CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

These is indeed a problem.

with 0.01 I got one point and with 0.011 I got a segment.

I'll have a closer look at what is going on.

If Gmp is available on your machine, use CGAL::Gmpq (which is a better 
choice anyway) as Number_type and the problem should no longer appear.

S.

nsoonhui wrote:
> I found that if pixel size is less than 1( but bigger than 0), then the pixel
> size is taken as 1, regardless of what you specify.
>
> Here's the code sample that demonstrates my point:
>
>
> #include 
> #include 
> #include 
> #include 
> #include 
> #include 
>
> typedef CGAL::Quotient           Number_type;
> typedef CGAL::Cartesian             Kernel;
> typedef CGAL::Snap_rounding_traits_2     Traits;
> typedef Kernel::Segment_2                        Segment_2;
> typedef Kernel::Point_2                          Point_2;
> typedef std::list                     Segment_list_2;
> typedef std::list                       Polyline_2;
> typedef std::list                    Polyline_list_2;
>
> int main()
> {
>   Segment_list_2 seg_list;
>   Polyline_list_2 output_list;
>
>   double mag = 1;
>   seg_list.push_back(Segment_2(Point_2(0.25*mag, 0.25*mag),
> Point_2(0.75*mag, 0.75*mag)));
>
>   // Generate an iterated snap-rounding representation, where the centers of
>   // the hot pixels bear their original coordinates, using 5 kd trees:
>   CGAL::snap_rounding_2
>     (seg_list.begin(), seg_list.end(), output_list, 0.01*mag, false, false,
> 1);
>
>   int counter = 0;
>   Polyline_list_2::const_iterator iter1;
>   for (iter1 = output_list.begin(); iter1 != output_list.end(); ++iter1) {
>     std::cout << "Polyline number " << ++counter << ":\n";
>     Polyline_2::const_iterator iter2;
>     for (iter2 = iter1->begin(); iter2 != iter1->end(); ++iter2)
>       std::cout << "    (" << iter2->x() << ":" << iter2->y() << ")\n";
>   }
>
>   return(0);
> }
>
>
> The output is that there is only 1 segment with 1 point (0.51/2, 0.51/2).
>
> However if I set mag=100, then there is 1 segment with2  points( 51/2,
> 0.51/2) and (151/2, 151/2).
>
> I cna't help but feel that for mag=1, there really should be 1 segment with
> 2 points, like the case with mag=100. Am I right?
>
> --
> View this message in context: 
> http://cgal-discuss.949826.n4.nabble.com/Snap-Rounding-Only-Works-when-Pixel-Size-1-tp3340736p3340736.html
> Sent from the cgal-discuss mailing list archive at Nabble.com.