Dakota  Version 7.0
Explore and Predict with Confidence
Public Member Functions | Private Attributes | List of all members
Study::ModelFactory Class Reference

Constructs models using an owning Study's runtime services. More...

Public Member Functions

 ModelFactory (const Study &study)
 Construct a factory borrowing study; study must outlive the factory and its products.
 
std::shared_ptr< SimulationModel > single (const IRStore &model_store, const Variables &variables, std::shared_ptr< Interface > interface, const Response &response) const
 Construct a simulation model from materialized configuration and injected components.
 
std::shared_ptr< SimulationModel > single (const nlohmann::json &model_json, const Variables &variables, std::shared_ptr< Interface > interface, const Response &response) const
 Validate a model-single fragment and construct a simulation model.
 
std::shared_ptr< NestedModel > nested (const IRStore &model_store, std::shared_ptr< Iterator > sub_iterator, std::shared_ptr< Interface > optional_interface, const Variables &variables, const Response &response) const
 Construct a nested model with a required sub-iterator and optional interface.
 
std::shared_ptr< NestedModel > nested (const IRStore &model_store, std::shared_ptr< Iterator > sub_iterator, std::shared_ptr< Interface > optional_interface, const Variables &variables, const Response &response, const Response &optional_interface_responses) const
 Construct a nested model whose optional interface returns a distinct response set (optional_interface_responses_pointer).
 
std::shared_ptr< NestedModel > nested (const nlohmann::json &model_json, std::shared_ptr< Iterator > sub_iterator, std::shared_ptr< Interface > optional_interface, const Variables &variables, const Response &response) const
 Validate a model-nested fragment and construct a nested model.
 
std::shared_ptr< NestedModel > nested (const nlohmann::json &model_json, std::shared_ptr< Iterator > sub_iterator, std::shared_ptr< Interface > optional_interface, const Variables &variables, const Response &response, const Response &optional_interface_responses) const
 Validate a model-nested fragment and construct a nested model whose optional interface returns a distinct response set.
 
std::shared_ptr< EnsembleSurrModel > ensemble_surrogate (const IRStore &model_store, std::shared_ptr< Model > truth_model, std::vector< std::shared_ptr< Model >> approximation_models, const Variables &variables, const Response &response) const
 Construct an ensemble from a truth model and approximation models.
 
std::shared_ptr< EnsembleSurrModel > ensemble_surrogate (const nlohmann::json &model_json, std::shared_ptr< Model > truth_model, std::vector< std::shared_ptr< Model >> approximation_models, const Variables &variables, const Response &response) const
 Validate an ensemble_surrogate fragment and construct the ensemble.
 
std::shared_ptr< DataFitSurrModel > global_surrogate (const IRStore &model_store, const Variables &variables, const Response &response, std::shared_ptr< Model > truth_model=nullptr, std::shared_ptr< Iterator > dace_iterator=nullptr) const
 Construct a global surrogate with optional truth-model and DACE dependencies.
 
std::shared_ptr< DataFitSurrModel > global_surrogate (const nlohmann::json &model_json, const Variables &variables, const Response &response, std::shared_ptr< Model > truth_model=nullptr, std::shared_ptr< Iterator > dace_iterator=nullptr) const
 Validate a global_surrogate fragment and construct the surrogate.
 
std::shared_ptr< DataFitSurrModel > local_surrogate (const IRStore &model_store, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Construct a local surrogate around a non-null truth model.
 
std::shared_ptr< DataFitSurrModel > local_surrogate (const nlohmann::json &model_json, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Validate a local_surrogate fragment and construct the surrogate.
 
std::shared_ptr< DataFitSurrModel > multipoint_surrogate (const IRStore &model_store, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Construct a multipoint surrogate around a non-null truth model.
 
std::shared_ptr< DataFitSurrModel > multipoint_surrogate (const nlohmann::json &model_json, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Validate a multipoint_surrogate fragment and construct the surrogate.
 
std::shared_ptr< ActiveSubspaceModel > active_subspace (const IRStore &model_store, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Construct an active subspace model over a non-null truth model.
 
std::shared_ptr< ActiveSubspaceModel > active_subspace (const nlohmann::json &model_json, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Validate an active_subspace fragment and construct the model.
 
std::shared_ptr< AdaptedBasisModel > adapted_basis (const IRStore &model_store, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Construct an adapted basis model over a non-null truth model.
 
std::shared_ptr< AdaptedBasisModel > adapted_basis (const nlohmann::json &model_json, std::shared_ptr< Model > truth_model, const Variables &variables, const Response &response) const
 Validate an adapted_basis fragment and construct the model.
 
std::shared_ptr< RandomFieldModel > random_field (const IRStore &model_store, std::shared_ptr< Model > propagation_model, const Variables &variables, const Response &response, std::shared_ptr< Iterator > dace_iterator=nullptr) const
 Construct a random field model over a non-null propagation model, with an optional DACE iterator that generates field realizations.
 
std::shared_ptr< RandomFieldModel > random_field (const nlohmann::json &model_json, std::shared_ptr< Model > propagation_model, const Variables &variables, const Response &response, std::shared_ptr< Iterator > dace_iterator=nullptr) const
 Validate a random_field fragment and construct the model.
 
std::shared_ptr< EnsembleSurrModel > ensemble_surrogate (const IRStore &model_store, std::vector< std::shared_ptr< Model >> ordered_models, const Variables &variables, const Response &response) const
 Construct an ensemble from models ordered low-to-high; the last is truth.
 
std::shared_ptr< EnsembleSurrModel > ensemble_surrogate (const nlohmann::json &model_json, std::vector< std::shared_ptr< Model >> ordered_models, const Variables &variables, const Response &response) const
 Validate an ensemble_surrogate fragment and construct an ordered ensemble.
 

Private Attributes

const Study & study
 

Detailed Description

Constructs models using an owning Study's runtime services.

JSON arguments contain the children of the selected model keyword, without a model array or selector wrapper. Child models, methods, interfaces, variables, and responses are injected as objects rather than resolved from pointer strings.


The documentation for this class was generated from the following files: