Spotting Differences - Data Points: Visualization That Means Something - page 264

FIGurE 7-31 Results of model-based clustering
That first plot on the top left shows the results of running an algorithm to
find the ideal number of clusters in the data. The remaining three plots
show the clusters. Pretty awesome. You get the two clusters, now more
well defined, showing high and low states.
This is usually when I tell you to bring your PDF files into Illustrator and
apply some touchups, but I'm not so sure I'd ever publish these for a gen-
eral audience. It's too abstract for the nontechnically minded to figure out
what's going on. They're good for data exploration; however, if you were so
inclined, all standard design principles apply. Figure out what you need to
tell a clear story, and strip out the rest.