CUTIE v0.2.0

Version 0.2.0 of CUTIE, a statistical software in Python 3.7.3 that computes pairwise correlations within a dataset and identifies correlations driven by influential points via a leave-one-out/jackknifing inspired method. It is capable of both true positive and false positive separation as well as true negative and false negative separation, in addition to detection of reverse sign correlations. Supported correlation coefficients include Pearson, Spearman, and Kendall. The known issue with Kendall's tau (from scipy.stats.kendalltau) has been resolved with updated dependencies. A brand new wiki and tutorial are available, along with analysis scripts and repo to accompany the paper figures. MINE, Jackknifing and bootstrapping (from 0.1.0) are no longer supported.

CUTIE v0.1.0

Version 0.1.0 of CUTIE, a statistical software in Python 3.7.3 that computes pairwise correlations within a dataset and identifies correlations driven by influential points via a leave-one-out/jackknifing inspired method. It is capable of both true positive and false positive separation as well as true negative and false negative separation. Supported correlation coefficients include Pearson, Spearman, Kendall, and MINE. Note that there is a known issue with Kendall's tau (from scipy.stats.kendalltau) that is currently undergoing development which should be resolved when scipy 1.3.0 is released (currently 1.2.1).