Skip to content

Lower _MINIMUM_COUNT to 1e-7 #644

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

Merged
merged 1 commit into from
Oct 14, 2017
Merged

Conversation

chihuahua
Copy link
Member

PR curve summaries had used 1.0 as the minimum division value while computing precision and recall.

That actually does not work because certain weights can cause TP, FP, TN, and FN values to be below 1.0.

This change sets that minimum to 1e-7 and adds a test.

PR curve summaries had used 1.0 as the minimum division value while computing precision and recall.

That actually does not work because certain weights can cause TP, FP, TN, and FN values to be below 1.0.

This change sets that minimum to 1e-7 and adds a test.
Copy link
Contributor

@wchargin wchargin left a comment

Choose a reason for hiding this comment

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

Great; thanks!

@chihuahua chihuahua merged commit 5806dd7 into tensorflow:master Oct 14, 2017
@chihuahua chihuahua deleted the summary branch October 14, 2017 22:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants