Nominal predictor to be used in regression

Questions
6

x_1 = shared(train.content_id).values because casting a pd.Series to theano.shared might not work.

x1.eval().shape would be the easiest way to check for theano.shared