How to figure out which column names are illegal in ranger?

Question

Here is a ranger call:

rf_fit <- 
  rf_mod %>% 
  fit(my_outcome_factor ~ ., data = data_train)

and the output:

Error in