« Less is more | Main | A picture is worth... »



I tried out Excel's new heatmaps and they look pretty good: http://www.westnet.net.au/balson/Temp/cht1.PNG

and this approach isn't bad either:


Dermot: Thanks for the examples.

The "heatmap" and contour plots certainly improve upon the data tables but you probably noticed that the value dimension got lost in the mix. The colors can't show the scale of the value dimension. Depending on the colors chosen, the order of the color and the width of each color "bin", many different versions of these plots can be created.

What is clear is that the small multiples principle forces us to pick one dimension to be spread across graphs (here, the Red dimension), and in some cases as in here, there is no obvious candidate for it.

Jon Peltier

When I worked as a research scientist, and later as production engineer, I would do the small multiple thing (I didn't know it had a name at the time) making several different views, so I could vary the dimension which was spread across the charts. I didn't have to pick. For each experiment or set of process conditions, a different arrangement might best show the behavior. My predecessor in one job would stare at a table of numbers and come up with his guidelines, whereas I drew a couple pages of charts and refuted all of his work, and I could use the charts to explain it to others in the group.

iowa dermatologist

At first i was really confused. But understanding every section clears everything out.

Matt Curtis

How did you make your so called "junkart" visualization?


Matt: That was done in R where one can control every little aspect of a chart.

The comments to this entry are closed.

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter