Inheritance diagram for BasisPolyApproximation:
Public Member Functions | |
| BasisPolyApproximation () | |
| default constructor | |
| BasisPolyApproximation (const ProblemDescDB &problem_db, const size_t &num_acv) | |
| standard constructor | |
| ~BasisPolyApproximation () | |
| destructor | |
| virtual const Real & | get_mean ()=0 |
| return the mean of the expansion, treating all variables as random | |
| virtual const Real & | get_mean (const RealVector &x)=0 |
| treating a subset of the variables as random | |
| virtual const RealBaseVector & | get_mean_gradient ()=0 |
| vector, treating all variables as random | |
| virtual const RealBaseVector & | get_mean_gradient (const RealVector &x, const UIntArray &dvv)=0 |
| and given DVV, treating a subset of the variables as random | |
| virtual const Real & | get_variance ()=0 |
| return the variance of the expansion, treating all variables as random | |
| virtual const Real & | get_variance (const RealVector &x)=0 |
| treating a subset of the variables as random | |
| virtual const RealBaseVector & | get_variance_gradient ()=0 |
| vector, treating all variables as random | |
| virtual const RealBaseVector & | get_variance_gradient (const RealVector &x, const UIntArray &dvv)=0 |
| vector and given DVV, treating a subset of the variables as random | |
| void | solution_approach (short soln_approach) |
| set expCoeffsSolnApproach | |
| short | solution_approach () const |
| get expCoeffsSolnApproach | |
| void | expansion_coefficient_flag (bool coeff_flag) |
| set expansionCoeffFlag | |
| bool | expansion_coefficient_flag () const |
| get expansionCoeffFlag | |
| void | expansion_gradient_flag (bool grad_flag) |
| set expansionGradFlag | |
| bool | expansion_gradient_flag () const |
| get expansionGradFlag | |
| void | integration_iterator (const Iterator &iterator) |
| set integrationRep | |
| void | random_variables_key (const BoolDeque &random_vars_key) |
| set randomVarsKey | |
Protected Member Functions | |
| int | num_constraints () const |
| return the number of constraints to be enforced via anchorPoint | |
| const RealVector & | approximation_coefficients () const |
| return the coefficient array computed by find_coefficients() | |
| void | approximation_coefficients (const RealVector &approx_coeffs) |
| computing with find_coefficients() | |
Protected Attributes | |
| short | expCoeffsSolnApproach |
| QUADRATURE, SPARSE_GRID, REGRESSION, or SAMPLING. | |
| bool | expansionCoeffFlag |
| flag for calculation of expansionCoeffs from response values | |
| bool | expansionGradFlag |
| flag for calculation of expansionCoeffGrads from response gradients | |
| NonDIntegration * | integrationRep |
| weight products | |
| BoolDeque | randomVarsKey |
| the active variables (used in all_variables mode) | |
| SizetList | randomIndices |
| variables (used in all_variables mode; defined from randomVarsKey) | |
| SizetList | nonRandomIndices |
| active variables (used in all_variables mode; defined from randomVarsKey) | |
| Real | expansionMean |
| expected value of the expansion | |
| RealBaseVector | expansionMeanGrad |
| gradient of the expected value of the expansion | |
| Real | expansionVariance |
| variance of the expansion | |
| RealBaseVector | expansionVarianceGrad |
| gradient of the variance of the expansion | |
| RealVector | expansionCoeffs |
| the coefficients of the expansion | |
| RealMatrix | expansionCoeffGrads |
| the gradients of the expansion coefficients | |
The BasisPolyApproximation class provides a global approximation based on basis polynomials. This includes orthogonal polynomials used for polynomial chaos expansions and interpolation polynomials used for stochastic collocation.
RealMatrix expansionCoeffGrads [protected] |
the gradients of the expansion coefficients
may be interpreted as either the gradients of the expansion coefficients or the coefficients of expansions for the response gradients. This array is used when sensitivities of moments are needed with respect to variables that do not appear in the expansion (e.g., with respect to design variables for an expansion only over the random variables).
1.5.1