CGAL users discussion list

[Laurent Saboret <>]

Hi Marc,

The "Incorrect Matrix" error below looks like a bug. I never got this error so far.

Let's be sure about your configuration:

* What is your OS ? Compiler ? Amount of physical RAM ? Amount of virtual memory ?
* Which version of TAUCS are you using? The latest version is "TAUS for CGAL 3.5" at https://gforge.inria.fr/frs/?group_id=52.
* Do you get the same error in Debug and Release mode?
* Do you get the same error with other point sets?

Best regards,
Laurent Saboret


-------- Message original --------

Sujet:	[cgal-discuss] Re: CGAL :: Poisson Surface Reconstruction
Date :	Mon, 8 Aug 2011 07:28:31 -0700 (PDT)
De :	Marc Klopfenstein
Répondre à :
Pour :

Hi Laurent,

I turned on traces and have this results:

*CONSOL OUTPUT*

Number of point : 502573 (own cout)
Creates Poisson triangulation…
Creates Poisson triangulation: 147.532 seconds, 189Mb allocates
Delaunay refinement…
Calls Delaunay_refinement(radius_edge_ratio_bound=2.500000,
cell_radius_bound=0.119956, max_vertices= 10000000, enlarge_ration=
1.500000)
Calls poisson_refine_triangulation()
	190Mb allocated 
	Creates queue
	Max allocation in Delaunay refinement = 565Mb
	252 Mb allocated
End of poisson_refine_triangulation()
End of Delaunay_refinement()
Delaunay refinement: added 279908 Steiner points, 1829.39 seconds, 252 Mb
allocated
Solve Poisson equation with normalized divergence…
Calls solve_poisson()
	252 Mb allocated
	Creates matrix…
Creates matrix: done (1536.51s)
	474Mb allocated
	Solve sparse linear system…
Error: cannot solve Poisson equation

*TAUCS LOG*

taucs_linsolve: preparing to factor
taucs_linsolve: ordering (llt=1, lu=0, ordering=metis)
taucs_linsolve: ordering time 2.18e+001 seconds (2.17e+001 seconds CPU time)
taucs_linsolve: starting LLT factorization
taucs_linsolve: pre-factorization permuting of A
taucs_linsolve: starting IC LLT factorization
taucs_linsolve: starting IC LLT MF factorization
		Elimination tree depth is 7020
		Symbolic Analysis of LL^T: 1.67e+008 nonzeros, 2.22e+011 flops, 1.50e+009
bytes in L
		Relaxed  Analysis of LL^T: 1.84e+008 nonzeros, 2.27e+011 flops, 1.71e+009
bytes in L
		Symbolic Analysis            =    398.503 seconds (398.422 cpu)
	Incorrect Matrix



Sorry I can't enclose the Bimba point set because the XYZ file is too large
but the procedure to build it is explained in a previous post.

I have add this point set simplification in the example to make my test:
points.erase(CGAL::grid_simplify_point_set(points.begin(), points.end(),
cell_size),points.end());
std::cout<<"Nombre de point restant"<<points.size()<<"\n";

Best regards,
Marc


--
View this message in context: http://cgal-discuss.949826.n4.nabble.com/CGAL-Poisson-Surface-Reconstruction-tp3616368p3727130.html
Sent from the cgal-discuss mailing list archive at Nabble.com.

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