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

Sub Branch class for the PEBBL-based Minimizer. More...

Inherits branchSub.

Public Member Functions

 PebbldBranchSub ()
 Constructor.
 
 ~PebbldBranchSub ()
 Destructor.
 
PebbldBranchingglobal () const
 Returns a reference to the corresponding main Branching object.
 
pebbl::branching * bGlobal () const
 Returns a reference to the corresponding main Branching object.
 
void setGlobalInfo (PebbldBranching *global_)
 Method that sets up the main Branching object.
 
void setRootComputation ()
 Method that is called when declaring the current node as a root node.
 
void boundComputation (double *controlParam)
 Method that does the Bounding Operation. In other words, it calls the optimization algorithm on the relaxed domain.
 
bool candidateSolution ()
 Method called after the bounding operation, returns true if the bounding resulted in a possible solution to the original non-relaxed problem.
 
pebbl::solution * extractSolution ()
 Method that returns a PEBBL-based solution.
 
int splitComputation ()
 Method that determines how many branches are created and how. Returns the number of branches created from this object.
 
pebbl::branchSub * makeChild (int whichChild)
 Method that returns a new PebbldBranchSub object based on Objective Function improvements and the number of branches.
 
void pebbldSubAsChildOf (PebbldBranchSub *parent, int splitVar, int whichChild, std::vector< double > _candidate_x, RealVector _lower_bounds, RealVector _upper_bounds)
 Method that creates a new PebbldBranching object.
 

Protected Attributes

PebbldBranchingglobalPtr
 Pointer referring to all info passed to subproblem.
 
Model subModel
 Model used for sub-problem.
 
Iterator subNLPSolver
 Solver used for sub-problems.
 
std::vector< double > candidate_x
 Candidate solution after solving sub-problem (also the bound)
 
double candidate_objFn
 Objective value at the candidate solution.
 
int splitVar
 Variable to branch on.
 
RealVector cont_vars
 Initial variable values for sub-problem.
 
RealVector lower_bounds
 Lower bounds for sub-problem.
 
RealVector upper_bounds
 Upper bounds for sub-problem.
 

Friends

class PebbldBranching
 

Detailed Description

Sub Branch class for the PEBBL-based Minimizer.


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