Hierarchical Bi-directional Incremental Compressed Row Storage scheme. More...

#include <HBICRS.hpp>

Inheritance diagram for HBICRS< _t_value >:

Collaboration diagram for HBICRS< _t_value >:

Public Member Functions
	~HBICRS ()
	Base deconstructor. More...

	HBICRS ()
	Base constructor. More...

	HBICRS (std::string file, _t_value zero=0)
	Base constructor. More...

	HBICRS (std::vector< Triplet< _t_value > > &input, LI m, LI n, _t_value zero)
	Base constructor. More...

	HBICRS (std::vector< std::vector< Triplet< _t_value > > > &input, signed char *group_type, LI m, LI n, _t_value zero=0)
	Base constructor. More...

virtual void	load (std::vector< Triplet< _t_value > > &input, LI m, LI n, _t_value zero)
	This function will rewrite the std::vector< Triplet > structure to one suitable for the other load function. More...

virtual void	load (std::vector< std::vector< Triplet< _t_value > > > &input, signed char *group_type, LI m, LI n, _t_value zero)
	Constructs the hierarchical part. More...

void	load (LI row, LI col, LI m, LI n, LI nb)
	Builds the BICRS structure. More...

virtual void	getFirstIndexPair (LI &i, LI &j)

virtual void	zxa (const _t_value __restrict__ x_p, _t_value __restrict__ y_p)
	Calculates y=xA, but does not allocate y itself. More...

virtual void	zax (const _t_value __restrict__ x_p, _t_value __restrict__ y_p)
	Calculates y=Ax, but does not allocate y itself. More...

virtual size_t	bytesUsed ()
	Function to query the amount of storage required by this sparse matrix. More...

Public Member Functions inherited from SparseMatrix< _t_value, LI >
	SparseMatrix ()
	Base constructor. More...

	SparseMatrix (const LInzs, const LInr, const LInc, const _t_valuezero)
	Base constructor. More...

virtual	~SparseMatrix ()
	Base deconstructor. More...

void	loadFromFile (const std::string file, const _t_valuezero=0)
	Function which loads a matrix from a matrix market file. More...

virtual unsigned long int	m ()
	Queries the number of rows this matrix contains. More...

virtual unsigned long int	n ()
	Queries the number of columns this matrix contains. More...

virtual unsigned long int	nzs ()
	Queries the number of nonzeroes stored in this matrix. More...

virtual _t_value *	mv (const _t_value *x)
	Calculates and returns z=Ax. More...

Public Member Functions inherited from Matrix< _t_value >
	Matrix ()
	Base constructor. More...

virtual	~Matrix ()
	Base deconstructor. More...

virtual void	zax (const _t_value __restrict__ x, _t_value __restrict__ z, const size_t k, const clockid_t clock_id=0, double *elapsed_time=NULL)
	Wrapper function to call the zax kernel multiple times successively, while timing the duration of the operation. More...

void	ZaX (const _t_value __restrict__ const __restrict__ const X, _t_value __restrict__ const __restrict__ const Z)
	In-place Z=AX function, where A is m x n, Z = m x k, and X is n x k. More...

virtual void	zxa (const _t_value __restrict__ x, _t_value __restrict__ z, const unsigned long int repeat, const clockid_t clock_id=0, double *elapsed_time=NULL)
	Wrapper function to call the zxa kernel multiple times successively, while timing the operation duration. More...

void	ZXa (const _t_value __restrict__ const __restrict__ const X, _t_value __restrict__ const __restrict__ const Z)
	In-place Z=XA function, where A is m x n, Z = k x n, and X is k x m. More...

Protected Attributes
size_t	jumps
	The number of row jumps. More...

LI *	r_inc
	Stores the row jumps; size is at maximum the number of nonzeros. More...

LI *	c_inc
	Stores the column jumps; size is exactly the number of nonzeros. More...

Matrix< _t_value > **	vals
	Stores the values individual storage schemes. More...

ULI	ntt
	Caches n times two. More...

Protected Attributes inherited from SparseMatrix< _t_value, LI >
LI	nor
	Number of rows. More...

LI	noc
	Number of columns.

LI	nnz
	Number of non-zeros. More...

Additional Inherited Members
Public Attributes inherited from SparseMatrix< _t_value, LI >
_t_value	zero_element
	The element considered to be zero. More...

Detailed Description

template<typename _t_value>
class HBICRS< _t_value >

Hierarchical Bi-directional Incremental Compressed Row Storage scheme.

Stores other SparseMatrix data structures, can include other HBICRS schemes, but this is not recommended with regards to efficiency. Supports jumping back and forward within columns. Supports jumping back and forward between rows. Main storage direction in column-wise. Storage requirements are 2nz plus the number of row jumps required, plus the storage requirements for each stored data structure, of course. Many row jumps are disadvantageous to storage as well as speed.

Parameters

_t_value The type of the nonzeros in the matrix.

This class is based on the BICRS as per revision 75. BICRS was chosen since it is an all-round improvement over Triplets, with the additional advantage of being pointer based (no jumps required when going recursive into deeper storage schemes).

Warning: this class uses assertions! For optimal performance, define the NDEBUG flag (e.g., pass -DNDEBUG as a compiler flag).

Constructor & Destructor Documentation

template<typename _t_value>

HBICRS< _t_value >::~HBICRS ( )

inline

Base deconstructor.

template<typename _t_value>

HBICRS< _t_value >::HBICRS ( )

inline

Base constructor.

template<typename _t_value>

HBICRS< _t_value >::HBICRS	(	std::string	file,
		_t_value	zero = `0`
	)

inline

Base constructor.

Will read in from Matrix Market file.

See also: SparseMatrix::SparseMatrix( file, zero )

template<typename _t_value>

HBICRS< _t_value >::HBICRS	(	std::vector< Triplet< _t_value > > &	input,
		LI	m,
		LI	n,
		_t_value	zero
	)

inline

Base constructor.

See also: SparseMatrix::SparseMatrix( input, m, n, zero )

template<typename _t_value>

HBICRS< _t_value >::HBICRS	(	std::vector< std::vector< Triplet< _t_value > > > &	input,
		signed char *	group_type,
		LI	m,
		LI	n,
		_t_value	zero = `0`
	)

inline

Base constructor.

See also: SparseMatrix::SparseMatrix( input, m, n, zero )

Member Function Documentation

template<typename _t_value>

virtual size_t HBICRS< _t_value >::bytesUsed ( )

inlinevirtual

Function to query the amount of storage required by this sparse matrix.

Returns: The size of the sparse matrix in bytes.

Implements Matrix< _t_value >.

Referenced by HBICRS< T >::bytesUsed(), and BlockHilbert< T >::bytesUsed().

template<typename _t_value>

virtual void HBICRS< _t_value >::getFirstIndexPair	(	LI &	i,
		LI &	j
	)

inlinevirtual

See also: SparseMatrix::getFirstIndexPair

Implements SparseMatrix< _t_value, LI >.

Referenced by BlockHilbert< T >::getFirstIndexPair().

template<typename _t_value>

virtual void HBICRS< _t_value >::load	(	std::vector< Triplet< _t_value > > &	input,
		LI	m,
		LI	n,
		_t_value	zero
	)

inlinevirtual

This function will rewrite the std::vector< Triplet > structure to one suitable for the other load function.

See also: load( row, col, val, m, n, nz ); SparseMatrix::load

Implements SparseMatrix< _t_value, LI >.

Referenced by HBICRS< T >::HBICRS(), and HBICRS< T >::load().

template<typename _t_value>

virtual void HBICRS< _t_value >::load	(	std::vector< std::vector< Triplet< _t_value > > > &	input,
		signed char *	group_type,
		LI	m,
		LI	n,
		_t_value	zero
	)

inlinevirtual

Constructs the hierarchical part.

Note that this function does not do anything with offsets, due to lack of support in the other datastructures used.

template<typename _t_value>

void HBICRS< _t_value >::load	(	LI *	row,
		LI *	col,
		LI	m,
		LI	n,
		LI	nb
	)

inline

Builds the BICRS structure.

template<typename _t_value>

virtual void HBICRS< _t_value >::zax	(	const _t_value *__restrict__	x_p,
		_t_value *__restrict__	y_p
	)

inlinevirtual

Calculates y=Ax, but does not allocate y itself.

Parameters

x_p	The input vector should be initialised and of correct measurements.
y_p	The output vector should be preallocated and of size m. Furthermore, y[i]=0 for all i, 0<=i<m.

Implements SparseMatrix< _t_value, LI >.

Referenced by BlockHilbert< T >::zax().

template<typename _t_value>

virtual void HBICRS< _t_value >::zxa	(	const _t_value *__restrict__	x_p,
		_t_value *__restrict__	y_p
	)

inlinevirtual

Calculates y=xA, but does not allocate y itself.

Parameters

x_p	The input vector should be initialised and of correct measurements.
y_p	The output vector should be preallocated and of size m. Furthermore, y[i]=0 for all i, 0<=i<m.

Implements SparseMatrix< _t_value, LI >.

Referenced by BlockHilbert< T >::zxa().