Designing R functions — where validation lives, when one function should be several, decomposing a calculation, and keeping nesting shallow. Use when writing or reviewing the internals of R functions, when a function is hard to follow or hard to test, when…
Conventions for structuring an R analysis project — packages and conflicts instead of pkg::fun(), fs over base file functions, air formatting, .Renviron, where generated files go, what the README must answer, and how to size files in R/. Use when setting up…
Write minimal reproducible examples for R bug reports, GitHub issues and questions. Use when drafting an issue about broken R code, writing up a bug you have found, turning an investigation into something someone else can run, or when asked for "a reprex", "a…
Write fewer, better code comments — the three-stage test for whether a comment earns its place, and an audit pass for comments already written. Use when writing or editing code in any language, when reviewing a diff before committing, or when asked to "check…
Structure an R analysis as a targets pipeline. Use when setting up a new targets project, converting scripts to a pipeline, writing new targets, reviewing or tidying an existing _targets.R, debugging an errored target or failed tar_make (workspaces,…
Migrate a blogdown/hugodown (Hugo) R blog to Quarto. Use when converting a Hugo-based R blog or website to a Quarto website - covers auditing the site, replicating the theme, migrating posts from pre-rendered output without re-executing R code, preserving old…