Can you help me about this bug?

User145718

• Exception: java.lang.IndexOutOfBoundsException
• Message: Index: 0, Size: 0
• Stack trace:
• java.util.ArrayList.rangeCheck(ArrayList.java:657)
• java.util.ArrayList.get(ArrayList.java:433)
• com.rapidminer.operator.features.Population.get(Population.java:111)
• com.rapidminer.gui.dialog.IndividualSelector.(IndividualSelector.java:103)
• com.rapidminer.gui.dialog.IndividualSelector.(IndividualSelector.java:81)
• com.rapidminer.gui.dialog.IndividualSelector.(IndividualSelector.java:77)
• com.rapidminer.operator.features.FeatureOperator.doWork(FeatureOperator.java:392)
• com.rapidminer.operator.features.selection.FeatureSelectionOperator.doWork(FeatureSelectionOperator.java:162)
• 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)
•  

IngoRM

Hey,

Can you help us describing what you have been trying to do and how?  Please also include the XML description of the process if possible - that typically helps a lot...

Many thanks,

Ingo

User145718

Hey there,
I have been trying to attack on the complicated problem (at least it is complicated for me). I have data with many attributes in itself and I'm trying to reduce its size with "Optimize Selection" operator using backward elimination. However, the problematic part is that I have a performance indicator which changes when we reduce the number of attributes. 

My current approach is utilizing"cross-validation" operation in the "Optimize selection"  via splitting on a batch. Hence my performance indicator is related to two exact attributes. Then in the training part, I use the default model. Then in the testing part, after labeling the data I generated my performance indicator using the two exact attributes. Moreover, I utilized the "extract performance" module to select my performance indicator.

My approach can be totally wrong, that's maybe why I get this error. Finally, let me clarify the problem once more. I have data with many attributes and two attributes that are related to performance. I want to reduce the number of attributes. However, I can only calculate the performance indicator after reducing the number of attributes hence my performance indicator is std deviation of the product of two attributes for whole data points. 

Even little help would be appreciated.

Thanks

varunm1

Hi @User145718

Are you saying that your model is trained on multiple attributes but you are trying to test (performance) only on two attributes in the model?

providing an XML code of your process is much appreciated.

User145718

Thanks for the reply,

Yes. However, the performance indicator changes when we reduce the size of the multiple attributes.

varunm1

 The performance values changes based on attribute size, that is correct. Are you getting any errors when running the process? Sorry, it is a bit difficult to understand without your process. 

User145718

There are no exact errors. When I run the model, it says the process failed. Index:0 and Size:0

varunm1

It might be because you were training model on a different number of attributes and testing them on a different number of attributes. You can check error in log (View --> Show Panel --> Log). 

User145718

Hi @varunm1

I tried to solve this problem with aggregating on attributes that I mentioned as performance indicators. Right now, what I need to figure out is how to select the best attributes to aggregate based on the performance indicators. I tried "optimize parameters (brute force)" operator, however I cannot insert aggregation attributes as input to that operator. Is there a way to handle this problem?

varunm1

Hello @User145718

If you want to keep only the aggregated attribute, you need to attach Generate aggregation operator inside optimize selection operation between input exa and validation operator. See below image. If you want old attributes as well with aggregated attribute then you need to select "keep all" highlighted in the image below. If you want only aggregated column as input then remove "keep all". Specify what kind of aggregation you need in an aggregate function. If you are aggregating on all performance vectors then select all in attribute filter type.

sgenzer

hi @User145718 thx for this bug report. I will need to reproduce this error in order to file the bug. Can you please send the rapidminer-studio.log file, XML process, and a data set (if needed)? You can send them to me via PM if you like.