EarthSim supports Python 3.6 on Linux, Windows, or Mac.
EarthSim itself is a pure Python package that itself would be easy to install, but it depends on several packages (gdal, geos, and others) that are linked against binary-code geospatial libraries. Unfortunately, incompatibly compiled versions of those libraries are available from different sources, which can make it difficult to assemble an appropriate environment for safely running EarthSim. For this reason, EarthSim is currently provided via a special set of installation steps based on conda and git that ensure only a compatible set of packages is used. Specifically:
Install the EarthSim package via conda.
conda install -c conda-forge earthsim
- Download the sample files.
python -m “earthsim examples”
If you are actively collaborating with the EarthSim developers and want to try out the latest pyviz work as it first appears (which is not necessarily functional or stable), you can install EarthSim from GitHub.
Clone the EarthSim git repository:
git clone git://github.com/pyviz/EarthSim.git
Set up an environment with all of the dependencies needed to run the examples:
cd EarthSim doit env_create -c pyviz/label/earthsim -c pyviz/label/dev -c defaults -c erdc -c conda-forge --name=earthsim --python=3.6
Activate the earthsim environment:
earthsimmodule into this environment:
pip install -e .
Download the sample files:
If you really want to be on the bleeding edge, you can instead get the absolute latest changes by cloning the earthsim, holoviews, geoviews, datashader, param, and panel Github repositories and running pip install -e . inside each one, pulling new changes from each of these libraries as needed. That’s what the main developers do, but it isn’t recommended for other users unless you are very skilled at debugging the broken environments that are likely to appear as packages change unpredictably over time.
Once you’ve installed EarthSim as above and are in the EarthSim directory, you can run the examples shown on the website using Jupyter:
cd earthsim-examples jupyter notebook
You should now be able to select one of the
topics notebooks and run it in Jupyter.
You can look through the User Guide and the Topics to see examples, but most of the documentation is at the sites for the packages like HoloViews that are used in the examples. A good way to get comfortable with those tools is to work through the tutorials at PyViz.org.