The operator class 'h2o:deep_learning' is unknown. Possibly you must install a plugin for operators

DekwoKybon

Hi all,

My RapidMiner process works fine in RapidMiner Studio.

However, if I try to run it from Java I get an error message saying "a dummy operator" cannot be executed, in my case it's the Deep Learning operator.

I had this problem before with Cross Validation Operator but by searching the posts here I found out that I had to register the extra plugins. That did solve my problem with the Cross Validation operator but unfortunately the problem with the H20 Deep Learning Operator remains.

 

 

RapidMiner.setExecutionMode(RapidMiner.ExecutionMode.COMMAND_LINE);
Plugin.addAdditionalExtensionDir("/opt/rapidminer/lib/plugins");
Plugin.setInitPlugins(true);
RapidMiner.init();

// loads the process from the repository (if you do not have one, see alternative below)
RepositoryLocation pLoc = new RepositoryLocation("//Local Repository/processes/STELA_Studieduur_Learner");
ProcessEntry pEntry = (ProcessEntry) pLoc.locateEntry();
String processXML = pEntry.retrieveXML();
Process myProcess = new Process(processXML);
myProcess.setProcessLocation(new RepositoryProcessLocation(pLoc));

IOContainer ioResult = myProcess.run();

 

The JAR file is present in the plugins directory however: /opt/rapidminer/lib/plugins/h2o-7.3.0-all.jar

I also tried to add it directly to my classpath without succes.

 

I don't know if this is a related issue but I also get an 'Caused by: java.lang.ClassNotFoundException: com.rapidminer.tools.nexus.NexusCommunicationException' and a warning : 'Plugin initializer com.rapidminer.PluginInitCloudExecution.initPlugin of Plugin Cloud Execution caused an error: null '

 

Any ideas on how to solve this?

 

Best regards,

Wouter

DekwoKybon

Hey Marcel,

 

I'm beginning to think that the error is unrelated to the H20 deep learning operator after all. 

I started over with a very small project that is based off the one I was using before but stripping it from all the whistles and bells. 

 

Seems the error already occurs earlier in the process during the preprocessing of my data.RapidMiner keeps complaining about an unconnected input. I use a "Set Macros" operator to make selections possible in subprocesses but I get warnings, nevertheless the process runs fine. 

 

Anyway, I'm first going to try to get process run without warnings. 

 

Thanks for your assistance,

and Merry Christmas! :-)

Wouter