Skip to content

[BUG] failure of GFO estimators with replacement=False #135

@fkiraly

Description

@fkiraly

A number of GFO estimator seem to fail with replcaement=False in what seems like a sensible call, e.g.,

from gradient_free_optimizers import LipschitzOptimizer
from hyperactive.experiment.integrations import SklearnCvExperiment

experiment = SklearnCvExperiment.create_test_instance()

opt = LipschitzOptimizer(
        **{
            "search_space": {
                "C": [0.01, 0.1, 1, 10],
                "gamma": [0.0001, 0.01, 0.1, 1, 10],
            },
            "max_sample_size": 1000,
            "replacement": True,
        }
)

opt.search(
    objective_function=experiment.score,
    n_iter=n_iter,
    max_time=max_time,
)

Same with ForestOptimizer, TreeStructuredParzenEstimators

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions