Dakota  Version 6.15
Explore and Predict with Confidence
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Pages
Public Member Functions | Public Attributes | Private Member Functions | Friends | List of all members
DataModelRep Class Reference

Body class for model specification data. More...

Public Member Functions

 ~DataModelRep ()
 destructor
 

Public Attributes

String idModel
 string identifier for the model specification data set (from the id_model specification in ModelIndControl)
 
String modelType
 model type selection: single, surrogate, or nested (from the model type specification in ModelIndControl)
 
String variablesPointer
 string pointer to the variables specification to be used by this model (from the variables_pointer specification in ModelIndControl)
 
String interfacePointer
 string pointer to the interface specification to be used by this model (from the interface_pointer specification in ModelSingle and the optional_interface_pointer specification in ModelNested)
 
String responsesPointer
 string pointer to the responses specification to be used by this model (from the responses_pointer specification in ModelIndControl)
 
bool hierarchicalTags
 whether this model and its children will add hierarchy-based tags to eval ids
 
String subMethodPointer
 pointer to a sub-iterator used for global approximations (from the dace_method_pointer specification in ModelSurrG) or by nested models (from the sub_method_pointer specification in ModelNested)
 
String solutionLevelControl
 (state) variable identifier that defines a set or range of solution level controls (space/time discretization levels, iterative convergence tolerances, etc.) for defining a secondary hierarchy of fidelity within the scope of a single model form (from solution_level_control specification; see also ordered_model_fidelities)
 
RealVector solutionLevelCost
 array of relative simulation costs corresponding to each of the solution levels (from solution_level_cost specification; see also solution_level_control); a scalar input is interpreted as a constant cost multiplier to be applied recursively
 
SizetSet surrogateFnIndices
 array specifying the response function set that is approximated
 
String surrogateType
 the selected surrogate type: local_taylor, multipoint_tana, global_(neural_network,mars,orthogonal_polynomial,gaussian, polynomial,kriging), or hierarchical
 
String actualModelPointer
 pointer to the model specification for constructing the truth model used in constructing surrogates (from the actual_model_pointer specification in ModelSurrL and ModelSurrMP)
 
StringArray ensembleModelPointers
 an ordered (low to high) or unordered (peer) set of model pointers corresponding to a ensemble of modeling fidelities (from the ordered_model_fidelities specification in ModelSurrH or the unordered_model_fidelities specification in ModelSurrNonH)
 
int pointsTotal
 user-specified lower bound on total points with which to build the model (if reuse_points < pointsTotal, new samples will make up the difference)
 
short pointsManagement
 points management configuration for DataFitSurrModel: DEFAULT_POINTS, MINIMUM_POINTS, or RECOMMENDED_POINTS
 
String approxPointReuse
 sample reuse selection for building global approximations: none, all, region, or file (from the reuse_samples specification in ModelSurrG)
 
String importBuildPtsFile
 the file name from the import_build_points_file specification in ModelSurrG
 
unsigned short importBuildFormat
 tabular format for the build point import file
 
bool importUseVariableLabels
 whether to parse/validate variable labels from header
 
bool importBuildActive
 whether to import active variables only
 
String exportApproxPtsFile
 the file name from the export_approx_points_file specification in ModelSurrG
 
unsigned short exportApproxFormat
 tabular format for the approx point export file
 
String exportApproxVarianceFile
 filename for surrogate variance evaluation export
 
unsigned short exportApproxVarianceFormat
 tabular format for the approx variance export file
 
bool exportSurrogate
 Option to turn on surrogate model export (export_model)
 
String modelExportPrefix
 the filename prefix for export_model
 
unsigned short modelExportFormat
 Format selection for export_model.
 
bool importSurrogate
 Option to turn on surrogate model import (import_model)
 
String modelImportPrefix
 the filename prefix for import_model
 
unsigned short modelImportFormat
 Format selection for import_model.
 
short approxCorrectionType
 correction type for global and hierarchical approximations: NO_CORRECTION, ADDITIVE_CORRECTION, MULTIPLICATIVE_CORRECTION, or COMBINED_CORRECTION (from the correction specification in ModelSurrG and ModelSurrH)
 
short approxCorrectionOrder
 correction order for global and hierarchical approximations: 0, 1, or 2 (from the correction specification in ModelSurrG and ModelSurrH)
 
bool modelUseDerivsFlag
 flags the use of derivatives in building global approximations (from the use_derivatives specification in ModelSurrG)
 
bool respScalingFlag
 flag to indicate bounds-based scaling of current response data set prior to surrogate build; important for data fits of decaying discrepancy data using regression with absolute tolerances
 
short polynomialOrder
 scalar integer indicating the order of the polynomial approximation (1=linear, 2=quadratic, 3=cubic; from the polynomial specification in ModelSurrG)
 
RealVector krigingCorrelations
 vector of correlations used in building a kriging approximation (from the correlations specification in ModelSurrG)
 
String krigingOptMethod
 optimization method to use in finding optimal correlation parameters: none, sampling, local, global
 
short krigingMaxTrials
 maximum number of trials in optimization of kriging correlations
 
RealVector krigingMaxCorrelations
 upper bound on kriging correlation vector
 
RealVector krigingMinCorrelations
 lower bound on kriging correlation vector
 
Real krigingNugget
 nugget value for kriging
 
short krigingFindNugget
 option to have Kriging find the best nugget value to use
 
short mlsWeightFunction
 weight function for moving least squares approximation
 
short rbfBases
 bases for radial basis function approximation
 
short rbfMaxPts
 maximum number of points for radial basis function approximation
 
short rbfMaxSubsets
 maximum number of subsets for radial basis function approximation
 
short rbfMinPartition
 minimum partition for radial basis function approximation
 
short marsMaxBases
 maximum number of bases for MARS approximation
 
String marsInterpolation
 interpolation type for MARS approximation
 
short annRandomWeight
 random weight for artificial neural network approximation
 
short annNodes
 number of nodes for artificial neural network approximation
 
Real annRange
 range for artificial neural network approximation
 
int numRestarts
 number of restarts for gradient-based optimization in GP
 
bool domainDecomp
 whether domain decomposition is enabled
 
String decompCellType
 type of local cell of domain decomp
 
int decompSupportLayers
 number of support layers for each local basis function
 
bool decompDiscontDetect
 whether discontinuity detection is enabled
 
Real discontJumpThresh
 function value (jump) threshold for discontinuity detection in domain decomp
 
Real discontGradThresh
 gradient threshold for discontinuity detection in domain decomp
 
String trendOrder
 scalar integer indicating the order of the Gaussian process mean (0= constant, 1=linear, 2=quadratic, 3=cubic); from the gaussian_process specification in ModelSurrG)
 
bool pointSelection
 flag indicating the use of point selection in the Gaussian process
 
StringArray diagMetrics
 List of diagnostic metrics the user requests to assess the goodness of fit for a surrogate model.
 
bool crossValidateFlag
 flag indicating the use of cross validation on the metrics specified
 
int numFolds
 number of folds to perform in cross validation
 
Real percentFold
 percentage of data to withhold for cross validation process
 
bool pressFlag
 flag indicating the use of PRESS on the metrics specified
 
String importChallengePtsFile
 the file name from the challenge_points_file specification in ModelSurrG
 
unsigned short importChallengeFormat
 tabular format of the challenge data file
 
bool importChalUseVariableLabels
 whether to parse/validate variable labels from header
 
bool importChallengeActive
 whether to import active variables only
 
String advancedOptionsFilename
 file containing advanced surrogate option overrides
 
String optionalInterfRespPointer
 string pointer to the responses specification used by the optional interface in nested models (from the optional_interface_responses_pointer specification in ModelNested)
 
StringArray primaryVarMaps
 the primary variable mappings used in nested models for identifying the lower level variable targets for inserting top level variable values (from the primary_variable_mapping specification in ModelNested)
 
StringArray secondaryVarMaps
 the secondary variable mappings used in nested models for identifying the (distribution) parameter targets within the lower level variables for inserting top level variable values (from the secondary_variable_mapping specification in ModelNested)
 
RealVector primaryRespCoeffs
 the primary response mapping matrix used in nested models for weighting contributions from the sub-iterator responses in the top level (objective) functions (from the primary_response_mapping specification in ModelNested)
 
RealVector secondaryRespCoeffs
 the secondary response mapping matrix used in nested models for weighting contributions from the sub-iterator responses in the top level (constraint) functions (from the secondary_response_mapping specification in ModelNested)
 
bool identityRespMap
 whether an identity response map is requested in lieu of explicit maps
 
int subMethodServers
 number of servers for concurrent sub-iterator parallelism
 
int subMethodProcs
 number of processors for each concurrent sub-iterator partition
 
short subMethodScheduling
 scheduling approach for concurrent sub-iterator parallelism: {DEFAULT,MASTER,PEER}_SCHEDULING
 
int initialSamples
 initial samples to build the subspace model
 
unsigned short subspaceSampleType
 sampling method for building the subspace model
 
IntVector refineSamples
 refinement samples to add in each batch
 
size_t maxIterations
 maximum number of subspace build iterations
 
Real convergenceTolerance
 convergence tolerance on build process
 
bool subspaceIdBingLi
 Flag to use Bing Li method to identify active subspace dimension.
 
bool subspaceIdConstantine
 Flag to use Constantine method to identify active subspace dimension.
 
bool subspaceIdEnergy
 Flag to use eigenvalue energy method to identify active subspace dimension.
 
bool subspaceBuildSurrogate
 Flag to build surrogate over active subspace.
 
int subspaceDimension
 Size of subspace.
 
unsigned short subspaceNormalization
 Normalization to use when forming a subspace with multiple response functions.
 
int numReplicates
 Number of bootstrap samples for subspace identification.
 
bool subspaceIdCV
 Flag to use cross validation to identify active subspace dimension.
 
Real relTolerance
 relative tolerance used by cross validation subspace dimension id method
 
Real decreaseTolerance
 decrease tolerance used by cross validation subspace dimension id method
 
int subspaceCVMaxRank
 maximum rank considered by cross validation subspace dimension id method
 
bool subspaceCVIncremental
 flag to use incremental dimension estimation in the cross validation metric
 
unsigned short subspaceIdCVMethod
 Contains which cutoff method to use in the cross validation metric.
 
short regressionType
 type of (regularized) regression: FT_LS or FT_RLS2
 
Real regressionL2Penalty
 penalty parameter for regularized regression (FT_RLS2)
 
size_t maxSolverIterations
 max iterations for optimization solver used in FT regression
 
int maxCrossIterations
 maximum number of cross iterations
 
Real solverTol
 optimization tolerance for FT regression
 
Real solverRoundingTol
 Rounding tolerance for FT regression.
 
Real statsRoundingTol
 arithmetic (rounding) tolerance for FT sums and products
 
bool tensorGridFlag
 sub-sample a tensor grid for generating regression data
 
unsigned short startOrder
 starting polynomial order
 
unsigned short kickOrder
 polynomial order increment when adapting
 
unsigned short maxOrder
 maximum order of basis polynomials
 
bool adaptOrder
 whether or not to adapt order by cross validation
 
size_t startRank
 starting rank
 
size_t kickRank
 rank increase increment
 
size_t maxRank
 maximum rank
 
bool adaptRank
 whether or not to adapt rank
 
size_t maxCVRankCandidates
 maximum number of cross-validation candidates for adaptRank
 
unsigned short maxCVOrderCandidates
 maximum number of cross-validation candidates for adaptOrder
 
short c3AdvanceType
 quantity to increment (start rank, start order, max rank, max order, max rank + max order) for FT (uniform) p-refinement
 
size_t collocationPoints
 number of data points used in FT construction by regression
 
Real collocationRatio
 ratio of number of points to nuber of unknowns
 
bool autoRefine
 whether automatic surrogate refinement is enabled
 
size_t maxFunctionEvals
 maximum evals in refinement
 
String refineCVMetric
 metric to use in cross-validation guided refinement
 
int softConvergenceLimit
 max number of iterations in refinement without improvement
 
int refineCVFolds
 number of cross-validation folds in guided refinement
 
unsigned short adaptedBasisSparseGridLev
 sparse grid level for low-order PCE used to compute rotation matrix within adapted basis approach to dimension reduction
 
unsigned short adaptedBasisExpOrder
 expansion order for low-order PCE used to compute rotation matrix within adapted basis approach to dimension reduction
 
Real adaptedBasisCollocRatio
 collocation ratio for low-order PCE used to compute rotation matrix within adapted basis approach to dimension reduction
 
short method_rotation
 
Real adaptedBasisTruncationTolerance
 
unsigned short randomFieldIdForm
 Contains which type of random field model.
 
unsigned short analyticCovIdForm
 Contains which type of analytic covariance function.
 
Real truncationTolerance
 truncation tolerance on build process: percent variance explained
 
String propagationModelPointer
 pointer to the model through which to propagate the random field
 
String rfDataFileName
 File from which to build the random field.
 

Private Member Functions

 DataModelRep ()
 constructor
 
void write (std::ostream &s) const
 write a DataModelRep object to an std::ostream
 
void read (MPIUnpackBuffer &s)
 read a DataModelRep object from a packed MPI buffer
 
void write (MPIPackBuffer &s) const
 write a DataModelRep object to a packed MPI buffer
 

Friends

class DataModel
 the handle class can access attributes of the body class directly
 

Detailed Description

Body class for model specification data.

The DataModelRep class is used to contain the data from a model keyword specification. Default values are managed in the DataModelRep constructor. Data is public to avoid maintaining set/get functions, but is still encapsulated within ProblemDescDB since ProblemDescDB::dataModelList is private.


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