blackjax.ns

Nested sampling algorithms.

Nested sampling is a Monte Carlo method for Bayesian computation, used for evidence (marginal likelihood) estimation and posterior sampling.

Available modules:

• base: Core components for Nested Sampling.

• adaptive: Adaptive nested sampling combining SMC-style adaptive tempering with per-step inner-kernel parameter tuning and evidence tracking.

• nss: Nested slice sampling, with hit-and-run (build_kernel) or slice-within-Gibbs (build_swig_kernel) inner kernels.

• integrator: NSIntegrator for tracking evidence integration.

• utils: Utility functions for processing nested sampling results.

• from_mcmc: Utilities to build nested sampling algorithms from MCMC kernels.

Submodules

• blackjax.ns.adaptive
• blackjax.ns.base
• blackjax.ns.from_mcmc
• blackjax.ns.integrator
• blackjax.ns.nss
• blackjax.ns.utils