đź“Ś R Workflow
I always love stumbling across a bookmark I completely forgot about, the kind that makes you feel oddly grateful to past-you for spotting something useful before you even knew you’d need it.
đź“Ś R Workflow by Frank E Harrell Jr
A concise, mindset-first guide to structuring R-based statistical analysis projects, with a focus on reproducibility, sustainability, and clear communication.
Reflections
This isn’t a how-to or tutorial, it’s a focused, opinionated take on how to think about analytical projects in R, especially when working in collaborative or research-driven contexts. What I appreciated most was how it made me pause and reevaluate some of the defaults I’ve accumulated over time.
A few things that stood out:
- Script-first workflows: no hidden logic in interactive consoles or notebooks. If it matters, it should be in the script.
- Project structure is part of reproducibility, not an afterthought. Folder organization, file naming, and code modularity all impact how easily others (and future-you) can navigate the work.
- Communication isn’t extra: comments, documentation, audit trails, they are the work. Especially in long-term or handoff-heavy settings.
I’m not following this workflow to the letter (yet), but it resonated with some practices I’ve already started to build around logging, commenting, and structuring code for clarity. It gave me a few nudges to go further and a bit of validation that I’m heading in the right direction, slowly, but intentionally.
Definitely one I’ll return to as I keep shaping my own standards for what “well-structured” means in practice.
This post builds on a recent LinkedIn #BookmarkDive reflection, feel free to join the conversation there.