Forecast of Coronavirus cases in Italy

Published: Mar 15, 2020 by marc

This post is an updated version of this article and will be updated on a daily basis.

Code available on GitHub.

Update: 25 March 2020

According to the logistic regression model, the predictions are the following:

Peak of the infection: 21 March 2020
End of the infection: 20 May 2020
Total cases at the end of infection: 113884 (min: 110957, max: 116812)

The image below shows the real data compared to the logistic model.

scraped

Update: 23 March 2020

According to the logistic regression model, the predictions are the following:

Peak of the infection: 22 March 2020
End of the infection: 22 May 2020
Total cases at the end of infection: 117476 (min: 112024, max: 122929)

The image below shows the real data compared to the logistic model.

scraped

Update: 22 March 2020

According to the logistic regression model, the predictions are the following:

Peak of the infection: 22 March 2020
End of the infection: 30 May 2020
Total cases at the end of infection: 121717 (min: 113798, max: 129636)

The image below shows the real data compared to the logistic model.

scraped

Update: 21 March 2020

According to the logistic regression model, the predictions are the following:

Peak of the infection: 21 March 2020
End of the infection: 28 May 2020
Total cases at the end of infection: 114633 (min: 105440, max: 123825)

The image below shows the real data compared to the logistic model.

scraped

Update: 20 March 2020

According to the logistic regression model, the predictions are the following:

Peak of the infection: 20 March 2020
End of the infection: 1 Jun 2020
Total cases at the end of infection: 94246 (min: 87915, max: 100576)

The image below shows the real data compared to the logistic model.

scraped

Update: 19 March 2020

According to the logistic regression model, the predictions are the following:

Peak of infections: 18 March 2020
End of infection: 10 May 2020
Total cases at the end of infection: 77924 (min: 74014, max: 81835)

The image below shows the real data compared to the logistic model.

scraped

Update: 18 March 2020

According to the logistic regression model, the predictions are the following:

Peak of infections: 17 March 2020
End of infection: 18 May 2020
Total cases at the end of infection: 67488 (min: 64775, max: 70201)

The image below shows the real data compared to the logistic model.

scraped

Update: 17 March 2020

According to the logistic regression model, the predictions are the following:

Peak of infections: 17 March 2020
End of infection: 10 May 2020
Total cases at the end of infection: 64710 (min: 61316, max: 68104)

The image below shows the real data compared to the logistic model.

scraped

Update: 16 March 2020

According to the logistic regression model, the predictions are the following:

Peak of infections: 17 March 2020
End of infection: 6 May 2020
Total cases at the end of infection: 67812 (min: ‬62402, max: 73220)

Observations: new cases and the total number are starting to slow down, this suggests that the infection is near the peak. However two regions were late in reporting the number of new cases.

The image below shows the real data compared to the logistic model.

scraped

Update: 15 March 2020

According to the logistic regression model, the predictions are the following:

Peak of infections: 18 March 2020
End of infection: 9 May 2020
Total cases at the end of infection: 73520 (min: 64093‬, max: 82947)

The image below shows the real data compared to the logistic model.

scraped

Update: 14 March 2020

According to the logistic model, the predictions are the following:

End of infection: 6 May 2020
Total cases at the end of infection: 63493 (min: 53689, max: 73297)

This image shows the real data compared to the logistic and exponential model.

scraped