The 'deciding where everything should go' was done in java -- the really heavy lifting-and-carrying was handled by the Barnes-Hut algorithm. The final.png was drawn by a quick little ruby/GD script. I suppose I ought to release the software some day, but it's still got lots of that hacked-together feel to it.
I've been investigating a similar mapping technique to the one these people used, nearly identical in fact, as applied to social networks. I've modelled people as antigravitationally interacting points, with friendships represented as springs.
Slashdot Mirror
The 'deciding where everything should go' was done in java -- the really heavy lifting-and-carrying was handled by the Barnes-Hut algorithm. The final .png was drawn by a quick little ruby/GD script. I suppose I ought to release the software some day, but it's still got lots of that hacked-together feel to it.
It's just generated directly from what's been published. It's not biased; this is just what people are working on.
I've been investigating a similar mapping technique to the one these people used, nearly identical in fact, as applied to social networks. I've modelled people as antigravitationally interacting points, with friendships represented as springs.
You can see an early render (deviantart.org), or one using the same data but with a slightly more sophisticated physics simulation (deviantart.org).