In FPRgroup/FPEMcountry: fpemlocal

knitr::opts_chunk$set(echo = TRUE)
devtools::load_all()

This script must run annually when new UNPD or track20 data is provided. Make sure to build package after!! At the end of the script All of the paths to data in this document will be replaced annually. Here is a list of datasets below which you may be replacing. You will probably want to replace the pre-existing raw data files as you go through this markdown. -contraceptive use data -track 20 data -population count data -geographic division data -global estimates -global parameters -fo2020 countries

UNPD contraceptive data

# Read and impute
contraceptive_use <- readr::read_csv(here::here("data-raw/contraceptive_use_2022-02-15.csv")) %>%
  fpemlocal::impute_packagedata_se() %>% 
  purrr::pluck("contraceptive_use_imputed") 
# Check the format compared to last year
# If the dev every intends to change the format going forward you need to edit the script
# data-raw/contraceptive_use_format.R which is fed to R/format_check
#fpemlocal::format_check(format_list = fpemlocal::contraceptive_use_format, 
#             data = contraceptive_use)
format_check(format_list = fpemlocal::contraceptive_use_format, 
             data = contraceptive_use)
# Save as package data if the format is correct
usethis::use_data(contraceptive_use, overwrite = TRUE)

track 20 data

# Read
contraceptive_use_track20 <- read.csv(here::here("data-raw/Track20Database121819_Fixed.csv"))
# check format
fpemlocal::format_check(format_list = fpemlocal::contraceptive_use_format, 
             data = contraceptive_use_track20)
usethis::use_data(contraceptive_use_track20, overwrite = TRUE)

Population count data

#read
population_counts <- readr::read_csv(here::here("data-raw/population_counts_2022-02-15.csv")) %>%
  #use of readdelimn here is to format as tibble
  dplyr::mutate(mid_year = floor(mid_year))
#check that format is the same
janitor::compare_df_cols(population_counts, fpemlocal::population_counts, return = "mismatch")
#save since it checks out
usethis::use_data(population_counts, overwrite = TRUE)

Geographic division data. Not sure if we update every year

# Note some name changes and hacks here

# Read the two datasets
data1 <- read.csv(here::here("data-raw/division_classifications.csv"))
divisions <-
  read.csv(here::here("data-raw/country_and_area_classification.csv"))
# CHange some names to package friendly names
divisions <- divisions %>%
  dplyr::rename(
    name_country = 'Country or area',
    name_sub_region = 'Region',
    name_region = "Major area"
  ) %>%
  dplyr::select(division_numeric_code,
                name_country,
                name_region,
                name_sub_region)
divisions <- left_join(divisions, data1)

#2020 HACKS
divisions <- fpemdata::divisions
divisions$is_unmarried_sexual_activity[which(divisions$name_country == "South Sudan")] <-
  "N"
divisions <- divisions %>%
  dplyr::filter(division_numeric_code != 732) # [1] "Western Sahara"
usethis::use_data(divisions, overwrite = TRUE)

Global estimates

mw_perc <- fpemlocal::read_rda(here::here("data-raw/globalrun_data_20220215/global_estimates_married.rda"))
uw_perc <- fpemlocal::read_rda(here::here("data-raw/globalrun_data_20220215/global_estimates_unmarried.rda"))
global_estimates_married <- mw_perc %>%
  dplyr::mutate(is_in_union = "Y")
global_estimates_unmarried <- uw_perc %>%
  dplyr::mutate(is_in_union = "N")
global_estimates <- dplyr::bind_rows(global_estimates_married,
                 global_estimates_unmarried) %>%
  tidyr::gather(year, value, "1970.5":"2030.5") %>% 
  dplyr::mutate(year = as.numeric(year)) %>%
  tidyr::spread(Percentile,  value) %>%
  dplyr::rename(division_numeric_code = Iso,
                indicator = par) %>%
  dplyr::rename_all(tolower) %>%
  dplyr::select(-name) %>%
  dplyr::mutate(indicator = dplyr::recode(indicator,
                                          unmet = "unmet_need_any",
                                          modern = "contraceptive_use_modern",
                                          traditional = "contraceptive_use_traditional")) %>%
  dplyr::rename("2.5%" = "0.025",
                "10%" = "0.1",
                "50%" = "0.5",
                "90%" = "0.9",
                "97.5%" = "0.975") %>%
  dplyr::mutate(model = "global")

#check is equal to current package version
janitor::compare_df_cols(global_estimates, fpemlocal::global_estimates, return = "mismatch")

# all checks out so lets save it
usethis::use_data(global_estimates, overwrite = TRUE)
rm(list = ls()) #clearing environment as it has too much data

Global parameters and global index data saved as internal Sysdata from processed from the global run

globalrun_input_m <- fpemlocal::read_rda(here::here("data-raw/globalrun_data_20220215/global_mcmc_meta_married.rda"))
globalrun_input_u <- fpemlocal::read_rda(here::here("data-raw/globalrun_data_20220215/global_mcmc_meta_unmarried.rda"))
globalrun_output_m <- fpemlocal::read_rda(here::here(path = "data-raw/globalrun_data_20220215/global_output_married.rda"))
globalrun_output_u <- fpemlocal::read_rda(here::here(path = "data-raw/globalrun_data_20220215/global_output_unmarried.rda"))

index_area <- fpemlocal::index_area(globalrun_input_m, run_type = "mwra")
index_datatype <- fpemlocal::index_datatype(globalrun_input_m)
index_m <- list(index_area_df = index_area,
                index_datatype = index_datatype)

index_area <- fpemlocal::index_area(globalrun_input_u, run_type = "uwra")
index_datatype <- fpemlocal::index_datatype(globalrun_input_u)
index_u <- list(index_area_df = index_area,
                index_datatype = index_datatype)


#compare all of these in some way? not sure what is best as they are lists
simple_list_check <- function(l1, l2) {
  namec <- names(l1) %in% names(l2) %>% all
  classc <- lapply(l1, class) %in% lapply(l2, class) %>% all
  if(namec & classc) {
    return("names and classes match")
  } else {
    return("names and classes dont match")
  }
}
simple_list_check(globalrun_input_m, fpemlocal:::globalrun_input_m)
simple_list_check(globalrun_input_m$winbugs.data, fpemlocal:::globalrun_input_m$winbugs.data)
simple_list_check(globalrun_input_u, fpemlocal:::globalrun_input_u)
simple_list_check(globalrun_output_m, fpemlocal:::globalrun_output_m)
simple_list_check(globalrun_output_u, fpemlocal:::globalrun_output_u)
simple_list_check(index_m, fpemlocal:::index_m)
simple_list_check(index_u, fpemlocal:::index_u)


usethis::use_data(index_m, index_u,
                  globalrun_input_m, 
                  globalrun_input_u,
                  globalrun_output_m,
                  globalrun_output_u, 
                  internal = TRUE,
                  overwrite= TRUE)

fp2020 countries

temp <- read.csv(here::here("data-raw/legacy.csv"))
fp2020 <- temp %>% filter(FP2020.country == "Yes") %>% select(division_numeric_code)
usethis::use_data(fp2020, overwrite = TRUE)

Build package

devtools::document()
devtools::build(vignettes = FALSE)

FPRgroup/FPEMcountry documentation built on April 24, 2023, 4:32 p.m.