gemseo / core

surrogate_disc module

Surrogate discipline baseclass

class gemseo.core.surrogate_disc.SurrogateDiscipline(surrogate, data=None, transformer=None, disc_name=None, default_inputs=None, input_names=None, output_names=None, **parameters)[source]

Bases: gemseo.core.discipline.MDODiscipline

Surrogate discipline class

Constructor

Parameters

  • surrogate (str or MLRegressionAlgo) – name of the surrogate model algorithm.

  • data (Dataset) – dataset to train the surrogate. If None, assumes that the surrogate is trained. Default: None.

  • transformer (dict(str)) – transformation strategy for data groups. If None, do not transform data. Default: None.

  • disc_name (str) – name of the surrogate discipline. If None, use surrogate.ABBR + data.name . Default: None

  • default_inputs (dict) – default inputs. If None, use the first sample from the dataset. Default: None.

  • input_names (list(str)) – list of input names. If None, use all inputs. Default: None.

  • output_names (list(str)) – list of output names. If None, use all outputs. Default: None.

  • parameters – surrogate model parameters.