olorenchemengine.external.GaussianProcess package#


olorenchemengine.external.GaussianProcess.main module#

class olorenchemengine.external.GaussianProcess.main.GaussianProcessClassifier(*args, **kwargs)#

Bases: BaseEstimator

Wrapper for sklearn GaussianProcessClassifier

class olorenchemengine.external.GaussianProcess.main.GaussianProcessModel(representation: BaseVecRepresentation, kernel: str = None, kernel_params: dict = {}, alpha: float = 1e-10, random_state: int = None, **kwargs)#

Bases: BaseSKLearnModel

Gaussian process model using SciPy’s implementation of Gaussian process models.

  • representation (BaseVecRepresentation) – Representation to use.

  • kernel (str) – Kernel to use. Default is None which uses SciPy’s default.

  • kernel_params (dict) – Parameters for the kernel. Default is {}, where no kernel parameters are passed and SciPy defaults are used.

  • alpha (float) – Value of alpha to use. Default is 1e-10.

  • random_state (int) – Random state to use. Default is None.

class olorenchemengine.external.GaussianProcess.main.GaussianProcessRegressor(*args, **kwargs)#

Bases: BaseEstimator

Wrapper for sklearn GaussianProcessRegressor

class olorenchemengine.external.GaussianProcess.main.Tanimoto(*args: Any, **kwargs: Any)#

Bases: Kernel

Tanimoto kernel.

Adapted from DotProduct kernel in sklearn.gaussian_process.kernels and from Ryan-Rhys Griffiths’ implementation of the Tanimoto kernel in https://towardsdatascience.com/gaussian-process-regression-on-molecules-in-gpflow-ee6fedab2130.

bew(A, B)#

Returns the diagonal of the kernel k(X, X). The result of this method is identical to np.diag(self(X)); however, it can be evaluated more efficiently since only the diagonal is evaluated. :param X: Left argument of the returned kernel k(X, Y). :type X: ndarray of shape (n_samples_X, n_features)


K_diag – Diagonal of kernel k(X, X).

Return type:

ndarray of shape (n_samples_X,)

property hyperparameter_sigma_0#

Returns whether the kernel is stationary.

Module contents#