Data audit (P1)
All SHA-256 values recomputed with shasum -a 256 and match data/README.md.
| file | data rows | cols | range | NaN | dup dates | min / max |
|---|---|---|---|---|---|---|
| current_vintage.csv | 318 | 2 | 1947-04-01 … 2026-07-01 | 0 | 0 | -28.0 / 34.9 |
| vintages/..._2024-01-31.csv | 307 | 2 | 1947-04-01 … 2023-10-01 | 0 | 0 | -28.0 / 34.8 |
| vintages/..._2024-04-30.csv | 308 | 2 | … 2024-01-01 | 0 | 0 | -28.0 / 34.8 |
| vintages/..._2024-07-31.csv | 309 | 2 | … 2024-04-01 | 0 | 0 | -28.0 / 34.8 |
| vintages/..._2024-10-31.csv | 310 | 2 | … 2024-07-01 | 0 | 0 | -28.1 / 35.2 |
| vintages/..._2025-01-31.csv | 311 | 2 | … 2024-10-01 | 0 | 0 | -28.1 / 35.2 |
| vintages/..._2025-04-30.csv | 312 | 2 | … 2025-01-01 | 0 | 0 | -28.1 / 35.2 |
| vintages/..._2025-07-31.csv | 313 | 2 | … 2025-04-01 | 0 | 0 | -28.1 / 35.2 |
| vintages/..._2025-12-31.csv | 314 | 2 | … 2025-07-01 | 0 | 0 | -28.0 / 34.9 |
| first_release.csv | 13 | 3 | quarters 2023-07-01 … 2026-07-01 | 0 | 0 | -0.3 / 4.9 |
Units: percent change from preceding quarter, SAAR (percentage points). Quarterly spacing 90–92 days, no gaps.
Anomalies / notes:
- 2020Q2 (≈-28) and 2020Q3 (≈+35) pandemic outliers in every file. Handling: excluded from the model's regression sample (pairs whose y or lag lies in 2020Q2–2020Q3); naive unaffected.
- Origins unevenly spaced: 2025Q3 advance estimate delayed by the 2025 shutdown, so origin 8 is 2025-12-31. Target 2025Q4 is the h=2 target of origin 7 and the h=1 target of origin 8; it enters both horizon samples.
- Revisions between vintages are visible (e.g. 2025Q2: 3.0 in 2025-07-31 vintage, 3.8 in 2025-12-31 vintage).
- Availability: current_vintage.csv violates the real-time rule for estimation; used only for the revision statistic. analysis.py asserts that each target's first_release_vintage_date is after its origin.
- Process incident: a working directory slip briefly created
data/out/state/(my own state files only); it was moved toout/state/and removed; data/ hashes re-verified unchanged afterwards.