Rv3 with interactive RSnippet under Linux

[Jonas]

Hi,

as the topic says: I want to use Rv3 with the interactive R Snippet node from KNIME labs under Linux. I read the according FAQ entry and set the variable in the KNIME ini. The rJava\jri folder does not contain a x64 subfolder. But it contains the libjri.so. This is the Linux equivalent to the jri.dll, right?

I also installed the R package and the contained example is working (with Rv3) when i execute it without KNIME.

 

Has anybody a souteion to get these R snippets running with Rv3?

thanks!

[Jonas]

has nobody here tried to use Rv3 in the interactive R Snippet nodes before?

[kilian.thiel]

Hi Jonas,

i got Rv3 running with the R interactive nodes. First in the KNIME preferences "R (labs)" the path to the Rv3 installation has to be specified. On my windows machine this is "C:\Program Files\R\R-3.0.2". For this R installation the rJava lib has to be installed. If the lib is installed and where it is can be found out with:

> installed.packages()

On my machine i get:

rJava        "rJava"        "C:/Users/Kilian/Documents/R/win-library/3.0" "0.9-6"

Now the path to the rJava lib has to be specified in the knime.ini file. The specified directory must contain the jri.dll (in case of a windows machine).

-Djava.library.path=C:\Users\Kilian\Documents\R\win-library\3.0\rJava\jri\x64

Finally restart KNIME to apply changes.

Hope this helps.

Cheers, Kilian

 

[thor]

On my Linux I can use R3 with the interactive node by just installing them. If the R specified in the preferences has rJava installed then everything works out of the box.