"""Contains the Emukit base class
The defined base class for the Emukit library.
"""
from collections.abc import Callable
import numpy as np
from emukit.core import ContinuousParameter, DiscreteParameter, ParameterSpace
from emukit.core.initial_designs import RandomDesign
from ..data_model import ParameterDataType
from .calibration_base import CalibrationWorkflowBase
[docs]
class EmukitBase(CalibrationWorkflowBase):
"""The Emukit base class."""
[docs]
def specify(self) -> None:
"""Specify the parameters of the model calibration procedure."""
parameters = []
self.names = []
self.data_types = []
self.bounds = []
parameter_spec = self.specification.parameter_spec.parameters
for spec in parameter_spec:
parameter_name = spec.name
data_type = spec.data_type
if (
data_type == ParameterDataType.DISCRETE
or data_type == ParameterDataType.CATEGORICAL
):
if data_type == ParameterDataType.DISCRETE:
bounds = self.get_parameter_bounds(spec)
else:
bounds = self.set_categorical_parameter(spec)
lower_bound, upper_bound = bounds
discrete_domain = np.arange(lower_bound, upper_bound + 1)
parameter = DiscreteParameter(parameter_name, discrete_domain)
elif data_type == ParameterDataType.CONSTANT:
parameter_value = spec.parameter_value
self.constants[parameter_name] = parameter_value
continue
else:
bounds = self.get_parameter_bounds(spec)
lower_bound, upper_bound = bounds
parameter = ContinuousParameter(
parameter_name, lower_bound, upper_bound
)
self.names.append(parameter_name)
self.data_types.append(data_type)
parameters.append(parameter)
self.bounds.append(bounds)
self.parameters = ParameterSpace(parameters)
[docs]
def get_X_Y(
self, n_init: int, target_function: Callable
) -> tuple[np.ndarray, np.ndarray]:
"""Get the X and Y matrices.
Args:
n_init (int): The number of samples to take
from the random design.
target_function (Callable):
The simulation function.
Returns:
tuple[np.ndarray, np.ndarray]: The X and Y matrices.
"""
design = RandomDesign(self.parameters)
X = self.specification.X
if X is None:
X = design.get_samples(n_init)
n_replicates = self.specification.n_replicates
if n_replicates > 1:
X = np.repeat(X, n_replicates, axis=0)
self.rng.shuffle(X)
Y = self.specification.Y
if Y is None:
Y = target_function(X)
return X, Y
[docs]
def sample_parameters(self, n_samples: int) -> np.ndarray:
"""Get new parameter samples.
Args:
n_samples (int): The number of samples.
Returns:
np.ndarray: The parameter samples.
"""
design = RandomDesign(self.parameters)
X = design.get_samples(n_samples)
return X