cotengra.hyperoptimizers.hyper_es

Hyper optimization using a steady-state diagonal evolutionary strategy.

Attributes

_CHI1

Classes

SteadyStateES	A steady-state diagonal evolutionary strategy (SSES) operating in raw
ESOptLib	Hyper-optimization using a steady-state diagonal ES.

Functions

_reflect(x)

Reflect a scalar into [-1, 1].

Module Contents

cotengra.hyperoptimizers.hyper_es._CHI1

cotengra.hyperoptimizers.hyper_es._reflect(x)

Reflect a scalar into [-1, 1].

class cotengra.hyperoptimizers.hyper_es.SteadyStateES(space, seed=None, population_size='auto', sigma=0.5, sigma_min=0.01, sigma_max=1.0, c_sigma='auto', d_sigma='auto', elite_ratio=0.5, use_mirror=True, inject_best_every='auto', restart_patience='auto', explore_prob=0.05, mean_lr=0.5, exponential_param_power=None)

A steady-state diagonal evolutionary strategy (SSES) operating in raw [-1, 1] parameter space.

Unlike a generational ES, this sampler has no synchronization barrier. Every ask draws from the current distribution immediately, and every tell updates the distribution state from a sliding window of recent results. Trials can be in-flight in any number and can complete in any order.

Step sizes are adapted per dimension using a simplified cumulative step-size adaptation (CSA) rule, giving the optimizer the character of a separable (diagonal) CMA-ES without any matrix operations.

Parameters:

• space (dict[str, dict]) – The search space for a single contraction method.

• seed (None or int, optional) – Random seed.

• population_size (int or "auto", optional) – Size of the sliding archive window. When "auto" it is chosen as max(8, 4 * ndim).

• sigma (float, optional) – Initial per-dimension step size in raw [-1, 1] space.

• sigma_min (float, optional) – Lower bound for per-dimension step sizes.

• sigma_max (float, optional) – Upper bound for per-dimension step sizes (also the restart value).

• c_sigma (float or "auto", optional) – Learning rate for the evolution path. When "auto" it is set to 1 / sqrt(ndim) (clamped to [0.05, 1.0]).

• d_sigma (float or "auto", optional) – Damping factor for step-size updates. When "auto" it is set to 1 + sqrt(ndim).

• elite_ratio (float, optional) – Fraction of the archive used as elites for weighted recombination.

• use_mirror (bool, optional) – If True, every fresh perturbation is paired with its antipodal mirror, halving gradient variance.

• inject_best_every (int or "auto", optional) – Inject the best-so-far point every this many asks. When "auto" it equals population_size.

• restart_patience (int or "auto", optional) – Number of tells without improvement before resetting step sizes and re-centering on the best-so-far point. When "auto" it equals 5 * population_size.

• explore_prob (float, optional) – Probability of issuing a uniform random point instead of the ES-directed point. Maintains global diversity throughout the search and helps escape local minima. Random points still feed into the archive normally.

• mean_lr (float, optional) – Learning rate for the mean update in (0, 1]. Each tell blends the current mean toward the elite weighted mean by this factor: mean = (1 - mean_lr) * mean + mean_lr * elite_mean. Values below 1.0 smooth out noise-driven mean drift. Default is 1.0 (full update, original behaviour).

• exponential_param_power (float, optional) – Passed through to the shared parameter mapping for FLOAT_EXP parameters.

rng

params = []

ndim

population_size = 'auto'

sigma0 = 0.5

sigma_min = 0.01

sigma_max = 1.0

c_sigma = 'auto'

d_sigma = 'auto'

elite_ratio = 0.5

use_mirror = True

explore_prob = 0.05

mean_lr = 0.5

inject_best_every = 'auto'

restart_patience = 'auto'

_weights

_mu

_mu_eff

_init_state()

Initialize or reset the mutable optimizer state.

ask()

Return the next candidate point.

Draws from the current distribution, optionally using mirrored sampling or best-so-far injection. No synchronization with outstanding trials is required.

tell(trial_number, score)

Record a completed trial and update the distribution.

Updates the sliding archive, recomputes the weighted mean from elites, adapts per-dimension step sizes via evolution paths, and checks for stalls.

_restart()

Reset step sizes and re-center.

Alternates between re-centering on the best-so-far point (exploitation restart) and a uniformly random point (exploration restart) to avoid repeated restarts converging to the same local minimum.

class cotengra.hyperoptimizers.hyper_es.ESOptLib

Bases: cotengra.hyperoptimizers.hyper.HyperOptLib

Hyper-optimization using a steady-state diagonal ES.

setup(methods, space, optimizer=None, population_size='auto', sigma=0.5, sigma_min=0.01, sigma_max=1.0, c_sigma='auto', d_sigma='auto', elite_ratio=0.5, use_mirror=True, inject_best_every='auto', restart_patience='auto', explore_prob=0.05, mean_lr=1.0, method_exploration=1.0, method_temperature=1.0, exponential_param_power=None, seed=None, **kwargs)

Initialize ES optimizers for each contraction method.

Parameters:

• methods (list[str]) – The contraction methods to optimize over.

• space (dict[str, dict[str, dict]]) – The per-method hyperparameter search space.

• optimizer (HyperOptimizer, optional) – The parent optimizer.

• population_size (int or "auto", optional) – Archive window size for each method-specific ES sampler.

• sigma (float, optional) – Initial per-dimension step size.

• sigma_min (float, optional) – Bounds for per-dimension step sizes.

• sigma_max (float, optional) – Bounds for per-dimension step sizes.

• c_sigma (float or "auto", optional) – Evolution path learning rate.

• d_sigma (float or "auto", optional) – Step-size damping factor.

• elite_ratio (float, optional) – Fraction of the archive used as elites.

• use_mirror (bool, optional) – Enable mirrored sampling.

• inject_best_every (int or "auto", optional) – Best-injection frequency.

• restart_patience (int or "auto", optional) – Tells without improvement before restart.

• explore_prob (float, optional) – Probability of issuing a uniform random point on each ask.

• mean_lr (float, optional) – Learning rate for the mean update in (0, 1].

• method_exploration (float, optional) – Exploration strength for the LCB-based method chooser.

• method_temperature (float, optional) – Noise temperature for the LCB-based method chooser.

• exponential_param_power (float, optional) – Passed to the shared parameter mapping for FLOAT_EXP.

• seed (None or int, optional) – Random seed.

get_setting()

Choose a contraction method, then request its next setting.

report_result(setting, trial, score)

Report a completed trial back to the method chooser and ES.