phlower.settings.GroupModuleSetting

class phlower.settings.GroupModuleSetting(*, name, inputs=<factory>, outputs=<factory>, modules=<factory>, destinations=<factory>, nn_type='Group', no_grad=False, solver_type='none', is_steady_problem=False, solver_parameters=<factory>, time_series_length=None)

Bases: BaseModel, IModuleSetting, IReadOnlyReferenceGroupSetting

Methods

find_module(name)
get_destinations()
get_input_keys()
get_name()
get_output_info()
get_output_keys()
resolve(*resolved_outputs[, is_first])
search_module_setting(name)

Attributes

model_computed_fields
model_config	Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].
model_extra	Get extra fields set during validation.
model_fields
model_fields_set	Returns the set of fields that have been explicitly set on this model instance.
name	name of group
inputs	definition of input variables
outputs	definition of output variables
modules	modules which belongs to this group
destinations	name of destination modules.
nn_type	name of neural network type.
no_grad	A Flag not to calculate gradient.
solver_type	Solver name calculating in iteration loop
is_steady_problem	When true, updating function in the group iteraion is defined as steady problems
solver_parameters	Parameters to pass iteration solver.
time_series_length	If feed integer value, do iteration to output time series tensor.

Parameters:

• name (str)

• inputs (list[GroupIOSetting])

• outputs (list[GroupIOSetting])

• modules (list[Annotated[Annotated[GroupModuleSetting, Tag(tag=GROUP)] | Annotated[ModuleSetting, Tag(tag=MODULE)], Discriminator(discriminator=~phlower.settings._group_setting._custom_discriminator, custom_error_type=invalid_union_member, custom_error_message=Invalid union member, custom_error_context={'discriminator': 'group_or_module'})]])

• destinations (list[str])

• nn_type (Literal['Group', 'GROUP'])

• no_grad (bool)

• solver_type (str)

• is_steady_problem (bool)

• solver_parameters (Annotated[IPhlowerIterationSolverSetting, PlainValidator(func=~phlower.settings._nonlinear_solver_setting._validate, json_schema_input_type=~typing.Any), PlainSerializer(func=~phlower.settings._nonlinear_solver_setting._serialize, return_type=PydanticUndefined, when_used=always)])

• time_series_length (int | None)

destinations: list[str]

name of destination modules.

inputs: list[GroupIOSetting]

definition of input variables

is_steady_problem: bool

When true, updating function in the group iteraion is defined as steady problems

model_config: ClassVar[ConfigDict] = {'extra': 'forbid'}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

modules: list[Annotated[Annotated[GroupModuleSetting, Tag(_DiscriminatorTag.GROUP.name)] | Annotated[ModuleSetting, Tag(_DiscriminatorTag.MODULE.name)], Discriminator(_custom_discriminator, custom_error_type='invalid_union_member', custom_error_message='Invalid union member', custom_error_context={'discriminator': 'group_or_module'})]]

modules which belongs to this group

name: str

name of group

nn_type: Literal['Group', 'GROUP']

name of neural network type. Fixed to “Group” or “GROUP”

no_grad: bool

A Flag not to calculate gradient. Defauls to False.

outputs: list[GroupIOSetting]

definition of output variables

solver_parameters: SolverParameters

Parameters to pass iteration solver. Contents depends on solver_type

solver_type: str

Solver name calculating in iteration loop

none: No iteration. simple: run iteration until calculated value is converged. bb: Barzilan-Borwein method is used to converge iteration results.

time_series_length: int | None

If feed integer value, do iteration to output time series tensor.