Clipping with Planes & Boxes

Clip/cut any dataset using using planes or boxes.

import pyvista as pv
from pyvista import examples

Clip with Plane

Clip any dataset by a user defined plane using the pyvista.DataSetFilters.clip() filter

dataset = examples.download_bunny_coarse()
dataset

Header	Data Arrays
PolyData Information N Cells 1000 N Points 872 N Strips 0 X Bounds -1.316e-01, 1.802e-01 Y Bounds -1.205e-01, 1.877e-01 Z Bounds -1.430e-01, 9.851e-02 N Arrays 1	Name Field Type N Comp Min Max Normals Points float32 3 -9.998e-01 9.977e-01

help(dataset.clip)

Help on method clip in module pyvista.core.filters.data_set:

clip(normal='x', origin=None, invert=True, value=0.0, inplace=False, return_clipped=False, progress_bar=False, crinkle=False) method of pyvista.core.pointset.PolyData instance
    Clip a dataset by a plane by specifying the origin and normal.

    If no parameters are given the clip will occur in the center
    of that dataset.

    Parameters
    ----------
    normal : tuple(float) or str, default: 'x'
        Length 3 tuple for the normal vector direction. Can also
        be specified as a string conventional direction such as
        ``'x'`` for ``(1,0,0)`` or ``'-x'`` for ``(-1,0,0)``, etc.

    origin : tuple(float), optional
        The center ``(x,y,z)`` coordinate of the plane on which the clip
        occurs. The default is the center of the dataset.

    invert : bool, optional
        Flag on whether to flip/invert the clip.

    value : float, optional
        Set the clipping value along the normal direction.
        The default value is 0.0.

    inplace : bool, optional
        Updates mesh in-place.

    return_clipped : bool, optional
        Return both unclipped and clipped parts of the dataset.

    progress_bar : bool, optional
        Display a progress bar to indicate progress.

    crinkle : bool, optional
        Crinkle the clip by extracting the entire cells along the
        clip. This adds the ``"cell_ids"`` array to the ``cell_data``
        attribute that tracks the original cell IDs of the original
        dataset.

    Returns
    -------
    pyvista.PolyData or tuple(pyvista.PolyData)
        Clipped mesh when ``return_clipped=False``,
        otherwise a tuple containing the unclipped and clipped datasets.

    Examples
    --------
    Clip a cube along the +X direction.  ``triangulate`` is used as
    the cube is initially composed of quadrilateral faces and
    subdivide only works on triangles.

    >>> import pyvista as pv
    >>> cube = pv.Cube().triangulate().subdivide(3)
    >>> clipped_cube = cube.clip()
    >>> clipped_cube.plot()

    Clip a cube in the +Z direction.  This leaves half a cube
    below the XY plane.

    >>> import pyvista as pv
    >>> cube = pv.Cube().triangulate().subdivide(3)
    >>> clipped_cube = cube.clip('z')
    >>> clipped_cube.plot()

    See :ref:`clip_with_surface_example` for more examples using this filter.

Perform a clip with a Y axis normal

clipped = dataset.clip('y', invert=False)
clipped

Header	Data Arrays
PolyData Information N Cells 427 N Points 248 N Strips 0 X Bounds -1.316e-01, 1.326e-01 Y Bounds 3.360e-02, 1.877e-01 Z Bounds -1.430e-01, 8.721e-02 N Arrays 1	Name Field Type N Comp Min Max Normals Points float32 3 -9.927e-01 9.977e-01

p = pv.Plotter()
p.add_mesh(dataset, style='wireframe', color='blue', label='Input')
p.add_mesh(clipped, label='Clipped')
p.add_legend()
p.camera_position = [(0.24, 0.32, 0.7), (0.02, 0.03, -0.02), (-0.12, 0.93, -0.34)]
p.show()

Clip with Bounds

Clip any dataset by a set of XYZ bounds using the pyvista.DataSetFilters.clip_box() filter.

dataset = examples.download_office()

help(dataset.clip_box)

Help on method clip_box in module pyvista.core.filters.data_set:

clip_box(bounds=None, invert=True, factor=0.35, progress_bar=False, merge_points=True, crinkle=False) method of pyvista.core.pointset.StructuredGrid instance
    Clip a dataset by a bounding box defined by the bounds.

    If no bounds are given, a corner of the dataset bounds will be removed.

    Parameters
    ----------
    bounds : tuple(float), optional
        Length 6 sequence of floats: (xmin, xmax, ymin, ymax, zmin, zmax).
        Length 3 sequence of floats: distances from the min coordinate of
        of the input mesh. Single float value: uniform distance from the
        min coordinate. Length 12 sequence of length 3 sequence of floats:
        a plane collection (normal, center, ...).
        :class:`pyvista.PolyData`: if a poly mesh is passed that represents
        a box with 6 faces that all form a standard box, then planes will
        be extracted from the box to define the clipping region.

    invert : bool, optional
        Flag on whether to flip/invert the clip.

    factor : float, optional
        If bounds are not given this is the factor along each axis to
        extract the default box.

    progress_bar : bool, optional
        Display a progress bar to indicate progress.

    merge_points : bool, optional
        If ``True`` (default), coinciding points of independently
        defined mesh elements will be merged.

    crinkle : bool, optional
        Crinkle the clip by extracting the entire cells along the
        clip. This adds the ``"cell_ids"`` array to the ``cell_data``
        attribute that tracks the original cell IDs of the original
        dataset.

    Returns
    -------
    pyvista.UnstructuredGrid
        Clipped dataset.

    Examples
    --------
    Clip a corner of a cube.  The bounds of a cube are normally
    ``[-0.5, 0.5, -0.5, 0.5, -0.5, 0.5]``, and this removes 1/8 of
    the cube's surface.

    >>> import pyvista as pv
    >>> cube = pv.Cube().triangulate().subdivide(3)
    >>> clipped_cube = cube.clip_box([0, 1, 0, 1, 0, 1])
    >>> clipped_cube.plot()

    See :ref:`clip_with_plane_box_example` for more examples using this filter.

Clip the dataset with a bounding box defined by the values in bounds (xmin, xmax, ymin, ymax, zmin, zmax)

bounds = [2, 4.5, 2, 4.5, 1, 3]
clipped = dataset.clip_box(bounds)
clipped

Header	Data Arrays
UnstructuredGrid Information N Cells 38690 N Points 9438 X Bounds 1.000e-02, 4.500e+00 Y Bounds 1.000e-02, 4.500e+00 Z Bounds 1.000e-02, 2.500e+00 N Arrays 2	Name Field Type N Comp Min Max scalars Points float32 1 -3.870e+00 7.186e-01 vectors Points float32 3 -6.729e-01 4.289e-01

p = pv.Plotter()
p.add_mesh(dataset, style='wireframe', color='blue', label='Input')
p.add_mesh(clipped, label='Clipped')
p.add_legend()
p.show()

Clip with Rotated Box

Clip any dataset by an arbitrarily rotated solid box using the pyvista.DataSetFilters.clip_box() filter.

mesh = examples.load_airplane()

# Use `pv.Box()` or `pv.Cube()` to create a region of interest
roi = pv.Cube(center=(0.9e3, 0.2e3, mesh.center[2]), x_length=500, y_length=500, z_length=500)
roi.rotate_z(33, inplace=True)

p = pv.Plotter()
p.add_mesh(roi, opacity=0.75, color="red")
p.add_mesh(mesh, opacity=0.5)
p.show()

Run the box clipping algorithm with the defined box geometry

extracted = mesh.clip_box(roi, invert=False)

p = pv.Plotter(shape=(1, 2))
p.add_mesh(roi, opacity=0.75, color="red")
p.add_mesh(mesh)
p.subplot(0, 1)
p.add_mesh(extracted)
p.add_mesh(roi, opacity=0.75, color="red")
p.link_views()
p.view_isometric()
p.show()

Crinkled Clipping

Crinkled clipping is useful if you don’t want the clip filter to truly clip cells on the boundary, but want to preserve the input cell structure and to pass the entire cell on through the boundary.

This option is available for pyvista.DataSetFilters.clip(), pyvista.DataSetFilters.clip_box(), and pyvista.DataSetFilters.clip_sruface(), but not available when clipping by scalar in pyvista.DataSetFilters.clip_scalar().

# Input mesh
mesh = pv.Wavelet()

Define clipping plane

normal = (1, 1, 1)
plane = pv.Plane(i_size=30, j_size=30, direction=normal)

Perform a standard clip

clipped = mesh.clip(normal=normal)

Perform a crinkled clip to compare

crinkled = mesh.clip(normal=normal, crinkle=True)

Plot comparison

p = pv.Plotter(shape=(1, 2))
p.add_mesh(clipped, show_edges=True)
p.add_mesh(plane.extract_feature_edges(), color='r')
p.subplot(0, 1)
p.add_mesh(crinkled, show_edges=True)
p.add_mesh(plane.extract_feature_edges(), color='r')
p.link_views()
p.show()