gw-remnant 0.4.0

Python package to extract remnant black hole properties from gravitational waveforms

gw_remnant

During binary-black-hole (BBH) mergers, energy and momenta are carried away from the binary system as gravitational radiation. Access to the radiated energy and momenta allows us to accurately predict the properties of the remnant black hole.

gw_remnant is an easy-to-use Python package to efficiently extract the remnant mass, remnant spin, peak luminosity and the final kick imparted on the remnant black hole directly from the gravitational radiation.

Documentation: https://tousifislam.github.io/gw_remnant/

Installation

Install from PyPI:

pip install gw_remnant

Install with optional dependencies for waveform surrogates:

pip install gw_remnant[surrogates]

Install with optional dependencies for numerical relativity catalogs:

pip install gw_remnant[nr]

Install all optional dependencies:

pip install gw_remnant[all]

Install from source (development):

git clone https://github.com/tousifislam/gw_remnant.git
cd gw_remnant
pip install -e .[dev]

Dependencies

Core (always installed):

Package	Version
Python	>= 3.8
numpy	>= 1.20.0
scipy	>= 1.7.0
matplotlib	>= 3.3.0
gwtools	—

Optional extras:

Install option	Packages	Purpose
pip install gw_remnant[surrogates]	gwsurrogate, surfinBH	Waveform surrogate models and remnant fits
pip install gw_remnant[nr]	sxs, mayawaves	Numerical relativity catalog access
pip install gw_remnant[dev]	pytest, pytest-cov, black, flake8, ipython, jupyter	Development and testing
pip install gw_remnant[docs]	sphinx, pydata-sphinx-theme, sphinx-autodoc-typehints, nbsphinx, pandoc, ipykernel	Building documentation
pip install gw_remnant[eob]	pyseobnr	Effective-one-body waveforms
pip install gw_remnant[all]	gwsurrogate, surfinBH, sxs, mayawaves, pyseobnr	All non-dev optional dependencies

Quick Start

import numpy as np
from gw_remnant.gw_remnant_calculator import GWRemnantCalculator

# Provide time array and waveform modes (e.g. from a surrogate model)
# time: 1D array in geometric units (M)
# h_dict: dictionary of complex waveform modes, e.g. {(2,2): h22(t), ...}
calc = GWRemnantCalculator(time, h_dict, q=2.0,
                           chi1=[0, 0, 0.5])

# Print summary
calc.print_remnants()

# Access individual properties
props = calc.get_remnant_properties()
print(f"Remnant mass: {props['remnant_mass']:.6f} M")
print(f"Remnant spin: {props['remnant_spin']:.6f}")

# Plot diagnostics (optionally save to file)
calc.plot_mass_energy(save_path='mass_energy.png')

See the Quick Start guide for more details.

Architecture

The package uses a mixin-based class hierarchy that composes into a single entry point:

GWRemnantCalculator   (gw_remnant_calculator.py)
 ├── GWPlotter                    — diagnostic plots (mass, momentum, spin, kick, trajectory)
 ├── PeakLuminosityCalculator     — peak dE/dt via spline interpolation
 ├── AngularMomentumCalculator    — J_dot, Joft, remnant spin (scalar + 3-vector)
 ├── TrajectoryCalculator         — center-of-mass displacement x(t) = ∫v dt
 ├── LinearMomentumCalculator     — P_dot, Poft, kick velocity v(t) = P/M(t)
 ├── RemnantMassCalculator        — dE/dt, E_rad(t), Bondi mass M(t), remnant mass
 └── InitialEnergyMomenta         — initial E, L from post-Newtonian (PN) expressions

Everything is computed eagerly in __init__ — constructing a GWRemnantCalculator runs the full pipeline bottom-up.

gw_remnant/
├── gw_remnant/                        # Main package directory
│   ├── __init__.py                    # Package initialization
│   ├── gw_remnant_calculator.py       # Main calculator class
│   ├── gw_waveform_generator.py       # Waveform generation utilities
│   │
│   ├── gw_utils/                      # Utility functions
│   │   ├── __init__.py
│   │   ├── waveform_generator.py      # Waveform generation helpers
│   │   └── gw_plotter.py             # Plotting utilities
│   │
│   └── remnant_calculators/           # Remnant property calculators
│       ├── __init__.py
│       ├── remnant_mass_calculator.py
│       ├── remnant_spin_calculator.py
│       ├── kick_velocity_calculator.py
│       ├── trajectory_calculator.py
│       ├── peak_luminosity_calculator.py
│       └── initial_energy_momenta.py
│
├── docs/                              # Sphinx documentation source
├── tutorials/                         # Example notebooks and tutorials
├── tests/                             # Unit tests
├── pyproject.toml                     # Packaging configuration
├── README.md                          # This file
├── LICENSE                            # License information
└── .gitignore                         # Git ignore rules

Supported Waveform Models

The package includes built-in waveform generators for NRHybSur3dq8, NRSur7dq4, BHPTNRSur1dq1e4, and BHPTNRSur2dq1e3. It can also accept waveforms generated by any other method — see Providing Waveform Data.

Tutorials

Example notebooks are provided in the tutorials/ directory:

#	Notebook	Waveform source
1	Remnant from default waveforms	Built-in surrogate generators (BHPTNRSur1dq1e4, NRHybSur3dq8)
2	Remnant from NRSur7dq4 waveforms	NRSur7dq4 (precessing surrogate)
3	Remnant from NRHybSur3dq8 waveforms	NRHybSur3dq8 (aligned-spin surrogate)
4	Remnant from SXS quasi-circular simulations	SXS NR catalog (non-spinning, aligned-spin, precessing)
5	Remnant from SXS eccentric simulations	SXS NR catalog (eccentric orbits)
6	Remnant from Maya catalog waveforms	Maya/Georgia Tech NR catalog
7	Remnant from pySEOBNR waveforms	pySEOBNR EOB models (SEOBNRv5HM/PHM/EHM)

These are also rendered in the online documentation.

BHPTNRremnant

gw_remnant has been used in developing an NR-tuned perturbation-based remnant model that provides faithful estimates of the remnant properties for binaries with mass ratios ranging from q=3 to q=1000.

Citation

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 gw_remnant package or BHPTNRremnant surrogate models please cite the following paper:

@article{Islam:2022laz,
    author = "Islam, Tousif and Field, Scott E. and Khanna, Gaurav.",
    title = "{Remnant black hole properties from numerical-relativity-informed perturbation theory and implications for waveform modelling}",
    eprint = "https://arxiv.org/abs/2301.07215",
    archivePrefix = "arXiv",
    primaryClass = "gr-qc",
    month = "1",
    year = "2023"
}