos_Superludist or Amesos_Mumps.

For instructions on how to install Trilinos for use with direct solvers other than KLU, see the link to the Trilinos installation instructions linked to from the deal.II ReadMe file.

Author

Martin Kronbichler, 2009, Uwe Köcher, 2014

Definition at line 611 of file trilinos_solver.h.

Constructor & Destructor Documentation

SolverDirect()

TrilinosWrappers::SolverDirect::SolverDirect	(	SolverControl &	cn,
		const AdditionalData &	data = AdditionalData()
	)

Constructor. Takes the solver control object and creates the solver.

Definition at line 710 of file trilinos_solver.cc.

~SolverDirect()

TrilinosWrappers::SolverDirect::~SolverDirect ( )

virtual

Destructor.

Definition at line 719 of file trilinos_solver.cc.

Member Function Documentation

initialize()

void TrilinosWrappers::SolverDirect::initialize

(

const SparseMatrix &

A

)

Initializes the direct solver for the matrix A and creates a factorization for it with the package chosen from the additional data structure. Note that there is no need for a preconditioner here and solve() is not called.

Definition at line 732 of file trilinos_solver.cc.

solve() [1/4]

void TrilinosWrappers::SolverDirect::solve	(	VectorBase &	x,
		const VectorBase &	b
	)

Solve the linear system Ax=b based on the package set in intialize(). Note the matrix is not refactorized during this call.

Definition at line 779 of file trilinos_solver.cc.

solve() [2/4]

void TrilinosWrappers::SolverDirect::solve	(	const SparseMatrix &	A,
		VectorBase &	x,
		const VectorBase &	b
	)

Solve the linear system Ax=b. Creates a factorization of the matrix with the package chosen from the additional data structure and performs the solve. Note that there is no need for a preconditioner here.

Definition at line 860 of file trilinos_solver.cc.

solve() [3/4]

void TrilinosWrappers::SolverDirect::solve	(	const SparseMatrix &	A,
		::Vector< double > &	x,
		const ::Vector< double > &	b
	)

Solve the linear system Ax=b. This class works with Trilinos matrices, but takes deal.II serial vectors as argument. Since these vectors are not distributed, this function does only what you expect in case the matrix is serial (i.e., locally owned). Otherwise, an exception will be thrown.

Definition at line 877 of file trilinos_solver.cc.

solve() [4/4]

void TrilinosWrappers::SolverDirect::solve	(	const SparseMatrix &	A,
		::LinearAlgebra::distributed::Vector< double > &	x,
		const ::LinearAlgebra::distributed::Vector< double > &	b
	)

Solve the linear system Ax=b for deal.II's own parallel vectors. Creates a factorization of the matrix with the package chosen from the additional data structure and performs the solve. Note that there is no need for a preconditioner here.

Definition at line 905 of file trilinos_solver.cc.

control()

SolverControl & TrilinosWrappers::SolverDirect::control

(

)

const

Access to object that controls convergence.

Definition at line 725 of file trilinos_solver.cc.

do_solve()

void TrilinosWrappers::SolverDirect::do_solve ( )

private

Actually performs the operations for solving the linear system, including the factorization and forward and backward substitution.

Definition at line 807 of file trilinos_solver.cc.

Member Data Documentation

solver_control

SolverControl& TrilinosWrappers::SolverDirect::solver_control

private

Reference to the object that controls convergence of the iterative solver. In fact, for these Trilinos wrappers, Trilinos does so itself, but we copy the data from this object before starting the solution process, and copy the data back into it afterwards.

Definition at line 741 of file trilinos_solver.h.

linear_problem

std_cxx11::shared_ptr<Epetra_LinearProblem> TrilinosWrappers::SolverDirect::linear_problem

private

A structure that collects the Trilinos sparse matrix, the right hand side vector and the solution vector, which is passed down to the Trilinos solver.

Definition at line 748 of file trilinos_solver.h.

solver

std_cxx11::shared_ptr<Amesos_BaseSolver> TrilinosWrappers::SolverDirect::solver

private

A structure that contains the Trilinos solver and preconditioner objects.

Definition at line 754 of file trilinos_solver.h.

additional_data

const AdditionalData TrilinosWrappers::SolverDirect::additional_data

private

Store a copy of the flags for this particular solver.

Definition at line 759 of file trilinos_solver.h.

---

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

• deal.II/lac/trilinos_solver.h
• source/lac/trilinos_solver.cc