pytransc

PyTransC: A Python library for Trans-conceptual Bayesian sampling.

PyTransC implements various trans-conceptual MCMC algorithms for Bayesian inference across models of different states. The package provides:

  • Product space sampling using emcee
  • State-jump sampling with pseudo-priors
  • Ensemble resampling methods
  • Analysis tools for trans-conceptual chains
  • Utility functions for pseudo-prior construction

Examples

Basic usage with product space sampler:

>>> from pytransc.samplers import run_product_space_sampler
>>> # Set up your log_posterior and log_pseudo_prior functions
>>> results = run_product_space_sampler(
...     nwalkers=32, nsteps=1000,
...     pos_initial=initial_positions,
...     pos_state=initial_states,
...     log_posterior=my_log_posterior,
...     log_pseudo_prior=my_log_pseudo_prior
... )

 1"""PyTransC: A Python library for Trans-conceptual Bayesian sampling.
 2
 3PyTransC implements various trans-conceptual MCMC algorithms for Bayesian inference
 4across models of different states. The package provides:
 5
 6- Product space sampling using emcee
 7- State-jump sampling with pseudo-priors
 8- Ensemble resampling methods
 9- Analysis tools for trans-conceptual chains
10- Utility functions for pseudo-prior construction
11
12Examples
13--------
14Basic usage with product space sampler:
15
16    >>> from pytransc.samplers import run_product_space_sampler
17    >>> # Set up your log_posterior and log_pseudo_prior functions
18    >>> results = run_product_space_sampler(
19    ...     nwalkers=32, nsteps=1000,
20    ...     pos_initial=initial_positions,
21    ...     pos_state=initial_states,
22    ...     log_posterior=my_log_posterior,
23    ...     log_pseudo_prior=my_log_pseudo_prior
24    ... )
25"""
26
27__version__ = "0.3.0a1"