Skip to content

[MRG+1] FIX: for changing mode#5143

Merged
agramfort merged 1 commit intoscikit-learn:masterfrom
larsoner:fix-r2
Aug 23, 2015
Merged

[MRG+1] FIX: for changing mode#5143
agramfort merged 1 commit intoscikit-learn:masterfrom
larsoner:fix-r2

Conversation

@larsoner
Copy link
Copy Markdown
Contributor

Due to the changes in r2_score, I was getting a lot of DeprecationWarnings. This should preserve the old behavior.

Alternatively, an argument could be added to score.

@larsoner larsoner changed the title MRG: Fix for changing mode [MRG] FIX: for changing mode Aug 20, 2015
@agramfort
Copy link
Copy Markdown
Member

did you remove any warning from the test suite?

@larsoner
Copy link
Copy Markdown
Contributor Author

It removed some warnings from the mne test-suite, I didn't look at the sklearn output very closely. I actually think this change probably needs to be done more to preserve backward compatibility / behavior (i.e., avoid API change) than suppress a warning. That's where the choice between adding a new parameter or not is necessary, because as it stands the user of RegressorMixin has no way of setting the algorithm to use to be what it was before.

@ogrisel
Copy link
Copy Markdown
Member

ogrisel commented Aug 21, 2015

this LGTM

@ogrisel ogrisel changed the title [MRG] FIX: for changing mode [MRG+1] FIX: for changing mode Aug 21, 2015
@agramfort
Copy link
Copy Markdown
Member

agramfort commented Aug 21, 2015 via email

agramfort added a commit that referenced this pull request Aug 23, 2015
[MRG+1] FIX: for changing mode
@agramfort agramfort merged commit 39bad0a into scikit-learn:master Aug 23, 2015
@agramfort
Copy link
Copy Markdown
Member

thanks @Eric89GXL

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants