# Data audit (P1)
| file | SHA-256 (matches README) | data rows | cols | range | NaN | dup dates | gaps >1q | min/max |
|---|---|---|---|---|---|---|---|---|
| current_vintage.csv | 1af2590d…0562 ✓ | 318 | 2 | 1947-04-01..2026-07-01 | 0 | 0 | 0 | -28.0/34.9 |
| first_release.csv | ab843c9f…95bb ✓ | 13 | 3 | 2023-07-01..2026-07-01 | 0 | 0 | 0 | -0.3/4.9 |
| vintage 2024-01-31 | 7ce218fb… ✓ | 307 | 2 | ..2023-10-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintage 2024-04-30 | 62c1dfce… ✓ | 308 | 2 | ..2024-01-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintage 2024-07-31 | 6f942591… ✓ | 309 | 2 | ..2024-04-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintage 2024-10-31 | c4c51edee… ✓ | 310 | 2 | ..2024-07-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintage 2025-01-31 | 10a5e02e… ✓ | 311 | 2 | ..2024-10-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintage 2025-04-30 | 1f19c975… ✓ | 312 | 2 | ..2025-01-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintage 2025-07-31 | f1e7c556… ✓ | 313 | 2 | ..2025-04-01 | 0 | 0 | 0 | -28.0/35.2 (max) |
| vintage 2025-12-31 | c111c070… ✓ | 314 | 2 | ..2025-07-01 | 0 | 0 | 0 | -28.0/34.9 |

All vintages start 1947-04-01. Units: percent, SAAR. All values have one decimal.
Anomalies: (1) COVID outliers 2020Q2 (≈-28) and 2020Q3 (≈+35) in every vintage — handled by excluding them from the AR estimation sample (decided a priori). (2) Vintages revise earlier quarters (e.g. 2024Q4 2.3→2.4→1.9; 2025Q2 3.0→3.8); per rules each origin uses its own vintage values. (3) 8th origin is 2025-12-31 (shutdown delayed 2025Q3 advance release); 2025Q4 first release dated 2026-02-28 (also delayed). Fine: actuals are only used for evaluation.
Availability: all first-release vintage dates of targets are after the respective origin (checked in analysis.py). current_vintage.csv is read only for the revision statistic. No input violates the rules.