Data audit (P1)
SHA-256 computed with shasum -a 256; all 10 match data/README.md.
| file | sha256 | data rows | cols | range | NaN | dup dates | gaps>1q | min/max |
|---|---|---|---|---|---|---|---|---|
| vintages/..._2024-01-31.csv | 7ce218fb…f243 | 307 | 2 | 1947-04-01..2023-10-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintages/..._2024-04-30.csv | 62c1dfce…97cb | 308 | 2 | ..2024-01-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintages/..._2024-07-31.csv | 6f942591…79da | 309 | 2 | ..2024-04-01 | 0 | 0 | 0 | -28.0/34.8 |
| vintages/..._2024-10-31.csv | c4c51ede…0003 | 310 | 2 | ..2024-07-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintages/..._2025-01-31.csv | 10a5e02e…aee2 | 311 | 2 | ..2024-10-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintages/..._2025-04-30.csv | 1f19c975…085d1a | 312 | 2 | ..2025-01-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintages/..._2025-07-31.csv | f1e7c556…57a9 | 313 | 2 | ..2025-04-01 | 0 | 0 | 0 | -28.1/35.2 |
| vintages/..._2025-12-31.csv | c111c070…a8d4a7 | 314 | 2 | ..2025-07-01 | 0 | 0 | 0 | -28.0/34.9 |
| 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 | quarters 2023-07-01..2026-07-01 | 0 | 0 | 0 | -0.3/4.9 |
Units: percent change from preceding quarter, SAAR (percentage points of annualised growth). Quarters dated by first day.
Anomalies / notes:
- COVID quarters 2020Q2/Q3 are extreme outliers in every vintage -> handled in model estimation (exclusion ledger).
- 8th origin 2025-12-31 (shutdown-delayed advance estimate); spacing irregular.
- Last value of each vintage equals first_release value of that quarter (checked in analysis.py; no mismatch).
- Every target quarter (2024Q1..2026Q1) has a first release with vintage date strictly after its origin (asserted in analysis.py).
Information-availability handling: current_vintage.csv is read only for the revision statistic; first_release.csv only for actuals. No input violates the rules as long as each origin uses only its own vintage file, which analysis.py enforces (one file per origin loop iteration). Model specification (AR(1), 1985Q1+ window, COVID rows excluded) was fixed before computing any forecast errors.