Customization of recursive feature elimination to do resampling inside rfFuncs

[shnsn]

Dear community,

I have been fighting with the customization of rfFuncs. There has been a topic about that, however the provided code here does not work:

Caret - recursive feature elimination (with upsamling?)

I'm not able to find the right answer in Applied Predictive Modelling or Caret documentation but maybe you guys could help. What would be the right way of doing RFE on a highly imbalanced classification problem so that the procedure could learn well patterns describing the minority class? The problem I facing right now is that with 5:95 target class ratio the outcome of RFE is not really representative of the patterns that I'm trying to discover. I couldn't find any possibility to use upsamling in the rfe function itself. Is there another way of doing that?

I have changed the rfFuncs$fit, which works outside rfe:
rf_fit <- function(x, y, first, last, ...){
loadNamespace("randomForest")

df_up <- caret::upSample(x, y)

randomForest::randomForest(
df_up[ , names(df_up) != "Class"],
df_up$Class,
importance = TRUE,
...)
}

However I get following error message when inserting in the rfe formula:
Error in { : task 1 failed - "argument is of length zero"

Can anybody help me??
Thx