add.hh File Reference

Algebraic domain decomposition based on the sparsity pattern of the matrix. More...

#include <cassert>
#include <vector>
#include "linalg/threadedMatrix.hh"
#include "mg/bddc.hh"
#include "utilities/timing.hh"

Go to the source code of this file.

Namespaces
namespace	Kaskade
namespace	Kaskade::BDDC

Functions
template<class Index >
std::vector< std::vector< Index > >	Kaskade::BDDC::algebraicDomainDecomposition (NumaCRSPattern< Index > const &A, int n)
	Creates a decomposition of indices into overlapping subsets that can be used for constructing nonoverlapping domain decomposition solvers. More...
template<class Scalar , class Index >
std::vector< NumaBCRSMatrix< Dune::FieldMatrix< Scalar, 1, 1 >, Index > >	Kaskade::BDDC::algebraicMatrixDecomposition (NumaBCRSMatrix< Dune::FieldMatrix< Scalar, 1, 1 >, Index > const &A, std::vector< std::vector< Index > > const &subIndices, Scalar tolerance, int maxIter)
template<class Scalar , int m, class Index >
std::tuple< std::vector< NumaBCRSMatrix< Dune::FieldMatrix< Scalar, m, m >, Index > >, std::vector< Dune::BlockVector< Dune::FieldVector< Scalar, m > > >, std::vector< std::vector< Index > >, std::vector< std::vector< LocalDof > > >	Kaskade::BDDC::matrixDomainDecommposition (NumaBCRSMatrix< Dune::FieldMatrix< Scalar, m, m >, Index > const &A, Dune::BlockVector< Dune::FieldVector< Scalar, m > > const &f, int n, Scalar tolerance=1000 *std::numeric_limits< Scalar >::epsilon(), int maxIter=100)

Detailed Description

Algebraic domain decomposition based on the sparsity pattern of the matrix.

Definition in file add.hh.