minCharacterizationCount #318
Description
We currently have the minCharacterizationMean which is really useful for reducing the number of covariates when running an aggregate analysis. I generally set this value of 1% and that returns a smaller subset of useful features. Sometimes I have small cohorts (<100) and then the threshold of 1% doesn't filter anything. It would be useful to have a minCharacterizationCount in addition to minCharacterizationMean where for smaller cohorts a minCharacterizationCount of 2 will remove covariates than only one person has in the smaller cohorts to reduce the set of covariates.