The Sankey Graph is one of the most famous data visualizations in history (in no small part because of Edward Tufte‘s proclamation of its greatness). Created in the 1890’s by Irish Captain Matthew Henry Phineas Riall Sankey to demonstrate the efficiency of steam engines, it was popularized during Napoleon’s Russian campaign as an efficient way to represent the disastrous consequences of that ill-fated military strategy. The skinnier it gets, the more troops died, up until the thin line that shows the surviving troops that made it back to Paris.
Lee Byron, creator of the streamgraph, used a Sankey Graph to show the attrition of a relationship (original here). Drawn from a variety of data sources, he uses this incredibly data-compressed (more on that in a moment) visualization in a very creative way.
(Via Fastcodedesign. Click to expand.)
What I find so awesome about the Sankey graph is the amount of data it can encode in such a small space. As I’ve argued elsewhere, the power of images lays in the way a picture can encode complex information–the very type of complex information that social scientists work with all the time. Notice how the relationship graph conveys both temporal direction (from the first date to death) as well as spatial (comparative by country) and social (types of breakups). We could learn a bit from graphic design.
- Stacked graphs – geometry & aesthetics (seeingcomplexity.wordpress.com)