I have updated the graphs for the latest week. There is a change in direction of both the New Case Slope and the Days to Double graphs. This may be a Holiday week effect like the weekend effect where fewer cases are reported. Hopefully, this change is a start of decline in the number of new infections because new restrictions are working.volvolugnut wrote: ↑Sun Nov 22, 2020 12:14 pmNew COVID-19 cases are growing in the US. Nearly every week brings new highs in the number of cases. The growth rates which I have tracked for months are rising again to levels similar to July or even April.volvolugnut wrote: ↑Sat Nov 14, 2020 11:30 am More statistics on US states and COVID-19 growth rates. My raw information is from Worldometer. Any errors or misrepresentations are my own.
To give some context to these changes, I have calculated the average growth rate for each week back to April. The growth rate is calculated as the number of new cases each day divided by the number of active cases for the previous day. There is a typical weekly pattern of rising cases each week and falling case numbers on the weekends. I have calculated the 7 day average of the growth rate for Saturday to Friday of each week.
Using the 7 day growth rate, a calculation can be made of the number of days for the new cases to double. This calculation uses (1+Growth Rate) to the exponential of 1/X and solved for when this equation equals 2 (doubling of new cases). Variable X is the number of days to double the new cases count. I have made this calculation using 14 day and 21 day growth rate averages as well.
A graph using the average for 7, 14, and 21 days and showing the days to double new case counts for April to middle of November is below. Another graph shows the 7 day average new virus cases counts for the same period.
Inspection of the graphs show that when the line for days to double becomes flat, there may soon be a decline or increase to the count of new cases. This has happened three times since April. When the slope of the line for doubling is steep, either going up or down, there is unlikely to be a change soon in the growth or decline in new cases.
The current lines for case doubling now shows some indication of going flat soon. I expect we will see the number of new virus cases continue to increase, but we may have a change of direction starting. Perhaps the changing recommendations for controlling the virus are beginning to have an impact.
There was an error in my original logic posted on November 14. During the upward slope periods of the Days to Double graph, the actual number of daily cases are declining. I reconsidered how to resolve this contradiction in logic.
I have added a new graph with a calculation of the slope of the new case numbers (current case count minus past case count divided by the number of days). The slope of a curve is the rate of change of the plotted line. If we were plotting velocity, the slope would be the acceleration. The slope is a better representation of the virus growth because the slope goes negative when the trend in cases starts going down.
volvolugnut
Stay safe out there.
volvolugnut