# Data audit

• File: data/firm_panel.csv, SHA-256 ae54378f63f4c9527cb2a1061d96d42eb1e45173bb8762a5998966cbfc9672e0
• 1200 rows (+header), 4 columns: firm_id int64, year int64, x float64, y float64 (dimensionless; synthetic).
• Firms: 100 (ids 1001–1100), years 2008–2019 (12); every firm has exactly 12 rows → balanced.
• Missing values: 0 in all columns; non-finite: 0. Exact duplicates: 0. Duplicate (firm_id, year) keys: 0.
• x: mean −0.162, sd 1.258, max 3.30; y: mean −0.424, sd 2.003, max 5.33. No anomalies detected.
• Rows are in random order (sorted in analysis; irrelevant for estimates).
• Information availability: single permitted input; no violations.