CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

As written here [1], when using search_traits_adapter Point_d is Key.
Thus the distance must operate on the tuple as the error message state.

From here [2], you have a guess of what you must use:
"Expects for the second template argument a model of the concept 
GeneralDistance. If Traits is 
CGAL::Search_traits_adapter<Key,PointPropertyMap,BaseTraits> the default 
type is 
CGAL::Distance_for_point_adapter<Key,PointPropertyMap,CGAL::Euclidean_distance<Traits> 
>, and CGAL::Euclidean_distance<Traits> otherwise."

So CGAL::Distance_for_point_adapter is what you were looking for.

Sebastien.


[1] 
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Spatial_searching_ref/Class_Search_traits_adapter.html

[2] 
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Spatial_searching_ref/Class_K_neighbor_search.html

[3] 
http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Spatial_searching_ref/Class_Distance_for_point_adapter.html

On 11/06/2012 05:22 PM, Daniel S. wrote:
> Dear CGAL Users,
>
> atm I'm using cgal to do some nearest neighbour searching.
>
> *Situation:*
> Up to now I used the sample for property maps stated here:
>
> http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Spatial_searching/Chapter_main.html
>
> ->61.3.6   Examples for Using an Arbitrary Point Type with Point Property
> Maps
> Using a Point and an Integer as Key Type
>
> Meanwhile I need to define my own metric/point and use it in the kd-Tree. So
> I try to combine the sample stated above with the example, also on this
> page, for
>
> 61.3.5   Example for User Defined Point and Distance Class
>
> *Problem:* If i use the kd Tree, it doesn't recognized, that the input are
> boost tuples (or he actually doesn't use the map ?)it throws following
> error:
>
> error C2664: 'double Distance::transformed_distance(const Point&,const
> Point&) const':
> Converting Parameter 2 from 'const boost::tuples::tuple<T0,T1>' in 'const
> Point&'
> not possible    C:\CGAL-4.0.2\include\CGAL\K_neighbor_search.h  120
>
> a little "work around" is, to wirte a function in distance.h that accepts
> tuples and knows, that the first parameter is the point. But this is
> certainly not the way to go. As it doesn't help with the rest tuple entries.
>
> *Following my simple code sample:*
>
> #include<experimental/motionPlanning2/TestingClasses/distance.h>
>
> typedef boost::tuple<Point,int>  Point_and_int;
>
> //definition of the property map
> struct My_point_property_map{
>    typedef Point value_type;
>    typedef const value_type&  reference;
>    typedef const Point_and_int&  key_type;
>    typedef boost::readable_property_map_tag category;
> };
>
> //get function for the property map
> My_point_property_map::reference
> get(My_point_property_map,My_point_property_map::key_type p)
> {return boost::get<0>(p);}
>
> typedef CGAL::Search_traits<double, Point, const double*,
> Construct_coord_iterator>         Traits_base;
> typedef
> CGAL::Search_traits_adapter<Point_and_int,My_point_property_map,Traits_base>
> Traits;
> typedef CGAL::K_neighbor_search<Traits, Distance>
> K_neighbor_search;
> typedef K_neighbor_search::Tree  Tree;
> typedef K_neighbor_search::Distance      DistanceD;
>
> class kdTreeTest3 {
> public:
>          static void testIt();
> };
>
>   void kdTreeTest3::testIt(){
>          const int D = 7;//set dimension
>          const double bound = 1.0;//and bounds;
>          const unsigned int K = 1;
>          CGAL::Random random;
>          Tree tree;
>
>          Point q1(1.0,1.0,1.0,1.0,0.0,0.0,0.0);
>          Point q2(0.5,0.5,0.5,0.0,0.0,0.0,1.0);
>
>          tree.insert(boost::make_tuple(q1,1));
>          tree.insert(boost::make_tuple(q2,2));
>
>    Point query(0.5, 0.5, 0.5,1.0,0.0,0.0,0.0);
>    DistanceD tr_dist;
>
>    // search K nearest neighbours
>    K_neighbor_search search(tree, query, K);
>    for(K_neighbor_search::iterator it = search.begin(); it != search.end();
> it++){
>      std::cout<<  " d(q, nearest neighbor)=
> "<<std::endl&lt;&lt;boost::get&lt;0>(it->first)<<"with "
>                <<  tr_dist.transformed_distance(it->second)<<  std::endl;
>
>    }
>    // search K furthest neighbour searching, with eps=0, search_nearest=false
>    K_neighbor_search search2(tree, query, K, 0.0, false);
>
>    for(K_neighbor_search::iterator it = search2.begin(); it != search2.end();
> it++){
>      std::cout<<  " d(q, furthest neighbor)=
> "<<std::endl&lt;&lt;boost::get&lt;0>(it->first)<<"with "
>                <<  tr_dist.transformed_distance(it->second)<<  std::endl;
>    }
> }
> //end code
>
> I would really appreciate some help, as im stuck on this problem. Thanks in
> advance.
>
> Greets from Germany
> Daniel
>
>
>
> --
> View this message in context: 
> http://cgal-discuss.949826.n4.nabble.com/Samples-Spatial-Searching-combining-Custom-distance-with-property-maps-tp4656160.html
> Sent from the cgal-discuss mailing list archive at Nabble.com.