If the page does not update, press shift + refresh

Last update Thu 11 Mar 14:16:32 CET 2021

For the second wave calculation, data from 10th October 2020 are considered only: at least for Italy this seems to be the date in which we had the divergence from a linear regime to an exponential one.

FORECASTS SHOWING DASHED LINES AND "NA" IN THEIR ESTIMATE ARE NOT RELIABLE SO FAR.

The non-linear model to forecast the evolution is reliable only after one third of its path (Winsor, PNAS, 1932). Hence errors are likely larger than currently estimated in situation where the outbreak is just started.

Saturation values are the alpha parameter of the Gompertz fit. Errors are estimated as a combination of a) 95% confidence interval of Gompertz fit parametrs obtained with confint2() function and b) in order to account for errors into the data reporting, the percentiles are computed 100 times adding a Gaussian noise independent for each daily value, that has mean 0 and standard deviation equal to the 20% of the daily mortality.

Forecasts are updated at least twice a day (07:00 and 18:00).

Some inconsistencies in Italian prediction may arise from slight data differences between the Civil Protection and the Johns Hopkins archives.

The code can be found on my GitHub repository, althought it is really badly written.

Forecasts at 2021-03-10

Forecasts at 2021-03-09

Forecasts at 2021-03-08

Forecasts at 2021-03-07

Forecasts at 2021-03-06

Forecasts at 2021-03-05

Forecasts at 2021-03-04

Forecasts at 2021-03-03

Forecasts at 2021-03-02

Forecasts at 2021-03-01

P. Davini (CNR-ISAC), October 2020