How to Assert Approximately Equal in pytest: Simple Guide

In pytest, you can assert approximate equality using pytest.approx(). This helper allows you to compare floating-point numbers within a tolerance, making tests reliable for values that may differ slightly due to rounding.

Syntax

The basic syntax to assert approximate equality in pytest is:

• assert actual == pytest.approx(expected, rel=relative_tolerance, abs=absolute_tolerance)

Here, actual is the value you got, expected is the value you want to compare against.

rel sets the relative tolerance (default 1e-6), and abs sets the absolute tolerance (default 0).

python

assert actual == pytest.approx(expected, rel=1e-6, abs=0)

Example

This example shows how to test that two floating-point numbers are approximately equal using pytest.approx(). It passes even if the numbers differ slightly due to rounding.

python

import pytest

def test_approx_equal():
    result = 0.1 + 0.2
    expected = 0.3
    assert result == pytest.approx(expected)

if __name__ == "__main__":
    pytest.main([__file__])

Output

============================= test session starts ==============================
collected 1 item

test_approx_equal.py .                                                     [100%]

============================== 1 passed in 0.01s ===============================

Common Pitfalls

Common mistakes include:

• Using simple == for floating-point numbers, which can fail due to tiny rounding errors.
• Not specifying tolerances when needed, causing tests to be too strict or too loose.
• Confusing relative and absolute tolerance; use rel for proportional differences and abs for fixed differences.

Always use pytest.approx() for floating-point comparisons instead of plain equality.

python

import pytest

def test_wrong():
    result = 0.1 + 0.2
    expected = 0.3
    # This will fail due to floating point precision
    # assert result == expected

    # Correct way
    assert result == pytest.approx(expected)

if __name__ == "__main__":
    pytest.main([__file__])

Output

============================= test session starts ==============================
collected 1 item

test_wrong.py .                                                             [100%]

============================== 1 passed in 0.01s ===============================

Quick Reference

Summary tips for using pytest.approx():

• Use assert actual == pytest.approx(expected) for approximate equality.
• Adjust rel and abs parameters to control tolerance.
• Works with floats, lists, tuples, and dicts of numbers.
• Helps avoid flaky tests caused by floating-point rounding.

Key Takeaways

Use pytest.approx() to assert approximate equality of floating-point numbers.

Adjust relative (rel) and absolute (abs) tolerances to fit your precision needs.

Avoid direct equality checks (==) for floats to prevent flaky tests.

pytest.approx() supports complex data structures like lists and dicts of numbers.

Always prefer pytest.approx() for reliable floating-point comparisons in tests.