Extending diagonalization in numpy to more than 2 dimensions

Question

I have a vector

>>> import numpy as np
>>> v = np.array([1, 2, 3]) 

in numpy. One uses np.diag to create a diagona