Categorical vs categorical

Contingency tables with row percentages and a stacked bar chart. Concepts: Crosstabs; proportions; chi-square test Project A: Chi-square + Cramér's V for every categorical vs target (your Cell 22): checking ≈ 0.34 strong, credit history ≈ 0.20, savings ≈ 0.19, purpose ≈ 0.15. WARNING: a 3-level feature with chi² = 0.00 and p = 1.000 (debtors_merged, job_merged in your run) usually means the .map() produced NaN or a single level — check value_counts(dropna=False) before trusting the test. Project B: Weather type × season; holiday × weekday (mostly a sanity check).