NestedModel Class Reference

Derived model class which performs a complete sub-iterator execution within every evaluation of the model. More...

Inheritance diagram for NestedModel:

Public Member Functions
	NestedModel (ProblemDescDB &problem_db)
	constructor
void	declare_sources () override
	Declare a model's sources to the evaluationsDB.
Public Member Functions inherited from Model
	Model ()
	default constructor More...
	Model (ProblemDescDB &problem_db)
	standard constructor More...
	Model (const Model &model)=delete
	copy constructor
Model &	operator= (const Model &model)=delete
	assignment operator
virtual void	resize_from_subordinate_model (size_t depth=SZ_MAX)
	resize vars/resp if needed from the bottom up
virtual void	update_from_subordinate_model (size_t depth=SZ_MAX)
	propagate vars/labels/bounds/targets from the bottom up More...
virtual void	primary_response_fn_weights (const RealVector &wts, bool recurse_flag=true)
	set the relative weightings for multiple objective functions or least squares terms
virtual bool	initialize_mapping (ParLevLIter pl_iter)
	Perform any global updates prior to individual evaluate() calls; returns true if the variables size has changed.
virtual bool	finalize_mapping ()
	restore state in preparation for next initialization; returns true if the variables size has changed
virtual void	stop_init_mapping (ParLevLIter pl_iter)
	called from IteratorScheduler::run_iterator() for iteratorComm rank 0 to terminate serve_init_mapping() on other iteratorComm processors
virtual int	serve_init_mapping (ParLevLIter pl_iter)
	called from IteratorScheduler::run_iterator() for iteratorComm rank != 0 to balance resize() calls on iteratorComm rank 0
virtual void	stop_finalize_mapping (ParLevLIter pl_iter)
	called from IteratorScheduler::run_iterator() for iteratorComm rank 0 to terminate serve_finalize_mapping() on other iteratorComm processors
virtual int	serve_finalize_mapping (ParLevLIter pl_iter)
	called from IteratorScheduler::run_iterator() for iteratorComm rank != 0 to balance resize() calls on iteratorComm rank 0
virtual void	cache_unmatched_response (int raw_id)
	migrate an unmatched response record from active response map (computed by synchronize() or synhronize_nowait()) to cached response map
virtual void	cache_unmatched_responses ()
	migrate remaining response records from responseMap to cachedResponseMap
virtual void	create_2d_plots ()
	create 2D graphics plots for automatic logging of vars/response data
virtual void	create_tabular_datastream ()
	create a tabular output stream for automatic logging of vars/response data
virtual void	derived_auto_graphics (const Variables &vars, const Response &resp)
	Update tabular/graphics data with latest variables/response data.
virtual void	active_view (short view, bool recurse_flag=true)
	update the Model's active view based on a higher level context
virtual void	inactive_view (short view, bool recurse_flag=true)
	update the Model's inactive view based on a higher level context
virtual bool	evaluation_cache (bool recurse_flag=true) const
	Indicates the usage of an evaluation cache by the Model. More...
virtual bool	restart_file (bool recurse_flag=true) const
	Indicates the usage of a restart file by the Model. More...
virtual void	eval_tag_prefix (const String &eval_id_str)
	set the hierarchical eval ID tag prefix More...
virtual bool	db_lookup (const Variables &search_vars, const ActiveSet &search_set, Response &found_resp)
	search the eval database (during derivative estimation); derived may need to reimplement due to problem transformations (RecastModel); return true if found in DB
virtual void	nested_variable_mappings (const SizetArray &c_index1, const SizetArray &di_index1, const SizetArray &ds_index1, const SizetArray &dr_index1, const ShortArray &c_target2, const ShortArray &di_target2, const ShortArray &ds_target2, const ShortArray &dr_target2)
	set primaryA{C,DI,DS,DR}VarMapIndices, secondaryA{C,DI,DS,DR}VarMapTargets (coming from a higher-level NestedModel context to inform derivative est.)
virtual const SizetArray &	nested_acv1_indices () const
	return primaryACVarMapIndices
virtual const ShortArray &	nested_acv2_targets () const
	return secondaryACVarMapTargets
virtual short	query_distribution_parameter_derivatives () const
	calculate and return derivative composition of final results w.r.t. distribution parameters (none, all, or mixed)
virtual void	activate_distribution_parameter_derivatives ()
	activate derivative setting w.r.t. distribution parameters
virtual void	deactivate_distribution_parameter_derivatives ()
	deactivate derivative setting w.r.t. distribution parameters
virtual Pecos::ProbabilityTransformation &	probability_transformation ()
	return probability transformation employed by the Model (forwarded along to ProbabilityTransformModel recasting)
virtual void	trans_U_to_X (const RealVector &u_c_vars, RealVector &x_c_vars)
	transform u-space variable values to x-space
virtual void	trans_X_to_U (const RealVector &x_c_vars, RealVector &u_c_vars)
	transform x-space variable values to u-space
virtual void	trans_grad_X_to_U (const RealVector &fn_grad_x, RealVector &fn_grad_u, const RealVector &x_vars)
	transform x-space gradient vector to u-space
virtual void	trans_grad_U_to_X (const RealVector &fn_grad_u, RealVector &fn_grad_x, const RealVector &x_vars)
	transform u-space gradient vector to x-space
virtual void	trans_grad_X_to_S (const RealVector &fn_grad_x, RealVector &fn_grad_s, const RealVector &x_vars)
	transform x-space gradient vector to gradient with respect to inserted distribution parameters
virtual void	trans_hess_X_to_U (const RealSymMatrix &fn_hess_x, RealSymMatrix &fn_hess_u, const RealVector &x_vars, const RealVector &fn_grad_x)
	transform x-space Hessian matrix to u-space
virtual size_t	qoi () const
	return number of unique response functions (managing any aggregations)
virtual std::shared_ptr< Model >	surrogate_model (size_t i=_NPOS)
	return the i-th approximation sub-model in surrogate models More...
virtual std::shared_ptr< const Model >	surrogate_model (size_t i=_NPOS) const
	return the i-th approximation sub-model in surrogate models
virtual std::shared_ptr< Model >	truth_model ()
	return the truth sub-model in surrogate models More...
virtual std::shared_ptr< const Model >	truth_model () const
	return the truth sub-model in surrogate models
virtual unsigned short	active_surrogate_model_form (size_t i) const
	return the model form of the i-th active surrogate model
virtual unsigned short	active_truth_model_form () const
	return the model form of the active truth model
virtual std::shared_ptr< Model >	active_surrogate_model (size_t i=_NPOS)
	return the i-th active approximation sub-model in surrogate models More...
virtual std::shared_ptr< const Model >	active_surrogate_model (size_t i=_NPOS) const
	return the i-th active approximation sub-model in surrogate models
virtual std::shared_ptr< Model >	active_truth_model ()
	return the active truth sub-model in surrogate models More...
virtual std::shared_ptr< const Model >	active_truth_model () const
	return the active truth sub-model in surrogate models
virtual void	active_model_key (const Pecos::ActiveKey &key)
	set the active model key within surrogate data, grid driver, and approximation classes that support the management of multiple approximation states within surrogate models
virtual const Pecos::ActiveKey &	active_model_key () const
	return the active model key (used by surrogate data, grid driver, and approximation classes to support the management of multiple approximation states within surrogate models)
virtual void	clear_model_keys ()
	reset by removing all model keys within surrogate data, grid driver, and approximation classes that support the management of multiple approximation states within surrogate models
virtual bool	active_truth_key () const
	return true if there is an active truth model indicated by truthModelKey
virtual size_t	active_surrogate_keys () const
	return the number of active surrogate models indicated by surrModelKeys
virtual void	build_approximation ()
	build a new SurrogateModel approximation
virtual bool	force_rebuild ()
	determine whether a surrogate model rebuild should be forced based on changes in the inactive data
virtual void	surrogate_function_indices (const SizetSet &surr_fn_indices)
	set the (currently active) surrogate function index set
virtual short	surrogate_response_mode () const
	return response computation mode used in SurrogateModels for forming currentResponse
virtual DiscrepancyCorrection &	discrepancy_correction ()
	return the DiscrepancyCorrection object used by SurrogateModels
virtual void	correction_type (short corr_type)
	set the correction type from the DiscrepancyCorrection object used by SurrogateModels
virtual short	correction_type () const
	return the correction type from the DiscrepancyCorrection object used by SurrogateModels
virtual short	correction_order () const
	return the correction order from the DiscrepancyCorrection object used by SurrogateModels
virtual unsigned short	correction_mode () const
	return correctionMode
virtual void	correction_mode (unsigned short corr_mode)
	set correctionMode
virtual bool	build_approximation (const Variables &vars, const IntResponsePair &response_pr)
	build a new SurrogateModel approximation using/enforcing anchor response at vars; rebuild if needed
virtual void	rebuild_approximation ()
	incremental rebuild of an existing SurrogateModel approximation
virtual void	rebuild_approximation (const IntResponsePair &response_pr)
	incremental rebuild of an existing SurrogateModel approximation
virtual void	rebuild_approximation (const IntResponseMap &resp_map)
	incremental rebuild of an existing SurrogateModel approximation
virtual void	update_approximation (bool rebuild_flag)
	replace the approximation data within an existing surrogate based on data updates propagated elsewhere
virtual void	update_approximation (const Variables &vars, const IntResponsePair &response_pr, bool rebuild_flag)
	replace the anchor point data within an existing surrogate
virtual void	update_approximation (const VariablesArray &vars_array, const IntResponseMap &resp_map, bool rebuild_flag)
	replace the data points within an existing surrogate
virtual void	update_approximation (const RealMatrix &samples, const IntResponseMap &resp_map, bool rebuild_flag)
	replace the data points within an existing surrogate
virtual void	append_approximation (bool rebuild_flag)
	append to the existing approximation data within a surrogate based on data updates propagated elsewhere
virtual void	append_approximation (const Variables &vars, const IntResponsePair &response_pr, bool rebuild_flag)
	append a single point to an existing surrogate's data
virtual void	append_approximation (const RealMatrix &samples, const IntResponseMap &resp_map, bool rebuild_flag)
	append multiple points to an existing surrogate's data
virtual void	append_approximation (const VariablesArray &vars_array, const IntResponseMap &resp_map, bool rebuild_flag)
	append multiple points to an existing surrogate's data
virtual void	append_approximation (const IntVariablesMap &vars_map, const IntResponseMap &resp_map, bool rebuild_flag)
	append multiple points to an existing surrogate's data
virtual void	replace_approximation (const IntResponsePair &response_pr, bool rebuild_flag)
	replace the response for a single point (based on eval id from response_pr) within an existing surrogate's data
virtual void	replace_approximation (const IntResponseMap &resp_map, bool rebuild_flag)
	replace the responses for a set of points (based on eval ids from resp_map) within an existing surrogate's data
virtual void	track_evaluation_ids (bool track)
	assigns a flag to track evaluation ids within surrogate data, enabling id-based lookups for data replacement
virtual void	pop_approximation (bool save_surr_data, bool rebuild_flag=false)
	remove the previous data set addition to a surrogate (e.g., due to a previous append_approximation() call); flag manages storing of surrogate data for use in a subsequent push_approximation()
virtual void	push_approximation ()
	push a previous approximation data state; reverse of pop_approximation
virtual bool	push_available ()
	query for whether a trial increment is restorable within a surrogate
virtual void	finalize_approximation ()
	finalize an approximation by applying all previous trial increments
virtual void	combine_approximation ()
	combine the current approximation with previously stored data sets
virtual void	combined_to_active (bool clear_combined=true)
	promote the combined approximation into the active approximation
virtual void	clear_inactive ()
	clear inactive approximations (finalization + combination completed)
virtual bool	advancement_available ()
	query the approximation for available advancement in resolution controls (order, rank, etc.); an input to adaptive refinement strategies
virtual bool	formulation_updated () const
	query the approximation for updates in formulation, requiring a rebuild even if no updates to the build data
virtual void	formulation_updated (bool update)
	assign the status of approximation formulation updates
virtual void	run_dace ()
	execute the DACE iterator (prior to building/appending the approximation)
virtual SharedApproxData &	shared_approximation ()
	retrieve the shared approximation data within the ApproximationInterface of a DataFitSurrModel
virtual std::vector< Approximation > &	approximations ()
	retrieve the set of Approximations within the ApproximationInterface of a DataFitSurrModel
virtual const Pecos::SurrogateData &	approximation_data (size_t fn_index)
	retrieve a SurrogateData instance from a particular Approximation instance within the ApproximationInterface of a DataFitSurrModel
virtual const RealVectorArray &	approximation_coefficients (bool normalized=false)
	retrieve the approximation coefficients from each Approximation within a DataFitSurrModel
virtual void	approximation_coefficients (const RealVectorArray &approx_coeffs, bool normalized=false)
	set the approximation coefficients for each Approximation within a DataFitSurrModel
virtual const RealVector &	approximation_variances (const Variables &vars)
	retrieve the prediction variances from each Approximation within a DataFitSurrModel
virtual void	discrepancy_emulation_mode (short mode)
	set discrepancy emulation mode used in SurrogateModels for approximating response differences
virtual bool	multifidelity () const
	identify if 1D hierarchy can be defined across model forms
virtual bool	multilevel () const
	identify if 1D hierarchy can be defined across resolution levels
virtual bool	multilevel_multifidelity () const
	identify if 2D hierarchy can be defined across both model forms and resolution levels
virtual short	ensemble_precedence () const
	return precedence for ensemble definition: model forms, resolution levels, or both
virtual void	ensemble_precedence (short mlmf_prec, bool update_default=false)
	assign precedence for ensemble definition (model forms or resolution levels or both) as determined from algorithm context
virtual void	single_apply (const Variables &vars, Response &resp, const Pecos::ActiveKey &paired_key)
	apply a DiscrepancyCorrection to correct an approximation within an EnsembleSurrModel
virtual void	recursive_apply (const Variables &vars, Response &resp)
	apply a sequence of DiscrepancyCorrections to recursively correct an approximation within an EnsembleSurrModel
virtual size_t	solution_levels () const
	number of discrete levels within solution control (SimulationModel) More...
virtual void	solution_level_cost_index (size_t index)
	activate a particular level within the solution level control (SimulationModel) More...
virtual size_t	solution_level_cost_index () const
	return currently active level within the solution level control (SimulationModel)
virtual RealVector	solution_level_costs () const
	return ordered cost estimates across solution levels (SimulationModel)
virtual Real	solution_level_cost () const
	return currently active cost estimate from solution level control (SimulationModel)
virtual short	solution_control_variable_type () const
	return type of solution control variable
virtual size_t	solution_control_variable_index () const
	return index of solution control variable within all variables
virtual size_t	solution_control_discrete_variable_index () const
	return index of solution control variable within all discrete variables
virtual int	solution_level_int_value () const
	return the active (integer) value of the solution control
virtual String	solution_level_string_value () const
	return the active (string) value of the solution control
virtual Real	solution_level_real_value () const
	return the active (real) value of the solution control
virtual size_t	cost_metadata_index () const
	return index of online cost estimates within metadata
virtual bool	resize_pending () const
	return true if a potential resize is still pending, such that sizing-based initialization should be deferred
ModelList &	subordinate_models (bool recurse_flag=true)
	return the sub-models in nested and surrogate models More...
void	evaluate ()
	Compute the Response at currentVariables (default ActiveSet).
void	evaluate (const ActiveSet &set)
	Compute the Response at currentVariables (specified ActiveSet).
void	evaluate_nowait ()
	Spawn an asynchronous job (or jobs) that computes the value of the Response at currentVariables (default ActiveSet).
void	evaluate_nowait (const ActiveSet &set)
	Spawn an asynchronous job (or jobs) that computes the value of the Response at currentVariables (specified ActiveSet).
const IntResponseMap &	synchronize ()
	Execute a blocking scheduling algorithm to collect the complete set of results from a group of asynchronous evaluations.
const IntResponseMap &	synchronize_nowait ()
	Execute a nonblocking scheduling algorithm to collect all available results from a group of asynchronous evaluations.
int	evaluation_id () const
	return Model's (top-level) evaluation counter, not to be confused with derived counter returned by derived_evaluation_id()
bool	mapping_initialized () const
void	init_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true)
	allocate communicator partitions for a model and store configuration in modelPCIterMap More...
void	init_serial ()
	for cases where init_communicators() will not be called, modify some default settings to behave properly in serial. More...
void	set_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true)
	set active parallel configuration for the model (set modelPCIter from modelPCIterMap)
void	free_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true)
	deallocate communicator partitions for a model
MPI_Comm	analysis_comm () const
	retrieve the MPI communicator on which this model is configured to conduct function evaluation analyses (provided for library clients)
void	stop_init_communicators (ParLevLIter pl_iter)
	called from IteratorScheduler::init_iterator() for iteratorComm rank 0 to terminate serve_init_communicators() on other iteratorComm processors
int	serve_init_communicators (ParLevLIter pl_iter)
	called from IteratorScheduler::init_iterator() for iteratorComm rank != 0 to balance init_communicators() calls on iteratorComm rank 0
void	estimate_message_lengths ()
	estimate messageLengths for a model More...
bool	manage_data_recastings ()
	initialize modelList and recastFlags for data import/export More...
bool	recastings () const
	return true if recastFlags is defined
void	user_space_to_iterator_space (Variables &vars)
	employ the model recursion to transform from bottom level user-space variables to top level iterator-space variables
void	user_space_to_iterator_space (const Variables &user_vars, const Response &user_resp, Variables &iter_vars, Response &iter_resp)
	employ the model recursion to transform from bottom level user-space data to top level iterator-space data
void	iterator_space_to_user_space (Variables &vars)
	employ the model recursion to transform from top level iterator-space variables to bottom level user-space variables
void	iterator_space_to_user_space (const Variables &iter_vars, const Response &iter_resp, Variables &user_vars, Response &user_resp)
	employ the model recursion to transform from top level iterator-space data to bottom level user-space data
Pecos::MultivariateDistribution &	multivariate_distribution ()
	return mvDist
const Pecos::MultivariateDistribution &	multivariate_distribution () const
	return mvDist
const Variables &	current_variables () const
	return the current variables (currentVariables) as const reference (preferred)
Variables &	current_variables ()
	return the current variables (currentVariables) in mutable form (special cases)
const Constraints &	user_defined_constraints () const
	return the user-defined constraints (userDefinedConstraints) as const reference (preferred)
Constraints &	user_defined_constraints ()
	return the user-defined constraints (userDefinedConstraints) in mutable form (special cases)
const Response &	current_response () const
	return the current response (currentResponse)
Response &	current_response ()
	return the current response (currentResponse)
ProblemDescDB &	problem_description_db () const
	return the problem description database (probDescDB)
ParallelLibrary &	parallel_library () const
	return the parallel library (parallelLib)
const String &	model_type () const
	return the model type (modelType)
const String &	surrogate_type () const
	return the surrogate type (surrogateType)
const String &	model_id () const
	return the model identifier (modelId)
size_t	num_primary_fns () const
	return number of primary functions (total less nonlinear constraints)
size_t	num_secondary_fns () const
	return number of secondary functions (number of nonlinear constraints)
const String &	gradient_type () const
	return the gradient evaluation type (gradientType)
const String &	method_source () const
	return the numerical gradient evaluation method source (methodSource)
const String &	interval_type () const
	return the numerical gradient evaluation interval type (intervalType)
bool	ignore_bounds () const
	option for ignoring bounds when numerically estimating derivatives
bool	central_hess () const
	option for using old 2nd-order scheme when computing finite-diff Hessian
const RealVector &	fd_gradient_step_size () const
	return the finite difference gradient step size (fdGradStepSize)
const String &	fd_gradient_step_type () const
	return the finite difference gradient step type (fdGradStepType)
const IntSet &	gradient_id_analytic () const
	return the mixed gradient analytic IDs (gradIdAnalytic)
const IntSet &	gradient_id_numerical () const
	return the mixed gradient numerical IDs (gradIdNumerical)
const String &	hessian_type () const
	return the Hessian evaluation type (hessianType)
const String &	quasi_hessian_type () const
	return the Hessian evaluation type (quasiHessType)
const RealVector &	fd_hessian_by_grad_step_size () const
	return gradient-based finite difference Hessian step size (fdHessByGradStepSize)
const RealVector &	fd_hessian_by_fn_step_size () const
	return function-based finite difference Hessian step size (fdHessByFnStepSize)
const String &	fd_hessian_step_type () const
	return the finite difference Hessian step type (fdHessStepType)
const IntSet &	hessian_id_analytic () const
	return the mixed Hessian analytic IDs (hessIdAnalytic)
const IntSet &	hessian_id_numerical () const
	return the mixed Hessian analytic IDs (hessIdNumerical)
const IntSet &	hessian_id_quasi () const
	return the mixed Hessian analytic IDs (hessIdQuasi)
void	primary_response_fn_sense (const BoolDeque &sense)
	set the optimization sense for multiple objective functions
const BoolDeque &	primary_response_fn_sense () const
	get the optimization sense for multiple objective functions
const RealVector &	primary_response_fn_weights () const
	get the relative weightings for multiple objective functions or least squares terms
const ScalingOptions &	scaling_options () const
	user-provided scaling options
short	primary_fn_type () const
	get the primary response function type (generic, objective, calibration)
void	primary_fn_type (short type)
	set the primary response function type, e.g., when recasting
bool	derivative_estimation ()
	indicates potential usage of estimate_derivatives() based on gradientType/hessianType
void	supports_derivative_estimation (bool sed_flag)
	set whether this model should perform or pass on derivative estimation
void	init_comms_bcast_flag (bool icb_flag)
	set initCommsBcastFlag
int	evaluation_capacity () const
	return the evaluation capacity for use in iterator logic
int	derivative_concurrency () const
	return the gradient concurrency for use in parallel configuration logic More...
bool	asynch_flag () const
	return the asynchronous evaluation flag (asynchEvalFlag)
void	asynch_flag (const bool flag)
	set the asynchronous evaluation flag (asynchEvalFlag)
short	output_level () const
	return the outputLevel
void	output_level (const short level)
	set the outputLevel
const IntArray &	message_lengths () const
	return the array of MPI packed message buffer lengths (messageLengths)
void	parallel_configuration_iterator (ParConfigLIter pc_iter)
	set modelPCIter
ParConfigLIter	parallel_configuration_iterator () const
	return modelPCIter
void	auto_graphics (const bool flag)
	set modelAutoGraphicsFlag to activate posting of graphics data within evaluate/synchronize functions (automatic graphics posting in the model as opposed to graphics posting at the strategy level).
bool	auto_graphics () const
	get modelAutoGraphicsFlag to activate posting of graphics data within evaluate/synchronize functions (automatic graphics posting in the model as opposed to graphics posting at the strategy level).
virtual String	root_model_id ()
	Return the model ID of the "innermost" model. For all derived Models except RecastModels, return modelId. The RecastModel override returns the root_model_id() of the subModel.
virtual ActiveSet	default_active_set ()

Protected Member Functions
void	derived_evaluate (const ActiveSet &set) override
	portion of evaluate() specific to NestedModel More...
void	derived_evaluate_nowait (const ActiveSet &set) override
	portion of evaluate_nowait() specific to NestedModel More...
const IntResponseMap &	derived_synchronize () override
	portion of synchronize() specific to NestedModel More...
std::shared_ptr< Iterator >	subordinate_iterator () override
	return subIterator
std::shared_ptr< Model >	subordinate_model () override
	return subModel
void	derived_subordinate_models (ModelList &ml, bool recurse_flag) override
	return subModel
std::shared_ptr< Interface >	derived_interface () override
	return optionalInterface
void	derived_interface (std::shared_ptr< Interface >) override
const RealVector &	error_estimates () override
	retrieve error estimates corresponding to the subIterator's response results (e.g., statistical MSE for subordinate UQ).
void	surrogate_response_mode (short mode) override
	pass a bypass request on to the subModel for any lower-level surrogates
void	component_parallel_mode (short mode) override
	update component parallel mode for supporting parallelism in optionalInterface and subModel
size_t	mi_parallel_level_index () const override
	return subIteratorSched.miPLIndex
short	local_eval_synchronization () override
	return optionalInterface synchronization setting More...
int	local_eval_concurrency () override
	return optionalInterface asynchronous evaluation concurrency More...
bool	derived_scheduler_overload () const override
	flag which prevents overloading the scheduler with a multiprocessor evaluation (forwarded to optionalInterface) More...
IntIntPair	estimate_partition_bounds (int max_eval_concurrency) override
	estimate the minimum and maximum partition sizes that can be utilized by this Model
void	derived_init_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true) override
	set up optionalInterface and subModel for parallel operations More...
void	derived_init_serial () override
	set up optionalInterface and subModel for serial operations.
void	derived_set_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true) override
	set active parallel configuration within subModel
void	derived_free_communicators (ParLevLIter pl_iter, int max_eval_concurrency, bool recurse_flag=true) override
	deallocate communicator partitions for the NestedModel (forwarded to optionalInterface and subModel)
void	serve_run (ParLevLIter pl_iter, int max_eval_concurrency) override
	Service optionalInterface and subModel job requests received from the scheduler. Completes when a termination message is received from stop_servers().
void	stop_servers () override
	Executed by the scheduler to terminate server operations for subModel and optionalInterface when iteration on the NestedModel is complete.
const String &	interface_id () const override
	return the optionalInterface identifier
int	derived_evaluation_id () const override
	Return the current evaluation id for the NestedModel. More...
void	set_evaluation_reference () override
	set the evaluation counter reference points for the NestedModel (request forwarded to optionalInterface and subModel)
void	fine_grained_evaluation_counters () override
	request fine-grained evaluation reporting within optionalInterface and subModel
void	print_evaluation_summary (std::ostream &s, bool minimal_header=false, bool relative_count=true) const override
	print the evaluation summary for the NestedModel (request forwarded to optionalInterface and subModel)
void	warm_start_flag (const bool flag) override
	set the warm start flag, including actualModel
void	initialize_iterator (int job_index)
void	pack_parameters_buffer (MPIPackBuffer &send_buffer, int job_index)
void	unpack_parameters_buffer (MPIUnpackBuffer &recv_buffer, int job_index)
void	unpack_parameters_initialize (MPIUnpackBuffer &recv_buffer, int job_index)
void	pack_results_buffer (MPIPackBuffer &send_buffer, int job_index)
void	unpack_results_buffer (MPIUnpackBuffer &recv_buffer, int job_index)
void	update_local_results (int job_index)
ActiveSet	default_interface_active_set ()
Protected Member Functions inherited from Model
	Model (const ShortShortPair &vars_view, const SharedVariablesData &svd, bool share_svd, const SharedResponseData &srd, bool share_srd, const ActiveSet &set, short output_level, ProblemDescDB &problem_db=dummy_db, ParallelLibrary &parallel_lib=dummy_lib)
	constructor initializing the base class part of letter classes (BaseConstructor overloading avoids infinite recursion in the derived class constructors - Coplien, p. 139) More...
	Model (ProblemDescDB &problem_db, ParallelLibrary &parallel_lib)
	constructor initializing base class for recast model instances More...
virtual const IntResponseMap &	derived_synchronize_nowait ()
	portion of synchronize_nowait() specific to derived model classes
void	update_model_active_variables (Model &model)
	update incoming (sub-)model with active values from currentVariables
IntResponseMap &	response_map ()
	return responseMap
void	initialize_distribution (Pecos::MultivariateDistribution &mv_dist, bool active_only=false)
	initialize distribution types from problemDescDB More...
void	initialize_active_types (Pecos::MultivariateDistribution &mv_dist)
	initialize distribution types from problemDescDB More...
void	initialize_distribution_parameters (Pecos::MultivariateDistribution &mv_dist, bool active_only=false)
	initialize distribution parameters from problemDescDB
void	set_ie_asynchronous_mode (int max_eval_concurrency)
	default logic for defining asynchEvalFlag and evaluationCapacity based on ie_pl settings
void	assign_max_strings (const Pecos::MultivariateDistribution &mv_dist, Variables &vars)
	assign all of the longest possible string values into vars
SSCIter	max_string (const StringSet &ss)
	return iterator for longest string value found in string set
SRMCIter	max_string (const StringRealMap &srm)
	return iterator for longest string value found in string map
SizetMultiArrayConstView	initialize_x0_bounds (const SizetArray &original_dvv, bool &active_derivs, bool &inactive_derivs, RealVector &x0, RealVector &fd_lb, RealVector &fd_ub) const
	Initialize data needed for computing finite differences (active/inactive, center point, and bounds)
Real	forward_grad_step (size_t num_deriv_vars, size_t xj_index, Real x0_j, Real lb_j, Real ub_j)
	Compute the forward step for a finite difference gradient; updates shortStep.
EvaluationsDBState	evaluations_db_state (const Interface &interface)
	Return the interface flag for the EvaluationsDB state.
EvaluationsDBState	evaluations_db_state (const Model &model)
	Return the model flag for the EvaluationsDB state.
void	asynch_eval_store (const Interface &interface, const int &id, const Response &response)
	Store the response portion of an interface evaluation. Called from rekey_response_map()
void	asynch_eval_store (const Model &model, const int &id, const Response &response)
	Exists to support storage of interface evaluations. No-op so that rekey_response_map<Model> can be generated.
template<typename MetaType >
void	rekey_response_map (MetaType &meta_object, IntIntMapArray &id_maps, IntResponseMapArray &resp_maps_rekey, bool deep_copy)
	rekey returned jobs matched in array of id_maps into array of resp_maps_rekey; unmatched jobs can be cached within the meta_object
template<typename MetaType >
void	rekey_response_map (MetaType &meta_object, IntIntMap &id_map, IntResponseMap &resp_map_rekey, bool deep_copy)
	rekey returned jobs matched in id_map into resp_map_rekey; unmatched jobs can be cached within the meta_object
template<typename MetaType >
void	rekey_synch (MetaType &meta_object, bool block, IntIntMapArray &id_maps, IntResponseMapArray &resp_maps_rekey, bool deep_copy=false)
	synchronize via meta_object and rekey returned jobs matched in array of id_maps into array of resp_maps_rekey; unmatched jobs are cached within the meta_object
template<typename MetaType >
void	rekey_synch (MetaType &meta_object, bool block, IntIntMap &id_map, IntResponseMap &resp_map_rekey, bool deep_copy=false)
	synchronize via meta_object and rekey returned jobs matched in id_map into resp_map_rekey; unmatched jobs are cached within the meta_object

Private Member Functions
void	init_sub_iterator ()
	init subIterator-based counts and init subModel with mapping data
PRPQueueIter	job_index_to_queue_iterator (int job_index)
	convert job_index to an eval_id through subIteratorIdMap and eval_id to a subIteratorPRPQueue queue iterator
void	initialize_iterator (const Variables &vars, const ActiveSet &set, int eval_id)
	lower level function shared by initialize_iterator(int) and unpack_parameters_initialize()
void	unpack (MPIUnpackBuffer &recv_buffer, int job_index, Variables &vars, ActiveSet &set, int &eval_id)
	lower level function shared by unpack_parameters_buffer() and unpack_parameters_initialize()
void	resolve_map1 (const String &map1, size_t &ac_index1, size_t &adi_index1, size_t &ads_index1, size_t &adr_index1, size_t curr_index, short &inactive_sm_view)
	compute variable mapping indices corresponding to map1 and update inactive view if necessary
void	resolve_real_variable_mapping (const String &map1, const String &map2, size_t curr_index, short &inactive_sm_view)
	for a named real mapping, resolve primary index and secondary target
void	resolve_integer_variable_mapping (const String &map1, const String &map2, size_t curr_index, short &inactive_sm_view)
	for a named integer mapping, resolve primary index and secondary target
void	resolve_string_variable_mapping (const String &map1, const String &map2, size_t curr_index, short &inactive_sm_view)
	for a named string mapping, resolve primary index and secondary target
void	real_variable_mapping (Real r_var, size_t av_index, short svm_target)
	insert r_var into appropriate recipient
void	integer_variable_mapping (int i_var, size_t av_index, short svm_target)
	insert i_var into appropriate recipient
void	string_variable_mapping (const String &s_var, size_t av_index, short svm_target)
	insert s_var into appropriate recipient
void	set_mapping (const ActiveSet &mapped_set, ActiveSet &interface_set, bool &opt_interface_map, ActiveSet &sub_iterator_set, bool &sub_iterator_map)
	define the evaluation requirements for the optionalInterface (interface_set) and the subIterator (sub_iterator_set) from the total model evaluation requirements (mapped_set)
void	response_mapping (const Response &interface_response, const Response &sub_iterator_response, Response &mapped_response)
	combine the response from the optional interface evaluation with the response from the sub-iteration using the primaryCoeffs/secondaryCoeffs mappings to create the total response for the model More...
void	interface_response_overlay (const Response &opt_interface_response, Response &mapped_response)
	assign the response from the optional interface evaluation within the total response for the model
void	iterator_response_overlay (const Response &sub_iterator_response, Response &mapped_response)
	overlay the sub-iteration response within the total response for the model using the primaryCoeffs/secondaryCoeffs mappings
void	iterator_error_estimation (const RealSymMatrix &sub_iterator_errors, RealVector &mapped_errors)
	combine error estimates from the sub-iteration to define mappedErrorEstimates
Response &	nested_response (int nested_cntr)
	locate existing or allocate new entry in nestedResponseMap
void	check_response_map (const ShortArray &mapped_asv)
	check function counts for the mapped_asv
void	update_inactive_view (short new_view, short &view)
	update inactive variables view for subIterator based on new_view
void	update_inactive_view (unsigned short type, short &view)
	update inactive variables view for subIterator based on type
void	update_sub_model (const Variables &vars, const Constraints &cons)
	update subModel with current variable values/bounds/labels

Private Attributes
int	nestedModelEvalCntr
	number of calls to derived_evaluate()/derived_evaluate_nowait()
bool	firstUpdate
	boolean to trigger one-time updates on first call to update_sub_model()
IntResponseMap	nestedResponseMap
	used to return a map of nested responses (including subIterator and optionalInterface contributions) for aggregation and rekeying at the base class level
RealVector	mappedErrorEstimates
	mapping of subIterator->response_error_estimates() through primary and secondary mappings
size_t	outerMIPLIndex
	the miPLIndex for the outer parallelism context, prior to any subIterator partitioning
std::shared_ptr< Iterator >	subIterator
	the sub-iterator that is executed on every evaluation of this model
std::shared_ptr< Model >	subModel
	the sub-model used in sub-iterator evaluations More...
PRPQueue	subIteratorPRPQueue
	job queue for asynchronous execution of subIterator jobs
IteratorScheduler	subIteratorSched
	scheduling object for concurrent iterator parallelism
String	subMethodPointer
	the sub-method pointer from the nested model specification
int	subIteratorJobCntr
	subIterator job counter since last synchronize()
IntIntMap	subIteratorIdMap
	mapping from subIterator evaluation counter to nested model counter (different when subIterator evaluations do not occur on every nested model evaluation due to variable ASV content)
size_t	numSubIterFns = 0
	number of sub-iterator response functions prior to mapping
size_t	numSubIterMappedIneqCon = 0
	number of top-level inequality constraints mapped from the sub-iteration results
size_t	numSubIterMappedEqCon = 0
	number of top-level equality constraints mapped from the sub-iteration results
std::shared_ptr< Interface >	optionalInterface
	the optional interface contributes nonnested response data to the total model response
String	optInterfacePointer
	the optional interface pointer from the nested model specification
Response	optInterfaceResponse
	the response object resulting from optional interface evaluations
IntIntMap	optInterfaceIdMap
	mapping from optionalInterface evaluation counter to nested model counter (different when optionalInterface evaluations do not occur on every nested model evaluation due to variable ASV content)
size_t	numOptInterfPrimary = 0
	number of primary response functions (objective/least squares/generic functions) resulting from optional interface evaluations
size_t	numOptInterfIneqCon = 0
	number of inequality constraints resulting from optional interface evaluations
size_t	numOptInterfEqCon = 0
	number of equality constraints resulting from the optional interface evaluations
IntSet	optInterfGradIdAnalytic
	analytic IDs for mixed gradients on the optional interface
IntSet	optInterfHessIdAnalytic
	analytic IDs for mixed Hessians on the optional interface
String	optInterfGradientType
	Gradient type for the optional interface.
String	optInterfHessianType
	Hessian type for the optional interface.
SizetArray	active1ACVarMapIndices
	"primary" variable mappings for inserting active continuous currentVariables within all continuous subModel variables. If there are no secondary mappings defined, then the insertions replace the subModel variable values.
SizetArray	active1ADIVarMapIndices
	"primary" variable mappings for inserting active discrete int currentVariables within all discrete int subModel variables. No secondary mappings are defined for discrete int variables, so the active variables replace the subModel variable values.
SizetArray	active1ADSVarMapIndices
	"primary" variable mappings for inserting active discrete string currentVariables within all discrete string subModel variables. No secondary mappings are defined for discrete string variables, so the active variables replace the subModel variable values.
SizetArray	active1ADRVarMapIndices
	"primary" variable mappings for inserting active discrete real currentVariables within all discrete real subModel variables. No secondary mappings are defined for discrete real variables, so the active variables replace the subModel variable values.
ShortArray	active2ACVarMapTargets
	"secondary" variable mappings for inserting active continuous currentVariables into sub-parameters (e.g., distribution parameters for uncertain variables or bounds for continuous design/state variables) within all continuous subModel variables.
ShortArray	active2ADIVarMapTargets
	"secondary" variable mappings for inserting active discrete int currentVariables into sub-parameters (e.g., bounds for discrete design/state variables) within all discrete int subModel variables.
ShortArray	active2ADSVarMapTargets
	"secondary" variable mappings for inserting active discrete string currentVariables into sub-parameters (e.g., bounds for discrete design/state variables) within all discrete string subModel variables.
ShortArray	active2ADRVarMapTargets
	"secondary" variable mappings for inserting active discrete real currentVariables into sub-parameters (e.g., bounds for discrete design/state variables) within all discrete real subModel variables.
SizetArray	complement1ACVarMapIndices
	"primary" variable mappings for inserting the complement of the active continuous currentVariables within all continuous subModel variables
SizetArray	complement1ADIVarMapIndices
	"primary" variable mappings for inserting the complement of the active discrete int currentVariables within all discrete int subModel variables
SizetArray	complement1ADSVarMapIndices
	"primary" variable mappings for inserting the complement of the active discrete string currentVariables within all discrete string subModel variables
SizetArray	complement1ADRVarMapIndices
	"primary" variable mappings for inserting the complement of the active discrete real currentVariables within all discrete real subModel variables
BitArray	extraCVarsData
	flags for updating subModel continuous bounds and labels, one for each active continuous variable in currentVariables
BitArray	extraDIVarsData
	flags for updating subModel discrete int bounds and labels, one for each active discrete int variable in currentVariables
BitArray	extraDSVarsData
	flags for updating subModel discrete string labels, one for each active discrete string variable in currentVariables
BitArray	extraDRVarsData
	flags for updating subModel discrete real bounds and labels, one for each active discrete real variable in currentVariables
bool	identityRespMap = false
	whether identity response mapping is active
size_t	subIterMappedPri = 0
	number of sub-iterator results functions mapped to nested model primary functions (cached for use with identity case)
size_t	subIterMappedSec = 0
	number of sub-iterator results functions mapped to nested model secondary functions (cached for use with identity case)
RealMatrix	primaryRespCoeffs
	"primary" response_mapping matrix applied to the sub-iterator response functions. For OUU, the matrix is applied to UQ statistics to create contributions to the top-level objective functions/least squares/ generic response terms.
RealMatrix	secondaryRespCoeffs
	"secondary" response_mapping matrix applied to the sub-iterator response functions. For OUU, the matrix is applied to UQ statistics to create contributions to the top-level inequality and equality constraints.

Friends
class	IteratorScheduler
	protect scheduler callback functions from general access

Additional Inherited Members
Static Public Member Functions inherited from Model
static void	active_variables (const RealVector &config_vars, Model &model)
	set the specified configuration to the Model's inactive vars, converting from real to integer or through index to string value as needed More...
static void	inactive_variables (const RealVector &config_vars, Model &model)
	set the specified configuration to the Model's inactive vars, converting from real to integer or through index to string value as needed More...
static void	inactive_variables (const RealVector &config_vars, Model &model, Variables &updated_vars)
static void	evaluate (const RealMatrix &samples_matrix, Model &model, RealMatrix &resp_matrix)
	Bulk synchronously evaluate the model for each column (of active variables) in the samples matrix and return as columns of the response matrix.
static void	evaluate (const VariablesArray &sample_vars, Model &model, RealMatrix &resp_matrix)
	Bulk synchronously evaluate the model for each entry (of active variables) in the samples vector and return as columns of the response matrix.
Static Protected Member Functions inherited from Model
static String	user_auto_id ()
	return the next available model ID for no-ID user methods More...
static String	no_spec_id ()
	return the next available model ID for on-the-fly methods More...
Protected Attributes inherited from Model
Variables	currentVariables
	the set of current variables used by the model for performing function evaluations
size_t	numDerivVars
	the number of active continuous variables used in computing most response derivatives (i.e., in places such as quasi-Hessians and response corrections where only the active continuous variables are supported)
Response	currentResponse
	the set of current responses that holds the results of model function evaluations
size_t	numFns
	the number of functions in currentResponse
Constraints	userDefinedConstraints
	Explicit constraints on variables are maintained in the Constraints class hierarchy. Currently, this includes linear constraints and bounds, but could be extended in the future to include other explicit constraints which (1) have their form specified by the user, and (2) are not catalogued in Response since their form and coefficients are published to an iterator at startup.
String	modelId
	model identifier string from the input file
String	modelType
	type of model: simulation, nested, or surrogate
String	surrogateType
	type of surrogate model: local_*, multipoint_*, global_*, or hierarchical
String	gradientType
	type of gradient data: analytic, numerical, mixed, or none
String	methodSource
	source of numerical gradient routine: dakota or vendor
String	intervalType
	type of numerical gradient interval: central or forward
String	hessianType
	type of Hessian data: analytic, numerical, quasi, mixed, or none
RealVector	fdGradStepSize
	relative finite difference step size for numerical gradients More...
String	fdGradStepType
	type of finite difference step to use for numerical gradient: relative - step length is relative to x absolute - step length is what is specified bounds - step length is relative to range of x
RealVector	fdHessByGradStepSize
	relative finite difference step size for numerical Hessians estimated using first-order differences of gradients More...
RealVector	fdHessByFnStepSize
	relative finite difference step size for numerical Hessians estimated using second-order differences of function values More...
String	fdHessStepType
	type of finite difference step to use for numerical Hessian: relative - step length is relative to x absolute - step length is what is specified bounds - step length is relative to range of x
bool	ignoreBounds
	option to ignore bounds when computing finite diffs
bool	centralHess
	option to use old 2nd-order finite diffs for Hessians
bool	warmStartFlag
	if in warm-start mode, don't reset accumulated data (e.g., quasiHessians)
bool	supportsEstimDerivs
	whether model should perform or forward derivative estimation
String	quasiHessType
	quasi-Hessian type: bfgs, damped_bfgs, sr1
IntSet	gradIdAnalytic
	analytic id's for mixed gradients
IntSet	gradIdNumerical
	numerical id's for mixed gradients
IntSet	hessIdAnalytic
	analytic id's for mixed Hessians
IntSet	hessIdNumerical
	numerical id's for mixed Hessians
IntSet	hessIdQuasi
	quasi id's for mixed Hessians
EvaluationsDBState	modelEvaluationsDBState
	Whether to write model evals to the evaluations DB.
EvaluationsDBState	interfEvaluationsDBState
	Whether to write interface evals to the evaluations DB.
IntArray	messageLengths
	length of packed MPI buffers containing vars, vars/set, response, and PRPair
bool	mappingInitialized
	track use of initialize_mapping() and finalize_mapping()
ProblemDescDB &	probDescDB
	class member reference to the problem description database More...
ParallelLibrary &	parallelLib
	class member reference to the parallel library
ParConfigLIter	modelPCIter
	the ParallelConfiguration node used by this Model instance
short	componentParallelMode
	the component parallelism mode: NO_PARALLEL_MODE, SURROGATE_MODEL_MODE,
bool	asynchEvalFlag
	flags asynch evaluations (local or distributed)
int	evaluationCapacity
	capacity for concurrent evaluations supported by the Model
short	outputLevel
	output verbosity level: {SILENT,QUIET,NORMAL,VERBOSE,DEBUG}_OUTPUT
Pecos::MultivariateDistribution	mvDist
	the multivariate random variable distribution (in probability space corresponding to currentVariables)
BoolDeque	primaryRespFnSense
	array of flags (one per primary function) for switching the sense to maximize the primary function (default is minimize)
RealVector	primaryRespFnWts
	primary response function weightings (either weights for multiobjective optimization or weighted least squares)
bool	hierarchicalTagging
	whether to perform hierarchical evalID tagging of params/results
ScalingOptions	scalingOpts
	user-provided scaling data from the problem DB, possibly modified by Recasting
String	evalTagPrefix
	cached evalTag Prefix from parents to use at evaluate time
EvaluationStore &	evaluationsDB
	reference to the global evaluation database

Detailed Description

Derived model class which performs a complete sub-iterator execution within every evaluation of the model.

The NestedModel class nests a sub-iterator execution within every model evaluation. This capability is most commonly used for optimization under uncertainty, in which a nondeterministic iterator is executed on every optimization function evaluation. The NestedModel also contains an optional interface, for portions of the model evaluation which are independent from the sub-iterator, and a set of mappings for combining sub-iterator and optional interface data into a top level response for the model.

Member Function Documentation

derived_evaluate()

void derived_evaluate ( const ActiveSet &  set )

overrideprotectedvirtual

portion of evaluate() specific to NestedModel

Update subModel's inactive variables with active variables from currentVariables, compute the optional interface and sub-iterator responses, and map these to the total model response.

Reimplemented from Model.

References NestedModel::active2ACVarMapTargets, Response::active_set(), ParallelLibrary::bcast(), ParallelLibrary::bcast_hs(), NestedModel::component_parallel_mode(), Model::currentResponse, Model::currentVariables, Model::evalTagPrefix, Model::evaluationsDB, Model::hierarchicalTagging, NestedModel::interface_id(), NestedModel::interface_response_overlay(), Model::interfEvaluationsDBState, NestedModel::iterator_response_overlay(), IteratorScheduler::iteratorCommSize, IteratorScheduler::iteratorScheduling, IteratorScheduler::messagePass, IteratorScheduler::miPLIndex, Model::modelId, Model::modelPCIter, NestedModel::nestedModelEvalCntr, NestedModel::optInterfaceResponse, NestedModel::optionalInterface, Model::outputLevel, ParallelLibrary::parallel_configuration_iterator(), Model::parallelLib, IteratorScheduler::peerAssignJobs, Response::reset(), IteratorScheduler::run_iterator(), NestedModel::set_mapping(), IteratorScheduler::stop_iterator_servers(), NestedModel::subIterator, NestedModel::subIteratorSched, NestedModel::update_sub_model(), and Model::userDefinedConstraints.

derived_evaluate_nowait()

void derived_evaluate_nowait ( const ActiveSet &  set )

overrideprotectedvirtual

portion of evaluate_nowait() specific to NestedModel

Asynchronous execution of subIterator on subModel and, optionally, optionalInterface.

Reimplemented from Model.

References Response::active_set(), Model::currentResponse, Model::currentVariables, Model::evaluationsDB, NestedModel::interface_id(), Model::interfEvaluationsDBState, Model::modelId, NestedModel::nestedModelEvalCntr, NestedModel::optInterfaceIdMap, NestedModel::optInterfaceResponse, NestedModel::optionalInterface, NestedModel::set_mapping(), NestedModel::subIterator, NestedModel::subIteratorIdMap, NestedModel::subIteratorJobCntr, and NestedModel::subIteratorPRPQueue.

derived_synchronize()

const IntResponseMap & derived_synchronize ( )

overrideprotectedvirtual

portion of synchronize() specific to NestedModel

Recovery of asynchronous subIterator executions and, optionally, asynchronous optionalInterface mappings.

Reimplemented from Model.

References NestedModel::component_parallel_mode(), NestedModel::interface_response_overlay(), NestedModel::iterator_response_overlay(), Model::modelPCIter, NestedModel::nested_response(), NestedModel::nestedResponseMap, IteratorScheduler::numIteratorJobs, NestedModel::optInterfaceIdMap, NestedModel::optInterfacePointer, NestedModel::optionalInterface, ParallelLibrary::parallel_configuration_iterator(), Model::parallelLib, IteratorScheduler::schedule_iterators(), NestedModel::subIterator, NestedModel::subIteratorIdMap, NestedModel::subIteratorJobCntr, NestedModel::subIteratorPRPQueue, and NestedModel::subIteratorSched.

local_eval_synchronization()

short local_eval_synchronization ( )

inlineoverrideprotectedvirtual

return optionalInterface synchronization setting

Used in setting Model::asynchEvalFlag. subModel synchronization is used for setting asynchEvalFlag within subModel.

Reimplemented from Model.

References NestedModel::optInterfacePointer, and NestedModel::optionalInterface.

local_eval_concurrency()

int local_eval_concurrency ( )

inlineoverrideprotectedvirtual

return optionalInterface asynchronous evaluation concurrency

Used in setting Model::evaluationCapacity. subModel concurrency is used for setting evaluationCapacity within subModel.

Reimplemented from Model.

References NestedModel::optInterfacePointer, and NestedModel::optionalInterface.

derived_scheduler_overload()

bool derived_scheduler_overload ( ) const

inlineoverrideprotectedvirtual

flag which prevents overloading the scheduler with a multiprocessor evaluation (forwarded to optionalInterface)

Derived scheduler overload for subModel is handled separately in subModel.evaluate() within subIterator->run().

Reimplemented from Model.

References IteratorScheduler::iteratorScheduling, NestedModel::optInterfacePointer, NestedModel::optionalInterface, IteratorScheduler::procsPerIterator, NestedModel::subIterator, and NestedModel::subIteratorSched.

derived_init_communicators()

void derived_init_communicators ( ParLevLIter  pl_iter, int  max_eval_concurrency, bool  recurse_flag = true  )

overrideprotectedvirtual

set up optionalInterface and subModel for parallel operations

Asynchronous flags need to be initialized for the subModel. In addition, max_eval_concurrency is the outer level iterator concurrency, not the subIterator concurrency that subModel will see, and recomputing the message_lengths on the subModel is probably not a bad idea either. Therefore, recompute everything on subModel using init_communicators().

Reimplemented from Model.

References Response::active_set(), IteratorScheduler::configure(), Model::currentVariables, ProblemDescDB::get_db_method_node(), ProblemDescDB::get_db_model_node(), IteratorScheduler::init_iterator(), NestedModel::init_sub_iterator(), IteratorScheduler::iterator_message_lengths(), IteratorScheduler::iteratorServerId, Model::messageLengths, IteratorScheduler::messagePass, Model::modelPCIter, IteratorScheduler::numIteratorServers, NestedModel::optInterfacePointer, NestedModel::optionalInterface, ParallelLibrary::parallel_configuration_iterator(), Model::parallelLib, IteratorScheduler::partition(), Model::probDescDB, MPIPackBuffer::reset(), ProblemDescDB::set_db_list_nodes(), ProblemDescDB::set_db_method_node(), ProblemDescDB::set_db_model_nodes(), MPIPackBuffer::size(), NestedModel::subIterator, NestedModel::subIteratorSched, NestedModel::subMethodPointer, NestedModel::subModel, and IteratorScheduler::update().

derived_evaluation_id()

int derived_evaluation_id ( ) const

inlineoverrideprotectedvirtual

Return the current evaluation id for the NestedModel.

return the top level nested evaluation count. To get the lower level eval count, the subModel must be explicitly queried. This is consistent with the eval counter definitions in surrogate models.

Reimplemented from Model.

References NestedModel::nestedModelEvalCntr.

response_mapping()

void response_mapping ( const Response &  opt_interface_response, const Response &  sub_iterator_response, Response &  mapped_response  )

inlineprivate

combine the response from the optional interface evaluation with the response from the sub-iteration using the primaryCoeffs/secondaryCoeffs mappings to create the total response for the model

In the OUU case,

optionalInterface fns = {f}, {g} (deterministic primary functions, constraints)
subIterator fns       = {S}      (UQ response statistics)

Problem formulation for mapped functions:
                  minimize    {f} + [W]{S}
                  subject to  {g_l} <= {g}    <= {g_u}
                              {a_l} <= [A]{S} <= {a_u}
                              {g}    == {g_t}
                              [A]{S} == {a_t}

where [W] is the primary_mapping_matrix user input (primaryRespCoeffs class attribute), [A] is the secondary_mapping_matrix user input (secondaryRespCoeffs class attribute), {{g_l},{a_l}} are the top level inequality constraint lower bounds, {{g_u},{a_u}} are the top level inequality constraint upper bounds, and {{g_t},{a_t}} are the top level equality constraint targets.

NOTE: optionalInterface/subIterator primary fns (obj/lsq/generic fns) overlap but optionalInterface/subIterator secondary fns (ineq/eq constraints) do not. The [W] matrix can be specified so as to allow

• some purely deterministic primary functions and some combined: [W] filled and [W].num_rows() < {f}.length() [combined first] or [W].num_rows() == {f}.length() and [W] contains rows of zeros [combined last]
• some combined and some purely stochastic primary functions: [W] filled and [W].num_rows() > {f}.length()
• separate deterministic and stochastic primary functions: [W].num_rows() > {f}.length() and [W] contains {f}.length() rows of zeros.

If the need arises, could change constraint definition to allow overlap as well: {g_l} <= {g} + [A]{S} <= {g_u} with [A] usage the same as for [W] above.

In the UOO case, things are simpler, just compute statistics of each optimization response function: [W] = [I], {f}/{g}/[A] are empty.

References Response::active_set_request_vector(), NestedModel::check_response_map(), NestedModel::interface_response_overlay(), and NestedModel::iterator_response_overlay().

Member Data Documentation

subModel

std::shared_ptr<Model> subModel

private

the sub-model used in sub-iterator evaluations

There are no restrictions on subModel, so arbitrary nestings are possible. This is commonly used to support surrogate-based optimization under uncertainty by having NestedModels contain SurrogateModels and vice versa.

Referenced by NestedModel::component_parallel_mode(), NestedModel::derived_init_communicators(), NestedModel::derived_init_serial(), NestedModel::derived_subordinate_models(), NestedModel::estimate_partition_bounds(), NestedModel::fine_grained_evaluation_counters(), NestedModel::integer_variable_mapping(), NestedModel::NestedModel(), NestedModel::print_evaluation_summary(), NestedModel::real_variable_mapping(), NestedModel::resolve_integer_variable_mapping(), NestedModel::resolve_map1(), NestedModel::resolve_real_variable_mapping(), NestedModel::resolve_string_variable_mapping(), NestedModel::serve_run(), NestedModel::set_mapping(), NestedModel::string_variable_mapping(), NestedModel::subordinate_model(), NestedModel::surrogate_response_mode(), NestedModel::update_inactive_view(), NestedModel::update_sub_model(), and NestedModel::warm_start_flag().

---

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

• NestedModel.hpp
• NestedModel.cpp