CODEWITHSUNDEEP
pythonnumpyscipy
Nicolas Gervais
Nicolas Gervais

Reputation: 36704

How do I use scipy.ndimage.interpolate to randomly rotate a numpy array in 3d?

I made a square in 3 dimensions that is essentially a 3d version of this:

        [[0., 0., 0., 0., 1., 1., 1., 0.],
         [0., 0., 0., 0., 1., 1., 1., 0.],
         [0., 0., 0., 0., 1., 1., 1., 0.],
         [0., 0., 0., 0., 0., 0., 0., 0.],
         [0., 0., 0., 0., 0., 0., 0., 0.],
         [0., 0., 0., 0., 0., 0., 0., 0.],
         [0., 0., 0., 0., 0., 0., 0., 0.],
         [0., 0., 0., 0., 0., 0., 0., 0.]]

You can see a 3x3 square of ones. In 3d, in a plot, it gives this:

import numpy as np
import matplotlib.pyplot as plt

square = np.ones((8, 8, 8)) 
x, y, z = np.where(square ==1)

fig = plt.figure(figsize=(6, 6))
ax = fig.add_subplot(111, projection='3d')
ax.scatter(x, y, z, alpha=.8, s=100, ec='k', color='#2FAA75')
ax.set_xlim(-5, 10), ax.set_ylim(-5, 10), ax.set_zlim(-5, 10)
plt.show()

enter image description here

What I want is this straight square to rotate of various angles, not only 90 degrees.

I know that scipy.spatial.transform.Rotation can do that. Unfortunately, I don't know how to implement it. Expected results: imagine the cube being rotated 45 degrees with respects to the x and z axes (for example).

Upvotes: 0

Views: 399

Answers (1)

Amadan
Amadan

Reputation: 198476

the cube being rotated 45 degrees with respects to the x and z axes 

# ...
coords = np.where(square == 1)
coords = np.transpose(coords)          # get coordinates into a proper shape

rot = Rotation.from_euler('xz', [45, 45], degrees=True)  # create a rotation
coords = rot.apply(coords)             # apply the rotation
coords = np.transpose(coords)          # get coordinates back to the matplotlib shape

# ...

ax.scatter(*coords, alpha=.8, s=100, ec='k', color='#2FAA75')

randomly rotate a numpy array in 3d?

Replace the rotation with

rot = Rotation.random()

Upvotes: 2

Related Questions