Closed whizzalan closed 9 years ago
It is easy. Just do something like:
generic.cv(
train=...,
predict=function(model, newdata) predict(model, newdata, type="response"),
metric=...., .....)
generic.cv
also works with normal functions besides those from PivotalR. Recently I found generic.cv
is very useful and flexible. It is easier to use than those similar functions in package 'caret'. Maybe I will send a pull request to add more examples to its manual.
Thanks a lot. If it can use similar in package 'caret', that help me a lot. It mean I can use a lot of algorithm in package 'caret'.
May I use predict function with other output type? like predict(object, newdata, type="response") How do I do in fucntion generic.cv?