EducationalTestingService / skll

SciKit-Learn Laboratory (SKLL) makes it easy to run machine learning experiments.
http://skll.readthedocs.org
Other
551 stars 67 forks source link

Do not return 0 if metric value is nan in correlation function #588

Closed mulhod closed 4 years ago

mulhod commented 5 years ago

Addresses #585.

mulhod commented 5 years ago

This will have to wait until #586 is finished.

codecov[bot] commented 5 years ago

Codecov Report

Merging #588 into master will not change coverage by %. The diff coverage is 100.00%.

Impacted file tree graph

@@           Coverage Diff           @@
##           master     #588   +/-   ##
=======================================
  Coverage   95.06%   95.06%           
=======================================
  Files          20       20           
  Lines        2977     2977           
=======================================
  Hits         2830     2830           
  Misses        147      147           
Impacted Files Coverage Δ
skll/metrics.py 97.14% <100.00%> (ø)

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 1f7a6fa...4d2a9a8. Read the comment docs.

mulhod commented 5 years ago

This is ready for review now.

mulhod commented 4 years ago

Does someone else have time to review this? @aoifecahill @bndgyawali @ananyaganesh @AVajpayeeJr @chaomenghsuan

desilinguist commented 4 years ago

@mulhod I merged master into this just to double check that nothing breaks with the new scikit-learn.

mulhod commented 4 years ago

Even though it says this passed all checks, there was one failure that still needs to be investigated. I'm not sure why it says all checks passed.

mulhod commented 4 years ago

This is good to go now.