arviz.PyMCSamplingWrapper#

class arviz.PyMCSamplingWrapper(model, idata_orig=None, log_lik_fun=None, is_ufunc=True, posterior_vars=None, sample_kwargs=None, idata_kwargs=None, log_lik_kwargs=None, apply_ufunc_kwargs=None)[source]#

PyMC (4.0+) sampling wrapper base class.

See the documentation on SamplingWrapper for a more detailed description. An example of PyMCSamplingWrapper usage can be found in the pymc_refitting notebook.

Warning

Sampling wrappers are an experimental feature in a very early stage. Please use them with caution.

Methods

`PyMCSamplingWrapper.__init__`(model[, ...])
`PyMCSamplingWrapper.check_implemented_methods`(methods)	Check that all methods listed are implemented.
`PyMCSamplingWrapper.get_inference_data`(...)	Return sampling result without modifying.
`PyMCSamplingWrapper.log_likelihood__i`(...)	Get the log likelilhood samples \(\log p_{post(-i)}(y_i)\).
`PyMCSamplingWrapper.sample`(...)	Update data and sample model on modified_observed_data.
`PyMCSamplingWrapper.sel_observations`(idx)	Select a subset of the observations in idata_orig.