R/wide_to_long.R

In InseeFrLab/rtauargus: Using Tau-Argus from R

#' Convert Metadata from Wide to Long Format
#'
#' This function transforms a metadata data frame from a wide format to a long format.
#' The goal is to consolidate spanning variables into a single column, making the table
#' easier to handle regardless of the number of spanning variables described.
#'
#' @param df_metadata A data frame in wide format. It must contain columns starting
#'   with "spanning" and "hrc_spanning", representing spanning variables and their
#'   hierarchical counterparts.
#'
#' @return A data frame in long format. The resulting data frame contains two columns:
#'   - `spanning`: The consolidated values of the spanning variables.
#'   - Any additional columns retained from the original data frame.
#'   Rows where the names of "spanning" and "hrc_spanning" do not match, or where
#'   `spanning` values are NA, are filtered out.
#'
#' @examples
#' \dontrun{
#' data(metadata_pizza_lettuce)
#'
#' # View the structure of the original data
#' str(metadata_pizza_lettuce)
#'
#' # Convert to long format
#' metadata_pizza_lettuce_long <- wide_to_long(metadata_pizza_lettuce)
#'
#' # View the structure of the resulting data
#' str(metadata_pizza_lettuce_long)
#' }
#'
wide_to_long <- function(df_metadata){
  df_metadata_long <- df_metadata %>%
    pivot_longer(cols = starts_with("spanning"),
                 names_to = "spanning_name",
                 values_to = "spanning") %>%
    pivot_longer(cols = starts_with("hrc_spanning"),
                 names_to = "hrc_spanning_name",
                 values_to = "hrc_spanning") %>%
    filter(sub("spanning", "", spanning_name) == sub("hrc_spanning", "", hrc_spanning_name)) %>%
    select(-spanning_name, -hrc_spanning_name) %>%
    filter(!is.na(spanning))
  return(df_metadata_long)
}