Calibrator

class Melodie.calibrator.Calibrator(config: Config, scenario_cls: Type[Scenario] | None, model_cls: Type[Model] | None, data_loader_cls: Type[DataLoader], processors=1)

Bases: BaseModellingManager

Calibrator calibrates the parameters of the scenario by minimizing the distance between model output and empirical evidence.

Parameters:

  • config – Config instance for current project.

  • scenario_cls – Scenario class for current project.

  • model_cls – Model class in current project.

  • data_loader_cls – DataLoader class in current project.

  • processors – Each path in current iteration will be computed parallelly, this parameter stands for processor cores used in parallel computation.

setup()

Setup method, be sure to inherit this method in custom calibrator class.

collect_data()

Set the agent and environment properties to be collected.

generate_scenarios() → List[Scenario]

Generate scenario objects by the parameter from static tables or scenarios_dataframe.

Returns:

A list of generated scenarios.

get_params_scenarios() → List

Get the parameters of calibrator parameters from the registered dataframe.

Returns:

A list of dict, and each dict contains parameters.

run()

The main method for calibrator.

run_once_new(scenario: Scenario, calibration_params: GACalibratorParams)

Run for one calibration path

Parameters:

  • scenario – The scenario to run.

  • calibration_params – calibration parameters.

Returns:

None

target_function(model: Model) → float | int

The target function to be minimized

Parameters:

env – Environment of the current model.

Returns:

distance(model: Model) → float

The optimization of calibrator is to minimize the distance.

Be sure to inherit this function in custom calibrator, and return a float value.

Parameters:

model – The current model after running the current parameter set.

Returns:

None

add_scenario_calibrating_property(prop: str)

Add a property to be tuned in the calibration, and prop should be a name of property in the environment.

Parameters:

prop – Property name

Returns:

None

add_environment_property(prop: str)

Add a property of environment to be recorded in the calibration voyage.

Parameters:

prop – Property name

Returns:

None