Home Forums GENERAL problem in optimization

Viewing 6 reply threads
  • Author
    Posts
    • #27818
      aasif bashir dar
      Participant

      below .osd file shows optimization aborted. I am unable to figure out why.

      Attachments:
    • #27821
      Alessandro Festa
      Participant

      The simulations runs on my computer…can you explain what error do you get?

      Regards,

      Alessandro

      • #27823
        alistu
        Participant

        Hi Alessandro,

        When you want to run the simulation, there is the option “run all optimizations” in the calculation window. When you enable this option and thn start the calculation, the line before the last line where the calculations process is being reported you can read “SPO optimization aborted”. I believe Aasif is referring to this.

        Regards

    • #27830
      Alessandro Festa
      Participant

      Hi Alitsu, thank for you note, I had never used that option! I am trying to run again and see why it does not work.

    • #27833
      Alessandro Festa
      Participant

      Hi Aasif, I understood where the problem is.
      You need to set the Result tolerance to a much lower level than 0.5, since you are using BER (which is usually very low!) as a result target.
      If you try to set the tolerance to 1e-30 you should get results which make more sense.

      Let me know if this works.

      Regards,

      Alessandro

    • #27845
      aasif bashir dar
      Participant

      thanks all;
      Alessandro Festa i will try ur opinion,will let u know if that works for me.

    • #27851
      aasif bashir dar
      Participant

      ur suggestion worked A. festa,thanks
      but what is its significance pls xplain

      • #27861
        Alessandro Festa
        Participant

        Hi Aasif, you are welcome.
        Here is the explanation I think is correct: BER is a value between 0 and 1, so putting the tolerance (which is the difference between the target and the value reached during iterations) at 0.5 does not make too much sense and probably causes the optimization to not converge. A very small value like the 1e-30 that I tried instead allows for the optimization to work.
        Let me know if my explanation is clear enough.

        Regards,

        Alessandro

    • #27863
      aasif bashir dar
      Participant

      yes got it,thanks for ur xplaination

      • #27867
        Alessandro Festa
        Participant

        You are welcome Aasif, happy to have helped!

Viewing 6 reply threads
  • You must be logged in to reply to this topic.