1# Data audit (P1)23| file | SHA-256 | rows (excl. header) | cols |4|---|---|---|---|5| data/wb_indicators.csv | 6823f02ed4c20fd12a54b8ca4454f85ba333f22337e2a9fc57c3cabccfdbfcd0 (matches README) | 2349 | 5 |6| data/wb_entities.csv | 06af6aec15122fecea5be5eb21464c505188ecc135bc55104a40f20f65040f8d (matches README) | 296 | 5 |7| data/README.md | 76cb5af488d97339fb6319a5c9a7ba6ec98112ab6a0a8116e7bec6acd1f9f67a (matches bundle.json) | – | – |89**Schema.** Indicators: iso3 (str), name (str), year (int, 2015–2023, 261 entities × 9 years, balanced), gdp_pc_ppp (float, const. 2021 PPP $, range 992–139,541), life_expectancy (float, years, range 18.8–86.4). Entities: iso3, name, region, income_level, lending_type; no missing.10**Duplicates.** None on (iso3, year) or entity iso3. Names agree between files for all rows.11**Matching.** Every indicator iso3 is in wb_entities. 35 metadata entities (all Aggregates-type codes) have no indicator rows.12**Aggregates.** 44 of 261 indicator entities have region == "Aggregates" → excluded (396 rows). 217 countries/economies remain.13**Missingness.** gdp_pc_ppp missing in 164 rows overall; life_expectancy never missing. Among countries, complete pairs: 199 per year 2015–2022, 197 in 2023. All years ≥150, so analysis year = 2023.14**2023 missing GDP (20):** ASM, CHI, CUB, ERI, GIB, GUM, IMN, LIE, MAF, MCO, MNP, NCL, PRK, PYF, SSD, SYR, VEN, VGB, VIR, YEM — mostly small territories plus several conflict/sanctioned economies (non-random missingness).15**Anomalies.** Central African Republic LE series is implausibly volatile (31.5 in 2019, 18.8 in 2022, 57.4 in 2023); 2023 value itself is plausible and kept (no rule permits removing it); flagged. West Bank and Gaza LE falls 11.5 years in 2023 (war; kept). LE complete for all territories → modelled (UN WPP) estimates, latest year likely provisional.16**Availability rules.** No input violates them; only provided files used; aggregates removed via metadata region field.17