toolero: An R package for research
toolero: An R package for research
Every research coding project begins with a blank slate and a cluster of low-stakes decisions — where does the data go, what do I call the scripts, do I track dependencies? The trouble is that these decisions rarely stay low-stakes. The cost surfaces later, when a collaborator tries to reproduce the analysis, when the project needs to move to a computing cluster, or when you return to it six months later and realize the outputs no longer match the code.
At the June 2026 UW-RSE monthly meeting, I presented toolero, a small, opinionated R package designed to reduce that friction. The motivating idea is simple: best practices are easier to adopt when the tooling makes them the path of least resistance. toolero does not impose a rigid framework — it provides sensible defaults and gets out of the way.
Here are a few functions I covered:
init_project() scaffolds a standard folder structure and optionally initializes renv and git from the start.
create_qmd() produces a reproducible Quarto document with a post-render hook that keeps the .R script derived from — rather than diverging from — the source document.
read_clean_csv() and write_clean_csv() make the raw/clean data distinction explicit and automatic.
write_by_group() splits a data frame into per-group files, producing a manifest that feeds directly into high-throughput computing job submission.
And detect_execution_context() handles the perennial headache of file path resolution across interactive, Quarto, and Rscript environments.
toolero is also the first step in a three-package family — containr and submitr extend the same philosophy toward containerization and CHTC job submission — but it stands on its own for anyone whose work lives entirely on their laptop.
The package is on CRAN. Documentation and examples are at erwinlares.github.io/toolero.
Questions?
If you any lingering questions about this resource, please feel free to post to the Nexus Q&A on GitHub. We will improve materials on this website as additional questions come in.
See also
- toolero on GitHub: Github toolero page.