Basically, the answer is to use dashboards for everything. Dashboards have become a lot more powerful. They are good for not only dynamically displaying results during a single model run for a presentation, they are also fantastic for multi-replication, multi-scenario model analysis. You can create Sample Sets within a dashboard widget that filter by scenario and/or replication of an experiment run. The Box Plot is a superb option when running an experiment, because it can provide: Min,Mean, Max, 25th Percentile, Median, 75th Percentile, and the Confidence Interval for 90, 95 or 99%, AND you get to view separate box plots for multiple periods throughout the day within the same chart window!
With the previous Experimenter, you were forced to define Performance Measures (PFMs) upfront before the simulation run, but now you can create dashboards after the experiment is finished and view the results from one or more of the scenarios and/or individual replication runs of the experiment! In addition, PFMs previously had to be defined by a single value reported at the end of each replication run, but now you can view time-based results throughout a run and across the entire experiment. Pretty cool, right? Good luck, and if you have any problems or questions figuring it out, please let us know.
|
