Consider augment() support for a parsnip model, but require newdata as an argument #401
{{ message }}
Consider
augment() support for a parsnip model, but require newdata as an argument #401
This gives us augment support, with the safety of not predicting on the training set