Python Random Module

Python random Module: Sampling, Shuffling, and Reproducible Seeds

Python's random module generates pseudo-random numbers for games, simulations, A/B splits, and test fixtures. Because it uses the Mersenne Twister algorithm, numbers look random and can be reproduced identically when you set a fixed seed.

Never use random for passwords, session tokens, or API keys—use Python's cryptographically secure secrets module for anything security-sensitive.

Picking Items, Numbers, and Shuffling

Use randint(a, b) for inclusive integers, choice() for one item, sample() for unique picks without replacement, and shuffle() to randomize a list in place.

import random

random.seed(42)  # Optional: makes output reproducible for tests

dice = random.randint(1, 6)
team = ["Amina", "Leo", "Tariq", "Maya"]
winner = random.choice(team)
pair = random.sample(team, k=2)

random.shuffle(team)
print(dice, winner, pair, team)

random vs secrets for Tokens

For OTP codes, password reset links, or API keys, always import Python's secrets module instead of random. Unlike Mersenne Twister, secrets pulls entropy directly from the operating system's cryptographic random source.

import secrets

token = secrets.token_urlsafe(16)
print(token)

Common Mistakes to Avoid

• Expecting random.shuffle(items) to return a new list: .shuffle() mutates items in place and returns None.
• Confusing randint(1, 10) and randrange(1, 10): randint(1, 10) can return 10 (inclusive), whereas randrange(1, 10) stops at 9.
• Generating password reset tokens with random: Mersenne Twister state can be predicted after observing outputs; always use secrets.

Frequently Asked Questions

When should I call random.seed()?

Set random.seed(42) inside automated tests, benchmarks, or machine learning scripts when you want identical random splits on every run. Omit it in games and production randomness.

What is the difference between random.choices() and random.sample()?

random.choices(items, k=3) picks with replacement (duplicates possible and supports weights=), while random.sample(items, k=3) picks unique items without replacement.

Can random.sample() pick more items than exist in the list?

No. Requesting k > len(items) raises ValueError.