I have some doubts about the representation of the COVID-19 epidemic by the John Hopkins platform.

In the most prominent map, they want show numbers of cases linked to a geographic position in a map. They represent this with a big red circle, varying in size for the number of cases. I find this confusing, since it appears that the whole area is infected or were hit by a atomic virus bomb and places with a couple of cases are almost invisible. But what other options are there to visualize where and how many cases?

As a side dish, they show 3 plots of actual, logarithmic, and daily increase. Is that a reasonable way of representing a growing epidemic? This question might slip too much in the opinion based terrain, but I'd like to know what are standards ways of representing this, and why these three are relevant.

What could be a plot for showing how fast and how far the cases are spreading or possible outbreaks?

  • 2
    $\begingroup$ For the main map, one option would be a 3D bar chart on a map, with vertical bars representing case counts. This isn't a great example, but you probably catch my drift. Those other charts are informative, and I think they are all stats that interest professionals in that field, But, man, the aesthetics are awful. $\endgroup$
    – ulfelder
    Commented Feb 16, 2020 at 11:34

1 Answer 1


Representing population data on a map is difficult and implies applying simplifications to understand the data « at a glance ». One of these is to use a scatter plot which is exactly what is shown in this data:

For the different data points, the size of the area is shown as a disk whose surface increases linearly with the number shown (here, infected people).

Indeed, this can lead to some confusion (it’s a data point on the map, it is not spread), and will difficulty apply to data which would be densely packed, leading to overlaps in the disks. For the latter, you may prefer heatmaps.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.