Plotting Point Clouds

This example shows you how to plot point clouds using PyVista using both the 'points' and 'points_gaussian' styles.

from __future__ import annotations

import numpy as np

import pyvista as pv
from pyvista import examples

Compare the Plotting methods

First, let’s create a sample point cloud using numpy.random.random().

# Seed rng for reproducibility
rng = np.random.default_rng(seed=0)
points = rng.random((1000, 3))
points

array([[0.63696169, 0.26978671, 0.04097352],
       [0.01652764, 0.81327024, 0.91275558],
       [0.60663578, 0.72949656, 0.54362499],
       ...,
       [0.1027894 , 0.60866026, 0.43809722],
       [0.95205703, 0.12833537, 0.77821681],
       [0.02483448, 0.20765295, 0.30052863]])

Basic Plot

We can simply plot this point cloud using the convenience pyvista.plot() function.

pv.plot(points)

Plot with Scalars

That’s quite boring, so let’s spice things up by adding color. We can either use a single scalar to plot the points. For example, the z coordinates.

For fun, let’s also render the points as spheres.

pv.plot(
    points,
    scalars=points[:, 2],
    render_points_as_spheres=True,
    point_size=20,
    show_scalar_bar=False,
)

Plot with RGBA

Alternatively, we can color the point cloud using an RGBA array. This has been normalized from (0, 1), but we could have also used a numpy.uint8 array from 0-255.

rgba = points - points.min(axis=0)
rgba /= rgba.max(axis=0)
pv.plot(points, scalars=rgba, render_points_as_spheres=True, point_size=20, cpos='xy', rgba=True)

Point Cloud Plot Styles

PyVista supports the 'points_gaussian' style, which renders points as individual soft sprites. You have the option of displaying these as tight “spheres” using render_points_as_spheres=True (default), or disabling it to create softer points at the expense of render performance.

Here’s the basic plot again, but with the style as 'points_gaussian':

pv.plot(points, style='points_gaussian', opacity=0.5, point_size=15)

Here’s a plotter with four combinations of the options side-by-side so you can see for yourself the different options available when plotting these points. PyVista tries to achieve sensible defaults, but should you find these insufficient for your needs, feel free to play around with the various options and find something that works for you.

pl = pv.Plotter(shape=(2, 2))

# Standard points
actor = pl.add_points(
    points,
    style='points',
    emissive=False,
    scalars=rgba,
    rgba=True,
    point_size=10,
    ambient=0.7,
)
pl.add_text('"points" not as spheres', color='w')

# Gaussian points
pl.subplot(0, 1)
actor = pl.add_points(
    points,
    render_points_as_spheres=False,
    style='points_gaussian',
    emissive=False,
    scalars=rgba,
    rgba=True,
    opacity=0.99,
    point_size=10,
    ambient=1.0,
)
pl.add_text('"points_gaussian" not as spheres\nemissive=False', color='w')

# Gaussian points with emissive=True
pl.subplot(1, 0)
actor = pl.add_points(
    points,
    render_points_as_spheres=False,
    style='points_gaussian',
    emissive=True,
    scalars=rgba,
    rgba=True,
    point_size=10,
)
pl.add_text('"points_gaussian" not as spheres\nemissive=True', color='w')

# With render_points_as_spheres=True
pl.subplot(1, 1)
actor = pl.add_points(
    points,
    style='points_gaussian',
    render_points_as_spheres=True,
    scalars=rgba,
    rgba=True,
    point_size=10,
)
pl.add_text('"points_gaussian" as spheres', color='w')

pl.background_color = 'k'
pl.link_views()
pl.camera_position = 'xy'
pl.camera.zoom(1.2)
pl.show()

Orbit a Point Cloud

Generate a plot orbiting around a point cloud. Color based on the distance from the center of the cloud using generate_orbital_path().

cloud = examples.download_cloud_dark_matter()
scalars = np.linalg.norm(cloud.points - cloud.center, axis=1)

pl = pv.Plotter(off_screen=True)
pl.add_mesh(
    cloud,
    style='points_gaussian',
    color='#fff7c2',
    scalars=scalars,
    opacity=0.25,
    point_size=4.0,
    show_scalar_bar=False,
)
pl.background_color = 'k'
pl.show(auto_close=False)
path = pl.generate_orbital_path(n_points=36, shift=cloud.length, factor=3.0)
pl.open_gif('orbit_cloud.gif')
pl.orbit_on_path(path, write_frames=True)
pl.close()

Tags: plot