image

🎉 🎉 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

Problems reading HierarchicalClusterModel

IgorIgor Member Posts: 6 Contributor II
Hello together,

I have created a model with AgglomerativeClustering and saved it as a file. It is a DendogramHierarchicalClusterModel as defined in AgglomerativeClustering. If I want to load the model with ClusterModelReader, I receive an exception:

[Fatal] Process failed: operator cannot be executed (com.rapidminer.operator.clustering.DendogramHierarchicalClusterModel cannot be cast to com.rapidminer.operator.clustering.ClusterModel). Check the log messages...

If I look into ClusterModelReader, I see that it casts ClusterModel on the loaded object (function read(), line "return (ClusterModel)ioReader.read();"). And DendogramHierarchicalClusterModel is not an extension of ClusterModel.

I have the version 4.5 and saw the same source for 4.6. Is it correct?

Regards,
Igor

Answers

  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Hi Igor,
    I would suggest using the IOWriter and Reader instead. I think the problem should be fixed anyway.

    Greetings,
      Sebastian
  • IgorIgor Member Posts: 6 Contributor II
    Hi Sebastian,

    thanks, it works now! Another suggestion: maybe you can handle the missing ID attribute for AgglomerativeClustering in a more intelligent /  understandable way - if I just apply this clustering for my data without ID attribute, I get a NullPointerException. Only after the exact look into the code I understood that IdTagging was required.

    For DBScanAlgorithm I see in the manual e.g. "If no id attribute is present, the operator will create one" - maybe it's a good way also for other clustering methods..

    Cheers,
    Igor
  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Hi,
    this has already be done for the RapidMiner 5.0

    Greetings,
      Sebastian
Sign In or Register to comment.