Form element-wise list from scalar and matrix

Question

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.

PaulS · Accepted Answer · 2024-11-26 00:47:54Z

2

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]]])

edited Nov 26, 2024 at 0:47

answered Nov 26, 2024 at 0:41

PaulS

27.1k3 gold badges18 silver badges40 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

hpaulj · Accepted Answer · 2024-11-26 00:52:00Z

0

The stack is fine. But since s needs to be repeated, a good alternative is to create a zero array of the desired shape, and allow broadcasting to fill in the columns:

In [2]: s = np.asarray(5)
   ...: 
   ...: m = np.asarray([[1,2],[3,4]])

In [3]: res = np.zeros((m.shape)+(2,)); res
Out[3]: 
array([[[0., 0.],
        [0., 0.]],

       [[0., 0.],
        [0., 0.]]])

In [4]: res[:,:,0] = s; res[:,:,1] = m

In [5]: res
Out[5]: 
array([[[5., 1.],
        [5., 2.]],

       [[5., 3.],
        [5., 4.]]])

answered Nov 26, 2024 at 0:52

hpaulj

233k14 gold badges260 silver badges392 bronze badges

Collectives™ on Stack Overflow

Form element-wise list from scalar and matrix

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related