Installation

Linux and Mac OSX

With Conda

If you haven’t already, install conda. You’ll also need to have the conda-forge channel enabled: if you don’t already, you can do so with conda config --add channels conda-forge.

Then run conda install geoplot and you’re done.

Without Conda

Things are much trickier if you don’t have access to conda, for whatever reason. You will need matplotlib, seaborn, geopandas, the proj4 (docs) and GEOS (docs) external libraries, and cartopy (which relies on proj4 and GEOS). Here’s an approximate code path:

# Use of a virtualenv is strongly encouraged.
mkdir my_project
cd my_project
virtualenv venv
# Easy installs that shouldn't cause any problems.
pip install matplotlib seaborn
# ... install proj4 and GEOS using brew, apt-get, etc. ...
# Install shapely with the no-binary option.
# See https://github.com/Toblerity/Shapely/issues/435
pip install --no-binary shapely
# Install the rest.
pip install geopandas
pip install cartopy
pip install geoplot

It’s difficult to succeed installing this way, due to conflicts deep within the stack with reference to the C dependencies. This, this, and this are reference issues.

Windows

Unforunately geoplot is not available on Windows yet due to an unresolved dependency issue. Sorry.