🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

Bug in new Deep learning Extension 0.9 sample process

wongcrwongcr Member Posts: 7 Contributor II
RM 9.1 and the new Deeplearning4J extension 0.9 has a bug in the 01 Airline Passenger regression with LSTM v1 sample process.

The issue is the conversion of the dataset from collection to Timeseries to Tensor. The message is "Insufficient capability: The operation Timeseries to Tensor does not have sufficient capabilities for the given data set: polynomial attributes not supported" error


varunm1
1
1 votes

Fixed and Released · Last Updated

DLE-52

Comments

  • sgenzersgenzer 12Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959  Community Manager
    @wongcr yes confirmed. Thank you for reporting. Moving to Product Feedback.
  • dias_1013dias_1013 Member Posts: 2 Newbie
    RM 9.1 and the new Deeplearning4J extension 0.9 has a bug in the Deep Learning +  Neural Net process.
    The message is: Could not initialize class org.nd4j.linalg.factory.Nd4j.
    Exception: java.lang.NoClassDefFoundError
  • pschlunderpschlunder Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, RMResearcher, Member Posts: 96  RM Research
    @wongcr sorry for the bug in the tutorial process, it's due to working on studio version with different attribute handling of the windowing operator. Please just limit the attributes to numerical ones to solve it. You can set the "attribute filter type" of the windowing operator to "value_type" and choose "numerica" as the value type.

    Hope this helps,
    Philipp
  • pschlunderpschlunder Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, RMResearcher, Member Posts: 96  RM Research
    can you share a process or a studio log please? Are you using a computer with an old 32bit system or 32bit java version?

    Regards,
    Philipp
Sign In or Register to comment.