v0.28.0 (#46): per-region stats — area×population-weighted-median Nation.stats (PR 1 of 2)

2026-10-15 00:17:19 +00:00
parent 327fffb789
commit 22805a3be6
19 changed files with 8743 additions and 29 deletions
@@ -0,0 +1,165 @@
+/**
+ * scripts/build-region-geo.mjs — build the region **geo** manifest:
+ * `public/region-geo.json` = `{ "<regionKey>": { "area": <number>, "population": <number> }, ... }`.
+ *
+ * For every sub-region (region key, see `public/region-keys.json`) this records
+ * two facts the v0.28 per-region stats model (and its area×population weighting)
+ * needs:
+ *   - `area`       — the region's **total** surface area in km², from
+ *                    `@turf/area` on the FINAL generated feature
+ *                    (`public/regions.geo.json`) — turf reports square metres,
+ *                    so the raw value is divided by 1e6 — rounded to 2 decimals.
+ *   - `population` — the region's share of its **parent country's** real
+ *                    population (Natural Earth 1:10m admin-0 `POP_EST`, from the
+ *                    committed `public/ne-admin0-pop.json`), distributed across
+ *                    the country's sub-regions **proportional to each region's
+ *                    area** (a first-order "uniform population density within a
+ *                    country" approximation — Natural Earth admin-1, which the
+ *                    region generator uses, carries no per-unit population).
+ *
+ * Population distribution uses **largest remainder** per country, so the sum of
+ * a country's sub-region populations is EXACTLY the country's POP_EST, and the
+ * sum over all 1,800 regions is EXACTLY the sum of the 177 repo countries'
+ * POP_EST (7,654,092,021) — no rounding drift.
+ *
+ * Method — plain turf on the pre-generated geo file (turf is allowed in .mjs
+ * scripts; it is deliberately kept OUT of the TS typecheck program, see the
+ * v0.17 note). The population source is the COMMITTED `public/ne-admin0-pop.json`
+ * (a small derived table produced once by `scripts/extract-admin0-pop.mjs` from
+ * the 13 MB Natural Earth admin-0 file, which we do NOT commit) — so this build
+ * is fully reproducible in CI.
+ *
+ * Output contract (asserted by scripts/region-geo.test.ts):
+ *   - one entry for **every** region key, keyed identically to
+ *     `public/region-keys.json` (same key set, 1,800 entries);
+ *   - `area` is a finite number >= 0 (km², <= 2 decimals);
+ *   - `population` is a non-negative integer;
+ *   - the total population across all entries is a non-negative integer.
+ *
+ * Deterministic + idempotent: fixed inputs (the geo file, the key manifest, and
+ * the committed population table), pure turf, largest-remainder with a
+ * (remainder desc, key asc) tie-break, and stable sorted key order → a re-run
+ * rewrites a byte-identical file (the test re-runs this script and compares
+ * bytes).
+ *
+ * Usage: `node scripts/build-region-geo.mjs`  (or `npm run build:region-geo`).
+ */
+import { readFileSync, writeFileSync } from 'node:fs';
+import path from 'node:path';
+import { fileURLToPath } from 'node:url';
+import area from '@turf/area';
+import { feature } from '@turf/helpers';
+
+const here = path.dirname(fileURLToPath(import.meta.url));
+const root = path.resolve(here, '..');
+
+const round2 = (v) => Math.round(v * 100) / 100;
+
+// --- load the inputs (plain parse) ------------------------------------------
+const geoPath = path.join(root, 'public', 'regions.geo.json');
+const keysPath = path.join(root, 'public', 'region-keys.json');
+const popPath = path.join(root, 'public', 'ne-admin0-pop.json');
+
+let geo, idToKey, popByCountry;
+try {
+  geo = JSON.parse(readFileSync(geoPath, 'utf8'));
+  idToKey = JSON.parse(readFileSync(keysPath, 'utf8'));
+  popByCountry = JSON.parse(readFileSync(popPath, 'utf8'));
+} catch (err) {
+  console.error(
+    `[build-region-geo] could not load the inputs (${geoPath}, ${keysPath}, ${popPath}). ` +
+      `Is public/ne-admin0-pop.json present? Regenerate it with scripts/extract-admin0-pop.mjs.`,
+  );
+  throw err;
+}
+
+/**
+ * One pass over the geo features: for each region key record its (unrounded)
+ * turf area and its parent ISO_A3, then group the keys by parent so the
+ * population can be distributed per country.
+ */
+const regionArea = new Map(); // key -> unrounded area (km²)
+const regionParent = new Map(); // key -> parent ISO_A3
+for (const f of geo.features ?? []) {
+  const id = f.id == null ? '' : String(f.id);
+  const key = idToKey[id];
+  if (!key) continue; // a geo feature with no manifest entry is skipped
+  // @turf/area reports square METRES; the manifest stores km².
+  const a = area(feature(f.geometry)) / 1e6;
+  regionArea.set(key, a);
+  regionParent.set(key, typeof f.properties?.parent === 'string' ? f.properties.parent : '');
+}
+
+// Group region keys by parent country (file order is stable, but we sort for
+// deterministic population distribution).
+const keysByCountry = new Map(); // parent -> string[] of keys
+for (const [key, parent] of regionParent) {
+  if (!parent) continue; // a region with no parent has no population source
+  const list = keysByCountry.get(parent);
+  if (list) list.push(key);
+  else keysByCountry.set(parent, [key]);
+}
+
+const population = new Map(); // key -> integer population
+for (const [parent, keys] of keysByCountry) {
+  keys.sort(); // deterministic ordering for the tie-break
+  const P = Number.isFinite(popByCountry[parent]) ? Math.max(0, popByCountry[parent]) : 0;
+  const totalArea = keys.reduce((s, k) => s + (regionArea.get(k) || 0), 0);
+  if (P <= 0 || totalArea <= 0) {
+    for (const k of keys) population.set(k, 0);
+    continue;
+  }
+  // Largest remainder: floor each share, then hand out the leftover 1s to the
+  // regions with the largest fractional remainders (ties broken by key asc).
+  const shares = keys.map((k) => {
+    const a = regionArea.get(k) || 0;
+    const ideal = (P * a) / totalArea;
+    return { k, floor: Math.floor(ideal), rem: ideal - Math.floor(ideal) };
+  });
+  const distributed = shares.reduce((s, x) => s + x.floor, 0);
+  let deficit = P - distributed; // 0 <= deficit < keys.length
+  const order = shares
+    .map((x, i) => ({ i, rem: x.rem, k: x.k }))
+    .sort((a, b) => (b.rem - a.rem) || (a.k < b.k ? -1 : a.k > b.k ? 1 : 0));
+  for (let i = 0; i < order.length && deficit > 0; i++) {
+    shares[order[i].i].floor += 1;
+    deficit -= 1;
+  }
+  for (const x of shares) population.set(x.k, x.floor);
+}
+
+// --- assemble the manifest (sorted keys) -------------------------------------
+const expectedKeys = Object.values(idToKey);
+const manifest = {};
+for (const k of [...expectedKeys].sort()) {
+  const a = regionArea.get(k);
+  const p = population.get(k);
+  if (a === undefined || p === undefined) {
+    throw new Error(`region key ${JSON.stringify(k)} has no computed area/population`);
+  }
+  manifest[k] = { area: round2(a), population: p };
+}
+
+// --- hard assertions (the test re-checks these against the committed file) ----
+if (Object.keys(manifest).length !== expectedKeys.length) {
+  throw new Error(
+    `expected ${expectedKeys.length} region-geo entries (one per region key), got ${Object.keys(manifest).length}`,
+  );
+}
+let totalPop = 0;
+let zeros = 0;
+for (const [k, v] of Object.entries(manifest)) {
+  if (!Number.isFinite(v.area) || v.area < 0) throw new Error(`bad area for ${k}: ${JSON.stringify(v)}`);
+  if (!Number.isInteger(v.population) || v.population < 0) throw new Error(`bad population for ${k}: ${JSON.stringify(v)}`);
+  if (v.area === 0) zeros += 1;
+  totalPop += v.population;
+}
+
+// --- write idempotently (sorted keys, 2-space indent, trailing newline) ------
+writeFileSync(path.join(root, 'public', 'region-geo.json'), JSON.stringify(manifest, null, 2) + '\n');
+
+// --- summary ------------------------------------------------------------------
+console.log(
+  `[build-region-geo] wrote public/region-geo.json — ${Object.keys(manifest).length} region keys` +
+    `${zeros ? ` (${zeros} zero-area regions)` : ''}, total population ${totalPop.toLocaleString('en-US')}`,
+);