kml_sparse_?csrmm

Compute the product of matrices. One of them is a sparse matrix in the CSR format. The operation is defined as follows:

C = alpha * A*B + beta * C
C = alpha * A^T * B + beta * C
C = alpha * A^H * B + beta * C

B and C are dense matrices, and A is a sparse matrix (m x k) stored in the CSR format.

Table 1 describes the relationship between layout and sparse matrix indexing.

**Table 1** Relationship between layout and sparse matrix indexing
Sparse matrix indexing	Dense matrix layout
KML_SPARSE_INDEX_BASE_ZERO	KML_SPARSE_LAYOUT_ROW_MAJOR
KML_SPARSE_INDEX_BASE_ONE	KML_SPARSE_LAYOUT_COLUMN_MAJOR

Interface Definition

C interface:

kml_sparse_status_t kml_sparse_scsrmm(const kml_sparse_operation_t opt, const KML_INT m, const KML_INT n, const KML_INT k, const float alpha, const char *matdescra,

const float *val, const KML_INT *indx, const KML_INT *pntrb, const KML_INT *pntre, const float *b, const KML_INT ldb, const float beta , float *c , const KML_INT ldc);

kml_sparse_status_t kml_sparse_dcsrmm(const kml_sparse_operation_t opt, const KML_INT m, const KML_INT n, const KML_INT k, const double alpha, const char *matdescra,

const double *val, const KML_INT *indx, const KML_INT *pntrb, const KML_INT *pntre, const double *b, const KML_INT ldb, const double beta , double *c , const KML_INT ldc);

kml_sparse_status_t kml_sparse_ccsrmm(const kml_sparse_operation_t opt, const KML_INT m, const KML_INT n, const KML_INT k, const KML_Complex8 alpha, const char *matdescra, const KML_Complex8 *val, const KML_INT *indx, const KML_INT *pntrb, const KML_INT *pntre, const KML_Complex8 *b, const KML_INT ldb, const KML_Complex8 beta , KML_Complex8 *c , const KML_INT ldc);

kml_sparse_status_t kml_sparse_zcsrmm(const kml_sparse_operation_t opt, const KML_INT m, const KML_INT n, const KML_INT k, const KML_Complex16 alpha, const char *matdescra, const KML_Complex16 *val, const KML_INT *indx, const KML_INT *pntrb, const KML_INT *pntre, const KML_Complex16 *b, const KML_INT ldb, const KML_Complex16 beta , KML_Complex16 *c , const KML_INT ldc);

Fortran interface:

RES = KML_SPARSE_SCSRMM(OPT, M, N K, ALPHA, MATDESCRA, VAL, INDX, PNTRB, PNTRE, B, LDB, BETA, C, LDC);

RES = KML_SPARSE_DCSRMM(OPT, M, N K, ALPHA, MATDESCRA, VAL, INDX, PNTRB, PNTRE, B, LDB, BETA, C, LDC);

RES = KML_SPARSE_CCSRMM(OPT, M, N K, ALPHA, MATDESCRA, VAL, INDX, PNTRB, PNTRE, B, LDB, BETA, C, LDC);

RES = KML_SPARSE_ZCSRMM(OPT, M, N K, ALPHA, MATDESCRA, VAL, INDX, PNTRB, PNTRE, B, LDB, BETA, C, LDC);

Parameters

Parameter	Type	Description	Input/Output
opt	Enumeration type kml_sparse_operation_t	Indicates whether to transpose. If opt = KML_SPARSE_OPERATION_NON_TRANSPOSE, then C = alphaAB + betaC. If opt = KML_SPARSE_OPERATION_TRANSPOSE, then C = alphaA^TB + betaC. If opt = KML_SPARSE_OPERATION_CONJUGATE_TRANSPOSE, then C = alphaA^HB + beta*C.	Input
m	Integer	Number of rows in matrix A. The value range is [1, MAX_KML_INT].	Input
n	Integer	Number of columns in matrix C. The value range is [1, MAX_KML_INT].	Input
k	Integer	Number of columns in matrix A. The value range is [1, MAX_KML_INT].	Input
alpha	For scsrmm, alpha is of the single-precision floating-point type. For dcsrmm, alpha is of the double-precision floating-point type. For ccsrmm, alpha is a single-precision complex number. For zcsrmm, alpha is a double-precision complex number.	Scalar alpha	Input
matdescra	Char pointer	Matrix operation attribute. For details, see the description of matdescra.	Input
val	For scsrmm, val is a single-precision floating-point array. For dcsrmm, val is a double-precision floating-point array. For ccsrmm, val is a single-precision complex number array. For zcsrmm, val is a double-precision complex number array.	Array values storing non-zero elements of matrix A in the CSR format. The length is pntre[m-1] - pntrb[0].	Input
indx	Integer array	Array columns in the CSR format, which contains the column indices for non-zero elements in matrix A	Input
pntrb	Integer array	Array of length m, containing row indices of matrix A. pntrb[i] - pntrb[0] indicates the subscript of the first non-zero element in row i in the val and indx arrays.	Input
pntre	Integer array	Array of length m, containing row indices of matrix A. pntre[i] - pntrb[0]-1 indicates the subscript of the last non-zero element in row i in the val and indx arrays.	Input
b	For scsrmm, b is a single-precision floating-point array. For dcsrmm, b is a double-precision floating-point array. For ccsrmm, b is a single-precision complex number array. For zcsrmm, b is a double-precision complex number array.	Array value of matrix B	Input
ldb	Integer	Size of the leading dimension of matrix B for one-based indexing Size of the second dimension of matrix B for zero-based indexing	Input
beta	For scsrmm, beta is of the single-precision floating-point type. For dcsrmm, beta is of the double-precision floating-point type. For ccsrmm, beta is a single-precision complex number. For zcsrmm, beta is a double-precision complex number.	Scalar beta	Input
c	For scsrmm, c is a single-precision floating-point array. For dcsrmm, c is a double-precision floating-point array. For ccsrmm, c is a single-precision complex number array. For zcsrmm, c is a double-precision complex number array.	Array value of matrix C	Input/Output
ldc	Integer	Size of the leading dimension of matrix C for one-based indexing Size of the second dimension of matrix C for zero-based indexing	Input

Table 2 describes the parameter constraints on matrix B.

**Table 2** Parameter constraints on matrix B
opt	Matrix Scale	Data Layout	Value Range
op(A) = A	k x n	Row-major order	k * ldbMAX_KML_INT
op(A) = A	k x n	Column-major order	ldb * n MAX_KML_INT
op(A) = A^T or A^H	m x n	Row-major order	m * ldbMAX_KML_INT
op(A) = A^T or A^H	m x n	Column-major order	ldb * n MAX_KML_INT

Table 3 describes the parameter constraints on matrix C.

**Table 3** Parameter constraints on matrix C
opt	Matrix Scale	Data Layout	Value Range
op(A) = A	m x n	Row-major order	m * ldcMAX_KML_INT
op(A) = A	m x n	Column-major order	ldc * n MAX_KML_INT
op(A) = A^T or A^H	k x n	Row-major order	k * ldc MAX_KML_INT
op(A) = A^T or A^H	k x n	Column-major order	ldc * n MAX_KML_INT

The function does not verify the integrity of parameters. Ensure that the elements in pntrb and pntre do not exceed the maximum index value of the input matrix.

Return Value

Function execution status. The enumeration type is kml_sparse_status_t.

Dependencies

C: "kspblas.h"

Fortran: "kspblas.f03"

Examples

C interface:

    kml_sparse_operation_t opt = KML_SPARSE_OPERATION_NON_TRANSPOSE; 
    KML_INT m = 3; 
    KML_INT n = 3; 
    KML_INT k = 3; 
    float alpha = 1.0; 
    float beta = 1.0; 
    char *matdescra = "G00C"; // General matrix with zero-based indexing
    float val[4] = {9, 8, 5, 2}; 
    KML_INT indx[4] = {2, 1, 2, 2}; 
    KML_INT pntrb[3] = {0, 2, 3}; 
    KML_INT pntre[3] = {2, 3, 4}; 
    float b[9] = {1, 5, 7, 4, 7, 7, 3, 3, 7}; 
    float c[9] = {0, 8, 8, 0, 2, 8, 3, 5, 6}; 
    KML_INT ldb = 3; 
    KML_INT ldc = 3; 
    kml_sparse_status_t status = kml_sparse_scsrmm(opt, m, n, k, alpha, matdescra, val, indx, pntrb, pntre, b, ldb, beta, c, ldc); 
    /* 
     *  Output c: 
     *     59.00  91.00  127.00  15.00  17.00  43.00  9.00  11.00  20.00 
     * 
     */

Fortran interface:

    INTEGER(C_INT) :: OPT = KML_SPARSE_OPERATION_NON_TRANSPOSE 
    INTEGER(C_INT) :: M = 3 
    INTEGER(C_INT) :: N = 3 
    INTEGER(C_INT) :: K = 3 
    REAL(C_FLOAT) :: ALPHA = 1.0 
    REAL(C_FLOAT) :: BETA = 1.0 
    INTEGER(C_INT) :: LDB = 3 
    INTEGER(C_INT) :: LDC = 3 
    INTEGER(C_INT) :: STATUS 
    CHARACTER(KIND=C_CHAR, LEN=4) :: MATDESCRA  = "G00C" ! General matrix with zero-based indexing
    REAL(C_FLOAT) :: VAL(4), B(9), C(9) 
    INTEGER(C_INT) :: INDX(4), PNTRB(4), PNTRE(3)   
    DATA VAL/9, 8, 5, 2/ 
    DATA B /1, 5, 7, 4, 7, 7, 3, 3, 7/ 
    DATA C /0, 8, 8, 0, 2, 8, 3, 5, 6/ 
    DATA INDX/2, 1, 2, 2/ 
    DATA PNTRB/0, 2, 3/ 
    DATA PNTRE/2, 3, 4/ 
    STATUS = KML_SPARSE_SCSRMM(OPT, M, N, K, ALPHA, MATDESCRA, VAL, INDX, PNTRB, PNTRE, B, LDB, BETA, C, LDC) 
    ! 
    !  OUTPUT C: 
    !     59.00  91.00  127.00  15.00  17.00  43.00  9.00  11.00  20.00 
    !

Parent topic: Sparse BLAS Level 2 and Level 3 Functions