Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

removing correlated features doesn't work for classification targets #140

Closed
ardunn opened this issue Dec 5, 2018 · 0 comments
Closed
Assignees

Comments

@ardunn
Copy link
Contributor

ardunn commented Dec 5, 2018

rm_correlated in FeatureReducer will error when the target column contains categorical variables

one way around this is to labelencode the target column temporarily and remove correlated features based on this. not the best practice from a rigorous standpoint but we can always test if it is giving a bad result down the line

@ardunn ardunn self-assigned this Dec 5, 2018
@ardunn ardunn added the ugrads label Dec 11, 2018
@ardunn ardunn closed this as completed Dec 20, 2018
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant