boax.core.models.gaussian_process.exact

Contents

boax.core.models.gaussian_process.exact#

boax.core.models.gaussian_process.exact(mean_fn, kernel_fn, likelihood_fn, observation_index_points=None, observations=None, jitter=1e-06)#

The exact gaussian process model.

Example

>>> model = gaussian_process(mean_fn, kernel_fn)
>>> mean, cov = model(xs)
Parameters:

  • mean_fn (Mean) – The process’ mean function.

  • kernel_fn (Kernel) – The process’ covariance function.

  • observation_index_points (Union[Array, ndarray, bool, number, None]) – The index points of the given observations.

  • observations (Union[Array, ndarray, bool, number, None]) – The observed values.

  • jitter (Union[Array, ndarray, bool, number, float, int]) – The scalar added to the diagonal of the covariance matrix to ensure positive definiteness.

Return type:

Model[TypeVar(T)]

Returns:

The gaussian process Model.

Contents