New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
FEA Allow passing custom covariance estimators to LDA #14446
FEA Allow passing custom covariance estimators to LDA #14446
Conversation
Overall looks pretty good to me. |
@adrinjalali Thanks for your comments. I made the requested changes. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A few comments for now, till the next round :) Thanks @hugorichard
Thanks @adrinjalali I have made the changes. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looking better now, thanks for the corrections @hugorichard
Hi @hugorichard , thanks for your patience! Do you mind triggering a new build either with an empty commit or syncing with upstream? The rendering of the documentation is not available anymore. Thanks! |
…into modular_discriminant_analysis
@cmarmo Thanks for your answer ! I have done what was asked. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @hugorichard minor comments waiting for a core-dev review.
@cmarmo Requested changes have been made |
Thanks @hugorichard! @NicolasHug might want to give a second review? Thanks! |
@NicolasHug @cmarmo Any news on this ? |
@NicolasHug, this is the oldest PR already approved waiting for reviewer and you already gave an LGTM (apart nitpicks.. :) ): may I ask you if you could find some time to finalize it? I'm happy to approve myself but it doesn't count for merging. Thanks for your patience! |
Thanks for the gentle reminders @hugorichard and @cmarmo @hugorichard also thanks again for your patience. I have committed a few minor changes in https://github.com/NicolasHug/scikit-learn/tree/my_modular_discriminant_analysis, which I cannot directly pushed here (looks like you have deactivated this option on this PR). I only added 1 test and made some cosmetic changes. Would you mind pulling the changes from my branch? LGTM then! Thanks again! EDIT: alternatively I opened a PR in your fork hugorichard#1 |
Minor updates to modular LDA
Hi @NicolasHug @cmarmo Thanks for your reviews. |
…into modular_discriminant_analysis
thanks a lot for your work @hugorichard ! |
You are welcome ! Thanks for the reviews ! |
…4446) Co-authored-by: hugo richard <hugo.richard@inria.fr> Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>
Fixes #14436
Plug-in any covariance estimator into Fisher discriminants
This PR allows the user to plug-in any sklearn covariance estimator into LDA and QDA. This option is only possible with the "lsqr" and "eigen" solver of LDA and for the "lsqr" solver of QDA (initially only an "svd" solver was available for QDA so I made this explicit and added the "lsqr").