def _sample_qmc(self, study: Study,
search_space: Dict[str, BaseDistribution]) -> np.ndarray:
# Lazy import because the `scipy.stats.qmc` is slow to import.
qmc_module = _LazyImport("scipy.stats.qmc")
sample_id = self._find_sample_id(study, search_space)
d = len(search_space)
if self._qmc_type == "halton":
qmc_engine = qmc_module.Halton(d,
seed=self._seed,
scramble=self._scramble)
elif self._qmc_type == "sobol":
qmc_engine = qmc_module.Sobol(d,
seed=self._seed,
scramble=self._scramble)
else:
raise ValueError("Invalid `qmc_type`")
forward_size = sample_id # `sample_id` starts from 0.
qmc_engine.fast_forward(forward_size)
sample = qmc_engine.random(1)
return sample
from typing import Union
import warnings
from cliff.app import App
from cliff.command import Command
from cliff.commandmanager import CommandManager
import yaml
import optuna
from optuna._imports import _LazyImport
from optuna.exceptions import CLIUsageError
from optuna.exceptions import ExperimentalWarning
from optuna.storages import RDBStorage
from optuna.trial import TrialState
_dataframe = _LazyImport("optuna.study._dataframe")
_DATETIME_FORMAT = "%Y-%m-%d %H:%M:%S"
def _check_storage_url(storage_url: Optional[str]) -> str:
if storage_url is None:
raise CLIUsageError("Storage URL is not specified.")
return storage_url
def _format_value(value: Any) -> Any:
# Format value that can be serialized to JSON or YAML.
if value is None or isinstance(value, (int, float)):
return value
__all__ = [
"Study",
"TYPE_CHECKING",
"Trial",
"TrialPruned",
"__version__",
"copy_study",
"create_study",
"create_trial",
"delete_study",
"distributions",
"exceptions",
"get_all_study_summaries",
"importance",
"integration",
"load_study",
"logging",
"multi_objective",
"pruners",
"samplers",
"storages",
"study",
"trial",
"version",
"visualization",
]
importance = _LazyImport("optuna.importance")
visualization = _LazyImport("optuna.visualization")
from optuna._imports import _LazyImport
from optuna.multi_objective import samplers # NOQA
from optuna.multi_objective import study # NOQA
from optuna.multi_objective import trial # NOQA
from optuna.multi_objective.study import create_study # NOQA
from optuna.multi_objective.study import load_study # NOQA
visualization = _LazyImport("optuna.multi_objective.visualization")
from typing import NamedTuple
from typing import Optional
from typing import Tuple
from typing import TYPE_CHECKING
import numpy as np
from optuna import distributions
from optuna._imports import _LazyImport
from optuna.distributions import BaseDistribution
if TYPE_CHECKING:
import scipy.special as special
import scipy.stats as stats
else:
special = _LazyImport("scipy.special")
stats = _LazyImport("scipy.stats")
EPS = 1e-12
SIGMA0_MAGNITUDE = 0.2
_DISTRIBUTION_CLASSES = (
distributions.CategoricalDistribution,
distributions.FloatDistribution,
distributions.IntDistribution,
)
class _ParzenEstimatorParameters(
NamedTuple(
"_ParzenEstimatorParameters",
__all__ = [
"Study",
"TYPE_CHECKING",
"Trial",
"TrialPruned",
"__version__",
"copy_study",
"create_study",
"create_trial",
"delete_study",
"distributions",
"exceptions",
"get_all_study_summaries",
"importance",
"integration",
"load_study",
"logging",
"multi_objective",
"pruners",
"samplers",
"storages",
"study",
"trial",
"version",
"visualization",
]
importance = _LazyImport("optuna.importance")
type_checking = _LazyImport("optuna.type_checking")
visualization = _LazyImport("optuna.visualization")