R to PMML Execute failed

An error occurred while executing a R to PMML node after a R Learning node. The message is:
*ERROR R To PMML 0:243 Execute failed: Error in R code: *
Error: subscript out of bounds
Error: object ‘knime.model.pmml’ not found

I think it depends on the data since I have used the same connection other times without problems. For this take a look at the attached workflow.

R_to_PMML_Error.knwf (34.7 KB)

There was no data in den example. I built one with the IRIS dataset and it seems to work with PMML:

Was there a specific reason why you defined an extra path and package for the rpart instead of relying on the ‘standard’ installation?


Hi, @mlauber71,
as I explained earlier, the problem depends on the data that you can find in the workflow, a Knime table…
There is no specific reason for the additional path, but the problem does not concern this node…

well if you could provide us with the data it might be possible to have a look.

Maybe I’m not able to explain myself well, the data is in the workflow!

1 Like

OK - I would suggest to use the syntax as workflow relative since my subfolder could be at another location. I renamed the variables with _ instead of blanks and brackets and now it seems to work. Always better not to use too fancy names.

And for the record: this is a thing about the R pmml package and not strictly speaking a KNIME issue :slight_smile:

R_to_PMML_Error.knwf (136.9 KB)

1 Like

Great! Thank you very much, @mlauber71, for your kind help!
All the best,


This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.