oneapi::mkl::sparse::set_matmat_data

Sets the appropriate

oneapi::mkl::sparse::matrix_view_descr

and

oneapi::mkl::transpose

values in the

oneapi::mkl::sparse::matmat_descr_t

object reflecting the

sparse::matmat

operation to be performed: $LaTex Math image.$ .

Description

API

Description

The

oneapi::mkl::sparse::set_matmat_data

routine allows user to set the desired sparse matrix - sparse matrix operation in the

oneapi::mkl::sparse::matmat_descr_t

object which will be used in the sparse::matmat routine.

The

oneapi::mkl::sparse::matrix_view_descr

enum class is defined in the

oneapi/mkl/spblas.hpp

header file

namespace oneapi::mkl::sparse {
    enum class matrix_view_descr : std::int32_t {
        general
    };
}

where general view assumes all data is populated in the

sparse::matrix_handle_t

object for both lower, diagonal and upper portions of the matrix. A common usage model for the

matmat

descriptor is the following:

using namespace oneapi::mkl;
sparse::matmat_descr_t descr = NULL;
sparse::init_matmat_descr(descr);

// example descriptor for general
// C = A * B
sparse::matrix_view_descr viewA = sparse::matrix_view_descr::general;
sparse::matrix_view_descr viewB = sparse::matrix_view_descr::general;
sparse::matrix_view_descr viewC = sparse::matrix_view_descr::general;
transpose opA = transpose::nontrans;
transpose opB = transpose::nontrans;

sparse::set_matmat_data(descr, viewA, opA, viewB, opB, viewC);

// use descr in sparse::matmat() api

sparse::release_matmat_descr(descr);

API

Syntax

namespace oneapi::mkl::sparse {
    void set_matmat_descr (
        sparse::matmat_descr_t    descr,
        sparse::matrix_view_descr viewA,
        transpose                 opA,
        sparse::matrix_view_descr viewB,
        transpose                 opB,
        sparse::matrix_view_descr viewC);
}

Include Files

oneapi/mkl/spblas.hpp

Input Parameters

viewA, viewB, viewC: sparse::matrix_view_descr
enum values describing how the $LaTex Math image.$ , $LaTex Math image.$ and $LaTex Math image.$ matrix representations should be viewed. Currently, only the general
type is supported.
opA, opB: Specifies operation
op()
on input matrix.

oneapi::mkl::transpose::nontrans

Non-transpose, $LaTex Math image.$ .

oneapi::mkl::transpose::trans

Transpose, $LaTex Math image.$ .

oneapi::mkl::transpose::conjtrans

Conjugate transpose, $LaTex Math image.$ .

Currently, the only supported case for operation is
oneapi::mkl::transpose::nontrans
.