Code to create COVID-19 attributed death trajectories by country using JHU data

coronaDeaths.R

# remotes::install_github('ramikrispin/coronavirus')
library(coronavirus)
update_datasets()
pacman::p_load(char=c('tidyverse','janitor', 'ggrepel'))

deaths <- coronavirus %>% 
  clean_names() %>% 
  filter(type=='death') %>% 
  group_by(country_region, date) %>% 
  summarise(cases = sum(cases)) %>% 
  ungroup() %>% 
  group_by(country_region) %>% 
  arrange(date) %>% 
  mutate(cum_cases = cumsum(cases)) %>% 
  ungroup() %>% 
  as_tibble()

top_ten <- deaths %>% 
  filter(date==max(date)) %>% 
  top_n(10, cum_cases)

deaths_top10 <- 
  deaths %>% 
  filter(country_region %in% top_ten$country_region) %>% 
  group_by(country_region) %>% 
  filter(cum_cases >= 10) %>% 
  mutate(days = as.numeric(date - min(date))) %>% 
  ungroup()

ggplot(deaths_top10, aes(x = days, y = cum_cases)) +
  geom_label_repel(data = deaths_top10 %>% 
                     filter(country_region %in% top_ten$country_region) %>% 
                     group_by(country_region) %>% 
                     filter(days == max(days)) %>% 
                     ungroup(),
                   aes(color = country_region, label = country_region),
                   hjust = 1, show.legend=F)+
  geom_point(data = deaths_top10 %>% 
               filter(country_region %in% top_ten$country_region) %>% 
               group_by(country_region) %>% 
               filter(days == max(days)) %>% 
               ungroup(),
             aes(color = country_region, size=cum_cases),
              show.legend=F)+
  geom_line(aes(color = country_region), show.legend = F)+
  scale_y_log10('Cumulative deaths')+
  labs(x = 'Days since 10th death')+
  theme_classic()