vega-datasets

What it is and what it does

vega_datasets is a data-loading utility that gives you quick access to a collection of sample datasets commonly used in data visualization and analysis. It wraps the Vega project's public dataset collection and returns them as Pandas DataFrames. The package bundles a half-dozen datasets (iris, cars, seattle-weather, and others) for offline use, and falls back to fetching additional datasets via HTTP when needed.

You import a single `data` object, call methods like `data.iris()` to get a DataFrame, and optionally inspect the source URL or local file path. The package is designed for exploratory analysis, visualization prototyping, and testing—situations where you need real but small, well-known datasets without setting up a data pipeline.

Use it for:

• Prototyping data visualizations with Altair or Matplotlib using well-known reference datasets
• Teaching or demonstrating data analysis workflows with standard, reproducible sample data
• Unit testing visualization or data-processing code with bundled datasets that require no network
• Quick exploratory analysis when you need a familiar dataset like iris or flights without downloading
• Building examples or documentation that reference canonical datasets from the Vega ecosystem

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides offline and online access to Vega datasets as Pandas DataFrames, with a half-dozen datasets bundled locally and fallback to HTTP for others.

Yes, if you need sample datasets for visualization or analysis prototyping and can tolerate that the package is no longer maintained. The low install friction, permissive license, and bundled offline data make it convenient for development and teaching. However, do not rely on it for production pipelines or expect updates to bundled datasets or compatibility with future pandas versions.

Install

vega-datasets on PyPI

pip install vega-datasets

uv add vega-datasets

poetry add vega-datasets

Installing vega-datasets

Before you install

Low install friction with a single pandas dependency and a pure-Python wheel. However, the package is archived and abandoned as of 2025-11-12, with no releases since 2020-11-26; it will not receive bug fixes or security updates.

License in practice

MIT license is permissive and poses no restrictions on use, modification, or redistribution in commercial or private projects.

Quickstart

pip install vega_datasets

from vega_datasets import data
df = data.iris()
print(df.head())

Verify before relying

• Whether bundled datasets remain current with upstream Vega datasets repository
• Compatibility with pandas versions released after 2020-11-26
• Whether HTTP fallback URLs remain valid and accessible

Package facts

Evidence: vega_datasets-0.9.0-py3-none-any.whl

Tags

More Information Analysis packages