R/rename.R

In tidymodels/recipes: Preprocessing and Feature Engineering Steps for Modeling

#' Rename variables by name using dplyr
#'
#' `step_rename()` creates a *specification* of a recipe step that will add
#' variables using [dplyr::rename()].
#'
#' @inheritParams step_pca
#' @inheritParams step_center
#' @param ... One or more unquoted expressions separated by commas. See
#'  [dplyr::rename()] where the convention is **`new_name = old_name`**.
#' @param inputs Quosure(s) of `...`.
#' @template step-return
#' @details When an object in the user's global environment is referenced in
#'  the expression defining the new variable(s), it is a good idea to use
#'  quasiquotation (e.g. `!!`) to embed the value of the object in the
#'  expression (to be portable between sessions).
#'
#'  # Tidying
#'
#' When you [`tidy()`][tidy.recipe()] this step, a tibble is returned with
#' columns `terms`, `value` , and `id`:
#'
#' \describe{
#'   \item{terms}{character, the selectors or variables selected}
#'   \item{value}{character, `rename` expression}
#'   \item{id}{character, id of this step}
#' }
#'
#' @template case-weights-not-supported
#'
#' @family dplyr steps
#' @export
#' @examples
#' recipe(~., data = iris) %>%
#'   step_rename(Sepal_Width = Sepal.Width) %>%
#'   prep() %>%
#'   bake(new_data = NULL) %>%
#'   slice(1:5)
#'
#' vars <- c(var1 = "cyl", var2 = "am")
#' car_rec <-
#'   recipe(~., data = mtcars) %>%
#'   step_rename(!!!vars)
#'
#' car_rec %>%
#'   prep() %>%
#'   bake(new_data = NULL)
#'
#' car_rec %>%
#'   tidy(number = 1)
step_rename <- function(recipe, ...,
                        role = "predictor",
                        trained = FALSE,
                        inputs = NULL,
                        skip = FALSE,
                        id = rand_id("rename")) {
  inputs <- enquos(..., .named = TRUE)

  add_step(
    recipe,
    step_rename_new(
      terms = terms,
      trained = trained,
      role = role,
      inputs = inputs,
      skip = skip,
      id = id
    )
  )
}

step_rename_new <-
  function(terms, role, trained, inputs, skip, id) {
    step(
      subclass = "rename",
      terms = terms,
      role = role,
      trained = trained,
      inputs = inputs,
      skip = skip,
      id = id
    )
  }

#' @export
prep.step_rename <- function(x, training, info = NULL, ...) {
  step_rename_new(
    terms = x$terms,
    trained = TRUE,
    role = x$role,
    inputs = x$inputs,
    skip = x$skip,
    id = x$id
  )
}

#' @export
bake.step_rename <- function(object, new_data, ...) {
  dplyr::rename(new_data, !!!object$inputs)
}

#' @export
print.step_rename <-
  function(x, width = max(20, options()$width - 35), ...) {
    title <- "Variable renaming for "
    trained_names <- names(x$inputs)

    untrained_terms <- rlang::parse_quos(
      trained_names %||% "",
      rlang::current_env()
    )
    print_step(trained_names, untrained_terms, x$trained, title, width)
    invisible(x)
  }

#' @rdname tidy.recipe
#' @export
tidy.step_rename <- function(x, ...) {
  var_expr <- map(x$inputs, quo_get_expr)
  var_expr <- map_chr(var_expr, quo_text, width = options()$width, nlines = 1)

  tibble(
    terms = names(x$inputs) %||% character(),
    value = unname(var_expr) %||% character(),
    id = rep(x$id, length(x$inputs))
  )
}