I have been successfully using the SVM nodes until a couple of days ago. Now what was an intermittant error has started to happen every time I try to run the "SVM Predictor" node. Here is the error:
Unable to clone input data at port 0: java.lang.RuntimeException: Parsing of PMML v 3.x/4.0 document failed.
Occaisionally, if I restart KNIME it will work again, but quickly returns to throwing off the error. I am running 2.10.1.
I am still working on a version of this workflow, so I will try and duplicate the error. At that point I will send the whole error message as well as the pmml file.
ERROR SVM Predictor Unable to clone input data at port 0: java.lang.RuntimeException: Parsing of PMML v 3.x/4.0 document failed.
I just recreated it, and it will not write the PMML model for the same reason (giving the same error). However, if I save the workflow, exit KNIME and restart, the rest of the workflow will work.
Any idea why this error is being generated or how to avoid it? Thanks,
I get a similar error. Did you guys find a solution for the same? If so, can you please share? What is being cloned here first of all.
This is the error i get :
ERROR Decision Tree Predictor Unable to clone input data at port 1 (Decision Tree Model): java.lang.RuntimeException: Parsing of PMML v 3.x/4.0 document failed.
Changes on PMML package for R, in my case Knime 3.4, R 3.4.2, and PMML file version package 1.5.2 result: "PMML version=“4.3” xmlns="http://www.dmg.org/PMML-4_3\ xml…"
This version is not supported by RtoPMML3.4.0, neither JPMML 3.4.0, so you must convert to 4.0 version or 3.x. PMML version is file version, not package version. There’s not too much difference as you can see at:http://dmg.org/pmml/v4-3/GeneralStructure.html.
Simple edit the file to remove new version or Recompile RtoPMML node to support 4.3 files.
J
.