Awesome

vega_datasets

A Python package for offline access to vega datasets.

This package has several goals:

• Provide straightforward access in Python to the datasets made available at vega-datasets.
• return the results in the form of a Pandas dataframe.
• wherever dataset size and/or license constraints make it possible, bundle the dataset with the package so that datasets can be loaded in the absence of a web connection.

Currently the package bundles a half-dozen datasets, and falls back to using HTTP requests for the others.

Installation

vega_datasets is compatible with Python 3.5 or newer. Install with:

$ pip install vega_datasets

Usage

The main object in this library is data:

>>> from vega_datasets import data

It contains attributes that access all available datasets, locally if available. For example, here is the well-known iris dataset:

>>> df = data.iris()
>>> df.head()
   petalLength  petalWidth  sepalLength  sepalWidth species
0          1.4         0.2          5.1         3.5  setosa
1          1.4         0.2          4.9         3.0  setosa
2          1.3         0.2          4.7         3.2  setosa
3          1.5         0.2          4.6         3.1  setosa
4          1.4         0.2          5.0         3.6  setosa

If you're curious about the source data, you can access the URL for any of the available datasets:

>>> data.iris.url
'https://cdn.jsdelivr.net/npm/vega-datasets@v1.29.0/data/iris.json'

For datasets bundled with the package, you can also find their location on disk:

>>> data.iris.filepath
'/lib/python3.6/site-packages/vega_datasets/data/iris.json'

Available Datasets

To list all the available datsets, use list_datasets:

>>> data.list_datasets()
['7zip', 'airports', 'anscombe', 'barley', 'birdstrikes', 'budget', 'budgets', 'burtin', 'cars', 'climate', 'co2-concentration', 'countries', 'crimea', 'disasters', 'driving', 'earthquakes', 'ffox', 'flare', 'flare-dependencies', 'flights-10k', 'flights-200k', 'flights-20k', 'flights-2k', 'flights-3m', 'flights-5k', 'flights-airport', 'gapminder', 'gapminder-health-income', 'gimp', 'github', 'graticule', 'income', 'iris', 'jobs', 'londonBoroughs', 'londonCentroids', 'londonTubeLines', 'lookup_groups', 'lookup_people', 'miserables', 'monarchs', 'movies', 'normal-2d', 'obesity', 'points', 'population', 'population_engineers_hurricanes', 'seattle-temps', 'seattle-weather', 'sf-temps', 'sp500', 'stocks', 'udistrict', 'unemployment', 'unemployment-across-industries', 'us-10m', 'us-employment', 'us-state-capitals', 'weather', 'weball26', 'wheat', 'world-110m', 'zipcodes']

To list local datasets (i.e. those that are bundled with the package and can be used without a web connection), use the local_data object instead:

>>> from vega_datasets import local_data
>>> local_data.list_datasets()

['airports', 'anscombe', 'barley', 'burtin', 'cars', 'crimea', 'driving', 'iowa-electricity', 'iris', 'seattle-temps', 'seattle-weather', 'sf-temps', 'stocks', 'us-employment', "wheat"]

We plan to add more local datasets in the future, subject to size and licensing constraints. See the local datasets issue if you would like to help with this.

Dataset Information

If you want more information about any dataset, you can use the description property:

>>> data.iris.description
'This classic dataset contains lengths and widths of petals and sepals for 150 iris flowers, drawn from three species. It was introduced by R.A. Fisher in 1936 [1]_.'

This information is also part of the data.iris doc string. Descriptions are not yet included for all the datasets in the package; we hope to add more information on this in the future.