RFFKernel¶

class baybe.kernels.basic.RFFKernel[source]¶

Bases: BasicKernel

A random Fourier features (RFF) kernel.

Public methods

`__init__`(num_samples[, lengthscale_prior, ...])	Method generated by attrs for class RFFKernel.
`from_dict`(dictionary)	Create an object from its dictionary representation.
`from_json`(string)	Create an object from its JSON representation.
`to_dict`()	Create an object's dictionary representation.
`to_factory`()	Wrap the kernel in a `baybe.surrogates.gaussian_process.kernel_factory.PlainKernelFactory`.
`to_gpytorch`(*[, ard_num_dims, batch_shape, ...])	Create the gpytorch representation of the kernel.
`to_json`()	Create an object's JSON representation.

Public attributes and properties

`num_samples`	The number of frequencies to draw.
`lengthscale_prior`	An optional prior on the kernel lengthscale.
`lengthscale_initial_value`	An optional initial value for the kernel lengthscale.

__init__(num_samples: int, lengthscale_prior: Prior | None = None, lengthscale_initial_value=None)¶

Method generated by attrs for class RFFKernel.

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

classmethod from_dict(dictionary: dict)¶

Create an object from its dictionary representation.

classmethod from_json(string: str)¶

Create an object from its JSON representation.

to_dict()¶

Create an object’s dictionary representation.

to_factory()¶

to_gpytorch(*, ard_num_dims: int | None = None, batch_shape: Size | None = None, active_dims: tuple[int, ...] | None = None)¶

Create the gpytorch representation of the kernel.

to_json()¶

Create an object’s JSON representation.

lengthscale_initial_value: float | None¶: An optional initial value for the kernel lengthscale.

lengthscale_prior: Prior | None¶: An optional prior on the kernel lengthscale.