Anaconda is the World’s Most Popular Python Data Science Distribution.
This demo shows how to install the full Anaconda data science stack on Platform.sh.
Conda is an open source package and environment management system. It was created for Python programs, but it can package and distribute software for any language. Conda is included in all versions of Anaconda, as well as Miniconda, which is essentially Anaconda without the packages included.
The build hook in your
.platform.app.yaml file lets you specify scripts to run when your code builds.
This script uses curl to get Miniconda. It then runs the Conda installer, adds Conda to the path,
and installs the full Anaconda environment.
In the web section of your
.platform.app.yaml you then specify a start command which runs this script.
The example shown activates the Conda environment and starts the Jupyter notebook, a popular interactive
tool for creating rich documents and running code.
When Platform.sh builds your code and manages dependencies, all of the required packages are written on a read-only disk image that is available to your application container.
R, PyPy, and Stackless Python are examples of other runtimes or tools that can be installed on Platform.sh by first installing Miniconda and then using it to install further packages.