Lucas Garron
Premium Member
Yeah, that's a covering (as opposed to a packing). Even Mathworld agrees for n=1. I was trying to find a covering of all Square-1 shapes using Jaap's graph in Mathematica, but that never finished. (And even kastellorizo, with a "Master's in Algebraic Graph Theory," couldn't helpA while ago I tried to compute how many cases need to be fully learned if you allow making one initial turn speed bld style. Don't remember the number, but maybe a dozen or less. With two initial turns this of course gets reduced further, and this should then be fairly easy to learn and do.
Darn. I know this is a set cover problem, but I thought it would be more precisely be a vertex cover problem. Checking that again, I noticed I was wrong. Is there a name for the problem of having an undirected graph, a vertex considered covered iff itself or one of its neighbours is marked, and the task is to mark as few vertices as needed to cover all?
And then I decided it was useless. Better memorize the few 6-shapes (much less to keep in mind/refresh) and speed BLD those 4 twists to them -maybe memo those twists, too, if they're not obvious.
But I'd still love to know the minimal coverings for distances n=1 to 5.
(1 and 7 are trivial, and so is 6 if you know the graph. I know it has a diameter of 7, is any shape removed from more than one other shape by 7 twists?)