scikit-posthocs: Pairwise multiple comparison tests in Python

Abstract

scikit-posthocs is a Python package providing multiple pairwise comparison tests (post hocs). Statisticians, data scientists, and researchers will find it useful in a statistical analysis routine to assess the differences between group levels if a statistically significant result of a parametric or nonparametric analysis of variance (ANOVA) test has been obtained. Most statistical software packages (such as SPSS and Stata) provide a number of post hoc tests. However, Python ecosystem still lacks implementation of these tests for factorial and block design data. Currently, Student, Wilcoxon, Mann-Whitney tests which are not adapted to multiple pairwise comparisons are offered by SciPy. Statsmodels package includes TukeyHSD test which needs extra actions to be fluently integrated into a data analysis pipeline. scikit-posthocs package is aimed at filling this gap by providing a number of nonpara- metric and parametric pairwise comparisons tests as well as outliers detection algorithms implemented in Python.