For some postprocessors, enable simple application of adjustments

[topepo]

If a tailor object has not tuning parameters or estimation, we should have an API that let's us just push predictions through.

fit.action_tailor() is where the columns names. are set. However, we can update the tailor objects with that information from the workflow.

For example:

get_output_columns <- function(x, syms = FALSE) {
  mod_res <- extract_fit_parsnip(x)
  mod_mode <- mod_res %>% purrr::pluck("spec") %>% purrr::pluck("mode")

  if (mod_mode == "regression") {
    res <- list(estimate = ".pred", probabilities = character(0))
  } else if (mod_mode == "classification") {
    res <-
      list(
        estimate = ".pred_class",
        probabilities = paste0(".pred_", mod_res$lvl)
      )
  } else {
    cli::cli_abort("Unsupported model mode: {mod_mode}")
  }
  res <- c(list(outcome = outcome_names(x)), res)

  if (syms) {
    res <- purrr::map(res, rlang::syms)
  }
  res
}

could be used to populate the columns.

The reason for this is that the tune package will have to step through an unneeded loop and have a loft of redundant computations.

[topepo]

This does not appear to be needed anymore.

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.