Doing frequentist statistics with Scipy
summary
Covers concrete commands and gotchas: use SciPy.stats for Kolmogorov Smirnov Shapiro Anderson tests (Anderson returns critical values not a p), t tests and ANOVA, Mann‑Whitney/Wilcoxon, chi‑square and Fisher exact, Pearson/Spearman correlations, and quick linear fits; SciPy returns statistic and p and lacks posthoc and regression diagnostics so use statsmodels OLS/logit for summaries residuals influence and odds ratios; shows effect‑size calculation (example Cohen's d = 1.90 for sepal width) and p thresholds like normality p > 0.1.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 3:01Loading Iris dataset
- 18:28Statistical assumptions
- 28:16Variance homogeneity
- 31:55Two-group T-tests
- 35:31Effect sizes and power
- 39:46Nonparametric alternatives
- 43:18ANOVA and post-hoc tests
- 45:49Chi-square and Fisher tests
- 52:21Correlation analysis
- 55:58Linear regression
- 1:03:54Logistic regression
- 1:17:03Model validation
- 1:19:36Summary and Q&A
related talks
01h 30mAn Introduction to Probability and Statistics
01h 18mA/B Testing with SciPy
01h 53mAnalyze your data in Python
30mStats Don't Have to Be Scary: Automatic a/b Test Analysis Using Python
29mIt might look normal but this distribution will ruin your stats
08mIt might look normal but this distribution will ruin your stats
01h 15mAB-testing by cluster
01h 10mLies damned lies and statistics in Python