Decision Optimization

Decision Optimization

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


#Analytics
#DecisionOptimization
#DecisionOptimization
 View Only
  • 1.  Using model output as input for next model (Incremental Build))

    Posted 03/04/15 04:59 AM

    Originally posted by: AMurshed


    Hello,

    I want to solve different models where each model depends on the output of the prevoius model.

    Is it OK to create the models as a serial format in the code where each model takes the output of the prevoius one?

    If there is a simple example on how to create such incremental MILP models, I would be thankful.

     

    Thanks.

    Ayman Murshed


    #DecisionOptimization
    #MathematicalProgramming-General


  • 2.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/06/15 01:31 AM

    I am not clear what you mean by "each model takes the output of the previous one". Can you please elaborate on that and explain in more detail what you are trying to do?


    #DecisionOptimization
    #MathematicalProgramming-General


  • 3.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/06/15 05:50 AM

    Originally posted by: AMurshed


    Thanks Daniel for your reply.

    Sorry for misunderstanding.

    I would like to create incremental MILP modleing using a loop, where each time the model initializes its constraints and creates new constraints in order to find the cost function as shown below.

    -------------------------------------------

    loop through messages

          schedule the message in LAN1

          schedule the message in LAN2

         ...

         schedule the message in LANn

    end loop

    -------------------------------------------

    I would like each time before scheduling the message, delete the constraints in LAN (x-1) in order to start feeding constraints of LAN (x).

     

     

    I hope that is clear.

     

    Thanks in advence.

    Ayman


    #DecisionOptimization
    #MathematicalProgramming-General


  • 4.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/10/15 05:00 AM

    Originally posted by: AMurshed


    The only thing I need to know is how to delete all constraints in the model in order to create new constraints of the next model.

     

    Thanks in advance.

     

    Ayman


    #DecisionOptimization
    #MathematicalProgramming-General


  • 5.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/12/15 02:31 AM

    How to delete constraints depends on the API. Which API (C, C++, Java, Python, ...) are you using?


    #DecisionOptimization
    #MathematicalProgramming-General


  • 6.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/12/15 06:04 AM

    Originally posted by: AMurshed


    I am using C++ API.

    I have found in the documentation that there is a command : cplex.clearModel();

    Does this command delete all constraints and variables in the model?


    #DecisionOptimization
    #MathematicalProgramming-General


  • 7.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/20/15 02:55 AM

    clearModel() unextracts the IloModel from the invoking IloCplex instance. That is, it removes the model from IloCplex but does not change the IloModel instance. In order to remove individual constraints from the IloModel instance you have to use IloModel::remove(). Note that if the IloModel is currently extracted to an IloCplex instance then this IloCplex instance will be notified about the change in the model and will adapt dynamically.


    #DecisionOptimization
    #MathematicalProgramming-General


  • 8.  Re: Using model output as input for next model (Incremental Build))

    Posted 03/24/15 06:44 AM

    Originally posted by: AMurshed


    Thanks Daniel,

    I have used clearModel() to delete all constraints in the model and then rebuild the model again.

     


    #DecisionOptimization
    #MathematicalProgramming-General