Class SplitRandomSelectFloat<T extends Var>

  • Type Parameters:
    T - type of variable being used in the search.
    All Implemented Interfaces:
    SelectChoicePoint<T>

    public class SplitRandomSelectFloat<T extends Var>
    extends SimpleSelect<T>
    It is simple and customizable selector of decisions (constraints) which will be enforced by search. However, it does not use P=c as a search decision but rather P <= c (potentially splitting the domain), unless c is equal to the maximal value in the domain of P then the constraint P < c is used.
    Version:
    4.7
    • Field Detail

      • leftFirst

        public boolean leftFirst
        It specifies if the left branch (values smaller or equal to the value selected) are first considered.
      • roundRobin

        public boolean roundRobin
      • currentIndex

        TimeStamp<java.lang.Integer> currentIndex
      • generator

        private final java.util.Random generator
    • Constructor Detail

      • SplitRandomSelectFloat

        public SplitRandomSelectFloat​(Store store,
                                      T[] variables,
                                      ComparatorVariable<T> varSelect)
        The constructor to create a simple choice select mechanism.
        Parameters:
        store - current store
        variables - variables upon which the choice points are created.
        varSelect - the variable comparator to choose the variable.
      • SplitRandomSelectFloat

        public SplitRandomSelectFloat​(Store store,
                                      T[] variables,
                                      ComparatorVariable<T> varSelect,
                                      ComparatorVariable<T> tieBreakerVarSelect)
        It constructs a simple selection mechanism for choice points.
        Parameters:
        store - current store
        variables - variables used as basis of the choice point.
        varSelect - the main variable comparator.
        tieBreakerVarSelect - secondary variable comparator employed if the first one gives the same metric.
    • Method Detail

      • getChoiceVariable

        public T getChoiceVariable​(int index)
        Description copied from class: SimpleSelect
        It returns the variable which is the base on the next choice point. Only if choice is of an X = C type. This function returns null if all variables have a value assigned or a choice point based on other type of constraint is being selected. The parameter index is the last value which have been return by this SelectChoicePoint object which has not been backtracked upon yet.
        Specified by:
        getChoiceVariable in interface SelectChoicePoint<T extends Var>
        Overrides:
        getChoiceVariable in class SimpleSelect<T extends Var>
        Parameters:
        index - the position of the last variable in selection choice point heuristic.
        Returns:
        variable based on which the choice needs to be created.
      • getChoiceConstraint

        public PrimitiveConstraint getChoiceConstraint​(int index)
        Description copied from class: SimpleSelect
        It always returns null as choice point is obtained by getChoiceVariable and getChoiceValue.
        Specified by:
        getChoiceConstraint in interface SelectChoicePoint<T extends Var>
        Overrides:
        getChoiceConstraint in class SimpleSelect<T extends Var>
        Parameters:
        index - the position of the last variable returned by selection choice point heuristic.
        Returns:
        primitive constraint which is a base of a choice point.
      • roundRobinVarSelection

        T roundRobinVarSelection​(int index)