Broadcasting

Definition: Broadcasting is how NumPy combines arrays of different shapes. It automatically "stretches" the smaller array across the larger one so the operation works without writing loops.

Example 1 — array and a single number

import numpy as np
a = np.array([1, 2, 3])
print(a + 10)   # 10 is broadcast to [10 10 10] -> [11 12 13]

Example 2 — 2D and 1D

A 1D row can be added to every row of a 2D array:

import numpy as np
m = np.array([[1, 2, 3],
              [4, 5, 6]])
row = np.array([10, 20, 30])
print(m + row)   # row is added to each row of m

A practical use — normalising data

import numpy as np
data = np.array([10, 20, 30, 40])
print(data - data.mean())   # centre the data around 0

Here data.mean() (a single number) is broadcast across the whole array.

💡 Tip: broadcasting is what lets you scale, shift, and normalise entire datasets in one clean line.