CGAL users discussion list

[Johannes Otepka <>]

Sophea,

With the existing code this cannot be done effeciently, since the 
K-Nearest neighbor object stores point copies. You have to adapt the 
Orthogonal_k_neighbor_search class in such a way that is stores point 
pointers (or point references). Not vey difficult to do.... After that 
the index can be computed by simple pointer arithmetic.

If you are interested, I can send my code which contains those changes 
and some additional <ende?lp=ende&p=/gQPUD&search=additional> 
optimisation <ende?lp=ende&p=/gQPUD&search=optimisation>s.
Cheers,
  Johannes

Chansophea Chuon wrote:

> Dear All:
>
> Suppose that we pass a vector of points(std::vector<Point_3 
> <http://www.cgal.org/Manual/3.2/doc_html/cgal_manual/Kernel_d_ref/Class_Point_d.html#Cross_link_anchor_361>> 
> points) to the (Tree tree(points.begin(), points.end()))
>
> Then we find the K-Nearest neighbors
>
> Neighbor_search search(tree, query, N);
> // This should sort all N points by increasing distance from origin
>
> for(Neighbor_search::iterator it = search.begin(); it != search.end(); 
> ++it){
>     std::cout << it->first << " "<< std::sqrt 
> <http://www.cgal.org/Manual/3.2/doc_html/cgal_manual/NumberTypeSupport_ref/Function_sqrt.html#Cross_link_anchor_1176>(it->second) 
> << std::endl;
> }
>
> We get the K-Nearest neighbors points and their distances. How to get 
> the index of the K
> -Nearest neighbors in the vector effeciently? Does the 
> Neighbor_search::iterator provide the indexes?
>
> Best regards,
>
> Sophea
>
> ------------------------------------------------------------------------
> Don't get soaked. Take a quick peek at the forecast 
> <http://tools.search.yahoo.com/shortcuts/?fr=oni_on_mail&#news>
> with theYahoo! Search weather shortcut. 
> <http://tools.search.yahoo.com/shortcuts/?fr=oni_on_mail&#news>