Merge pull request #374 from hpcleuven/feature/R_devtools

Feature/r devtools

Author: WouterVanAssche

source/software/r_devtools.rst

@@ -0,0 +1,76 @@
+.. _r_devtools:
+
+Installing R packages with devtools
+===================================
+
+Introduction
+~~~~~~~~~~~~
+
+The installation of some R packages may require the use of devtools.
+Devtools is an R package that facilitates the installation of other
+R packages from github, gitlab, bitbucket or other repositories.
+In what follows github will be used as an example. Please consult the
+devtools_ documentation for examples of other repositories.
+
+Depending on how your R library is managed, you will need a slightly different
+approach to use and install devtools.
+
+.. note::
+
+  When consulting the devtools documentation, make sure that it is the correct version!
+  It should match the devtools version included in the R module. To check which devtools version is installed:
+
+  .. code-block:: r
+    
+    library(devtools)
+    sessioninfo::session_info()
+
+Installing in a local R library
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+If you manage your R packages in a :ref:`local R library<r_package_management_standard_lib>` under ``$VSC_DATA/R``
+while using a centrally installed R module, you can use the devtools package included in the module.
+You will need to execute the following commands in the R console:
+
+.. code-block:: r
+
+   > # First check that the R library path points to your local R library:
+   > .libPaths()
+   > # Set the R library path if this is not the case.
+   > .libPaths("/data/leuven/XXX/vscXXXXX/R/")
+   > # Load devtools and e.g. install your package from github:
+   > library(devtools)
+   > devtools::install_github("Developer/Package")
+
+.. note::
+
+  The devtools package is **not** included in "-bare" R modules, e.g. R/4.0.2-foss-2018a-bare.
+
+Installing in a conda environment
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+If you are using conda to manage your R packages, you should first install
+devtools in your conda environment. The following steps assume that you 
+already have a conda environment named "science". If you do not yet have
+a conda environment, First create a :ref:`conda environment<r_package_management_conda>`. 
+In the following example, it is assumed that your miniconda environment is installed in ``$VSC_DATA/miniconda3``.
+
+.. code-block:: bash
+
+   $ # Activate your conda environment and install devtools
+   $ source activate science
+   $ conda install -c conda-forge r-devtools
+   $ # Launch R
+   $ R
+
+.. code-block:: r
+
+   > # Check that the R library path points to your conda R library
+   > .libPaths()
+   > # Set the R library path if this was not the case.
+   > .libPaths("/data/leuven/XXX/vscXXXXX/miniconda3/envs/science/lib/R/library")
+   > # Load devtools and e.g. install your package from github:
+   > library(devtools)
+   > devtools::install_github("Developer/Package")
+
+.. _devtools: https://www.rdocumentation.org/packages/devtools

source/software/r_package_management.rst

@@ -1,16 +1,18 @@
+.. _R_package_management:
+
 R package management
 ====================
 
 Introduction
 ------------
 
-Most of the useful R packages come in the form of packages that can be
-installed separately. Some of those are part of the default installation
-on VSC infrastructure. Given the astounding number of packages, it is
-not sustainable to install each and everyone system wide. Since it is
-very easy for a user to install them just for himself, or for his
-research group, that is not a problem though. Do not hesitate to contact
-support whenever you encounter trouble doing so.
+Most of the useful R packages can be installed separately. Some of those are 
+part of the centrally installed R modules. However, given the astounding number of 
+packages, it is not sustainable to install each and everyone of them system wide. 
+Fortunately, it is very easy for users to install R packages themselves.
+If you do encounter problems when doing so, do not hesitate to contact support.
+
+.. _r_package_management_standard_lib:
 
 Standard R package installation
 -------------------------------
@@ -45,8 +47,12 @@ library, .e.g., ::
 
    $ module load GSL/2.5-GCC-6.4.0-2.28
 
-Note that R packages often depend on the specific R version they were installed
-for, so you may need to reinstall them for other versions of R.
+.. note::
+
+  R packages often depend on the specific R version they were installed
+  for, so you may need to reinstall them for other versions of R.
+
+.. _r_package_management_conda:
 
 Installing R packages using conda
 ---------------------------------
@@ -176,4 +182,6 @@ original state, use the following command
 More information
 ~~~~~~~~~~~~~~~~
 
-Additional information about conda can be found on its `documentation site <https://conda.readthedocs.io/en/latest/>`__.
+Additional information about conda can be found on its `documentation site <https://docs.conda.io/en/latest/>`__.
+
+For installing R packages from github or other repositories see also :ref:`R devtools<r_devtools>`:

source/software/software_development.rst

@@ -39,4 +39,4 @@ Libraries
    perl_package_management
    python_package_management
    r_package_management
-
+   r_devtools