Evolution of Covid-19 in France, Sweden, USA and World (last updated: 29dec20 11:00)


This page displays various Covid-19 France, USA and World epidemic forecasts for the next few days to come.

These forecasts are based on various mathematical models which seem to well fit the historical Covid-19 data.

They have no other claims than show a trend and allow us to roughly see where we stand with the Covid-19 epidemic.

This page will be updated if possible daily, based on the availability of new data.

The French Health Ministry has stopped publishing data during the week end. The next available data are published on Monday evening at around 21h.

Sources: "Santé Publique France" for the new daily data and https://fr.wikipedia.org/wiki/Pandémie_de_Covid-19_en_France for France historical data.

Download the data (Covid19.csv)   (missing values are represented by -99999 in this file)
 

Sources: https://ourworldindata.org/covid-deaths for Sweden, USA and World historical data.

Download the data(Covid19USW.csv)   (missing values are represented by -99999 in this file)
 

See notes about these charts

 

Notes:

  1. All regression computations have been done with APL+Win and the Excel Solver
  2. Charts are produced by a C#+DevExpress application that calls APL+Win for the calculations
  3. For easier reading by most everyone, the best fit regression equations displayed in the charts have been converted from APL syntax to Python syntax
  4. Some data vary a lot depending on the day. This is the case of "Number of New Confirmed Cases per day" and "Number of New Deaths per day". One of the reasons is that the data is not well regularly enough transmitted from the ground up to the government every day, notably during the weekends. For this reason, it makes more sense to do regressions on a 7-days moving average of these variables.
  5. The curves decrease rate seems better predicted by LogNormal regressions
  6. To best fit historical data, 3 Sigmoids have been combined for cumulative curves and 3 LogNormal equations have been combined for daily curves
  7. Thanks to Mircea T. Sofonea (Expert and Lecturer in Epidemiology and Evolution of Infectious Diseases at Montpellier University) for his numerous precious pieces of advice
  8. And thanks to Pierre Windal (Director at Pierre Windal Conseil) for his precious help concerning the cumulative curves regressions and forecasts