Skip to main content

Rapid On-ramps to Reproducible Research for R Users

Awareness of the importance of improving the reproducibility of research is spreading, however slowly and patchily. Journals in many fields now require data and code to be available to accompany published articles. However it is not always obvious how we should do this, and there are few established conventions on how to ensure our work is as reproducible as possible. In this seminar I will present two recently developed R packages (currently available on GitHub) that provide instructions, templates, and functions for quickly getting started with compendium suitable for doing reproducible research with R. They allow us to quickly set up version control, continuous integration, isolation of the computational environment and other best practices, with sensible defaults and minimal fuss. This allows us to concentrate on our writing and analysis, without falling into time-consuming rabbit holes of configuration problems. The first package, rrtools, is optimised for writing journal articles and reports, and the second package, huskydown, is specifically for writing PhD dissertations and books.