R/recalculate_concerted_cols.R

In mhkhan27/illuminate: Weighted survey analysis

#' This function recreates the concerted columns for select multiple questions
#' @param df data frame
#' @param uuid Unique ID column(Default is X_uuid).
#' @export

recalculate_concerted_col_for_select_multiple <- function(df, uuid = "X_uuid"){


  names(df) <- sub("(\\..*?)\\.", "\\1_", names(df))
  names(df) <- sub("(\\..*?)\\.", "\\1_", names(df))
  cols_order <- df %>% names()

  select_multiple <- auto_sm_parent_child(df)
  if(nrow(select_multiple) > 0){
    select_multiple_list <- list()
    for (i in select_multiple$sm_parent) {
      select_multi_single <- select_multiple %>% filter(sm_parent == i)
      concat_col <- select_multi_single$sm_parent %>% unique()
      choice_cols <- select_multi_single$sm_child %>% unique()

      df_only_cols <- df %>% select(choice_cols,uuid)

      pivot_long <-  df_only_cols %>% mutate_at(names(df_only_cols),as.character)

      final_df <- pivot_long %>% pivot_longer(cols = !uuid,names_to = "cols" ,values_to = "value") %>%
        filter(value == 1 | value == TRUE | value == "1" | value == "TRUE") %>% group_by(!!sym(uuid)) %>% summarise(
          !!sym(concat_col):= paste0(cols,collapse = ",")
        )

      final_df[[concat_col]] <- final_df[[concat_col]] %>% str_replace_all(paste0(concat_col,"."),"")

      select_multiple_list[[concat_col]] <- final_df
    }

    final_df_for_export<- purrr::reduce(select_multiple_list, dplyr::full_join, by = uuid)
    concat_col_names_from_fina_export <-  final_df_for_export %>% select(-uuid) %>% names()

    data_with_fix_concat <- df %>% select(-concat_col_names_from_fina_export) %>%
      left_join(final_df_for_export) %>% select(cols_order)

    return(data_with_fix_concat)
  }

  if(nrow(select_multiple)== 0){
    return(df)
  }

}