animate

animate.R

library(tidyverse)
library(countrycode) # install.packages("countrycode")

# get the data
  cases_wide <- read.csv("https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv", stringsAsFactors = FALSE)

# pivot longer
  cases <- 
  cases_wide %>%
  # sum subnational data to get national totals
    select(-Lat, -Long) %>%                 # drop these columns
    group_by(Country.Region) %>%            # group by country
    mutate_at(vars(starts_with("X")),       # sum each daily column
              funs(sum)) %>%                #  by county
    ungroup() %>%
    distinct(Country.Region,                # countries with subnational
             .keep_all = TRUE) %>%          #  have repetitive data now 
    select(-Province.State) %>%             #  so drop duplicates
  # pivot wide to long
    pivot_longer(cols = starts_with("X"),   # take the daily cases
                 names_to = "date",         #  put col headings in date
                 names_prefix = "X",        #  pattern is X variables
                 values_to = "cases") %>%   #  put case values in case
    mutate(date = lubridate::mdy(date)) %>% # create proper date var
    rename("country"="Country.Region") %>%  # give nicer var name
    filter(country!="Diamond Princess")     # drop diamond princess

# filter to top 10 per day
  leaderboard <- 
  cases %>%
    group_by(date) %>%
    mutate(rank = rank(-cases),
           relative_to_1 = cases/cases[rank == 1],
           label = paste0(" ", cases)) %>%
  # lots of countries tied for zero at the start
  # so order by rank and then alphabetical by country
    arrange(rank, country) %>%
  # give new ranks to break ties with alphabetical country names
    mutate(rank = 1:n()) %>%
  # limit to top 10 per day
    filter(rank <= 10) %>%
    ungroup() %>%
  # give a day label
    mutate(month = lubridate::month(date, label=TRUE, abbr=TRUE),
           day = lubridate::day(date),
           monthDay = paste(month, day, sep=" ")) %>%
  # convert country name to code
    mutate(cc = countrycode(country, origin="country.name", 
                            destination = "genc2c"),
           cc = tolower(cc))

# get vector of unique monthDay
  monthDayLevels <- 
  leaderboard %>% 
    distinct(date, .keep_all = TRUE) %>% 
    pull(monthDay)



  df <- 
  leaderboard %>%
# convert to factor
  mutate(monthDay = factor(monthDay, 
                           levels = monthDayLevels,
                           labels = monthDayLevels))