limp-cbc-0.3.2.0: cbits/coin/CbcHeuristicGreedy.hpp
/* $Id: CbcHeuristicGreedy.hpp 1585 2011-01-11 19:04:34Z forrest $ */
// Copyright (C) 2005, International Business Machines
// Corporation and others. All Rights Reserved.
// This code is licensed under the terms of the Eclipse Public License (EPL).
#ifndef CbcHeuristicGreedy_H
#define CbcHeuristicGreedy_H
#include "CbcHeuristic.hpp"
/** Greedy heuristic classes
*/
class CbcHeuristicGreedyCover : public CbcHeuristic {
public:
// Default Constructor
CbcHeuristicGreedyCover ();
/* Constructor with model - assumed before cuts
Initial version does not do Lps
*/
CbcHeuristicGreedyCover (CbcModel & model);
// Copy constructor
CbcHeuristicGreedyCover ( const CbcHeuristicGreedyCover &);
// Destructor
~CbcHeuristicGreedyCover ();
/// Clone
virtual CbcHeuristic * clone() const;
/// Assignment operator
CbcHeuristicGreedyCover & operator=(const CbcHeuristicGreedyCover& rhs);
/// Create C++ lines to get to current state
virtual void generateCpp( FILE * fp) ;
/// update model (This is needed if cliques update matrix etc)
virtual void setModel(CbcModel * model);
using CbcHeuristic::solution ;
/** returns 0 if no solution, 1 if valid solution.
Sets solution values if good, sets objective value (only if good)
We leave all variables which are at one at this node of the
tree to that value and will
initially set all others to zero. We then sort all variables in order of their cost
divided by the number of entries in rows which are not yet covered. We randomize that
value a bit so that ties will be broken in different ways on different runs of the heuristic.
We then choose the best one and set it to one and repeat the exercise.
*/
virtual int solution(double & objectiveValue,
double * newSolution);
/// Validate model i.e. sets when_ to 0 if necessary (may be NULL)
virtual void validate() ;
/// Resets stuff if model changes
virtual void resetModel(CbcModel * model);
/* Algorithm
0 - use current upper bounds
1 - use original upper bounds
If 10 added perturb ratios more
if 100 added round up all >=0.5
*/
inline int algorithm() const {
return algorithm_;
}
inline void setAlgorithm(int value) {
algorithm_ = value;
}
// Only do this many times
inline int numberTimes() const {
return numberTimes_;
}
inline void setNumberTimes(int value) {
numberTimes_ = value;
}
protected:
/// Guts of constructor from a CbcModel
void gutsOfConstructor(CbcModel * model);
// Data
// Original matrix by column
CoinPackedMatrix matrix_;
// original number of rows
int originalNumberRows_;
/* Algorithm
0 - use current upper bounds
1 - use original upper bounds
If 10 added perturb ratios more
*/
int algorithm_;
/// Do this many times
int numberTimes_;
};
class CbcHeuristicGreedyEquality : public CbcHeuristic {
public:
// Default Constructor
CbcHeuristicGreedyEquality ();
/* Constructor with model - assumed before cuts
Initial version does not do Lps
*/
CbcHeuristicGreedyEquality (CbcModel & model);
// Copy constructor
CbcHeuristicGreedyEquality ( const CbcHeuristicGreedyEquality &);
// Destructor
~CbcHeuristicGreedyEquality ();
/// Clone
virtual CbcHeuristic * clone() const;
/// Assignment operator
CbcHeuristicGreedyEquality & operator=(const CbcHeuristicGreedyEquality& rhs);
/// Create C++ lines to get to current state
virtual void generateCpp( FILE * fp) ;
/// update model (This is needed if cliques update matrix etc)
virtual void setModel(CbcModel * model);
using CbcHeuristic::solution ;
/** returns 0 if no solution, 1 if valid solution.
Sets solution values if good, sets objective value (only if good)
We leave all variables which are at one at this node of the
tree to that value and will
initially set all others to zero. We then sort all variables in order of their cost
divided by the number of entries in rows which are not yet covered. We randomize that
value a bit so that ties will be broken in different ways on different runs of the heuristic.
We then choose the best one and set it to one and repeat the exercise.
*/
virtual int solution(double & objectiveValue,
double * newSolution);
/// Validate model i.e. sets when_ to 0 if necessary (may be NULL)
virtual void validate() ;
/// Resets stuff if model changes
virtual void resetModel(CbcModel * model);
/* Algorithm
0 - use current upper bounds
1 - use original upper bounds
If 10 added perturb ratios more
if 100 added round up all >=0.5
*/
inline int algorithm() const {
return algorithm_;
}
inline void setAlgorithm(int value) {
algorithm_ = value;
}
// Fraction of rhs to cover before branch and cut
inline void setFraction(double value) {
fraction_ = value;
}
inline double fraction() const {
return fraction_;
}
// Only do this many times
inline int numberTimes() const {
return numberTimes_;
}
inline void setNumberTimes(int value) {
numberTimes_ = value;
}
protected:
/// Guts of constructor from a CbcModel
void gutsOfConstructor(CbcModel * model);
// Data
// Original matrix by column
CoinPackedMatrix matrix_;
// Fraction of rhs to cover before branch and cut
double fraction_;
// original number of rows
int originalNumberRows_;
/* Algorithm
0 - use current upper bounds
1 - use original upper bounds
If 10 added perturb ratios more
*/
int algorithm_;
/// Do this many times
int numberTimes_;
};
/** Greedy heuristic for SOS and L rows (and positive elements)
*/
class CbcHeuristicGreedySOS : public CbcHeuristic {
public:
// Default Constructor
CbcHeuristicGreedySOS ();
/* Constructor with model - assumed before cuts
Initial version does not do Lps
*/
CbcHeuristicGreedySOS (CbcModel & model);
// Copy constructor
CbcHeuristicGreedySOS ( const CbcHeuristicGreedySOS &);
// Destructor
~CbcHeuristicGreedySOS ();
/// Clone
virtual CbcHeuristic * clone() const;
/// Assignment operator
CbcHeuristicGreedySOS & operator=(const CbcHeuristicGreedySOS& rhs);
/// Create C++ lines to get to current state
virtual void generateCpp( FILE * fp) ;
/// update model (This is needed if cliques update matrix etc)
virtual void setModel(CbcModel * model);
using CbcHeuristic::solution ;
/** returns 0 if no solution, 1 if valid solution.
Sets solution values if good, sets objective value (only if good)
We leave all variables which are at one at this node of the
tree to that value and will
initially set all others to zero. We then sort all variables in order of their cost
divided by the number of entries in rows which are not yet covered. We randomize that
value a bit so that ties will be broken in different ways on different runs of the heuristic.
We then choose the best one and set it to one and repeat the exercise.
*/
virtual int solution(double & objectiveValue,
double * newSolution);
/// Validate model i.e. sets when_ to 0 if necessary (may be NULL)
virtual void validate() ;
/// Resets stuff if model changes
virtual void resetModel(CbcModel * model);
/* Algorithm
Bits
1 bit - use current model, otherwise original
2 - use current solution as starting point, otherwise pure greedy
4 - as 2 but use merit not merit/size
8 - use duals to modify greedy
16 - use duals on GUB/SOS in special way
*/
inline int algorithm() const {
return algorithm_;
}
inline void setAlgorithm(int value) {
algorithm_ = value;
}
// Only do this many times
inline int numberTimes() const {
return numberTimes_;
}
inline void setNumberTimes(int value) {
numberTimes_ = value;
}
protected:
/// Guts of constructor from a CbcModel
void gutsOfConstructor(CbcModel * model);
// Data
// Original RHS - if -1.0 then SOS otherwise <= value
double * originalRhs_;
// Original matrix by column
CoinPackedMatrix matrix_;
// original number of rows
int originalNumberRows_;
/* Algorithm
*/
int algorithm_;
/// Do this many times
int numberTimes_;
};
#endif