Data audit (P1)
| file | SHA-256 | rows | cols |
|---|---|---|---|
| data/series.csv | c29dde22e215b14fe04d1786c9e9634c18863c78d1c283636e9cbfb31dabc514 | 400 (+header) | 3 |
| data/memo.md | 0486b6a6e26671a53d02e9cf0ceee063479ea0a05bd07aadf76aa71e9d2a0568 | text | – |
Hashes match bundle.json.
series.csv schema: month (str YYYY-MM), x (float, dimensionless), y (float, dimensionless).
- Date range 1990-01 to 2023-04; monotone increasing; 0 gaps; 0 duplicated months; 0 duplicated rows.
- Missing values: 0 in every column.
- x: mean -0.134, sd 1.005, range [-2.69, 2.36]; y: mean 2.079, sd 0.904, range [0.0036, 4.81]. No obvious outliers/anomalies.
- Lag-1 autocorrelation: x 0.662, y 0.624 (persistent, as stated in TASK.md).
memo.md: previous analyst's numbers (intercept 2.0925, slope 0.1017, SE 0.0448, t 2.27, p 0.0238, n 400, iid SEs). Used only for comparison after independent recomputation.
Information availability: both files are permitted inputs; nothing violates the rules. No exclusions needed (ledger records a no-op step).