Large-scale structure for the Inference of the Galaxy field with Hamiltonian Monte-Carlo
LIGHT is a Bayesian framework for reconstructing the galaxy field and underlying dark matter field using Hamiltonian Monte Carlo (HMC), and more specifically NumPyro. This package is designed to improve the inference of cosmological parameters with gravitational waves and galaxy catalogs, a.k.a. the dark siren method.
To create a new environment and install LIGHT, follow these steps:
conda env create -f environment.yml
conda activate light-env
pip install -e .Keep in mind that the code can run on GPUs which might require a more taylored install of jax.
For examples, please navigate to the examples folder, and follow the readme there.
If you use this software, please cite:
Cosmic Cartography: Bayesian reconstruction of the galaxy density informed by large-scale structure
and
Cosmic Cartography II: completing galaxy catalogs for gravitational-wave cosmology
@article{Leyde:2024tov,
author = "Leyde, Konstantin and Baker, Tessa and Enzi, Wolfgang",
title = "{Cosmic cartography: Bayesian reconstruction of the galaxy density informed by large-scale structure}",
eprint = "2409.20531",
archivePrefix = "arXiv",
primaryClass = "astro-ph.CO",
doi = "10.1088/1475-7516/2024/12/013",
journal = "JCAP",
volume = "12",
pages = "013",
year = "2024"
}
@article{Leyde:2025rzk,
author = "Leyde, Konstantin and Baker, Tessa and Enzi, Wolfgang",
title = "{Cosmic Cartography II: completing galaxy catalogs for gravitational-wave cosmology}",
eprint = "2507.12171",
archivePrefix = "arXiv",
primaryClass = "astro-ph.CO",
month = "7",
year = "2025"
}If you have any questions, feedback, or would like to discuss this project, please don't hesitate to reach out:
Email: Konstantin.Leyde@gmail.com