• 338 views | 5 messages Discussion: LEAP
    Topic: Optimization failed: GLPSOL: GLPK LP/MIP Solver,Subscribe | Previous | Next
  • Laura Zordeh 4/25/2023

    Like

    Hello Community,

    Please i need assistance to solve this error. since i updated my leap software to the latest version, i have been unable to run my leap model which was previously working well. I have added a print screen of the error message as well as the Opdebug message below:


    GLPSOL: GLPK LP/MIP Solver, v4.65
    Parameter(s) specified in the command line:
    --mipgap 0.01 --model OSeMOSYS.TXT --data OpData2.txt --output SelectedResults2.txt
    Reading model section from OSeMOSYS.TXT ...
    498 lines were read
    Reading data section from OpData2.txt...
    OpData2.txt:97097: REMinProductionTarget must have 2 subscripts, not 3
    Context: ...aram REMinProductionTarget default 0.000 := [ * , * , 2020 ]
    MathProg model processing error

    I would be grateful if you could assist me.

    Thank you.
    leap error.png
  • Jason Veysey 4/26/2023
      Best Response

    Like

    Hi, Laura -

    I'm not sure why that OSeMOSYS error is occurring, but perhaps one of my colleagues can elucidate it.

    However, I suspect you'd have a better experience with optimization if you used NEMO instead of OSeMOSYS. NEMO is a high-performance, open-source optimization tool that we develop here at SEI specifically to integrate with LEAP. You can download an installation file for NEMO through the LEAP website's Download page - https://leap.sei.org/default.asp?action=download.

    Once you have NEMO set up, try changing Transformation\Electricity Supply:Optimize in your model to NEMO(HiGHS). This will use NEMO and the HiGHS solver to calculate the model. HiGHS is an open-source solver that performs much better than GLPK, the solver you've been using.

    Regards,

    Jason
  • Amadu Mamuda 5/19/2023
      Best Response

    Like

    hello there,
    I am currently having this same problem. Has a solution been found?

    Best,
    Amadu
  • Pravesh Raghoo 7/9/2023
      Best Response

    Like
    1 Like

    I am having the same problem. I need to use OSeMOSYS - it is from previous work that I need to rectify.

    How do I solve this problem?

    Pravesh
  • Charlie Heaps 7/10/2023
      Best Response

    Like

    Hi Pravesh,

    Can you try using NEMO instead of OSeMOSYS? We are gradually phasing out support for OSeMOSYS now, so I'm afraid this may not get fixed quickly. NEMO has much better support for flexible time-slicing and is required for modeling storage. It also supports a a much wider range of solvers including fast open-source solvers like Cbc and HIGHS. You can get NEMO here.

    Thanks!

    Charlie