20. Data cleaning and transformation

Correct types, standardize categories, parse dates, handle missing values, remove duplicates, and reconcile inconsistent sources. Build repeatable transformations while preserving raw data and documenting assumptions.