R_tamu

R_tamu is an extension built on top of R and provides a large number of additionally installed packages not found in the base R version. In addition, R_tamu can also act as an R-project environment manager. Loading R_tamu will automatically load the matching R module so there is no need to load the R module separately. HPRC will update the installed packages regularly to make sure all packages are up-to-date with the latest bug fixes and improvements.

Loading the Module

To see a listing of all available R_tamu versions:

[NetID@cluster ~]$module spider R_tamu

To load a particular version of R_tamu (Example: 3.4.2 with the iomkl toolchain):

[NetID@cluster ~]$module load R_tamu/3.4.2-iomkl-2017A-Python-2.7.12-default-mt

Using R and Rscript

Using R and Rscript works exactly the same as the base R module. To start the R interpreter and to execute an R script, see example below.

[NetID@cluster ~]$ module load R_tamu/3.4.2-iomkl-2017A-Python-2.7.12-default-mt
[NetID@cluster ~]$ R
> # execute R commands
> q()
[NetID@cluster ~]$ Rscript myprog.R

For a more detailed introduction how to run R/Rscript on the login nodes and through the batch system, see the R Wiki page

Viewing Available Packages

As mentioned, R_tamu provides a large number of installed packages; currently, around 300 (including bio-conductor packages). Use the R library() function to see all installed packages. Alternatively, installed.packages() will list all installed packages.

Installing New Packages

In case a package is missing there are two ways to make it available:

Request HPRC to install package
Install locally in ${SCRATCH} directory

Request Additional packages

To request installation of additional R package, send email to help@hprc.tamu.edu and specify the package that is requested, a short description of its use, and any relevant documentation (if any).HPRC will install the package globally if the package might be useful to the general public.

Install Locally

R_tamu automatically sets up the environment for users to install packages locally by setting environmental variable R_LIBS_USER (set to /scratch/$user/R_LIBS_USER/, where is the R version/toolchain). R_tamu will create the $R_LIBS_USER directory if needed. Use the R function install.packages() function (or any other approach) to install the required package. The packages will automatically be installed locally in the $R_LIBS_USER directory. NOTE: HPRC can provide assistance to install the package to a user's personal directory as well.

Installing packages from Github

Whenever packages need to be loaded from repositories other than CRAN (e.g. github), directly loading the packages will result in errors. This is because packages are loaded by default in systemwide R_LIBS which is not writable by users. To handle this issue, set R_LIBS to R_LIBS_USER temporarily using following command:

export R_LIBS=$SCRATCH/R_LIBS_USER/

For example, if you load R_tamu/4.0.3-foss-2020a-recommended-mt, you can use the following command before loading R packages from Github:

export R_LIBS=$SCRATCH/R_LIBS_USER/4.0.3-foss-2020a-recommended-mt

R-project environments

R_tamu can also be used as a package manager for R projects. Using R_tamu, R and Rscript accept an additional flag --rtamuenvs= which sets up the environment for R to search for packages in a specific location. It will also create the directory structure to store the packages. This feature can be useful when a user is working on multiple projects each requiring the use of a certain number of R packages with specific versions. Using a project environment will also allow for reproducibility (e.g. user can run R simulations using the same environment again at a later time when needed). NOTE the globally installed packages will still be available, so there is no need to install every package again.

Example

Suppose, a user is working on two different R projects (let's call them projectA and projectB), When user is working on project A, environment for projectA will be used, when user switches to project B, environment for projectB will be used.

[NetID@cluster ~]$ module load R_tamu/3.4.2-iomkl-2017A-Python-2.7.12-default-mt
[NetID@cluster ~]$ R --rtamenvs=ProjectA
>
> #install packages for projectA
> install.packages(c('mypack1.tar.gz',mypack2.tar.gz'))
> library(mypack1)
> # use packages
> q()

[NetID@cluster ~]$ R --rtamenvs=ProjectB
>
> # packages installed under projectA will not be visible
> library(mypack1)
Error in library(mypack1) : there is no package called ‘mypack1’
> # install packages needed for projectB and use them
> q()

[NetID@cluster ~]$ R --rtamenvs=ProjectA
>
> # packages installed under projectA will be visible
> library(mypack1)
> # use library mypack1, mypack2, etc
> q()

[NetID@cluster ~]$ Rscript --rtamuenvs=ProjectA myAscript.R

Notes

The --rtamuenvs flag can accept a list of project environments (separated by :). For example: R --rtamuenvs=NewProj:ProjA. This can be useful when working on a new project that is based on an old project with some additional packages.