B1 US Infection reports: timing

Vincent J. Carey, stvjc at channing.harvard.edu

2023-01-07

Introduction

In this section we will work with time series of reports of new cases of COVID-19 infection.

Another time series resource of interest records date, place and assigned cause of COVID-19 related deaths. At this time we will not be analyzing this data.

A snapshot of data from “USAfacts”

The usafacts_confirmed function produces a table with rows corresponding to US counties and columns corresponding to dates. Entries of the table are counts of new COVID-19 cases.

Here’s a quick look at the table:

library(teachCovidData)
cc = usafacts_confirmed() # confirmed cases by county
library(dplyr)
MSdata = cc |> filter(State=="MS")
head(MSdata[,1:10])

##   countyFIPS           County Name State StateFIPS 2020-01-22 2020-01-23
## 1          0 Statewide Unallocated    MS        28          0          0
## 2      28001         Adams County     MS        28          0          0
## 3      28003        Alcorn County     MS        28          0          0
## 4      28005         Amite County     MS        28          0          0
## 5      28007        Attala County     MS        28          0          0
## 6      28009        Benton County     MS        28          0          0
##   2020-01-24 2020-01-25 2020-01-26 2020-01-27
## 1          0          0          0          0
## 2          0          0          0          0
## 3          0          0          0          0
## 4          0          0          0          0
## 5          0          0          0          0
## 6          0          0          0          0

tail(names(MSdata)) # why

## [1] "2022-12-28" "2022-12-29" "2022-12-30" "2022-12-31" "2023-01-01"
## [6] "2023-01-02"

We have to massage the column names to simplify programming with reocrd dates.

dates = lubridate::as_date(names(MSdata[-c(1:4)]))
cumulative = apply(MSdata[,-c(1:4)],2,sum)
head(cumulative)

## 2020-01-22 2020-01-23 2020-01-24 2020-01-25 2020-01-26 2020-01-27 
##          0          0          0          0          0          0

We can now visualize cumulative case counts:

library(ggplot2)
mydf1 = data.frame(cumulative, dates)
ggplot(mydf1, aes(x=dates, y=cumulative)) + geom_point() + ggtitle("Cumulative Confirmed Covid cases in MS")

“Incidence” refers to the rate at which new cases arise, frequently reported as number of cases (in a given period of time) per 100000 individuals at risk at the start of the interval.

We use differences in cumulative counts to visualize time and size of outbreaks.

incid = diff(cumulative)
ndate = dates[-1]
mydf2 = data.frame(incid, ndate)
ggplot(mydf2, aes(x=ndate, y=incid)) + geom_point() + ggtitle("New daily cases in MS")