GaussianProcessSurrogate

class baybe.surrogates.gaussian_process.core.GaussianProcessSurrogate

Bases: Surrogate

A Gaussian process surrogate model.

Public methods

__init__([kernel_or_factory])	Method generated by attrs for class GaussianProcessSurrogate.
fit(searchspace, train_x, train_y)	Train the surrogate model on the provided data.
from_dict(dictionary)	Create an object from its dictionary representation.
from_json(string)	Create an object from its JSON representation.
from_preset()	Create a Gaussian process surrogate from one of the defined presets.
posterior(candidates)	Evaluate the surrogate model at the given candidate points.
to_botorch()	Create the botorch-ready representation of the model.
to_dict()	Create an object's dictionary representation.
to_json()	Create an object's JSON representation.

Public attributes and properties

kernel_factory	The factory used to create the kernel of the Gaussian process.
joint_posterior	Class variable encoding whether or not a joint posterior is calculated.
supports_transfer_learning	Class variable encoding whether or not the surrogate supports transfer learning.

__init__(kernel_or_factory: Kernel | KernelFactory = NOTHING)

Method generated by attrs for class GaussianProcessSurrogate.

For details on the parameters, see Public attributes and properties.

fit(searchspace: SearchSpace, train_x: Tensor, train_y: Tensor)

Train the surrogate model on the provided data.

Parameters:

• searchspace (SearchSpace) – The search space in which experiments are conducted.

• train_x (Tensor) – The training data points.

• train_y (Tensor) – The training data labels.

Raises:

• ValueError – If the search space contains task parameters but the selected surrogate model type does not support transfer learning.

• NotImplementedError – When using a continuous search space and a non-GP model.

Return type:

None

classmethod from_dict(dictionary: dict)

Create an object from its dictionary representation.

Parameters:

dictionary (dict) – The dictionary representation.

Return type:

TypeVar(_T)

Returns:

The reconstructed object.

classmethod from_json(string: str)

Create an object from its JSON representation.

Parameters:

string (str) – The JSON representation of the object.

Return type:

TypeVar(_T)

Returns:

The reconstructed object.

classmethod from_preset()

Create a Gaussian process surrogate from one of the defined presets.

Return type:

GaussianProcessSurrogate

posterior(candidates: Tensor)

Evaluate the surrogate model at the given candidate points.

Parameters:

candidates (Tensor) – The candidate points, represented as a tensor of shape (*t, q, d), where t denotes the “t-batch” shape, q denotes the “q-batch” shape, and d is the input dimension. For more details about batch shapes, see: https://botorch.org/docs/batching

Return type:

tuple[Tensor, Tensor]

Returns:

The posterior means and posterior covariance matrices of the t-batched candidate points.

to_botorch()

Create the botorch-ready representation of the model.

Return type:

Model

to_dict()

Create an object’s dictionary representation.

Return type:

dict

to_json()

Create an object’s JSON representation.

Return type:

str

Returns:

The JSON representation as a string.

joint_posterior: ClassVar[bool] = True

Class variable encoding whether or not a joint posterior is calculated.

kernel_factory: KernelFactory

The factory used to create the kernel of the Gaussian process.

Accepts either a baybe.kernels.base.Kernel or a kernel_factory.KernelFactory. When passing a baybe.kernels.base.Kernel, it gets automatically wrapped into a kernel_factory.PlainKernelFactory.

supports_transfer_learning: ClassVar[bool] = True

Class variable encoding whether or not the surrogate supports transfer learning.