API Reference

This section provides the API documentation for the various components of jaxpe.

• core: Priors, unconstraining transforms, and the InferenceProblem interface.
• gw: Detectors, PSDs, data handling, FD likelihood, and GW priors.
• kernels: Local MCMC kernels + step-size/mass adaptation.
• sampler: The global-local orchestration loop.
• flows: Normalizing-flow wrapper and trainer.
• diagnostics: R-hat, ESS, corner/trace plots.
• surrogate: Active-learning GP surrogates for expensive, non-differentiable likelihoods.
• config: The declarative run configuration — prior and injection distributions, conditioning, budgets, seeds.
• drivers: The shared end-to-end PE schedule behind the bin/ relative-binning benchmarks — MAP+Laplace init, warmup/equilibration/production, convergence gating.

---

Table of contents

• gw
• kernels
• flows
• sampler
• core
• diagnostics
• surrogate
• config
• drivers