Broadcasting for outer products in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is broadcasting in NumPy?

Broadcasting is a way NumPy performs operations on arrays of different shapes by automatically expanding the smaller array to match the shape of the larger one without copying data.

beginner

How does broadcasting help in calculating outer products?

Broadcasting lets you multiply a column vector by a row vector directly, creating a matrix of all pairwise products without explicit loops.

intermediate

Given two 1D arrays `a` and `b`, how can you use broadcasting to compute their outer product?

Reshape `a` to a column vector using `a[:, None]` and multiply by `b` as a row vector. NumPy broadcasts these to produce the outer product matrix.

beginner

What shape does the result have when you compute the outer product of arrays with shapes (m,) and (n,)?

The result is a 2D array with shape (m, n), where each element is the product of elements from the first and second arrays.

intermediate

Why is broadcasting more efficient than using loops for outer products?

Broadcasting uses optimized, low-level operations and avoids explicit Python loops, making calculations faster and code simpler.

What does broadcasting allow you to do in NumPy?

A Copy arrays to match shapes before operations

B Perform operations on arrays of different shapes without explicit loops

C Only add arrays of the same shape

D Convert arrays to lists automatically

Broadcasting lets NumPy perform operations on arrays with different shapes by expanding the smaller array logically without copying data.

How do you reshape a 1D array `a` to a column vector for broadcasting?

A a[None, :]

B a.T

C a.reshape(-1)

D a[:, None]

Using `a[:, None]` adds a new axis making `a` a column vector suitable for broadcasting.

What is the shape of the outer product of arrays with shapes (3,) and (4,)?

A (3,)

B (4, 3)

C (3, 4)

D (4,)

The outer product results in a 2D array with shape (3, 4), combining each element of the first array with each element of the second.

Which of these is a correct way to compute the outer product using broadcasting?

A a[:, None] * b

B a * b

C a.T * b.T

D a + b

Reshaping `a` to a column vector with `a[:, None]` and multiplying by `b` uses broadcasting to compute the outer product.

Why is broadcasting preferred over loops for outer products?

A It is faster and uses optimized operations

B It is slower but easier to read

C It uses more memory

D It requires manual copying of arrays

Broadcasting uses optimized low-level operations, making it faster and more efficient than explicit loops.

Explain how broadcasting works to compute the outer product of two 1D arrays in NumPy.

Think about adding a new axis to one array and how NumPy matches shapes.

Describe why broadcasting is useful compared to using loops when calculating outer products.

Consider speed and code simplicity.