NumPyro likelihoods, sampling and storage

See Units, shapes and model contracts for units, shapes, domains and backend boundaries, and Quickstart for executable minimal examples.

AxisymmetricJeansLikelihoodModel(dsph_model, ...)	NumPyro likelihood for signed sky coordinates and zero mean streaming.
JeansLikelihoodModel(dsph_model, ...[, ...])	Callable spherical NumPyro model for line-of-sight velocity inference.
NumPyroSampler(mcmc, *, output_dir[, ...])	Composition-based helper around numpyro.infer.MCMC.
ParameterSpec(sample_name, distribution[, ...])	Describe a parameter site and its physical representation.
SamplerRunResult(resumed, checkpoint_path, ...)	Report one NumPyroSampler run and persistence request.
StorageBackend	Accepted sample-store names; install the dependencies for the selected backend.