withdrawal data

The withdrawal data is defined at the rally level for each category and identifies the teams/crews that have withdrawn during each stage of the rally.

# Load in the required packages
import pandas as pd
from jupyterlite_simple_cors_proxy import furl, xurl

# Generate the API URL pattern
dakar_api_template = "https://www.dakar.live.worldrallyraidchampionship.com/api/{path}"

# Define the year
YEAR = 2025
# Define the category
CATEGORY = "A"

# Define the API path to the withdrawal resource
# Use a Python f-string to instantiate variable values directly
withdrawal_path = f"withdrawal-{YEAR}-{CATEGORY}"

# Define the URL
withdrawal_url = dakar_api_template.format(path=withdrawal_path)

# Preview the path and the URL
withdrawal_path, withdrawal_url

('withdrawal-2025-A',
 'https://www.dakar.live.worldrallyraidchampionship.com/api/withdrawal-2025-A')

Within the withdrawal data, we get a list of withdrawals by stage.

# Load in data
# Use furl() to handle CORS issues in Jupyterlite
withdrawal_df = pd.read_json(furl(withdrawal_url))

# Use the stage number as the index
withdrawal_df.set_index("stage", drop=False, inplace=True)
withdrawal_df

	_id	_bind	_updatedAt	_parent	list	stage
stage
8	withdrawal-2025-A-8	withdrawal-2025-A	1736877543094	stage-2025-A:d7ab3acebdb4cd17e8d4af9f122fc425	[{'bib': 619, 'reason': '1', 'team': {'bib': 6...	8
7	withdrawal-2025-A-7	withdrawal-2025-A	1736877541255	stage-2025-A:fadf454f8df8f758e326ad268007a286	[{'bib': 202, 'reason': '1', 'team': {'bib': 2...	7
5	withdrawal-2025-A-5	withdrawal-2025-A	1736877542722	stage-2025-A:e2b95b95a4ae9063eb4924fc2d7cf7e7	[{'bib': 208, 'reason': '1', 'team': {'bib': 2...	5
3	withdrawal-2025-A-3	withdrawal-2025-A	1736877554109	stage-2025-A:127ed31c3eff071ba6a7e88fe01083ef	[{'bib': 225, 'reason': '1', 'team': {'bib': 2...	3
9	withdrawal-2025-A-9	withdrawal-2025-A	1736943602310	stage-2025-A:5f83172bf9f0bb97d14b83a7c0099fa8	[{'bib': 213, 'reason': '1', 'team': {'bib': 2...	9
6	withdrawal-2025-A-6	withdrawal-2025-A	1736877552621	stage-2025-A:33ee273dbd14a1657513f7dab64a635d	[{'bib': 239, 'reason': '1', 'team': {'bib': 2...	6
2	withdrawal-2025-A-2	withdrawal-2025-A	1736877545655	stage-2025-A:9afdc50df69247c38b23cef4ed50bd14	[{'bib': 223, 'reason': '1', 'team': {'bib': 2...	2
4	withdrawal-2025-A-4	withdrawal-2025-A	1736877547536	stage-2025-A:8002af597f2d091dc91450266fab52d6	[{'bib': 219, 'reason': '2', 'team': {'bib': 2...	4
10	withdrawal-2025-A-10	withdrawal-2025-A	1736960963500	stage-2025-A:91f8c3e34c474fe774498af9a05e08c9	[{'bib': 205, 'reason': '1', 'team': {'bib': 2...	10

# Tweak the displayu of the dataframes, wrapping the display over
# several lines if the df is too wide
from dakar_rallydj.stylers import enable_split_display

enable_split_display(width=80, max_rows=10)
# Params: width, max_colwidth, max_rows

ERROR: max_colwidth (30) greater than width (30); resetting down.

withdrawals_by_stage = withdrawal_df["list"].explode()
withdrawals_by_stage_index = withdrawals_by_stage.index

withdrawals_by_stage_df = pd.json_normalize(withdrawals_by_stage)

withdrawals_by_stage_df["stage"] = withdrawals_by_stage_index
withdrawals_by_stage_df#.head();

bib	reason	team.bib
619	1	619
202	1	202
206	1	206
220	1	220
231	1	231
...	...	...
439	1	439
609	1	609
205	1	205
322	1	322
640	1	640

team.brand
MAN
MINI
TOYOTA
HAN WEI MOTORSPORT
TOYOTA
...
BRP
IVECO
TOYOTA
MMP
MAN

team.model
TGA
JCW RALLY 3.0I
HILUX IMT EVO
HWM T1+
HILUX OVERDRIVE
...
CAN-AM MAVERICK XRS TURBO RR
POWERSTAR
HILUX IMT EVO
RALLY RAID
TGA

team.vehicle
Eagle Racing Team
X-RAID MINI JCW TEAM
TOYOTA GAZOO RACING
212 HANWEI MOTORSPORT TEAM
OVERDRIVE RACING
...
PEDREGÀ TEAM
De Groot Sport
TOYOTA GAZOO RACING
MMP
TH-TRUCKS TEAM

team.vehicleImg
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
https://img.aso.fr/core_ap...
nan

team.clazz
fa2b8d6b8d84480b3fd11a2fd8...
96c0869600e0013dbf5f86f60e...
96c0869600e0013dbf5f86f60e...
96c0869600e0013dbf5f86f60e...
96c0869600e0013dbf5f86f60e...
...
058d77cc7db191813c30a902a8...
fa2b8d6b8d84480b3fd11a2fd8...
96c0869600e0013dbf5f86f60e...
a0a6386a4b9a61b73b036a5096...
ec2f26ebeb14824160c7204618...

team.w2rc
False
False
False
False
False
...
False
False
False
False
False

team.competitors	stage
[{'name': 'P. CALABRIA', '...	8
[{'name': 'G. CHICHERIT', ...	7
[{'name': 'G. DE VILLIERS'...	7
[{'name': 'W. HAN', 'first...	7
[{'name': 'T. PRICE', 'fir...	7
...	...
[{'name': 'R. MUÑOZ CAMARA...	4
[{'name': 'B. DE GROOT', '...	4
[{'name': 'G. BOTTERILL', ...	10
[{'name': 'R. POREM', 'fir...	10
[{'name': 'A. ALONSO ROMER...	10

The withdrawal data also includes a reason column, although I haven’t (yet!) found a metadata feed that explains the reason code values.

Crew (that is, competitors) data is provided as a list of details for each crew member. We can extract this data into its own table, and widen it by adding the stage and reason data, to provide an alternative table for looking up withdrawal information by competitor name.

withdrawn_competitors_df = (
        withdrawals_by_stage_df[['stage', 'reason', 'bib', 'team.competitors']]
        .explode('team.competitors')
        .reset_index(drop=True)
    )

# Normalize the dictionary contents and combine with competitors
withdrawn_competitors_df = pd.concat([
    withdrawn_competitors_df[['stage', 'bib', 'reason']],
    pd.json_normalize(withdrawn_competitors_df['team.competitors'])
    ], axis=1)

withdrawn_competitors_df.head()

	stage	bib	reason	name	firstName	lastName	role	gender	nationality	profil	profil_sm	podium	aid
0	4	219	2	S. LOEB	SEBASTIEN	LOEB	P	m	fra	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	5dd0184f-5f90-4b42-b675-88d6501b9965
1	4	219	2	F. LURQUIN	FABIAN	LURQUIN	C	m	bel	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	c640056a-79d0-4454-8e6a-c8b303620020
2	4	236	1	E. AMOS	EUGENIO	AMOS	P	m	ita	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	0575de36-4a38-431c-8892-ac9514c8ccd9
3	4	236	1	P. CECI	PAOLO	CECI	C	m	ita	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	fdacd829-75f9-4cfa-9992-636c8244481e
4	4	258	1	M. WALCHER	MARKUS	WALCHER	P	m	deu	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	93be84ca-0b2b-4da0-933a-e394a8a29cdb

We note, however, that much of this information is generic and could exist in a separate competitors data table, with just a unique key reference value, such as the name, linking the withdrawn competitors table to the competitors data table.

A similar approach could be applied to the team values, extracting the data to a separate teams data table and using the team.bib as the unique key.

team_cols = [c for c in withdrawals_by_stage_df.columns if c.startswith("team")]

# Create a new dataframe, rather than a reference, by using .copy()
withdrawn_teams_df = withdrawals_by_stage_df[team_cols].copy()
withdrawn_teams_df.drop("team.competitors", axis=1, inplace=True)

withdrawn_teams_df.head()

	team.bib	team.brand	team.model	team.vehicle	team.vehicleImg	team.clazz	team.w2rc
0	219	DACIA	SANDRIDER	THE DACIA SANDRIDERS	https://img.aso.fr/core_app/img-motorSports-da...	96c0869600e0013dbf5f86f60e5c4da4	True
1	236	MD	OPTIMUS	MD RALLYE SPORT	https://img.aso.fr/core_app/img-motorSports-da...	f00d7ec8d2d96e9cf11aa515109376cf	False
2	258	RED-LINED	NAVARRA VK 56	WALCHER RACING TEAM	https://img.aso.fr/core_app/img-motorSports-da...	f666973e89db183ecfefc75c3af8ffb1	False
3	337	BRP	CAN-AM MAVERICK X3	COLORADO MOTORSPORT	NaN	a0a6386a4b9a61b73b036a50966345c0	False
4	408	BRP	CAN-AM MAVERICK R	SOUTH RACING CAN-AM	https://img.aso.fr/core_app/img-motorSports-da...	9a68ed3c41c5c7a1642df5d93458baa6	True

Create a simpler withdrawals table:

# If we have more than one team member, we will get duplicate rows
withdrawals_df = withdrawn_competitors_df[["stage", "bib", "reason"]].drop_duplicates()
withdrawals_df.head()

	stage	bib	reason	name
0	4	219	2	S. LOEB
1	4	219	2	F. LURQUIN
2	4	236	1	E. AMOS
3	4	236	1	P. CECI
4	4	258	1	M. WALCHER

And tidy up the competitors table to just leave competitor (and team bib) data:

withdrawn_competitors_df.drop("stage", axis=1, inplace=True)
withdrawn_competitors_df.drop("reason", axis=1, inplace=True)

withdrawn_competitors_df.head()

	bib	name	firstName	lastName	role	gender	nationality	profil	profil_sm	podium	aid
0	219	S. LOEB	SEBASTIEN	LOEB	P	m	fra	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	5dd0184f-5f90-4b42-b675-88d6501b9965
1	219	F. LURQUIN	FABIAN	LURQUIN	C	m	bel	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	c640056a-79d0-4454-8e6a-c8b303620020
2	236	E. AMOS	EUGENIO	AMOS	P	m	ita	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	0575de36-4a38-431c-8892-ac9514c8ccd9
3	236	P. CECI	PAOLO	CECI	C	m	ita	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	fdacd829-75f9-4cfa-9992-636c8244481e
4	258	M. WALCHER	MARKUS	WALCHER	P	m	deu	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	https://img.aso.fr/core_app/img-motorSports-da...	93be84ca-0b2b-4da0-933a-e394a8a29cdb