graphdot.model.gaussian_process.base module

class graphdot.model.gaussian_process.base.GaussianProcessRegressorBase(kernel, normalize_y, regularization, kernel_options)[source]

Bases: object

Base class for all Gaussian process regression (GPR) models.

X

The input values of the training set.

load(path, filename='model.pkl')[source]

Load a stored GaussianProcessRegressor model from a pickle file.

Parameters:
  • path (str) – The directory where the model is saved.
  • filename (str) – The file name for the saved model.
static mask(iterable)[source]
save(path, filename='model.pkl', overwrite=False)[source]

Save the trained GaussianProcessRegressor with the associated data as a pickle.

Parameters:
  • path (str) – The directory to store the saved model.
  • filename (str) – The file name for the saved model.
  • overwrite (bool) – If True, a pre-existing file will be overwritten. Otherwise, a runtime error will be raised.
y

The output/target values of the training set.