CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

The code responsible for providing the approximation points is in:
include/CGAL/Hyperbola_2.h

line 266 : void generate_points

At least 100 points should be generated per hyperbola.

Try switching to Exact_predicates_inexact_constructions_kernel as Kernel.
If this still does not work try printing the points in the
aforementioned function to check that you don't have a bug in
your drawing code.

About the error with MP_FLoat, you should have used 
CGAL::Quotient<CGAL::MP_Float> (which is not recommended if you
have gmp on your machine as CGAL::Gmpq is a better alternative
for non-filtered rational computations).

Sebastien.

On 02/18/2015 05:32 AM, Narusaki wrote:
> Hi everyone,
>
> I'm totally new to CGAL and recently I want to use the Apollonius Graph in
> the library to construct the weighted Voronoi diagram under CGAL 4.5 and
> Microsoft Visual Studio 2010. For the rendering purpose I define a stream
> class whose main part is the operator<< with Segment_2 as parameter, and I
> use it as the parameter of the draw_dual_edge() method of the
> Apollonius_graph_2 class. However, I found the weird thing as follow:
>
> If I use the following two sites (in the format of (x, y, w) where x and y
> are coordinates and w is weight):
> (0.80410481369434716 0.89234204587973531 0)
> (0.064896832669279911 -0.51119635852810019 -0.72435163677780512)
>
> in which case the diagram should be a hyperbola, it shows like three rays as
> follow:
> <http://cgal-discuss.949826.n4.nabble.com/file/n4660456/p3.png>
>
> if I multiple each x, y and w by 10^4, it shows as follow, with a reasonable
> hyperbola and an extra ray:
> <http://cgal-discuss.949826.n4.nabble.com/file/n4660456/p1.png>
>
> Now I have two questions:
>
> 1. I don't know too much about this but I guess it's sth about the accuracy.
> Here I'm just using the double type as follow:
>
> typedef double NT;
> typedef CGAL::Simple_cartesian<NT> K;
>
> I try to use the MP_float type, the draw_dual_edge() method of
> Apollonius_graph_2 seems to have some compiling error with the division(/)
> operator as follow:
>
> error C2676: binary '/' : 'CGAL::MP_Float' does not define this operator or
> a conversion to a type acceptable to the predefined operator
>
> So is it an accuracy issue and what should I do about it?
>
> 2. Why there is an extra ray starting from the center of this branch of
> hyperbola?
>
>
>
> --
> View this message in context: 
> http://cgal-discuss.949826.n4.nabble.com/Dual-edge-of-Apollonius-Graph-tp4660456.html
> Sent from the cgal-discuss mailing list archive at Nabble.com.