18 Feb
2013
18 Feb
'13
2:35 a.m.
On 18/02/2013 06:46, Dan Asimov wrote: [me:]
Why would one do that, rather than using one of the deterministic linear-time algorithms mentioned on the Wikipedia page?
[Dan:]
Well, I'd probably try it, if compute time were not at issue, just to see what the graph, if planar, looks like when embedded in the plane.
Several of the methods listed on that page do compute a planar embedding when there is one. (They don't do the actual *layout*; you'd still need to work out just where to put the vertices. But finding a planar layout is easy once you've got the topology, and finding a *nice* planar layout is probably best done by first finding any planar layout and then improving it.) -- g