Jump to content

Fatma

Members
  • Content Count

    30
  • Joined

  • Last visited

1 Follower

About Fatma

  • Rank
    Advanced Member

Recent Profile Visitors

1200 profile views
  1. I would first check if your input files are written correctly. You can do this by comparing them with each other and making sure that the DV values are changing. If they are written correctly, next is to check if the simulation is running correctly. Are you pointing directly to the executable or are you running a script? If you are running a script, there may be places that you had to use HyperStudy environment variables. Hope these help as a starting point
  2. In HyperStudy terminology, input and validation matrices correspond to training and testing data. A validation matrix is not obligatory but it is suggested as the fit diagnostics on input matrix is not as reliable. In cases where methods such as HyperKriging and RBF are used; it is very much needed. More information can be found in HyperStudy > How to Use HyperStudy > Study Contents > Approaches > Approach Steps: Select Matrices (Fit) As for your second question; I am afraid I am not clear about it. I read specific instructions in the tutorial for the fit method Can you give more details?
  3. In online help, under "Invoking HyperStudy" section batch commands are listed for different OS. As for picking what to run in batch; in the UI; if you click on the study name; you will see that one of the tabs is called "Batch Tasks". In this tab; you can pick which approaches you'd like to run batch. Hope this helps.
  4. Which HyperWorks version are you using? Thanks.
  5. I put an e before the sign before parameterizing the value. Make sure to adjust the columns by deleting a space after the value if you choose to do this workaround.
  6. What version of HyperStudy are you using? If it is older than HyperStudy ver 14.0; you need to use the resvector function. See HS-2000: DOE Method Comparison: Arm Model Study as an example.
  7. Please see HyperStudy tutorial HS-1040.
  8. Hello Do you still run into this issue? Can you try a newer version of HyperStudy than 12? Thanks.
  9. Hi Tanut Details of GRSM can be found in HyperStudy help. GRSM would work well with noisy responses as it is global search algorithm and good in capturing the global trends of the response functions (as opposed to being trapped in local minimums/maximums). Hope this helps. Thanks Fatma
  10. Hi Stefan If you can create responses that gets the minimum element index and maximum element index, you can use these responses in your resvector definition such as max(resvector(getenv("HST_APPROACH_RUN_PATH") + "/m_1/beam.h3d",1,0,r_6,0,0,0,1)) In this example, minimum element index is always 0 but maximum element index is another response (r_6). This response, r_6, in this case, reads the number of elements from the OptiStruct output file. Hope this helps Fatma
  11. Which files are you reading your responses from? If they are .mer files, please use the "ASCII Extracts" tab in the Expression Builder.
  12. You can use the "parametrized input file" model in HyperStudy instead of the "HyperMesh" model.
  13. You can add any responses to HyperStudy as long as the solver (AcuSolve in this case) writes out what is needed to calculate that response.
×
×
  • Create New...