Code for comparing danish covid deaths with others deaths

gistfile1.txt

library(tidyverse)
library(lubridate)

# JH Data

newnames <- c("province", "country", "lat", "lon")

url_global_deaths = "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_global.csv"

global_deaths_raw <- read_csv(url_global_deaths)

names(global_deaths_raw)[1:4] <- newnames

global_deaths <- global_deaths_raw %>%
  gather(date, deaths, -newnames) %>%
  mutate(date = lubridate::mdy(date)) %>%
  group_by(country, province, lat, lon) %>%
  arrange(date) %>%
  mutate(new_deaths = deaths - lag(deaths)) %>%
  ungroup()

deaths <- global_deaths %>%
  filter(country == "Denmark", is.na(province)) %>%
  select(date, covid = new_deaths)

# SST Data
url <- "https://api.statbank.dk/v1/data/DODC1/CSV?Tid=*"

data_raw <- read_csv2(url)

data = data_raw %>%
  set_names(c("date", "deaths")) %>%
  mutate(date = date %>% str_replace_all("M|D", "-"),
         date = date %>% ymd(),
         y = year(date),
         m = month(date),
         d = day(date)) %>%
  group_by(y) %>%
  mutate(doy = 1:n()) %>%
  ungroup()

# Combine data

pd = data %>%
  left_join(deaths) %>%
  mutate(deaths_no_covid = ifelse(is.na(covid), deaths, deaths - covid)) %>%
  mutate(covid_rm = c(rep(NA, 3),  # Adding a rolling mean
                      zoo::rollmean(x = covid, k = 7), 
                      rep(NA, 3))) %>%
  mutate(deaths_no_covid_rm = c(rep(NA, 3),  # Adding a rolling mean
                                zoo::rollmean(x = deaths_no_covid, k = 7), 
                                rep(NA, 3)))

death_minmax <- pd %>% 
  filter(y != 2020) %>% 
  group_by(doy) %>%
  summarise(mindeaths = min(deaths_no_covid_rm, na.rm = T),
            maxdeaths = max(deaths_no_covid_rm, na.rm = T),
            avgdeaths = mean(deaths_no_covid_rm, na.rm = T)) %>%
  ungroup()
  
bardata = pd %>% 
  filter(y == 2020) %>% 
  select(doy, covid_rm, deaths_no_covid_rm)


  # gather(type, count, -doy) %>%
  # drop_na()

# Plot the data
ggplot() + 
  geom_ribbon(data = death_minmax, aes(x = doy, ymin = 0, ymax = maxdeaths), fill = "lightgrey") + 
  geom_ribbon(data = bardata, aes(x = doy, ymin = 0, ymax = deaths_no_covid_rm), fill = "blue") +
  geom_ribbon(data = bardata, aes(x = doy, ymin = deaths_no_covid_rm, ymax = deaths_no_covid_rm + covid_rm), fill = "red") +
  # geom_line(data = death_minmax, aes(x = doy, y = mindeaths)) + 
  # geom_line(data = death_minmax, aes(x = doy, y = maxdeaths)) + 
  geom_line(data = death_minmax, aes(x = doy, y = avgdeaths), color = "white", linetype = "dashed") + 
  theme_minimal() +
  labs(y = "Døde per dag i Danmark",
       x = "Årets dage",
       caption = "Kilde: Danmarks Statistik (tabel: DODC1) og Johns Hopkins University (https://github.com/CSSEGISandData/COVID-19)",
       title = "Døde per dag i Danmark",
       subtitle = "
       Den bagvedliggende grå farve er det maksimale dødstal registreret på den pågældende dag siden 2007.
       Den blå farve er antal døde i Danmark i 2020, der IKKE er døde 'med COVID'.
       Den røde farve er antal døde i Danmark i 2020, der er døde 'med COVID'.
       Den hvide linje er gennemsnitsdødstallet på den pågældende dag siden 2007.")

h = 5
w = h * 1.61803398875

ggsave("deaths_dk2.png", width = w, height = h)