Data audit (P1)
| file | SHA-256 | rows (excl. header) | cols |
|---|---|---|---|
| data/design.csv | 2f459c539378cbe54d830c18af466f2fa4230ce48a609aa6464f8334a0d36ecb | 7 | 3 (parameter, value, description) |
Parameters: n=30, rho=0.8, sigma=1.0, intercept=yes, initial_condition=stationary (y_1 ~ N(0, sigma^2/(1-rho^2))), n_replications=20000, seed=20261029.
- Missing values: none. Duplicated parameter names: none. No dates (simulation design only).
- Units: dimensionless; sigma in units of y.
- Anomalies: none. |rho|<1 so stationary start is well defined.
- Note (not an error): with n=30 observations the regression uses n-1=29 pairs. The task defines bias_analytical at n=30 ("design values"); we follow that and report the n-1 variant only as a supplementary sensitivity.
- Availability: the file only contains design parameters; no violations.
- Exclusions: none expected (degenerate regressions checked in code; ledger in out/exclusions.csv).