Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

"R 2.12"

OK, I know this has been asked many times (because the R interface is a great addition, but for those of us that dont really understand JAVA and environment variables (windows), the install is painful).

I had R 2.11and RM worked great. Then I updated to R 2.12 and all hell broke loose! :-\
I *thought* I followed the instruction when RM loaded but I get either errors about dependent libraries not being found or incompatible architecture.

Can anyone explain what needs to be set up?

I have environment variables: classpath, path, JAVA_HOME and R_HOME. What needs to be in each?

i am using 64 bit windows, R and RM.

The R.dll is under ..bin\x64
The library folder is up a few folders (where all the package libraries are at for R). I have rjava (jri.dll) downloaded and think I pointed RM to this file under 64 bit.

Anyone know how to fix?

Many thanks! Brian

Find more posts tagged with

AI Studio

Accepted answers

All comments

dragoljub

In RapidMiner go to tools-->preferences-->R tab and set the path to the new location of 64bit R dll. I believe R combined their 32 and 64bit installers. I guess since RM is pretty much designed for 64bit their next update will check the 64bit folder too.

-Gagi

B_Miner

Gagi,

Thanks for the reply -I made that change (the preference tab was pointing to the jri.dll prior) but the result was the same. Something else must be screwed up.

B_Miner

Finally got it working

gusr

Hi B_Miner

How did you sort it out? What was it you had to change in the end to get it to work?

All the best,
gusr

B_Miner

I am using 64 bit Windows 7.

In the environment variables
JAVA_HOME I have C:\Program Files\Java\jre6
R_HOME I have C:\Program Files\R\R-2.12.0
Path I have C:\Program Files\R\R-2.12.0\library
and finally in Rapid Miner under preferences and R and have C:\Program Files\R\R-2.12.0\library\rJava\jri\x64\jri.dll

land

Hi,
we are going to make the installation procedure more simpler by bundling R directly, but we have still some legal issues as well as architectural problems, because the bundling would not be platform independent...

Greetings,
Sebastian