Subdivide Cells

Download Python source code | Download Jupyter notebook

Increase the number of triangles in a single, connected triangular mesh.

The pyvista.PolyDataFilters.subdivide() filter utilizes three different subdivision algorithms to subdivide a mesh’s cells: butterfly, loop, or linear.

import pyvista as pv
from pyvista import examples

First, let’s load a triangulated mesh to subdivide. We can use the pyvista.DataObjectFilters.triangulate() filter to ensure the mesh we are using is purely triangles.

mesh = examples.download_bunny_coarse().triangulate().clean()

cpos = pv.CameraPosition(
    position=(-0.02788175062966399, 0.19293295656233056, 0.4334449972621349),
    focal_point=(-0.053260899930287015, 0.08881197167521734, -9.016948161029588e-05),
    viewup=(-0.10170607813337212, 0.9686438023715356, -0.22668272496584665),
)

Now, lets do a few subdivisions with the mesh and compare the results. Below is a helper function which collects the meshes and labels for the comparison plot of the three different subdivisions.

def subdivisions(mesh, a, b):
    """Return the original mesh and its subdivisions, one row per subfilter."""
    datasets = pv.MultiBlock()
    for subfilter in ['linear', 'butterfly', 'loop']:
        datasets.append(mesh, 'Original Mesh')
        for n_subdivisions in [a, b]:
            datasets.append(
                mesh.subdivide(n_subdivisions, subfilter=subfilter),
                f'{subfilter} subdivision of {n_subdivisions}',
            )
    return datasets

Run the subdivisions for 1 and 3 levels and compare them with plot_compare(). The block names are used as labels.

datasets = subdivisions(mesh, 1, 3)

pv.plot_compare(
    datasets,
    shape=(3, 3),
    show_edges=True,
    color=True,
    cpos=cpos,
)

Static Scene

Interactive Scene

Tags: filter