# Data audit
| file | SHA-256 | rows (excl header) | cols | range | missing | dup dates |
|---|---|---|---|---|---|---|
| GDPC1.csv | 1c6c6e9f59e383e20c849dd299175c7a307526c09abdc18f18da9754fc5b3343 (matches README/bundle) | 319 | 2 (observation_date, GDPC1 float, bn chained 2017 $ SAAR) | 1947-01-01 – 2026-07-01, quarterly, no gaps | 0 | 0 |
| UNRATE.csv | 621ff6fadf50910cb7a21ee20ba610c247cb043b29b4e05c4c1a72fadca2a06c (matches) | 945 | 2 (observation_date, UNRATE float, %, SA) | 1948-01-01 – 2026-09-01, monthly, no date gaps | 1 (2025-10-01 empty value) | 0 |

- Window 1959Q4–2019Q4: no missing values in either series (UNRATE 1959-01..2019-12 732 months complete; GDPC1 complete).
- Anomaly: UNRATE 2025-10 missing (outside question window; irrelevant, recorded in data_issues).
- Availability: both files extend past 2019Q4 (to 2026). Handling: truncate to ≤2019Q4 before any computation. GDPC1 is the current (2026) vintage, i.e. revised data — acceptable since the question is historical, not real-time; noted as limitation.
- Values range: GDPC1 2172–24611; UNRATE 2.5–14.8. Plausible.
- `gate_spec.json` and `bundle.json` in workspace are not listed inputs; not used for analysis (bundle.json used only to cross-check hashes).