Broadcasting is a powerful feature in NumPy that allows arrays of different shapes to perform mathematical operations together.<\/p>\n\n\n\n
It automatically expands the smaller array to match the shape of the larger array without copying data. This makes calculations faster and memory-efficient.<\/p>\n\n\n\n
First, import NumPy:<\/p>\n\n\n\n
import numpy as np<\/pre>\n\n\n\nSimple Example of Broadcasting<\/h2>\n\n\n\n
Create an array:<\/p>\n\n\n\n
arr = np.array([1, 2, 3, 4])<\/pre>\n\n\n\nAdd a scalar value:<\/p>\n\n\n\n
arr + 5<\/pre>\n\n\n\nOutput:<\/p>\n\n\n\n
[6 7 8 9]<\/pre>\n\n\n\nHere, the number 5 is automatically broadcasted to all elements.<\/p>\n\n\n\n
Broadcasting with Two Arrays<\/h2>\n\n\n\n
Example:<\/p>\n\n\n\n
a = np.array([1, 2, 3])
b = np.array([10, 20, 30])a + b<\/pre>\n\n\n\nOutput:<\/p>\n\n\n\n
[11 22 33]<\/pre>\n\n\n\nElement-wise addition happens because both arrays have the same shape.<\/p>\n\n\n\n
Broadcasting with Different Shapes<\/h2>\n\n\n\n
Example:<\/p>\n\n\n\n
a = np.array([[1, 2, 3],
[4, 5, 6]])b = np.array([10, 20, 30])a + b<\/pre>\n\n\n\nOutput:<\/p>\n\n\n\n
[[11 22 33]
[14 25 36]]<\/pre>\n\n\n\nHere,
b<\/code> is broadcasted across each row ofa<\/code>.<\/p>\n\n\n\nBroadcasting Rules<\/h2>\n\n\n\n
NumPy follows these rules:<\/p>\n\n\n\n
\n
- If arrays have different dimensions, the smaller one is expanded.<\/li>\n\n\n\n
- Dimensions are compatible if:\n
\n
- They are equal, or<\/li>\n\n\n\n
- One of them is 1.<\/li>\n<\/ul>\n<\/li>\n\n\n\n
- If dimensions are not compatible, an error occurs.<\/li>\n<\/ol>\n\n\n\n
Example of incompatible shapes:<\/p>\n\n\n\n
a = np.array([[1, 2, 3],
[4, 5, 6]])b = np.array([10, 20])a + b<\/pre>\n\n\n\nThis will produce a shape mismatch error.<\/p>\n\n\n\n
Broadcasting with Column Vector<\/h2>\n\n\n\n
Example:<\/p>\n\n\n\n
a = np.array([[1, 2, 3],
[4, 5, 6]])b = np.array([[10],
[20]])a + b<\/pre>\n\n\n\nOutput:<\/p>\n\n\n\n
[[11 12 13]
[24 25 26]]<\/pre>\n\n\n\nHere,
b<\/code> is broadcasted across columns.<\/p>\n\n\n\nWhy Broadcasting is Important in Data Analytics<\/h2>\n\n\n\n
Broadcasting allows you to:<\/p>\n\n\n\n
Normalize data
Scale features
Perform fast matrix operations
Apply transformations to entire datasets
Avoid writing loops<\/p>\n\n\n\nKey Takeaway<\/h2>\n\n\n\n
Broadcasting automatically expands smaller arrays to match larger ones during operations. It makes NumPy powerful, efficient, and ideal for large-scale data analysis.<\/p>\n\n\n