Data release accompanying the PixelPop papers, analyzing gravitational wave populations.The first dataset (in gwtc3_result_files) is the posterior samples for the runs presented in analysis of LIGO--Virgo--KAGRA data, following the third gravitational wave catalog, see https://arxiv.org/abs/2406.16844. We include a python notebook (example_plot.ipynb) showing how to create the plots presented in this paper.In v2, we also include samples from the predictive distributions. Due to the large uncertainties, marginalizing over the hyperposterior may be a poor representation of the inferred distribution, and so instead we provide samples from the median predictive distribution. That is, samples from the distribution shown in the central panels of the figures. The second dataset (in o4inj_result_files) is the posterior samples accompanying the runs presented in the technical background paper, see https://arxiv.org/abs/2406.16813.

Data release accompanying the PixelPop papers, analyzing gravitational wave populations.The first dataset (in gwtc3_result_files) is the posterior samples for the runs presented in analysis of LIGO--Virgo--KAGRA data, following the third gravitational wave catalog, see https://arxiv.org/abs/2406.16844. We include a python notebook (example_plot.ipynb) showing how to create the plots presented in this paper.In v2, we also include samples from the predictive distributions. Due to the large uncertainties, marginalizing over the hyperposterior may be a poor representation of the inferred distribution, and so instead we provide samples from the median predictive distribution. That is, samples from the distribution shown in the central panels of the figures. The second dataset (in o4inj_result_files) is the posterior samples accompanying the runs presented in the technical background paper, see https://arxiv.org/abs/2406.16813.

Data release accompanying the PixelPop papers, analyzing gravitational wave populations.The first dataset (in gwtc3_result_files) is the posterior samples for the runs presented in analysis of LIGO--Virgo--KAGRA data, following the third gravitational wave catalog, see https://arxiv.org/abs/2406.16844. We include a python notebook (example_plot.ipynb) showing how to create the plots presented in this paper.The second dataset (in o4inj_result_files) is the posterior samples accompanying the runs presented in the technical background paper, see https://arxiv.org/abs/2406.16813.

This is the data release associated with Vitale et al 2209.06978 Samples.zip: Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter (Follow the references in the Method section of the paper for a description of each sub-model): Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
[...]
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
Figures.zip: Contains PDFs for all figures in the paper, plus individual figures for p(costau) and dR/dcostau for each model. Drop me (Salvatore Vitale) an email if anything doesn't work, is missing, or if you spot issues. Thanks!

This is the data release associated with https://arxiv.org/abs/2209.06978 Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter: Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
{'alpha_chi': Uniform(minimum=1, maximum=5, name='alphachi', latex_label='$\alpha_\chi$', unit=None, boundary=None),
'beta_chi': Uniform(minimum=1, maximum=5, name='betachi', latex_label='$\beta_\chi$', unit=None, boundary=None),
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'power_law_slope_mass_ratio': Uniform(minimum=-2, maximum=7, name=None, latex_label='$\beta$', unit=None, boundary=None),
'power_law_slope_m1': Uniform(minimum=1, maximum=6, name=None, latex_label='$\alpha$', unit=None, boundary=None),
'mu_0_costilt': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'mu_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'sigma_0_costilt': DeltaFunction(peak=1, name=None, latex_label=None, unit=None),
'sigma_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'peak_branchingratio_m1': Uniform(minimum=0, maximum=0.25, name=None, latex_label='$\lambda$', unit=None, boundary=None),
'power_law_slope_redshift': Uniform(minimum=-2, maximum=10, name='lamb', latex_label='$\lambda_z$', unit=None, boundary=None),
'peak_mean_m1': Uniform(minimum=20, maximum=45, name=None, latex_label='$\mu_{m}$', unit=None, boundary=None),
'peak_sigma_m1': Uniform(minimum=1, maximum=10, name=None, latex_label='$\sigma_{m}$', unit=None, boundary=None),
'minmass_m1': Uniform(minimum=2, maximum=10, name=None, latex_label='$m_{\min}$', unit=None, boundary=None),
'maxmass_m1': Uniform(minimum=60, maximum=100, name=None, latex_label='$m_{\max}$', unit=None, boundary=None),
'low_end_smoothing_m1': Uniform(minimum=0, maximum=10, name=None, latex_label='$\delta_{m}$', unit=None, boundary=None)} Drop me (Salvatore Vitale) an email if anything doesn't work or if you spot issues. Thanks!

This is the data release associated with https://arxiv.org/abs/2209.06978 Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter: Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
{'alpha_chi': Uniform(minimum=1, maximum=5, name='alphachi', latex_label='$\alpha_\chi$', unit=None, boundary=None),
'beta_chi': Uniform(minimum=1, maximum=5, name='betachi', latex_label='$\beta_\chi$', unit=None, boundary=None),
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'power_law_slope_mass_ratio': Uniform(minimum=-2, maximum=7, name=None, latex_label='$\beta$', unit=None, boundary=None),
'power_law_slope_m1': Uniform(minimum=1, maximum=6, name=None, latex_label='$\alpha$', unit=None, boundary=None),
'mu_0_costilt': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'mu_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'sigma_0_costilt': DeltaFunction(peak=1, name=None, latex_label=None, unit=None),
'sigma_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'peak_branchingratio_m1': Uniform(minimum=0, maximum=0.25, name=None, latex_label='$\lambda$', unit=None, boundary=None),
'power_law_slope_redshift': Uniform(minimum=-2, maximum=10, name='lamb', latex_label='$\lambda_z$', unit=None, boundary=None),
'peak_mean_m1': Uniform(minimum=20, maximum=45, name=None, latex_label='$\mu_{m}$', unit=None, boundary=None),
'peak_sigma_m1': Uniform(minimum=1, maximum=10, name=None, latex_label='$\sigma_{m}$', unit=None, boundary=None),
'minmass_m1': Uniform(minimum=2, maximum=10, name=None, latex_label='$m_{\min}$', unit=None, boundary=None),
'maxmass_m1': Uniform(minimum=60, maximum=100, name=None, latex_label='$m_{\max}$', unit=None, boundary=None),
'low_end_smoothing_m1': Uniform(minimum=0, maximum=10, name=None, latex_label='$\delta_{m}$', unit=None, boundary=None)} Drop me (Salvatore Vitale) an email if anything doesn't work or if you spot issues. Thanks!

This is the data release associated with https://arxiv.org/abs/2209.06978 Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter: Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
{'alpha_chi': Uniform(minimum=1, maximum=5, name='alphachi', latex_label='$\alpha_\chi$', unit=None, boundary=None),
'beta_chi': Uniform(minimum=1, maximum=5, name='betachi', latex_label='$\beta_\chi$', unit=None, boundary=None),
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'power_law_slope_mass_ratio': Uniform(minimum=-2, maximum=7, name=None, latex_label='$\beta$', unit=None, boundary=None),
'power_law_slope_m1': Uniform(minimum=1, maximum=6, name=None, latex_label='$\alpha$', unit=None, boundary=None),
'mu_0_costilt': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'mu_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'sigma_0_costilt': DeltaFunction(peak=1, name=None, latex_label=None, unit=None),
'sigma_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'peak_branchingratio_m1': Uniform(minimum=0, maximum=0.25, name=None, latex_label='$\lambda$', unit=None, boundary=None),
'power_law_slope_redshift': Uniform(minimum=-2, maximum=10, name='lamb', latex_label='$\lambda_z$', unit=None, boundary=None),
'peak_mean_m1': Uniform(minimum=20, maximum=45, name=None, latex_label='$\mu_{m}$', unit=None, boundary=None),
'peak_sigma_m1': Uniform(minimum=1, maximum=10, name=None, latex_label='$\sigma_{m}$', unit=None, boundary=None),
'minmass_m1': Uniform(minimum=2, maximum=10, name=None, latex_label='$m_{\min}$', unit=None, boundary=None),
'maxmass_m1': Uniform(minimum=60, maximum=100, name=None, latex_label='$m_{\max}$', unit=None, boundary=None),
'low_end_smoothing_m1': Uniform(minimum=0, maximum=10, name=None, latex_label='$\delta_{m}$', unit=None, boundary=None)} Drop me (Salvatore Vitale) an email if anything doesn't work or if you spot issues. Thanks!

This is the data release associated with https://arxiv.org/abs/2209.06978 Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter: Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
{'alpha_chi': Uniform(minimum=1, maximum=5, name='alphachi', latex_label='$\alpha_\chi$', unit=None, boundary=None),
'beta_chi': Uniform(minimum=1, maximum=5, name='betachi', latex_label='$\beta_\chi$', unit=None, boundary=None),
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'power_law_slope_mass_ratio': Uniform(minimum=-2, maximum=7, name=None, latex_label='$\beta$', unit=None, boundary=None),
'power_law_slope_m1': Uniform(minimum=1, maximum=6, name=None, latex_label='$\alpha$', unit=None, boundary=None),
'mu_0_costilt': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
'mu_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'sigma_0_costilt': DeltaFunction(peak=1, name=None, latex_label=None, unit=None),
'sigma_1_costilt': DeltaFunction(peak=0.0, name=None, latex_label=None, unit=None),
'peak_branchingratio_m1': Uniform(minimum=0, maximum=0.25, name=None, latex_label='$\lambda$', unit=None, boundary=None),
'power_law_slope_redshift': Uniform(minimum=-2, maximum=10, name='lamb', latex_label='$\lambda_z$', unit=None, boundary=None),
'peak_mean_m1': Uniform(minimum=20, maximum=45, name=None, latex_label='$\mu_{m}$', unit=None, boundary=None),
'peak_sigma_m1': Uniform(minimum=1, maximum=10, name=None, latex_label='$\sigma_{m}$', unit=None, boundary=None),
'minmass_m1': Uniform(minimum=2, maximum=10, name=None, latex_label='$m_{\min}$', unit=None, boundary=None),
'maxmass_m1': Uniform(minimum=60, maximum=100, name=None, latex_label='$m_{\max}$', unit=None, boundary=None),
'low_end_smoothing_m1': Uniform(minimum=0, maximum=10, name=None, latex_label='$\delta_{m}$', unit=None, boundary=None)} The figures folder contains PDFs for all figures in the paper, plus individual figures for p(costau) and dR/dcostau for each model. Drop me (Salvatore Vitale) an email if anything doesn't work, is missing, or if you spot issues. Thanks!

This is the data release associated with Vitale et al 2209.06978 Samples.zip: Contains all of the hyper posterior samples for the runs listed in Tables G.1. The files are in json format. Bilby offers a dedicated routine to read them in import bilby
data= bilby.core.result.read_in_result(path_to_json) See the Bilby documentation for what is contained in the result object. For each run, we report the posterior hyper samples for the mass model, reshift model, spin magnitude model, spin tilt model and merger rate [Gpc^-3 yr^-1] Here the name used to store and a short description of each parameter (Follow the references in the Method section of the paper for a description of each sub-model): Primary mass model (Power Law + Peak for all runs) power_law_slope_m1, slope of the primary mass power law component minmass_m1, minimum BH mass maxmass_m1, maximum BH mass low_end_smoothing_m1, smoothing at the low-mass end peak_branchingratio_m1, branching ratio between Gaussian peak and power law (1= 100% peak) peak_mean_m1, mean of the Gaussian peak peak_sigma_m1, sigma of the Gaussian peak Mass ratio model (power law for all runs) power_law_slope_mass_ratio, slope of the mass ratio Redshift (power law for all runs) power_law_slope_redshift, slope of the redshift Spin magnitude (IID beta distributions for all runs) alpha_chi, first argument of beta distribution beta_chi, second argument of beta distribution Cosine of tilt angle Gaussian models mu_0_costilt, for Gaussian models w/o correlation, the mean of the left (or only) Gaussian sigma_0_costilt, for Gaussian models w/o correlation, the sigma of the left (or only) Gaussian mu_1_costilt, for Gaussian models w/o correlation, the mean of the right Gaussian sigma_1_costilt, for Gaussian models w/o correlation, the sigma of the right Gaussian mu_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian mean mu_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian mean sigma_a_costilt, for Gaussian model with correlation, the constant part of the Gaussian sigma sigma_b_costilt, for Gaussian model with correlation, the coefficient of the linearly evolving part of the Gaussian sigma Beta models alpha_a_costilt, for all Beta models, the constant part of the first parameter of the Beta distribution alpha_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the first parameter of the Beta distribution beta_a_costilt, for all Beta models, the constant part of the second parameter of the Beta distribution beta_b_costilt, for all Beta models, the coefficient of the linearly evolving part of the second parameter of the Beta distribution Tukey models: tukey_x0, the center of the Tukey as defined in appendix E of the paper tukey_k, Tk as defined in appendix E of the paper tukey_r, Tk as defined in appendix E of the paper Branching ratios: spin_mixture_0, for 2-component models, this is the branching ratio of the non-isotropic component spin_mixture_1, for Isotropic + Gaussian + Tukey and Isotropic + Gaussian + Beta this is the branching ratio of the Gaussian component; for Isotropic + 2 Gaussian this is the branching ratio of the Gaussian on the right. Merger rate rates, merger rate per unit Gpc cubed per unit year Note that some of the parameters for the tilt models might not be used, but still stored (and fixed to - usually - zero). This can be checked by verifying what priors were used for each parameter. For example the Isotropic run was obtained from the Isotropic + Gaussian model by setting the branching ratio of the Gaussian component to zero (at which point the values of mu and sigma costitl are irrelevant) > data['prior']
[...]
'spin_mixture_0': DeltaFunction(peak=0, name=None, latex_label=None, unit=None),
Figures.zip: Contains PDFs for all figures in the paper, plus individual figures for p(costau) and dR/dcostau for each model. Drop me (Salvatore Vitale) an email if anything doesn't work, is missing, or if you spot issues. Thanks!

Neutron star-black hole (NSBH) mergers detected in gravitational waves have the potential to shed light on supernova physics, the dense matter equation of state, and the astrophysical processes that power their potential electromagnetic counterparts. We use the population of four candidate NSBH events detected in gravitational waves so far with a false alarm rate ≤1 yr−1 to constrain the mass and spin distributions and multimessenger prospects of these systems. We find that the black holes in NSBHs are both less massive and more slowly spinning than those in black hole binaries. We also find evidence for a mass gap between the most massive neutron stars and least massive black holes in NSBHs at 98.6% credibility. We consider both a Gaussian and a power-law pairing function for the distribution of the mass ratio between the neutron star and black hole masses but find no statistical preference between the two. Using an approach driven by gravitational-wave data rather than binary simulations, we find that fewer than 14% of NSBH mergers detectable in gravitational waves will have an electromagnetic counterpart. Finally, we propose a method for the multimessenger analysis of NSBH mergers based on the nondetection of an electromagnetic counterpart and conclude that, even in the most optimistic case, the constraints on the neutron star equation of state that can be obtained with multimessenger NSBH detections are not competitive with those from gravitational-wave measurements of tides in binary neutron star mergers and radio and X-ray pulsar observations.