Make insightful dot density visualizations in CARTO!

Explore our most recent blog post about dot density visualization and 4 more types of spatial data visualization. Plus, learn how to create them using SQL.

This map shows the ethnic diversity of the US by showing one point, for every 100 people of a given ethnicity.

To make the map we used data from the census along with a method called dot density. For each census block group we know from the census the number of people who reported themselves as Asian, Black, White and Hispanic of Latino. This allows us to make choropleth maps in the same way you are probably used to showing each ethnicity. For example here is a map showing the counts of people who indicated they are Black on the census:

What if we want to show all the ethnicities on the same map? What if we wanted to have a map that gets much closer to conveying the number of people who are reporting in the census along with their make up? Well thats when dot density can be useful. By using this following PLPGSQL function which you can add to you own account simply by copying and pasting it in to the SQL panel on CARTO, we can turn geometries and counts in to points in those geometries.

We can run this function for each ethnic group in our dataset as so:

This will return a table of points, one for every 100 people in the census tagged by their etnhnic group. The result (after some considerable computation time) is this map, which conveys the shear number of people in the census along with who the diversity of ethnic groups across the US.

It took quite a while on one of our largest plans to make the dot density for the entire US so we recommend you use this
method for smaller areas where it can still be really powerful!

How people get to work?

Of course you aren't limited to just making dot density maps of ethnic groups. You can use any data that is a number count of people in a geometry. For example Daniel McGlone over at Azavea produced this blog post about using the same function to map out how people in Phillidelphia.