Rotations in N-Dimensions

The following code creates a rotation matrix in 2D taking the angle as 30 degrees.

import numpy as np

theta = np.radians(30)

r = np.array(( (np.cos(theta), -np.sin(theta)),
               (np.sin(theta),  np.cos(theta)) ))

print('rotation matrix:')
print(r)

Output:

rotation matrix:
[[ 0.8660254 -0.5      ]
 [ 0.5        0.8660254]]

We can add a single function that takes the angle and dimension as the input and returns a rotation matrix. I want to work on this.

Reference: https://github.com/numpy/numpy/issues/17043

[rkern]

@rgommers As noted in the NumPy issue, the functionality you point to is for rotations in 3D space, not 2D.

[davidblitz]

I was investigating general rotations in n dimensions this weekend. Based on this I build this script (together with tests): https://github.com/davidblitz/torch-rot/blob/main/torch_rot/rotations.py It is based on formula (15) from this website: https://analyticphysics.com/Higher Dimensions/Rotations in Higher Dimensions.htm

The script is written with torch functions but it can easily be changed to use only numpy. Let me know if you have any questions or comments.