R/getDHSindicator.R

In surveyPrev: Mapping the Prevalence of Binary Indicators using Survey Data in Small Areas

#' Process DHS data
#'
#' This function processes DHS data from getDHSdata function.
#'
#' @param Rdata Result from getDHSdata function, the raw DHS survry data from get_datasets.
#' @param indicator Indicator of interests.
#' @param FUN a function to process the DHS data into a binary indicator if not using one of the implemented indicators. See surveyPrev::AN_ANEM_W_ANY for an example function to obtain the indicator for women classified as having any anemia.
#' @return The function returns processed survey data that contains the indicator of interests.
#'
#' @importFrom naniar replace_with_na
#' @import dplyr
#' @import tidyverse
#' @importFrom sjlabelled set_label add_labels
#' @import labelled
#' @author Qianyu Dong
#' @examples
#' \dontrun{
#' dhsData1 <- getDHSdata(country = "Zambia",
#'                                  indicator = "ancvisit4+",
#'                                  year = 2018)
#' data1 <- getDHSindicator(dhsData1, indicator = "ancvisit4+")
#'
#'
#' # User-specified function to process the data
#' # For example see the internal function surveyPrev::AN_ANEM_W_ANY
#' dhsData2 <- getDHSdata(country = "Zambia",
#'                                  indicator = NULL,
#'                                  year = 2018)
#' data2 <- getDHSindicator(dhsData2, indicator = NULL,
#'                          FUN = surveyPrev::AN_ANEM_W_ANY)
#' # which should be identical to the following
#' dhsData3 <- getDHSdata(country = "Zambia",
#'                                  indicator = "womananemia",
#'                                  year = 2018)
#' data3 <- getDHSindicator(dhsData3, indicator = "womananemia")
#'
#' }
#'
#' @export
getDHSindicator <- function(Rdata, indicator = NULL, FUN = NULL) {
  Rdata<-Rdata

  if(is.null(indicator))
  {
    raw.dat.tmp <- FUN(Rdata)
  }
  else if(indicator == "unmet_family"||indicator == "FP_NADA_W_UNT"){
    # IRdata <- Rdata %>%
    #   mutate(wt = v005/1000000)
    #
    # IRdata <- IRdata %>%
    #   mutate(n_un_fam_plan =
    #            case_when(
    #              v626 == 1 ~ 1,
    #              v626 == 2 ~ 1,
    #              v626 == 3 ~ 0,
    #              v626 == 4 ~ 0,
    #              v626 %in% c(5, 6, 7, 8, 9) ~ NA)) %>%
    #   set_value_labels(n_un_fam_plan = c("Yes" = 1, "No"=0  )) %>%
    #   set_variable_labels(n_un_fam_plan = "Unmet need for family planning")
    #
    # raw.dat.tmp<-IRdata
    # colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'n_un_fam_plan'] <- "value"
    raw.dat.tmp <- fp_unmet_tot( Rdata)
  }
  else if(indicator == "FP_CUSA_W_MOD"){
    raw.dat.tmp <- fp_cruse_mod( Rdata)

  }

  else if(indicator == "nmr"||indicator == "CM_ECMR_C_NNR"){
    # BRdata <- Rdata %>%
    #   mutate(wt = v005/1000000)
    # BRdata$value<- ifelse(BRdata$b7==0, 1, 0)
    # BRdata$value[is.na( BRdata$value)] <- 0
    #
    # raw.dat.tmp=BRdata
    raw.dat.tmp <- NMR(Rdata)
  }

  # else if(indicator == "sanitation"){
  #   PRdata <- Rdata %>%
  #     mutate(wt = hv005/1000000)
  #   PRdata <- PRdata %>%
  #     mutate(n_improved_sani =
  #              case_when(
  #                # household using improved sanitation = 1
  #                hv205 %in% c(10, 11, 12, 13, 14, 15, 20, 21, 22, 41) ~ 1,
  #                # household using unimproved sanitation = 0
  #                hv205 %in% c(14, 23, 42, 43, 96, 31, 30) ~ 0)) %>%
  #     #replace_with_na(replace = list(n_improved_sani = c(99))) %>%
  #     set_value_labels(n_improved_sani = c("Yes" = 1, "No"=0  )) %>%
  #     set_variable_labels(n_improved_sani = "Household using improved sanitation")
  #
  #   raw.dat.tmp<-PRdata
  #   colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'n_improved_sani'] <- "value"
  #
  # }


  else if(indicator =="ancvisit4+"||indicator == "RH_ANCN_W_N4P"){
    # IRdata <- Rdata%>%
    #   mutate(wt = v005/1000000) %>%
    #   mutate(period = 60)
    #
    # # age of child. If b19_01 is not available in the data use v008 - b3_01
    # if ("TRUE" %in% (!("b19_01" %in% names(IRdata))))
    #   IRdata [[paste("b19_01")]] <- NA
    # if ("TRUE" %in% all(is.na(IRdata $b19_01)))
    # { b19_included <- 0} else { b19_included <- 1}
    #
    # if (b19_included==1) {
    #   IRdata <- IRdata %>%
    #     mutate(age = b19_01)
    # } else {
    #   IRdata <- IRdata %>%
    #     mutate(age = v008 - b3_01)
    # }
    # # //Number of ANC visits in 4 categories that match the table in the final report
    # IRdata <- IRdata %>%
    #   mutate(rh_anc_numvs =
    #            case_when(
    #              m14_1 == 0 ~ 0 ,
    #              m14_1 == 1 ~ 1 ,
    #              m14_1  %in% c(2,3)   ~ 2 ,
    #              m14_1>=4 & m14_1<=90  ~ 3 ,
    #              m14_1>90  ~ 9 ,
    #              age>=period ~ 99 )) %>%
    #   replace_with_na(replace = list(rh_anc_numvs = c(99))) %>%
    #   set_value_labels(rh_anc_numvs = c(none = 0, "1" = 1, "2-3"=2, "4+"=3, "don't know/missing"=9  )) %>%
    #   set_variable_labels(rh_anc_numvs = "Number of ANC visits")
    #
    # # //4+ ANC visits
    # IRdata <- IRdata %>%
    #   mutate(rh_anc_4vs =
    #            case_when(
    #              rh_anc_numvs==3 ~ 1,
    #              rh_anc_numvs %in% c(0,1,2,9)   ~ 0 )) %>%
    #   set_value_labels(rh_anc_4vs = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(rh_anc_4vs = "Attended 4+ ANC visits")
    # raw.dat.tmp<-IRdata
    # colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'rh_anc_4vs'] <- "value"
    # raw.dat.tmp <- raw.dat.tmp %>% filter( age<60)
    raw.dat.tmp <- rh_anc_4vs(Rdata)


  }
  else if(indicator == "RH_DELA_C_SKP"){
    # BRdata <- Rdata %>%
    #   mutate(wt = v005/1000000)
    # BRdata$value<- ifelse(BRdata$b7==0, 1, 0)
    # BRdata$value[is.na( BRdata$value)] <- 0
    #
    # raw.dat.tmp=BRdata
    raw.dat.tmp <- rh_del_pvskill(Rdata)
  }


  else if(indicator == "DPT3"||indicator == "CH_VACC_C_DP3"){
    # KRdata <- Rdata %>%
    #   mutate(wt = v005/1000000)
    #
    # # age of child. If b19 is not available in the data use v008 - b3
    # if ("TRUE" %in% (!("b19" %in% names(KRdata))))
    #   KRdata [[paste("b19")]] <- NA
    # if ("TRUE" %in% all(is.na(KRdata$b19)))
    # { b19_included <- 0} else { b19_included <- 1}
    #
    # if (b19_included==1) {
    #   KRdata <- KRdata %>%
    #     mutate(age = b19)
    # } else {
    #   KRdata <- KRdata %>%
    #     mutate(age = v008 - b3)
    # }
    #
    # # *** Two age groups used for reporting.
    # KRdata <- KRdata %>%
    #   mutate(agegroup =
    #            case_when(
    #              age>=12 & age<=23 ~ 1,
    #              age>=24 & age<=35 ~ 2  )) %>%
    #   set_value_labels(agegroup = c("12-23" = 1, "24-35"=2)) %>%
    #   set_variable_labels(agegroup = "age group of child for vaccination")
    #
    # # Selecting children
    # # Create subset of KRfile to select for children for VAC indicators
    # # Select agegroup 1 or agegroup 2
    # KRvac <- KRdata %>%
    #   subset(agegroup==1 & b5==1) # select age group and live children
    #
    # # *******************************************************************************
    #
    # # Source of vaccination information. We need this variable to code vaccination indicators by source.
    # KRvac <- KRvac %>%
    #   mutate(source =
    #            case_when(h1==1 ~ 1, h1==0 | h1==2 | h1==3 ~ 2  )) %>%
    #   set_value_labels(source = c("card" = 1, "mother"=2)) %>%
    #   set_variable_labels(source = "source of vaccination information")
    #
    # # *** Pentavalent ***
    # # //DPT 1, 2, 3 either source
    # KRvac <- KRvac %>%
    #   mutate(dpt1 = case_when(h3%in%c(1,2,3) ~ 1, h3%in%c(0,8) ~ 0  )) %>%
    #   mutate(dpt2 = case_when(h5%in%c(1,2,3) ~ 1, h5%in%c(0,8) ~ 0  )) %>%
    #   mutate(dpt3 = case_when(h7%in%c(1,2,3) ~ 1, h7%in%c(0,8) ~ 0  )) %>%
    #   mutate(dptsum = dpt1 + dpt2 + dpt3)
    # # This step is performed for multi-dose vaccines to take care of any gaps in the vaccination history.
    # # See DHS guide to statistics for further explanation
    # KRvac <- KRvac %>%
    #   mutate(ch_pent1_either = case_when(dptsum >=1 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent1_either = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent1_either = "Pentavalent 1st dose vaccination according to either source") %>%
    #   mutate(ch_pent2_either = case_when(dptsum >=2 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent2_either = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent2_either = "Pentavalent 2nd dose vaccination according to either source") %>%
    #   mutate(ch_pent3_either = case_when(dptsum >=3 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent3_either = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent3_either = "Pentavalent 3rd dose vaccination according to either source")
    #
    # # //DPT 1, 2, 3 mother's report
    # KRvac <- KRvac %>%
    #   mutate(ch_pent1_moth = case_when(dptsum >=1 & source==2~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent1_moth = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent1_moth = "Pentavalent 1st dose vaccination according to mother") %>%
    #   mutate(ch_pent2_moth = case_when(dptsum >=2 & source==2 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent2_moth = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent2_moth = "Pentavalent 2nd dose vaccination according to mother") %>%
    #   mutate(ch_pent3_moth = case_when(dptsum >=3 & source==2 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent3_moth = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent3_moth = "Pentavalent 3rd dose vaccination according to mother")
    #
    # # //DPT 1, 2, 3 by card
    # KRvac <- KRvac %>%
    #   mutate(ch_pent1_card = case_when(dptsum >=1 & source==1~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent1_card = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent1_card = "Pentavalent 1st dose vaccination according to card") %>%
    #   mutate(ch_pent2_card = case_when(dptsum >=2 & source==1 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent2_card = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent2_card = "Pentavalent 2nd dose vaccination according to card") %>%
    #   mutate(ch_pent3_card = case_when(dptsum >=3 & source==1 ~ 1, TRUE ~ 0  )) %>%
    #   set_value_labels(ch_pent3_card = c("Yes" = 1, "No"=0)) %>%
    #   set_variable_labels(ch_pent3_card = "Pentavalent 3rd dose vaccination according to card")
    #
    # raw.dat.tmp<-KRvac
    # colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'ch_pent3_either'] <- "value"
    raw.dat.tmp <- ch_pent3_either(Rdata)


  }
  else if(indicator== "CH_VACC_C_MSL") {

    raw.dat.tmp <- ch_meas_either(Rdata)

  }
  else if(indicator== "PCV3") {

    raw.dat.tmp <- ch_pneumo3_either(Rdata)

  }
  else if(indicator== "RotaC1") {

    raw.dat.tmp <- ch_rotav1_either(Rdata)

  }


  else if(indicator =="CH_VACC_C_DP1"){
    raw.dat.tmp <- ch_pent1_either(Rdata)
  }
  else if(indicator =="CH_VACC_C_BAS"){
    raw.dat.tmp <- ch_allvac_either(Rdata)
  }
  else if(indicator =="CH_VACC_C_NON"){
    raw.dat.tmp <- ch_novac_either(Rdata)
  }


  else if(indicator =="CH_DIAT_C_ORT"){
    raw.dat.tmp <- ch_diar_ors_rhf(Rdata)
  }

  else if(indicator == "wasting"||indicator=="CN_NUTS_C_WH2"){
  #   PRdata <- Rdata %>%
  #   mutate(wt = hv005/1000000)
  # PRdata <- PRdata %>%
  #   mutate(nt_ch_wast =
  #            case_when(
  #              hv103==1 &  hc72< -200  ~ 1 ,
  #              hv103==1 &  hc72>= -200 ~ 0 ,
  #              hc72>=9996 ~ 99)) %>%
  #   replace_with_na(replace = list(nt_ch_wast = c(99))) %>%
  #   set_value_labels(nt_ch_wast = c("Yes" = 1, "No"=0  )) %>%
  #   set_variable_labels(nt_ch_wast = "Wasted child under 5 years")
  #
  #   raw.dat.tmp<-PRdata
  #   colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'nt_ch_wast'] <- "value"
    raw.dat.tmp <- nt_ch_wast(Rdata)

  }
  else if(indicator == "stunting"||indicator=="CN_NUTS_C_HA2") {
  #   PRdata <- Rdata %>%
  #   mutate(wt = hv005/1000000)
  # PRdata <- PRdata %>%
  #   mutate(nt_ch_stunt =
  #            case_when(
  #              hv103==1 &  hc70< -200  ~ 1 ,
  #              hv103==1 &  hc70>= -200 ~ 0 ,
  #              hc70>=9996 ~ 99)) %>%
  #   replace_with_na(replace = list(nt_ch_stunt = c(99))) %>%
  #   set_value_labels(nt_ch_stunt = c("Yes" = 1, "No"=0  )) %>%
  #   set_variable_labels(nt_ch_stunt = "Stunted child under 5 years")
  #
  #   raw.dat.tmp<-PRdata
  #   colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'nt_ch_stunt'] <- "value"
    raw.dat.tmp <- nt_ch_stunt(Rdata)

  }
  else if (indicator == "womananemia"|| indicator == "AN_ANEM_W_ANY"){

  #
  # *** Anemia indicators ***
  #
  # //Any anemia
  # IRdata <- IRdata %>%
  #   mutate(nt_wm_any_anem =
  #            case_when(
  #              v042==1 & v457<4 ~ 1 ,
  #              v042==1 &  v455==0 ~ 0)) %>%
  #   set_value_labels(nt_wm_any_anem = c("Yes" = 1, "No"=0  )) %>%
  #   set_variable_labels(nt_wm_any_anem = "Any anemia - women")

  #  raw.dat.tmp<-IRdata
  #   colnames(raw.dat.tmp)[colnames(raw.dat.tmp) == 'nt_wm_any_anem'] <- "value"
  raw.dat.tmp <- nt_wm_any_anem(Rdata)

  }

  else if(indicator=="CN_BRFS_C_EXB"){

    raw.dat.tmp <- nt_ebf(Rdata)

  }
  else if(indicator=="CN_ANMC_C_ANY") {

    raw.dat.tmp <- nt_ch_any_anem(Rdata)

  }
  else if(indicator== "AN_NUTS_W_THN") {

    raw.dat.tmp <- nt_wm_thin(Rdata)

  }

  else if(indicator== "ML_NETP_H_IT2") {

    raw.dat.tmp <- ml_hhaccess(Rdata)

  }
  else if(indicator== "ML_PMAL_C_RDT") {

    raw.dat.tmp <- ml_test_rdtmal(Rdata)

  }

  else if(indicator== "HA_HIVP_B_HIV") {

    raw.dat.tmp <- hv_hiv_pos(Rdata)

  }


  else if(indicator== "WS_TLET_H_IMP"||indicator== "sanitation") {

    raw.dat.tmp <- ph_sani_improve(Rdata)

  }
  else if(indicator== "WS_TLET_P_BAS") {

    raw.dat.tmp <- ph_sani_basic(Rdata)

  }

  else if(indicator== "WS_SRCE_P_BAS") {

    raw.dat.tmp <- ph_wtr_basic(Rdata)

  }


  if("hv001" %in% colnames(raw.dat.tmp)){
    pre <- "h"
  }else if("v001" %in% colnames(raw.dat.tmp)){
    pre <- ""
  }
  strat <- attr(raw.dat.tmp[, paste0(pre, "v025")], which='labels')
  names(strat) <- tolower(names(strat))
  raw.dat.tmp[, paste0(pre, "v025")] <- ifelse(raw.dat.tmp[, paste0(pre, "v025")] == strat["urban"][[1]],'urban','rural')
  raw.dat.tmp[, paste0(pre, "v025")] <- factor(raw.dat.tmp[, paste0(pre, "v025")], levels = c('urban','rural'))
  raw.dat.tmp[, paste0(pre, "v024")] <- factor(labelled::unlabelled(raw.dat.tmp[, paste0(pre, "v024")]))
  dat.tmp<-  raw.dat.tmp %>%
      dplyr::  select(c(cluster= paste0(pre, "v001"),
                        householdID= paste0(pre, "v002"),
                        v024= paste0(pre, "v024"),
                        weight= paste0(pre, "v005"),
                        strata= paste0(pre, "v025"),
                        value="value"))


  return(dat.tmp)

}