#include <triplet.hh>

Detailed Description

template<class Scalar, class SparseIndexInt = int>
class Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >

Sparse Matrix in triplet format.

A simple sparse matrix format consisting of scalar-valued (row,col,value) entries. Use this as the least common denominator of sparse linear algebra interfaces.

Multiple entries with same row and col are semantically summed up.

Template Parameters

Scalar	(double or float, field_type)
SparseIndexInt	(int,long)

long has to be used as SparseIndexInt if the number of nonzeros of the sparse matrix exceeds INT_MAX 2,147,483,647 or if the umfpack direct solver is used and the internal workspace exceeds this number, what will happen much earlier.

Definition at line 56 of file triplet.hh.

Public Types
typedef Scalar	value_type
	STL-compliant typedefs. More...

typedef std::vector< Scalar >::iterator	iterator

typedef std::vector< Scalar >::const_iterator	const_iterator

typedef std::vector< SparseIndexInt >::iterator	index_iterator

typedef std::vector< SparseIndexInt >::const_iterator	const_index_iterator

Public Member Functions
	MatrixAsTriplet ()
	Constructor. More...

	MatrixAsTriplet (size_t s)
	Constructor that allocates memory directly. More...

template<int n, int m, class Allocator >
	MatrixAsTriplet (Dune::BCRSMatrix< Dune::FieldMatrix< Scalar, n, m >, Allocator > const &other)
	Constructor from a Dune::BCRSMatrix. More...

template<int n, int m, class Index >
	MatrixAsTriplet (NumaBCRSMatrix< Dune::FieldMatrix< Scalar, n, m >, Index > const &other)
	Constructor from a NumaBCRSMatrix. More...

	MatrixAsTriplet (std::vector< SparseIndexInt > &&row, std::vector< SparseIndexInt > &&col, std::vector< Scalar > &&value)
	Move constructor taking raw data. More...

	MatrixAsTriplet (SparseIndexInt nrows, SparseIndexInt ncols, Scalar const &value)
	Constructs a matrix of size (nrows,ncols), filled with value. If value==0.0, then it is an empty matrix. More...

	MatrixAsTriplet (SparseIndexInt nrows, SparseIndexInt ncols, SparseIndexInt shift, Scalar const &value)
	Constructs a diagonal matrix of size (nrows,ncols). The diagonal is shifted down shift rows (if shift is negative, it is shifted up) More...

	MatrixAsTriplet (SparseIndexInt nrows, SparseIndexInt ncols, SparseIndexInt shift, std::vector< Scalar > const &value)
	Constructs a diagonal matrix of size (nrows,ncols). The diagonal is shifted down shift rows (if shift is negative, it is shifted up) More...

	MatrixAsTriplet (MatrixAsTriplet const &other)=default
	Copy constructor. More...

	MatrixAsTriplet (MatrixAsTriplet &&other)=default
	Move constructor. More...

template<class OtherScalar , class OtherIndex >
	MatrixAsTriplet (MatrixAsTriplet< OtherScalar, OtherIndex > const &A)
	Copy constructor from a triplet matrix with different entry and/or index types. More...

void	clear ()
	Deletes the data, leaving an empty matrix of size 0 x 0. More...

void	setStartToZero ()
	Shift the row and column indices such that \( A_{00} \) is (structurally) nonzero. More...

SparseIndexInt	nrows () const
	Returns number of rows (computes them, if not known) More...

SparseIndexInt	N () const
	Returns number of rows (computes them, if not known) This is a Dune ISTL compatibility method. More...

SparseIndexInt	ncols () const
	Returns number of cols (computes them, if not known) More...

SparseIndexInt	M () const
	Returns number of columns (computes them, if not known) This is a Dune ISTL compatibility method. More...

void	resize (size_t s)
	Resizes the memory. More...

void	reserve (size_t s)
	Reserve memory. More...

size_t	nnz () const

template<class X , class Y >
void	axpy (Y &out, X const &in, Scalar alpha=1.0, int nvectors=1) const
	Matrix-vector multiplication: out += alpha * (this) in. More...

template<class X , class Y >
void	atxpy (Y &y, X const &x, Scalar alpha=1.0) const
	Matrix-vector multiplication. More...

template<class X , class Y >
void	usmtv (Scalar const alpha, X const &x, Y &y) const
	Matrix-vector multiplication. More...

template<class X , class Y >
void	usmv (Scalar const alpha, X const &x, Y &y) const
	scaled matrix-vector multiplication More...

template<class X , class Y >
void	ax (Y &out, X const &in) const
	Matrix-vector multiplication: out = (this) in. More...

template<class X , class Y >
void	mv (X const &x, Y &y) const
	matrix-vector multiplication More...

void	shiftIndices (SparseIndexInt row, SparseIndexInt col)
	Shifts matrix indices. Can be used together with += to concatenate sparese matrices. More...

void	addEntry (SparseIndexInt row, SparseIndexInt col, Scalar const &value)
	Adds the given value to the specified matrix entry. The entry is created in case it has been structurally zero before. More...

MatrixAsTriplet &	operator+= (MatrixAsTriplet const &m)
	Matrix addition: (*this) += m (works also for matrices with non-matching sparsity pattern) More...

MatrixAsTriplet &	operator-= (MatrixAsTriplet const &m)

MatrixAsTriplet &	operator*= (Scalar const &scalar)
	Scaling. More...

MatrixAsTriplet &	operator/= (Scalar const &scalar)
	Scaling. More...

MatrixAsTriplet &	operator= (MatrixAsTriplet const &m)=default
	Assignment. More...

MatrixAsTriplet &	operator= (MatrixAsTriplet &&other)=default
	Move assignment. More...

void	scaleRows (std::vector< Scalar >const &scaling)

MatrixAsTriplet &	transpose ()
	Transposition. More...

void	deleteLowerTriangle ()
	Deletes all sub-diagonal entries. More...

std::ostream &	print (std::ostream &out=std::cout, double threshold=0.0) const
	Output as Matlab source code. More...

void	toColumns (std::vector< std::vector< Scalar > > &colsvector) const
	transform into a set of column vectors More...

void	toRows (std::vector< std::vector< Scalar > > &rows) const

void	toVector (std::vector< Scalar > &colsvector)

void	addColumn (std::vector< Scalar > &colsvector, size_t position)
	add a column More...

template<class Mat >
void	addToMatrix (Mat &mat)
	add to a dense matrix More...

void	erase (SparseIndexInt i)

MatrixAsTriplet	lumped ()

iterator	begin ()
	get iterator on data field More...

const_iterator	begin () const
	get const iterator on data field More...

iterator	end ()
	get iterator on end of data field More...

const_iterator	end () const
	get const iterator on end of data field More...

size_t	size () const
	get number of entries in sparse matrix More...

void	toFile (const char *filename, unsigned int precision=10) const
	write to text file More...

template<int bcrsN = 1>
std::unique_ptr< Dune::BCRSMatrix< Dune::FieldMatrix< Scalar, bcrsN, bcrsN > > >	toBCRS () const
	copy to Dune::BCRSMatrix More...

bool	isSymmetric () const

std::pair< size_t, bool >	findEntry (SparseIndexInt row, SparseIndexInt col) const

Public Attributes
std::vector< SparseIndexInt >	ridx
	row indices More...

std::vector< SparseIndexInt >	cidx
	column indices More...

std::vector< Scalar >	data
	data More...

Member Typedef Documentation

◆ const_index_iterator

template<class Scalar , class SparseIndexInt = int>

typedef std::vector<SparseIndexInt>::const_iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::const_index_iterator

Definition at line 64 of file triplet.hh.

◆ const_iterator

template<class Scalar , class SparseIndexInt = int>

typedef std::vector<Scalar>::const_iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::const_iterator

Definition at line 62 of file triplet.hh.

◆ index_iterator

template<class Scalar , class SparseIndexInt = int>

typedef std::vector<SparseIndexInt>::iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::index_iterator

Definition at line 63 of file triplet.hh.

◆ iterator

template<class Scalar , class SparseIndexInt = int>

typedef std::vector<Scalar>::iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::iterator

Definition at line 61 of file triplet.hh.

◆ value_type

template<class Scalar , class SparseIndexInt = int>

typedef Scalar Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::value_type

STL-compliant typedefs.

Definition at line 60 of file triplet.hh.

Constructor & Destructor Documentation

◆ MatrixAsTriplet() [1/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( )

inline

Constructor.

Definition at line 67 of file triplet.hh.

◆ MatrixAsTriplet() [2/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( size_t s )

inlineexplicit

Constructor that allocates memory directly.

Definition at line 70 of file triplet.hh.

◆ MatrixAsTriplet() [3/11]

template<class Scalar , class SparseIndexInt = int>

template<int n, int m, class Allocator >

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( Dune::BCRSMatrix< Dune::FieldMatrix< Scalar, n, m >, Allocator > const & other )

inlineexplicit

Constructor from a Dune::BCRSMatrix.

Definition at line 74 of file triplet.hh.

◆ MatrixAsTriplet() [4/11]

template<class Scalar , class SparseIndexInt = int>

template<int n, int m, class Index >

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( NumaBCRSMatrix< Dune::FieldMatrix< Scalar, n, m >, Index > const & other )

inlineexplicit

Constructor from a NumaBCRSMatrix.

Definition at line 81 of file triplet.hh.

◆ MatrixAsTriplet() [5/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet	(	std::vector< SparseIndexInt > &&	row,
		std::vector< SparseIndexInt > &&	col,
		std::vector< Scalar > &&	value
	)

inline

Move constructor taking raw data.

The size of row, col, and value has to be identical. All entries in row and col need to be nonnegative. The supplied containers are empty after construction.

Definition at line 94 of file triplet.hh.

◆ MatrixAsTriplet() [6/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet	(	SparseIndexInt	nrows,
		SparseIndexInt	ncols,
		Scalar const &	value
	)

inline

Constructs a matrix of size (nrows,ncols), filled with value. If value==0.0, then it is an empty matrix.

Definition at line 107 of file triplet.hh.

◆ MatrixAsTriplet() [7/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet	(	SparseIndexInt	nrows,
		SparseIndexInt	ncols,
		SparseIndexInt	shift,
		Scalar const &	value
	)

inline

Constructs a diagonal matrix of size (nrows,ncols). The diagonal is shifted down shift rows (if shift is negative, it is shifted up)

Definition at line 124 of file triplet.hh.

◆ MatrixAsTriplet() [8/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet	(	SparseIndexInt	nrows,
		SparseIndexInt	ncols,
		SparseIndexInt	shift,
		std::vector< Scalar > const &	value
	)

inline

Constructs a diagonal matrix of size (nrows,ncols). The diagonal is shifted down shift rows (if shift is negative, it is shifted up)

Definition at line 143 of file triplet.hh.

◆ MatrixAsTriplet() [9/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( MatrixAsTriplet< Scalar, SparseIndexInt > const & other )

default

Copy constructor.

◆ MatrixAsTriplet() [10/11]

template<class Scalar , class SparseIndexInt = int>

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( MatrixAsTriplet< Scalar, SparseIndexInt > && other )

default

Move constructor.

◆ MatrixAsTriplet() [11/11]

template<class Scalar , class SparseIndexInt = int>

template<class OtherScalar , class OtherIndex >

Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::MatrixAsTriplet ( MatrixAsTriplet< OtherScalar, OtherIndex > const & A )

inline

Copy constructor from a triplet matrix with different entry and/or index types.

Definition at line 168 of file triplet.hh.

Member Function Documentation

◆ addColumn()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::addColumn	(	std::vector< Scalar > &	colsvector,
		size_t	position
	)

inline

add a column

Definition at line 536 of file triplet.hh.

◆ addEntry()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::addEntry	(	SparseIndexInt	row,
		SparseIndexInt	col,
		Scalar const &	value
	)

inline

Adds the given value to the specified matrix entry. The entry is created in case it has been structurally zero before.

Definition at line 349 of file triplet.hh.

Referenced by Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::lumped().

◆ addToMatrix()

template<class Scalar , class SparseIndexInt = int>

template<class Mat >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::addToMatrix ( Mat & mat )

inline

add to a dense matrix

Definition at line 549 of file triplet.hh.

◆ atxpy()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::atxpy	(	Y &	y,
		X const &	x,
		Scalar	alpha = `1.0`
	)		const

inline

Matrix-vector multiplication.

\( x = x + \alpha A^{T}y \)

Definition at line 283 of file triplet.hh.

Referenced by Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::usmtv().

◆ ax()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::ax	(	Y &	out,
		X const &	in
	)		const

inline

Matrix-vector multiplication: out = (*this) * in.

Definition at line 312 of file triplet.hh.

Referenced by Kaskade::Bridge::DirectInnerSolver< DirectSolver >::ax().

◆ axpy()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::axpy	(	Y &	out,
		X const &	in,
		Scalar	alpha = `1.0`,
		int	nvectors = `1`
	)		const

inline

Matrix-vector multiplication: out += alpha * (*this) * in.

?????: Was ist denn nr? number of blocks/rows?? Damit kann man mehrere Vektoren gleichzeitig axpy'en

Definition at line 262 of file triplet.hh.

Referenced by Kaskade::Bridge::KaskadeLinearization< Functional >::d2axpy(), Kaskade::Bridge::NormalStepLinearization< Functional, stateId, adjointId >::d2axpy(), Kaskade::Bridge::TangentialStepLinearization< Functional, stateId, adjointId >::d2axpy(), Kaskade::Bridge::KaskadeLinearization< Functional >::d2taxpy(), Kaskade::Bridge::NormalStepLinearization< Functional, stateId, adjointId >::ddxpy(), Kaskade::Bridge::TangentialStepLinearization< Functional, stateId, adjointId >::ddxpy(), Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::mv(), Kaskade::EnergyScalarProduct< X >::operator()(), and Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::usmv().

◆ begin() [1/2]

template<class Scalar , class SparseIndexInt = int>

iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::begin ( )

inline

get iterator on data field

Definition at line 586 of file triplet.hh.

◆ begin() [2/2]

template<class Scalar , class SparseIndexInt = int>

const_iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::begin ( ) const

inline

get const iterator on data field

Definition at line 589 of file triplet.hh.

◆ clear()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::clear ( )

inline

Deletes the data, leaving an empty matrix of size 0 x 0.

Definition at line 182 of file triplet.hh.

◆ deleteLowerTriangle()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::deleteLowerTriangle ( )

inline

Deletes all sub-diagonal entries.

Definition at line 461 of file triplet.hh.

Referenced by Kaskade::HigherOrderRecovery< Grid, Space >::getErrorFunction().

◆ end() [1/2]

template<class Scalar , class SparseIndexInt = int>

iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::end ( )

inline

get iterator on end of data field

Definition at line 592 of file triplet.hh.

◆ end() [2/2]

template<class Scalar , class SparseIndexInt = int>

const_iterator Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::end ( ) const

inline

get const iterator on end of data field

Definition at line 595 of file triplet.hh.

◆ erase()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::erase ( SparseIndexInt i )

inline

Definition at line 555 of file triplet.hh.

◆ findEntry()

template<class Scalar , class SparseIndexInt = int>

std::pair< size_t, bool > Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::findEntry	(	SparseIndexInt	row,
		SparseIndexInt	col
	)		const

inline

Returns: std::pair<size_t,bool> retVal. retVal.first: index. retVal.second: true, if entry was found

Definition at line 658 of file triplet.hh.

Referenced by Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::isSymmetric().

◆ isSymmetric()

template<class Scalar , class SparseIndexInt = int>

bool Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::isSymmetric ( ) const

inline

Definition at line 642 of file triplet.hh.

◆ lumped()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::lumped ( )

inline

Definition at line 562 of file triplet.hh.

◆ M()

template<class Scalar , class SparseIndexInt = int>

SparseIndexInt Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::M ( ) const

inline

Returns number of columns (computes them, if not known) This is a Dune ISTL compatibility method.

Definition at line 225 of file triplet.hh.

◆ mv()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::mv	(	X const &	x,
		Y &	y
	)		const

inline

matrix-vector multiplication

Definition at line 328 of file triplet.hh.

◆ N()

template<class Scalar , class SparseIndexInt = int>

SparseIndexInt Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::N ( ) const

inline

Returns number of rows (computes them, if not known) This is a Dune ISTL compatibility method.

Definition at line 212 of file triplet.hh.

◆ ncols()

template<class Scalar , class SparseIndexInt = int>

SparseIndexInt Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::ncols ( ) const

inline

Returns number of cols (computes them, if not known)

Definition at line 215 of file triplet.hh.

◆ nnz()

template<class Scalar , class SparseIndexInt = int>

size_t Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::nnz ( ) const

inline

Number of non-zeros. This is the number of entries in the triplet format and can be larger than the number of nonzero entries in the represented sparse matrix.

Definition at line 254 of file triplet.hh.

Referenced by Kaskade::AdditiveSchwarzPreconditioner< Op >::AdditiveSchwarzPreconditioner(), Kaskade::ARMSPreconditioner< Op >::ARMSPreconditioner(), Kaskade::BoomerAMG< Op >::BoomerAMG(), Kaskade::Euclid< Op >::Euclid(), Kaskade::getFactorization(), Kaskade::ILUKPreconditioner< Op >::ILUKPreconditioner(), Kaskade::ILUTPreconditioner< Op >::ILUTPreconditioner(), Kaskade::AdditiveSchwarzPreconditioner< Op >::init(), Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::lumped(), and Kaskade::PartialDirectPreconditioner< Op >::PartialDirectPreconditioner().

◆ nrows()

template<class Scalar , class SparseIndexInt = int>

SparseIndexInt Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::nrows ( ) const

inline

Returns number of rows (computes them, if not known)

Definition at line 202 of file triplet.hh.

◆ operator*=()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator*= ( Scalar const & scalar )

inline

Scaling.

Definition at line 426 of file triplet.hh.

◆ operator+=()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator+= ( MatrixAsTriplet< Scalar, SparseIndexInt > const & m )

inline

Matrix addition: (*this) += m (works also for matrices with non-matching sparsity pattern)

Definition at line 357 of file triplet.hh.

◆ operator-=()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator-= ( MatrixAsTriplet< Scalar, SparseIndexInt > const & m )

inline

Definition at line 389 of file triplet.hh.

◆ operator/=()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator/= ( Scalar const & scalar )

inline

Scaling.

Definition at line 435 of file triplet.hh.

◆ operator=() [1/2]

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator= ( MatrixAsTriplet< Scalar, SparseIndexInt > && other )

default

Move assignment.

◆ operator=() [2/2]

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::operator= ( MatrixAsTriplet< Scalar, SparseIndexInt > const & m )

default

Assignment.

◆ print()

template<class Scalar , class SparseIndexInt = int>

std::ostream & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::print	(	std::ostream &	out = `std::cout`,
		double	threshold = `0.0`
	)		const

inline

Output as Matlab source code.

Definition at line 490 of file triplet.hh.

Referenced by Kaskade::operator<<().

◆ reserve()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::reserve ( size_t s )

inline

Reserve memory.

Definition at line 242 of file triplet.hh.

◆ resize()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::resize ( size_t s )

inline

Resizes the memory.

Existing values are preserved if s >= nnz(), and the new entries are undefined. For s<nnz(), all the values are undefined.

Definition at line 234 of file triplet.hh.

Referenced by Kaskade::BlockLUFactorization< Factorization >::BlockLUFactorization().

◆ scaleRows()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::scaleRows ( std::vector< Scalar >const & scaling )

inline

Definition at line 446 of file triplet.hh.

◆ setStartToZero()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::setStartToZero ( )

inline

Shift the row and column indices such that \( A_{00} \) is (structurally) nonzero.

Definition at line 193 of file triplet.hh.

◆ shiftIndices()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::shiftIndices	(	SparseIndexInt	row,
		SparseIndexInt	col
	)

inline

Shifts matrix indices. Can be used together with += to concatenate sparese matrices.

Definition at line 334 of file triplet.hh.

Referenced by Kaskade::Bridge::KaskadeLinearization< Functional >::getDiscreteMatrixBlocks(), Kaskade::Bridge::NormalStepLinearization< Functional, stateId, adjointId >::getMatrixBlocks(), Kaskade::Bridge::TangentialStepLinearization< Functional, stateId, adjointId >::getMatrixBlocks(), Kaskade::LagrangeOperator< NormalStepAssembler, TangentialStepAssembler, stateId, controlId, adjointId >::getTriplet(), Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::setStartToZero(), and Kaskade::Bridge::DirectInnerSolver< DirectSolver >::solveAdjAndNormal().

◆ size()

template<class Scalar , class SparseIndexInt = int>

size_t Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::size ( ) const

inline

get number of entries in sparse matrix

Definition at line 598 of file triplet.hh.

Referenced by Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toBCRS().

◆ toBCRS()

template<class Scalar , class SparseIndexInt = int>

template<int bcrsN = 1>

std::unique_ptr< Dune::BCRSMatrix< Dune::FieldMatrix< Scalar, bcrsN, bcrsN > > > Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toBCRS ( ) const

inline

copy to Dune::BCRSMatrix

Definition at line 613 of file triplet.hh.

◆ toColumns()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toColumns ( std::vector< std::vector< Scalar > > & colsvector ) const

inline

transform into a set of column vectors

Definition at line 500 of file triplet.hh.

◆ toFile()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toFile	(	const char *	filename,
		unsigned int	precision = `10`
	)		const

inline

write to text file

Definition at line 601 of file triplet.hh.

◆ toRows()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toRows ( std::vector< std::vector< Scalar > > & rows ) const

inline

Definition at line 513 of file triplet.hh.

◆ toVector()

template<class Scalar , class SparseIndexInt = int>

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::toVector ( std::vector< Scalar > & colsvector )

inline

Definition at line 526 of file triplet.hh.

◆ transpose()

template<class Scalar , class SparseIndexInt = int>

MatrixAsTriplet & Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::transpose ( )

inline

Transposition.

Definition at line 453 of file triplet.hh.

Referenced by Kaskade::Bridge::KaskadeLinearization< Functional >::d2taxpy(), and Kaskade::HierarchicalBasisErrorEstimator2< Functional, VariableSetDescription, ExtensionVariableSetDescription, ExtensionSpace, NormFunctional, LinearSolverLA, LinearSolverHA, LinearSolverLM, LinearSolverHM, lumpM, RefinementStrategy >::operator()().

◆ usmtv()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::usmtv	(	Scalar const	alpha,
		X const &	x,
		Y &	y
	)		const

inline

Matrix-vector multiplication.

\( x = x + \alpha A^{T}y \)

Definition at line 296 of file triplet.hh.

Referenced by Kaskade::HierarchicalBasisErrorEstimator2< Functional, VariableSetDescription, ExtensionVariableSetDescription, ExtensionSpace, NormFunctional, LinearSolverLA, LinearSolverHA, LinearSolverLM, LinearSolverHM, lumpM, RefinementStrategy >::operator()().

◆ usmv()

template<class Scalar , class SparseIndexInt = int>

template<class X , class Y >

void Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::usmv	(	Scalar const	alpha,
		X const &	x,
		Y &	y
	)		const

inline

scaled matrix-vector multiplication

Definition at line 304 of file triplet.hh.

Referenced by Kaskade::HierarchicalBasisErrorEstimator2< Functional, VariableSetDescription, ExtensionVariableSetDescription, ExtensionSpace, NormFunctional, LinearSolverLA, LinearSolverHA, LinearSolverLM, LinearSolverHM, lumpM, RefinementStrategy >::operator()().

Member Data Documentation

◆ cidx

template<class Scalar , class SparseIndexInt = int>

std::vector<SparseIndexInt> Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::cidx

column indices

Definition at line 671 of file triplet.hh.

◆ data

template<class Scalar , class SparseIndexInt = int>

std::vector<Scalar> Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::data

data

Definition at line 673 of file triplet.hh.

◆ ridx

template<class Scalar , class SparseIndexInt = int>

std::vector<SparseIndexInt> Kaskade::MatrixAsTriplet< Scalar, SparseIndexInt >::ridx

row indices

Definition at line 669 of file triplet.hh.

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

Detailed Description

Public Types

Public Member Functions

Public Attributes

Member Typedef Documentation

◆ const_index_iterator

◆ const_iterator

◆ index_iterator

◆ iterator

◆ value_type

Constructor & Destructor Documentation

◆ MatrixAsTriplet() [1/11]

◆ MatrixAsTriplet() [2/11]

◆ MatrixAsTriplet() [3/11]

◆ MatrixAsTriplet() [4/11]

◆ MatrixAsTriplet() [5/11]

◆ MatrixAsTriplet() [6/11]

◆ MatrixAsTriplet() [7/11]

◆ MatrixAsTriplet() [8/11]

◆ MatrixAsTriplet() [9/11]

◆ MatrixAsTriplet() [10/11]

◆ MatrixAsTriplet() [11/11]

Member Function Documentation

◆ addColumn()

◆ addEntry()

◆ addToMatrix()

◆ atxpy()

◆ ax()

◆ axpy()

◆ begin() [1/2]

◆ begin() [2/2]

◆ clear()

◆ deleteLowerTriangle()

◆ end() [1/2]

◆ end() [2/2]

◆ erase()

◆ findEntry()

◆ isSymmetric()

◆ lumped()

◆ M()

◆ mv()

◆ N()

◆ ncols()

◆ nnz()

◆ nrows()

◆ operator*=()

◆ operator+=()

◆ operator-=()

◆ operator/=()

◆ operator=() [1/2]

◆ operator=() [2/2]

◆ print()

◆ reserve()

◆ resize()

◆ scaleRows()

◆ setStartToZero()

◆ shiftIndices()

◆ size()

◆ toBCRS()

◆ toColumns()

◆ toFile()

◆ toRows()

◆ toVector()

◆ transpose()

◆ usmtv()

◆ usmv()

Member Data Documentation

◆ cidx

◆ data

◆ ridx