currently uses kendall tau between the trend and the sort by the data sorted by the statfeature (goes to feat1). maybe should be something related to confidence interval? how can that generalize for more than 2?
for two it uses % of confidence interval that is shared.
currently uses kendall tau between the trend and the sort by the data sorted by the statfeature (goes to feat1). maybe should be something related to confidence interval? how can that generalize for more than 2?
for two it uses % of confidence interval that is shared.