bayesian_inference
Bayesian inference module for pulsar timing array analysis.
This module provides the main interface for performing Bayesian parameter estimation on pulsar timing array data. It serves as the orchestration layer that coordinates prior model specifications, parameter sampling, and NUTS inference.
The module handles parameters like: - Gravitational wave background amplitude (ha) and spectral index (γa) - Pulsar-specific red noise parameters (γp, σp) - White noise parameters (EFAC, EQUAD)
The implementation uses the Hellings-Downs correlation pattern for the gravitational wave background and models pulsar red noise as an Ornstein-Uhlenbeck process.
Parameters
Define a struct to store the parameters of the Kalman filter model.
display_prior_summary
Display a readable summary of all prior distributions.
Parameters
prior_specs : dict Dictionary containing prior distributions from get_prior_model_specs() n_pulsars : int Number of pulsars (for vector parameter information) logger : logging.Logger, optional Logger object for output. If None, gets the centralized argus logger.