The ExponentialSolver struct solves exp(prefactor*mat)*vec. More...

#include <classic_krylov.h>

Collaboration diagram for syten::Krylov::ClassicImpl::ExponentialSolver< Solver >:

Public Types
using	Vector = typename Solver::Vector_type
	The vector type of the calling solver. More...

Public Member Functions
template<typename MatrixScalar , typename Vectors , typename Scalar >
void	finalizeSolution (Solver const &, IterativeConfig const &, std::pair< DenseTensor< 2, MatrixScalar >, Vectors > &&mat_v, SRDef initialNorm, Scalar, Vector &vec)
	as the name suggests, this actually solves after the test has run More...

template<typename MatrixScalar , typename Vectors , typename Scalar >
SRDef	solveForConvergence (Solver const &, IterativeConfig const &, std::pair< DenseTensor< 2, MatrixScalar >, Vectors > &&mat_v, SRDef, Scalar prefactor, Vector &)
	operator () solve `exp(prefactormat)vec` and calculate the error, does not actually return the solution. More...

Private Member Functions
template<typename MatrixScalar , typename Scalar >
auto	solve (DenseTensor< 2, MatrixScalar > mat, Scalar prefactor)
	Solves the Lanczos subproblem and calculates the current result. More...

Private Attributes
SRDef	normPrev = NAN
	Norm of the result from the previous call. More...

DenseTensor< 1, SDef >	prev
	Result from previous call. More...

boost::any	prevPrefactor
	Prefactor of the previous call to this solver. More...

Detailed Description

template<typename Solver>
struct syten::Krylov::ClassicImpl::ExponentialSolver< Solver >

The ExponentialSolver struct solves exp(prefactor*mat)*vec.

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

Public Types