I have a zero-dimensional numpy scalar s and a two-dimensional numpy matrix m. I want to form a matrix of vectors in which all the elements of m are paired with s as in the following example:
import numpy as np
s = np.asarray(5)
m = np.asarray([[1,2],[3,4]])
# Result should be as follows
array([[[5, 1],
[5, 2]],
[[5, 3],
[5, 4]]])
In other words, I want to vectorize the operation np.asarray([s, m]) element-wise at the lowest level of m. Is there an obvious way to do that for any multidimensional array m within numpy?
I'm sure this is somewhere, but I have trouble expressing it in words and cannot find it. If you can find it, please feel free to redirect me there.
A possible solution, which uses broadcast_to and stack functions to combine two arrays, s and m, into a single array along a new axis. The steps are:
First, np.broadcast_to(s, m.shape) expands the shape of array s to match that of array m without copying data.
Then, np.stack([np.broadcast_to(s, m.shape), m], axis=-1) joins the broadcasted s and m along a new last axis
np.stack([np.broadcast_to(s, m.shape), m], axis=-1)
Output:
array([[[5, 1],
[5, 2]],
[[5, 3],
[5, 4]]])