Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner > Class Template Reference

Implements a preconditioned stabilized BiCG linear solver. More...

#include <bicgstabsolver.hh>

Public Member Functions
	BiCGStabSolver (Preconditioner &preconditioner, ConvergenceCriterion &convergenceCriterion, Dune::ScalarProduct< Vector > &scalarProduct)
void	setMaxIterations (unsigned value)
	Set the maximum number of iterations before we give up without achieving convergence. More...
unsigned	maxIterations () const
	Return the maximum number of iterations before we give up without achieving convergence. More...
void	setVerbosity (unsigned value)
	Set the verbosity level of the linear solver. More...
unsigned	verbosity () const
	Return the verbosity level of the linear solver. More...
void	setLinearOperator (const LinearOperator *A)
	Set the matrix "A" of the linear system. More...
void	setRhs (const Vector *b)
	Set the right hand side "b" of the linear system. More...
bool	apply (Vector &x)
	Run the stabilized BiCG solver and store the result into the "x" vector. More...
void	setConvergenceCriterion (ConvergenceCriterion &crit)
const SolverReport &	report () const

Detailed Description

template<class LinearOperator, class Vector, class Preconditioner>
class Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >

Implements a preconditioned stabilized BiCG linear solver.

This solves a linear system of equations Ax = b, where the matrix A is sparse and may be unsymmetric.

See https://en.wikipedia.org/wiki/Biconjugate_gradient_stabilized_method, (article date: December 19, 2016)

Constructor & Destructor Documentation

BiCGStabSolver()

template<class LinearOperator , class Vector , class Preconditioner >

Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::BiCGStabSolver ( Preconditioner &  preconditioner, ConvergenceCriterion &  convergenceCriterion, Dune::ScalarProduct< Vector > &  scalarProduct  )

inline

Member Function Documentation

apply()

template<class LinearOperator , class Vector , class Preconditioner >

bool Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::apply ( Vector &  x )

inline

Run the stabilized BiCG solver and store the result into the "x" vector.

References Opm::Linear::SolverReport::converged(), Opm::Linear::ConvergenceCriterion< Vector >::converged(), Opm::Linear::ConvergenceCriterion< Vector >::failed(), Opm::Linear::SolverReport::increment(), Opm::Linear::SolverReport::iterations(), Opm::Linear::ConvergenceCriterion< Vector >::print(), Opm::Linear::ConvergenceCriterion< Vector >::printInitial(), Opm::Linear::SolverReport::reset(), Opm::Linear::SolverReport::setConverged(), Opm::Linear::ConvergenceCriterion< Vector >::setInitial(), Opm::Timer::start(), Opm::Linear::SolverReport::timer(), and Opm::Linear::ConvergenceCriterion< Vector >::update().

maxIterations()

template<class LinearOperator , class Vector , class Preconditioner >

unsigned Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::maxIterations ( ) const

inline

Return the maximum number of iterations before we give up without achieving convergence.

report()

template<class LinearOperator , class Vector , class Preconditioner >

const SolverReport & Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::report ( ) const

inline

setConvergenceCriterion()

template<class LinearOperator , class Vector , class Preconditioner >

void Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::setConvergenceCriterion ( ConvergenceCriterion &  crit )

inline

setLinearOperator()

template<class LinearOperator , class Vector , class Preconditioner >

void Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::setLinearOperator ( const LinearOperator *  A )

inline

Set the matrix "A" of the linear system.

setMaxIterations()

template<class LinearOperator , class Vector , class Preconditioner >

void Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::setMaxIterations ( unsigned  value )

inline

Set the maximum number of iterations before we give up without achieving convergence.

setRhs()

template<class LinearOperator , class Vector , class Preconditioner >

void Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::setRhs ( const Vector *  b )

inline

Set the right hand side "b" of the linear system.

setVerbosity()

template<class LinearOperator , class Vector , class Preconditioner >

void Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::setVerbosity ( unsigned  value )

inline

Set the verbosity level of the linear solver.

The levels correspont to those used by the dune-istl solvers:

• 0: no output
• 1: summary output at the end of the solution proceedure (if no exception was thrown)
• 2: detailed output after each iteration

verbosity()

template<class LinearOperator , class Vector , class Preconditioner >

unsigned Opm::Linear::BiCGStabSolver< LinearOperator, Vector, Preconditioner >::verbosity ( ) const

inline

Return the verbosity level of the linear solver.

---

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

• bicgstabsolver.hh