Report a bug
If you spot a problem with this page, click here to create a GitHub issue.
Improve this page
Quickly fork, edit online, and submit a pull request for this page. Requires a signed-in GitHub account. This works well for small changes. If you'd like to make larger changes you may want to consider using a local clone.


Level 2

This is a submodule of mir.glas.
The Level 2 BLAS perform matrix-vector operations.

Note GLAS is singe thread for now.

Matrix-vector operations

Function Name Description
gemv general matrix-vector multiplication, partially optimized

Ilya Yaroshenko
nothrow @nogc @system void gemv(A, B, C, SliceKind kindA, SliceKind kindB, SliceKind kindC)(C alpha, Slice!(const(A)*, 2, kindA) asl, Slice!(const(B)*, 1, kindB) xsl, C beta, Slice!(C*, 1, kindC) ysl)
if (allSatisfy!(isNumeric, A, B, C));
DRAFT Performs general matrix-vector multiplication.

Pseudo code y := alpha A × x + beta y.

C alpha scalar
Slice!(const(A)*, 2, kindA) asl m ⨉ n matrix
Slice!(const(B)*, 1, kindB) xsl n ⨉ 1 vector
C beta scalar. When beta is supplied as zero then the vector ysl need not be set on input.
Slice!(C*, 1, kindC) ysl m ⨉ 1 vector

Note GLAS does not require transposition parameters. Use transposed  to perform zero cost Slice transposition.

BLAS SGEMV, DGEMV, (CGEMV, ZGEMV are not implemented for now)

import mir.ndslice;

auto a = slice!double(3, 5);
a[] =
    [[-5,  1,  7, 7, -4],
     [-1, -5,  6, 3, -3],
     [-5, -2, -3, 6,  0]];

auto b = slice!double(5);
b[] =

auto c = slice!double(3);

gemv!(double, double, double)(1.0, a, b, 0.0, c);

assert(c ==