Distribution fitting enhancement

[bdartigues]

Distribution fitting (for z-scores, does it make sense to test all distributions? For example, von Mises for circular data)
=> Consider calculating the robust z-score (which uses the median and MAD to avoid the prevalence of outliers)