Data audit (P1)
| file | SHA-256 | rows (excl. header) | cols | range | missing | duplicates |
|---|---|---|---|---|---|---|
| data/cpi_quarterly.csv | c8652dec1567ec35e3b84ee8fa55545e9bf0ae4990a4d6d05b89a6d22f6801ea (matches README) | 127 | 3 (quarter, cpi_avg, infl_annualised) | 1995-01-01 … 2026-07-01 | infl_annualised empty for 1995Q1 only (by construction) | 0 |
| data/cpi_monthly.csv | 855d069176369d30d02cbbe0cc6456b76d2f3967cdaabc1b362c2461ef9015a2 (matches README) | 381 | 2 (month, cpi) | 1995-01 … 2026-09 | 0 | 0 |
- Units: cpi index 2002=100; infl_annualised in percent at annual rate (400·Δln).
- Regularity: quarterly and monthly dates contiguous (no gaps). Every quarter has 3 months.
- Consistency: recomputed quarterly means from monthly file match cpi_avg (max abs diff 3.3e-7, rounding); recomputed 400·Δln(cpi_avg) matches infl_annualised (max abs diff 3.3e-6, rounding).
- Summary of target: mean 2.13, sd 2.39, min −6.04 (2008Q4), max 12.15 (2022Q2). Strong seasonality: quarter means Q1 2.32, Q2 4.18, Q3 1.66, Q4 0.30.
- Anomalies: none erroneous; notable outliers 2008Q4 (oil collapse), 2020Q2 (COVID), 2022Q1–Q2 (inflation surge) — genuine, kept.
- Information availability: the file contains quarters after each origin (up to 2026Q3); these are used only as targets. The analysis script slices data to quarters ≤ origin before any estimation/selection. Last needed target 2025Q3+4 = 2026Q3 exists. No input violates the rules per se.
- Note: monthly file used only for provenance/consistency checking; not an analysis input.