Only one label error message

sschult

I'm using the Discretize by User Specification operator to turn my label from numerical data into nominal with 5 different classes, those being lowest, low, on target, high, and highest.  After doing this all 5 classes are showing up in the data. I am then using a Sliding Window Validation using various models but am getting an error message for several models saying "Only one label.  The learning scheme "Model Name used" does not have sufficient capabilities for handling an example set with only one label value."  I have one label with 5 classes showing up in the example set just before going into the sliding window.  Anyone know why I'm getting this error?  I've tried Random Forest, k-nn, Generalized Linear Model, and Decision Tree all giving me this error message.

sschult

I have separately used Generate Attributes to make the same nominal attribute with 5 classes and still get the same error message.

sschult

The issue was a Session ID as a 2nd special attribute.  The model saw ID, which it new was an ID, Session ID as a special attribute, and the label.  It was not seeing the session ID change inside some windows resulting in the one label error.

sschult

Actually this still isn't working.  Tried the above with a breakpoint after the GLM inside the sliding window and it would step through the GLM one training set at a time.  Once I removed the breakpoint it gave me the same error message.