Organizing Projects
Laying out projects with a clean, well-organized structure, with careful and complete recording of dependencies and workflows, makes them a lot easier to work on, collaborate on, extend, and maintain.
This set of notes is primarily focused on analytics or modeling projects, where our primary outcome is results and/or trained models. Reusable libraries, command-line tools, production services, etc. have their own considerations, but many of the points here will still apply.
Tip
My project starter has key pieces ready-to-go for Python-based projects.