us_covid19_plotting.R

library(tidyverse)
library(scales)
library(ggthemes)
library(lubridate)
library(zoo)
library(covid19us)
library(datasets)

tsa <- c('NY','CT','NJ')
states <- tibble(state_name = state.name,
                 state = state.abb)

us <- get_states_daily() %>%
  mutate(grouping = ifelse(state %in% tsa,'NY Tri-State Area','Rest of the US')) %>%
  arrange(state,date) %>%
  mutate(pct_positive = positive_increase / total_test_results_increase) %>%
  group_by(state) %>%
  mutate(rolling_positives = rollapply(positive_increase,7,mean,align='right',fill=NA),
         rolling_deaths = rollapply(death_increase,7,mean,align='right',fill=NA),
         rolling_pct_positive = rollapply(pct_positive,7,mean,align='right',fill=NA),
         rolling_tests = rollapply(total_test_results_increase,7,mean,align='right',fill=NA)) %>%
  left_join(states,by='state') %>%
  select(date,state,state_name,everything())

us2 <- us %>% 
  group_by(grouping,date) %>%
  summarize(death_increase = sum(death_increase,na.rm=TRUE)) %>%
  ungroup() %>%
  arrange(grouping,date) %>%
  group_by(grouping) %>%
  mutate(rolling_deaths = rollapply(death_increase,7,mean,align='right',fill=NA)) %>%
  filter(date >= '2020-03-01')

us2_plot <- us2 %>% 
  ggplot(aes(x=date,y=rolling_deaths,col=grouping))+
  geom_line()+
  scale_y_continuous(labels = comma)+
  scale_x_date()+
  theme_tufte()+
  ggtitle('US rolling 7-day average for confirmed Covid-19 deaths, New York Tri-State Area vs Rest of the USA')+
  xlab('Date') +
  ylab('Rolling 7-day average of deaths')+
  labs(col='Grouping')