Heuristic Solution Generation in CPLEX
At every node, CPLEX gives you the opportunity to attempt to convert a fractional solution into an integer solution with the HeuristicCallback. In addition, CPLEX periodically uses its own heuristics, as described in the manual, to convert fractional solutions to heuristic ones. If you use a HeuristicCallback, the diagram below shows were it will be called.
Interestingly, the Heuristic callback can be called multiple times at a node, once for each round of cuts added. If your heuristic is computationally expensive, be sure to keep this in mind.
Warning
Observe that any solution generated by a HeuristicCallback will not be checked against the lazy constraints. The programmer is responsible for ensuring feasibility.
Implementing HeuristicCallback in CPLEX
HeuristicCallback works similarly to LazyConstraintCallback and UserCutCallback. Mathematically, you are passing CPLEX the following function:
- Input: A fractional solution to the LP relaxation of your problem, potentially after some lazy constraints, CPLEX generated cuts, or user cuts have been added
- Output: Either a single integral feasible solution or nothing
The Javadoc for HeuristicCallback can be found here, but we summarize the important methods in the table below (the interface is very similar to the other callbacks}):
Method Name |
Return Type |
Arguments |
Description |
|---|---|---|---|
getValue |
double |
IloNumVar var |
Returns the solution value of var at the current node. |
getValues |
double[] |
IloNumVar[] vars |
Returns the solution values for vars at the current node. |
setSolution |
void |
IloIntVar[] vars, double[] vals |
Injects a solution to be used as the potential new incumbent. The injected solution is specified by providing solution values for all variables. If a user heuristic is successful in finding a new candidate for an incumbent, it can be passed to IloCplex by the method setSolution. IloCplex analyzes the solution and, if it is both feasible and better than the current incumbent, uses it as the new incumbent. A solution is specified using arrays vars and vals, where valsi specifies the solution value for varsi. Do not call this method multiple times. Calling it again overwrites any previously specified solution. |