Experimenter results different from 3D simulation

Question

question

Lucas Klein asked May 21, '18 Jordan Johnson edited May 21, '18

Experimenter results different from 3D simulation

In the Experimenter tool, it is possible to repeat the stream of certain replication. A customer is using this tool to debug his model and the behavior of new logics in the model. He contacted us reporting that the results obtained in some performance measures are discrepant in relation to the 3D model when he repeats the stream.

I.E.: He informed that the performance measure X presented a value of 0(zero) in certain replications. Assuming that the new logic added to the model had some error, he repeated the stream and ran the 3D model to observate and to try to discover the source of the error, but at the end of the simulation, the logic worked fine and the measurement was correct.

There is any issue with this tool of repeating the stream, that the 3D model runs different from the Experimenter replication? Any sort of variable or code that the Experimenter have trouble in calculate or set?

Thanks in advance.

Software Version:

FlexSim 18.0.4

experimenter 3d model replication stream repeat streams

Answer 1 · 2018-05-21T19:48:04Z

Jordan Johnson answered May 21, '18 Jordan Johnson edited May 21, '18

Check out this article on model repeatability:

https://answers.flexsim.com/articles/19380/model-repeatability-14.html

Especially be sure that the draw code of an object doesn't change model state, or use random numbers. The Experimenter does not run draw code, as the replications run without a view.

question