Manage project dependencies with conda environments

TL;DR: Create a conda environment for each project, capture exact versions when possible, automate activation and updating with a bash function.

I often work on several different projects involving software: Python libraries, papers, presentations, posters, this website, etc. Each project has different dependencies and there is a non-zero chance that these dependencies might be in conflict with each other. For example, I need Python 2.7 to work on a tesseroid modeling paper with a student, while my current work on

(continued...)

from Planet SciPy
read more