Skip to content

PyVista Analysis

pyvista_analysis: Geometric analysis and attribute computation via PyVista: curvature estimation, distance computation, glyph generation, vector magnitude, and scalar statistics.
Methods pv_curvature, pv_distance, pv_glyph, pv_vector_magnitude, pv_scalar_stats, pv_extract_feature_edges
Input files .ply, .obj, .stl, .vtk, .vtu
Bundled datasets to try open3d_bunny, open3d_icp_demo, kitti_raw_2011_09_26_drive_0001, medshapenet, c3vd_endoscopy_splats, hf_objaverse_glb_shard
CLI pardesline run pyvista_analysis.<method> input.ext [--param value]

How it works

PyVista analysis tools compute derived geometric quantities:

Curvature estimates surface curvature (mean, Gaussian, min, max) at each vertex. Distance computes point-to-surface distances between two meshes. Glyphs visualize vector fields by placing oriented symbols at data points. Vector Magnitude computes \(\|\mathbf{v}\|\) for each vector in a field. Scalar Statistics provides summary statistics of scalar arrays.

Methods

pv_curvature

Curvature types: mean \(H = \frac{\kappa_1 + \kappa_2}{2}\), Gaussian \(K = \kappa_1 \kappa_2\), maximum \(\kappa_1\), minimum \(\kappa_2\).

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_curvature")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.curvature --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

Parameter Type Default Range What it does Why this default, when to change it
curvature_type string mean Curvature type: mean, gaussian, maximum, or minimum. mean curvature is the one that reads as 'bumps and dents' on an organ or a part; Gaussian separates saddles, maximum/minimum are the principal curvatures.

pv_distance

Computes the closest-point distance from each vertex to a reference surface using a KD-tree.

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_distance")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.distance --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

No tunable parameters.

pv_glyph

Creates oriented arrow glyphs at each point, scaled by a factor and oriented along a vector field.

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_glyph")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.glyph --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

Parameter Type Default Range What it does Why this default, when to change it
scale_factor float 0.01 0.0001 to 1.0 Scale factor for glyph size. 0.01 in file units draws normal arrows of 1 % of a unit object, visible without hiding the surface.
vector_name string Normals Vector array to orient glyphs (e.g. Normals). Normals is the array every mesh has after compute_normals.
scale (advanced) bool false Scale glyphs by scalar/vector magnitude. Off: all arrows the same length, so direction is what you see.
tolerance (advanced) float -1.0 -1.0 to 1.0 Merge tolerance as fraction of bbox (-1 = auto/None). -1 lets PyVista choose the glyph merge tolerance.
absolute (advanced) bool false Interpret tolerance as an absolute distance. Off: the tolerance is a fraction of the bounding box.
clamping (advanced) bool false Clamp glyph scaling to the rng range. Off: no scaling range clamp since scale is off.

pv_vector_magnitude

Computes \(\|\mathbf{v}_i\| = \sqrt{v_{ix}^2 + v_{iy}^2 + v_{iz}^2}\) for each vector and stores as a scalar.

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_vector_magnitude")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.vector_magnitude --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

Parameter Type Default Range What it does Why this default, when to change it
vector_name string Normals Name of the vector array to compute magnitude for. Normals is the array every mesh has.

pv_scalar_stats

Computes min, max, mean, median, std, and stores a normalized version of the scalar array.

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_scalar_stats")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.scalar_stats --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

Parameter Type Default Range What it does Why this default, when to change it
scalars string "" Scalar array name (empty = first available). Empty takes the first available array.

pv_extract_feature_edges

Extracts edges where the surface normal changes sharply. A feature edge is identified when the dihedral angle between adjacent faces exceeds the threshold. Also detects boundary edges (single-face adjacency) and non-manifold edges (three or more face adjacency).

import pardesline as pl

obj = pl.load("input.ply")
result = obj.run("pyvista_analysis.pv_extract_feature_edges")  # catalog defaults
result.provenance[-1]  # the step, every parameter used, duration, version

Output: .ply. Script: python -m pardesline.ops.pyvista_analysis.extract_feature_edges --input ... --output ....

Defaults validated on: generated meshes in tests/test_ops_smoke.py

Parameter Type Default Range What it does Why this default, when to change it
feature_angle float 30.0 0.0 to 180.0 Dihedral angle threshold for feature edge detection (degrees). 30 degrees is VTK's default dihedral threshold: it catches machined edges and misses the facets of a smooth organ.
boundary_edges string true Include boundary edges (true/false). true: open boundaries are what you look for on a scan with holes.
non_manifold_edges string true Include non-manifold edges (true/false). true: non-manifold edges are the defects to fix before printing or simulation.
manifold_edges string false Include regular manifold edges (true/false). false: regular edges would be every edge of the mesh.
feature_edges (advanced) bool true Include feature (sharp dihedral-angle) edges. On: sharp edges are included.
clear_data (advanced) bool false Drop point/cell data arrays from the output edges. Off: keep the scalar arrays on the edges.

Generated from src/pardesline/catalog/data/pyvista_analysis.json and _defaults.json on 2026-10-10.