This site compiles the various analyses and plots that have been useful for sharing. For more information, see About.
Why do reported new daily cases fluctuate by the day of the week? This has been attributed to fluctuations in the daily variation in testing. Typically, reported cases are higher Thursday, Friday, and Saturday because testing is available early in the week. Testing may not be available on the weekend in many locations, leading to lower reporting early in the week.
Here, I explore how individual states contribute to the total USA case count. I focus on MA, NY, and NJ because 1) this is where cases were high early in the epidemic and 2) many of my friends and family live in these states.
In the NYTimes anomaly notes they describe the various anomalies that occurred in MA data collection and reporting—including things like releases of backlogs of data (causing large spikes) and dates of system maintenance (causing missing data for a day). They also describe incidents of MA state revising its methodology for case counts, leading to a “decrease” in new daily case numbers as they adjust their overall numbers on that specific day. For more information, see the anomaly notes.
For information about the current % positive COVID19 tests in each Boston neighborhood, see the BPHC blog here: https://www.bphc.org/onlinenewsroom/Blog/Lists/Posts/Post.aspx?ID=1282
Data for these analyses are obtained from the NYTimes.
Terms of use of the New York Times data here: https://github.com/nytimes/covid-19-data
https://github.com/nytimes/covid-19-data/blob/master/LICENSE
No medical or scientific conclusions should be made using the analyses within this repository.
The code to create this page and the analyses herein can be found at https://github.com/vcannataro/COVID19_data_explore
Special dates like Thanksgiving and Christmas have anomalies in reporting. I do not currently take this fact into account when generating the 7-day rolling mean.