sampled_khatri_rao

# Import statement

from hottbox.core import sampled_khatri_rao

sampled_khatri_rao(matrices, sample_size=None, skip_matrix=None)

Sampled Khatri-Rao product of a list of matrices.

Parameters

matriceslist[np.ndarray]

List of matrices. Each matrix should have the same number of columns

skip_matrixint

Index of a matrix (from the matrices) to be skipped. By default none are skipped

reversebool

If True, perform khatri-rao product on the list of matrices in the reversed order

Returns

resultnp.ndarray

The result of the Khatri-Rao product of the sampled matrices

indicestuple list

list of indices for all modes

References

• Battaglino, C., Ballard, G., & Kolda, T. G. (2018). A Practical Randomized CP Tensor Decomposition. SIAM Journal on Matrix Analysis and Applications, 39(2), 876–901. http://doi.org/10.1137/17m1112303