I am quite frustrated with corona graphs in the news, since most reporters seem to have skipped math classes back then. For instance, just plotting the number of confirmed infections at the respective dates does not tell you anything due to the different time point of outbreak. So lets fix that first.

The following charts are based on live data from https://github.com/pomber/covid19, which in turn sources jhu.edu.

Countries to compare

This is still pretty useless as the countries differ in total population and population density, lets turn to something actually comparable.

Spreading rate measures

The values are smoothed using a moving average over four days. This numbers actually allow comparing how well the different countries are doing in their corona countermeasurements, regardless of the population size.

# The same data can be visualized more intuitively as the doubling time, i.e. the time it takes for the confirmed count to double. It is given below for reference.

However, this representation is less numerically stable as it diverges to infinity as the infection rate approaches to 1.0.

# Yet another way to represent the same data is the effective reproduction number R.

Actually, it is very similar what we defined as the infection rate above. However, instead of comparing the the absolute number of confirmed from day to day, the change in confirmed over a period of time is used. Therefore, R becomes 0.0 if there are no new infections.

# However, confirmed does not mean the same between different countries and even in the same country at different time points of the epidemic. This is due to the sampling bias induced by the limited amount of corona test kits. In the early days of the epidemic there are enough kits to test everyone, so many cases that do not yet show symptoms are tested. With ongoing spread, we hit limits on test-kit and health-system capacities and the focus shifts to testing severe cases only. This in turn pushes the fatality rate.

A significant increase of the fatality rate indicates that

  1. the confirmed count is being under-estimated.
  2. the health-system capacites being exhausted.
# In contrast to the fatality rate, the mortality rate below is shown in dead per million inhabitants. This makes it independant of whether the confirmed count is estimated correctly. In most cases both will be correlated.

However, the mortality rate is a better indicator of the influence of the pandepic on a countries society and economy - especially when the health-system is exhausted.

# To close with something positive, lets look at the absolute number of infections. That is, cases which are neither recovered nor died yet and are potentially infectious. If there are no such cases left, we can assume the epidemic to be stopped. Notably, when this graph reaches zero you can not assume that the epidemic is stopped due to the stochastic nature of the data. However, a rise of the graph after a decline can indicate that a second wave is coming.

Source-code: https://github.com/paroj/arewedeadyet
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.