CGAL users discussion list

["Laurent Rineau (CGAL/GeometryFactory)" <>]

Le Tuesday 12 November 2013 18:29:20 

 a écrit :
> Enough context, here are the actual questions :
> 
> * is the constrained delaunay property too weak to satisfy the mathematical
> requirements of the query logics in Point_set_2 ? (so that their is no point
> hoping to plug the later on the former)

Yes, the properties of a constrained triangulation are too weak for the logic 
of Point_set_2. Even if the code of Point_set_2 was adapted, there 
performance 
of queries would be guaranteed to be similar to what you get with a Delaunay 
triangulation.

> * is there a way I could adapt / extend the library and contribute a way to
> answer queries on Constrained_Delaunay_2 instances

In the package "2D Conforming Triangulations and Meshes" of CGAL, there is a 
section about conforming triangulations:
  http://doc.cgal.org/latest/Mesh_2/index.html

Using that tool, you can transform a CDT 
(Constrained_Delaunay_triangulation_2) to a CDT that has more vertices, and 
is 
mathematically a Delaunay triangulation. That might help you using the 
Point_set_2 algorithms on your data.

Can we see a screenshot of the type of isolines you have? Are your typical 
CDTs far from having the Delaunay property?

> * is there other accelerating query structures usable for constrained /not-
> really-Delaunay/ triangulations ?

You should tell us more about the type of queries you need.

-- 
Laurent Rineau, PhD
R&D Engineer at GeometryFactory           http://www.geometryfactory.com/
Release Manager of the CGAL Project       http://www.cgal.org/