R data analysis template

My template for automatically setting up a new data analysis project using R. This template uses cookiecutter, a Python templating tool, to setup a directory structure, set up an renv reproducible environment, and initialize git.

Acknowledgements

Modified from lazappi/cookiecutter-r-analysis, bdcalf/cookiecutter-r-data-analysis, and khughitt/cookiecutter-rmd-data-science

Requirements

Install cookiecutter using pip:

pip install cookiecutter

OR

Install cookiecutter using homebrew:

brew install cookiecutter

Usage

Generate a new project using:

cookiecutter gh:pgcudahy/cookiecutter-r-analysis-template

You will then be asked some questions to set up your project. Leaving answers blank will select the default (shown in [brackets]).

Structure

The resulting data analysis project will have the following structure:

project_title
├── README.md
├── artifacts - All results (eg figures, tables)
├── data - Intermediate data files (eg cleaned dataframes). These may be modified or deleted.
├── dataraw - All original input data. These may not be modified or deleted. 
├── notebooks - Jupyter notebooks documenting all analysis.
└── r - Reusable R code (eg convenience functions).

Environment

After setting up the directory environment:

• renv is installed and used to manage a reproducible R environment
• Git is initialized and an initial commit is made

Workflow

• Raw input data is put in the dataraw directory and should not be modified
• Data is analyzed by Jupyter notebooks in the notebooks directory
• Custom R source code used in the data analysis is put in the r folder
• Transformed datasets are saved in the data directory. Must be able to be reproduced from the raw data and notebooks.
• Output for publication or dissemination is put in the artifacts folder

License

This project is licensed under the terms of the MIT License