CGAL users discussion list

[Stefan Witzel <>]

Hi,

Am Do., 20. Juni 2019 um 14:46 Uhr schrieb Stefan Witzel 
<>:
> Here is an obscure issue: I have two points p and q for which
> Construct_hyperbolic_segment_2 does not terminate. The points are
> approximately Hyperbolic_point_2(0.9666536103464638,-0.0141826279439945)
> and Hyperbolic_point_2(0.9666536103464638,0.0141826279439945) but I
> don't know how I can dump them exactly. They arise as vertices of a
> hyperbolic Voronoi diagram. Any suggestions?

Actually it did terminate after minutes. So this must have to do with
the exact calculations, climbing up some tower of quadratic
extensions?
I will use the occasion to ask a bit more about exact/inexact calculations:
1. My points do actually come from a quadratic extension of Q (but
from a different program); is there a good way to feed them exactly
into cgal?
2. What would be a good alternative to the time-consuming exact
calculations? If I use

  typedef CGAL::Exact_predicates_inexact_constructions_kernel K;
  typedef CGAL::Hyperbolic_Delaunay_triangulation_traits_2<K> Gt;

the program is very fast but also the outcome is very unsatisfactory.
Am I right to assume that I should use

  typedef CGAL::Hyperbolic_Delaunay_triangulation_CK_traits_2<> Gt;

? In that case, can someone indicate how to cast a
Hyperbolic_segment_2 to a segment or a circular arc? In principle I
appreciate this abstraction between interface and implementation but
in practice I find that the interface doesn't provide what I need (see
a previous discussion) so I need to access the actual implementation
and then digging into a template with three nested template parameters
is beyond me.

Best,
Stefan