amult#
Purpose#
Performs matrix multiplication on the planes described by the two trailing dimensions of N-dimensional arrays.
Format#
- y = amult(a, b)#
- Parameters:
a (N-dimensional array)
b (N-dimensional array)
- Returns:
y (N-dimensional array) – containing the product of the matrix multiplication of the planes described by the two trailing dimensions of a and b.
Examples#
/*
** Create an additive sequence from 1-12 and reshape it into
** a 2x3x2 dimensional
*/
a = areshape(seqa(1, 1, 12), 2|3|2);
b = areshape(seqa(1, 1, 16), 2|2|4);
/*
** Multiply the two 3x2 matrices in 'a' by the corresponding
** 2x4 matrices in 'b'
*/
y = amult(a, b);
a is a 2x3x2 array, such that: [1,1,1] through [1,3,2] =
1.0000000 2.0000000
3.0000000 4.0000000
5.0000000 6.0000000
[2,1,1] through [2,3,2] =
7.0000000 8.0000000
9.0000000 10.000000
11.000000 12.000000
b is a 2x2x4 array, such that: [1,1,1] through [1,2,4] =
1.0000000 2.0000000 3.0000000 4.0000000
5.0000000 6.0000000 7.0000000 8.0000000
[2,1,1] through [2,2,4] =
9.0000000 10.000000 11.000000 12.000000
13.000000 14.000000 15.000000 16.000000
y will be a 2x3x4 array, such that: [1,1,1] through [1,3,4] =
11.000000 14.000000 17.000000 20.000000
23.000000 30.000000 37.000000 44.000000
35.000000 46.000000 57.000000 68.000000
[2,1,1] through [2,3,4] =
167.00000 182.00000 197.00000 212.00000
211.00000 230.00000 249.00000 268.00000
255.00000 278.00000 301.00000 324.00000
Remarks#
The multiplication operator, *, performs the same operation for arrays as amult().
All leading dimensions must be strictly conformable, and the two trailing dimensions of each array must be matrix-product conformable.