PyDoit Examples Walkthrough

These examples are available in the inclass_examples repository: finm-32900/inclass_examples . Clone the repository and navigate to the pydoit/ directory to follow along.

Quick Start

pip install doit
cd pydoit/01_simplest
doit

The Examples

01_simplest

The most basic PyDoit example: three tasks that run separate Python scripts in sequence (create_data.py → process_data.py → create_report.py). Demonstrates actions, file_dep, and targets. See the README in this directory for details.

02_example_project

A realistic data science workflow with proper project structure. Pulls GDP and unemployment data from FRED, processes it, and creates charts. Demonstrates how to organize a project with a src/ folder and settings.py. See the README in this directory for details.

03_handling_notebooks

Shows how to handle regular .ipynb Jupyter notebooks in a doit workflow. Uses a two-stage approach: convert notebooks to .py files for change detection, then execute only when the code (not metadata) changes. This is a clone of blank_project_simple. See the README in this directory for details.

04_jupytext_notebooks

Demonstrates Jupytext percent-format notebooks (stored as .py files) which are version-control friendly. Includes Sphinx documentation generation. This is a copy of fred_charts. See the README in this directory for details.

Learning Progression

Example	Key Concept
01	Basic task definitions and dependencies
02	Project structure and external data
03	Two-stage notebook workflow
04	Jupytext and documentation