1.7.1 The data🔗ℹ

The post’s file is the flights table of the R package nycflights13 (Hadley Wickham, released under CC0; the data come from the US Bureau of Transportation Statistics), as the TSV at https://www.travishinkelman.com/data/nycflights.tsv: 336,776 rows × 19 columns. The examples here run while the manual builds, with no network, so they read "nycflights-sample.tsv", which ships with the package: the header and the 5,434 flights on the first day of January, March, May, July, September and November, their lines unchanged (492 KB).

  awk -F'\t' 'NR == 1 || ($3 == 1 && $2 % 2 == 1)' nycflights.tsv

Counts below are the sample’s; the prose gives the full file’s, from nix run .#bench in the repository. The bench fetches the post’s file, checks its hash, checks the post’s steps and complaints on it (the scoreboard), and times the operations against Python polars; its ratio table closes the chapter.