Rapidminer is a cool toy to play with machine-learning/data-mining algorithms and it can interface with R. However, it was a bit problematic for me to get the R extension working properly on Mac OS X Leopard for R 2.11. Here is what works for me at the moment:

1) get rapidminer  (obviously 🙂 )

2) install rJava and JavaGD in R through install.packages()
JavaGD is necessary so that Rapidminer R extension doesn’t keep asking for a CRAN repository at the start up, also after installing JavaGD I was able to plot using R extension. Some people reported R tcltk package is also necessary. Maybe get that as well.

3) Set up the following lines at your ~/.bash_profile

export R_HOME=/Library/Frameworks/R.framework/Resources
export JAVA_HOME=/Library/Java/Home

I assume your R and Java installed in these locations.

4) At the first run of Rapidminer with R extension you will be asked to point the location of JRI dynamic-link library. Full path to mine is this:
/Library/Frameworks/R.framework/Resources/library/rJava/jri/libjri.jnilib

5) Run Rapidminer using “java -jar [RAPIDMINER_HOME]/lib/rapidminer.jar” on the terminal.
Clicking on rapidminer.jar starts the Rapidminer but then R extension doesn’t work. So you have to run it through the terminal as shown above.