Today we’re talking about treemaps!

Treemaps are perfect for visualising data with some hierarchical structure. The visualisation consists of rectangles that are sized by a quantitative variable. The hierarchy is shown when initial larger rectangles (first level branches of the hierarchy) are then tiled with smaller rectangles from sub-branches of the hierarchy. The rectangles within the map are typically sorted by size. Colour is generally used to show the top level of the hierarchy, but can also be used like a heat map to show quantitative values.

Our World in Data has a great treemap showing who emits the most CO2 with region as the top level of the hierarchy.

Here is an example from Data-To-Viz showing world population by region.

Treemaps work extremely well when:

1. You have hierarchical data and want to visualise parts within categories
2. You have a large amount of data and limited space
3. Exact comparisons between branches or categories aren’t too important

The impact and ease of interpretation of treemaps is evident by their common use in mainstream media. The idea of the treemap was invented by Ben Shneiderman at the University of Maryland Huma - Computer Interaction Lab in the early 1990’s when he wanted to find a way to visualise the file directory on a computer. Some of Shneiderman’s beautiful treemaps can be found on his blog.

Image ref.