Reader BR pointed me to a, er, unique representation of the proportion of the proportion of the population in each of the 50 US states with a BMI above 30: The United States of Obesity. Here's a small part of the graphic:

Obesity
I can't help but feel that this charts represents an opportunity lost (and I'm not talking about the dietary habits of the US). If we're going to go all anthropomorphic in a chart, at least let's visualize more than one dimension of data. Seems like a perfect opportunity to deploy Chernoff faces (such as with the faces function in the aplpack package for R): we could put little faces on each of those bodies and represent data like average age in the state, soda intake per capita, recreation time spent outdoors, that kind of thing. (Maybe some enterprising R user wants to take a stab at it?) Maybe we'd find that a certain range of Chernoff expressions is associated with a large BMI.



Link to original post