Using itkwidgets with PyVista

Note

As of version 0.32.0, itkwidgets does not support Jupyterlab 3. Attempting to run the following will return a Model not found error within jupyterlab.

Track the progress of this in Issue 405.

PyVista has an interface for visualizing plots in Jupyter. The pyvista.PlotterITK class allows you interactively visualize a mesh within a jupyter notebook. For those who prefer plotting within jupyter, this is an great way of visualizing using VTK and pyvista.

Special thanks to @thewtex for the itkwidgets library.

Installation

To use PlotterITK you’ll need to install itkwidgets>=0.25.2. Follow the installation steps here.

You can install everything with pip if you prefer not using conda, but be sure your juptyerlab is up-to-date. If you encounter problems, uninstall and reinstall jupyterlab using pip.

Example Plotting with ITKwidgets

The following example shows how to create a simple plot that shows a simple sphere.

import pyvista as pv

# create a mesh and identify some scalars you wish to plot
mesh = pv.Sphere()
z = mesh.points[:, 2]

# Plot using the ITKplotter
pl = pv.PlotterITK()
pl.add_mesh(mesh, scalars=z, smooth_shading=True)
pl.show(True)
../../_images/itk_plotting_sphere.png

ITKwidgets with pyvista

For convenience, figures can also be plotted using the plot_itk function:

import pyvista as pv

# create a mesh and identify some scalars you wish to plot
mesh = pv.Sphere()
z = mesh.points[:, 2]

# Plot using the itkwidget
pv.plot_itk(mesh, scalars=z)

Attributes

background_color

Return the background color of the plotter.

camera_position

Return camera position of the plotter as a list.

Methods

add_mesh(mesh[, color, scalars, opacity, …])

Add a PyVista/VTK mesh or dataset.

add_points(points[, color, point_size])

Add points to plotter.

show([ui_collapsed, rotate, show_bounds])

Show itkwidgets plotter in cell output.
class pyvista.PlotterITK(**kwargs)

Bases: object

ITKwidgets plotter.

Used for plotting interactively within a jupyter notebook. Requires itkwidgets>=0.25.2. For installation see:

https://github.com/InsightSoftwareConsortium/itkwidgets#installation

Examples

>>> import pyvista
>>> mesh = pyvista.Sphere()
>>> pl = pyvista.PlotterITK()  
>>> pl.add_mesh(mesh, color='w')  
>>> pl.background_color = 'k'  
>>> pl.show()
add_mesh(mesh, color=None, scalars=None, opacity=1.0, smooth_shading=False)

Add a PyVista/VTK mesh or dataset.

Adds any PyVista/VTK mesh that itkwidgets can wrap to the scene.

Parameters

  • mesh (pyvista.DataSet or pyvista.MultiBlock) – Any PyVista or VTK mesh is supported. Also, any dataset that pyvista.wrap() can handle including NumPy arrays of XYZ points.

  • color (string or 3 item list, optional, defaults to white) – Use to make the entire mesh have a single solid color. Either a string, RGB list, or hex color string. For example: color='white', color='w', color=[1, 1, 1], or color='#FFFFFF'. Color will be overridden if scalars are specified.

  • scalars (str or numpy.ndarray, optional) – Scalars used to “color” the mesh. Accepts a string name of an array that is present on the mesh or an array equal to the number of cells or the number of points in the mesh. Array should be sized as a single vector. If both color and scalars are None, then the active scalars are used.

  • opacity (float, optional) – Opacity of the mesh. If a single float value is given, it will be the global opacity of the mesh and uniformly applied everywhere - should be between 0 and 1. Default 1.0

  • smooth_shading (bool, optional) – Smooth mesh surface mesh by taking into account surface normals. Surface will appear smoother while sharp edges will still look sharp. Default False.

add_points(points, color=None, point_size=3.0)

Add points to plotter.

Parameters

  • points (np.ndarray or pyvista.DataSet) – n x 3 numpy array of points or pyvista dataset with points.

  • color (string or 3 item list, optional. Color of points (if visible)) –

    Either a string, rgb list, or hex color string. For example:

    color='white' color='w' color=[1, 1, 1] color='#FFFFFF'

  • point_size (float, optional) – Point size of any nodes in the dataset plotted. Also applicable when style=’points’. Default 3.0

Examples

Add 10 random points to the plotter

>>> add_points(np.random.random((10, 3)), 'r', 10)
property background_color

Return the background color of the plotter.

property camera_position

Return camera position of the plotter as a list.

show(ui_collapsed=True, rotate=False, show_bounds=False, **kwargs)

Show itkwidgets plotter in cell output.

Parameters

  • ui_collapsed (bool, optional) – Plot with the user interface collapsed. UI can be enabled when plotting. Default False.

  • rotate (bool, optional) – Rotate the camera around the scene. Default False. Appears to be computationally intensive.

  • show_bounds (bool, optional) – Show the bounding box. Default False

  • point_size (int, optional) – Size of the points displayed in the

Returns

viewer – ITKwidgets viewer.

Return type

itkwidgets.Viewer