francksteel: at which age does one learn to use a fonction like f(x) = L / (1 + exp(-k(x-x0)) (this kind of fonction describe well an
7PCGamer: hi,
i am trying to model the covid-19 spread (as a passtime... games no longer soothe my nerves)
somehow the function above is not consistent with single peak curves that i have been looking at... can you link some source material please?
thanks.
For a very simplistic, introductory model, you might take the logistic curve, but you can do better.
Nowadays, the standard is the SIR model for spread of sickness, that considers infected, recovered and susceptible people. It is so standarized that it is used as an example in introductory courses to differential equations.
Also, one thing is considering the spread of the sickness if unchecked, and another how the sickness behaves once measures start being taken. You will see a delay of about 1,5 the duration of the incubation period (15 days) between the measures and the effect in the infection curve.
Also, you have to consider changes of policy in the testing. Like, availability of tests and whom are the test applied to. This policy can change over time, for example, when there are few tests, they can be reserved for the patients in worse condition, like it happened in two regions of Spain for few days after march the 7th, saturday, which led to an apparent fdent in the curve that was later recovered.
The problem is fascinating and it is the work of epidemiologists to consider all the factors in the real world.
Would suggest to start with data from Italy and Spain, that were transparent, and compare daily data with the news. That might prepare yourself better for analysis in other countries.
Currently the daily increase is diminishing in Italy and Spain, with the peak to be reached soon.
In any case, the behaviour of the curve in the initial phase might be grossly assimilated to an exponential function. The daily increase duplicated every 3 days (approximately) in Italy and Spain. In the USA the curve is being steeper compared to the same stage in those European countries.
Finally, Spain, then Italy, are the European countries with the most severe measures against this sickness. Being very strict has worked greatly in South Korea and other countries. To be sure, they had the SARS as a precedent for this. Any approach that differs widely, meaning the lax measures in some countries like the Netherlands, etc, is the equivalent of making their population the subject of a risky experiment.
Edit: The simplest of approaches is considering the sickness growing exponentially. In the beginning, it will be good enough. Also, drawing it on a logarithmic scale (in base ten) will help you greatly.