Every hub running the standard image has RStudio available alongside JupyterLab. You do not need a separate account or a different hub address.
Opening it¶
Start your server as usual, then from the JupyterLab launcher choose RStudio. It opens in a new browser tab, served from the same session.
It is the same home directory¶
This is the part worth understanding. JupyterLab and RStudio are two windows onto the same
files. A CSV you upload in JupyterLab is immediately visible to RStudio, and an .Rmd you write in
RStudio appears in the JupyterLab file browser.
The same limits apply to both — one 5 GiB home directory, one memory ceiling shared by everything in your session. See Memory and Storage Limits.
R packages¶
About 65 R packages are pre-installed, including tidyverse, tidymodels, ggplot2 and the wider
gg* family (ggpubr, ggrepel, ggridges, ggthemes, GGally), mosaic, infer, car,
janitor, lubridate, forecast, leaflet, sf, terra, stars, quarto, rmarkdown,
knitr, and teaching datasets such as palmerpenguins, nycflights13, openintro and
Lock5Data.
The authoritative list, with versions, is install.R in the image repository.
IRkernel is installed too, so you can work in R from a Jupyter notebook instead if you prefer.
Installing a package yourself¶
install.packages("<package>")It installs into your home directory and persists across sessions. As with Python packages, it is yours alone — your students will not have it. Anything a class depends on should go in the image instead; see Requesting Packages.
Shiny¶
Shiny apps run on the hub through the launcher. Quarto and R Markdown are installed, so documents knit to HTML and PDF without extra setup.