pyvista.DataSetFilters.split_values#

Split mesh into separate sub-meshes using point or cell data.

By default, this filter generates a separate mesh for each unique value in the data array and combines them as blocks in a MultiBlock dataset. Optionally, specific values and/or ranges of values may be specified to control which values to split from the input.

This filter is a convenience method for extract_values() with split set to True by default. Refer to that filter’s documentation for more details.

Added in version 0.44.

Parameters:

valuesnumber | array_like | dict, optional

Value(s) to extract. Can be a number, an iterable of numbers, or a dictionary with numeric entries. For dict inputs, either its keys or values may be numeric, and the other field must be strings. The numeric field is used as the input for this parameter, and if split is True, the string field is used to set the block names of the returned MultiBlock.

Note

When extracting multi-component values with component_mode=multi, each value is specified as a multi-component scalar. In this case, values can be a single vector or an array of row vectors.

rangesarray_like | dict, optional

Range(s) of values to extract. Can be a single range (i.e. a sequence of two numbers in the form [lower, upper]), a sequence of ranges, or a dictionary with range entries. Any combination of values and ranges may be specified together. The endpoints of the ranges are included in the extraction. Ranges cannot be set when component_mode=multi.

For dict inputs, either its keys or values may be numeric, and the other field must be strings. The numeric field is used as the input for this parameter, and if split is True, the string field is used to set the block names of the returned MultiBlock.

Note

Use +/- infinity to specify an unlimited bound, e.g.:

[0, float('inf')] to extract values greater than or equal to zero.
[float('-inf'), 0] to extract values less than or equal to zero.

scalarsstr, optional

Name of scalars to extract with. Defaults to currently active scalars.

preferencestr, default: ‘point’

When scalars is specified, this is the preferred array type to search for in the dataset. Must be either 'point' or 'cell'.

component_modeint | ‘any’ | ‘all’ | ‘multi’, default: ‘all’

Specify the component(s) to use when scalars is a multi-component array. Has no effect when the scalars have a single component. Must be one of:

number: specify the component number as a 0-indexed integer. The selected component must have the specified value(s).
'any': any single component can have the specified value(s).
'all': all individual components must have the specified values(s).
'multi': the entire multi-component item must have the specified value.

**kwargsdict, optional

Additional keyword arguments passed to extract_values().

Returns:

pyvista.MultiBlock: Composite of split meshes with pyvista.UnstructuredGrid blocks.

See also

extract_values, split_bodies, partition

Examples

Load image with labeled regions.

>>> import numpy as np
>>> import pyvista as pv
>>> from pyvista import examples
>>> image = examples.load_channels()
>>> np.unique(image.active_scalars)
pyvista_ndarray([0, 1, 2, 3, 4])

Split the image into its separate regions. Here, we also remove the first region for visualization.

>>> multiblock = image.split_values()
>>> _ = multiblock.pop(0)  # Remove first region

Plot the regions.

>>> plot = pv.Plotter()
>>> _ = plot.add_composite(multiblock, multi_colors=True)
>>> _ = plot.show_grid()
>>> plot.show()

Static Scene

../../../_images/pyvista-DataSetFilters-split_values-1_00_00.png

Interactive Scene

Note that the block names are generic by default.

>>> multiblock.keys()
['Block-01', 'Block-02', 'Block-03', 'Block-04']

To name the output blocks, use a dictionary as input instead.

Here, we also explicitly omit the region with 0 values from the input instead of removing it from the output.

>>> labels = dict(region1=1, region2=2, region3=3, region4=4)
>>>
>>> multiblock = image.split_values(labels)
>>> multiblock.keys()
['region1', 'region2', 'region3', 'region4']

Plot the regions as separate meshes using the labels instead of plotting the MultiBlock directly.

Clear scalar data so we can color each mesh using a single color >>> _ = [block.clear_data() for block in multiblock] >>> >>> plot = pv.Plotter() >>> plot.set_color_cycler(‘default’) >>> _ = [ … plot.add_mesh(block, label=label) … for block, label in zip(multiblock, labels) … ] >>> _ = plot.add_legend() >>> plot.show()

Static Scene

../../../_images/pyvista-DataSetFilters-split_values-1_01_00.png

Interactive Scene