hsanchezp/us-dot-flight-delays-2015
收藏资源简介:
# US DOT Flight Delays — 2015 (Parquet Version) This dataset contains 5,819,079 records of commercial flights in the United States during 2015. It has been converted to Parquet for efficient analytics in Python, DuckDB, Ibis, Spark, and Polars. ## Files included - flights.parquet (main fact table) - airlines.parquet (carrier info) - airports.parquet (airport geolocation & metadata) - cancellation_codes.parquet (mapping table) ## Source U.S. Department of Transportation — Bureau of Transportation Statistics Public Domain (U.S. Government Work) Original dataset: https://www.transtats.bts.gov/ ## Notes - Original data downloaded from the U.S. DOT (via Maven Analytics frontend). - This version is provided as a Parquet file for efficient loading in Python, DuckDB, Polars, and Ibis. ## Schema - 39 fields in flights.parquet - Includes departure/arrival times, delays, distance, carrier, and airport identifiers ## Recommended Usage ### Python (Polars) ```python import polars as pl flights = pl.read_parquet("flights.parquet") airlines = pl.read_parquet("airlines.parquet") airports = pl.read_parquet("airports.parquet")



