calc_log_posterior
- zdm.MCMC.calc_log_posterior(param_vals, state, params, surveys_sep, Pn=False, pNreps=True, ptauw=False, log_halo=False, lin_host=False, ind_surveys=False, g0info=None)[source]
Calculate log-posterior probability for a parameter vector.
This is the main function called by emcee samplers. It evaluates the log-posterior (proportional to log-likelihood for uniform priors) by building grids and computing likelihoods for all surveys.
- Parameters:
param_vals (ndarray) – Array of parameter values for this MCMC step.
state (parameters.State) – State object to be updated with new parameter values.
params (dict) – Dictionary defining parameters to vary. Each key is a parameter name, with value dict containing ‘min’ and ‘max’ for prior bounds.
surveys_sep (list) – Two-element list: [non_repeater_surveys, repeater_surveys].
Pn (bool, optional) – Include Poisson likelihood for total number of FRBs. Default False.
pNreps (bool, optional) – Include likelihood for number of repeaters. Default True.
ptauw (bool, optional) – Include p(tau, width) likelihood. Default False.
log_halo (bool, optional) – Use log-uniform prior on DMhalo. Default False.
lin_host (bool, optional) – Use linear-uniform prior on host DM mean. Default False.
ind_surveys (bool, optional) – If True, return list of individual survey likelihoods. Default False.
g0info (list, optional) – Pre-computed [zDMgrid, zvals, DMvals] for speedup.
- Returns:
Log-posterior value. Returns -inf if parameters outside prior bounds. If ind_surveys=True, returns (llsum, ll_list) with individual likelihoods.
- Return type: