Variance calculations are used in these computations within each outcome class. For this reason, zero-variance predictors (i.e., with a single unique value) within each class should be eliminated before fitting the model.



topepo/parsnip documentation built on April 16, 2024, 3:23 a.m.