These functions are the interface to the various model types for mod_eval()
, and through
that to all the other mod_
functions that need to evaluate models, e.g. mod_effect()
, mod_cv()
, and so on.
mod_eval_fun(model, data = NULL, interval = "none", ...)
model | A model object of the classes permitted |
---|---|
data | Usually, a data table specifying the inputs to the model. But if not specified, the training data will be used. |
interval | One of "none", "confidence", or "prediction". Not all model types support "prediction" or even "confidence". |
... | additional arguments |
All of the eval_
functions are ex
These functions return a numerical vector (for regression types) or
a matrix of probabilities (for classifiers)