Working effectively in notebooks

Notebooks are excellent for exploring and terrible at hidden state. Restart and run all is the only honest test of a notebook.

What this lesson covers

• Structure a notebook so it reads top to bottom
• Avoid out-of-order execution and stale state
• Move stable code out into importable modules

This lesson sits in Reproducible Workflows, part of Python for Data Work. It assumes what came before it and leads directly into the next lesson in the module.