V2 has been updated such as the outlier detection returns outputs for 3 different methods - however the skipped correlation only uses the boxplot rule ensuring a type one error rate of 5% (v1 used the MAD but the type 1 error is not well controlled)
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
V2 has been updated such as the outlier detection returns outputs for 3 different methods - however the skipped correlation only uses the boxplot rule ensuring a type one error rate of 5% (v1 used the MAD but the type 1 error is not well controlled)