date: 2020-04-10

SUMMARY

New COVID-19 cases in Washington have strongly declined in the last two days. Total cases in Greater Washington are still about a factor of five below the number of cases in Puget Sound. #COVID2019

GET DATA

Data are retrieved from the NYTimes Github Repo.

county_data_file <- "https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-counties.csv"

county_data <- read_csv(county_data_file, 
                        col_types = cols(
                          date = col_date(format = ""),
                          county = col_character(),
                          state = col_character(),
                          fips = col_character(),
                          cases = col_double(),
                          deaths = col_double()
                          )
                        )

The latest available data are from 2020-04-09.

METRO DATA SUMMARIES

Create regions by specficying specific counties. The Puget Sound is where 4.2 Million (56%) of Washington’s 7.4 Million people live.

## define region by county

puget_region <- c("Thurston", "Island", "Kitsap", "Pierce", "King", "Snohomish", "Whatcom", "Skagit" )



state_metro_data <- county_data %>%
  filter(state == "Washington") %>%
  mutate(region = ifelse(county %in% puget_region, "Puget Sound", "Greater Washington")) %>%
  yo

Plots of Cases by Region

Plot data is created by summarizing data by metro area. Daily case rates are also computed adn shown as the gray area plot. (the colored line is essnetially the integral of the grey area).

Linear view of the data shows rapidly escalating case number

Cummulative cases are rising rapidly, with the number of daily cases shown in grey below the curve.

Logarithmic view reveals changing growth behavior

Slopes representing different doubling times shown for reference. Longer doubling times mean slower case progression.

RATE CONSTANTS

Growth rates analyzed below with highly anomolous data (three sigma) removed to improve fit behavior.

Doubling Time Trend

Here is the (noisy) doubling time data computed on a daily basis with a smoothed trend line overlayed. Although there is a fair amount of noise in the data, trends appear.

Longer doubling times are better. (As new cases drop to zero the doubling time will trend to infinity).

Estimated Doubling Rate

An estimate of the doubling rate with uncertainy, by averaging last week of data.


fin