[vtkusers] Delaunay3D Unable to factor linear system?

[Kevin H. Hobbs]

On 04/10/2010 04:54 PM, David Doria wrote:
> Can anyone explain why this code:
> http://www.vtk.org/Wiki/VTK/Examples/ConvexHull_vtkDelaunay3D_RealData
> 
> Works on Input.vtp (a sphere with 50 points) , but not HardInput.vtp
> (a sphere with 290 points)
> 
> (data available here http://www.rpi.edu/~doriad/VTK_List/Delaunay3D/)?
> 
> I am getting "Unable to factor linear system" when I try it with HardInput.vtp.
> 
> Thanks,
> 
> David
> 

My understanding is that numerical problems occur when there are
triangles with very acute angles.

For your HardInput.vtp I suspect that the Delaunay filter is having
trouble around the poles.

I believe there is a note in the documentation that randomizing the
order of input points often helps.

To play around with that I sent the output of ParaView's sphere source,
to a clean filter with a tolerance, then to the mask points filter in
random mode. This didn't really randomize the points but it more evenly
distributed them. I was then able to run the delaunay filter without the
warning but with far more points than are in the HardInput.vtp.

I suspect a better (than random) thing to do would be to divide the
points up into regions and add them to the delaunay triangulation in a
sort of round robin by region way.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20100415/f6f7ac1a/attachment.pgp>