TST Extend tests for scipy.sparse.*array in test_nmf.py#27100
TST Extend tests for scipy.sparse.*array in test_nmf.py#27100adrinjalali merged 5 commits intoscikit-learn:mainfrom
scipy.sparse.*array in test_nmf.py#27100Conversation
|
I think the only place that it's expected this could return a sparse type is 
However, I'm now thinking that Maybe the docs should be updated? |
|
I convinced myself |
| Returns | ||
| ------- | ||
| X : {ndarray, sparse matrix} of shape (n_samples, n_features) | ||
| X : ndarray of shape (n_samples, n_features) |
There was a problem hiding this comment.
Indeed, this is in line with the docstring of _fit_transform that should always return H as numpy arrays for both solvers (coordinate descent and multiplicative update).
ogrisel
left a comment
There was a problem hiding this comment.
LGTM (assuming CI will turn out green).
…earn#27100) Co-authored-by: Adrin Jalali <adrin.jalali@gmail.com>
…earn#27100) Co-authored-by: Adrin Jalali <adrin.jalali@gmail.com>
…earn#27100) Co-authored-by: Adrin Jalali <adrin.jalali@gmail.com>
Reference Issues/PRs
Towards #27090
What does this implement/fix? Explain your changes.
This modifies the NMF transformer to accept scipy.sparse.sparray classes.
Any other comments?
Still needs tests to check that the NMF returns the correct type. I am not 100% sure it's going to be obvious what the desired return type is.