Bitwise Operators

Nerd Cafe

What Are Bitwise Operators?

Bitwise operators allow you to manipulate individual bits of integers. These are fast, low-level operations often used in:

• Image processing

• Feature engineering (e.g., bitmasking)

• Performance-critical ML tasks

• Cryptography

• Embedded systems

Bitwise Operators in Python

Operator	Name	Example	Description
&	AND	a & b	1 if both bits are 1
|	OR	a | b	1 if either bit is 1 (or both)
^	XOR	a ^ b	1 if bits are different
~	NOT	~a	Inverts all bits
<<	Left Shift	a << n	Shift bits left, adding 0s on right
>>	Right Shift	a >> n	Shift bits right, discards right bits

Step-by-Step with Examples

Step 1: Understand Binary Format in Python

Before applying bitwise ops, you should know how integers are stored in binary.

a = 10
b = 4

print(bin(a))  # 0b1010
print(bin(b))  # 0b100

bin() function returns the binary string of a number prefixed with 0b.

Step 2: Bitwise AND (&)

a = 10        # 1010
b = 4         # 0100

result = a & b
print(result)  # 0b0000 => 0

Use Case in ML: Masking selected bits in a bit pattern (e.g., enabling specific features only).

Step 3: Bitwise OR (|)

a = 10        # 1010
b = 4         # 0100

result = a | b
print(result)  # 0b1110 => 14

Use Case in ML: Combining binary feature flags.

Step 4: Bitwise XOR (^)

a = 10        # 1010
b = 4         # 0100

result = a ^ b
print(result)  # 0b1110 => 14

Use Case in ML: Change detection, feature toggles — highlight differences between two feature sets.

Step 5: Bitwise NOT (~)

a = 10         # 1010

result = ~a
print(result)  # -11

Use Case: Inverting bits, e.g., selecting the complement of a feature mask.

Step 6: Bitwise Left Shift (<<)

a = 3         # 0011

result = a << 2
print(result)  # 001100 => 12

Use Case: Efficient multiplication (e.g., a << 1 is like a * 2)

Step 7: Bitwise Right Shift (>>)

a = 8         # 1000

result = a >> 2
print(result)  # 0010 => 2

Use Case: Efficient division (e.g., a >> 1 is like a / 2)

Machine Learning Use Cases

1. Encoding Categorical Values Compactly

Bitwise encoding is sometimes used to store categorical or binary data compactly.

# Represent 'Red' = 0b0001, 'Blue' = 0b0010, 'Green' = 0b0100
RED = 1
BLUE = 2
GREEN = 4

# Let's say an object has RED and GREEN properties
color = RED | GREEN  # => 0b0101

# Check if it contains RED
print(bool(color & RED))   # True
# Check if it contains BLUE
print(bool(color & BLUE))  # False

2. Fast Image Processing (Pillow + NumPy + Bitwise)

import numpy as np
from PIL import Image

img = Image.open("your_image.jpg").convert("L")  # grayscale
np_img = np.array(img)

# Highlight dark pixels using a bitwise threshold
mask = (np_img & 0b11110000) == 0
np_img[mask] = 255  # highlight them in white

Image.fromarray(np_img).show()

Practical Notes

Topic	Note
Speed	Bitwise operations are faster than arithmetic counterparts
Compatibility	Only works on integers (not floats)
NumPy Support	Use np.bitwise_and, np.left_shift, etc. for vectorized ops
Safety	Always make sure shifts don't go beyond bit width
ML Use	Useful in binary encoding, feature flags, data compression

Keywords

bitwise operators, python bitwise, bitwise and, bitwise or, bitwise xor, bitwise not, left shift, right shift, binary operations, bitmasking, feature flags, fast computation, binary encoding, low-level operations, image processing, numpy bitwise, feature selection, binary logic, machine learning optimization, python binary tools, nerd cafe