US Flight Delays & On
The live successor to the famous frozen-at-2015 Kaggle flight-delays dataset. US DOT's Bureau of Transportation Statistics publishes ~560k US domestic flights every month; we aggregate them into one tidy panel - for each airline × origin airport × month: the flight count, on-time / cancelled / diverted rates, average departure & arrival delay, and the average minutes attributable to each delay cause (carrier, weather, national-airspace, security, late aircraft). One row = a carrier's monthly reliability at an airport. Public domain.
- Source
- U.S. BTS
- Licence
- Redistributable open data
- Quality
- QC clean · 0 failures
- Format
- Apache Parquet + dictionary
Safe to train on
Built only from a named, redistributable official source under a documented open licence - not scraped web data. No copyright grey zone, no personal data. Ships machine-readable Croissant metadata (ML Commons - loads in Hugging Face / Kaggle / Google), an AI training-licence manifest documenting source, licence and provenance for your model's data governance, and a machine-readable data dictionary (drop it into an agent / RAG prompt so the model knows every column) - all inspectable before you buy.
What's included
- Monthly on-time / cancelled / diverted rates per airline × origin airport, 2018→present
- Average departure & arrival delay + delay-cause breakdown (carrier/weather/NAS/security/late-aircraft)
- The live, updated version of the most-used flight-delays dataset on Kaggle
- One compact row per (carrier × airport × month) - ~150k rows, easy to pivot/join
- Public domain (U.S. DOT / BTS) - no personal data, no endorsement implied
- Source: U.S. DOT, Bureau of Transportation Statistics - TranStats On-Time Performance
Schema
| Column | Type | Description | Filled | Distinct |
|---|---|---|---|---|
| carrier | str | Reporting airline code | 100.0% | 19 |
| origin | str | Origin airport (IATA) | 100.0% | 389 |
| origin_city | str | Origin city | 100.0% | 382 |
| origin_state | str | Origin US state | 100.0% | 53 |
| year | int | Year | 100.0% | 9 |
| month | int | Month (1-12) | 100.0% | 12 |
| n_flights | int | Scheduled flights | 100.0% | 5,407 |
| pct_on_time | float | % of operated flights arriving <15 min late | 99.9% | 732 |
| pct_cancelled | float | % of flights cancelled | 100.0% | 720 |
| pct_diverted | float | % of flights diverted | 100.0% | 517 |
| avg_dep_delay | float | Average departure delay (minutes) | 99.9% | 7,424 |
| avg_arr_delay | float | Average arrival delay (minutes) | 99.9% | 7,529 |
| avg_carrier_delay | float | Avg minutes of carrier-caused delay per operated flight | 99.9% | 4,645 |
| avg_weather_delay | float | Avg minutes of weather delay per operated flight | 99.9% | 2,444 |
| avg_nas_delay | float | Avg minutes of national-airspace delay per operated flight | 99.9% | 2,955 |
| avg_security_delay | float | Avg minutes of security delay per operated flight | 99.9% | 495 |
| avg_late_aircraft_delay | float | Avg minutes of late-aircraft delay per operated flight | 99.9% | 4,335 |
Sample & preview
Every purchase ships as Apache Parquet with a data dictionary and the full QC report. A free sample (first rows + schema) is downloadable here - confirm fit before you buy. Source: U.S. BTS (redistributable open data; attribution passes through - see our Licence).
Don't trust screenshots - drop the sample into your notebook right now.
import pandas as pd df = pd.read_parquet("us-flight-delays_sample.parquet") df.info() # typed columns, gap-aware, ready to join⬇ Download free Parquet sample or CSV for Excel
Related datasets
European EV Charging Buildout (EAFO)
The installed base of electric-vehicle recharging points across Europe, per country over time - the European mirror of the US EV-charging buildout. From the European Alternative Fuels Observatory (EAFO/DG MOVE): the AFIR-methodology quarterly AC/DC split (2020→present) plus the pre-AFIR annual totals (2012-2019), stitched into one clean multi-country Parquet keyed by country code. Note: AFIR counts recharging POINTS (connectors), not stations/sites.
European Transport
The land-transport pulse of Europe in one panel: rail freight (thousand tonnes) and rail passengers (million passenger-km), plus the passenger-car fleet and new-car registrations split by fuel (total / petrol / diesel). Cleaned from Eurostat transport statistics into one tidy annual per-country table - the mobility and fleet-transition signal that complements our road-freight and air-passenger products.
European Road Freight
The road-logistics pulse of Europe in one clean Parquet: annual road freight moved (million tonne-kilometres) by each country's hauliers, split into national and international transport, one row per country × series × year since 2000. A read on trade flows, supply-chain activity and the freight economy; joins to macro, trade and energy on country code.