Statistics with Python
Pitfalls
ttest_indassumes equal or unequal variance: By default, scipy uses Welch's t-test (unequal variance), which is appropriate for most bioinformatics comparisons. If you have reason to believe variances are equal, passequal_var=True.mannwhitneyurequires thealternativeargument: Omitting it uses a two-sided test by default in newer scipy, but older versions behave differently. Always specifyalternative='two-sided'or'greater'expl
[Description truncada. Veja o README completo no GitHub.]