API Reference

This page documents every public class in bayes_halving_search_cv and every parameter each one accepts. The text below is generated directly from each class’s own docstring (NumPyDoc format) — the source of truth lives in the docstrings themselves (src/bayes_halving_search_cv/_search.py and src/bayes_halving_search_cv/_bayes.py), not here.

Estimators

PatternSearchCV(estimator, param_grid, *[, ...])	Hooke-Jeeves pattern search over a discrete hyperparameter grid.
BayesHalvingSearchCV(estimator, param_grid, *)	Gaussian-Process Bayesian search over a discrete hyperparameter grid, with bullseye multi-fidelity data growth and scatter-search multi-start.

Search space

Both estimators above share the exact same search-space standard — see PatternSearchCV_SPEC.md and BAYESHALVINGSearchCV_SPEC.md §3.1 — built on the two classes below.

Space(param_grid)	Ordered collection of Dimensions; all point math lives here.
Dimension(name, spec)	One axis of the grid: sorted values, integer index addressing.