import json,pickle,datetime as dt,statistics,collections,csv,os OUT=r'C:\dev\FactHub-Tester\thanksgiving\busiest-travel-day\data' tsa=json.load(open('tsa/tsa.json')) def tg(y): x=dt.date(y,11,1) while x.weekday()!=3: x+=dt.timedelta(1) return x+dt.timedelta(21) def label(d): t=tg(d.year) if d.month>=11 else None if d.month==12 and d.day<=3: t=tg(d.year) if not t: return 'Other day' off=(d-t).days return {-1:'Wednesday before Thanksgiving',0:'Thanksgiving Day',3:'Sunday after Thanksgiving'}.get(off,'Other day') # table 1 tsa daily with open(os.path.join(OUT,'tsa_daily.csv'),'w',newline='') as f: w=csv.writer(f); w.writerow(['date','weekday','travelers_screened','thanksgiving_day_type']) for k,v in sorted(tsa.items()): d=dt.date.fromisoformat(k); w.writerow([k,d.strftime('%A'),v,label(d)]) # table 2 tsa ranks rows=[] for y in range(2019,2026): yd={k:v for k,v in tsa.items() if k.startswith(str(y))}; rk=sorted(yd,key=lambda k:-yd[k]); t=tg(y) for off,nm in [(-2,'Tuesday before'),(-1,'Wednesday before'),(0,'Thanksgiving Day'),(1,'Friday after'),(2,'Saturday after'),(3,'Sunday after'),(4,'Monday after')]: k=str(t+dt.timedelta(off)) rows.append(dict(year=y,day=nm,date=k,travelers_screened=yd[k],rank_in_year=rk.index(k)+1,days_in_year=len(yd),days_busier=rk.index(k),busiest_day_of_year=rk[0],busiest_day_travelers=yd[rk[0]])) with open(os.path.join(OUT,'tsa_thanksgiving_ranks.csv'),'w',newline='') as f: w=csv.DictWriter(f,fieldnames=list(rows[0])); w.writeheader(); w.writerows(rows) json.dump(rows,open('tsa_ranks.json','w')) # road index road_rows=[]; road_ranks=[]; strict=[] for year in (2023,2024): data,fsys=pickle.load(open(f'tmas/agg_{year}.pkl','rb')) nd=(dt.date(year+1,1,1)-dt.date(year,1,1)).days days=[str(dt.date(year,1,1)+dt.timedelta(i)) for i in range(nd)] res={}; meta={} for lab,flt in [('all_roads',lambda f:True),('rural_interstates',lambda f:f=='1R'),('urban_interstates',lambda f:f=='1U')]: idx=collections.defaultdict(list); n=0; st=set() for k,v in data.items(): if len(v)0 and min(vals)>=0.1*statistics.median(vals) and max(vals)<=5*statistics.median(vals))([data[k][d] for d in days])] tot={d:sum(data[k][d] for k in good) for d in days}; rk=sorted(days,key=lambda d:-tot[d]); w=str(tg(year)-dt.timedelta(1)) strict.append(dict(year=year,lanes=len(good),states=len({k[0] for k in good}),wednesday_vehicles=tot[w],wednesday_rank=rk.index(w)+1,busiest_day=rk[0],busiest_day_vehicles=tot[rk[0]])) print(year,meta) with open(os.path.join(OUT,'road_daily_index.csv'),'w',newline='') as f: w=csv.DictWriter(f,fieldnames=list(road_rows[0])); w.writeheader(); w.writerows(road_rows) with open(os.path.join(OUT,'road_thanksgiving_ranks.csv'),'w',newline='') as f: w=csv.DictWriter(f,fieldnames=list(road_ranks[0])); w.writeheader(); w.writerows(road_ranks) json.dump(road_ranks,open('road_ranks.json','w')); json.dump(strict,open('road_strict.json','w')) print(strict) for r in road_ranks: if r['day'] in('Wednesday before','Sunday after','Thanksgiving Day'): print(r['year'],r['road_group'],r['day'],r['rank_in_year'],r['traffic_index'],r['busiest_day_of_year'],r['busiest_day_index'],r['lanes'],r['states'])