IMPORTANT ANNOUNCEMENT (9/21/2022): EMRISurrogate has moved to BHPTNRSurrogate. We will continue to make bug fixes to the EMRISurrogate package. This package will continue to host the EMRISur1dq1e4 model while future models can be found in the BHPTNRSurrogate package.
Copyright (c) 2020 Black Hole Perturbation Toolkit Team
The EMRISurrogate package provides access to a surrogate gravitational waveform model. Currently, this package supports one model, EMRISur1dq1e4, for non-spinning black hole binary systems with mass-ratios varying from 3 to 10000. This surrogate model is trained on waveform data generated by point-particle black hole perturbation theory (ppBHPT), with the total mass rescaling parameter tuned to NR simulations according to equation 4 of paper 1 (see citations below).
This total mass rescaling is applied in the function slog_surrogate, and so to generate point-particle perturbation theory waveforms without any tuning to NR set alpha = 1 in the relevant function. Note that alpha approaches unity in the extreme mass ratio limit.
Available modes are [(2,2), (2,1), (3,3), (3,2), (3,1), (4,4), (4,3), (4,2), (5,5), (5,4), (5,3)]. The m<0 modes are deduced from the m>0 modes.
Model details can be found in paper 1 (see citations below).
The latest development version will always be available from the project git repository:
git clone https://github.com/BlackHolePerturbationToolkit/EMRISurrogate.gitThis package requires Python 3 and the sklearn package. Parts of the accompanying Jupyter notebook will require gwsurrogate, which can be installed with
pip install gwsurrogateNote that you do not need gwsurrogate to evalulate the EMRI surrogate model or run most parts of the notebook.
- Clone the repository
- Download the datafile (hosted on Zenodo)
wget https://zenodo.org/record/3612600/files/EMRISur1dq1e4.h5- Create a static link from in the main directory to the EMRISur1dq1e4.h5 file. Or simply move the EMRISur1dq1e4.h5 file into the main directory.
Please see the accompanying Jupyter notebook
jupyter notebook EMRISur1dq1e4.ipynbExamples are included in the Jupyter notebook.
Known bugs are recorded in the project bug tracker:
https://github.com/BlackHolePerturbationToolkit/EMRISurrogate/issues
This code is distributed under the MIT License. Details can be found in the LICENSE file.
Scott Field
Tousif Islam
Gaurav Khanna
Nur Rifat
Vijay Varma
If you make use of any module from the Toolkit in your research please acknowledge using:
This work makes use of the Black Hole Perturbation Toolkit.
If you make use of the EMRI surrogate model, EMRISur1dq1e4, please cite Paper 1:
@article{rifat2019surrogate,
title={A Surrogate Model for Gravitational Wave Signals from Comparable-to Large-Mass-Ratio Black Hole Binaries},
author={Rifat, Nur EM and Field, Scott E and Khanna, Gaurav and Varma, Vijay},
journal={arXiv preprint arXiv:1910.10473},
year={2019}
}