trying the script

[nickoodn]

Hello,
Very interesting model and matlab codes. I have to tried to use it for Italy, France and other countries today but it is not so accurate than your examples. Do I need to change some parameters ?
(I have used the example3 with modifying country)
Thanks in advance and congratulations for your job

[ECheynet]

Hi nickoodn,
Thank for your message!
The fit may be sensitive to the choice of initial conditions and the subset considered for fitting. Which country did you try and which period?

[nickoodn]

Hi Echeynet,
Thanks for your rapid answer.
I would like, for example, to model France and Italy with all available data (from January to now)

[ECheynet]

I do not think the data recorded in January (or even February) for France and Italy are reliable. The way the deceased were counted was different for example. If you want to use with several months of data, I would start from the end of February or beginning of March. This should ensure a more reliable fit.

[nickoodn]

Thank you. For these nice curves did you manually adjust some parameters? I will try again with starting later the first data.

[ECheynet]

Hi, the parameters are adjusted automatically. However, after fitting, one should check if the death and recovery rate are properly approximated by the functions kappa(t) or lambda(t), respectively. I also check the sensitivity of the period considered on the final fit, especially for the death rate.

[nickoodn]

It is great. I will retry, I hope it was not due to my modifications in matlab (due to missing pack tools). Do you know if it is also written in pyhton?

[ECheynet]

I don't know but this should be feasible to have something in python. Most of the Matlab functions are not exclusive to Matlab. Maybe Python's implementations are not as effective but they may be good enough.