# Data audit
- `data/life_expectancy_panel.csv` SHA-256 b3f2f774ed6797034acf62853286ca37fc02334ba3ce70e745e4f3d5c533c3ac; 439 data rows + header, 4 columns (country_name str, iso3 str, year int, life_expectancy float, years).
- (The README's SHA-256 refers to the raw API response, not this CSV; not comparable.)
- Years 2000–2022. 20 distinct iso3, 21 distinct country_name.
- TUR appears as "Turkey" and "Turkiye", 23 rows each, all 23 year pairs have identical values -> exact duplicates by (iso3,year); deduplicated.
- BGD, ETH, IND, NGA: rows only for 2011–2022 (12 rows each) -> not balanced.
- CHE 2007: empty value (only missing cell) -> CHE not balanced.
- Value range 51.49 (NGA) – 84.56 (JPN); largest year-to-year jumps are 2020–2022 (COVID) movements, plausible; no unit anomalies.
- "Korea, Rep." quoted correctly in CSV. No whitespace issues in iso3.
- No input violates the information rules.