Node that partitions the data into train- and test set. The dialog allows to specify the training data as in the sampling node. The test data is then the remainder.