Pivot Grid results: more than 1 avg, min, max?

[mmashaie]

Hello,

When I run an model and filter on the results in the pivot grid I want to see, I end up with more than 1 value for average, minimum, and maximum values. Based on doing some traces while the model is running, I am sure I have identified the correct average, min, and max values (circled in attached figure), but why do have all the extra columns?

Thanks in advance.

[mmashaie]

I think I may have figured it out with my colleague's help. I failed to mention that I was running an experiment, replicated 10 times, so I believe the extra max and min columns are the infimums (min of mins) and supremums (max of maxs) from all the 10 replicationed runs.

Does this make sense?

[ASagan]

That is correct.

You have the average, min, max from within a run. And the average of the average, the average of the min, the average of the max... the min of the min/average/max, and the max of the min/average/max.

Looking at your results, I see that the average time in station of ScreeningLineL1.Xray.Processing ranged from 12.6974 to 13.3549 seconds, while the min ranged from 0.1229 to 0.8086 seconds, and so on and so forth.

[dsturrock]

The later posts are both correct. You must realize that stochastic simulation does not give an exact answer but rather gives you an estimate of the answer. Hence it would be inadequate to say

The simulation predicts that the average ScreeningLineL1.Xray.Processing time will be 13.0725 seconds.

If you have run this with default settings a better interpretation would be:

I am 95% confident that the true mean ScreeningLineL1.Xray.Processing time is between 12.6974 to 13.3549 seconds.

If you create a response for this item, you can also see that information (and lots more) graphically in the Response Results tab.

If the range is too large, you can generally reduce the range (create a tighter prediction) by running additional replications.