GWForge.population_fisher.model¶

The population models GWForge forecasts on, with their analytic scores.

Attributes¶

`MASS_GRID_NODES`
`POWER_LAW_SERIES_THRESHOLD`
`DEFAULT_MAXIMUM_SPIN`
`DEFAULT_MINIMUM_COSINE_TILT`
`NORMALISATION_NODES`
`OBSERVABLE_KEYS`

Classes¶

`BrokenPowerLawTwoPeakMass`	BGP primary mass and power-law mass ratio, with analytic scores.
`DefaultSpin`	GWTC-5.0 Default BBH spins: truncated-Gaussian magnitudes, mixture tilts.
`MadauDickinsonRedshift`	Merger-rate density following the Madau-Dickinson shape.
`SpectralSirenModel`	Source-frame population observed in the detector frame, cosmology free.

Functions¶

`mass_grid`(mmin, maximum_mass[, nodes])	The shared normalisation grid: quadratically clustered towards `mmin`.
`broken_power_law_norm`(alpha_1, alpha_2, m_break, mmin, ...)	Normalisation of `GWForge.population._smoothed_mass.broken_power_law()`.
`broken_power_law_norm_derivatives`(alpha_1, alpha_2, ...)	`dN/dLambda` for the broken power law, as a dict.
`left_truncated_normal_dlog`(mass, mu, sigma, mmin)	`d ln N_lt / d(mu, sigma)` for the left-truncated normal.
`smoothing_ddelta_m`(mass, mmin, mmax, delta_m)	`dS/d(delta_m)` for the Planck taper.
`smoothing_dmass`(mass, mmin, mmax, delta_m)	`dS/dm` for the Planck taper, at fixed hyper-parameters.

Module Contents¶

GWForge.population_fisher.model.MASS_GRID_NODES = 20000[source]¶

GWForge.population_fisher.model.POWER_LAW_SERIES_THRESHOLD = 1e-08[source]¶

GWForge.population_fisher.model.mass_grid(mmin, maximum_mass, nodes=MASS_GRID_NODES)[source]¶

The shared normalisation grid: quadratically clustered towards mmin.

Parameters:

mmin (float)
maximum_mass (float)
nodes (int)

Return type:

numpy.ndarray

GWForge.population_fisher.model.broken_power_law_norm(alpha_1, alpha_2, m_break, mmin, m_high)[source]¶

Normalisation of GWForge.population._smoothed_mass.broken_power_law().

Return type:: float

GWForge.population_fisher.model.broken_power_law_norm_derivatives(alpha_1, alpha_2, m_break, mmin, m_high)[source]¶

dN/dLambda for the broken power law, as a dict.

Returns:: Keys "alpha_1", "alpha_2", "m_break".
Return type:: dict

GWForge.population_fisher.model.left_truncated_normal_dlog(mass, mu, sigma, mmin)[source]¶

d ln N_lt / d(mu, sigma) for the left-truncated normal.

Returns:: Keys "mu", "sigma", each an array shaped like mass.
Return type:: dict

GWForge.population_fisher.model.smoothing_ddelta_m(mass, mmin, mmax, delta_m)[source]¶

dS/d(delta_m) for the Planck taper.

Return type:: numpy.ndarray

GWForge.population_fisher.model.smoothing_dmass(mass, mmin, mmax, delta_m)[source]¶

dS/dm for the Planck taper, at fixed hyper-parameters.

Return type:: numpy.ndarray

class GWForge.population_fisher.model.BrokenPowerLawTwoPeakMass(mmin=BGP_PARAMETERS['mmin'], m_high=BGP_PARAMETERS['m_high'], maximum_mass=BGP_PARAMETERS['maximum_mass'], mmin_2=BGP_PARAMETERS['mmin_2'], delta_m_2=BGP_PARAMETERS['delta_m_2'], fiducial=None, nodes=MASS_GRID_NODES)[source]¶

Bases: GWForge.population_fisher.base.PopulationModel

BGP primary mass and power-law mass ratio, with analytic scores.

Variables:

parameter_names (list of str) – The eleven free hyper-parameters, using GWForge.population.mass.Mass’s spelling so that a generation config and a Fisher config read the same.
maximum_mass (mmin, m_high,) – Support parameters, fixed by construction: hard cutoffs have a constant score, so their centred score vanishes and they are not measurable within Term I.

parameter_names = ['alpha_1', 'alpha_2', 'm_break', 'lam_0', 'lam_1', 'mpp_1', 'sigpp_1', 'mpp_2', 'sigpp_2',...[source]¶

event_keys = ['mass_1_source'][source]¶

mmin[source]¶

mmin_2[source]¶

m_high[source]¶

maximum_mass[source]¶

fiducial[source]¶

delta_m_2[source]¶

grid[source]¶

dlog_prob_dmasses(events, parameters)[source]¶

(d ln p / d m_1, d ln p / d m_2) at fixed hyper-parameters.

Returns:: (d ln p / d m_1, d ln p / d m_2), nan outside the support.
Return type:: tuple of numpy.ndarray

log_prob(events, parameters)[source]¶: See GWForge.population_fisher.base.PopulationModel.log_prob().

analytic_score(events, parameters)[source]¶: Closed-form score, (n_events, 11).

GWForge.population_fisher.model.DEFAULT_MAXIMUM_SPIN = 1.0[source]¶

GWForge.population_fisher.model.DEFAULT_MINIMUM_COSINE_TILT = -1.0[source]¶

class GWForge.population_fisher.model.DefaultSpin(fiducial=None, maximum_spin=DEFAULT_MAXIMUM_SPIN, minimum_cosine_tilt=DEFAULT_MINIMUM_COSINE_TILT)[source]¶

Bases: GWForge.population_fisher.base.PopulationModel

GWTC-5.0 Default BBH spins: truncated-Gaussian magnitudes, mixture tilts.

Variables:: parameter_names (list of str) – ["mu_chi", "sigma_chi", "mu_t", "sigma_t", "xi_spin"] – the names GWForge.population.spin.Spin takes in spin-parameters. The posterior files call mu_t/sigma_t mu_spin/sigma_spin.

parameter_names = ['mu_chi', 'sigma_chi', 'mu_t', 'sigma_t', 'xi_spin'][source]¶

event_keys = ['a_1', 'a_2'][source]¶

fiducial[source]¶

maximum_spin[source]¶

minimum_cosine_tilt[source]¶

isotropic_density[source]¶

check_events(events)[source]¶: Raise if the magnitudes or either spelling of the tilts is missing.

log_prob(events, parameters)[source]¶: See GWForge.population_fisher.base.PopulationModel.log_prob().

analytic_score(events, parameters)[source]¶: Closed-form score, (n_events, 5).

GWForge.population_fisher.model.NORMALISATION_NODES = 4000[source]¶

class GWForge.population_fisher.model.MadauDickinsonRedshift(maximum_redshift=10.0, cosmology=None, fiducial=None, nodes=NORMALISATION_NODES)[source]¶

Bases: GWForge.population_fisher.base.PopulationModel

Merger-rate density following the Madau-Dickinson shape.

Variables:: parameter_names (list of str) – ["gamma", "kappa", "z_peak"] – the names GWForge.population.redshift.Redshift uses, so a generation config and a Fisher config read the same.

parameter_names = ['gamma', 'kappa', 'z_peak'][source]¶

event_keys = ['redshift'][source]¶

maximum_redshift[source]¶

cosmology = None[source]¶

fiducial[source]¶

nodes[source]¶

log_shape(redshift, parameters)[source]¶: ln psi(z): the Madau-Dickinson shape alone, without the measure.

log_prob(events, parameters)[source]¶: See GWForge.population_fisher.base.PopulationModel.log_prob().

shape_gradient(redshift, parameters)[source]¶: d ln psi / dLambda, shape (n, 3).

analytic_score(events, parameters)[source]¶: Closed-form score, (n_events, 3).

GWForge.population_fisher.model.OBSERVABLE_KEYS = ('mass_1', 'mass_2', 'luminosity_distance')[source]¶

class GWForge.population_fisher.model.SpectralSirenModel(mass_model, redshift_model, spin_model=None, cosmology=None, cosmology_parameters=COSMOLOGY_PARAMETERS, nodes=None)[source]¶

Bases: GWForge.population_fisher.base.PopulationModel

Source-frame population observed in the detector frame, cosmology free.

Variables:

parameter_names (list of str) – Mass, then redshift, then spin, then the cosmology parameters.
maximum_redshift (float) – Upper edge of the redshift support, taken from the redshift model.

mass_model[source]¶

redshift_model[source]¶

spin_model = None[source]¶

cosmology_parameters = ['H0', 'Om0', 'w0'][source]¶

maximum_redshift[source]¶

blocks[source]¶

parameter_names = [][source]¶

event_keys = ['mass_1', 'mass_2', 'luminosity_distance'][source]¶

nodes[source]¶

property fiducial[source]¶: Block fiducials plus the cosmology, merged live.

log_prob(events, parameters)[source]¶: See GWForge.population_fisher.base.PopulationModel.log_prob().

analytic_score(events, parameters)[source]¶: Closed-form score, (n_events, len(parameter_names)).