data-raw/africa_pr.R

In njtierney/yahtsee: Yet Another Hierachical Time Series Extension and Expansion

## code to prepare `africa_pr` dataset goes here
library(malariaAtlas)
library(clock)
library(tidyverse)
library(yahtsee)
library(tsibble)

# getPR(continent = "Africa", species = "Pf") %>% autoplot()

malaria_africa_raw <- getPR(continent = "Africa", species = "Pf") %>%
  as_tibble()

malaria_africa_tidy <- malaria_africa_raw %>%
  filter(
    year_start == year_end,
    month_start == month_end
  ) %>%
  select(
    -year_end,
    -month_end
  ) %>%
  rename(
    year = year_start,
    month = month_start
  ) %>%
  relocate(
    country,
    country_id,
    continent_id,
    year,
    month
  ) %>%
  mutate(
    date = date_build(
      year = year,
      month = month
    ),
    .after = month
  ) %>%
  arrange(
    country,
    date
  )

# subset down to one measurement per country per year per month
malaria_africa_reduced <- malaria_africa_tidy %>%
  select(
    -dhs_id,
    -site_id,
    -site_name,
    -rdt_type,
    -latitude,
    -longitude,
    -malaria_metrics_available,
    -location_available,
    -permissions_info,
    -citation1,
    -citation2,
    -citation3,
    -rural_urban,
    -pcr_type
  )

malaria_africa_extra_cols <- malaria_africa_reduced %>%
  select(
    country,
    country_id,
    continent_id,
    date,
    species
  ) %>%
  distinct() %>%
  arrange(country, date)

malaria_africa_summarised <- malaria_africa_reduced %>%
  group_by(
    country_id,
    date
  ) %>%
  summarise(
    total_positive = sum(positive),
    total_examined = sum(examined),
    avg_lower_age = mean(lower_age),
    avg_upper_age = mean(upper_age),
    pr = total_positive / total_examined
  ) %>%
  right_join(malaria_africa_extra_cols,
    by = c(
      "country_id",
      "date"
    )
  ) %>%
  relocate(continent_id,
    country,
    .before = country_id
  ) %>%
  mutate(
    year = get_year(date),
    month = get_month(date),
    .after = date
  )
# sum positive
# sum examined
# recompute the PR, based on the sum of positive and examined

malaria_africa_ts <- as_tsibble(
  x = malaria_africa_summarised,
  key = country,
  index = date
) %>%
  # who_regions data is exported in this package
  left_join(who_regions,
    by = c(
      "country_id" = "country_iso_code",
      "country"
    )
  ) %>%
  dplyr::ungroup() %>%
  relocate(who_region, who_subregion, .before = country) %>%
  mutate_if(is.character, as.factor) %>%
  tidyr::drop_na() %>%
  # as_tibble() %>%
  # add_group_id(date) %>%
  mutate(month_num = 12 * (year - min(year) + month)) %>%
  select(who_region,
         who_subregion,
         country,
         date,
         month_num,
         positive = total_positive,
         examined = total_examined,
         pr,
         avg_lower_age,
         everything())
  # tsibble::as_tsibble(key = country,
  #                     index = .date_id)

usethis::use_data(malaria_africa_ts, overwrite = TRUE)