np.newaxis for adding dimensions in NumPy - Step-by-Step Execution

Concept Flow - np.newaxis for adding dimensions

Start with 1D array

↓

Use np.newaxis to add axis

↓

Array shape changes

↓

New dimension added

↓

Result: Higher dimensional array

Start with a 1D array, apply np.newaxis to add a new dimension, resulting in a higher dimensional array.

Execution Sample

NumPy

import numpy as np
arr = np.array([1, 2, 3])
arr2d = arr[:, np.newaxis]
print(arr2d.shape)
print(arr2d)

This code adds a new axis to a 1D array to make it 2D with shape (3,1).

Execution Table

Step	Action	Array Shape	Array Content
1	Create 1D array arr	(3,)	[1 2 3]
2	Apply arr[:, np.newaxis]	(3, 1)	[[1] [2] [3]]
3	Print shape	(3, 1)	N/A
4	Print array	(3, 1)	[[1] [2] [3]]
5	End	N/A	N/A

💡 Finished printing the new array with added dimension.

Variable Tracker

Variable	Start	After Step 2	Final
arr	[1 2 3]	[1 2 3]	[1 2 3]
arr2d	N/A	[[1] [2] [3]]	[[1] [2] [3]]

Key Moments - 2 Insights

Why does arr2d have shape (3, 1) instead of (1, 3)?

Because np.newaxis was used in the second dimension position (arr[:, np.newaxis]), it adds a new axis as a column, making shape (3,1). See execution_table step 2.

What happens if np.newaxis is used in the first dimension like arr[np.newaxis, :]?

It adds a new axis as the first dimension, changing shape to (1, 3). This is different from arr[:, np.newaxis].

Visual Quiz - 3 Questions

Look at the execution_table at step 2, what is the shape of arr2d?

A (3,)

B (3, 1)

C (1, 3)

D (1,)

💡 Hint

Check the 'Array Shape' column at step 2 in execution_table.

At which step is the new axis added to the array?

A Step 2

B Step 1

C Step 3

D Step 4

💡 Hint

Look for the step where arr2d is created with shape (3, 1).

If we change arr2d = arr[np.newaxis, :], what would be the new shape?

A (3,)

B (3, 1)

C (1, 3)

D (1,)

💡 Hint

Recall that np.newaxis at the first dimension adds a new row dimension.

Concept Snapshot

np.newaxis adds a new dimension to numpy arrays.
Use it inside indexing like arr[:, np.newaxis] to add a column dimension.
This changes shape from (n,) to (n,1).
Use arr[np.newaxis, :] to add a row dimension, shape (1,n).
Useful for reshaping without copying data.

Full Transcript

We start with a 1D numpy array arr with shape (3,). Using np.newaxis inside indexing, like arr[:, np.newaxis], adds a new dimension. This changes the shape to (3, 1), turning the array into a 2D column vector. The variable arr2d holds this new array. Printing arr2d shows the new shape and content. This technique helps reshape arrays easily by adding dimensions where needed.