1
32%
master: 32%

Ran 02 Mar 2015 07:19PM UTC

Files 82

Run time 9s

Badge

Embed ▾

Committed 02 Mar 2015 06:55PM UTC coverage: 96.819% (+0.03%) from 96.786%

Job # 740.1

Build Type

push

travis-ci

Committed by

ddemidov

Commit Message

Implemented tensor dot product operation

Given two tensors (arrays of dimension greater than or equal to one),
A and B, and a list of axes pairs (where each pair represents corresponding
axes from two tensors), sums the products of A's and B's elements over the
given axes.

Inspired by python's numpy.tensordot operation.

For example, with tensordot a dense matrix-matrix product may be represented as:

    slicer<2> Adim(extents[N][M]);
    slicer<2> Bdim(extents[M][L]);

    // C = A * B
    C = tensordot(Adim[_](A), Bdim[_](B), axes_pairs(1, 0));

This should perform better than the following combination of reshape and
reduce (see #31):

    C = reduce<SUM>(
                extents[N][M][L],
                reshape(A, extents[N][M][L], extents[0][1]) *
                reshape(B, extents[N][M][L], extents[1][2]),
                extents[1]
                );

Run Details

12478 of 12888 relevant lines covered (96.82%)

43145.26 hits per line

ddemidov / vexcl / 740 / 1
32%
master: 32%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 740.1

ddemidov / vexcl / 740 / 1 32% master: 32%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 740.1

ddemidov / vexcl / 740 / 1
32%
master: 32%

README BADGES
x