def get_bounds_and_task(
search_space: SearchSpace, param_names: List[str]
) -> Tuple[List[Tuple[float, float]], List[int], Dict[int, float]]:
"""Extract box bounds from a search space in the usual Scipy format.
Identify integer parameters as task features.
"""
bounds: List[Tuple[float, float]] = []
task_features: List[int] = []
target_fidelities: Dict[int, float] = {}
for i, p_name in enumerate(param_names):
p = search_space.parameters[p_name]
# Validation
if not isinstance(p, RangeParameter):
raise ValueError(f"{p} not RangeParameter")
elif p.log_scale:
raise ValueError(f"{p} is log scale")
# Set value
bounds.append((p.lower, p.upper))
if p.parameter_type == ParameterType.INT:
task_features.append(i)
if p.is_fidelity:
if not isinstance(not_none(p.target_value), (int, float)):
raise NotImplementedError(
"Only numerical target values currently supported.")
target_fidelities[i] = float(
checked_cast_to_tuple((int, float), p.target_value))
return bounds, task_features, target_fidelities
def _make_range_param(name: str, representation: TParameterRepresentation,
parameter_type: Optional[str]) -> RangeParameter:
assert "bounds" in representation, "Bounds are required for range parameters."
bounds = representation["bounds"]
assert isinstance(bounds, list) and len(bounds) == 2, (
f"Cannot parse parameter {name}: for range parameters, json representation "
"should include a list of two values, lower and upper bounds of the bounds."
)
return RangeParameter(
name=name,
parameter_type=_to_parameter_type(bounds, parameter_type, name,
"bounds"),
lower=checked_cast_to_tuple((float, int), bounds[0]),
upper=checked_cast_to_tuple((float, int), bounds[1]),
log_scale=checked_cast(bool, representation.get("log_scale", False)),
is_fidelity=checked_cast(bool,
representation.get("is_fidelity", False)),
)
def _make_range_param(
name: str, representation: TParameterRepresentation, parameter_type: Optional[str]
) -> RangeParameter:
assert "bounds" in representation, "Bounds are required for range parameters."
bounds = representation["bounds"]
assert isinstance(bounds, list) and len(bounds) == 2, (
f"Cannot parse parameter {name}: for range parameters, json representation "
"should include a list of two values, lower and upper bounds of the range."
)
return RangeParameter(
name=name,
parameter_type=_to_parameter_type(bounds, parameter_type, name, "bounds"),
lower=checked_cast_to_tuple((float, int), bounds[0]),
upper=checked_cast_to_tuple((float, int), bounds[1]),
log_scale=checked_cast(bool, representation.get("log_scale", False)),
digits=representation.get("digits", None), # pyre-ignore[6]
is_fidelity=checked_cast(bool, representation.get("is_fidelity", False)),
# pyre-fixme[6]: Expected `Union[None, bool, float, int, str]` for 8th param
# but got `Union[None, List[Union[None, bool, float, int, str]], bool, float,
# int, str]`.
target_value=representation.get("target_value", None),
)
def extract_search_space_digest(search_space: SearchSpace,
param_names: List[str]) -> SearchSpaceDigest:
"""Extract basic parameter prpoerties from a search space."""
bounds: List[Tuple[Union[int, float], Union[int, float]]] = []
ordinal_features: List[int] = []
categorical_features: List[int] = []
discrete_choices: Dict[int, List[Union[int, float]]] = {}
task_features: List[int] = []
fidelity_features: List[int] = []
target_fidelities: Dict[int, Union[int, float]] = {}
for i, p_name in enumerate(param_names):
p = search_space.parameters[p_name]
if isinstance(p, ChoiceParameter):
if p.is_task:
task_features.append(i)
elif p.is_ordered:
ordinal_features.append(i)
else:
categorical_features.append(i)
# at this point we can assume that values are numeric due to transforms
discrete_choices[i] = p.values # pyre-ignore [6]
bounds.append((min(p.values), max(p.values))) # pyre-ignore [6]
elif isinstance(p, RangeParameter):
if p.log_scale:
raise ValueError(f"{p} is log scale")
if p.parameter_type == ParameterType.INT:
ordinal_features.append(i)
d_choices = list(range(int(p.lower), int(p.upper) + 1))
discrete_choices[i] = d_choices # pyre-ignore [6]
bounds.append((p.lower, p.upper))
else:
raise ValueError(f"Unknown parameter type {type(p)}")
if p.is_fidelity:
if not isinstance(not_none(p.target_value), (int, float)):
raise NotImplementedError(
"Only numerical target values are supported.")
target_fidelities[i] = checked_cast_to_tuple((int, float),
p.target_value)
fidelity_features.append(i)
return SearchSpaceDigest(
feature_names=param_names,
bounds=bounds,
ordinal_features=ordinal_features,
categorical_features=categorical_features,
discrete_choices=discrete_choices,
task_features=task_features,
fidelity_features=fidelity_features,
target_fidelities=target_fidelities,
)
def extract_search_space_digest(search_space: SearchSpace,
param_names: List[str]) -> SearchSpaceDigest:
"""Extract basic parameter prpoerties from a search space."""
bounds: List[Tuple[Union[int, float], Union[int, float]]] = []
ordinal_features: List[int] = []
categorical_features: List[int] = []
task_features: List[int] = []
fidelity_features: List[int] = []
target_fidelities: Dict[int, Union[int, float]] = {}
for i, p_name in enumerate(param_names):
p = search_space.parameters[p_name]
if isinstance(p, ChoiceParameter):
if p.parameter_type != ParameterType.INT:
raise ValueError(f"Choice parameter {p} not of integer type")
if p.is_task:
task_features.append(i)
elif p.is_ordered:
ordinal_features.append(i)
else:
categorical_features.append(i)
bounds.append(
(int(p.values[0]), int(p.values[-1]))) # pyre-ignore [6]
elif isinstance(p, RangeParameter):
if p.log_scale:
raise ValueError(f"{p} is log scale")
if p.parameter_type == ParameterType.INT:
task_features.append(i)
# TODO: modify to ordinal_features.append(i) and adjust models
# that rely on the current behavior
bounds.append((p.lower, p.upper))
else:
raise ValueError(f"Unknown parameter type {type(p)}")
if p.is_fidelity:
if not isinstance(not_none(p.target_value), (int, float)):
raise NotImplementedError(
"Only numerical target values are supported.")
target_fidelities[i] = checked_cast_to_tuple((int, float),
p.target_value)
fidelity_features.append(i)
return SearchSpaceDigest(
feature_names=param_names,
bounds=bounds,
ordinal_features=ordinal_features,
categorical_features=categorical_features,
task_features=task_features,
fidelity_features=fidelity_features,
target_fidelities=target_fidelities,
)
