I have some issues at the execution of the decision tree ruleset with the following output:
ERROR Decision Tree to Ruleset 0:24 Execute failed: (“NullPointerException”): null
I do not know what to do because I am not experienced in KNIME, I do need that for a student task in which I am a little bit under time pressure
The Workflow is attached.
Happy for some Feedback
BI2-Steffen.knwf (3.9 MB)
Hi @Hsbcoeeer and welcome to the KNIME forum
Nice workflow well commented. Would it be possible for you to post please the Excel data or the workflow already run until where it fails ? Otherwise it will be dificult to exactly reproduce the error.
Thanks & regards,
Hi @aworker thanks for your reply.
I am not sure where or which Excel data you need. Do you mean the input data or something exported from the workflow?
The workflow was exported with the Reset option checked and since the path to the data is absolute (not relative to the workflow with the knime://knime.workflow/" option) I thought the data was not included.
But just checked inside the workflow and it is there. I will fix the acces to the data, run the workflow and eventually let you know if I find why you get this error.
thanks for the support. Attached you will find the the excel-data. The workflow-data without the reset option is too big for the upload.
I am looking forwar to your reply.
Thanks and regards
vTargetMail.xlsx (3.0 MB)
The problem was with the Numeric Binner node:
For a reason I do not understand, it is not binning correctly and you only get one class (BUYER) in the end. This screws up everything else
I have tided up a bit the reading and saving of data in your workflow by creating two local data directories inside the workflow (_LOCAL_DATA & _LOCAL_RESULTS) from/to which the data is read and written. To access them, you need to use the following setting:
Not knowing initially what was the problem, I also replaced the Model Reader / Writer nodes by MMPL writer and reader nodes which in this case makes sense (although both should be ok).
BI2-Steffen_V2.knwf (3.7 MB)
Hope this helps.
Good luck with your student task !
many thanky to you, you helped me many ways out
Thank you very much, this is awesome.
@Hsbcoeeer My pleasure to help !
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.
I just wanted to let you know that the NullPointerException you saw initially was fixed and is part of our 4.4.1 release.
As you guys already figured out, it was caused by the decision tree containing only a single node.
Such trees will now be turned into a very simple ruleset that always predicts the majority class of the training data.