Decision Optimization

 View Only

CPLEX Error 1004: Null Pointer for required Data

  • 1.  CPLEX Error 1004: Null Pointer for required Data

    Posted Wed January 13, 2021 11:45 AM
    Hello,
    I am a graduate student trying to handle a large optimization for my dissertation, first time posting and an inexperienced CPLEX user.
    I am using the CPLEX Matlab connector to solve a LP on my university's computing cluster and received the error " CPLEX Error 1004: Null Pointer for required Data" and was hoping someone here could help me out.
    I was able to successfully run a scaled down version of my full problem locally on my computer and was able to replicate the results on the computing cluster without any errors. However, when I scale it up I receive the error. The details of the problem that is causing the error are:
    - 42,881 constraints
    - 283,649 variables
    - The inequality matrix is a sparse matrix in Matlab to conserve memory (roughly 90 gigs of data)
    - There are no equality constraints
    - There are upper and lower bounds on the variables
    - I am using CPLEX version 12.9
    - At this time I have not changed any of the LP Options from their default values
    - The core on the computing cluster has 512 gigs of memory dedicated to the problem
    - I am unsure on how many threads to request, however the error occurred both when utilizing 1 and 4 threads
    - The optimization is set up to always find a feasible solution by design
    - I submit my job to the computing cluster using a batch scheduler called SLURM

    Any guidance would be very much appreciated. If there is any additional information I could provide that would help clarify the issue I will do my best to provide it. I know Matlab isn't the best way to approach a problem of this size but unfortunately I discovered this fact after I had already written all the code. Any recommendations on which Options to set using 'cplexoptimset' or how to troubleshoot the issue would also be greatly appreciated.
    - Thank your time and any assistance you can provide,
    Chris

    ------------------------------
    Christopher Horstman
    ------------------------------

    #DecisionOptimization