R/munge_unaids.R

In USAID-OHA-SI/mindthegap: Mind the Gap

#' Cleaning UNAIDS Data
#'
#' Deprecated. This function fetches and cleans UNAIDS Estimates/Test and Treat Data
#'
#' @param return_type Returns either 'HIV Estimates' or 'HIV Test & Treat' Data
#' @param indicator_type Returns either 'Integer' or 'Percent' indicator values

#' @return df
#' @export
#'
#' @examples
#'  \dontrun{
#'    munge_unaids(return_type = "HIV Test & Treat", indicator_type = "Percent")
#' }
#'
munge_unaids <- function(return_type, indicator_type) {

  lifecycle::deprecate_warn("2.0.0", "munge_unaids()", "munge_edms()",
                             details = "Data should be pulled from the UNAIDS EDMS Database rather than the Excel file posted to AIDSInfo. This function has stopped being maintained and may not provide valid results.")

  if (!requireNamespace('googlesheets4', quietly = TRUE))
    cli::cli_abort("{.pkg googlesheets4} needed for this function to work. Please install it.")

  # Google Sheet ID to original
  #sheet_id <- googledrive::as_id("1tkwP532mPL_yy7hJuHNAHaZ1_K_wd7zo_8AjeOe7fRs")

  #sheet_id_names <- "1vaeac7hb7Jb6RSaMcxLXCeTyim3mtTcy-a1DQ6JooCw"

  #Read Data from googlesheet and validate columns
  gdrive_df <- read_rename(return_type)

  #Munge
  gdrive_df_clean <-
    gdrive_df %>%
    #dplyr::mutate(dplyr::across(dplyr::contains("_"), ~gsub(" |<|>", "", .))) %>%
    #dplyr::mutate(dplyr::across( dplyr::contains("_"), as.numeric)) %>%
    dplyr::mutate(region = ifelse(country %in% regions, country, NA)) %>%
    tidyr::fill(region) %>% #get regions column
    tidyr::pivot_longer(-c(year, iso, country, region),
                        names_to = c("indicator")) %>%
    tidyr::separate(indicator, sep = "_", into = c("indicator", "age", "sex", "stat")) %>%
    tidyr::pivot_wider(names_from = 'stat', values_from = "value") %>%
    dplyr::mutate(estimate_flag = ifelse(stringr::str_detect(est, "<|>"), TRUE, FALSE)) %>% #estimate flag
    dplyr::mutate(dplyr::across(c(est:high), ~gsub(" |<|>", "", .))) %>% #replace special characters
    dplyr::mutate(dplyr::across(c(est:high), ~ gsub("m","00000", .x))) %>% #replace unit values
    dplyr::mutate(dplyr::across(c(est:high), ~ ifelse(grepl("\\.\\d+00000$", .x), gsub("\\.", "", .x), .x)))

  #Add sheet and indicator type variable
  gdrive_df_clean <- gdrive_df_clean %>%
    dplyr::mutate(sheet = return_type,
                  sex = ifelse(indicator == "pmtct", "female", sex),
                  indic_type = dplyr::case_when(
                    indicator %in% c("prev", "incidence", # HIV Estimates indicators
                                     "knownstatus", "plhivOnArt", "knownstatusOnArt",
                                     "plhivVLS", "onArtVLS", "pmtctArtPct" #T&T indicaotrs
                    ) ~ "percent_indics",
                    TRUE ~ "integer_indics"
                  )
    )

  #Recode indicators and rename columns
  gdrive_df_clean <- gdrive_df_clean %>%
    dplyr::mutate(indicator = dplyr::recode(indicator,
                                            "prev" = "Percent Prevalence",
                                            "deaths" = "Number AIDS Related Deaths",
                                            "plhiv" = "Number PLHIV",
                                            "incidence" = "Percent Incidence",
                                            "pmtct" = "Number PMTCT Needing ART",
                                            "newhiv" = "Number New HIV Infections",
                                            "knownstatus" = "Percent Known Status of PLHIV", #T&T indicators
                                            "plhivOnArt" = "Percent on ART of PLHIV",
                                            "knownstatusOnArt" = "Percent on ART with Known Status",
                                            "plhivVLS" = "Percent VLS of PLHIV",
                                            "onArtVLS" = "Percent VLS on ART",
                                            "knownstatusNum" = "Number Known Status of PLHIV",
                                            "onArtNum" = "Number on ART of PLHIV",
                                            "vlsNum" = "Number VLS of PLHIV",
                                            "pmtct" = "PMTCT", #what to call this
                                            "pmtctArt" = "Number PMTCT on ART",
                                            "pmtctArtPct" = "Percent PMTCT on ART")) %>%
    dplyr::rename(estimate = est,
           lower_bound = low,
           upper_bound = high)

  #add PEPFAR grouping category
  gdrive_df_clean <-  pepfar %>%
    dplyr::select(countryname = country_pepfar, iso = iso3) %>%
    dplyr::left_join(gdrive_df_clean, ., by = "iso") %>%
    dplyr::mutate(country = ifelse(is.na(countryname), country, countryname),
                  pepfar = ifelse(is.na(countryname), FALSE, TRUE)) %>%
    dplyr::select(-countryname)

#Export final df
  final_df <- suppressWarnings(
    gdrive_df_clean %>%
      dplyr::mutate(across(estimate:upper_bound, ~as.numeric(.x)),
                    indic_type =stringr::str_remove(indic_type, "_indics") %>% stringr::str_to_title(),
                    across(estimate:upper_bound, ~dplyr::case_when(indic_type == "Integer" ~ round(.x),
                                                                   indic_type == "Percent" ~ .x)),
                    across(age:sex, ~stringr::str_to_title(.x))) %>%
      dplyr::filter(indic_type == indicator_type)

  )

  return(final_df)

}