Details
Description
In the example of trainlogistic, if we add a category feature. It will throw a NullPointerException while printing out the model.
The root cause is for category feature, the code try to use the feature name as the predictor rather than feature=value as the predictor while get the weight from the trained LR model.