Clip Volume Widget

If you have a structured dataset like a pyvista.ImageData or pyvista.RectilinearGrid, you can clip it using the pyvista.Plotter.add_volume_clip_plane() widget to better see the internal structure of the dataset.

Create the Dataset

Create a dense pyvista.ImageData with dimensions (200, 200, 200) and set the active scalars to distance from the center of the grid.

import numpy as np

import pyvista as pv

grid = pv.ImageData(dimensions=(200, 200, 200))
grid['scalars'] = np.linalg.norm(grid.center - grid.points, axis=1)
grid

Header	Data Arrays
ImageData Information N Cells 7880599 N Points 8000000 X Bounds 0.000e+00, 1.990e+02 Y Bounds 0.000e+00, 1.990e+02 Z Bounds 0.000e+00, 1.990e+02 Dimensions 200, 200, 200 Spacing 1.000e+00, 1.000e+00, 1.000e+00 N Arrays 1	Name Field Type N Comp Min Max scalars Points float64 1 8.660e-01 1.723e+02

Generate the Opacity Array

Create a banded opacity array such that our dataset shows “rings” at certain values. Have this increase such that higher values (values farther away from the center) are more opaque.

opacity = np.zeros(100)
opacity[::10] = np.geomspace(0.01, 0.75, 10)

Plot a Single Clip Plane Dataset

Plot the volume with a single clip plane.

Reverse the opacity array such that portions closer to the center are more opaque.

pl = pv.Plotter()
pl.add_volume_clip_plane(grid, normal='-x', opacity=opacity[::-1], cmap='magma')
pl.show()

Plot Multiple Clip Planes

Plot the dataset using the pyvista.Plotter.add_volume_clip_plane() with the output from pyvista.Plotter.add_volume() Enable constant interaction by setting the interaction_event to 'always'.

Disable the arrows to make the plot a bit clearer and flip the opacity array.

pl = pv.Plotter()
vol = pl.add_volume(grid, opacity=opacity)
vol.prop.interpolation_type = 'linear'
pl.add_volume_clip_plane(
    vol,
    normal='-x',
    interaction_event='always',
    normal_rotation=False,
)
pl.add_volume_clip_plane(
    vol,
    normal='-y',
    interaction_event='always',
    normal_rotation=False,
)
pl.show()