Introduction
I looked at extracted the Ciovid-19 statistics from Johns Hopkins University (JHU) . There are raw data for confirmed cases and deaths by date for each county in each state. The data also includes. For each state, I also extracted state population, area and population density from https://www.states101.com/populations on each state from
...
. I wrote a Perl script covid-us.pl to gather the above information and cast it in a suitable form for the www.charte.ca motion charts and correlation data.
Results
| US | US with fit | ||
|---|---|---|---|
| Deaths vs Confirmed cases | Here | Here | |
| Deaths vs % population confirmed with Corona-19 virus | Here | Here |
Notes:
Deaths vs Confirmed:
- Straight-line in log-log plot followed by most states. This indicates it follows a power law (exponential increase in both measures (deaths and % confirmed)).
- The first confirmed cases seen by CA towards the end of January, followed by IL at the end of January.
- High outliers are NM and NH
- Washington state started getting confirmed cases first starting at the end of February.Washinton state started having deaths in early March
- First deaths reported for Washington State at the start of March.
- Deaths start to really increase in the second week in March.
- By the start of April, NY followed by NJ are leading the way in both deaths and confirmed cases.
Deaths vs %population confirmed
- Note that since it is a log-log scale no bubble appears for a state until there is at least 1 confirmed case and 1 death for the state.
- Wide dispersion (not all states on line as seen in the log-log plot)
- % Confirmed and deaths both low for AK, VT, NH, ID
- Cluster of DE, DC and RI with low deaths compared to the % confirmed cases
- NY, NJ, MA, DE, CT, LA, RI DC have the highest % confirmed cases.
- By March 14th, WA, NY, CA, FL were reporting deaths.
Possible further work
Look at the California counties data from JHU using the same analysis and visualization.
Maybe look at red states vs blue states.