Datasets:
Rideshare hotspot forecasting: model-grain tables
Generated 2026-10-27T21:59:27+00:00 from commit bbe6ffb. Full column types in MANIFEST.json.
| file | rows | MB | time range | description |
|---|---|---|---|---|
nyc_zone_hour.parquet |
17,475,824 | 207.6 | 2019-02-01 00:00:00 β 2026-08-31 23:00:00 | Per NYC taxi zone (1-263) x local hour: pickups and dropoffs (zero-filled); median request-to-pickup wait, fare+tip, driver pay and trip miles of pickups. Source of hotspot labels and demand lags. |
chicago_zone_hour.parquet |
5,286,512 | 47.8 | 2018-11-01 00:00:00 β 2026-08-31 23:00:00 | Per Chicago community area (1-77) x local hour: pickups and dropoffs (zero-filled); median fare+tip and trip miles of pickups (no wait or driver pay published). |
nyc_city_hour.parquet |
66,448 | 0.5 | 2019-02-01 00:00:00 β 2026-08-31 23:00:00 | Citywide pickups per hour, pickups with no valid zone (264/265), dst_merged_hour. |
chicago_city_hour.parquet |
68,656 | 0.5 | 2018-11-01 00:00:00 β 2026-08-31 23:00:00 | Citywide pickups per hour, pickups with NULL community area, dst_merged_hour. |
nyc_drive_times.parquet |
18,630,100 | 62.6 | 2019 β 2026 | Median trip seconds per OD zone pair x year x hour of day x weekend. Use year y-1 for year y. |
chicago_drive_times.parquet |
2,201,421 | 7.3 | 2018 β 2026 | Median trip seconds per OD community-area pair x year x hour of day x weekend. |
zones_nyc.parquet |
263 | 1.4 | GeoParquet: taxi zone polygons, borough, name, area, centroid, subway station complexes. | |
zones_chicago.parquet |
77 | 0.7 | GeoParquet: community area polygons, name, area, centroid, 'L' stations. | |
zone_adjacency.parquet |
1,716 | 0.0 | Neighbouring zone pairs per city (polygons within 20 m), both directions, with centroid distance in km. | |
event_venues.parquet |
93 | 0.0 | Event venues (sports + concerts) with coordinates, their zone (NULL outside the zones) and, for some out-of-zone venues, the zones their crowds enter through. | |
events.parquet |
22,388 | 0.6 | 2018-08-21 11:00:00 β 2026-10-29 19:30:00 | Sports games (pro, college, US Open matches) and concert nights at NYC- and Chicago-area venues, with local start, estimated end, zone and attendance. |
weather_asos.parquet |
462,884 | 6.5 | 2018-01-01 00:00:00 β 2026-10-27 18:53:00 | ASOS observations (ORD, MDW, LGA, JFK, Central Park), routine + special METARs. |
nws_warnings.parquet |
85,657 | 1.5 | 2018-01-01 03:00:00 β 2026-10-28 16:00:00 | NWS VTEC products from LOT (Chicago) and OKX (NYC); gtype C rows carry UGC zone codes. |
bts_airport_hour.parquet |
346,352 | 6.1 | 2018-01-01 04:00:00 β 2026-09-01 20:00:00 | Flights per airport x local hour: scheduled/cancelled by scheduled time, delays by actual time. |
mta_zone_hour.parquet |
8,489,096 | 23.8 | 2020-07-01 00:00:00 β 2026-10-14 23:00:00 | Subway entries and transfers per taxi zone x hour, from 2020-07 (zones with a station only). |
mta_alerts_hour.parquet |
165,663 | 0.7 | 2018-01-01 00:00:00 β 2026-08-12 01:00:00 | MTA service alert counts per agency x hour (subway, bus, LIRR, MNR, bridges & tunnels). |
cta_area_day.parquet |
131,628 | 0.4 | 2018-01-01 β 2026-07-31 | CTA 'L' entries per community area x day (areas with a station only). Use the previous day. |
Conventions
- Times are naive local wall-clock (America/New_York, America/Chicago) unless the column ends in
_utc. - Hour grids skip the nonexistent spring-forward hour; the fall-back 01:00 hour holds two real hours
(
dst_merged_hourin*_city_hour). - Tables are indexed by when things happened. Lag every feature by >= 1 hour before predicting hour t (CTA: previous day; drive times: previous year).
- 0 means an observed zero; a missing row or NULL means no data. MTA ridership starts 2020-07; zones
without a station are absent from
mta_zone_hour/cta_area_day(see the station counts in the zones tables). Alert and flight tables only have rows for hours with activity. - Zone-hour
median_*columns describe that hour's pickups and are NULL when it had none. Dropoffs are counted where and when trips ended. Like pickups, all of them are only known once the hour is over, so lag them >= 1 hour. - Events: schedules are known in advance, so they may describe the target hour. ESPN gives start
times; end times are estimates (typical length per league). setlist.fm gives dates only, so
concerts are placed at 19:30-23:00. ESPN attendance is reported after the game. Canceled and
postponed games are dropped. Out-of-zone venues have
zoneNULL; some listentry_zoneswhere their crowds enter the city (e.g. MetLife -> Penn Station, Port Authority). - Splits are by time only: train 2019-2022 excluding 2020-03..2021-06, early-stop 2023, calibrate 2024, test 2025+.
Sources
NYC TLC HVFHV trip records; City of Chicago TNP trips, community areas and CTA ridership;
Iowa Environmental Mesonet (ASOS, NWS VTEC archive); BTS Reporting Carrier On-Time Performance;
MTA via data.ny.gov (subway hourly ridership, service alerts). Built by scripts/export.py.
- Downloads last month
- 6