How to Sample Random Items Without Replacement in Python

Select k random unique items from a sequence using random.sample for uniform, non-repeating selection.

Easy Python 3.9+ Aug 9, 2026 Algorithms & data structures 15 views 0 copies

random sampling algorithms stdlib

Python code

15 lines

Python 3.9+

import random

def sample_without_replacement(population, k):
    """Return k random items from population without replacement."""
    if k > len(population):
        raise ValueError("k cannot exceed population size")
    # Use random.sample for O(k) time, no mutation of the original
    return random.sample(population, k)

if __name__ == "__main__":
    # Example: sample 3 fruits from a list of 6
    fruits = ["apple", "banana", "cherry", "date", "elderberry", "fig"]
    result = sample_without_replacement(fruits, 3)
    print("Sampled items:", result)
    print("Original list unchanged:", fruits)

Output

stdout

Sampled items: ['cherry', 'fig', 'apple']
Original list unchanged: ['apple', 'banana', 'cherry', 'date', 'elderberry', 'fig']

How it works

random.sample performs sampling without replacement in O(k) time by using a partial Fisher-Yates shuffle on a copy of the population, leaving the input sequence unmodified. It raises a ValueError if k exceeds the population length, which matches the Python API for random sampling. The function uses the system's randomness source by default, making it suitable for most non-security applications. This approach is both concise and idiomatic, avoiding manual index tracking or destructive list population.

Common mistakes

Forgetting to check if k exceeds population size before sampling, causing a ValueError from random.sample
Mutating the original list when implementing a manual shuffle-based approach
Using random.choices instead, which allows duplicates by design

Variations

Use `random.sample(range(n), k)` for sampling k distinct integers from 0 to n-1
Implement a manual Fisher-Yates shuffle on a list copy for educational clarity

Real-world use cases

Selecting a random subset of users for A/B testing or survey distribution.
Choosing k training rows from a dataset for mini-batch stochastic gradient descent.
Picking random support tickets for QA auditing without revisiting the same ticket twice.

How to Sample Random Items Without Replacement in Python

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from Algorithms & data structures

Tutorials

Quizzes

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from Algorithms & data structures

Keep learning

Tutorials

Quizzes