Data Science project template

Note: This template uses poetry and python 3.9+

This repository is a cookiecutter template for a data science project.

The necessary tools

poetry dependency management. cookiecutter for new project generation based on the template directory.

Features

Once the project is generated based on the template, you will get out-of-the-box support for:

• Dependency management using poetry.
• Linting and formatting done via ruff.
• Helper development tasks (lint, format, test, etc) run via Makefile (run make help to see the available tasks).
• Pre-commit hooks using pre-commit package and plugins to automate code quality aspects.

Usage guide

You first need to have cookiecutter available in your machine. There are several alternatives to this. Here are 2:

Installation via homebrew (for MacOs users): brew install cookiecutter

brew install cookiecutter

Or, create a python environment with cookiecutter in it. Assuming you have a workspace directory in you USER HOME, you can then add the cookiecutter environment into it like so:

python -m venv ~/workspace/cookiecutter
source ~/workspace/cookiecutter/bin/activate
python -m pip install -U pip cookiecutter 

Reminder: if you opted for the python environment approach, you will only be able to use cookiecutter as a command if you activate the environment.

Now you are ready to create a new project from the template, using the cookiecutter utility:

cookiecutter https://github.com:jmetzz/data-science-template

This will use the latest version in the main file. To generate a project from a different version, use the -c, --checkout <checkout> option. See details at Cookiecutter Command Line Options.

Enter in your new project directory, install the dependencies and create the first commit:

cd <your-project-directory>
make deps license
make lint test
git add .
git commit -m "Initial commit"

Enjoy your new project.