8,000 rows · 10 columns

Flight delays

Eight thousand domestic flights across six carriers and fourteen airports, with departure and arrival delay, cancellations, and a delay distribution that is long-tailed rather than normal.

Download CSV 361 KB Open in Explorer Query it in SQL

What is in it, and what is wrong with it

Cancelled flights carry null delays rather than zeros, so an average that silently drops nulls tells a different story from one that treats a cancellation as a very long delay. Delay is not symmetric: the median flight leaves a minute or two early, while the mean is dragged positive by a small number of very bad days. This is the dataset to use when someone asks for "average delay".

Schema

Column	Type	Description
flight_date	date	Scheduled date of departure.
carrier	category	Two-letter carrier code.
flight_no	integer	Flight number.
origin	category	Origin airport code.
dest	category	Destination airport code.
sched_dep_hour	integer	Scheduled departure hour, 0–23.
dep_delay_min	integer	Departure delay in minutes; negative means early. Null if cancelled.
arr_delay_min	integer	Arrival delay in minutes. Null if cancelled.
distance_mi	integer	Great-circle distance in miles.
cancelled	boolean	Whether the flight was cancelled.

First twelve rows

flight_date	carrier	flight_no	origin	dest	sched_dep_hour	dep_delay_min	arr_delay_min
2023-01-01	WN	4324	DFW	ATL	17	15	28
2023-01-01	UA	875	DTW	ORD	14	227	226
2023-01-01	AS	795	SEA	SFO	7	3	-1
2023-01-01	WN	6229	DEN	ORD	9	28	21
2023-01-01	WN	5021	MCO	MSP	9	3	-1
2023-01-01	UA	6211	PHX	JFK	8	1	-2
2023-01-01	AA	3212	PHX	JFK	14	126	123
2023-01-01	AS	6303	ATL	BOS	12	6	4
2023-01-01	AS	4752	PHX	SFO	20	—	—
2023-01-01	WN	871	ORD	BOS	21	12	-7
2023-01-01	AS	5276	BOS	JFK	9	-4	0
2023-01-01	B6	3571	SEA	DFW	9	-17	-30

Showing the first 8 of 10 columns. Open the full table in the Explorer.

Grain

One row per scheduled flight.

Things to try

Show why the mean and median delay disagree, with a histogram.
Rank carriers on on-time rate — then check whether route mix explains it.
Do delays propagate through the day? Look at delay by scheduled hour.
Build a simple classifier for "will arrive more than 15 minutes late".

Load it

python

import pandas as pd

df = pd.read_csv(
  "https://datanemics.com/datasets/data/flight-delays.csv"
)

Licence

Public domain, CC0 1.0. Synthetic data — no attribution required.