Coronavirus is all anyone can think about these days, and the New York Times has become a repository for USA case data for some reason. They have been publishing the data on Github for now.

Data Loading

I loaded the data directly from the NY Times repository. The repository has county and state lists but because I was interested in the different counties in Hawaii, I used the county file.

library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(ggplot2)
library(scales)

covid <- readr::read_csv('https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-counties.csv')

## Parsed with column specification:
## cols(
##   date = col_date(format = ""),
##   county = col_character(),
##   state = col_character(),
##   fips = col_character(),
##   cases = col_double(),
##   deaths = col_double()
## )

timestamp()

## ##------ Tue Apr 14 19:25:55 2020 ------##

Data pull time

Because the repository keeps getting updated but this post reflects the situation on 4/4/2020, I filtered all the data after this for 4/4/2020 and earlier.

glimpse(covid)

## Observations: 56,541
## Variables: 6
## $ date   <date> 2020-01-21, 2020-01-22, 2020-01-23, 2020-01-24, 2020-01-24, 2…
## $ county <chr> "Snohomish", "Snohomish", "Snohomish", "Cook", "Snohomish", "O…
## $ state  <chr> "Washington", "Washington", "Washington", "Illinois", "Washing…
## $ fips   <chr> "53061", "53061", "53061", "17031", "53061", "06059", "17031",…
## $ cases  <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
## $ deaths <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…

covid$county <- factor(covid$county)
covid$state <- factor(covid$state)
covid$fips <- as.integer(covid$fips)

Total Cases in USA as of 2020-04-04

Here is the total nationwide cases as of 2020-04-04.

x <- covid %>% filter(date == "2020-04-04") %>% select(cases)
sum(x$cases)

## [1] 310842

Total Cases by State

covid %>% filter(date == "2020-04-04") %>%
  group_by(state) %>%
  summarize(total = sum(cases)) %>%
  arrange(-total)

## # A tibble: 55 x 2
##    state          total
##    <fct>          <dbl>
##  1 New York      114996
##  2 New Jersey     34124
##  3 Michigan       14225
##  4 California     13796
##  5 Louisiana      12492
##  6 Massachusetts  11736
##  7 Florida        11537
##  8 Illinois       10358
##  9 Pennsylvania   10110
## 10 Washington      6788
## # … with 45 more rows

covid %>% filter(date == "2020-04-04") %>%
  group_by(state) %>%
  summarize(total = sum(cases)) %>% 
  filter(total > median(total)) %>% 
  arrange(-total) %>%
  ggplot(aes(x = reorder(state, total), y = total)) + 
  geom_bar(stat = 'identity') +
  scale_y_continuous(trans=log10_trans()) +
  coord_flip()

covid %>% filter(date == "2020-04-04") %>%
  group_by(state) %>%
  summarize(total = sum(cases)) %>% 
  filter(total < median(total)) %>% 
  arrange(-total) %>%
  ggplot(aes(x = reorder(state, total), y = total)) + 
  geom_bar(stat = 'identity') +
  scale_y_continuous(trans=log10_trans()) +
  coord_flip()

Total Cases in Hawaii

In Hawaii, the number of cases has been growing steadily.

covid %>% filter(state == "Hawaii", date <= "2020-04-04") %>%
  group_by(date) %>%
  summarize(total = sum(cases)) %>%
  arrange(-total)

## # A tibble: 30 x 2
##    date       total
##    <date>     <dbl>
##  1 2020-04-04   349
##  2 2020-04-03   317
##  3 2020-04-02   283
##  4 2020-04-01   256
##  5 2020-03-31   224
##  6 2020-03-30   199
##  7 2020-03-29   173
##  8 2020-03-28   150
##  9 2020-03-27   120
## 10 2020-03-26   106
## # … with 20 more rows

covid %>% filter(state == "Hawaii", date <= "2020-04-04") %>%
  group_by(date) %>%
  summarize(total = sum(cases)) %>% 
  ggplot(aes(x = date, y = total)) +
  geom_point() +  
  geom_line()

Cases by County in Hawaii

Honolulu county has the most cases by far (as it has by far the largest population).

covid %>% filter(state == "Hawaii", date <= "2020-04-04") %>%
  ggplot(aes(x = date, y = cases, color = county)) +
  geom_point() +  
  geom_line()

Cases by County other than Honolulu

To see the trend for the other counties, it is best to remove Honolulu County.

covid %>% filter(state == "Hawaii" & county != "Honolulu", date <= "2020-04-04") %>%
  ggplot(aes(x = date, y = cases, color = county)) +
  geom_point() +  
  geom_line()

Change per Day

Here’s the change in our state per day. The growth rate in positive tests may still be accelerating.

covid %>% filter(state == "Hawaii", date <= "2020-04-04") %>% 
  group_by(date) %>%
  summarize(total = sum(cases)) %>% mutate(day_before_total = lag(total)) %>%
  mutate(change = total - day_before_total) %>%
  filter(!is.na(change)) %>%
  ggplot(aes(x = date, y = change)) +
  geom_point() + geom_line() + geom_smooth()

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

coronavirus