np-stats 0.1.0

Educational non-parametric statistics lab toolkit with printed interpretations

np-stats

np-stats is an educational non-parametric statistics lab toolkit. It wraps common SciPy tests and prints both numeric results and plain-English interpretations.

Install:

pip install np-stats

Import:

import np_stats as nps

Example:

result = nps.mann_whitney_test(
    [23, 45, 12, 67, 34, 56, 22, 48],
    [34, 56, 78, 45, 89, 62, 41, 57],
)

Every test returns a dictionary and prints the result by default. Use print_result=False to suppress printing.

If you want the function to ask for required parameters interactively, call it with ask=True and omit the data arguments:

nps.binomial_test(ask=True)

Included tests:

• Binomial test
• One-sample Kolmogorov-Smirnov test
• Chi-square goodness-of-fit
• McNemar test
• Wilcoxon signed-rank test
• Sign test
• Paired permutation test
• Fisher exact test
• Chi-square test for independent samples
• Median test for two or k samples
• Mann-Whitney U test
• Two-sample Kolmogorov-Smirnov test
• Siegel-Tukey test
• Two-independent-samples permutation test
• Cochran Q test
• Friedman test
• Page trend test
• Chi-square test for k independent samples
• Kruskal-Wallis test
• Jonckheere-Terpstra trend test