R/unwrap_fn.R

In unheadr: Handle Data with Messy Header Rows and Broken Values

#' Unwrap values and clean up NAs used as padding
#'
#' @param df A data frame with wrapped values and an inconsistent number of NA
#'   values used to as within-group padding.
#' @param groupingVar Name of the variable describing the observational units.
#' @param separator Character string defining the separator that will delimit
#'   the elements of the unwrapped value.
#'
#' @return A summarized tibble. Order is preserved in the grouping variable by
#'   making it a factor.
#'
#' @details This is roughly the opposite of `tidyr::separate_rows()`.
#'
#' @examples
#' data(primates2017_wrapped)
#' # using commas to separate elements
#' unwrap_cols(primates2017_wrapped, scientific_name, ", ")
#'
#' # separating with semicolons
#' df <- data.frame(
#'   ounits = c("A", NA, "B", "C", "D", NA),
#'   vals = c(1, 2, 2, 3, 1, 3)
#' )
#' unwrap_cols(df, ounits, ";")
#' @importFrom rlang :=
#' @export
unwrap_cols <- function(df, groupingVar, separator) {
  groupingVar <- dplyr::enquo(groupingVar)
  groupVarFctName <- rlang::as_name(groupingVar)

  dffilled <- tidyr::fill(df, !!groupingVar)
  dffilled_grpd <- dplyr::group_by(
    dffilled,
    !!groupVarFctName := forcats::fct_inorder(!!groupingVar)
  )
  dfsummarised <- dplyr::summarise_all(
    dffilled_grpd,
    ~ paste(na.omit(.), collapse = separator)
  )
  dplyr::mutate(
    dfsummarised,
    !!groupVarFctName := as.character(forcats::fct_inorder(!!groupingVar))
  )
}