Make models into model_list object
as.model_list( ..., listed_models = NULL, target = ".outcome", model_class, tuned = TRUE, recipe = NULL, positive_class = NULL, model_name = NULL, best_levels = NULL, original_data_str, versions )
... |
|
---|---|
listed_models | Use this if your models are already in a list |
target | Quoted name of response variable |
model_class | "classification" or "regression". Will be determined if not provided |
tuned | Logical; if FALSE, will have super-class untuned_models |
recipe | recipe object from prep+_data, or NULL if the data didn't go through prep_data |
positive_class | If classification, the positive outcome class, otherwise NULL |
model_name | Quoted, name of the model. Defaults to the name of the outcome variable. |
best_levels | best_levels list as attached to data frames from add_best_levels |
original_data_str | zero-row data frame with names and classes of all columns except the outcome as they came into either the model training function such as tune_models or prep_data |
versions | A list containing the following environmental variables from model training: r_version, hcai_version, and other_packages (a tibble). If not provided, will be extracted from the current session. See healthcareai:::attach_session_info for details |
A model_list with child class type_list