Skip to content

RidgeClassifierCV is not supporting properly multilabel-indicator #19858

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

Closed
glemaitre opened this issue Apr 10, 2021 · 9 comments · Fixed by #19869
Closed

RidgeClassifierCV is not supporting properly multilabel-indicator #19858

glemaitre opened this issue Apr 10, 2021 · 9 comments · Fixed by #19869
Assignees
Labels

Comments

@glemaitre
Copy link
Member

Our documentation state that RidgeClassifierCV is supporting multilabel-indicator.
However, a call to predict, will output an array of (n_samples,) containing n_classes instead of a matrix of (n_samples, n_classes). The method score` will fail in this case.

It looks like we should use a LabelBinarizer before to predict (I did not look at code thought).

@glemaitre
Copy link
Member Author

This is only the case when cv=None

@hxyue1
Copy link
Contributor

hxyue1 commented Apr 11, 2021

Hello! I can try fixing this if nobody else wants to?

@glemaitre
Copy link
Member Author

You will also need to investigate a couple of things because I find it fishy that we are supposed to provide support for multilabel-indicator for RegressionClassifierCV with cv=None and not provide the same support for RegressionClassifier

@glemaitre
Copy link
Member Author

My bad cv=None is the using the gcv_mode.

@glemaitre
Copy link
Member Author

@hxyue1 You would like to take this issue intro consideration. It will be indeed easier to deal with due to my ongoing work in #19859.

@glemaitre
Copy link
Member Author

take

@hxyue1
Copy link
Contributor

hxyue1 commented Apr 12, 2021

@hxyue1 You would like to take this issue intro consideration. It will be indeed easier to deal with due to my ongoing work in #19859.

Sorry, I'm not following...?

@glemaitre
Copy link
Member Author

Ups sorry, I did a typo. I would like to take this issue as part of solving the consistency with multilabel-indicator common test.

@hxyue1
Copy link
Contributor

hxyue1 commented Apr 12, 2021

No worries, I'll leave you to it!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants