Add/Update Operations

Operations for summing or updating existing containers. More...

Functions
template<typename T_Scalar>
void	culite::ops::add (T_Scalar alpha, const dns::XxVector< T_Scalar > &x, T_Scalar beta, const dns::XxVector< T_Scalar > &y, dns::XxVector< T_Scalar > &z, CuBlasHandler &cublasHandler=globalCuBlasHandler())
	Adds two compatible scaled dense vectors.
template<typename T_Scalar>
void	culite::ops::add (op_t opA, T_Scalar alpha, const dns::XxMatrix< T_Scalar > &A, op_t opB, T_Scalar beta, const dns::XxMatrix< T_Scalar > &B, dns::XxMatrix< T_Scalar > &C, CuBlasHandler &cublasHandler=globalCuBlasHandler())
	Adds two compatible scaled dense matrices with optional transpose operations.
template<typename T_Int, typename T_Scalar>
csr::XxMatrix< T_Int, T_Scalar >	culite::ops::add (T_Scalar alpha, const csr::XxMatrix< T_Int, T_Scalar > &A, T_Scalar beta, const csr::XxMatrix< T_Int, T_Scalar > &B, CuSparseHandler &cusparseHandler=globalCuSparseHandler())
	Adds two compatible sparse matrices in CSR format.
template<typename T_Int, typename T_Scalar>
csc::XxMatrix< T_Int, T_Scalar >	culite::ops::add (T_Scalar alpha, const csc::XxMatrix< T_Int, T_Scalar > &A, T_Scalar beta, const csc::XxMatrix< T_Int, T_Scalar > &B, CuSparseHandler &cusparseHandler=globalCuSparseHandler())
	Adds two compatible sparse matrices in CSC format.
template<typename T_Scalar>
void	culite::ops::update (T_Scalar alpha, const dns::XxVector< T_Scalar > &x, dns::XxVector< T_Scalar > &y, CuBlasHandler &cublasHandler=globalCuBlasHandler())
	Update a dense vector with a compatible scaled dense vector.
template<typename T_Scalar>
void	culite::ops::update (op_t opA, T_Scalar alpha, const dns::XxMatrix< T_Scalar > &A, dns::XxMatrix< T_Scalar > &B, CuBlasHandler &cublasHandler=globalCuBlasHandler())
	Update a dense matrix with a compatible scaled dense matrix with optional transposition.
template<typename T_Scalar>
void	culite::ops::update (T_Scalar alpha, const dns::XxMatrix< T_Scalar > &A, dns::XxMatrix< T_Scalar > &B, CuBlasHandler &cublasHandler=globalCuBlasHandler())
	Update a dense matrix with a compatible scaled dense matrix.
template<typename T_Int, typename T_Scalar>
void	culite::ops::update (T_Scalar alpha, const csr::XxMatrix< T_Int, T_Scalar > &A, csr::XxMatrix< T_Int, T_Scalar > &B, CuSparseHandler &cuSparseHandler=globalCuSparseHandler())
	Update a sparse matrix with a compatible scaled sparse matrix.
template<typename T_Int, typename T_Scalar>
void	culite::ops::update (T_Scalar alpha, const csc::XxMatrix< T_Int, T_Scalar > &A, csc::XxMatrix< T_Int, T_Scalar > &B, CuSparseHandler &cuSparseHandler=globalCuSparseHandler())
	Update a sparse matrix with a compatible scaled sparse matrix.

Detailed Description

Operations for summing or updating existing containers.

Function Documentation

add() [1/4]

template<typename T_Scalar>

void culite::ops::add	(	T_Scalar	alpha,
		const dns::XxVector< T_Scalar > &	x,
		T_Scalar	beta,
		const dns::XxVector< T_Scalar > &	y,
		dns::XxVector< T_Scalar > &	z,
		CuBlasHandler &	cublasHandler = globalCuBlasHandler() )

Adds two compatible scaled dense vectors.

Performs the operation \( \alpha \cdot x + \beta \cdot y \).

Template Parameters

T_Scalar

The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient for x.
[in]	x	The first input dense vector.
[in]	beta	The scaling coefficient for y.
[in]	y	The second input dense vector.
[out]	z	The output dense vector.
[in]	cublasHandler	The cuBLAS handler for GPU operations (defaults to global handler).

Examples

ex02l_dense_matrix_algebra_add.cpp, and ex06l_sparse_matrix_algebra_add.cpp.

add() [2/4]

template<typename T_Scalar>

void culite::ops::add	(	op_t	opA,
		T_Scalar	alpha,
		const dns::XxMatrix< T_Scalar > &	A,
		op_t	opB,
		T_Scalar	beta,
		const dns::XxMatrix< T_Scalar > &	B,
		dns::XxMatrix< T_Scalar > &	C,
		CuBlasHandler &	cublasHandler = globalCuBlasHandler() )

Adds two compatible scaled dense matrices with optional transpose operations.

Performs the operation \( \alpha \cdot op_A(A) + \beta \cdot op_B(B) \), where \( op_A \) and \( op_B \) can be no-transpose, transpose, or conjugate transpose.

Template Parameters

T_Scalar

The scalar type (e.g., float, double, complex).

Parameters

[in]	opA	The operation to apply to matrix A (no-transpose, transpose, or conjugate transpose).
[in]	alpha	The scaling coefficient for A.
[in]	A	The first input dense matrix.
[in]	opB	The operation to apply to matrix B (no-transpose, transpose, or conjugate transpose).
[in]	beta	The scaling coefficient for B.
[in]	B	The second input dense matrix.
[out]	C	The output dense matrix.
[in]	cublasHandler	The cuBLAS handler for GPU operations (defaults to global handler).

add() [3/4]

template<typename T_Int, typename T_Scalar>

csr::XxMatrix< T_Int, T_Scalar > culite::ops::add	(	T_Scalar	alpha,
		const csr::XxMatrix< T_Int, T_Scalar > &	A,
		T_Scalar	beta,
		const csr::XxMatrix< T_Int, T_Scalar > &	B,
		CuSparseHandler &	cusparseHandler = globalCuSparseHandler() )

Adds two compatible sparse matrices in CSR format.

Performs the operation \( \alpha \cdot A + \beta \cdot B \) on GPU using cuSPARSE.

Template Parameters

T_Int	The integer type for indexing.
T_Scalar	The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient for A.
[in]	A	The first input sparse matrix in CSR format.
[in]	beta	The scaling coefficient for B.
[in]	B	The second input sparse matrix in CSR format.
[in]	cusparseHandler	The cuSPARSE handler for GPU operations (defaults to global handler).

Returns

The result of the operation \( \alpha \cdot A + \beta \cdot B \) as a CSR matrix.

add() [4/4]

template<typename T_Int, typename T_Scalar>

csc::XxMatrix< T_Int, T_Scalar > culite::ops::add	(	T_Scalar	alpha,
		const csc::XxMatrix< T_Int, T_Scalar > &	A,
		T_Scalar	beta,
		const csc::XxMatrix< T_Int, T_Scalar > &	B,
		CuSparseHandler &	cusparseHandler = globalCuSparseHandler() )

Adds two compatible sparse matrices in CSC format.

Performs the operation \( \alpha \cdot A + \beta \cdot B \) on GPU using cuSPARSE.

Template Parameters

T_Int	The integer type for indexing.
T_Scalar	The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient for A.
[in]	A	The first input sparse matrix in CSC format.
[in]	beta	The scaling coefficient for B.
[in]	B	The second input sparse matrix in CSC format.
[in]	cusparseHandler	The cuSPARSE handler for GPU operations (defaults to global handler).

Returns

The result of the operation \( \alpha \cdot A + \beta \cdot B \) as a CSC matrix.

update() [1/5]

template<typename T_Scalar>

void culite::ops::update	(	T_Scalar	alpha,
		const dns::XxVector< T_Scalar > &	x,
		dns::XxVector< T_Scalar > &	y,
		CuBlasHandler &	cublasHandler = globalCuBlasHandler() )

Update a dense vector with a compatible scaled dense vector.

Performs the operation \( y = y + \alpha \cdot x \).

Template Parameters

T_Scalar

The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient.
[in]	x	The input dense vector.
[in,out]	y	The dense vector to be updated.
[in]	cublasHandler	The cuBLAS handler for GPU operations (defaults to global handler).

Examples

ex02l_dense_matrix_algebra_add.cpp, and ex06l_sparse_matrix_algebra_add.cpp.

update() [2/5]

template<typename T_Scalar>

void culite::ops::update	(	op_t	opA,
		T_Scalar	alpha,
		const dns::XxMatrix< T_Scalar > &	A,
		dns::XxMatrix< T_Scalar > &	B,
		CuBlasHandler &	cublasHandler = globalCuBlasHandler() )

Update a dense matrix with a compatible scaled dense matrix with optional transposition.

Performs the operation \( B = B + \alpha \cdot op(A) \), where \( op(A) \) can be \( A \), \( A^T \), or \( A^H \).

Template Parameters

T_Scalar

The scalar type (e.g., float, double, complex).

Parameters

[in]	opA	The operation to apply to matrix A (N for none, T for transpose, H for conjugate transpose).
[in]	alpha	The scaling coefficient.
[in]	A	The input dense matrix.
[in,out]	B	The dense matrix to be updated.
[in]	cublasHandler	The cuBLAS handler for GPU operations (defaults to global handler).

update() [3/5]

template<typename T_Scalar>

void culite::ops::update	(	T_Scalar	alpha,
		const dns::XxMatrix< T_Scalar > &	A,
		dns::XxMatrix< T_Scalar > &	B,
		CuBlasHandler &	cublasHandler = globalCuBlasHandler() )

Update a dense matrix with a compatible scaled dense matrix.

Performs the operation \( B = B + \alpha \cdot A \).

Template Parameters

T_Scalar

The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient.
[in]	A	The input dense matrix.
[in,out]	B	The dense matrix to be updated.
[in]	cublasHandler	The cuBLAS handler for GPU operations (defaults to global handler).

update() [4/5]

template<typename T_Int, typename T_Scalar>

void culite::ops::update	(	T_Scalar	alpha,
		const csr::XxMatrix< T_Int, T_Scalar > &	A,
		csr::XxMatrix< T_Int, T_Scalar > &	B,
		CuSparseHandler &	cuSparseHandler = globalCuSparseHandler() )

Update a sparse matrix with a compatible scaled sparse matrix.

Performs the operation \( B = B + \alpha \cdot A \).

Template Parameters

T_Int	The integer type for indexing.
T_Scalar	The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient.
[in]	A	The input sparse matrix.
[in,out]	B	The sparse matrix to be updated.
[in]	cuSparseHandler	The cuSPARSE handler for GPU operations (defaults to global handler, currently unused).

Warning

This function currently does not utilize the cuSparseHandler and may not be optimized for performance. Use with caution for large sparse matrices as it allocates memory on the device internally.

update() [5/5]

template<typename T_Int, typename T_Scalar>

void culite::ops::update	(	T_Scalar	alpha,
		const csc::XxMatrix< T_Int, T_Scalar > &	A,
		csc::XxMatrix< T_Int, T_Scalar > &	B,
		CuSparseHandler &	cuSparseHandler = globalCuSparseHandler() )

Update a sparse matrix with a compatible scaled sparse matrix.

Performs the operation \( B = B + \alpha \cdot A \).

Template Parameters

T_Int	The integer type for indexing.
T_Scalar	The scalar type (e.g., float, double, complex).

Parameters

[in]	alpha	The scaling coefficient.
[in]	A	The input sparse matrix.
[in,out]	B	The sparse matrix to be updated.
[in]	cuSparseHandler	The cuSPARSE handler for GPU operations (defaults to global handler, currently unused).

Warning

This function currently does not utilize the cuSparseHandler and may not be optimized for performance. Use with caution for large sparse matrices as it allocates memory on the device internally.