alltime_athletics_python

Scrapes Peter Larsson's website Alltime Athletics. Check out my blog to see how this data can be visualized.

Just give me the data

The latest data frame can be found here as a csv, or here in parquet format. Or run

import pandas as pd

df = pd.read_csv(
     "https://media.githubusercontent.com/media/thomascamminady/alltime_athletics_python/main/dataframes/latest_version_alltime_athletics.csv"
)

As an example, here are the women's world record performances (as of 2023-06-12), sorted by the date of the world record.

	event	name	result	date of event
0	800 metres	Jarmila Kratochvílová	1:53.28	1983-07-26
1	400 metres	Marita Koch	47.60	1985-10-06
2	100 metres	Florence Griffith-Joyner	10.49	1988-07-16
3	200 metres	Florence Griffith-Joyner	21.34	1988-09-29
4	60 metres	Irina Privalova	6.92	1993-02-11
5	3000 metres	Wang Junxia	8:06.11	1993-09-13
6	60 metres	Irina Privalova	6.92	1995-02-09
7	10 km race walk	Yelena Nikolayeva	41:04	1996-04-20
8	1000 metres	Svetlana Masterkova	2:28.98	1996-08-23
9	5000 metres track walk	Gillian O'Sullivan	20:02.60	2002-07-14
10	2 Miles	Meseret Defar	8:58.58	2007-09-14
11	20 km race walk	Yelena Lashmanova	1:23:39	2018-06-09
12	3000m steeplechase	Beatrice Chepkoech	8:44.32	2018-07-20
13	300 metres	Shaunae Miller-Uibo	34.41	2019-06-20
14	1 Mile	Sifan Hassan	4:12.33	2019-07-12
15	2000m steeplechase	Gesa Felicitas Krause	5:52.80	2019-09-01
16	marathon	Brigid Kosgei	2:14:04	2019-10-13
17	15km road	Letesenbet Gidey	44:20	2019-11-17
18	50 km race walk	Yelena Lashmanova	3:50:42	2020-09-05
19	10000 metres	Letesenbet Gidey	29:01.03	2021-06-08
20	2000 metres	Francine Niyonsaba	5:21.56	2021-09-14
21	half-marathon	Letesenbet Gidey	62:52	2021-10-24
22	20km road	Letesenbet Gidey	59:46+	2021-10-24
23	10km road	Yalemzerf Yehualaw	29:14	2022-02-27
24	400m hurdles	Sydney McLaughlin-Levrone	50.68	2022-07-22
25	100m/110m hurdles	Oluwatobiloba Amusan	12.12	2022-07-24
26	30km road	Ruth Chepngetich	1:34:01+	2022-10-09
27	1500 metres	Faith Kipyegon	3:49.11	2023-06-02
28	5000 metres	Faith Kipyegon	14:05.20	2023-06-09

You would get this table by using polars and running

import polars as pl

df = pl.read_csv(
     "https://media.githubusercontent.com/media/thomascamminady/alltime_athletics_python/main/dataframes/latest_version_alltime_athletics.csv"
)

(
 df.filter(pl.col("rank") == 1)
 .filter(pl.col("sex") == "female")
 .select("event", "name", "result", "date of event")
 .sort("date of event")
)

Download

If you have cloned the source code, you can run

poetry run python alltime_athletics_python/app.py

If you installed this package from PyPI, run

from alltime_athletics_python.io import download_data
download_data()

Note that download_data() reads data from Alltime Athletics AS IS. You will definitely need to do some postprocessing.

Postprocessing

To read the processed data, run

from alltime_athletics_python.io import import_running_only_events
df = import_running_only_events("./data")

Development

To set up the project, simply run

make init

Credits

This tool does not take credit for the amazing effort by Peter Larsson, who compiles Alltime Athletics. Alltime Athletics is an amazing collection of track and field results with a lot of work that must have gone into it. Thank you, Peter Larsson.

The only functionality that this tool provides is to have an easier way to read data from Alltime Athletics.

This package was created with Cookiecutter and thomascamminady/cookiecutter-pypackage, a fork of the audreyr/cookiecutter-pypackage project template.