analysis/data/raw_data/server_results/archive/ecdc/20200627-230215-c5fcb17b/script.R

In mrc-ide/global-lmic-meffs: Analysis of LMIC COVID-19 Meffs

date <- as.Date(date, "%Y-%m-%d")
if (is.na(date)) {
  stop("Date must be provided in ISO format (i.e., YYYY-MM-DD)")
}

fmt <- "https://www.ecdc.europa.eu/sites/default/files/documents/COVID-19-geographic-disbtribution-worldwide-%s.xlsx"
date_iso <- as.character(date)
url <- sprintf(fmt, date_iso)

url_page <- "https://www.ecdc.europa.eu/en/publications-data/download-todays-data-geographic-distribution-covid-19-cases-worldwide"
tryCatch({
  code <- download.file(url, "ecdc.xlsx", mode = "wb")
  if (code != 0) {
    stop("Error downloading file")
  }
},
error = function(e) {
  stop(sprintf("Error downloading file '%s': %s, please check %s",
               url, e$message, url_page))
})



d <- readxl::read_excel("ecdc.xlsx", progress = FALSE)
names(d)[1] <- "dateRep"

names(d)[names(d) %in% c("Countries and territories", "countriesAndTerritories")] <- "Region"

# sort out the date malarkey that happened on the 16th May
if(is.na(as.Date(d$dateRep[1], "%Y-%m-%d"))) {
  d$dateRep <- as.Date(d$dateRep, format = "%d/%m/%Y")
}

# Data corrections

# fix panama's negative deaths
d[which(d$countryterritoryCode=="PAN" & as.Date(d$dateRep)=="2020-06-04"),]$deaths <-
  d[which(d$countryterritoryCode=="PAN" & as.Date(d$dateRep)=="2020-06-04"),]$deaths +
  d[which(d$countryterritoryCode=="PAN" & as.Date(d$dateRep)=="2020-06-03"),]$deaths 
d[which(d$countryterritoryCode=="PAN" & as.Date(d$dateRep)=="2020-06-03"),]$deaths <- 0

# fix italy's negative deaths and missing date
d$deaths[d$countryterritoryCode == "ITA" & d$deaths<0] <- 0

# fix spain's negative deaths and missing date
d$deaths[d$countryterritoryCode == "ESP" & d$deaths<0] <- 0

# spain seems to have stopped reporting now too...
esp_miss <- unique(d$dateRep)[which(!unique(d$dateRep) %in% d$dateRep[d$countryterritoryCode=="ESP"])]
if(length(esp_miss) > 0) {
  df_esp <- data.frame("dateRep" = esp_miss, 
                       "day" = as.numeric(format(as.Date(esp_miss), "%d")),
                       "month" = as.numeric(format(as.Date(esp_miss), "%m")),
                       "year" = as.numeric(format(as.Date(esp_miss), "%Y")),
                       "cases" = 0, 
                       "deaths" = 0,
                       "Region" = "Spain", 
                       "geoId" = "ES",
                       "countryterritoryCode" = "ESP",
                       "popData2018" = 46723749,
                       "continentExp" = "Europe", 
                       stringsAsFactors = FALSE)
  colnames(df_esp) <- colnames(d)
  d <- rbind(df_esp, d)
}

# save 
saveRDS(d, "ecdc_all.rds")