Module sampler 

Hyperparameter samplers for optimization studies.

• GridSampler — exhaustive cartesian product, lazy index-based
• RandomSampler — uniform sampling with deterministic seeds
• BayesianSampler — TPE (Tree-Parzen Estimator) with explore/exploit

Re-exports

pub use bayesian::BayesianSampler;

Modules

bayesian

Structs

GridSampler

Exhaustive grid search over all combinations.

RandomSampler

Random search: sample uniformly from each dimension.

Traits

Sampler

A sampler produces hyperparameter configurations from a search space.

Functions

hash_u64

Simple hash combiner for generating unique RNG states.

pseudo_random

Simple deterministic pseudo-random (public for use by BayesianSampler): hash-based, returns [0.0, 1.0).

sample_float

Sample a float from [low, high] given t in [0, 1), respecting scale.

splitmix64

SplitMix64 hash function.