KenSciResearch/fairMLHealth

Feature Request: Add Flexibility to Bootstrapping Utility

Opened this issue · 0 comments

Checklist

  • I am running the latest version
  • I checked the documentation and troubleshooting resources and found no answer
  • I have read the Contributing guidelines
  • I verified that this issue has not already been filed by checking the Issues Board

Current Behavior

Bootstrapping utility returns true/false if a sufficient number of p-values contra-indicate the null hypothesis for a given statistical test.

Desired Behavior

Per @kenviku: "[O]ne flexibility we might add in next iteration is that the func does not need to be a statistical test. It can be an aggregator too. For instance, func can be an average function. In that case, the bootstrapping can validate that in how many trials the average of the group a is larger or smaller than group b."