R/class.R

In recipes: Preprocessing and Feature Engineering Steps for Modeling

#' Check Variable Class
#'
#' `check_class` creates a *specification* of a recipe
#'  check that will check if a variable is of a designated class.
#'
#' @inheritParams check_missing
#' @param class_nm A character vector that will be used in `inherits` to
#'  check the class. If `NULL` the classes will be learned in `prep`.
#'  Can contain more than one class.
#' @param allow_additional If `TRUE` a variable is allowed to
#'  have additional classes to the one(s) that are checked.
#' @param class_list A named list of column classes. This is
#'  `NULL` until computed by [prep()].
#' @template check-return
#'
#' @family checks
#' @export
#' @details
#' This function can check the classes of the variables
#'  in two ways. When the `class` argument is provided
#'  it will check if all the variables specified are of the
#'  given class. If this argument is `NULL`, the check will
#'  learn the classes of each of the specified variables in `prep`.
#'  Both ways will break `bake` if the variables are not of
#'  the requested class. If a variable has multiple
#'  classes in `prep`, all the classes are checked. Please note
#'  that in `prep` the argument `strings_as_factors` defaults to
#'  `TRUE`. If the train set contains character variables
#'  the check will be break `bake` when `strings_as_factors` is
#'  `TRUE`.
#'
#'  # Tidying
#'
#'  When you [`tidy()`][tidy.recipe()] this check, a tibble with columns
#'  `terms` (the selectors or variables selected) and `value` (the type)
#'  is returned.
#'
#' @template case-weights-not-supported
#'
#' @examplesIf rlang::is_installed("modeldata")
#' library(dplyr)
#' data(Sacramento, package = "modeldata")
#'
#' # Learn the classes on the train set
#' train <- Sacramento[1:500, ]
#' test <- Sacramento[501:nrow(Sacramento), ]
#' recipe(train, sqft ~ .) %>%
#'   check_class(everything()) %>%
#'   prep(train, strings_as_factors = FALSE) %>%
#'   bake(test)
#'
#' # Manual specification
#' recipe(train, sqft ~ .) %>%
#'   check_class(sqft, class_nm = "integer") %>%
#'   check_class(city, zip, type, class_nm = "factor") %>%
#'   check_class(latitude, longitude, class_nm = "numeric") %>%
#'   prep(train, strings_as_factors = FALSE) %>%
#'   bake(test)
#'
#' # By default only the classes that are specified
#' #   are allowed.
#' x_df <- tibble(time = c(Sys.time() - 60, Sys.time()))
#' x_df$time %>% class()
#' \dontrun{
#' recipe(x_df) %>%
#'   check_class(time, class_nm = "POSIXt") %>%
#'   prep(x_df) %>%
#'   bake_(x_df)
#' }
#'
#' # Use allow_additional = TRUE if you are fine with it
#' recipe(x_df) %>%
#'   check_class(time, class_nm = "POSIXt", allow_additional = TRUE) %>%
#'   prep(x_df) %>%
#'   bake(x_df)
check_class <-
  function(recipe,
           ...,
           role = NA,
           trained = FALSE,
           class_nm = NULL,
           allow_additional = FALSE,
           skip = FALSE,
           class_list = NULL,
           id = rand_id("class")) {
    add_check(
      recipe,
      check_class_new(
        terms            = enquos(...),
        trained          = trained,
        role             = role,
        class_nm         = class_nm,
        allow_additional = allow_additional,
        class_list       = class_list,
        skip             = skip,
        id               = id
      )
    )
  }

## Initializes a new object
check_class_new <-
  function(terms, role, trained, class_nm,
           allow_additional, class_list, skip, id) {
    check(
      subclass         = "class",
      terms            = terms,
      role             = role,
      skip             = skip,
      trained          = trained,
      class_nm         = class_nm,
      allow_additional = allow_additional,
      class_list       = class_list,
      skip             = skip,
      id               = id
    )
  }

#' @export
prep.check_class <- function(x,
                             training,
                             info = NULL,
                             ...) {
  col_names <- recipes_eval_select(x$terms, training, info)

  # vapply requires a very specific return here
  # class can give back multiple values, return shape
  # is not predetermined. Thats why we use lapply instead.
  if (is.null(x$class_nm)) {
    class_list <- lapply(training[, col_names], class)
  } else {
    class_list <- rep(list(x$class_nm), length(col_names))
    names(class_list) <- col_names
  }

  check_class_new(
    terms            = x$terms,
    role             = x$role,
    skip             = x$skip,
    trained          = TRUE,
    class_nm         = x$class_nm,
    allow_additional = x$allow_additional,
    class_list       = class_list,
    id               = x$id
  )
}

# we don't use inherits() because class_nm
# can be of length > 1. inherits will result
# in TRUE if just one of the classes in class_nm
# is present in x.
bake_check_class_core <- function(x,
                                  class_nm,
                                  var_nm,
                                  aa = FALSE) {
  classes <- class(x)
  missing <- setdiff(class_nm, classes)
  if (length(missing) > 0) {
    rlang::abort(
      paste0(
        var_nm,
        " should have the class(es) ",
        paste(class_nm, collapse = ", "),
        " but has the class(es) ",
        paste(classes, collapse = ", "),
        "."
      )
    )
  }

  extra <- setdiff(classes, class_nm)
  if (length(extra) > 0 && !aa) {
    rlang::abort(
      paste0(
        var_nm,
        " has the class(es) ",
        paste(classes, collapse = ", "),
        ", but only the following is/are asked ",
        paste(class_nm, collapse = ", "),
        ", allow_additional is FALSE."
      )
    )
  }
}

#' @export
bake.check_class <- function(object, new_data, ...) {
  col_names <- names(object$class_list)
  check_new_data(col_names, object, new_data)

  mapply(
    bake_check_class_core,
    new_data[, col_names],
    object$class_list,
    col_names,
    MoreArgs = list(aa = object$allow_additional)
  )

  new_data
}

print.check_class <-
  function(x, width = max(20, options()$width - 30), ...) {
    title <- "Checking the class(es) for "
    print_step(names(x$class_list), x$terms, x$trained, title, width)
    invisible(x)
  }


#' @rdname tidy.recipe
#' @export
tidy.check_class <- function(x, ...) {
  if (is_trained(x)) {
    values <- vapply(
      unname(x$class_list),
      FUN = function(x) paste0(x, collapse = "-"),
      FUN.VALUE = character(1)
    )
    res <- tibble(terms = names(x$class_list), value = values)
  } else {
    term_names <- sel2char(x$terms)
    res <- tibble(terms = term_names, value = na_chr)
  }
  res$id <- x$id
  res
}