SDSC "R" roll

Overview

This Rocks roll bundles the R statistical computing package.

R is a free software environment for statistical computing and graphics. For more information about R please visit the official web page.

Requirements

To build/install this roll you must have root access to a Rocks development machine (e.g., a frontend or development appliance).

If your Rocks development machine does not have Internet access you must download the appropriate R source file(s) using a machine that does have Internet access and copy them into the src/R directory on your Rocks development machine.

Dependencies

yum install tk-devel readline-devel

tk-devel is not needed by R itself; however, a number of R modules depend on R having been compiled with it present.

readline-devel is needed to build R.

The sdsc-roll must be installed on the build machine, since the build process depends on make include files provided by that roll.

The roll sources assume that modulefiles provided by SDSC compiler rolls are available, but it will build without them as long as the environment variables they provide are otherwise defined.

If they are present, the build process uses the Intel MKL libraries and assumes that the mkl modulefile provided by the SDSC intel-roll is available. It will build without the modulefile as long as the environment variables it provides are otherwise defined. If MKL is not present, the build process uses R's internal BLAS implementation.

Building

To build the R-roll, execute this on a Rocks development machine (e.g., a frontend or development appliance):

% make 2>&1 | tee build.log

A successful build will create the file R-*.disk1.iso. If you built the roll on a Rocks frontend, proceed to the installation step. If you built the roll on a Rocks development appliance, you need to copy the roll to your Rocks frontend before continuing with installation.

This roll source supports building with different compilers; by default, it builds using the gnu compilers. To build for a different configuration, use the ROLLCOMPILER make variable, e.g.,

% make ROLLCOMPILER=intel

This roll source supports building with different compilers. The ROLLCOMPILER make variable can be used to specify the names of compiler modulefiles to use for building the software, e.g.,

% make ROLLCOMPILER=gnu/4.8.1

If the environment variable MKL_ROOT is set either before the build or by the ROLLCOMPILER module, the build process will incorporate the Math Kernel Libraries into the R installation.

Building the R-roll under Rocks version 6 requires additional system libraries that are bundled with an earlier version of the roll sources. To use this version, first make a git clone of the R-roll source, then execute the following before following the build instructions given above.

cd R-roll
git checkout ROCKSv6_FINAL

Installation

To install, execute these instructions on a Rocks frontend:

% rocks add roll *.iso
% rocks enable roll R
% cd /export/rocks/install
% rocks create distro

Subsequent installs of compute and login nodes will then include the contents of the R-roll. To avoid cluttering the cluster frontend with unused software, the R-roll is configured to install only on compute and login nodes. To force installation on your frontend, run this command after adding the R-roll to your distro

% rocks run roll R host=NAME | bash

where NAME is the DNS name of a compute or login node in your cluster.

In addition to the software itself, the roll installs R environment module files in:

/opt/modulefiles/applications/R

Testing

The R-roll includes a test script which can be run to verify proper installation of the roll documentation, binaries and module files. To run the test scripts execute the following command(s):

% /root/rolltests/R.t