Corona Analytics

Motivation

The motivation behind this web application is that we had difficulties finding data analysis for comparing different regions in a meaningful way. In order to do so, one has to take the population of the different regions into account, as well as the time from when the virus took a foothold.

Version History



Version	Date	Description

1.0.0	May 15, 2020	Comparison of normalized total Covid-19 deaths between various health regions. Trends of new cases, new deaths, total cases and total deaths per region. A simple SIR model fit to data (total deaths) per region. This basic model does reasonably good job at modelling the first wave.

1.1.0	March 11, 2022	Increased space allocated for graphs. Moving average filtering of daily data. Separation of regional trends and SIR modelling. Region selection remembered between pages. sr(t) and β(t) are estimated from daily data. Time varying sr(t) and β(t) result in good SIR modelling of the waves and variants seen thus far.

Known Issues

The data cleaning and filtering do not fully resolve raw data issues for France and Spain (no reported data for several days and negative new cases). This results in spikes in estimates for sr(t) and β(t).

Contributors

Johan Backström has a M.Sc. degree in Systems Engineering from the Royal Institute of Technology (KTH) and over 20 years experience in modelling and control of dynamic systems. Current R&D interests are system identification, advanced controls and optimization, analytics, and contemporary web application development. Johan is a registered professional engineer in B.C., Canada.

Paul Dooner is a software engineer in Vancouver, Canada.

Justin Johnson is a software developer in Vancouver focusing on web applications supporting large service design projects in government and industry.