Creating a new data science project

When creating a new data science project, there are a lot of little details to remember in order to set up a sane project structure that supports great practices. PyDS lets you set all of that up with one interactive command:

pyds project initialize

Behind the scenes, we create a new Python project with an opinionated directory structure. There's a place to put your notebooks/. There's a custom source directory named after your project_name/, where you can easily refactor out notebook code. Your project_name/ source also comes with basic command line interface capabilities that you can expand on in case you wish to make a CLI tool for others to use. There are a smattering of code quality tools automatically installed as part of your pre-commit hooks. You default to getting a conda environment automagically, and you have a custom Python package (project_name/) installed into the environment too, ensuring portability of your data analysis code.