CGAL users discussion list

[pierre <>]

Well, sorry for being so confusing (and for the typo in your name).

To make things simple, I must keep my Vertex_handles untouched at all cost.

Actually, all I need is to find the set of proper tests for detecting and preventing any combinatorial changes from happening if I move my vertex so that I can backtrack.
But I just cannot seem to have those tests right. I must forget something ... :S

On Wed, Apr 7, 2010 at 4:26 PM, pierre <> wrote:

Hi Sebastian,

Actually I need to keep my Vertex_handles intact.

This algorithm is basically some kind of a greedy triangulation growing thing (with a lot of computations going around)
and I need to keep track of the recently-added vertices so that I can process them later on. I typically store the to-be-processed
Vertex_handles in a priority_queue from which I pop the vertex to consider.

Am afraid that using remove(V) and insert(P) will generate another handle and compromise the validity of the pointers stored in my
queue.
But thanks for your quick reply anyway. :)

   Pierre.

On Wed, Apr 7, 2010 at 4:20 PM, Sebastien Loriot (GeometryFactory) <sloriot.ml@gmail.com> wrote:

Hello,

have you tried to use functions
remove(Vertex_handle v) and insert(Point p)
to mimic the move? is it to slow?

S.

piR wrote:

Hi everyone,

This is I guess a pretty stupid question but I just cannot seem to make it
work properly.
I need to move a point within a triangulation. For that aim, I use the Vertex_handle->setpoint(Point) function a bit
recklessly.
Usually the new positions are almost the same as the previous ones (it is a
relaxation step), so it seems to work fine.
But of course, since I do not properly check whether the predicates are
still verified after the move, there are cases where my triangulation gets
corrupted.

Would you mind helping me doing this properly?
I tried to backtrack according to the result of the is_valid() functions,
but to no avail.
I swear I looked around, and only found a solution for a more restrictive
Delaunay triangulation somewhere.
Am a total CGAL newbie and this needs to work asap.
MANY THANKS!

--

You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://lists-sop.inria.fr/wws/info/cgal-discuss