I have here a training and testing dataset. The training set has more unique categorical values than the testing set and I would like to filter out those rows f