[pgrouting-dev] Alternative paths

[Stephen Woodbridge]

On 1/9/2012 9:14 PM, Daniel Kastl wrote:
>
>
> On Tue, Jan 10, 2012 at 9:07 AM, Stephen Woodbridge
> <woodbri at swoodbridge.com <mailto:woodbri at swoodbridge.com>> wrote:
>
>     So the question about alternative paths came up on the users list
>     and this seems like a good paper discussing how to implement that.
>
>     http://algo2.iti.kit.edu/ download/altgraph_tapas_ extended.pdf
>     <http://algo2.iti.kit.edu/download/altgraph_tapas_extended.pdf>
>
>     This would be a cool GSoC project or just something someone might
>     want to tackle.
>
>     Did Jay or someone implement k-shortest paths? I see we have a
>     ticket for this some maybe not:
>
>     https://github.com/pgRouting/ pgrouting/issues/11
>     <https://github.com/pgRouting/pgrouting/issues/11>
>
>
>
> Hi Steve,
>
> This hasn't been done yet. And it could be interesting project for GSoC,
> I agree.
>
> It's also not clear to me, if k-shortest path doesn't mean often just a
> tiny permutation in the path. This wouldn't be really what we want, right?
> I guess, that alternate routes as Google provides them are probably
> calculated with different costs.

Correct, I would agree the k-shortest path just does little 
permutations, so read the pdf link above. That is what we need for what 
google is doing.

-Steve

> Daniel
>
>
>
>
>
>
>
>     -Steve
>     ______________________________ _________________
>     pgrouting-dev mailing list
>     pgrouting-dev at lists.osgeo.org <mailto:pgrouting-dev at lists.osgeo.org>
>     http://lists.osgeo.org/ mailman/listinfo/pgrouting-dev
>     <http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
>
>
>
>
> --
> Georepublic UG & Georepublic Japan
> eMail: daniel.kastl at georepublic.de <mailto:daniel.kastl at georepublic.de>
> Web: http://georepublic.de <http://georepublic.de/>
>
>
> _______________________________________________
> pgrouting-dev mailing list
> pgrouting-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pgrouting-dev