gemseo.core.mdo_functions.discipline_adapter_generator module

A class to create MDOFunction objects from an Discipline.

class DisciplineAdapterGenerator(discipline, names_to_sizes=mappingproxy({}))

Bases: object

A generator of discipline adapter.

Given a discipline, an DisciplineAdapter computes specific outputs from specific inputs.

Parameters:

• discipline (Discipline) -- The discipline from which to generate discipline adapters.

• names_to_sizes (MutableMapping[str, int]) --

  The sizes of the input variables. If empty, determine them from the default inputs and local data of the discipline.

  By default it is set to {}.

get_function(input_names, output_names, default_input_data=mappingproxy({}), is_differentiable=True, differentiated_input_names_substitute=())

Build a function executing a discipline for some inputs and outputs.

Parameters:

• input_names (Sequence[str]) -- The discipline input names defining the function input vector. If empty, use all the discipline inputs.

• output_names (Sequence[str]) -- The discipline output names defining the function output vector. If empty, use all the discipline outputs.

• default_input_data (Mapping[str, ndarray]) --

  The default values of the input variables. If empty, use the default input values of the discipline.

  By default it is set to {}.

• is_differentiable (bool) --

  Whether the function is differentiable.

  By default it is set to True.

• differentiated_input_names_substitute (Sequence[str]) --

  The names of the inputs with respect to which to differentiate the functions. If empty, use input_names. This argument is not used when is_differentiable is False.

  By default it is set to ().

Returns:

The function.

Raises:

ValueError -- When either an input name is not a discipline input name, a differentiated input name is not a discipline input name or an output name is not a discipline output name.

Return type:

DisciplineAdapter

discipline: Discipline

The discipline from which to generate discipline adapters.