Waveform Models

JenpyROQ separates ROQ construction from waveform generation through wrapper classes in JenpyROQ/waveform_wrappers.py. A wrapper is responsible for accepting a parameter dictionary and returning h_+ and h_x on the requested frequency grid.

Diagram of the JenpyROQ waveform wrapper interface.

The ROQ algorithm calls one common wrapper interface while the wrapper handles library-specific waveform generation.

The wrapper contract is:

class MyWrapper:
    def __init__(self, approximant, additional_waveform_params=None):
        ...

    def generate_waveform(self, p, deltaF, f_min, f_max, distance):
        ...
        return hp, hc

Available Approximant Names

The command-line help lists the intended user-facing approximants. The source registers a wrapper only when the corresponding optional dependency can be imported.

Approximant family

Names

Dependency

LALSimulation

IMRPhenomD, IMRPhenomPv2, IMRPhenomPv3, IMRPhenomPv3HM, IMRPhenomXHM, IMRPhenomXPHM, TaylorF2Ecc, IMRPhenomPv2_NRTidal, IMRPhenomNSBH

lal and lalsimulation

TEOBResumS

teobresums-giotto

EOBRun_module

MLGW-BNS standalone

mlgw-bns-standalone

mlgw_bns

bajes inspiral

mlgw-bns

bajes

NRPMw and hybrids

nrpmw, nrpmw-recal, nrpmw-merger, nrpmw-recal-merger, teobresums-spa-nrpmw, teobresums-spa-nrpmw-recal, mlgw-bns-nrpmw, mlgw-bns-nrpmw-recal

bajes and the relevant waveform backend

If an optional module is missing, the import prints a warning and that wrapper is not registered. A run requesting an unregistered wrapper raises ValueError('Unknown approximant requested.').

Spin Handling

spins and spin-sph decide which spin variables are sampled. The wrapper still decides which of those variables the underlying waveform model can handle.

Setting

Active variables

Notes

spins = no-spins

no spin variables

Wrappers fill missing spin components with zero where supported.

spins = aligned

s1z, s2z

The common path for BNS examples.

spins = precessing and spin-sph = 0

s1x, s1y, s1z, s2x, s2y, s2z

LAL precessing models can use these. MLGW-BNS and bajes wrappers reject non-zero in-plane spins.

spins = precessing and spin-sph = 1

s1s1 through s2s3

Converted internally to Cartesian components before waveform calls.

Tidal And Eccentricity Parameters

tides = 1 activates lambda1 and lambda2. Several BNS wrappers require these parameters. MLGW-BNS checks that the values lie between 5 and 5000. The bajes wrappers reject values above 5000.

eccentricity = 1 activates the ecc parameter only when paired with a waveform model that supports the requested eccentric domain.

Post-Merger Parameters

post-merger = 1 activates:

Parameter

Default range

Meaning

nrpmw-tcoll

[0, 3000]

Collapse time parameter for NRPMw models.

nrpmw-df2

[-1e-5, 1e-5]

NRPMw frequency-derivative parameter.

nrpmw-phi

[0, 2*pi]

Post-merger phase used by attached or merger models.

For recalibrated NRPMw approximants, the code tries to import recalibration parameter names and bounds from bajes. Those extra parameters are added as nrpmw-<name> only when the import succeeds and the approximant name contains nrpmw-recal.

TEOB Dynamics Parameters

dynamics = 1 activates:

Parameter

Default range

Meaning

TEOBResumS_a6c

[-100, -20]

EOB dynamics parameter mapped to a6c before calling TEOBResumS.

TEOBResumS_cN3LO

[-100, -20]

EOB dynamics parameter mapped to cN3LO before calling TEOBResumS.