Approximate comparisons (pytest.approx) - Framework Patterns

Framework Mode - Approximate comparisons (pytest.approx)

Folder Structure

test_project/
├── tests/
│   ├── test_math_operations.py
│   └── test_utils.py
├── src/
│   └── math_functions.py
├── conftest.py
├── pytest.ini
└── requirements.txt

Test Framework Layers

• Tests Layer: Contains test files like test_math_operations.py where pytest.approx is used for approximate value assertions.
• Source Code Layer: Application code under src/ folder, e.g., math functions to be tested.
• Fixtures & Config Layer: conftest.py for shared fixtures and setup.
• Utilities Layer: Helper functions or custom assertion utilities if needed.
• Configuration Layer: pytest.ini to configure pytest options.

Configuration Patterns

• pytest.ini: Configure markers, addopts, and test paths.

  [pytest]
  addopts = -v --tb=short
  markers =
      slow: marks tests as slow
  

• Environment Handling: Use environment variables or pytest command line options to switch test parameters if needed.
• Approximate Comparison Settings: Use pytest.approx with relative or absolute tolerance directly in tests, no global config needed.

Test Reporting and CI/CD Integration

• Use pytest built-in verbose reporting with -v option for clear pass/fail output.
• Integrate with CI tools (GitHub Actions, Jenkins) to run tests on push or pull requests.
• Generate JUnit XML reports with --junitxml=report.xml for CI dashboards.
• Use pytest plugins like pytest-html for human-readable HTML reports.
• Approximate comparison failures show clear diffs with pytest's built-in assertion introspection.

Best Practices

• Use pytest.approx for floating-point comparisons: Avoid exact equality to handle rounding errors.
• Set tolerances explicitly: Use rel or abs parameters in approx to control precision.
• Keep tests readable: Write assertions like assert result == pytest.approx(expected, rel=1e-6) for clarity.
• Organize tests by feature: Group approximate comparison tests logically in test files.
• Use fixtures for setup: Share common test data or parameters via conftest.py fixtures.

Self Check

Where in this folder structure would you add a new test file to verify approximate comparisons for a new math function?

Key Result

Use pytest.approx in test files under tests/ to handle floating-point approximate comparisons with clear, configurable tolerances.