Setting up Nova to work with R

The NOVA platform is able to communicate with the statistical analysis software R. So that, it is possible to call R functions from NOVA as well as interchange of data with R. It just requires some modifications to connect the two softwares.

First, you need to install in R the package rJava. If Java JDK is already installed, install.packages(“rJava”) in R will do the trick.

Second, to talk to R through NOVA you have to specify environmental variables. First of all you need to publish your R installation path “R_HOME” and to add the native JRI-library included in the rJava folder (in your R library) to your path “R_LIBRARY_DIR”. Make sure that the files JRI.jar, JRIEngine.jar and REngine.jar are included into the rJava folder. Otherwise, you can download them from here:


For Windows (from NOVA build 560):
In the environment variables, add:

R_HOME: C:\Program Files\R\R-3.0.2
R_LIBRARY_DIR: %USERPROFILE%\Documents\R\win-library\3.0\rJava\jri\x64

Note that, you also have to specify the path of your R library trough R_LIBS and R_LIBS_USER. For example:

R_LIBS: %R_HOME%\library
R_LIBS_USER: % USERPROFILE %\Documents\R\win-library\3.0

Then, just add %R_HOME%/bin/x64;%R_LIBRARY_DIR%;%R_LIBS%;%R_LIBS_USER% to your PATH.
For example:

Now, it”s time to reboot”¦


For Mac (it”s simpler”¦as always):
In your file .profile, you just have to add:

export R_HOME=/Library/Frameworks/R.framework/Resources
export R_DOC_DIR=/Library/Frameworks/R.framework/Resources/doc
export R_INCLUDE_DIR=/Library/Frameworks/R.framework/Resources/include
export R_LIBRARY_DIR=/Library/Frameworks/R.framework/Versions/Current/Resources/library/rJava/jri


For Linux (it”s even more simple):

Assuming your are using Java-7-Oracle (just check which java JDK you”re using with the command javac ”“version in a terminal), just put in a terminal:

export R_HOME=/usr/lib/R
export JAVA_HOME=/usr/lib/jvm/java-7-oracle
export R_LIBRARY_DIR=/usr/lib/R/library/rJava/jri

The latter commands should now be in the file /etc/environment.

Et voilà !



R Core Team (2013). R: A language and environment for statistical computing. Foundation for Statistical Computing, Vienna, Austria. URL