scikit-posthocs 0.3.9

Statistical post-hoc analysis and outlier detection algorithms

This Python package provides statistical post-hoc tests for pairwise multiple comparisons and outlier detection algorithms.

Contents:

• Features

• Dependencies

• Compatibility

• Install

• Examples

  • List or NumPy array

  • Pandas DataFrame

  • Significance plots

• Credits

• Donate

Features

• Pairwise multiple comparisons parametric and nonparametric tests:

  • Conover, Dunn, and Nemenyi tests for use with Kruskal-Wallis test.

  • Conover, Nemenyi, Siegel, and Miller tests for use with Friedman test.

  • Quade, van Waerden, and Durbin tests.

  • Student, Mann-Whitney, Wilcoxon, and TukeyHSD tests.

  • Anderson-Darling test.

  • Mack-Wolfe test.

  • Nashimoto and Wright’s test (NPM test).

  • Scheffe test.

  • Tamhane T2 test.

• Plotting functionality (e.g. significance plots).

• Outlier detection algorithms:

  • Simple test based on interquartile range (IQR).

  • Grubbs test.

  • Tietjen-Moore test.

  • Generalized Extreme Studentized Deviate test (ESD test).

  All tests are capable of p adjustments for multiple pairwise comparisons.

Dependencies

• NumPy and SciPy packages

• Statsmodels

• Pandas

• Matplotlib

• Seaborn

Compatibility

Package is compatible with Python 2 and Python 3.

Install

You can install the package with: pip install scikit-posthocs

Examples

List or NumPy array

import scikit_posthocs as sp
x = [[1,2,3,5,1], [12,31,54], [10,12,6,74,11]]
sp.posthoc_conover(x, p_adjust = 'holm')

array([[-1.        ,  0.00119517,  0.00278329],
       [ 0.00119517, -1.        ,  0.18672227],
       [ 0.00278329,  0.18672227, -1.        ]])

Pandas DataFrame

Columns specified with val_col and group_col args must be melted prior to making comparisons.

import scikit_posthocs as sp
import pandas as pd
x = pd.DataFrame({"a": [1,2,3,5,1], "b": [12,31,54,62,12], "c": [10,12,6,74,11]})
x = x.melt(var_name='groups', value_name='values')

sp.posthoc_conover(x, val_col='values', group_col='groups', p_adjust = 'fdr_bh')

Significance plots

P values can be plotted using a heatmap:

pc = sp.posthoc_conover(x, val_col='values', group_col='groups')
heatmap_args = {'linewidths': 0.25, 'linecolor': '0.5', 'clip_on': False, 'square': True, 'cbar_ax_bbox': [0.80, 0.35, 0.04, 0.3]}
sp.sign_plot(pc, **heatmap_args)

Custom colormap applied to a plot:

pc = sp.posthoc_conover(x, val_col='values', group_col='groups')
# Format: diagonal, non-significant, p<0.001, p<0.01, p<0.05
cmap = ['1', '#fb6a4a',  '#08306b',  '#4292c6', '#c6dbef']
heatmap_args = {'cmap': cmap, 'linewidths': 0.25, 'linecolor': '0.5', 'clip_on': False, 'square': True, 'cbar_ax_bbox': [0.80, 0.35, 0.04, 0.3]}
sp.sign_plot(pc, **heatmap_args)

Credits

Thorsten Pohlert, PMCMR author and maintainer

Donate