qatarcars 0.2.0

A package containing data on Qatari cars for pedagogical purposes

QatarCars – Modern Passenger Cars Data

QatarCars is a lightweight CSV describing modern passenger cars from around the world. The aim is to give students a small, well‑structured dataset that still covers a range of data‑science tasks (merging, filtering, statistics, machine‑learning). It was originally created in 2025 by Paul Musgrave and his International Politics statistics class at Georgetown University. This package ports the dataset into Python and allows users to decide whether to engage with the data as either a pandas or polars dataframe. QatarCars is available in its original form as a csv, in Stata, and as an R Package.

Installation

From pypi:

pip install qatarcars

From GitHub:

pip install git+https://github.com/prlitics/qatarcars.git

What’s Included?

Column	Description
origin	A string denoting nation of origin of the car
make	A string denoting the manufacturer/brand of the car
model	A string denoting the specific type of the car
length	A float denoting the car’s length (in meters)
width	A float denoting the car’s width (in meters)
height	A float denoting the car’s height (in meters)
seating	An integer denoting how many seats are within the car
trunk	An integer denoting the trunk’s volume (in liters)
economy	A float denoting how many liters of fuel is required to travel 100km
performance	A float denoting how many seconds it takes to accelerate to 100km/h from a dead stop
mass	A float of the car’s mass (in kg)
price	An integer of the vehicles’ price (QAR)
horsepower	An integer denoting the car’s horsepower
type	A string denoting the body-type of the car
enginetype	A string denoting the type of fuel/energy used by the engine

Polars or Pandas? You decide!

Pandas is pretty much ubiquitous in Python data analytics, often considered the “default” implementation of a dataframe. However, Polars is gaining in popularity because of its more consistent and (especially comming to Python from R) intuitive syntax. It’s also very fast, which can be great when dealing with large data. However, the syntax of Polars compared to Pandas is often quite different. So, for pedagogical/learning purposes, I’ve made it so that the core function for the package get_qatar_cars can return either a pandas or polars dataframe. It’s up to you!

Examples

Let’s look at the first few observations. We’ll visualize using plotnine which is a great import of the ggplot2 package into Python.

from qatarcars import get_qatar_cars
import plotnine as p9

from qatarcars import get_qatar_cars
df = get_qatar_cars("pandas")   # or "polars"

df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } .dataframe tbody tr th { vertical-align: top; } .dataframe thead th { text-align: right; } </style>

	origin	make	model	length	width	height	seating	trunk	economy	horsepower	price	mass	performance	type	enginetype
0	Germany	BMW	8 Series Gran Coupe	5.082	1.932	1.407	4	440	7.2	340	435000	2090	5.2	Coupe	Petrol
1	Italy	Bugatti	Centodieci	4.544	2.047	1.200	2	0	22.5	1600	33000000	1976	2.4	Coupe	Petrol
2	Germany	Porsche	991 Carrera S	4.510	1.852	1.300	4	132	11.2	473	530000	1495	3.3	Coupe	Petrol
3	Japan	Lexus	RC F	4.695	2.069	1.395	4	66	11.5	471	320000	1770	4.5	Coupe	Petrol
4	Japan	Toyota	Supra	4.379	1.854	1.292	2	290	7.2	382	220000	1410	4.1	Coupe	Petrol

I’m curious how much a vehicle’s weight impacts its ability to get to 100 km/hr from a dead stop?

(p9.ggplot(df) +
p9.aes(x = 'mass', y = 'performance') +
p9.geom_point() +
p9.geom_smooth() +
p9.labs(x = 'Mass (kg)', y = 'Time to 100km/h \n(seconds)'))

Does this change based on what kind of drivetrain it has?

(p9.ggplot(df) +
p9.aes(x = 'mass', y = 'performance', group = 'enginetype', color = 'enginetype') +
p9.geom_point() +
p9.geom_smooth() +
p9.labs(x = 'Mass (kg)', y = 'Time to 100km/h \n(seconds)'))

Interesting! Electric cars take more of a performance hit the heavier they get whereas hybrids and petrol-based vehicles tend to flatten out in their performance.

Let’s check out the distribution of cars by body type in the dataset.

(p9.ggplot(df) +
p9.aes(x = 'type') +
p9.geom_bar())

A lot of SUVs!

Have fun! 🚗

License

The qatarcars python package is licsensed with a CC0 1.0 Universal (Creative Commons) license. Have fun with it!

Bibliography

• Musgrave, Paul. 2025. “Introducing the Qatar Cars Dataset.” July 8. https://musgrave.substack.com/p/introducing-the-qatar-cars-dataset
• Musgrave, Paul. 2025. “qatarcars” https://github.com/profmusgrave/qatarcars/tree/main

See also:

• {qatarcars} R package version from Andrew Heiss and Paul Musgrave