#!/usr/bin/env python3 """Generate and seed ~2000 worker personas for the climate-risk pilot. Augments the 30 hand-crafted personas from populate_worker_personas.py with a programmatically-generated population distributed proportionally across the 15 Dar es Salaam neighborhoods by `worker_population_est`. Deterministic (seed=42) so re-running produces the same population. Run once (or any time — idempotent via ON CONFLICT DO UPDATE): DATABASE_URL=... python3 scripts/populate_worker_population.py [TOTAL] Default TOTAL = 2000. """ import os import random import sys import psycopg2 TARGET_TOTAL = int(sys.argv[1]) if len(sys.argv) > 1 else 2000 # 15 Dar zones with approximate enrollment basis (worker_population_est). # Counts are distributed proportionally so Mbagala (40k pop) gets more # synthesized personas than Masaki (4k pop). ZONES = [ ("DAR-BUG", "Buguruni", "informal", 22000), ("DAR-JAN", "Jangwani", "informal", 25000), ("DAR-KAR", "Kariakoo", "commercial", 12000), ("DAR-KIG", "Kigamboni", "mixed", 15000), ("DAR-KIN", "Kinondoni", "formal", 15000), ("DAR-MAG", "Magomeni", "mixed", 18000), ("DAR-MAS", "Masaki", "formal", 4000), ("DAR-MBA", "Mbagala", "informal", 40000), ("DAR-MIK", "Mikocheni", "formal", 8000), ("DAR-MNZ", "Manzese", "informal", 35000), ("DAR-MSA", "Msasani", "mixed", 8000), ("DAR-TAN", "Tandale", "informal", 28000), ("DAR-TEM", "Temeke", "mixed", 30000), ("DAR-UBU", "Ubungo", "mixed", 20000), ("DAR-VIN", "Vingunguti", "informal", 20000), ] MALE_FIRST = [ "Juma", "Hassan", "Mustafa", "Emmanuel", "Baraka", "Saidi", "Rashid", "Omari", "Peter", "Ibrahim", "Ally", "Athumani", "Joseph", "Hamisi", "Mohamed", "Daudi", "Yusuf", "Suleiman", "Rajabu", "Salim", "Amani", "Kassim", "Fred", "John", "George", "Samuel", "Henry", "James", "Paul", "Thomas", "Patrick", "Michael", "Charles", "Francis", "Robert", "Anthony", "Philip", "David", "Martin", "Simon", "Elias", "Joshua", "Samson", "Solomon", "Abraham", "Isaac", "Jacob", "Noah", "Joel", "Abdallah", "Issa", "Ramadhani", "Shabani", "Rweyongera", "Magufuli", "Mkapa", ] FEMALE_FIRST = [ "Amina", "Fatuma", "Mariam", "Zainabu", "Halima", "Rehema", "Asha", "Khadija", "Tatu", "Bahati", "Rose", "Grace", "Neema", "Mwanahamisi", "Esther", "Fausta", "Lucia", "Anna", "Mary", "Tabitha", "Magdalena", "Joyce", "Agnes", "Imani", "Sara", "Sofia", "Upendo", "Salma", "Tumaini", "Husna", "Zena", "Mwajabu", "Mwanajuma", "Riziki", "Subira", "Jane", "Ruth", "Eva", "Sarah", "Elizabeth", "Rachel", "Rebecca", "Hannah", "Naomi", "Deborah", "Martha", "Lydia", "Priscilla", "Phoebe", "Mwajuma", "Asia", "Halima", "Shamsa", "Latifa", "Nuru", "Aisha", ] SURNAMES = [ "Mwakalinga", "Shaaban", "Msuya", "Mushi", "Kimaro", "Kibona", "Maerere", "Mbwiliza", "Ngolwa", "Swai", "Ishengoma", "Mlelwa", "Athumani", "Hassan", "Juma", "Kilonzo", "Makawa", "Mlima", "Nkya", "Rweyemamu", "Sanga", "Temba", "Kaniki", "Mariki", "Lyimo", "Kikwete", "Mkumbo", "Msafiri", "Kateka", "Kimaryo", "Msigwa", "Mbowe", "Lema", "Mrema", "Kivuyo", "Bujiku", "Chijika", "Kilama", "Mbeya", "Matonya", "Njokopa", "Minja", "Mpemba", "Kilumbe", "Lugazia", "Chamshama", "Mkasa", "Mngumi", "Luhanga", "Mtelewa", "Kagaruki", "Ndimbo", "Ndaki", "Ngowi", "Chimanga", "Mollel", ] OCCUPATIONS = { "informal": [ "waste picker", "charcoal seller", "water seller", "water carrier", "day laborer", "street food cook", "charcoal trader", "fish vendor", "street tailor", "market porter", "domestic worker", "secondhand clothes vendor", "vegetable seller", "fruit hawker", "chapati cook", "sand carrier", "sack seller", "metalworker", "cobbler", "motorbike repair", ], "commercial": [ "market porter", "stall vendor", "food vendor", "tailor", "shoe repairer", "small shop keeper", "hardware hawker", "spice vendor", "fabric seller", "cloth hawker", "fruit vendor", "fish seller", "wholesale porter", "grain seller", "appliance hawker", "cosmetics vendor", ], "mixed": [ "boda-boda driver", "mama lishe", "car washer", "construction laborer", "gardener", "tailor", "fishmonger", "kiosk operator", "masonry helper", "painter", "welder", "glass cutter", "food stall operator", "yogurt seller", "nyama choma cook", "coconut vendor", "banana vendor", ], "formal": [ "security guard", "gardener", "driver", "housekeeper", "cleaner", "office messenger", "office gardener", "maintenance worker", "receptionist", "groundskeeper", "caretaker", "watchman", "janitor", "valet", "doorman", ], } # Mobile money share in Dar roughly follows M-Pesa dominance + Tigo + Airtel. MOBILE_MONEY = [("M-Pesa", 55), ("Tigo Pesa", 27), ("Airtel Money", 18)] # TASAF enrollment varies by settlement — higher in informal settlements. TASAF_PROB = {"informal": 0.40, "commercial": 0.15, "mixed": 0.20, "formal": 0.05} CREATE_TABLE = """ CREATE TABLE IF NOT EXISTS workers ( worker_id TEXT PRIMARY KEY, name TEXT NOT NULL, name_swahili TEXT, nida_id TEXT, phone TEXT NOT NULL, zone_id TEXT NOT NULL REFERENCES zones(zone_id), occupation TEXT NOT NULL, age INTEGER, years_outdoor INTEGER, household_size INTEGER, mobile_money TEXT, tasaf_enrolled BOOLEAN DEFAULT false, enrolled_at TIMESTAMPTZ DEFAULT NOW() ); CREATE INDEX IF NOT EXISTS idx_workers_zone ON workers (zone_id); """ UPSERT = """ INSERT INTO workers ( worker_id, name, name_swahili, nida_id, phone, zone_id, occupation, age, years_outdoor, household_size, mobile_money, tasaf_enrolled ) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s) ON CONFLICT (worker_id) DO UPDATE SET name = EXCLUDED.name, name_swahili = EXCLUDED.name_swahili, nida_id = EXCLUDED.nida_id, phone = EXCLUDED.phone, zone_id = EXCLUDED.zone_id, occupation = EXCLUDED.occupation, age = EXCLUDED.age, years_outdoor = EXCLUDED.years_outdoor, household_size = EXCLUDED.household_size, mobile_money = EXCLUDED.mobile_money, tasaf_enrolled = EXCLUDED.tasaf_enrolled """ def weighted_choice(rng: random.Random, weighted: list[tuple[str, int]]) -> str: total = sum(w for _, w in weighted) r = rng.uniform(0, total) for item, w in weighted: r -= w if r <= 0: return item return weighted[-1][0] def random_phone(rng: random.Random) -> str: prefix = rng.choice([12, 13, 14, 15, 16, 17, 18]) middle = rng.randint(0, 9) tail = rng.randint(1000, 9999) return f"+2557{prefix}{middle}XX{tail}" def random_nida(rng: random.Random) -> str: year = rng.randint(1965, 2004) month = rng.randint(1, 12) tail = rng.randint(1000, 9999) return f"{year}{month:02d}-XXXX-XXXX-{tail}" def generate_worker(rng: random.Random, zone_id: str, settlement: str, index: int) -> tuple: is_female = rng.random() < 0.48 first = rng.choice(FEMALE_FIRST if is_female else MALE_FIRST) surname = rng.choice(SURNAMES) name = f"{first} {surname}" age = rng.randint(20, 60) years_outdoor = rng.randint(1, min(max(age - 16, 1), 35)) household_size = rng.randint(2, 10) occupation = rng.choice(OCCUPATIONS[settlement]) mobile_money = weighted_choice(rng, MOBILE_MONEY) tasaf_enrolled = rng.random() < TASAF_PROB[settlement] worker_id = f"{zone_id}-P{index:04d}" return ( worker_id, name, first, random_nida(rng), random_phone(rng), zone_id, occupation, age, years_outdoor, household_size, mobile_money, tasaf_enrolled, ) def main() -> int: db_url = os.environ.get("DATABASE_URL") if not db_url: print("ERROR: DATABASE_URL not set. Export your Neon connection string.", file=sys.stderr) return 1 total_pop = sum(z[3] for z in ZONES) rng = random.Random(42) workers: list[tuple] = [] for zone_id, _name, settlement, pop in ZONES: count = max(1, round(TARGET_TOTAL * pop / total_pop)) for i in range(count): workers.append(generate_worker(rng, zone_id, settlement, i)) with psycopg2.connect(db_url) as conn: with conn.cursor() as cur: cur.execute(CREATE_TABLE) for row in workers: cur.execute(UPSERT, row) conn.commit() print(f"Seeded {len(workers)} worker personas across {len(ZONES)} " f"neighborhoods (target was {TARGET_TOTAL})") return 0 if __name__ == "__main__": raise SystemExit(main())