Tet remeshing - add sizing field #7830
Conversation
|
At this point, remeshing can use the same sizing field as the one given to Mesh_3 for |
…ithub.com/janetournois/cgal into Tet_remeshing-with_sizing_field-jtournois # Conflicts: # Tetrahedral_remeshing/examples/Tetrahedral_remeshing/mesh_and_remesh_with_sizing.cpp # Tetrahedral_remeshing/include/CGAL/Tetrahedral_remeshing/internal/tetrahedral_adaptive_remeshing_impl.h # Tetrahedral_remeshing/include/CGAL/tetrahedral_remeshing.h
weights are 1/squared_distance_to(query point)
+ add default NP template type
…zing_field-jtournois # Conflicts: # Tetrahedral_remeshing/include/CGAL/Tetrahedral_remeshing/internal/tetrahedral_adaptive_remeshing_impl.h
…zing_field-jtournois
and use facet_edges() helper function
|
The full ctest of examples and tests for the Tetrahedral_remeshing package passes on my machine, far below the 1200 sec timeout. |
| typename Gt::Construct_cross_product_vector_3 cross_product = | ||
| gt.construct_cross_product_vector_3_object(); | ||
| typename Gt::Compute_scalar_product_3 scalar_product = | ||
| gt.compute_scalar_product_3_object(); | ||
|
|
||
| typedef typename Gt::FT FT; | ||
| typedef typename Gt::Vector_3 Vector_3; | ||
| const Vector_3 ij = vector(i, j); | ||
| const Vector_3 ik = vector(i, k); | ||
| const Vector_3 ij(i, j); |
There was a problem hiding this comment.
Important
vector was the function object Construct_vector_3 from the kernel.
Was there really a copy? That might be important to understand, and fix, for all the other implementation all over CGAL, which might use similar ways to initialize kernel objects.
| @@ -319,34 +319,101 @@ Dihedral_angle_cosine cos_dihedral_angle(const typename Gt::Point_3& i, | |||
| return Dihedral_angle_cosine(CGAL::sign(num), CGAL::square(num), sqden); | |||
| } | |||
|
|
|||
| // cosine of dihedral angle, computed from pre-computed and non-zero normals | |||
| template<typename Gt> | |||
| Dihedral_angle_cosine cos_dihedral_angle(const typename Gt::Vector_3& n1, | |||
There was a problem hiding this comment.
Dihedral_angle_cosine might be useful for other tet-related packages. Do you plan to make it more official one way or another?
There was a problem hiding this comment.
I agree it could be useful elsewhere, for example in perturb_mesh_3() and exude_mesh_3().
I had not planned to document it but we should do it once this PR is merged
|
Successfully tested in CGAL-6.0-Ic-237 (only two timeout, well done!) |
|
This pull-request was previously marked with the label |
|
/force-build:v2 |
|
There was an error while building the doc: |
|
/force-build:v2 |
|
The documentation is built. It will be available, after a few minutes, here: https://cgal.github.io/7830/v2/Manual/index.html |
sloriot
force-pushed
the
Tet_remeshing-with_sizing_field-jtournois
branch
from
89328b0
to
d257b65
Compare
lrineau
added
the
rm only: ready for master
lrineau
removed
the
rm only: ready for master
Summary of Changes
This PR adds a sizing field as input for
CGAL::tetrahedral_isotropic_remeshing(), that can be passed as second parameter, instead oftarget_edge_length.Release Management
CGAL::tetrahedral_isotropic_remeshing()#6056** reference manual
** user manual
Todo
useMesh_sizing_fieldlike the optimizers do (adapt it to be aMeshDomainField_3) and find an API for itMeshDomainField_3)