Reference · no installation

Run R here, now, before you install anything.

Every code block in this library can be pasted into the console below and run for real. It is a genuine R interpreter, not a simulation — the same language, the same error messages. Use it to try things on a train, to check a function's behaviour, or to work through volume 1 while your university IT department thinks about your install request.

Console Cmd/Ctrl + Enter to run · first run downloads about 30 MB
{{ timing }}
{{ output }}
Preloaded

Snippets, one per volume

Click one to load it into the console. Each uses only base R and datasets that ship with it, so nothing needs installing.

Honest limits

What this console cannot do

WebR is remarkable and it is not a replacement for R on your machine. Four real constraints, stated plainly so you are not surprised halfway through a volume.

slow first load
The R runtime is roughly 30 MB and downloads the first time you press Run. Ten to forty seconds on a decent connection, and it is cached afterwards.
not every package
Many CRAN packages are available as WebAssembly builds, but anything needing compiled system libraries or Stan — brms, rstan, V8 — will not install. Volume 7 needs a real R.
no project on disk
There is a virtual filesystem, but no RStudio panes, no .Rproj, no Git, no Quarto render. Volume 9 is unavoidably local work.
memory and speed
Fine for a few thousand rows and a mixed model. A 5,000-iteration simulation or a large dataset will be noticeably slower than native R and may exhaust the tab's memory.
Day 1, really

Install R properly, in fifteen minutes

1. R itself cran.r-project.org → your platform → latest release 2. RStudio Desktop posit.co/download/rstudio-desktop (free tier) 3. Open RStudio, then in the console: install.packages(c("tidyverse", "here", "haven", "readxl")) 4. File → New Project → New Directory → give it a name. That folder is now your working directory. Never use setwd() again. 5. Windows only, later, for packages that compile: install Rtools macOS only: xcode-select --install

Then start volume 1

Seven days from the four panes to a complete pipeline: raw file in, cleaned data out, one plot, one table, saved. Come back here whenever you want to test something without opening RStudio.

Start day 1 →