Dataset

airline-disruption-data

Foot Traffic & Mobility Download 1 source

airline-disruption-data is a tabular dataset of U.S. domestic flight records published on the datasets hub, tagged as covering between 10 million and 100 million rows in Parquet format. The schema describes per-flight scheduling, delay, cancellation, and routing fields.

Open on Hugging Face

Data preview

Subset

Loading preview...

Dataset card

What this dataset contains

The single train split holds 45,136,631 rows of flight-level records covering Year, Month, DayofMonth, DayOfWeek, and FlightDate. Carrier and routing are captured through Airline, FlightNumber, OperatingAirline, OperatingAirlineCode, Origin, OriginCityName, OriginState, Dest, DestCityName, and DestState. Operational outcomes are described by CRSDepTime, DepDelay, DepDelayMinutes, DepDel15, TaxiOut, TaxiIn, CRSArrTime, ArrDelay, ArrDelayMinutes, ArrDel15, Cancelled, CancellationCode, Diverted, AirTime, and Distance. The license, file size, and a textual README description are not stated by the publisher.

Structure and fields

All fields are typed as scalar Value columns, mixing integer codes (Year, DepDel15, ArrDel15, Cancelled, Diverted), floating-point measurements (DepDelay, ArrDelay, TaxiOut, TaxiIn, AirTime, Distance), timestamps (FlightDate), and string identifiers (Airline, Origin, Dest, CancellationCode). Numeric granularities range from int8 flags to int64 flight numbers, with delay magnitudes stored as float32. The dataset is exposed through the datasets, dask, polars, and mlcroissant libraries, indicating it is distributed as a Parquet table loadable into multiple query engines.

How analysts use it

Analysts can aggregate DepDelayMinutes, ArrDelayMinutes, and Cancelled by Airline or Origin-Dest pair to benchmark on-time performance and disruption frequency over time. The schema also supports seasonality studies using Month, DayOfWeek, and FlightDate, and route-economics work combining Distance and AirTime. Because the publisher has not stated a license or documented provenance, downstream use for investment signals should be paired with independent verification of the source.

Documentation & references 1