Sovereign 8s.is mirror of standardgalactic/Firefly-ringdown
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
2025-12-09 11:25:30 +08:00
config [FIX] plot comparison 2025-05-17 17:52:51 +08:00
data [FIX] plot comparison 2025-05-17 17:52:51 +08:00
docs [MOD] runtime test readme 2025-06-22 11:45:43 +08:00
firefly [FIX] plot comparison 2025-05-17 17:52:51 +08:00
.gitignore [FIX] plot comparison 2025-05-17 17:52:51 +08:00
LICENSE Initial commit 2025-05-04 10:36:56 +08:00
README.md [MOD] README 2025-05-10 16:51:30 +08:00
requirements.txt [Update] install requirements 2025-12-09 11:25:30 +08:00

Firefly-ringdown


Firefly-ringdown is a Python program for accelerating the Bayesian analysis of ringdown signals involving multiple quasi-normal modes (QNMs), serving the ringdown signal analysis of future gravitational-wave detectors.

It implements the marginalization of the amplitude parameters under a freely chosen QNM amplitude prior, dramatically reducing computation time while maintaining high sampling quality.

🛠️ Development install

We recommend using conda to configure the environment; please run the following command in the terminal.

conda create -n firefly python=3.10.16
conda activate firefly
pip install -r requirements.txt

🚀 Quick start

Make sure that firefly is in your current operating path, run the following command in the terminal,

python -m firefly.cli.firefly_ringdown --config_path config/NR0305_221_firefly_config.yaml

For comparison with full-parameters estimation, run the following command in the terminal,

python -m firefly.cli.fullparams_ringdown --config_path config/NR0305_221_fullparams_config.yaml

We offer a script to compare posteriors of Firefly and full-parameter sampling,

python -m firefly.cli.posterior_comparison --config_path config/posterior_comparison.yaml

and a script to simulate QNM signals,

python -m firefly.cli.simulate_signals --config_path config/simulate_signals.yaml

💡 Algorithm

Based on the likelihood-driven importance-sampling scheme design, we extended the marginalization acceleration approach of the $\mathcal{F}$-statistic to arbitrary prior scenarios, significantly accelerating the computational efficiency of traditional full-parameter Bayesian inference.

🔗 Attribution

If you have used or referred to the code from Firefly-ringdown, please cite Y. Dong & Z. Wang et al. (2025).

@article{Dong:2025igh,
    author = "Dong, Yiming and Wang, Ziming and Wang, Hai-Tian and Zhao, Junjie and Shao, Lijing",
    title = "{A practical Bayesian method for gravitational-wave ringdown analysis with multiple modes}",
    eprint = "2502.01093",
    archivePrefix = "arXiv",
    primaryClass = "gr-qc",
    month = "2",
    year = "2025"
}

🌟 Acknowledgement

We gratefully acknowledge the developers of the following gravitational-wave analysis libraries, which have been instrumental to this project, bilby, pycbc, lalsuite, qnm, gwpy and pykerr.

We gratefully acknowledge the following papers, upon whose theoretical foundations the Firefly methodology is built.

R. Prix (2016) pioneered leveraging the Gaussian nature of the likelihood to accelerate ringdown signal analysis. M. Isi & W. M. Farr (2021) laid the theoretical groundwork for the full-parameter sampling employed in this project, and introduced a sampler-friendly parameterization that further exploits Gaussian characteristics. The code of $\mathcal{F}$-statistic-based auxillary inference is adpated from the project of H.-T. Wang et al. (2024).