I would like to ask for some help with graph layouts. I have a large graph, with ~4k vertices and 300k+ edges, and I use the DRL layout to visualize it. It works fine, however the resulting graph is very sparse, and I wonder which parameters should I tinker with to make it a bit less sparse. I played with init.iterations, expansion iterations, but so far without real success, the resulting graphs are not more denser, just more messy. Any suggestions would be great.
PS.: the graph is fully connected, so no isolated clusters, I simply use a low alpha to visualize its structure.
George, I know this is not the answer you are looking for, but I would use the equivalent OpenOrd implementation in Gephi. It works quite well at exposing structure and you can follow it up with a brief Force Atlas 2 with nonoverlap filter on to clean it up. This has been my go-to visualization combo for large networks for a long time. – DS
Thanks, I will try Gephi.