Rows with missing values on predictor variables are excluded. An unused column for the dependent variable is added to avoid errors.
prepDatForPredict(formula, newdata)Model formula
data.frame containing all variables that appear in the model
Object of class na.exclude
A column of ones for the dependent variable is added. Otherwise
model.matrix tries to take it from the formula's
environment, which is the original data. This usually results in an error due
to unequal variable length. This column is however not used.