NumPy broadcasting error when trying to add 2D array and 1D array

Question

I'm trying to perform an element-wise addition between a 2D NumPy array and a 1D array (like adding a row vector to each row of a matrix), but I'm running into a broadcasting error.

import numpy as np

# Create a 2D array with shape (3, 4)
matrix = np.array([
    [1, 2, 3, 4],
    [5, 6, 7, 8],
    [9, 10, 11, 12]
])

# Create a 1D array with shape (3,) – I intended this to be a row vector
row_vector = np.array([1, 2, 3])

# Try to add the row vector to each row of the matrix
result = matrix + row_vector

When I run, I am getting following error:

ValueError: operands could not be broadcast together with shapes (3,4) (3,)

Any Ideas?

I thought NumPy would automatically broadcast the 1D array across the rows, but it doesn’t work here.

I tried reshaping the row_vector but not sure what the correct shape should be.

Looked at the documentation on broadcasting, but still not fully clear.

Answer 1

Probably you could use np.newaxis or None like below

mat = np.array([
    [1, 2, 3, 4],
    [5, 6, 7, 8],
    [9, 10, 11, 12]
])

vec = np.array([1,2,3])

print(mat + vec[:,np.newaxis]) # or `print(mat + vec[:,None])`

and you will obtain

[[ 2  3  4  5]
 [ 7  8  9 10]
 [12 13 14 15]]