PIConfiguration — PIConfiguration • ospsuite.parameteridentification

Encapsulates configurations such as optimization algorithm choice, and evaluation settings for parameter identification.

Active bindings

printEvaluationFeedback: Boolean. If TRUE, prints objective function value after each evaluation. Default is FALSE.
simulationRunOptions: Object of SimulationRunOptions for simulation runs. If NULL, default options are used.
objectiveFunctionOptions: Settings for model fit evaluation, affecting error metrics and cost calculation. Defaults in ObjectiveFunctionOptions. Partial lists are merged with the current settings; only the provided keys are updated and validated.
algorithm: Optimization algorithm name. See Algorithms for a list of supported algorithms. Defaults to BOBYQA. Changing the algorithm resets algorithmOptions to the new algorithm's defaults.
ciMethod: Confidence interval estimation method. See CIMethods for available options. Defaults to hessian. Changing the method resets ciOptions to the new method's defaults.
algorithmOptions: Named list of user-defined overrides for the selected algorithm's settings. Returns NULL if no overrides are set; in that case algorithm defaults (AlgorithmOptions_XYZ) are used automatically. Partial lists are merged with previously stored overrides. Unknown keys produce a warning and are ignored. Set to NULL to clear all overrides.
ciOptions: Named list of settings for the selected CI method. Refer to CIOptions for default settings per method (e.g., CIOptions_XYZ where XYZ corresponds to the method name). If NULL, CI method's default settings are returned. Partial lists are merged with the current settings; only the provided keys are validated. Unknown keys produce a warning and are ignored. Set to NULL to reset to defaults.
autoEstimateCI: Logical. If TRUE, confidence intervals are automatically estimated after optimization. If FALSE, the step is skipped and can be triggered manually by calling the estimateCI() method on the ParameterIdentification object.
modelCostField: Read-only field name in the cost object used as the optimization target. Currently, only modelCost is supported.

Methods

`PIConfiguration$new()`

Initialize a new instance of the class.

Usage

PIConfiguration$new()

Returns

A new PIConfiguration object.

`PIConfiguration$print()`

Prints a summary of the PIConfiguration.

Usage

PIConfiguration$print()

`PIConfiguration$clone()`

The objects of this class are cloneable with this method.

Usage

PIConfiguration$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.