R/LoadGoogleSourced.R

In sjbeckett/subregionalcovid19: Processing and mapping COVID-19 case data at subnational scales

#' LoadGoogleSourced
#'
#' @description Reads in subnational data for various regions in the Google COVID-19 dataset to calculate most recent estimate of per capita active COVID-19 cases. Use with LoadData() is recommended.
#'
#' @note
#' Data aggregated from local health resources in the Google COVID-19 Open Data datasets \url{https://goo.gle/covid-19-open-data}.
#'
#' @references
#' O. Wahltinez and others. 2020. COVID-19 Open-Data: curating a fine-grained, global-scale data repository for SARS-CoV-2. Work in progress \url{https://goo.gle/covid-19-open-data}.
#'
#' @return A simple feature returning the date of most recent data (DateReport), a unique region code (geoid), the region name (RegionName) and country name (Country), the number of active cases per capita (pInf) and the regions geometry (geometry).
#'
#' @examples
#' \dontrun{
#' GS <- LoadGoogleSourced()
#' }
#' @seealso [LoadData()]
#' @export
LoadGoogleSourced <- function() { # takes a long time to process.
  # @article{Wahltinez2020,
  #  author = "O. Wahltinez and others",
  #  year = 2020,
  #  title = "COVID-19 Open-Data: curating a fine-grained, global-scale data repository for SARS-CoV-2",
  #  note = "Work in progress",
  #  url = {https://goo.gle/covid-19-open-data},
  # }

  # list with recent subnational data as of 8th August 2021. Note other useable datasets may exist.
  LIST <- c("Argentina", "Colombia", "Afghanistan", "Mozambique")
  utils::data(list = c("geomArgentina", "geomAfghanistan", "geomColombia", "geomMozambique"), envir = environment())
  geomArgentina <- sf::st_as_sf(geomArgentina)
  geomAfghanistan <- sf::st_as_sf(geomAfghanistan)
  geomColombia <- sf::st_as_sf(geomColombia)
  geomMozambique <- sf::st_as_sf(geomMozambique)

  # index file
  INDEX <- vroom::vroom("https://storage.googleapis.com/covid19-open-data/v3/index.csv", col_types = vroom::cols(aggregation_level = vroom::col_double(), .default = vroom::col_character()), show_col_types = FALSE, progress = FALSE)
  KEYS <- c()
  LOCALES <- c()
  COUNTRY <- c()
  for (aa in 1:length(LIST)) {
    Inds <- which(INDEX$country_name == LIST[aa] & INDEX$aggregation_level == 1)
    KEYS <- c(KEYS, INDEX$location_key[Inds])
    LOCALES <- c(LOCALES, INDEX$subregion1_name[Inds])
    COUNTRY <- c(COUNTRY, rep(LIST[aa], length(Inds)))
  }

  DateReport <- c()
  CaseDiff <- c()
  Pop <- c()
  for (bb in 1:length(KEYS)) {
    DAT <- vroom::vroom(paste0("https://storage.googleapis.com/covid19-open-data/v3/location/", KEYS[bb], ".csv"), guess_max = 1000, show_col_types = FALSE, progress = FALSE)
    naIND <- which(is.na(DAT$cumulative_confirmed))
    DateReport[bb] <- as.character(max(DAT$date[-naIND]))
    curr <- DAT$cumulative_confirmed[which(DAT$date == as.Date(DateReport[bb]))]
    past <- DAT$cumulative_confirmed[which(DAT$date == as.Date(DateReport[bb]) - 14)]
    if (is.na(past)) { # if past is NA try interpolation.
      interpolated <- stats::approx(DAT$date[-naIND], DAT$cumulative_confirmed[-naIND], xout = DAT$date)
      past <- interpolated[[2]][which(DAT$date == as.Date(DateReport[bb]) - 14)]
    }
    CaseDiff[bb] <- (10 / 14) * (curr - past)
    if ("population" %in% names(DAT)) {
      Pop[bb] <- DAT$population[1]
    } else {
      Pop[bb] <- NA
    }
  }


  IndexTable <- data.frame(KEYS, LOCALES, Country = COUNTRY, DateReport, CaseDiff, Pop)

  # 2017 census
  IndexTable$Pop[IndexTable$KEYS == "MZ_A"] <- 1865976 # Niassa Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_A"] <- "Nassa"
  IndexTable$Pop[IndexTable$KEYS == "MZ_B"] <- 1911237 #            Manica Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_B"] <- "Manica"
  IndexTable$Pop[IndexTable$KEYS == "MZ_G"] <- 1446654 #             Gaza Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_G"] <- "Gaza"
  IndexTable$Pop[IndexTable$KEYS == "MZ_I"] <- 1496824 #       Inhambane Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_I"] <- "Inhambane"
  IndexTable$Pop[IndexTable$KEYS == "MZ_L"] <- 2507098 #           Maputo Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_L"] <- "Maputo"
  IndexTable$Pop[IndexTable$KEYS == "MZ_MPM"] <- 1101170 #                    Maputo Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_MPM"] <- "Maputo City"
  IndexTable$Pop[IndexTable$KEYS == "MZ_N"] <- 6102867 #          Nampula Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_N"] <- "Nampula"
  IndexTable$Pop[IndexTable$KEYS == "MZ_P"] <- 2333278 #     Cabo Delgado Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_P"] <- "Cabo Delgado"
  IndexTable$Pop[IndexTable$KEYS == "MZ_Q"] <- 5110787 #         Zambezia Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_Q"] <- "Zambezia"
  IndexTable$Pop[IndexTable$KEYS == "MZ_S"] <- 2221803 #                    Sofala Mozambique
  IndexTable$Pop[IndexTable$KEYS == "MZ_T"] <- 2764169 #             Tete Province Mozambique
  IndexTable$LOCALES[IndexTable$KEYS == "MZ_T"] <- "Tete"

  # 2015 census
  IndexTable$Pop[IndexTable$KEYS == "AF_JOW"] <- 540255 #   Jowzjan Afghanistan
  IndexTable$LOCALES[IndexTable$KEYS == "AF_JOW"] <- "Jawzjan"
  IndexTable$Pop[IndexTable$KEYS == "AF_KAN"] <- 1226593 #  Kandahar Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_KHO"] <- 574582 #    Khost Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_LOG"] <- 392045 #     Logar Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_NAN"] <- 1517388 # Nangarhar Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_PAN"] <- 371902 # Panjshir Afghanistan
  IndexTable$LOCALES[IndexTable$KEYS == "AF_PAN"] <- "Panjsher"
  IndexTable$Pop[IndexTable$KEYS == "AF_SAM"] <- 387928 #  Samangan Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_TAK"] <- 983336 #    Takhar Afghanistan
  IndexTable$Pop[IndexTable$KEYS == "AF_URU"] <- 386818 #  Urozgan Afghanistan
  IndexTable$LOCALES[IndexTable$KEYS == "AF_URU"] <- "Uruzgan"
  IndexTable$Pop[IndexTable$KEYS == "AF_ZAB"] <- 304126 #     Zabul Afghanistan


  IndexTable$LOCALES[IndexTable$KEYS == "AF_WAR"] <- "Maydan Wardak"
  IndexTable$LOCALES[IndexTable$KEYS == "AF_HEL"] <- "Hilmand"
  IndexTable$LOCALES[IndexTable$KEYS == "AF_HER"] <- "Hirat"
  IndexTable$LOCALES[IndexTable$KEYS == "AF_NIM"] <- "Nimroz"
  IndexTable$LOCALES[IndexTable$KEYS == "AF_SAR"] <- "Sari Pul"
  IndexTable$LOCALES[IndexTable$KEYS == "AF_PIA"] <- "Paktya"



  IndexTable$pInf <- IndexTable$CaseDiff / IndexTable$Pop
  IndexTable$MATCH <- paste(IndexTable$LOCALES, IndexTable$Country, sep = ", ")

  # geomArgentina = st_read("https://github.com/deldersveld/topojson/raw/master/countries/argentina/argentina-provinces.json")
  # data("geomArgentina")
  geomArgentina$micro_name[which(geomArgentina$micro_name == "Buenos Aires")] <- "Buenos Aires Province"
  geomArgentina$micro_name[which(geomArgentina$micro_name == "Ciudad de Buenos Aires")] <- "City of Buenos Aires"
  geomArgentina$RegionName <- paste(geomArgentina$micro_name, geomArgentina$country_name, sep = ", ")


  # geomColombia = st_read("https://www.acolgen.org.co/wp-content/uploads/geo-json/colombia.geo.json")
  # data("geomColombia")

  geomColombia$micro_name <- stringr::str_to_title(geomColombia$micro_name)
  geomColombia$micro_name[2] <- LOCALES[26]
  geomColombia$micro_name[3] <- LOCALES[27]
  geomColombia$micro_name[4] <- LOCALES[28]
  geomColombia$micro_name[5] <- LOCALES[29]
  geomColombia$micro_name[7] <- LOCALES[31]
  geomColombia$micro_name[10] <- LOCALES[34]
  geomColombia$micro_name[12] <- LOCALES[36]
  geomColombia$micro_name[18] <- LOCALES[42]
  geomColombia$micro_name[19] <- LOCALES[43]
  geomColombia$micro_name[29] <- LOCALES[54]
  geomColombia$micro_name[31] <- LOCALES[56]
  geomColombia$micro_name[33] <- LOCALES[52]
  geomColombia$RegionName <- paste(geomColombia$micro_name, geomColombia$country_name, sep = ", ")


  # geomAfghanistan = st_read("https://gist.github.com/notacouch/246dcbb684571b8dff41fc3ed325972f/raw/5e1f3c66b4213fb0424743cd1eb036e1c8c7fb23/afghanistan_provinces_geometry--cities-demo.json")
  # data("geomAfghanistan")
  geomAfghanistan$RegionName <- paste(geomAfghanistan$micro_name, geomAfghanistan$country_name, sep = ", ")

  # geomBangladesh = st_read("https://github.com/mapmeld/mro-map/raw/master/bangladesh-divisions.geojson")

  # geomMozambique = st_read("https://geonode.ingc.gov.mz/geoserver/ows?service=WFS&version=1.0.0&request=GetFeature&typename=geonode%3Amoz_adm1&outputFormat=json&srs=EPSG%3A4326&srsName=EPSG%3A4326")
  # data("geomMozambique")
  geomMozambique$RegionName <- paste(geomMozambique$micro_name, geomMozambique$country_name, sep = ", ")

  geo <- dplyr::bind_rows(geomArgentina, geomColombia, geomAfghanistan, geomMozambique)

  GoogleMap <- dplyr::inner_join(geo, IndexTable, by = c("RegionName" = "MATCH"))
  # GoogleMap$RegionName = GoogleMap$MATCH

  GOOGLE_DATA <- subset(GoogleMap, select = c("DateReport", "geoid", "RegionName", "Country", "pInf", "geometry"))

  return(GOOGLE_DATA)
}