knitr::opts_chunk$set( collapse = TRUE, comment = "#>" )
Most programming languages agree on where files belong. Python has its package layout. Node has src/, public/, package.json. Others give you a directory tree on day one.
R has strong conventions for packages (R/, man/, inst/), but analytical projects have not always had the same structure. Marwick, Boettiger, and Mullen (2018) proposed the research compendium as a standard way to organize the digital materials of a project so that others can inspect, reproduce, and extend the work. Their layout uses familiar R package conventions and adds an analysis/ directory for the project's computational work.
{froggeR} builds on this idea, adapting it for Quarto-based projects:
R/ for scriptsanalysis/ for Quarto (and .Rmd) documentswww/ for stylesheets and assetsdata/ for data fileslogos/ for brand imagesThe same convention maps to R package structure, to Shiny frameworks like rhino, and to web frameworks in other languages.
init()One command builds the full scaffold:
froggeR::init(path = "my_project")
my_project/ ├── R/ │ ├── _data_dictionary.R # Variable labels and metadata │ ├── _libraries.R # Centralized package loading │ └── _load.R # Sources everything. Your entry point. ├── analysis/ │ ├── index.qmd # Main Quarto document │ └── references.bib # Bibliography ├── www/ │ ├── custom.scss # Custom styling │ └── tables.js # Table enhancements ├── logos/ # Brand logos ├── data/ # Data files (gitignored) ├── _brand.yml # Quarto brand configuration ├── _quarto.yml # Quarto project configuration ├── _variables.yml # Author metadata ├── .gitignore # Opinionated git protection ├── .pre-commit-config.yaml # Pre-commit hook configuration └── README.md
init() creates the target directory if it does not exist, downloads the latest template from frogger-templates, and copies only files that are not already present. Existing files are never overwritten. If you have previously saved global config with save_variables() or save_brand(), your metadata and branding are applied automatically.
R/ contains R scripts. _load.R is the entry point: it sources _libraries.R (package loading) and _data_dictionary.R (variable labels and metadata). Add your own scripts here.
analysis/ contains Quarto documents. index.qmd is the default landing page. references.bib holds your bibliography. Add new analysis files with write_quarto().
www/ contains web assets. custom.scss controls document styling. tables.js provides table enhancements. Keep stylesheets, images, and scripts here.
data/ is for data files. It is gitignored by default, so you can keep raw and processed data here without worrying about accidentally committing it.
logos/ holds brand images referenced by _brand.yml.
The files in the project root are configuration:
_quarto.yml defines project-level rendering options. See Quarto Project Basics._brand.yml defines your visual identity: colors, logos, typography. See Quarto Brand._variables.yml stores author metadata: name, email, etc..gitignore keeps R artifacts, rendered output, and data files out of version control..pre-commit-config.yaml runs code quality checks before every commit. See below.The template includes a .pre-commit-config.yaml that integrates with pre-commit to run automated checks before every git commit. This catches problems early: unstyled code, lint violations, debug statements left in, large files accidentally staged.
From {precommit}:
browser() callsdebug() and debugonce() callsFrom pre-commit-hooks:
Local:
.Rhistory, .RData, .Rds, and .rds filesInstall the R packages and initialize:
install.packages(c("precommit", "styler", "lintr")) precommit::use_precommit()
{precommit} handles installing the pre-commit framework itself. See pre-commit.com for manual installation options.
The hooks are a starting point. Edit .pre-commit-config.yaml to add, remove, or reconfigure hooks for your workflow. If you do not want pre-commit at all, delete the file. No other part of {froggeR} depends on it.
{froggeR} stores configuration in two places:
_variables.yml and _brand.yml in your project directory. These are committed to version control and used by the current project.rappdirs package.~/.config/froggeR/C:\Users\<username>\AppData\Local\froggeR\Global config is set once and applied to every future project created with init(). Edit the project copy directly when you need something specific.
If you have configured global settings before running init(), your new project inherits them automatically:
_variables.yml: Global config overwrites the template version_brand.yml: Global config overwrites the template versionlogos/: Global logos supplement the template (adds missing logos without removing existing ones)Use write_variables() and write_brand() to create and edit your configuration files, then save_variables() and save_brand() to persist them globally. Every future init() call starts with your configuration in place.
Marwick, B., Boettiger, C., & Mullen, L. (2018). Packaging data analytical work reproducibly using R (and friends). The American Statistician, 72(1), 80-88. doi:10.1080/00031305.2017.1375986
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.