Provides a convergence criterion for the linear solvers which looks at the weighted maximum of the difference between two iterations. More...

#include <fixpointcriterion.hh>

Inheritance diagram for Ewoms::FixPointCriterion< Vector, CollectiveCommunication >:

Public Member Functions
	FixPointCriterion (const CollectiveCommunication &comm)

	FixPointCriterion (const CollectiveCommunication &comm, const Vector &weightVec, Scalar reduction)

void	setWeight (const Vector &weightVec)
	Sets the relative weight of a primary variable. More...

Scalar	weight (int outerIdx, int innerIdx) const
	Return the relative weight of a primary variable. More...

void	setTolerance (Scalar tol)
	Set the maximum allowed weighted maximum difference between two iterations. More...

Scalar	tolerance () const
	Return the maximum allowed weighted maximum difference between two iterations. More...

void	setInitial (const Vector &curSol, const Vector &curResid)
	Set the initial solution of the linear system of equations. More...

void	update (const Vector &curSol, const Vector &curResid)
	Update the internal members of the convergence criterion with the current solution. More...

bool	converged () const
	Returns true if and only if the convergence criterion is met. More...

Scalar	accuracy () const
	Returns the accuracy of the solution at the last update. More...

virtual void	printInitial (std::ostream &os=std::cout) const
	Prints the initial information about the convergence behaviour. More...

virtual void	print (Scalar iter, std::ostream &os=std::cout) const
	Prints the information about the convergence behaviour for the current iteration. More...

Detailed Description

template<class Vector, class CollectiveCommunication>
class Ewoms::FixPointCriterion< Vector, CollectiveCommunication >

Provides a convergence criterion for the linear solvers which looks at the weighted maximum of the difference between two iterations.

For the FixPointCriterion, the error of the solution is defined as

$e^k = \max_i\{ \left| w_i \delta^k_i \right| \}\;,$

where $\delta = x^k - x^{k + 1}$ is the difference between two consequtive iterative solution vectors $x^k$ and $x^{k + 1}$ and $w_i$ is the weight of the $i$ -th degree of freedom.

This criterion requires that the block type of the vector is a Dune::FieldVector

Constructor & Destructor Documentation

template<class Vector , class CollectiveCommunication >

Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::FixPointCriterion ( const CollectiveCommunication & comm )

inline

template<class Vector , class CollectiveCommunication >

Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::FixPointCriterion	(	const CollectiveCommunication &	comm,
		const Vector &	weightVec,
		Scalar	reduction
	)

inline

Member Function Documentation

template<class Vector , class CollectiveCommunication >

Scalar Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::accuracy ( ) const

inlinevirtual

Returns the accuracy of the solution at the last update.

A value of zero means that the solution was exact.

Implements Ewoms::ConvergenceCriterion< Vector >.

template<class Vector , class CollectiveCommunication >

bool Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::converged ( ) const

inlinevirtual

Returns true if and only if the convergence criterion is met.

Implements Ewoms::ConvergenceCriterion< Vector >.

References Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::tolerance().

template<class Vector >

virtual void Ewoms::ConvergenceCriterion< Vector >::print	(	Scalar	iter,
		std::ostream &	os = `std::cout`
	)		const

inlinevirtualinherited

Prints the information about the convergence behaviour for the current iteration.

Parameters

iter	The iteration number. The semantics of this parameter are chosen by the linear solver.
os	The output stream to which the message gets written.

Reimplemented in Ewoms::WeightedResidualReductionCriterion< Vector, CollectiveCommunication >, and Ewoms::ResidReductionCriterion< Vector >.

template<class Vector >

virtual void Ewoms::ConvergenceCriterion< Vector >::printInitial ( std::ostream & os = std::cout ) const

inlinevirtualinherited

Prints the initial information about the convergence behaviour.

This method is called after setInitial() if the solver thinks it's a good idea to be verbose. In practice, "printing the initial information" means printing column headers and the initial state.

Parameters

os	The output stream to which the message gets written.

Reimplemented in Ewoms::WeightedResidualReductionCriterion< Vector, CollectiveCommunication >, and Ewoms::ResidReductionCriterion< Vector >.

template<class Vector , class CollectiveCommunication >

void Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::setInitial	(	const Vector &	curSol,
		const Vector &	curResid
	)

inlinevirtual

Set the initial solution of the linear system of equations.

This version of the method does NOT take the two-norm of the residual as argument. If the two-norm of the defect is available for the linear solver, the version of the update() method with it should be called.

Parameters

curSol	The current iterative solution of the linear system of equations
curResid	The residual vector of the current iterative solution of the linear system of equations

Implements Ewoms::ConvergenceCriterion< Vector >.

template<class Vector , class CollectiveCommunication >

void Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::setTolerance ( Scalar tol )

inline

Set the maximum allowed weighted maximum difference between two iterations.

template<class Vector , class CollectiveCommunication >

void Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::setWeight ( const Vector & weightVec )

inline

Sets the relative weight of a primary variable.

For the FixPointCriterion, the error of the solution is defined as

$e^k = \max_i\{ \left| w_i \delta^k_i \right| \}\;,$

where $\delta = x^k - x^{k + 1}$ is the difference between two consequtive iterative solution vectors $x^k$ and $x^{k + 1}$ and $w_i$ is the weight of the $i$ -th degree of freedom.

This method is specific to the FixPointCriterion.

Parameters

weightVec A Dune::BlockVector<Dune::FieldVector<Scalar, n> > with the relative weights of the degrees of freedom

template<class Vector , class CollectiveCommunication >

Scalar Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::tolerance ( ) const

inline

Return the maximum allowed weighted maximum difference between two iterations.

Referenced by Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::converged().

template<class Vector , class CollectiveCommunication >

void Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::update	(	const Vector &	curSol,
		const Vector &	curResid
	)

inlinevirtual

Update the internal members of the convergence criterion with the current solution.

This version of the method does NOT take the two-norm of the residual as argument. If the two-norm of the defect is available for the linear solver, the version of the update() method with it should be called.

Parameters

curSol	The current iterative solution of the linear system of equations
curResid	The residual vector of the current iterative solution of the linear system of equations

Implements Ewoms::ConvergenceCriterion< Vector >.

References Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::weight().

template<class Vector , class CollectiveCommunication >

Scalar Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::weight	(	int	outerIdx,
		int	innerIdx
	)		const

inline

Return the relative weight of a primary variable.

For the FixPointCriterion, the error of the solution is defined as

$e^k = \max_i\{ \left| w_i \delta^k_i \right| \}\;,$

where $\delta = x^k - x^{k + 1}$ is the difference between two consequtive iterative solution vectors $x^k$ and $x^{k + 1}$ and $w_i$ is the weight of the $i$ -th degree of freedom.

This method is specific to the FixPointCriterion.

Parameters

outerIdx	The index of the outer vector (i.e. Dune::BlockVector)
innerIdx	The index of the inner vector (i.e. Dune::FieldVector)

Referenced by Ewoms::FixPointCriterion< Vector, CollectiveCommunication >::update().

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

fixpointcriterion.hh

Public Member Functions

Detailed Description

template<class Vector, class CollectiveCommunication> class Ewoms::FixPointCriterion< Vector, CollectiveCommunication >

Constructor & Destructor Documentation

Member Function Documentation

template<class Vector, class CollectiveCommunication>
class Ewoms::FixPointCriterion< Vector, CollectiveCommunication >