I would think it would pick the first edge with the same weight as the edge on the optimal path; does it do something else?

 well, it can pick any edge if that is the shortest path to its endpoint vertex. this edge need not be in shortest paths to all vertex.

I think I'll use a patched version as you suggested. Thanks Jeremiah.

-Shyam

On Mon, Oct 12, 2009 at 8:34 PM, Jeremiah Willcock <jewillco@osl.iu.edu> wrote:
On Mon, 12 Oct 2009, Shyamendra Solanki wrote:

This doesn't work. It always selects the first outgoing edge's endpoint vertex, which may not be in path.

Was the predecessor matrix avoided to meet some design goal? i don't understand.

I would think it would pick the first edge with the same weight as the edge on the optimal path; does it do something else?  As to why there is no predecessor matrix, I do not know; the code is very old and I did not write it.  I do not think it would be too hard to add if you wanted to (to either algorithm).


-- Jeremiah Willcock
_______________________________________________
Boost-users mailing list
Boost-users@lists.boost.org
http://lists.boost.org/mailman/listinfo.cgi/boost-users