Inheritance diagram for EffGlobalOptimizer:
Public Member Functions | |
EffGlobalOptimizer (Model &model) | |
standard constructor | |
~EffGlobalOptimizer () | |
alternate constructor for instantiations "on the fly" destructor | |
void | find_optimum () |
Redefines the run virtual function for the optimizer branch. | |
Private Member Functions | |
void | find_optimum_on_model () |
called by find_optimum for setUpType == "model" | |
void | get_best_sample () |
imporovement function | |
Real | expected_improvement (const RealVector &expected_values, const RealVector &c_variables) |
expected improvement function for the GP | |
RealVector | expected_violation (const RealVector &variances, const RealVector &c_variables) |
expected violation function for the constraint functions | |
void | update_penalty () |
initialize and update the penaltyParameter | |
void | update_augmented_lagrange_multipliers (const RealVector &fn_vals) |
initialize and update the Lagrange multipliers for augmented Lagrangian | |
Real | augmented_lagrangian_merit (const RealVector &fn_vals, const RealVector &nln_ineq_l_bnds, const RealVector &nln_ineq_u_bnds, const RealVector &nln_eq_tgts) |
compute an augmented Lagrangian function from a set of function values | |
Real | objective (const RealVector &fn_vals) |
compute a composite objective value from one or more objective functions | |
Real | constraint_violation (const RealVector &fn_vals, const Real &constraint_tol) |
compute the constraint violation from a set of function values | |
Static Private Member Functions | |
static void | EIF_objective_eval (const Variables &sub_model_vars, const Variables &recast_vars, const Response &sub_model_response, Response &recast_response) |
Expected Improvement (EIF) problem formulation for PMA. | |
static void | merit_objective_eval (const Variables &sub_model_vars, const Variables &recast_vars, const Response &sub_model_response, Response &recast_response) |
defines an augmented Lagrangian merit function from the sub_model_response | |
Private Attributes | |
String | setUpType |
(user-supplied functions mode for "on the fly" instantiations). | |
Real | Pi |
pi = 3.1415926535897932385 | |
Model | meritModel |
recast model to create an augmented Lagrangian merit function | |
Model | fHatModel |
if constraints present | |
Model | eifModel |
recast model used in solving the max(EIF) sub-problem | |
Iterator | eifOptimizer |
iterator used to solve the max(EIF) sub-problem | |
Real | meritFnStar |
minimum penalized response from among true function evaluations | |
RealVector | truthFnStar |
true function values corresponding to the minimum penalized response | |
RealVector | varStar |
point that corresponds to the optimal value meritFnStar | |
RealVector | augLagrangeMult |
Lagrange multipliers for augmented Lagrangian calculations. | |
Real | penaltyParameter |
penalty calculations; increased in update_penalty() | |
Real | eta |
constant used in etaSequence updates | |
Real | alphaEta |
power for etaSequence updates when updating penalty | |
Real | betaEta |
power for etaSequence updates when updating multipliers | |
Real | etaSequence |
Lagrangian updates (refer to Conn, Gould, and Toint, section 14.4). | |
size_t | numFns |
number of response functions | |
size_t | numVars |
number of active continuous variables | |
size_t | numObjFns |
number of objective functions | |
size_t | numNonlinIneqConstr |
number of nonlinear inequality constraints | |
size_t | numNonlinEqConstr |
number of nonlinear equality constraints | |
RealVector | nonlinIneqLowerBnds |
nonlinear inequality constraint lower bounds | |
RealVector | nonlinIneqUpperBnds |
nonlinear inequality constraint upper bounds | |
RealVector | nonlinEqTargets |
nonlinear equality constraint targets | |
Real | bigRealBoundSize |
cutoff value for continuous bounds | |
size_t | approxIters |
iteration counter | |
Static Private Attributes | |
static EffGlobalOptimizer * | effGlobalInstance |
functions in order to avoid the need for static data |
The EffGlobalOptimizer class provides an implementation of the Efficient Global Optimization algorithm developed by Jones, Schonlau, & Welch.
The user input mappings are as follows:
~EffGlobalOptimizer | ( | ) |
alternate constructor for instantiations "on the fly" destructor
This is an alternate constructor for instantiations on the fly using a Model but no ProblemDescDB.
Real objective | ( | const RealVector & | fn_vals | ) | [private] |
compute a composite objective value from one or more objective functions
The composite objective computation sums up the contributions from one of more objective functions using the multiobjective weights.
Real constraint_violation | ( | const RealVector & | fn_vals, | |
const Real & | constraint_tol | |||
) | [private] |
compute the constraint violation from a set of function values
Compute the quadratic constraint violation defined as cv = g+^T g+ + h+^T h+. This implementation supports equality constraints and 2-sided inequalities. The constraint_tol allows for a small constraint infeasibility (used for penalty methods, but not Lagrangian methods).