A leave-one-out apply helper

I had to compare how each listener differed from the mean of the other listeners, so I cooked up:

leave_one_out_apply <- function(xs, f, ...) {
  lapply(seq_along(xs), function(i) f(xs[-i], ...) )
}