Hello Heinz,
<br/>
Before comparing with John Hopkins' data and try to identify the different parameters I wanted to 'improve' the model. I have read several interesting articles on the web. <a href="http://www.madore.org/~david/weblog/d.2020-04-02.2648.html#d.2020-04-02.2648" target="_top" rel="nofollow" link="external">One of them</a>, written by David Madore, explains (in french) that the recovery model would be better at a "constant recovery time" instead of using an "exponential distribution of probability whose expected value is 1/γ" like in the traditional SIR model. <br/>I agree with him because I could recognized a "simple time delay" in his explanations <img class='smiley' src='/images/smiley/smiley_thinking.gif' /><br/>
So in the below diagram I replaced the gamma (γ) feedback on the infected integrator by a variable delay and add a super function diagram to model the healthcare system with:<br/>
- One input: infected people per day.<br/>
- Two outputs: Recovered and unfortunately dead people per day.<br/>
The other parameters are percentages of 'symtomatic', 'hospitabized', 'Reanimated' and 'dead' people on one hand and delays before being 'healed', 'hopitalized' and 'reanimated', but also 'incubation' time on the other hand.
<br/>
<img src="http://mailinglists.scilab.org/file/t498052/SEIRM_SystSant%C3%A9_01.png" border="0"/>
<br/>
Here are some of my results:
<br/>
<img src="http://mailinglists.scilab.org/file/t498052/Figure_n%C2%B020023_SEIRM_SystSante_01.png" border="0"/>
<br/>
Funny 'infected' curve and its derive right ?<br/>
I am thinking of studying the FFT of the real John Hopkings' curves to identify the delays to put inside my healthcare system model.
<br/>
But I'm running out of time before I have to work back next monday.
<br/>
Hervé de Foucault<br/>
<br/><hr align="left" width="300" />
Sent from the <a href="http://mailinglists.scilab.org/Scilab-users-Mailing-Lists-Archives-f2602246.html">Scilab users - Mailing Lists Archives mailing list archive</a> at Nabble.com.<br/>