Weekly COVID-19 Infections
July 3, 2020 11:31 AM Subscribe
Weekly COVID-19 Infections
I have assembled tables of COVID-19 virus data using the daily reports from each state of the United States. From this database of approximately 20,000 entries, I have assembled weekly statistics and rankings for each state.
I am presenting the data in forms that I have not seen elsewhere and would have like to seen, for example, I am ranking the states and showing how those ranks change over time.
Why is ranking the states important? Beyond having states compete with one another for better strategies for dealing with the virus, ranking can allow for non-parametric analyses. Ranking over time is also a good way of grasping trends and outbreaks.
These trends illuminate such things as how New York has gone from first place to fortieth place.
I plan to add more analyses and make weekly updates. (This has been a crucial week.)
The posts so far are:
Most recent: Those States Where Coronavirus Infection Rates Are Dropping.
(Includes graphs for all fifty-one states and DC through June 27).
States where the COVID-19 Virus is Raging. Focuses on the top ten and bottom ten in infection rates through June 20. Discusses some of the methodology.
A Problem with COVID-19 Virus Antibody Testing. Looks at the ancillary issue of antibody testing.
Discrepancies in Florida COVID-19 Data Reporting. Looks at the ancillary issue of Florida gaming its numbers.
Excel files are available upon request.
I have assembled tables of COVID-19 virus data using the daily reports from each state of the United States. From this database of approximately 20,000 entries, I have assembled weekly statistics and rankings for each state.
I am presenting the data in forms that I have not seen elsewhere and would have like to seen, for example, I am ranking the states and showing how those ranks change over time.
Why is ranking the states important? Beyond having states compete with one another for better strategies for dealing with the virus, ranking can allow for non-parametric analyses. Ranking over time is also a good way of grasping trends and outbreaks.
These trends illuminate such things as how New York has gone from first place to fortieth place.
I plan to add more analyses and make weekly updates. (This has been a crucial week.)
The posts so far are:
Most recent: Those States Where Coronavirus Infection Rates Are Dropping.
(Includes graphs for all fifty-one states and DC through June 27).
States where the COVID-19 Virus is Raging. Focuses on the top ten and bottom ten in infection rates through June 20. Discusses some of the methodology.
A Problem with COVID-19 Virus Antibody Testing. Looks at the ancillary issue of antibody testing.
Discrepancies in Florida COVID-19 Data Reporting. Looks at the ancillary issue of Florida gaming its numbers.
Excel files are available upon request.
Role: Project Leader, Data Collector, Analyzer
I added an update for the new cases for the week ending July 4th. I've spent two months putting together my database and the numbers for the past two weeks were shocking.
posted by dances_with_sneetches at 10:56 AM on July 9, 2020
posted by dances_with_sneetches at 10:56 AM on July 9, 2020
I added an update regarding the percent increases in cases over the past months including the percent increases for all states.
I also looked at the question of whether these were blue states or red states.
(16 out 20 of the states with the biggest increases are states that voted for Trump. 10 out of 10 of the states with the biggest decreases are those that voted for Clinton.)
posted by dances_with_sneetches at 2:07 PM on July 16, 2020
I also looked at the question of whether these were blue states or red states.
(16 out 20 of the states with the biggest increases are states that voted for Trump. 10 out of 10 of the states with the biggest decreases are those that voted for Clinton.)
posted by dances_with_sneetches at 2:07 PM on July 16, 2020
Very beautiful use of colors and you've indeed made data look beautiful a la Reddit style. I wonder why Google chose not to display per state data graphs in the searches because country-wise display is just too broad to be useful only just satisfies the itch to look at big numbers without being of much actual use. The way you've displayed it is much more useful I would argue. Also interesting to see the states that have follow a normal distribution curve which flatten over time versus states where the cases don't follow a normal distribution. An epidemiologist can collaborate with a data scientist and probably make their graphs much more interactive and people friendly and thereby more useful. Great job overall and loved your website too.
posted by MetaSquid at 12:36 AM on July 30, 2020
posted by MetaSquid at 12:36 AM on July 30, 2020
« Older Full Spectrum Resistance... | Maps of the Lost... Newer »
Interestingly, considering Trump's recent visit, South Dakota has the lowest testing rate per population of any state (for the week ending July 4). Trump's next stop, New Hampshire, ranks second-to-last.
posted by dances_with_sneetches at 3:37 PM on July 8, 2020