Shape, grain, dtypes and column names

Report rows, columns, time span, and fix column names and data types. Concepts: Grain check; dtype audit; snake_case names Project A: 1,000 rows; 7 integer columns (duration_months, credit_amount, installment_rate_pct, residence_years, age, num_existing_credits, num_dependents) and 13 text columns. installment_rate_pct, residence_years, num_dependents are integers but behave like ordinal categories (only 2–4 values) — note this, it changes outlier and encoding choices later. Project B: time.is_unique should be True after the DST fix; expect 4 years × 8,760 ≈ 35,064 hours; check the time span.