trimmesh_subdivide_loop exception

[Achillx]

Describe the bug The function returns exception:

Message: iteration over non-sequence of type Mesh

Traceback:
  line 581, in trimesh_subdivide_loop, "/Users/axydis/PROJECTS/python/compas-dev/src/compas/topology/subdivision.py"
  line 12, in <module>, "/Users/axydis/PROJECTS/PhD/DDAD/src/frontend/voronoi.py"

To Reproduce Steps to reproduce the behavior:

from compas.datastructures import Mesh
from compas.topology import trimesh_subdivide_loop
from compas.topology import delaunay_from_points
from compas.topology import voronoi_from_delaunay

from compas.geometry import pointcloud_xy

points = pointcloud_xy(50, (0, 100))
faces = delaunay_from_points(points)
delaunay = Mesh.from_vertices_and_faces(points, faces)

mesh_sub = trimesh_subdivide_loop(delaunay, 1, False)

Expected behavior To return a subdivided mesh.

Desktop (please complete the following information):

• OS: MacOS Mojave v.10.14
• Python version: 2.7.15 :: Anaconda, Inc.
• Compas commit: 9f94cce0
• IDE: Atom 1.33.0 x64
• Rhino for Mac 5.5.2

[tomvanmele]

nice bug report :) will get right on it...

[tomvanmele]

there was indeed something wrong with the scheme: it only worked for closed meshes. should be fixed since 6e5e0e3

i have tested the following successfully:

from compas.datastructures import Mesh
from compas.topology import trimesh_subdivide_loop
from compas.plotters import MeshPlotter

points = [[0.0, 0.0, 0.0], [1.0, 0.0, 0.0], [0.5, 1.0, 0.0]]
faces = [[0, 1, 2]]
mesh = Mesh.from_vertices_and_faces(points, faces)

subd = trimesh_subdivide_loop(mesh, k=3)

plotter = MeshPlotter(subd, figsize=(10, 7))
plotter.draw_vertices(radius=0.1)
plotter.draw_faces()
plotter.show()

from compas.datastructures import Mesh
from compas.geometry import pointcloud_xy
from compas.topology import delaunay_from_points
from compas.topology import trimesh_subdivide_loop
from compas.plotters import MeshPlotter

points = pointcloud_xy(10, (0, 100))
faces = delaunay_from_points(points)
mesh = Mesh.from_vertices_and_faces(points, faces)

subd = trimesh_subdivide_loop(mesh, k=3)

plotter = MeshPlotter(subd, figsize=(10, 7))
plotter.draw_vertices(radius=0.1)
plotter.draw_faces()
plotter.show()

[tomvanmele]

however, higher numbers of starting points in the pointcloud seem to sometimes generate overlaps... the algorithm is also clearly not yet optimised for speed :)

[Achillx]

Yes, that fixed it!