venpopov / jeeves

Tools for reproducible research workflows in R

jeeves

Jeeves assists in creating, managing and sharing reproducible research projects in R. It provides tools for creating standard project structures, managing dependencies, and generating reports with ease.

Installation

You can install the development version of jeeves like so:

if (!requireNamespace("remotes", quietly = TRUE)) {
  install.packages("remotes")
} 
remotes::install_github("venpopov/jeeves")

Example

This is a basic example which shows you how to solve a common problem:

library(jeeves)
## basic example code

Comparison with other packages

jeeves builds heavily on usethis to create research projects. Whereas usethis is modular and general-purpose, jeeves provides one-line functions for a complete project setup. After running one of the main jeeves functions, you will be immediately ready to start working on the code rather than setting up the project. By default, jeeves sets up a project with:

• standard project directory structure
• package skeleton with filled out DESCRIPTION, NAMESPACE, R/, tests/, LICENSE, README.Rmd, NEWS.md, .Rbuildignore, .gitignore files
• version control via git and GitHub. A new repository is created on your GitHub account and the local project is linked to it
• package dependencies management via renv and installation of default development packages
• a pkgdown or a Quarto website for the package or the project. The website is automatically deployed to GitHub pages with a GitHub action
• continuous integration via GitHub actions for testing and deployment
• a skeleton targets pipeline for reproducible data analysis

Jeeves sets opinionated defaults for each of these steps, but you can customize them by passing arguments to the main functions.

usethis is a fantastic package due to its modularity and flexibility. At the same time, if you are like me, you might find yourself running the same set of 20-30+ usethis, git and other functions every time you start a new project. jeeves is an attempt to automate this process and provide a one-stop-shop for setting up a new reproducible project.