PackLab 0.6.4

Features

• Random sequential adsorption of mono- and polydisperse spheres.

• Periodic or finite box domains with configurable stopping criteria.

• Radius samplers for constant, uniform, normal, log-normal, and discrete distributions.

• Pair-correlation estimates, packing statistics, and Matplotlib plots.

• Percus–Yevick mixture solver with automatic, resolution-aware wavenumber grids.

• Optional PyMieSim integration for scattering and phase-function workflows.

• Unit-aware quantities throughout, via TypedUnit.

Installation

Install the core package from PyPI:

pip install packlab

Install optional scattering support:

pip install "packlab[scattering]"

The conda package is also available:

conda install -c martinpdes packlab

Verify that the compiled extensions are available with the interpreter you will use for simulations:

python -c "import PackLab; print(PackLab.__version__)"

First RSA packing

Create a periodic domain, choose a radius distribution, configure the RSA stopping conditions, and run the simulation. Dimensional inputs carry units.

from PackLab import monte_carlo, samplers, ureg

domain = monte_carlo.PackingDomain(
    length_x=6 * ureg.micrometer,
    length_y=6 * ureg.micrometer,
    length_z=6 * ureg.micrometer,
    use_periodic_boundaries=True,
)
radii = samplers.UniformRadiusSampler(
    minimum_radius=100 * ureg.nanometer,
    maximum_radius=200 * ureg.nanometer,
    bins=12,
)
options = monte_carlo.RSAOptions()
options.random_seed = 42
options.maximum_attempts = 100_000
options.target_packing_fraction = 0.15

result = monte_carlo.RSASimulator(domain, radii, options).run()
print(result.statistics.packing_fraction_geometry)
result.plot_slice_2d()

Analytical reference

Use the analytical solver for a fast Percus–Yevick reference. With wavenumber="auto", PackLab chooses a zero-inclusive wavenumber grid from the particle radii and requested distance range.

import numpy as np

from PackLab import analytical, ureg

domain = analytical.PercusYevickDomain(
    size=10 * ureg.micrometer,
    radii=[100, 150] * ureg.nanometer,
    volume_fraction=0.15,
    number_fractions=[0.7, 0.3],
)
distances = np.linspace(0.2, 1.5, 300) * ureg.micrometer
solver = analytical.PercusYevickSolver(
    densities=domain.particle_densities_per_radius,
    radii=domain.radii,
    wavenumber="auto",
)
result = solver.compute(distances)
print(result.wavenumber)

For an explicit grid, use analytical.make_wavenumber_grid(...). PackLab warns when a manually supplied grid is too coarse for the requested distances.

Choosing a workflow

• Use PackLab.monte_carlo when individual centres, sampled radii, box boundaries, or finite-size effects are important.

• Use PackLab.analytical for fast parameter sweeps and an analytical correlation reference.

• Use the validation gallery examples to compare a matching RSA configuration against the analytical model.

Documentation and examples

The online documentation contains theory, API reference, and executable examples organised into Monte-Carlo, analytical, and validation workflows.

Building from source

For development, clone the repository and install it in editable mode. A C++20 compiler and CMake are required to build the native extensions.

git clone https://github.com/MartinPdeS/PackLab.git
cd PackLab
pip install -e ".[testing,documentation]"

Testing

Run the test suite with:

pytest

Citing PackLab

If PackLab contributes to academic work, cite the archived Zenodo release you used. Release metadata is included in .zenodo.json.

Contributing and contact

Issues and pull requests are welcome. For questions or collaborations, contact Martin Poinsinet de Sivry-Houle.