Decision Optimization

Decision Optimization

Delivers prescriptive analytics capabilities and decision intelligence to improve decision-making.


#Analytics
#DecisionOptimization
#DecisionOptimization
 View Only
  • 1.  MILP to LP automatically?

    Posted 01/12/11 02:57 PM

    Originally posted by: kc78


    Hi

    I was wondering if i fixed all the binary variables to the values returned after i solved it as a MILP, does CPLEX automatically switch it to solving it as a LP?

    Thanks!
    #CPLEXOptimizers
    #DecisionOptimization


  • 2.  Re: MILP to LP automatically?

    Posted 01/13/11 05:28 AM

    Originally posted by: SystemAdmin


    No, the problem remains a MIP even all integer variables are fixed. It is even possible to have a MIP that does not have any integer variables.
    This means, that after solving this fixed model you still cannot access LP solution features like a dual solution or reduced costs. If you want this, then you need to explicitly change the problem type to LP.

    Tobias
    #CPLEXOptimizers
    #DecisionOptimization


  • 3.  Re: MILP to LP automatically?

    Posted 02/15/11 05:00 PM

    Originally posted by: SystemAdmin


    Hi all,

    Tobias said:

    > you need to explicitly change the problem type to LP.

    How can I do it in OPL? How can I specify (in OPL) that I want a MIP model to be solved as LP problem?

    Thanks in advance,
    Paolo
    #CPLEXOptimizers
    #DecisionOptimization


  • 4.  Re: MILP to LP automatically?

    Posted 02/15/11 06:20 PM

    Originally posted by: SystemAdmin


    Hello again,

    I didn't notice OPL has its own forum:

    http://www.ibm.com/developerworks/forums/forum.jspa?forumID=2053

    so I suppose my topic resulted to be OT. I'm sorry.

    By the way, browsing the OPL forum I found how to solve the LP relaxation of a MIP problem, that is by using convertAllIntVars() method of IloOplModel class:

    dvar ...
    minimize ...
    subject to {...}

    main {
    thisOplModel.generate();
    thisOplModel.convertAllIntVars();
    cplex.solve();
    }

    Thanks anyway to everyone,
    Paolo
    #CPLEXOptimizers
    #DecisionOptimization