Skip to content
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

[MRG+2] Change default of ColumnTransformer remainder from passthrough to drop #11603

Conversation

jorisvandenbossche
Copy link
Member

Follow-up on #9012

The motivation is that silently passing through uniquely identifying columns (e.g. a user id) or the target will lead to catastrophic overfitting that beginner users will have a hard time to debug.

@jorisvandenbossche jorisvandenbossche changed the title Change default of ColumnTransformer remainder from passthrough to drop [MRG] Change default of ColumnTransformer remainder from passthrough to drop Jul 17, 2018
@jorisvandenbossche jorisvandenbossche added this to the 0.20 milestone Jul 17, 2018
@jorisvandenbossche
Copy link
Member Author

cc @GaelVaroquaux

@jorisvandenbossche
Copy link
Member Author

Travis is passing except of the sag failing one

Copy link
Member

@amueller amueller left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm

@GaelVaroquaux
Copy link
Member

I'm sorry, but I don't think that it's a good idea to merge with the failing SAG. The risk is to end up with a failing test in master, which isn't a good situation, in particular for a sprint.

@GaelVaroquaux GaelVaroquaux changed the title [MRG] Change default of ColumnTransformer remainder from passthrough to drop [MRG+2] Change default of ColumnTransformer remainder from passthrough to drop Jul 17, 2018
Copy link
Member

@GaelVaroquaux GaelVaroquaux left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM.

+1 for merge once CI is green.

@GaelVaroquaux
Copy link
Member

I've restarted the failing travis job, hoping that it will go away.

@jorisvandenbossche
Copy link
Member Author

@GaelVaroquaux It's a random failure that is already happening on master as well (I think Guillaume is working on a fix #11606)

@GaelVaroquaux
Copy link
Member

OK. However, we just got a very different error, that does not seem random to me:
https://travis-ci.org/scikit-learn/scikit-learn/jobs/405060141

@jorisvandenbossche
Copy link
Member Author

Ah, now it is another one (collection of the test), and that is just fixed in master normally, restarting again.

@jorisvandenbossche
Copy link
Member Author

Merged master in the hope this will then be green

@GaelVaroquaux
Copy link
Member

Everything is passing aside from appveyor that is pending. I'll merge!

@GaelVaroquaux GaelVaroquaux merged commit 277e125 into scikit-learn:master Jul 18, 2018
@jorisvandenbossche jorisvandenbossche deleted the columntransformer-change-remainder-default branch July 18, 2018 12:35
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.

None yet

3 participants