Implemented block HSIC lasso #9
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
extended gitignore excluded files generated during package installation i.e. eggs, dist, build deprecated-to-be .as_matrix() replaced by .values
…on _run_hsic_lasso
fixed block tests: before blocks were not really tested
|
As agreed with @myamada0321, I fixed a crucial bug in pyHSIC lasso, where normalisation was dividing by np.sqrt(n - (1 / n)) instead of np.sqrt(float(n - 1) / n). Additionally, I implemented a multiprocess version for parallel computation of blocks. However, I ended up not using them as block lasso computation is faster than initialising a whole new process + Python environment. |
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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 implemented here block HSIC lasso developed by @myamada0321. It is a variant of HSIC lasso that reduces memory usage by estimating HSIC on samples of the data. If blocks are not specified, by setting B = 0 in both HSICLasso.classification or HSICLasso.regression, conventional HSIC lasso is used. I added unit tests for the new block estimator, and assessed that the results don't change wildly, although they do, especially when only 5 features are selected.
Additionally I cleaned a bit the repository from unnecessary files (.egg, diet, build), and updated .gitignore to ignore them in the future.