Home

Corona virus

Here are the fits using data from the John Hopkins CSSE University site. Graphs and fits are automatically generate every day.

Jump to the table of contents

World wide without data from China, we are currently seeing a purely exponential growth with a doubling time of 4 days. We are in a process of fast exponential growth and will see large numbers for the next weeks. A typical phenomenon of exponential growth.

The numbers outside China have surpassed the numbers of China by several factors.

Any hospital system will be overrun by exponential growth, even when we are slowing down the growth, because when 15% have severe conditions and require treatment for 2-3 weeks, capacities quickly fill up. A sufficient slow down will take several years, which is not practical. A solution would be to implement heaving testing options as artifical herd immunity. This way we can set the symptomless carriers in quarantine and stop the growth process. China and South Korea have show this method to be successful.

Three data sections are available: Worldwide, China only, and Worldwide without China. Also available are the fits for each individual region or country or continental areas. An advanced Gompertz function has been implemented. The old data are available with the information „simple Gompertz“.

Applied maths is explained here.


Table of Contents

Worldwide

China only

Worldwide without China

Fits and +10 days predictions for each individual region or country.

Fits and +10 days predictions by continental regions.

Fits and +10 days predictions on death rates.

Compare with Ebola data 2014/2015

Simple Gompertz: Fits and +10 days predictions for each individual region or country.

Simple Gompertz: Fits and +10 days predictions by continental regions.


If you like to support my work:
(by click on image you accept third party cookies, required by paypal)

World

Here, you can see beautifully the world overtaking now the data of china with a second exponential rise. No response yet on the recovery curve which will take roughly 3 weeks to follow. The 6 days delay of of death curve just starts to rise up. The fits will be screwed by this progress. Therefore, data from China are important for learning the basic parameters, see below.

The classical way to calculate the death rate, taking the data of the actual day. However, there is a delay between diagnosis and death. Therefore the death should be counted to the day of diagnosis, which is barely done. In the graph above, we seem from the turning point an average delay between diagnosis and death of about 6 days.


China only

What you see below is a rise at the end. It is caused by the delay of the death curve of about 6 days: So when the blue curve above rises quickly, the read death curve lacks behind causing lower calculated death rates. Once the blue curve fades out asymptoticially, the rise is over and now the second rise of the red death curve follows that is visible below. After all the process has passed, the final death rate can be calculated. However, if we fit the data and have passed the turning point of the red curve, the maximum numbers become more certain and provide a better prediction.


World without China

We can read from the data outside China, that they are growing on an expected exponential curve. No sign of any control, otherwise we would see a turning point approaching and the curve to tip over to the right. The T2 appears here much bigger, as we are far away from the turning point, so Nmax has a huge error. Fitted with the logistic function, we see a doubling time of about 4 days, like in the data in China. This is extremely fast and leaves little time for reaction. Most governments have started to react now and setup restrictions. Once the restriction is in place, we will see the data increasing for the next 6 days as if there were no restrictions. The reason is the delay of the latent infected people. The average incubation time is about 5-6 days. So when we stop all contacts, all those within the incubation time will pop up in the data by that delay. Therefore we need patience and wait effects to show up, when a restriction was set in place. (22.03.2020)


.·. Impressum