Cookiecutter Data Project Repositories Structure Template

cookiecutter is a 'logical, reasonably standardized, but flexible project structure for doing and sharing data science work.' It helps with projects reproducibility, code quality and collaboration within data science teams. This fork of the main structure is a barebone structure that can be further customized to answer not only data science needs but also analytics engineering or even ad-hoc analysis projects as well.

This custom template is released under the MIT license so feel free to use it for all yours projects.

A few choices have been made regarding its content:

• The pipenv library for Python dependencies management is used by default.
• To check the quality of our code follows PEP8 standards, we use flake8.
• We use a Makefile to run recurring commands (linter, requirements). More can be added once installed.

Prerequisites

---

• cookiecutter

$ pip install cookiecutter

Install the structure in a new empty project folder

---

$ cookiecutter https://github.com/GuillaumeLegoy/data-projects-template.git

Once inside your project, run the following to install dev dependencies:

---

$ make requirements

To make sure your code follows pep8 standards run the following inside your project:

---

$ make lint

If you are developing a Python library, sync your Pipfile Python libraries with setup.py:

---

$ make setup

Directory structure

---

├── /src                <- Source code for use in this project.
│   └── __init__.py     <- Makes src a Python module
│
├── /test
│
├── Makefile           <- Makefile with commands like `make data` or `make train`
├── README.md          <- The top-level README for developers using this project.
│
├── .flake8            <- Flake8 linter parameters
├── .gitignore         <- Gitignore for Python projects
├── Makefile           <- Makefile including 3 commands: `requirements`, `lint`.
├── Pipfile            
├── Pipfile.lock       <- The requirements files for reproducing the analysis environment
├── README.md
└── setup.py           <- Setup file to package your library or application.