Janssen Design Principles

Contributing to Janssen

Thank you for your interest in contributing to Janssen! This document provides guidelines and best practices for contributing to the project.

Table of Contents

Development Setup

Prerequisites

  • Python 3.12 or higher

  • Virtual environment (recommended)

Installation

# Clone the repository
git clone https://github.com/yourusername/janssen.git
cd janssen

# Create and activate virtual environment
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install in development mode with all dependencies
pip install -e ".[dev]"

Running Quality Checks

# Run tests
pytest

# Run type checking
mypy src/

# Run linting
ruff check .

# Format code
ruff format .

Code Style and Conventions

General Principles

  1. Follow PEP 8 with a line length limit of 100 characters

  2. Use meaningful variable names - prefer descriptive names over short abbreviations

  3. Keep functions focused - each function should do one thing well

  4. Avoid magic numbers - use named constants for repeated values

  5. No commented-out code - remove dead code rather than commenting it out

Import Organization

Organize imports in the following order:

  1. Standard library imports

  2. Type annotation imports from typing

  3. Third-party imports (numpy, jax, etc.)

  4. Local application imports

Example:

from typing import Any

import chex
import jax
import jax.numpy as jnp
import pytest
from absl.testing import parameterized
from jaxtyping import Array, Complex

from janssen.lenses import create_lens_phase, lens_focal_length
from janssen.utils import LensParams, make_lens_params

JAX-Specific Conventions

  1. Enable 64-bit precision when needed:

    jax.config.update("jax_enable_x64", True)

  2. Use PyTree structures for complex data types:

    @register_pytree_node_class
class LensParams(NamedTuple):
    focal_length: Float[Array, " "]
    diameter: Float[Array, " "]

  3. Prefer functional programming - avoid mutable state where possible

Type Hinting Rules

Mandatory Type Hints

All functions must include type hints for:

  • All parameters (including self in methods is implicit)

  • Return types (use -> None for functions that don’t return a value)

Type Annotation Guidelines

  1. Import necessary types:

    from typing import Any, Optional, Tuple, Union
from jaxtyping import Array, Bool, Complex, Float, Int, Num

  2. Use specific JAX array types when possible:

    def process_field(field: Complex[Array, "height width"]) -> Float[Array, "height width"]:
    return jnp.abs(field)

  3. Use type aliases for complex types:

    from typing import TypeAlias

ScalarFloat: TypeAlias = Union[float, Float[Array, " "]]
ScalarComplex: TypeAlias = Union[complex, Complex[Array, " "]]

  4. Annotate test methods:

    def test_lens_thickness_profile(self) -> None:
    """Test lens thickness profile calculation."""
    ...

  5. Handle unused parameters in parameterized tests:

    # Prefix with underscore if parameter is needed but unused
def test_something(self, used_param: float, _unused_param: float) -> None:
    ...

Factory Functions

Always use factory functions with runtime type checking:

# Good - use factory function
params = make_lens_params(
    focal_length=0.01,
    diameter=0.005,
    n=1.5,
    center_thickness=0.001,
    r1=0.01,
    r2=0.01,
)

# Bad - direct instantiation
params = LensParams(0.01, 0.005, 1.5, 0.001, 0.01, 0.01)

Testing Guidelines

Test Structure

Tests should be organized in a parallel structure to the source code:

src/janssen/lenses/lens_elements.py
tests/test_janssen/test_lenses/test_lens_elements.py

Writing Tests

1. Use Descriptive Test Names
class TestLensElements(chex.TestCase, parameterized.TestCase):
    def test_lens_thickness_profile_double_convex(self) -> None:
        """Test thickness profile for double convex lens."""
        ...
    
    def test_lens_focal_length_plano_concave(self) -> None:
        """Test focal length calculation for plano-concave lens."""
        ...
2. Use Named Parameters for Clarity
@parameterized.named_parameters(
    ("double_convex_equal_radii", 1.5, 0.01, 0.01, 0.01),
    ("plano_convex", 1.5, 0.01, jnp.inf, 0.01),
    ("specific_focal_length", 1.5, 0.1, 0.3, 0.15),
)
def test_lens_focal_length(
    self, n: float, r1: float, r2: float, expected_f: float
) -> None:
    ...
3. Test with JAX Transformations

Use chex.variants to test both JIT-compiled and non-JIT versions:

@chex.variants(with_jit=True, without_jit=True)
def test_lens_thickness_profile(self) -> None:
    var_lens_thickness = self.variant(lens_thickness_profile)
    thickness = var_lens_thickness(self.r, r1, r2, ct, d)
    ...
4. Common Test Setup

Use setUp method for common initialization:

def setUp(self) -> None:
    super().setUp()
    self.nx = 128
    self.ny = 128
    self.dx = 1e-6
    self.wavelength = 500e-9
    x = jnp.arange(-self.nx // 2, self.nx // 2) * self.dx
    y = jnp.arange(-self.ny // 2, self.ny // 2) * self.dx
    self.xx, self.yy = jnp.meshgrid(x, y)
5. Use Chex Assertions

Prefer chex assertions for numerical testing:

# Shape assertions
chex.assert_shape(output, (128, 128))
chex.assert_trees_all_equal_shapes(output, input_field)

# Value assertions
chex.assert_trees_all_close(f, expected, rtol=1e-5)
chex.assert_scalar_positive(float(params.r1))
chex.assert_scalar_negative(float(params.r2))
chex.assert_tree_all_finite(values)

# Type assertions
chex.assert_type(output, jnp.complex128)
6. Test Edge Cases

Always include tests for:

  • Boundary conditions (zero, infinity, negative values)

  • Special cases (plano lenses, meniscus configurations)

  • Different data types and precisions

  • JAX transformations (jit, vmap, grad)

Example edge case testing:

@parameterized.named_parameters(
    ("zero_field", jnp.zeros((128, 128), dtype=complex)),
    ("complex_uniform", jnp.ones((128, 128)) * (1 + 1j)),
    ("phase_uniform", jnp.exp(1j * jnp.ones((128, 128)))),
)
def test_propagate_edge_cases(
    self, input_field: Complex[Array, "128 128"]
) -> None:
    ...
7. Test JAX Transformations

Ensure your code works with JAX transformations:

def test_jax_transformations_on_thickness_profile(self) -> None:
    # Test JIT compilation
    @jax.jit
    def jitted_thickness(
        r: Array, r1: float, r2: float, ct: float, d: float
    ) -> Array:
        return lens_thickness_profile(r, r1, r2, ct, d)
    
    # Test gradient computation
    def loss_fn(r1: float) -> Array:
        thickness = lens_thickness_profile(self.r, r1, 0.01, 0.001, 0.005)
        return jnp.sum(thickness**2)
    
    grad_fn = jax.grad(loss_fn)
    grad = grad_fn(0.01)
    chex.assert_scalar_non_negative(abs(float(grad)))
    
    # Test vmap
    vmapped_create = jax.vmap(
        lambda f: double_convex_lens(f, 0.005, 1.5, 0.001, 1.0)
    )
    focal_lengths = jnp.array([0.01, 0.02, 0.03])
    params_batch = vmapped_create(focal_lengths)
    chex.assert_shape(params_batch.focal_length, (3,))

Running Tests

# Run all tests
pytest

# Run specific test file
pytest tests/test_janssen/test_lenses/test_lens_elements.py

# Run with coverage
pytest --cov=janssen --cov-report=html

# Run with verbose output
pytest -v

# Run only failed tests
pytest --lf

Documentation Standards

Docstring Format

Use NumPy-style docstrings:

def lens_focal_length(n: float, r1: float, r2: float) -> float:
    """Calculate the focal length of a lens using the lensmaker's equation.
    
    Parameters
    ----------
    n : float
        Refractive index of the lens material
    r1 : float
        Radius of curvature of the first surface (positive for convex)
    r2 : float
        Radius of curvature of the second surface (positive for convex)
    
    Returns
    -------
    float
        Focal length of the lens in meters
    
    Notes
    -----
    Uses the thin lens approximation. For thick lenses, corrections
    may be necessary.
    
    Examples
    --------
    >>> focal_length = lens_focal_length(1.5, 0.01, 0.01)
    >>> print(f"Focal length: {focal_length:.3f} m")
    Focal length: 0.010 m
    """

Module Documentation

Each module should have a docstring explaining its purpose:

"""
Module: janssen.lenses.lens_elements
------------------------------------
Differentiable optical lens elements and propagation functions.

This module provides functions for creating and simulating various
types of optical lenses including convex, concave, plano, and meniscus
configurations.
"""

Commit Messages

Follow conventional commit format:

<type>(<scope>): <subject>

<body>

<footer>

Types:

  • feat: New feature

  • fix: Bug fix

  • docs: Documentation changes

  • style: Code style changes (formatting, etc.)

  • refactor: Code refactoring

  • test: Adding or modifying tests

  • chore: Maintenance tasks

Example:

feat(lenses): add meniscus lens configuration support

- Implemented meniscus_lens factory function
- Added comprehensive tests for convex/concave first configurations
- Updated documentation with usage examples

Pull Request Process

  1. Create a feature branch:

    git checkout -b feature/your-feature-name

  2. Make your changes following all guidelines above

  3. Run all quality checks:

    # Format code
ruff format .

# Check linting
ruff check .

# Run tests
pytest

# Check types (if configured)
mypy src/

  4. Ensure all tests pass including any new tests you’ve added

  5. Update documentation if you’ve added new features or changed APIs

  6. Create a pull request with:

    • Clear description of changes

    • Reference to any related issues

    • Screenshots/examples if applicable

  7. Address review feedback promptly

PR Checklist

Before submitting a PR, ensure:

  • [ ] All tests pass

  • [ ] Code follows style guidelines

  • [ ] Type hints are complete and correct

  • [ ] Documentation is updated

  • [ ] Commit messages are clear

  • [ ] No debugging code or print statements remain

  • [ ] New features have corresponding tests

Questions?

If you have questions about contributing, please:

  1. Check existing issues and discussions

  2. Open a new issue for clarification

  3. Reach out to maintainers

Thank you for contributing to Janssen!