search

Display-Lab / precision-feedback-pipeline

Apache License 2.0
1 stars 0 forks source link

when one measure is top10% and others are not, it choose top25% #204

Closed icejag closed 5 months ago

icejag commented 9 months ago

Problem: When we test more than one measure, where one measure is above top 10% and others are not, The top 10% measure is not selected at all.: when one measure is top10% and others are not, it choose top25%: The test case to test this bug [5,"GLU01","2023-12-01",99,16,100,85,88,96,90], [5,"BP03","2023-12-01",86,16,100,85,88,96,90], [5,"TOC01","2023-12-01",86,16,100,85,88,96,90]

Suggested solution: The issue 's solution is the solution for this problem.Instead of choosing the highest gap value, the lowest gap value should be selected and also the first record must be selected.

icejag commented 9 months ago

The issue persists when this test case was used.

[5,"GLU01","2023-12-01",98,16,100,85,88,96,90], [5,"BP03","2023-12-01",88,16,100,85,88,96,90], [5,"TOC01","2023-12-01",88,16,100,85,88,96,90]

Proposed solution: When the social better causal-pathway is selected, when the gap sizes are zero for couple of measures, we should select the highest gap size of the other measure.