the input of pytetgen

[tftangming]

Now I only have some point saved by a .ply file, and I want to use the pytetgen to generate the tetrahedra and visualize.Is it possible?

[akaszynski]

Now I only have some point saved by a .ply file

I'm assuming you have a surface. If so, I'd follow the steps in the README to load and tetrahedralize the surface. See the sphere example, but substitute your mesh for the sphere by first loading your mesh with pyvista.read.

[banesullivan]

To be extra clear, try this:

import pyvista as pv
import tetgen

surface = pv.read('mysurface.ply')

tet = tetgen.TetGen(surface)
tet.tetrahedralize()
grid = tet.grid

grid.plot(show_edges=True)

grid.clip().plot(show_edges=True)

[banesullivan]

Once you tetrahedralize it, then you should head over to PyVista's documentation to learn how to do all the visualization: https://docs.pyvista.org/examples/index.html

[chenzhaiyu]

Hi, is it possible for this interface to generates the Delaunay tetrahedralization from points only, instead of a surface?

With C++ tetgen this functionality is there (Figure 12 of http://wias-berlin.de/software/tetgen/1.5/doc/manual/manual005.html).

To be extra clear, try this:

import pyvista as pv
import tetgen

surface = pv.read('mysurface.ply')

tet = tetgen.TetGen(surface)
tet.tetrahedralize()
grid = tet.grid

grid.plot(show_edges=True)

grid.clip().plot(show_edges=True)

Tried loading a ply file using this example with point coordinates only (no face record) but failed:

RuntimeError: Failed to tetrahedralize.
May need to repair surface by making it manifold

[banesullivan]

Hi, is it possible for this interface to generates the Delaunay tetrahedralization from points only, instead of a surface?

Nope, the input must be a manifold, triangular surface

[chenzhaiyu]

Hi, is it possible for this interface to generates the Delaunay tetrahedralization from points only, instead of a surface?

With C++ tetgen this functionality is there (Figure 12 of http://wias-berlin.de/software/tetgen/1.5/doc/manual/manual005.html).

Thanks @banesullivan. For others looking for this, scipy.spatial.Delaunay supports Delaunay tessellation in N dimensions.