NumPy ndarray.squeeze() Method

NumPy ndarray.squeeze()

The numpy.ndarray.squeeze() method removes single-dimensional (size 1) entries from the shape of an array. It can operate on the entire array or selectively along a specified axis.

Syntax

</>

Copy

ndarray.squeeze(axis=None)

Parameters

Parameter	Type	Description
`axis`	None or int or tuple of ints, optional	The axis or axes to squeeze. If specified, only size-1 dimensions along these axes are removed. If `None`, all size-1 dimensions are removed.

Return Value

Returns a new array with size-1 dimensions removed. If no changes occur, it returns the original array.

Examples

1. Removing All Size-1 Dimensions

In this example, we create a multi-dimensional array with some size-1 dimensions and remove them using squeeze().

</>

Copy

import numpy as np

# Creating a 3D array with shape (1, 3, 1)
arr = np.array([[[1], [2], [3]]])

# Removing all size-1 dimensions
squeezed_arr = arr.squeeze()

# Displaying the original and squeezed shapes
print("Original shape:", arr.shape)
print("Squeezed shape:", squeezed_arr.shape)

# Displaying the squeezed array
print("Squeezed array:", squeezed_arr)

Output:

Original shape: (1, 3, 1)
Squeezed shape: (3,)
Squeezed array: [1 2 3]

Since all size-1 dimensions are removed, the shape changes from (1, 3, 1) to (3,).

2. Using the `axis` Parameter in ndarray.squeeze()

Here, we remove a size-1 dimension along a specified axis.

</>

Copy

import numpy as np

# Creating a 3D array with shape (1, 3, 1)
arr = np.array([[[10], [20], [30]]])

# Removing only the first dimension (axis=0)
squeezed_axis0 = arr.squeeze(axis=0)

# Removing only the last dimension (axis=2)
squeezed_axis2 = arr.squeeze(axis=2)

# Displaying shapes
print("Original shape:", arr.shape)
print("Shape after squeezing axis=0:", squeezed_axis0.shape)
print("Shape after squeezing axis=2:", squeezed_axis2.shape)

# Displaying squeezed arrays
print("Squeezed array (axis=0):", squeezed_axis0)
print("Squeezed array (axis=2):", squeezed_axis2)

Output:

Original shape: (1, 3, 1)
Shape after squeezing axis=0: (3, 1)
Shape after squeezing axis=2: (1, 3)
Squeezed array (axis=0): [[10]
 [20]
 [30]]
Squeezed array (axis=2): [[10 20 30]]

Using axis=0 removes the first dimension, changing the shape from (1,3,1) to (3,1). Using axis=2 removes the last dimension, resulting in a shape of (1,3).

3. Attempting to Squeeze a Non-Singleton Dimension

If we specify an axis that does not have a size-1 dimension, an error occurs.

</>

Copy

import numpy as np

# Creating a 2D array with shape (2, 3)
arr = np.array([[5, 10, 15], 
                [20, 25, 30]])

try:
    # Attempting to squeeze a non-singleton axis (axis=1, which has size 3)
    squeezed_arr = arr.squeeze(axis=1)
    print("Squeezed array:", squeezed_arr)
except ValueError as e:
    print("Error:", e)

Output:

Error: cannot select an axis to squeeze out which has size not equal to one

Since axis 1 has size 3, it cannot be squeezed, resulting in an error.

TutorialKart

NumPy ndarray.squeeze() Method