One of the perennial challenges in visualizing complex networks is dealing with hairballs: how do you draw a network that is so large and densely interconnected that any full rendering of it tends to turn into an inscrutable mess? There are various approaches to addressing this problem: BioFabric, Hive Plots, and many others. Most involve very different visual abstractions for the network.
There is something compelling, however, in seeing the full, messy complexity of a network laid out in one image. Many of the alternate approaches have the disadvantage of being less intuitive. Most people are accustomed to inferring network structure from a collection of dots and lines; not so much from a matrix representation. I wondered if there wasn't a way to retain the immediacy and intuitiveness of, say, a force-directed layout, while somehow ordering it and stretching it out in a way that would give the important elements room to breathe. In this blog post I will describe an effort to find this middle ground.