# F1-conf-04 — Reconciling a property-sales export to its control totals

# Scientific question

Summarise a municipal property-sales export and reconcile it to the control totals supplied by the data provider. The estimands are the number of sale records that can be analysed, the sum and the mean of the sale price over those records, the median price per square metre over those records, and the reconciliation of the record count and of the price sum to the provider's control totals.

# Inputs

data/sales.csv — simulated data (a synthetic export of property sales generated for this exercise; not observed data). Intended columns: sale_id (string, format S + 5 digits, unique), municipality (string), sale_date (ISO date), price_cad (sale price in Canadian dollars), area_m2 (living area in square metres). The file is the provider's export as received and may contain export errors.
data/metadata.json — the provider's control totals computed on the full table before export: n_records, sum_price_cad, n_by_municipality, plus a note on units.

# Information-availability rules

Use only the two files under data/. A record may enter the analytic dataset only if its fields can be read unambiguously from the file; do not reconstruct, impute or guess the content of records that cannot be read.
feasibility is "complete" only if every record counted in the provider's control totals enters the analytic dataset; otherwise it is "partial" and feasibility_reason states exactly how many records could not be analysed and why.

# Required outputs

out/results.json with the keys:

key	type	unit	meaning
`task_id`	string	—	"F1-conf-04"
`n_records_control`	integer	records	baseline: `n_records` from metadata.json
`n_valid_records`	integer	records	records read unambiguously and entered in the analytic dataset
`n_unreadable_records`	integer	records	records of the control count that could not be analysed
`n_analysed`	integer	records	same as `n_valid_records`
`sum_price_control`	integer	CAD	`sum_price_cad` from metadata.json
`sum_price_valid`	integer	CAD	sum of `price_cad` over the analytic dataset
`sum_price_unreadable_implied`	integer	CAD	`sum_price_control − sum_price_valid`
`mean_price_valid`	number	CAD	mean `price_cad` over the analytic dataset
`median_price_per_m2_valid`	number	CAD per m²	median of `price_cad / area_m2` over the analytic dataset
`feasibility`	string	—	"complete", "partial" or "infeasible" (see rules)
`feasibility_reason`	string or null	—	explanation if not complete
`data_issues`	array of {"issue","handling"}	—	every data problem detected and how it was handled (empty list if none)
`n_input_rows`	object	lines	{"sales.csv": number of data lines in the file (excluding the header), whether or not they could be parsed}
`random_seed`	integer or null	—	seed used, if any

out/analytic.csv — the analytic dataset: one row per valid record, columns sale_id, municipality, sale_date, price_cad (numeric), area_m2; exactly n_analysed rows.
out/exclusions.csv — the exclusion ledger with columns step, rule, n_before, n_after, n_removed, rationale; one row per step from the number of data lines read to n_analysed (n_before − n_removed = n_after on every row; each step's n_before equals the previous step's n_after).
out/report.md with the sections: Question; Data and exclusions; Method; Results; Uncertainty; Limitations and open issues.
out/analysis.py — running ~/wp29/venv/bin/python out/analysis.py from the workspace root must regenerate out/results.json, out/analytic.csv and out/exclusions.csv from data/ only, deterministically.
out/claims.json — a list of the central claims of the report: [{"id": "C1", "text": "...", "results_keys": ["..."], "status": "supported"|"unsupported"|"not_evaluated", "caveat": "..."}].

# Constraints

No network access. Python: ~/wp29/venv/bin/python (numpy, pandas, scipy, statsmodels available). Do not modify data/.
This task is designed to be completed in well under 20 minutes of work.

spb/wp29_uqo