[pgrouting-users] Problem with way restrictions in shooting star if you have more then one rule for an edge

Max Weninger max.weninger at gmail.com
Wed Jan 11 11:45:05 EST 2012


Hi

On Wed, 11 Jan 2012 11:28:38 -0500
Stephen Woodbridge <woodbri at swoodbridge.com> wrote:

> On 1/11/2012 11:06 AM, Max Weninger wrote:
> > Hi
> >
> > Thanks for your replay
> >
> > I just confirmed that the "order" is indeed the problem
> > by creating a "hacked" version that changes the order "on the
> > fly" :)
> 
> What are you referring to with respect to "order"?
> The order of the edge_ids in the rules? or something else?

Yes the order of the rules as they are stored in adjacent_edges
in the shooting_star_boost_wrapper.cpp file based on the DB entries

It always respects only the first edge entry.
Doing an "internal" revert made my "example" work correctly but
now others will fail (that worked before) of course.

I tried to "understand" the C++ code of the "boost part" but the
last time I used C++ is definitely too far away and I always "hated"
using it :)

But I am pretty sure that the problem is somewehre in the relax method
of shooting_star_relax.hpp since there it iterates over the rule
entries (around line 99)

Regards

max

> Regarding shooting star, there are definitely bugs in the current
> code so I would recommend that people test their use cases before
> relying on it or just avoid it until we can sort our what the issues
> are.
> 
> > I will take a look at the "new" code asap :)
> >
> > BTW: just about "speed". I tried some different graph based
> > routing "solutions" now. By far the fastest was one based on the
> > igraph package. http://igraph.sourceforge.net/index.html
> 
> Thanks, I will look at this. I'm always interested in what other
> people are doing for routing.
> 
> > The problem is that it is "just" a "static" Dijkstra
> > without any support for turn restrictions which is
> > definitly a "must" for me as the "final" goal
> 
> This is exactly why the new code was developed. It is going into a 
> production web site that is targeted at routing for trucks and trucks 
> typically have more restrictions the cars because of things like low 
> bridges, weight restrictions, width restrictions, low maneuverability 
> areas, tight turns, legal restrictions, etc, etc.
> 
> Also the trsp code is Boost free so it might be a little easier to 
> follow how it works if you plan to dig into the guts of it. ;-)
> 
> Regards,
>    -Steve
> 
> > Regards
> >
> > max
> >
> > On Wed, 11 Jan 2012 10:52:54 -0500
> > Stephen Woodbridge<woodbri at swoodbridge.com>  wrote:
> > ...
> >> The good news:
> >>
> >> If you need to support multiple turn restrictions as just developed
> >> and added an new branch to git:
> >>     https://github.com/pgRouting/pgrouting/tree/trsp
> >> Sorry there is no doc for this yet, but it works and is only
> >> marginally slow than Dijkstra.
> >>
> > ...
> >
> > _______________________________________________
> > Pgrouting-users mailing list
> > Pgrouting-users at lists.osgeo.org
> > http://lists.osgeo.org/mailman/listinfo/pgrouting-users
> 
> _______________________________________________
> Pgrouting-users mailing list
> Pgrouting-users at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pgrouting-users



More information about the Pgrouting-users mailing list