Colin Nielsen ha scritto: > The speed of r.cost depends not just on the number of cells but also > the complexity of the surface. Smoother surfaces seem to take longer > than rough ones, I think because there are more possibilities for the > algorithm to explore.
Thanks for this Colin, I'm going to test. However, the matrix is very small (250x370), and I have the same results with or without the Knight's move. In the meantime, I "solved" the dissolve issue: it turned out that a few NULL fields triggered the event. I think this is a bug; I can share the file if someone wants to do further testing. All the best. -- Paolo Cavallini: http://faunalia.it/pc _______________________________________________ grass-user mailing list grass-user@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-user