Joseph Knox
Joseph Knox
Move to a more object oriented framework for easier access of attributes of our algorithms. - add classes compatible with scikit-learn estimators - add `compute_` prefix to current functions This...
Add support for arrays of type ``scipy.sparse.[coo_matrix, csr_matrix, ...]``. This has a lot of potential gotchya's, however it would be nice to add support for them in the algorithms in...
Standardize the behavior of the algorithms when ``n > m``. Choices include: - 'flipped' (current): initially transpose input array ``A``, then transpose the returned results at the end. - 'axis':...
Index out of bounds error thrown: https://github.com/AllenInstitute/mouse_connectivity_models/blob/22b96e2f45bbdca4717e7b3722a1ee23aa71ac7b/mcmodels/models/homogeneous/subset_selection.py#L69
Would fix confusion in #6, and provides more utility in general.