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

Default the guess column to the most common class #8

Closed
rhiever opened this issue Nov 12, 2015 · 4 comments
Closed

Default the guess column to the most common class #8

rhiever opened this issue Nov 12, 2015 · 4 comments
Assignees
Labels

Comments

@rhiever
Copy link
Contributor

rhiever commented Nov 12, 2015

By default, the guess column is assigned to 0. In many use cases, there may not even be a 0 class. Instead, assign guess to the most common class in the training set.

@rhiever rhiever added the bug label Nov 12, 2015
@rhiever rhiever self-assigned this Nov 12, 2015
@rasbt
Copy link
Contributor

rasbt commented Nov 12, 2015

Related to that, the "positive" class label is always always assumed to be 1, right? (E.g., for computing metrics other than accuracy like precision, recall, F1 and so forth)?

@rhiever
Copy link
Contributor Author

rhiever commented Nov 12, 2015

Yes that's right. That's how it worked previously.

@rasbt
Copy link
Contributor

rasbt commented Nov 12, 2015

Okay, maybe -- in far future -- it would be worthwhile adding a feature for accepting a custom scorer via scikit-learn (since everything depends on scikit-learn) so that the user can choose whatever performance metric, positive label, and "greater is better: True/False" setting the user may prefer.

@rhiever
Copy link
Contributor Author

rhiever commented Nov 12, 2015

Definitely!

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

No branches or pull requests

2 participants