R and RStudio

R is installed as a module with a number of R packages installed. It can be useful to always use the same version of R to avoid the possibility of changes that affect your calculations between different versions.

Installing libraries

Because R allows users to have their own library of installed packages you will be able to install most packages for yourself. You can see the packages that you have installed along with those that come with the version of R that you have loaded by using the following command.

    >library()

Most of the time you will be able to install R packages from CRAN using the following command which will automatically look after all of the package dependencies for you.

    > install.packages('mypackage')
        ...
    * DONE (mypackage)

If one or more of the packages that you have installed for yourself rely on accessing software via the module command then you will need to load the module before starting R. If you forgot or are using RStudio where it is not possible to load the module ahead of time you can load the module from within R by making use of the file /usr/share/Modules/init/r.R in the following way to both load and unload modules.

    > load("/usr/share/Modules/init/r.R")
    > module("add","gdal/3.5.3")
    > module("del","gdal/3.5.3")

If you have problems installing a package and would like help installing it please email restech.support@unsw.edu.au.

RStudio and JupyterLab in Katana OnDemand

If you would like to use RStudio on Katana, we recommend you use Katana OnDemand which allows you to specify the version of R that you wish to use including any packages that you have previously installed. Depending on the packages that you have installed you may need to use the module command decribed above.

It is also possible to use the R Kernel within JuypterLab including having access to the packages that you have installed but we recommend using RStudio instead as it has been written specificlly for R. If you wish to use the R Kernel within JuypterLab then you can use the following commands within R before you start JupyterLab.

    > install.packages("devtools")
    > devtools::install_github("IRkernel/IRkernel")
    > IRkernel::installspec()

R and RStudio in Conda

Whilst it is possible to install R and RStudio within Conda you will probably not be able to install the version of R and supporting software that you wish to use. If the version that you wish to use is not currently used on Katana then you should email the Restech team at restech.support@unsw.edu.au and they will install it for you. The instructions on installing R within Conda, if required, are given below.

Instructions for using R within Conda

!!! warning Using Conda to install complex software with large number of dependencies such as R or Rstudio may not provide provide you a recent version. If you wish to install R as part of a Conda environment then you can use the following commands to install R in a new repository called R_software after . This may be the easiest approach to take if you need to use a version of R that is not installed on Katana or a specialised collection of packages like [BioConductor](https://www.bioconductor.org/). If you are having trouble figuring out which approach to take please email the Restech team at [restech.support@unsw.edu.au](mailto:restech.support@unsw.edu.au).

    conda create --name R_software
    conda activate R_software
    conda install R

If you then want to use RStudio you can install it using the following command:

    conda install RStudio

If there is a conflict between the installed Conda packages then you can create a new Conda environment using the following commands:

    conda create --name RStudio
    conda activate RStudio
    conda install RStudio

Once RStudio has been installed you can start a **FastX Desktop** with your required resources, open a terminal window by clicking on the icon at the top of the screen. Once the terminal opens you can then activate your Conda environment and then use the following command to start RStudio:

    rstudio