Execute R Script ARIMA Process Fails

pedrodomingosdvpedrodomingosdv Member Posts: 17 Contributor I
edited December 2018 in Help
Hi,
I'm trying to use R to build ARIMA model forecast with it. The script works well in R software, but when I try to use in Rapidminer with Execute R I always get an error message (bellow the detailed message).

I try to adapt an example process with R script to train and forecast with ARIMA, I found in Rapidminer.
Both, my script and the example, return the same error message.

Note: With other processes that use Execute R, like the tutorials, seems that there is no issue. 

Enclosed the process.

Thanks for your help,
Pedro

Error message details:

Exception: com.rapidminer.operator.OperatorException
Message: Script terminated abnormally.
Stack trace:

  com.rapidminer.extension.rscripting.operator.scripting.AbstractScriptRunner.run(AbstractScriptRunner.java:166)
  com.rapidminer.extension.rscripting.operator.scripting.AbstractScriptingLanguageOperator.doWork(AbstractScriptingLanguageOperator.java:90)
  com.rapidminer.extension.rscripting.operator.scripting.r.RScriptingOperator.doWork(RScriptingOperator.java:73)
  com.rapidminer.operator.Operator.execute(Operator.java:1025)
  com.rapidminer.operator.execution.SimpleUnitExecutor.execute(SimpleUnitExecutor.java:77)
  com.rapidminer.operator.ExecutionUnit$2.run(ExecutionUnit.java:812)
  com.rapidminer.operator.ExecutionUnit$2.run(ExecutionUnit.java:807)
  java.security.AccessController.doPrivileged(Native Method)
  com.rapidminer.operator.ExecutionUnit.execute(ExecutionUnit.java:807)
  com.rapidminer.operator.OperatorChain.doWork(OperatorChain.java:428)
  com.rapidminer.operator.Operator.execute(Operator.java:1025)
  com.rapidminer.Process.execute(Process.java:1322)
  com.rapidminer.Process.run(Process.java:1297)
  com.rapidminer.Process.run(Process.java:1183)
  com.rapidminer.Process.run(Process.java:1136)
  com.rapidminer.Process.run(Process.java:1131)
  com.rapidminer.Process.run(Process.java:1121)
  com.rapidminer.gui.ProcessThread.run(ProcessThread.java:65)
Tagged:

Best Answer

  • pedrodomingosdvpedrodomingosdv Member Posts: 17 Contributor I
    Solution Accepted
    Hi hughesfleming68, many thanks for your help. It is solved.
    RM is indeed a very good software, but it is a community like this that led me to persuade my bosses to acquire RM.

    Thank 

Answers

  • hughesfleming68hughesfleming68 Member Posts: 323 Unicorn
    edited December 2018
    Try putting library(forecast) before the rm_main = function(data) and take out the require(forecast). See if that works. It is helpful to include your data as well as it makes it easier to reproduce your error. Also some details like OS, R version etc.




  • pedrodomingosdvpedrodomingosdv Member Posts: 17 Contributor I
    Hi hughesfleming68, I put library(forecast), but got the same error message.
    OS: Windows 7 Professional / Version 6.1.7601 Service Pack Build 7601
    RM Version: RM Studio Free 9.0.003

    Enclosed the data
  • hughesfleming68hughesfleming68 Member Posts: 323 Unicorn
    Thanks. I will get back to you later today. Just to make sure...In the preferences, does the test for the RScript executable work correctly? It should say something like Detected R scripting front-end. Mine is R-3.5.1. What does yours say?
  • pedrodomingosdvpedrodomingosdv Member Posts: 17 Contributor I
    Hey, yes all ok with the test.
    I tried to run the "Execute R" tutorials and all good also. No issue at all.
    I feel that must be something with the code, but not sure


  • hughesfleming68hughesfleming68 Member Posts: 323 Unicorn
    edited December 2018
    Take a look at the fixed process. The problem started with the date format. The Date to Nominal operator fixed that.
  • pedrodomingosdvpedrodomingosdv Member Posts: 17 Contributor I
    Hi hughesfleming68, many thanks for your help. It is solved.
    RM is indeed a very good software, but it is a community like this that led me to persuade my bosses to acquire RM.

    Thank as a lot!
  • jczogallajczogalla Employee, Member Posts: 144 RM Engineering
    Hi @hughesfleming68 and @pedrodomingosdv!
    We know that the date conversion in the R extension is a problem right now and are working on it. But it's good to see that you found a workaround. :smile:
    Cheers
    Jan
Sign In or Register to comment.