Struct BohbSampler 

pub struct BohbSampler { /* private fields */ }

A BOHB sampler that combines TPE with HyperBand budget awareness.

BOHB filters trial history by budget level before delegating to TPE, so the surrogate model is conditioned on trials evaluated at the same resource level. This produces better-calibrated parameter proposals than using a single global model across all budgets.

Use BohbSampler::matching_pruner to create a HyperbandPruner with matching HyperBand parameters.

Examples

use optimizer::parameter::{FloatParam, Parameter};
use optimizer::sampler::bohb::BohbSampler;
use optimizer::{Direction, Study};

let bohb = BohbSampler::builder()
    .min_resource(1)
    .max_resource(27)
    .reduction_factor(3)
    .seed(42)
    .build()
    .unwrap();
let pruner = bohb.matching_pruner(Direction::Minimize);
let study: Study<f64> = Study::with_sampler_and_pruner(Direction::Minimize, bohb, pruner);

Implementations

impl BohbSampler

pub fn new() -> Self

Creates a new BOHB sampler with default settings.

Defaults:

• min_resource: 1
• max_resource: 81
• reduction_factor: 3
• min_points_in_model: 10
• TPE: default settings

pub fn builder() -> BohbSamplerBuilder

Creates a builder for configuring a BOHB sampler.

Examples

use optimizer::sampler::bohb::BohbSampler;

let sampler = BohbSampler::builder()
    .min_resource(1)
    .max_resource(27)
    .reduction_factor(3)
    .min_points_in_model(5)
    .seed(42)
    .build()
    .unwrap();

pub fn matching_pruner(&self, direction: Direction) -> HyperbandPruner

Creates a HyperbandPruner with matching Hyperband parameters.

This ensures the pruner’s budget schedule is consistent with the budget levels used by BOHB for model conditioning.

Trait Implementations

impl Default for BohbSampler

fn default() -> Self

Returns the “default value” for a type.

impl Sampler for BohbSampler

fn sample( &self, distribution: &Distribution, trial_id: u64, history: &[CompletedTrial], ) -> ParamValue

Samples a parameter value from the given distribution.