Update 2020-03-26 – updated animated map
I often see maps in the news that I find extremely lacking. I considered showing some particularly shameful maps, but the world has enough negativity in it right now. I am hoping that I can teach by example and show some nice techniques and maybe in my small way change the conversation.
Designing maps that display data clearly and honestly is exceptionally difficult. At first it seems so easy, but there are so many ways to go wrong. For example, we have all seen the political maps that make the entire country light up in one color, even though the candidate of the other color won the popular vote. In this particular case, the map is mistaking land area for popular support. If only mountains could vote.
So in general, whenever you paint a map with colors, be careful not to just make a plot of land area. Generally you can do this in 2 ways.
- Plot a rate or a ratio or something similar.
- Make all the geographies the same size (see here or below.)
Either, it takes land area out of the equation making a rural area on equal footing with a city. In the below map, I chose to show the rate of infections per 100,000 people. I feel that it is very easy to understand and conveys a very accurate picture of the problem facing our country. This is among the simpler maps that Alteryx can do and should be the default type of map unless you have another idea.
So where did the above map fail? It has a difficult time showing the intensity of the problem in New York City. The city (as as today, March 24th) is nearing a 1% infection rate! It is really horrifying. We could make the colors brighter or darker, but since the city is so small on the scale of things, that doesn’t help very much. Really the only way to address this is with physical scale. Many maps will paint blobs all over the map. In order to have the visual impact needed, they need to be quite large. Unfortunately, this means they almost certainly will overlap
So for the next map, I am making 2 changes. The first is that I am going to show a dot for each person that is currently infected with COVID-19. This is easy to understand. Since I don’t have actual point data of the real people being infected, I am picking a random point in their state. Going out to the state level leaves more room to disperse the NYC points and show the severity of the problem.
The other major thing that I changed is adding movement. Our eyes are amazing at picking out movement and that really helps draw us in to where the problem is. Did you know Alteryx can do animated maps? It sortof can’t, but its not that hard. This is a batch report writing to all separate PNG files that I combined with Image Magick. But since Alteryx can run command line tools it all happens seamlessly with one click of the run button.