R/add_n_range_to_chapter_structure.R

In saros.base: Base Tools for Semi-Automatic Reporting of Ordinary Surveys

#' Add Sample Size Range to Chapter Structure
#'
#' Takes `chapter_structure` and `data` and returns the `chapter_structure` with
#' an attached variable containing a string with the sample size-range (or
#' single value if min=max). Allows specifying the glue_template_1 (single value)
#' and glue_template_2 (for min and max values).
#'
#'
#' @param chapter_structure A grouped tibble. If not grouped, will give a warning
#' and continue with rowwise processing, which is unlikely what you want.
#'
#' @param data The raw data, with matching column names as in
#' `chapter_structure$.variable_name_dep`.
#'
#' @param glue_template_1,glue_template_2 Glue templates.
#' @param variable_name String, name of new variable to attach.
#' Defaults to ".n_range"
#'
#' @return chapter_structure with a new variable added. Grouped as before.
#' @keywords internal
#'
add_n_range_to_chapter_structure <-
  function(chapter_structure, data,
           glue_template_1 = "{n}",
           glue_template_2 = "[{n[1]}-{n[2]}]",
           variable_name = ".n_range") {
    chapter_structure |>
      dplyr::group_map(.keep = TRUE, .f = ~ {
        deps <- as.character(unique(.x$.variable_name_dep))
        deps <- deps[!is.na(deps)]
        n <-
          lapply(deps, function(v) {
            length(data[[v]][!is.na(data[[v]])])
          }) |>
          unlist()
        if (!is.null(n)) {
          n <-
            range(n, na.rm = TRUE) |>
            unique()
          if (all(is.na(n))) {
            .x[[variable_name]] <- 0
          } else {
            template <- if (length(n) == 1) glue_template_1 else glue_template_2
            tryCatch(.x[[variable_name]] <- glue::glue(template),
              error = function(cnd) glue_err(cnd = cnd, arg_name = "glue_template_*")
            )
          }
        }
        .x
      }) |>
      dplyr::bind_rows() |>
      dplyr::grouped_df(dplyr::group_vars(chapter_structure))
  }