theandyb/nyt_covid

scripts and analyses with the NYTime's COVID dataset

Some COVID-related Scripts

Scripts and shiny apps I've used to play around with the New York Times' COVID-19 data

Get the data

cd data
git clone https://github.com/nytimes/covid-19-data.git

Update daily

cd data/covid-19-data/
git pull

Cumulative cases by county for each day

The file cumulative_counts_by_county.R generates an excel document with the cumulative cases by counties across the dates in which there is data.

Accompanying Shiny App

The cumulative_by_state directory contains a simple shiny app to navigate the county-level cumulative counts by state and date. Either run from rStudio, or from an open R session in this directory run

library(rshiny)
runApp('cumulative_by_state')

Growth over time

The directory growth_over_time has a simple R shiny app to look at how case numbers have grown over time in each county.

library(shiny)
runApp('growth_over_time')

Anomalies

The script anomaly_detection.r and the accompanying shiny app anomaly_table produce a list of counties and dates where each row represents a county whose case and/or death count on one day was less than the previous date. Results from me having run this are available in weird_counties.csv (note: this might not be up to date with latest NYT data)

Indiana Data

The state no longer provides this file (boo). JSON version should still work.

The Indiana_Cumulative_Counts directory contains an RShiny app that parses the COVID-19 CASE DATA csv file available at Indiana's Data Hub and returns the cumulative case count in each county by day. I've included a dataset in this directory, but likely you'll want to download something more recent from the Data Hub.

The indiana_json directory has an Rshiny app and some additional code for pulling data from the state's JSON feed.