np.broadcast_to() for explicit broadcasting in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What does np.broadcast_to() do in NumPy?

np.broadcast_to() takes a smaller array and "stretches" it to a larger shape without copying data. It allows you to explicitly broadcast an array to a new shape.

beginner

Why is broadcasting useful in data science?

Broadcasting lets you perform operations on arrays of different shapes easily, like adding a single value to every element in a matrix, without writing loops.

intermediate

What happens if you try to broadcast to an incompatible shape with np.broadcast_to()?

NumPy will raise a ValueError because the shapes cannot be matched according to broadcasting rules.

beginner

Example: What is the output shape of np.broadcast_to(np.array([1, 2, 3]), (3, 3))?

The output shape is (3, 3). The 1D array [1, 2, 3] is broadcasted to 3 rows, repeating the original array in each row.

intermediate

Does np.broadcast_to() copy data or create a view?

It creates a view with broadcasted shape, so no new data is copied. This is memory efficient.

What is the main purpose of np.broadcast_to()?

A To sort an array

B To reshape an array by changing its data

C To copy an array multiple times

D To explicitly broadcast an array to a new shape

np.broadcast_to() explicitly broadcasts an array to a new shape without copying data.

If you have arr = np.array([5]), what will np.broadcast_to(arr, (2,3)) produce?

A An error because shapes don't match

B (2,3) array filled with 5

C (1,) array with 5

D (3,2) array filled with 5

Scalar 5 is broadcasted to fill the (2,3) shape.

What error occurs if broadcasting shapes are incompatible?

A ValueError

B TypeError

C IndexError

D KeyError

NumPy raises a ValueError when shapes cannot be broadcasted.

Does np.broadcast_to() create a copy of the data?

A Yes, it copies data

B Only for 2D arrays

C No, it creates a view

D Only if the array is large

It creates a view with broadcasted shape, saving memory.

Which of these shapes can np.broadcast_to() broadcast arr.shape = (3,) to?

A (3, 3)

B (4, 4)

C (3, 4)

D (2, 2)

Shape (3,) can broadcast to (3, 3) by repeating along the new axis.

Explain how np.broadcast_to() helps when working with arrays of different shapes.

Think about how you can stretch a small array to match a bigger one without loops.

Describe a real-life example where you might use np.broadcast_to() in data analysis.

Imagine adding a tax rate to every price in a table.