unable to apply learning algorithm as process document is making regular to text
I have the following process :
1. Process documents from files (where I load the text files with respective 6 classes ) --> this connects to set Role (which changes text attribute to REGULAR attribute to allow machine learning) -> Process documents from data ( I dont need the word vectors so I uncheck that, I keep text, within this process I tokenize, stopwords, stemming etc.) and then I feed this into validation operator. (bayes/svm)
What is happening here is in the example set the text column is going back to type "TEXT" from regular after running Process documents from Data. And hence I get the error Input ExampleSet has no attributes as there are zero regular attributes. And this is causing the process to fail. I have no idea why. I try to set the role again after this but then the error says "No examples in example set"