covid

rm(list = ls())
library(tidyverse)
library(ggrepel)
covid <-
  read_csv(file = "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_19-covid-Confirmed.csv")

covid <-
  covid %>% filter(`Country/Region` != "Cruise Ship")

covid_long <-
  covid %>%
  pivot_longer(names_to = "date",
               values_to = "cases",
               cols = ends_with("20")) %>%
  mutate(date = as.Date(date, "%m/%d/%y")) %>%
  group_by(date, `Country/Region`) %>%
  summarise(cases = sum(cases))

covid_long <-
  covid_long %>%
  group_by(`Country/Region`) %>%
  mutate(seen_100_cases = max(cases) > 100) %>%
  filter(seen_100_cases) %>%
  mutate(dayof100 = min(date[cases > 100]),
         days_since_100 = date - dayof100) %>%
  filter(days_since_100 > 0)

label_df <-
  covid_long %>%
  group_by(`Country/Region`) %>%
  arrange(-days_since_100) %>%
  slice(1) %>%
  filter(days_since_100 > 5)

thirty_three <- function(x)
  100 * (4 / 3) ^ x

line_df <-
  tibble(days_since_100 = 1:50,
         cases = thirty_three(days_since_100))

g <-
  ggplot(label_df, aes(days_since_100, cases, group = `Country/Region`)) +
  geom_text_repel(aes(label = `Country/Region`)) +
  geom_point() +
  geom_line(alpha = 0.25, data = covid_long) +
  geom_line(alpha = 0.25,
            linetype = "dashed",
            aes(group = NULL),
            data = line_df) +
  scale_y_log10(labels = comma) +
  annotate("text", 30, thirty_three(30), label = "33% daily increase") +
  theme_bw() +
  coord_cartesian(ylim = c(100, 1000000)) +
  labs(
    y = "Confirmed cases",
    x = "Days since 100th case reported",
    caption = "Data: https://github.com/CSSEGISandData/COVID-19
       Code: ",
    title =  "Cumulative cases since 100th case reported",
    subtitle = "Graph design inspired by @jburnmurdoch
    https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441"
  )

ggsave(
  filename = "covid_graph.pdf",
  plot = g,
  width = 6.5,
  height = 4.5
)

updated link : https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv