## Cleaning report - Input rows: 745,119 (all retained; exclusions are flags) - Screens: ratio band [0.2, 5.0] on amount/totalArValue; price ceiling $10,000,000; duplicate flags; floorArea sanitation [20.0, 2000.0] m². - Provider pre-filter verified: no transaction below $50,000 exists in the raw file. ### Exclusion table (outputs/tables/exclusions.csv) - assessment_ratio_outside_band: 24,976 (3.352%) - exact_duplicate: 665 (0.089%) - addr_date_amount_duplicate: 107 (0.014%) - price_above_10m: 46 (0.006%) - TOTAL_EXCLUDED: 25,794 (3.462%) - RETAINED_IN_RESEARCH_SAMPLE: 719,325 (96.538%) ### Characteristic sanitation (row retained, value nulled) - floorArea outside [20.0, 2000.0] m²: 576 - yearBuilt > sale year + 1: 19 ### Research sample - Clean rows incl. indéterminé: 719,325 - Headline sample (3 determined types): 626,127 (84.03% of raw) - unifamilial: 439,118 - condo: 94,843 - plex: 92,166 - Weeks with >=1 research-sample transaction: 291 / 291 - Median weekly research-sample volume: 1,999 ### Missingness in headline research sample - floorArea: 1.85% - yearBuilt: 5.4% - buildingType: 27.61% - ownerType: 23.7%