Broadcasting rules in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is broadcasting in NumPy?

Broadcasting is a way NumPy allows arrays of different shapes to work together in arithmetic operations by automatically expanding the smaller array to match the shape of the larger one.

beginner

What are the three main rules of broadcasting in NumPy?

1. If arrays have different numbers of dimensions, prepend 1s to the shape of the smaller array.<br>2. Arrays are compatible in a dimension if they are equal or one of them is 1.<br>3. Arrays can be broadcast together if they are compatible in all dimensions.

intermediate

Given arrays with shapes (3, 1) and (1, 4), what will be the shape after broadcasting?

The arrays will broadcast to shape (3, 4). The 1 in each dimension is stretched to match the other array's size.

beginner

Why does broadcasting help in data science computations?

Broadcasting lets you perform operations on arrays of different shapes without writing loops, making code simpler and faster, especially for large datasets.

beginner

What happens if two arrays are not compatible for broadcasting?

NumPy raises a ValueError because it cannot automatically align the arrays for element-wise operations.

Which of these pairs of shapes can be broadcast together?<br>A) (3, 4) and (4)<br>B) (2, 3) and (3, 2)<br>C) (5, 1) and (5, 3)<br>D) (1, 3) and (2, 3, 1)

A (3, 4) and (4)

B (2, 3) and (3, 2)

C (5, 1) and (5, 3)

D (1, 3) and (2, 3, 1)

Options A, C, and D work because their dimensions are compatible (equal or one is 1 after padding leading 1s). A: (4) as (1,4) with (3,4). C: (5,1) with (5,3). D: (1,3) padded to (1,1,3) vs (2,3,1). B incompatible.

What does NumPy do if one array has shape (3, 1) and the other (3, 4)?

A Broadcasts the (3, 1) array to (3, 4)

B Raises an error

C Broadcasts the (3, 4) array to (3, 1)

D No broadcasting, uses only first array

The (3, 1) array is stretched along the second dimension to match (3, 4).

If two arrays have shapes (2, 3, 4) and (3, 1), what shape will result after broadcasting?

A (3, 3, 4)

B (2, 3, 1)

C (2, 3, 4)

D Broadcasting not possible

The smaller array shape is treated as (1, 3, 1) and broadcasts to (2, 3, 4).

Which condition must be true for two dimensions to be compatible in broadcasting?

A They must both be even numbers

B They must both be prime numbers

C They must be multiples of each other

D They must be equal or one must be 1

Broadcasting requires dimensions to be equal or one to be 1 to stretch the smaller array.

What is the first step NumPy takes when broadcasting arrays with different numbers of dimensions?

A It removes extra dimensions

B It prepends 1s to the smaller array's shape

C It transposes the arrays

D It reshapes both arrays to 1D

NumPy adds 1s to the front of the smaller array's shape to align dimensions.

Explain the three main broadcasting rules in NumPy and why they matter.

Think about how shapes align and stretch.

Describe a real-life example where broadcasting can simplify data calculations.

Imagine you have a list of prices and one tax rate.