I understand your meaning now. Thank you very much.
--- In email@example.com
, "Brandon Corfman" <bcorfman@a...>
> A TSP does not require a complete graph, but it becomes more
> to solve it otherwise**. For instance, a fast but effective
> like 2-Opt (that hill-climbs by inverting subsequences along the
> would be rendered ineffective if the graph was not complete. You
> have to start checking constraints on the path at each exchange,
> requires different techniques (perhaps a constraint satisfaction
> Best regards,
> ** Source: Michalwicz, Zbigniew and Fogel, David B.; How to Solve
> Modern Heuristics, Springer-Verlag, 1999.
> --- In firstname.lastname@example.org, "chenyu468" <chenyu468@y...>
> > hello,
> > I am asking "why is there a requirement for a complete graph?"