Implementation:Scikit learn Scikit learn TestingUtils

Knowledge Sources	Scikit_learn Scikit-learn Docs
Domains	Machine Learning, Testing
Last Updated	2026-02-08 15:00 GMT

Overview

Concrete utility module for testing helpers and assertion functions provided by scikit-learn.

Description

The _testing module provides a comprehensive collection of testing utilities for scikit-learn. It includes assertion functions like assert_allclose, context managers like ignore_warnings, and re-exports from NumPy testing such as assert_array_equal and assert_array_almost_equal. The module also provides utilities for running Python scripts and validating estimator behavior.

Usage

Use these utilities when writing unit tests for scikit-learn estimators and functions. The module provides standardized assertion functions and warning management tools that account for scikit-learn-specific edge cases.

Code Reference

Source Location

• Repository: scikit-learn
• File: sklearn/utils/_testing.py

Signature

def ignore_warnings(obj=None, category=Warning):
    ...

def assert_allclose(actual, desired, rtol=None, atol=0., ...):
    ...

def assert_run_python_script_without_output(source_code, timeout=60):
    ...

# Re-exported from numpy.testing:
# assert_almost_equal, assert_array_almost_equal, assert_array_equal, assert_array_less

Import

from sklearn.utils._testing import assert_allclose, ignore_warnings

I/O Contract

Inputs

Name	Type	Required	Description
actual	array-like	Yes	Actual values to check
desired	array-like	Yes	Expected values to compare against
rtol	float	No	Relative tolerance for comparison
atol	float	No	Absolute tolerance for comparison
category	Warning class	No	Warning category to ignore

Outputs

Name	Type	Description
None	None	Raises AssertionError if values do not match within tolerance

Usage Examples

Basic Usage

import numpy as np
from sklearn.utils._testing import assert_allclose, ignore_warnings

# Assert arrays are close
actual = np.array([1.0, 2.0, 3.0])
desired = np.array([1.001, 2.001, 3.001])
assert_allclose(actual, desired, atol=0.01)

# Suppress warnings in a test function
@ignore_warnings(category=DeprecationWarning)
def test_something():
    pass

Related Pages

• Principle:Scikit_learn_Scikit_learn_Metric_Evaluation