Extract randomized PCA impl in a dedicated toplevel class #30
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I wanted to make PCA able to handle sparse data (scipy.sparse matrices) using the fast_svd implementation. Since computing PCA for sparse dataset (with big n_features) is only feasable with truncated SVD the API of the existing PCA module (with automated "mle" strategy for finding n_components) is not very well suited to such an evolution.
I hence decided to wrap the fast_svd method in a dedicated RandomizedPCA that makes it explicit that it is able to handle both sparse and dense input provided that you are willing to truncate the singular spectrum to an arbitrary level.
Here is a branch that does just that along with updated docstring, tests and examples along with a renaming of n_comp to n_components to be consistent with the overall scikit-learn naming conventions for dimension parameters.