Debian Codegraph @

examples/01-filter/connectivity.py

raw
1 """
2 .. _connectivity_example:
3
4 Connectivity
5 ~~~~~~~~~~~~
6
7 This example highlights some applications of the
8 :func:`~pyvista.DataSetFilters.connectivity` filter.
9
10 """
11
12 # %%
13 # Remove Noisy Isosurfaces
14 # ~~~~~~~~~~~~~~~~~~~~~~~~
15 #
16 # Use connectivity to remove noisy isosurfaces.
17 #
18 # This section is similar to `this VTK example <https://kitware.github.io/vtk-examples/site/Python/VisualizationAlgorithms/PineRootConnectivity/>`__.
19
20 # sphinx_gallery_thumbnail_number = 2
21 import numpy as np
22 import pyvista as pv
23 from pyvista import examples
24
25 # %%
26 # Load a dataset with noisy isosurfaces.
27 pine_roots = examples.download_pine_roots()
28
29 # Plot the raw data
30 cpos = pv.CameraPosition(
31 position=(40.6018, -280.533, 47.0172),
32 focal_point=(40.6018, 37.2813, 50.1953),
33 viewup=(0.0, 0.0, 1.0),
34 )
35 pl = pv.Plotter()
36 pl.add_mesh(pine_roots, color='#965434')
37 pl.add_mesh(pine_roots.outline())
38 pl.show(cpos=cpos)
39
40 # %%
41 # The plotted mesh is very noisy. We can extract the largest connected
42 # isosurface using the ``'largest'`` ``extraction_mode`` of the
43 # :func:`~pyvista.DataSetFilters.connectivity` filter. Equivalently,
44 # :func:`~pyvista.DataSetFilters.extract_largest` may also be used.
45
46 # Grab the largest connected volume present
47 largest = pine_roots.connectivity('largest')
48 # or: largest = mesh.extract_largest()
49
50 pl = pv.Plotter()
51 pl.add_mesh(largest, color='#965434')
52 pl.add_mesh(pine_roots.outline())
53 pl.camera_position = cpos
54 pl.show()
55
56
57 # %%
58 # Extract Small Regions
59 # ~~~~~~~~~~~~~~~~~~~~~
60 #
61 # Use connectivity to extract the smaller 'noisy' regions that were
62 # removed in the remove noisy isosurfaces example above.
63 #
64 # First, get a list of all region ids.
65 all_regions = pine_roots.connectivity('all')
66 region_ids = np.unique(all_regions['RegionId'])
67
68 # %%
69 # Since the region IDs are sorted in descending order (by cell count),
70 # we can extract all regions *except* for the largest one using the
71 # ``'specified'`` ``extraction_mode`` of the :func:`~pyvista.DataSetFilters.connectivity`
72 # filter.
73 noise_region_ids = region_ids[1::] # All region ids except '0'
74 noise = pine_roots.connectivity('specified', noise_region_ids)
75
76 # %%
77 # Plot the noisy regions using :meth:`~pyvista.DataSetFilters.color_labels`.
78 # For context, also plot the largest region.
79 pl = pv.Plotter()
80 pl.add_mesh(noise.color_labels())
81 pl.add_mesh(largest, color='lightgray', opacity=0.1)
82 pl.add_mesh(pine_roots.outline())
83 pl.camera_position = cpos
84 pl.show()
85
86
87 # %%
88 # Label Disconnected Regions
89 # ~~~~~~~~~~~~~~~~~~~~~~~~~~
90 #
91 # Use connectivity to label all disconnected regions.
92 #
93 # This section is similar to `this VTK example <https://examples.vtk.org/site/Cxx/PolyData/ColorDisconnectedRegionsDemo/>`__.
94 #
95 # First, load a dataset with disconnected regions.
96 mesh = examples.download_foot_bones()
97
98 # %%
99 # Extract all regions.
100 conn = mesh.connectivity('all')
101
102 # %%
103 # Plot the labeled regions using :meth:`~pyvista.DataSetFilters.color_labels`.
104
105 colored = conn.color_labels()
106 cpos = pv.CameraPosition(
107 position=(10.5, 12.2, 18.3), focal_point=(0.0, 0.0, 0.0), viewup=(0.0, 1.0, 0.0)
108 )
109 colored.plot(cpos=cpos)
110
111
112 # %%
113 # Extract Regions From Seed Points
114 # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
115 #
116 # Use connectivity to extract regions of interest using scalar data and
117 # seed points.
118 #
119 # First, create a dataset with salient features. Here, we create hills
120 # and use curvature to define its peaks and valleys.
121 mesh = pv.ParametricRandomHills()
122 mesh['Curvature'] = mesh.curvature()
123
124 # %%
125 # Visualize the peaks and valleys.
126 # Peaks have large positive curvature (i.e. are convex).
127 # Valleys have large negative curvature (i.e. are concave).
128 # Flat regions have curvature close to zero.
129 mesh.plot(
130 clim=[-0.5, 0.5],
131 cmap='bwr',
132 below_color='blue',
133 above_color='red',
134 )
135
136 # %%
137 # Extract a region of interest using the
138 # ``'point_seed'`` ``extraction_mode`` of the :func:`~pyvista.DataSetFilters.connectivity`
139 # filter. Let's extract the steepest peak using a seed point where the
140 # curvature is maximized.
141
142 # Get seed point
143 peak_point_id = np.argmax(mesh['Curvature'])
144
145 # Define a scalar range of the region to extract
146 data_min, data_max = mesh.get_data_range()
147 peak_range = [0.2, data_max] # Peak if curvature > 0.2
148
149 peak_mesh = mesh.connectivity('point_seed', peak_point_id, scalar_range=peak_range)
150
151 # %%
152 # Let's also extract the closest valley to the steepest peak using the
153 # ``'closest'`` ``extraction_mode`` of the :func:`~pyvista.DataSetFilters.connectivity`
154 # filter.
155 valley_range = [data_min, -0.2] # Valley if curvature < -0.2
156 peak_point = mesh.points[peak_point_id]
157 valley_mesh = mesh.connectivity('closest', peak_point, scalar_range=valley_range)
158
159 # %%
160 # Plot extracted regions.
161 pl = pv.Plotter()
162 _ = pl.add_mesh(mesh, style='wireframe', color='lightgray')
163 _ = pl.add_mesh(peak_mesh, color='red', label='Steepest Peak')
164 _ = pl.add_mesh(valley_mesh, color='blue', label='Closest Valley')
165 _ = pl.add_legend()
166 pl.show()
167 # %%
168 # .. tags:: filter