Installers for core Scientific Python packages

Installers for the core Scientific Python packages for macOS, Windows, and Linux.

Who should use these installers?

These installers are double click (macOS, Windows) or simple command line installers (Linux) for the standard packages for general scientific Python and data science. The installer includes packages such as:

• Jupyter (for interactive notebooks);
• Numpy (for arrays);
• Scipy (extended scientific computing);
• Matplotlib (plotting);
• Pandas (data science);
• Statsmodels (statistics);
• Scikit-Image (image processing);
• Scikit-Learn (machine learning and statistics);

as well as many others. See the ./recipes/scientific-python/construct.yaml file for details.

We imagine these installers will be particularly useful for beginner students taking courses in science, statistics or data science, to get them up and running with a single download and install step.

Where can I get the installers?

We will make a release soon. When we have done so, please visit the release page to download an installer for your OS.

Development

Locally, installers can be built using tools/build_local.sh. Steps:

1. Set up and activate a conda env with a forked version of constructor:

   $ conda env create -f environment.yml
   $ conda activate constructor-env

2. Run ./tools/build_local.sh.
3. Install the environment for your platform.
4. Test it using the tests/.