R and RStudio - Computing in Statistics at Berkeley

We provide R and RStudio, including a variety of popular libraries.

RStudio¶

The primary access to RStudio is through a web browser using our JupyterHub. You can use your SCF username and password to login, or you can login to a shared project account.

By default, your RStudio session will be spawned onto the first available standalone Linux server, however you may optionally start your session on a cluster node in case you need access to more processing power. You can also pass Slurm job submission options to your server and specify prologue commands that will run prior to your RStudio startup.

Once your Jupyter session is active, click RStudio from the Jupyter Lab launcher. If you don’t see the launcher, click File > New Launcher. All R packages installed on the system, as well as packages installed by users within their home directories should be available.

In contrast with running RStudio directly on SCF machines and forwarding the display over X11, this should be much more responsive.

R package installation¶

First, note that in general, SCF staff will install an R package on the system on request.

User installation¶

However, you can also install packages locally within your home directory. So if you need a package quickly or on a one-time basis, or if the package is particularly specialized, you might install it locally.

By default, user-installed packages are installed in the R subdirectory in your home directory, in ~/R/x86_64-pc-linux-gnu-library-ubuntu-24.04/4.5. However, if that directory does not exist, you may get an error message like this:

mkdir: cannot create directory '/server/linux/lib/R/site-library/00LOCK': Permission denied
ERROR: failed to lock directory  '/server/linux/lib/R/site-library' for modifying

The simplest solution is to create the directory:

mkdir -p ~/R/x86_64-pc-linux-gnu-library-ubuntu-24.04/4.5

You can use .libPaths() to check that the user-level directory is the default location (it should be the first result printed, as seen below) where R will install packages.

> .libPaths()
[1] "/accounts/grad/username/R/x86_64-pc-linux-gnu-library-ubuntu-24.04/4.5"
[2] "/system/linux/lib/R-24.04/4.5.0/x86_64/site-library"
[3] "/usr/lib/R/site-library"
[4] "/usr/lib/R/library"

Library path management¶

If you need more control over the library path, R provides a number of methods for controlling the library path to accommodate just about any user’s need.

Temporarily changing the library path¶

You can modify R’s notion of your library path on a one-time basis by specifying the lib= argument to install.packages. Suppose there is a directory called MyRlibs in your home directory. The command:

install.packages('caTools',lib='~/MyRlibs')

will install the specified package in your local directory. To access it, the lib.loc= argument of library must be used:

library('caTools',lib='~/MyRlibs')

One problem with this scheme is that if a local library invokes the library() function, it won’t know to also search the local library

Changing the library path for a session¶

The .libPaths() function accepts a character vector naming the libraries which are to be used as a search path. Note that it does not automatically retain directories which are already on the search path. Since the .libPaths() function returns the current search path when called with no arguments, a call like

.libPaths(c('~/MyRlibs',.libPaths()))

will put your local directory at the beginning of the search path. This means that install.packages() will automatically put packages there, and the library() function will find libraries in your local directory without additional arguments.

Permanently changing the library path¶

The environmental variable R_LIBS is set by the script that invokes R, and can be overridden (in a shell startup file, for example) to customize your library path. This variable should be set to a colon-separated string of directories to search. Since it’s always set inside of an R session, the easiest way to get a starting point for it is to use Sys.getenv():

> Sys.getenv('R_LIBS')
[1] "/usr/local/linux/lib/R/site-library:/usr/local/lib/R/site-library:/usr/lib/R/site-library:/usr/lib/R/library"

You could then make a copy of this path, modify it, and set the R_LIBS environmental variable to that value in the shell or a startup script.