The main functions of cvms are:
The difference between
cvms only provided the option to cross-validate Gaussian and binomial regression models, fitting the models internally with the
glmer() functions. The
cross_validate() function has thus been designed specifically to work with those functions.
To allow cross-validation of custom model functions like support-vector machines, neural networks, etc., the
cross_validate_fn() function has been added. You provide a model function and (if defaults fail) a predict function, and it does the rest (see examples below).