11
$\begingroup$

I implemented a cplex model and I am convinced that the model should allow a better solution on a specific instance. However, when I impose the variable values of the solution onto the model, it returns no solution.

Is there a simple way to find "the inequality that makes the solution infeasible"? I think, I once had cplex give me a violated inequality?

| improve this question | |
$\endgroup$
10
$\begingroup$

CPLEX has a conflict refiner. The instructions for invoking it are at How to invoke the conflict refiner.

| improve this answer | |
$\endgroup$
7
$\begingroup$

With CPLEX_CMD : execute the following commands

read(your_problem.lp)      

optimize 

tools

conflict

display

conflict 

all

This will give you the smallest set of equations that make the problem infeasible.

| improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ May the downvoter explain the downvote ? The set of command lines provides exactly what OP is asking if I am not crazy. I use this all the time. $\endgroup$ – Kuifje Nov 14 '19 at 17:52
  • 1
    $\begingroup$ I didn't down-vote, but the question is how to find out why a particular hypothetical solution is infeasible, not why the model is infeasible. (The model in fact seems to be feasible, put possibly missing an optimal solution.) $\endgroup$ – prubin Nov 14 '19 at 22:19
  • $\begingroup$ @prubin thank you for your feedback you are right. In this case I would add the variable values (or a subset of them) as constraints of the program and follow the same procedure. $\endgroup$ – Kuifje Nov 15 '19 at 9:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.