#!/usr/bin/env python3 """Build the LVC sample from an ONS private-rent snapshot and an ONSPD postcode spine.""" from __future__ import annotations import argparse import csv import math import subprocess import sys from pathlib import Path SOURCE_URL = ( "https://www.ons.gov.uk/economy/inflationandpriceindices/datasets/" "priceindexofprivaterentsukmonthlypricestatistics" ) SCOTTISH_RENTAL_MARKET_AREA = { "Aberdeen City": "Aberdeen and Shire", "Aberdeenshire": "Aberdeen and Shire", "Argyll and Bute": "Argyll and Bute", "East Ayrshire": "Ayrshires", "North Ayrshire": "Ayrshires", "South Ayrshire": "Ayrshires", "Dumfries and Galloway": "Dumfries and Galloway", "Angus": "Dundee and Angus", "Dundee City": "Dundee and Angus", "East Dunbartonshire": "East Dunbartonshire", "Fife": "Fife", "Clackmannanshire": "Forth Valley", "Falkirk": "Forth Valley", "Stirling": "Forth Valley", "East Renfrewshire": "Greater Glasgow", "Glasgow City": "Greater Glasgow", "Highland": "Highland and Islands", "Moray": "Highland and Islands", "Na h-Eileanan Siar": "Highland and Islands", "Orkney Islands": "Highland and Islands", "Shetland Islands": "Highland and Islands", "City of Edinburgh": "Lothian", "East Lothian": "Lothian", "Midlothian": "Lothian", "North Lanarkshire": "North Lanarkshire", "Perth and Kinross": "Perth and Kinross", "Inverclyde": "Renfrewshire/Inverclyde", "Renfrewshire": "Renfrewshire/Inverclyde", "Scottish Borders": "Scottish Borders", "South Lanarkshire": "South Lanarkshire", "West Dunbartonshire": "West Dunbartonshire", "West Lothian": "West Lothian", } def read_csv(path): with open(path, newline="", encoding="utf-8-sig") as handle: return list(csv.DictReader(handle)) def write_csv(path, rows, fields): with open(path, "w", newline="", encoding="utf-8") as handle: writer = csv.DictWriter(handle, fieldnames=fields) writer.writeheader() writer.writerows(rows) def number(row, field): try: return float(row[field]) except (KeyError, TypeError, ValueError): return None def derived_monthly_rent(row): """Estimate the unpublished 3-bed/semi cross-tab from three published marginals.""" all_rent = number(row, "Rental price") three_bed = number(row, "Rental price three bed") semi = number(row, "Rental price semidetached") if not all((all_rent, three_bed, semi)): return None return three_bed * semi / all_rent def quantile(values, q): values = sorted(values) x = (len(values) - 1) * q low, high = math.floor(x), math.ceil(x) if low == high: return values[low] return values[low] * (high - x) + values[high] * (x - low) def choose_source(postcode, by_code, by_name): code = postcode.get("local_authority_code", "") if code in by_code: return by_code[code], "local authority" rental_market_area = SCOTTISH_RENTAL_MARKET_AREA.get(postcode.get("local_authority_name", "")) if rental_market_area in by_name: return by_name[rental_market_area], "Scottish rental market area" region = postcode.get("region", "") if region in by_name: return by_name[region], "region" country_code = {"England": "E92000001", "Wales": "W92000004", "Scotland": "S92000003", "Northern Ireland": "N92000002"}.get(postcode.get("nation", "")) if country_code in by_code: return by_code[country_code], "country" raise ValueError(f"No usable published rent source for {postcode['postcode']}") def main(): parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--postcodes", default="postcodes-2026.csv") parser.add_argument("--rent-source", default="ons-private-rents-source-2026-08.csv") parser.add_argument("--model", default="lvc-register.py") parser.add_argument("--output-dir", default="published-output") parser.add_argument("--year", type=int, default=2026) parser.add_argument("--floor-area-sqm", type=float, default=90.0, help="Assumed representative floor area for a three-bedroom semi") parser.add_argument("--floor-quantile", type=float, default=0.10) parser.add_argument("--ceiling-quantile", type=float, default=0.90) args = parser.parse_args() if args.floor_area_sqm <= 0: raise SystemExit("--floor-area-sqm must be positive") postcodes = read_csv(args.postcodes) rents = read_csv(args.rent_source) by_code = {row["Area code"]: row for row in rents} by_name = {row["Area name"]: row for row in rents} output = Path(args.output_dir) output.mkdir(parents=True, exist_ok=True) evidence = [] for postcode in postcodes: source, geography_level = choose_source(postcode, by_code, by_name) estimate = derived_monthly_rent(source) if estimate is None: raise ValueError(f"Incomplete source measures for {source['Area name']}") evidence.append({ "postcode": postcode["postcode"], "use_class": "residential", "property_type": "semi", "bedroom_band": "3", "floor_area_sqm": f"{args.floor_area_sqm:.2f}", "annual_rent_gbp": f"{estimate * 12:.2f}", "year": args.year, "source": "ONS Price Index of Private Rents", "quality_weight": "1", "confidence_grade_override": "P", "method_override": "published_area_modelled_postcode", "estimated": "yes", "source_release": "19 August 2026", "source_period": source["Time period"], "source_url": SOURCE_URL, "source_geography_code": source["Area code"], "source_geography_name": source["Area name"], "source_geography_level": geography_level, "published_monthly_all_rent_gbp": source["Rental price"], "published_monthly_three_bed_rent_gbp": source["Rental price three bed"], "published_monthly_semidetached_rent_gbp": source["Rental price semidetached"], "estimated_monthly_three_bed_semi_rent_gbp": f"{estimate:.2f}", "floor_area_status": "user-adjustable assumption; not observed for this postcode", "combination_method": "three-bed rent x semi-detached rent / all-property rent", }) evidence_path = output / f"evidence_{args.year}.csv" write_csv(evidence_path, evidence, list(evidence[0])) # Set UK thresholds from unique published GB local-authority estimates, plus the # latest usable Northern Ireland national estimate. This avoids weighting the # threshold by however many sample postcodes happen to be selected in each area. population = [] for row in rents: code = row["Area code"] if code.startswith(("E06", "E07", "E08", "E09", "W06", "S12")) or code == "N92000002": monthly = derived_monthly_rent(row) if monthly is not None: population.append(monthly * 12 / args.floor_area_sqm) floor_value = quantile(population, args.floor_quantile) ceiling_value = quantile(population, args.ceiling_quantile) command = [ sys.executable, args.model, "build", "--postcodes", args.postcodes, "--evidence", str(evidence_path), "--year", str(args.year), "--floor-value", str(floor_value), "--ceiling-value", str(ceiling_value), "--output-dir", str(output), ] subprocess.run(command, check=True) print(f"Published-source areas used for thresholds: {len(population)}") print(f"Floor £{floor_value:.4f}/m2/year; ceiling £{ceiling_value:.4f}/m2/year") if __name__ == "__main__": main()