Numpy bitwise_invert() Function

The NumPy bitwise_invert() function is used to perform a bitwise NOT operation which inverts all the bits of each element in the input array.

This function flips each bit of the binary representation of the integers which results in the bitwise complement.

For unsigned integers it effectively computes the maximum value for the bit width minus the current value and where as for signed integers it uses two's complement representation.

This function handles inputs of various integer types and returns an array of the same shape and type with inverted bits. It is equivalent to bitwise_not().

Syntax

Following is the syntax of Numpy bitwise_invert() function −

numpy.bitwise_invert(x)

Parameters

The Numpy bitwise_invert() function takes a single parameter namely, x that takes input array or value on which the bitwise inversion is to be performed. The input should be an integer type array such as int8, int16, int32, int64, uint8, uint16, uint32, uint64 etc minus;

Return value

This function returns array with the bitwise inversion applied to each element of the input array. The output array has the same shape and type as the input array.

Example 1

Following is the basic example of Numpy bitwise_invert() function in which this provides a clear view of how bitwise inversion operates on a simple array of integers −

import numpy as np

# Create a basic integer array
x = np.array([0, 1, 2, 3], dtype=np.uint8)

# Apply bitwise NOT operation
result = np.bitwise_invert(x)
print(result)   

Below is the output of bitwise_invert() function applied on the array −

[255 254 253 252]

Example 2

When working with integer arrays of different data types in NumPy, the behavior of bitwise operations, including inversion, can vary based on the integer type used. Here is the example of it −

import numpy as np

# Define arrays with different data types
array_uint8 = np.array([0, 1, 2, 255], dtype=np.uint8)
array_uint16 = np.array([0, 1, 2, 65535], dtype=np.uint16)
array_int8 = np.array([0, 1, -1, -2], dtype=np.int8)
array_int16 = np.array([0, 1, -1, -2], dtype=np.int16)

# Perform bitwise inversion
result_uint8 = np.bitwise_invert(array_uint8)
result_uint16 = np.bitwise_invert(array_uint16)
result_int8 = np.bitwise_invert(array_int8)
result_int16 = np.bitwise_invert(array_int16)

print('Original uint8 array:', array_uint8)
print('Inverted uint8 array:', result_uint8)
print('\nOriginal uint16 array:', array_uint16)
print('Inverted uint16 array:', result_uint16)
print('\nOriginal int8 array:', array_int8)
print('Inverted int8 array:', result_int8)
print('\nOriginal int16 array:', array_int16)
print('Inverted int16 array:', result_int16)

Following is the output for the above example −

Original uint8 array: [  0   1   2 255]
Inverted uint8 array: [255 254 253   0]

Original uint16 array: [    0     1     2 65535]
Inverted uint16 array: [65535 65534 65533     0]

Original int8 array: [ 0  1 -1 -2]
Inverted int8 array: [-1 -2  0  1]

Original int16 array: [ 0  1 -1 -2]
Inverted int16 array: [-1 -2  0  1]

Note: In few versions bitwise_invert() function won't work, in such cases we can use bitwise_not() function to do the same operation.