3 ms·
Why can't the asymmetric version of TSP be re-coded into TSP but with twice the number of nodes, each of which have a "distance" of zero to their counterparts?
by gerdesj 9y ago
Why can't the asymmetric version of TSP be re-coded into TSP but with twice the number of nodes, each of which have a "distance" of zero to their counterparts?
Ahhh, I think I've just noticed the fatal flaw here - it makes the graph directed. Perhaps fiddling with the notion of +0 and -0 might fix it ... 8)
- Tarrosion 9y agoIt can, see eg Bill Cook's "in pursuit of the traveling salesman." I've also wondered how this approximation algorithm relates to the standard asymmetric to symmetric transformation. I should read the paper I guess.