Hello,
I’m working with Rhino 8 and have set up an optimization workflow using Opossum (version 3.1.2). In this workflow, a parameter in my model is changed, and then a hydrodynamic simulation is run and evaluated. Two of the evaluation results serve as objectives for the Opossum component. I’m using the RBFMOpt optimization algorithm. Theoretically, 20 iterations should be performed. However, yesterday I noticed that only 9 iterations were calculated. The tenth iteration was essentially complete, but it wasn’t recorded in the log file, and nothing happened for several hours (not even a background calculation—I checked the Task Manager). No errors occurred during the calculation and evaluation of the simulation, and I was also able to view the results manually in the Grasshopper canvas. There was also no message indicating that Opossum was finished, and under “Run Optimization,” the only option was to stop. However, the process didn’t stop even after I clicked “Stop.” I had to completely close Rhino to terminate Opossum. So it looked to me as if Opossum had frozen for some reason. Does anyone have any idea what might be causing this?
I started a new run overnight and have completed 16 iterations so far; everything is working as it should at the moment. However, my simulations are currently running on a very coarse mesh and take about 30 minutes, so having to restart is annoying but not quite as critical. Once the process is complete, I’d like to start a new run with a finer resolution, and those simulations take about 14 hours. A stable workflow would be very helpful there.
Is there a way to feed previously calculated results into the Opossum component? Yesterday, after the process stalled after 9 iterations, I had to start over from the beginning. Theoretically, though, I already had nine results and could have saved time on the second attempt if those results could somehow be incorporated into the calculation.
Thanks for your help
Best,
Anne