R/make_mdb_df.R

In USAID-OHA-SI/selfdestructin5: Creates SI OHA Mission Director Briefers

#' Create a base table of MDB indicators stacked long 
#' 
#' @description 
#' Makes a single table of core MDB indicators for all OUs, countries and USAID.
#' Uses [fetch_indicators()] to filter indicators and resolves known issues as a default.
#' Output from this function feeds into [reshape_mdb_df()]
#' 
#' @param df data frame from which MDB tables will be constructed for core indicators
#' @param resolve_issues logical that fetches troublesome mechs and omits them from df
#' 
#' @return data frame 
#' @export
#' @seealso [reshape_mdb_df()] to reshape into gt ready data frame;
#'   [fetch_indicators()] to filter indicators
#'   
#' @family data frame munging
#' 
#' 
#' @examples
#' \dontrun{
#'  mdb_df <- make_mdb_df(ou_im, resolve_issues = F)}
#'  
#'  

make_mdb_df <- function(df, resolve_issues = T) {
  
  pd <- fetch_metadata()
  
  # Check if pd exists and is a list
  if (missing(pd) || !is.list(pd)) {
    stop(glue::glue("The 'pd' parameter is missing. Use set_metadata() to create it."))
  }
  
  if (resolve_issues == TRUE) {
    df <- gophr::resolve_knownissues(df)
  } 
  
 
  
  # Get the indicator info you need 
  indicators <- fetch_indicators()
  indicator_fltr <- indicators %>% 
    dplyr::distinct(indicator) %>% 
    dplyr::pull()

  # Group by columns
  group_base <- c("fiscal_year", "agency", "indicator", "operatingunit")
  group_base_cntry <- c(group_base, "country")
  
  # Create OU dataframe to be bound with regional and USAID only
  df_ou <- df %>% 
    dplyr::filter(operatingunit %in% c(unique(glamr::pepfar_country_list$operatingunit), "Minoria")) %>% 
    collapse_base_tbl(indicator_fltr, group_base) %>% 
    label_aggregation(type = "OU")
  
  # Create regional
  df_reg <- df %>% 
    dplyr::filter(stringr::str_detect(operatingunit, "Region")) %>% 
    collapse_base_tbl(indicator_fltr, group_base_cntry) %>% 
    label_aggregation(type = "Regional") %>% 
    dplyr::mutate(agency = as.character(agency)) # in case there is no region, coerce agency to a character
  
  # Create USAID dataframe
  df_usaid <- df %>% 
    dplyr::mutate(operatingunit = ifelse(funding_agency == "USAID", "USAID", "ALL OTHER AGENCIES")) %>% 
    collapse_base_tbl(indicator_fltr, group_base) %>% 
    label_aggregation(type = "Agency") 
  
  # Bind the tables, set indicator levels
  md_table <- dplyr::bind_rows(df_ou, df_reg, df_usaid) %>% 
    dplyr::left_join(indicators %>% dplyr::select(indicator, indicator_plain)) %>% 
    dplyr::mutate(indicator = forcats::fct_relevel(indicator, indicator_fltr))
  
  return(md_table)
}