net.sf.cpsolver.ifs.heuristics
Interface VariableSelection

All Known Implementing Classes:

DbtVariableSelection, ExamUnassignedVariableSelection, GeneralVariableSelection, LectureSelection, RouletteWheelRequestSelection

public interface VariableSelection

Variable selection criterion.

The IFS algorithm requires a function that selects a variable to be (re)assigned during the current iteration step. This problem is equivalent to a variable selection criterion in constraint programming. There are several guidelines for selecting a variable. In local search, the variable participating in the largest number of violations is usually selected first. In backtracking-based algorithms, the first-fail principle is often used, i.e., a variable whose instantiation is most complicated is selected first. This could be the variable involved in the largest set of constraints or the variable with the smallest domain, etc.

We can split the variable selection criterion into two cases. If some variables remain unassigned, the “worst” variable among them is selected, i.e., first-fail principle is applied. This may, for example, be the variable with the smallest domain or with the highest number of hard and/or soft constraints.

The second case occurs when all variables are assigned. Because the algorithm does not need to stop when a complete feasible solution is found, the variable selection criterion for such case has to be considered as well. Here all variables are assigned but the solution is not good enough, e.g., in the sense of violated soft constraints. We choose a variable whose change of a value can introduce the best improvement of the solution. It may, for example, be a variable whose value violates the highest number of soft constraints.

It is possible for the solution to become incomplete again after such an iteration because a value which is not consistent with all hard constraints can be selected in the value selection criterion. This can also be taken into account in the variable selection heuristics.

Version:

IFS 1.1 (Iterative Forward Search)
Copyright (C) 2006 Tomáš Müller
muller@unitime.org
Lazenska 391, 76314 Zlin, Czech Republic

This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version.

This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

See Also:

Solver

Method Summary

void	init(Solver solver) Initialization
Variable	selectVariable(Solution solution) Variable selection

Method Detail

init

void init(Solver solver)

Initialization

selectVariable

Variable selectVariable(Solution solution)

Variable selection

Parameters:

solution - current solution