Thanks to visit codestin.com
Credit goes to github.com

Skip to content

DOC: Add an example for 2D images in 3D plots #28280

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 23, 2024
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
88 changes: 88 additions & 0 deletions galleries/examples/mplot3d/imshow3d.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,88 @@
"""
===============
2D images in 3D
===============

This example demonstrates how to plot 2D color coded images (similar to
`.Axes.imshow`) as a plane in 3D.

Matplotlib does not have a native function for this. Below we build one by relying
on `.Axes3D.plot_surface`. For simplicity, there are some differences to
`.Axes.imshow`: This function does not set the aspect of the Axes, hence pixels are
not necessarily square. Also, pixel edges are on integer values rather than pixel
centers. Furthermore, many optional parameters of `.Axes.imshow` are not implemented.

Multiple calls of ``imshow3d`` use independent norms and thus different color scales
by default. If you want to have a single common color scale, you need to construct
a suitable norm beforehand and pass it to all ``imshow3d`` calls.

A fundamental limitation of the 3D plotting engine is that intersecting objects cannot
be drawn correctly. One object will always be drawn after the other. Therefore,
multiple image planes can well be used in the background as shown in this example.
But this approach is not suitable if the planes intersect.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this is a good paragraph to include, this "2.5D" renderer behavior is non-obvious, and we should mention it as often as possible IMO.

Copy link
Member

@story645 story645 May 23, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeah, I also think this is great - it explains why this needs to be a separate example and so will be nice to link to off the intersecting example (and they can all have the same tag https://matplotlib.org/devdocs/devel/tag_guidelines.html )

"""

import matplotlib.pyplot as plt
import numpy as np

from matplotlib.colors import Normalize


def imshow3d(ax, array, value_direction='z', pos=0, norm=None, cmap=None):
"""
Display a 2D array as a color-coded 2D image embedded in 3d.

The image will be in a plane perpendicular to the coordinate axis *value_direction*.

Parameters
----------
ax : Axes3D
The 3D Axes to plot into.
array : 2D numpy array
The image values.
value_direction : {'x', 'y', 'z'}
The axis normal to the image plane.
pos : float
The numeric value on the *value_direction* axis at which the image plane is
located.
norm : `~matplotlib.colors.Normalize`, default: Normalize
The normalization method used to scale scalar data. See `imshow()`.
cmap : str or `~matplotlib.colors.Colormap`, default: :rc:`image.cmap`
The Colormap instance or registered colormap name used to map scalar data
to colors.
"""
if norm is None:
norm = Normalize()
colors = plt.get_cmap(cmap)(norm(array))

if value_direction == 'x':
nz, ny = array.shape
zi, yi = np.mgrid[0:nz + 1, 0:ny + 1]
xi = np.full_like(yi, pos)
elif value_direction == 'y':
nx, nz = array.shape
xi, zi = np.mgrid[0:nx + 1, 0:nz + 1]
yi = np.full_like(zi, pos)
elif value_direction == 'z':
ny, nx = array.shape
yi, xi = np.mgrid[0:ny + 1, 0:nx + 1]
zi = np.full_like(xi, pos)
else:
raise ValueError(f"Invalid value_direction: {value_direction!r}")
ax.plot_surface(xi, yi, zi, rstride=1, cstride=1, facecolors=colors, shade=False)


fig = plt.figure()
ax = fig.add_subplot(projection='3d')
ax.set(xlabel="x", ylabel="y", zlabel="z")

nx, ny, nz = 8, 10, 5
data_xy = np.arange(ny * nx).reshape(ny, nx) + 15 * np.random.random((ny, nx))
data_yz = np.arange(nz * ny).reshape(nz, ny) + 10 * np.random.random((nz, ny))
data_zx = np.arange(nx * nz).reshape(nx, nz) + 8 * np.random.random((nx, nz))

imshow3d(ax, data_xy)
imshow3d(ax, data_yz, value_direction='x', cmap='magma')
imshow3d(ax, data_zx, value_direction='y', pos=ny, cmap='plasma')

plt.show()
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Image from the built docs:

image

It may be more impactful if you manually set the axlims so the images line up with the backing planes, but your call. Can self merge if you prefer as-is.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I slightly favor leaving it as is:

  • Less optional configuration code keeps the example more readable
  • I like that the planes are floating above the axes cube. This is a slight visual hint that the images are not tied to the cube and can theoretically be anywhere.