Creating the whatif package

In this submodule, you will learn about creating a Python package based on our whatif work so far. In particular, we will:

  • review the basics of Python packaging,

  • use a new cookiecutter better aligned with projects in which we intend to create deployable packages,

  • create a PyCharm project for whatif,

  • redesign our whatif code:

    • create an abstract Model base class containing our data_table, goal_seek, and simulate functions implemented as class methods,

    • learn how to create new model classes that inherit attributes from Model,

  • create and install our whatif package,

  • use our installed package,

  • learn how to deal with code changes during package development.

Downloads and other resources

This downloads file will be used throughout all of the activities in this submodule.

Here’s a really nice overview of Python packaging done by the same author who created the RTWDS Jupyter book.

Activities

You’ll be working through two separate notebooks. One is focused on creating deployable Python packages and the other is an introduction to creating documentation.

Python packaging

Inside the downloads file you’ll find a folder named whatif. Inside that folder is a notebooks folder and in there you’ll find what_if_4_project_packaging.ipynb. You’ll be going through this notebook for this submodule. Within the notebook there are links to a few short screencasts for specific sections of the notebook.

Creating documentation

Within that same folder you’ll find what_if_5_documentation.ipynb. This notebook covers the basics of:

  • code commenting and docstrings,

  • creating project readme files,

  • creating and generating documentation with Sphinx and reStructuredText.

Here’s a single screencast that demonstrates and illustrates important documentation concepts from this notebook. By the way, as I describe in this blog post, my course websites are all created with Sphinx and reStructuredText. In the screencast I show you how my course websites work. You’ll also see how Sphinx can autogenerate API documentation from the docstrings within our code.