colour.utilities.dot_matrix

colour.utilities.dot_matrix(a, b)[source]

Convenient wrapper around np.einsum() with the following subscripts: ‘…ij,…jk->…ik’.

It performs the dot product of two arrays where a parameter is expected to be an array of 3x3 matrices and parameter b another array of of 3x3 matrices.

Parameters
  • a (array_like) – Array of 3x3 matrices.

  • b (array_like) – Array of 3x3 matrices.

  • dtype (object) – Type to use for conversion, default to the type defined by the colour.constant.DEFAULT_FLOAT_DTYPE attribute.

Returns

Return type

ndarray

Examples

>>> a = np.array(
...     [[0.7328, 0.4296, -0.1624],
...      [-0.7036, 1.6975, 0.0061],
...      [0.0030, 0.0136, 0.9834]]
... )
>>> a = np.reshape(np.tile(a, (6, 1)), (6, 3, 3))
>>> b = a
>>> dot_matrix(a, b)  
array([[[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]],

       [[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]],

       [[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]],

       [[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]],

       [[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]],

       [[ 0.2342420...,  1.0418482..., -0.2760903...],
        [-1.7099407...,  2.5793226...,  0.1306181...],
        [-0.0044203...,  0.0377490...,  0.9666713...]]])