Simulation Settings Window

 

 

 

The simulation settings window allows you to control how ModelRisk will run a simulation. It is accessed by clicking  on the ModelRisk ribbon. The window has three tabs: Model Settings, which controls the running of the specific model; Macro Settings, which allows one to run macros before and/or after each simulation and/or model sample; and Application Settings which defines the default settings for a number of ModelRisk parameters.

Model Settings tab

Simulation Setup group

Samples determines how many samples will be run for your model. 5000 samples is adequate to produce results of sufficient precision for most problems. If you need a specific level of precision in your results, use the Precision Control feature;

Simulations determines how many simulations will be run, each with the number of samples defined by the Samples dialog. It is more common to run just one simulation but, in conjunction with the VoseSimTable and/or the VoseCurrentSim functions, you can build models that will run several different scenarios together using this feature;
 

Simulation names optionally allows you to give a name to each simulation for easier identification. Either enter a list of names, e.g.

or refer to a spreadsheet range which will copy the list in for you. If used, the name will appear in the Simulation # drop-down control in the Simulation Results window.

Options group

Refresh every … samples will update the Excel screen with the current simulated values every … samples. The option is switched off by default because it will slow down the simulation speed considerably, particularly if a small value (e.g. less than 1% of total samples to be run). If you wish to see numbers change on the screen (which can certainly capture people’s attention, particularly if one has imbedded graphs that change too), consider leaving this switched off for longer simulation runs and using the Update Screen control (the blue button at the bottom of the Simulation Progress control) which allows you to toggle switch screen updating on and off:

Stop on output error is a useful feature to debug your model. ModelRisk will stop simulating and show the scenario that produced the error.

Show results window at end of simulation will automatically pop up the Simulation Results window once ModelRisk has completed a simulation. This should normally be switched on, but you may want to deselect this option if, for example, you are only interested in the mean simulation result displayed by using VoseSimMean.

Seed group

Seed values are used to control how ModelRisk generates random samples for its stochastic variables. They are particularly useful if you wish to reproduce a set of simulation results.

Seed generating: Random will randomly select a different seed value. You can use this option, for example, to run multiple simulations of the same model with different seeds to see if there is any appreciable effect on the simulation results by using different random numbers;

Seed generating: Manual will use the seeds specified in the Seed Value(s) list

Seed Value(s) allows you to enter specific seed values. If you enter a list of values, the first value will be used in the first simulation run, etc. Seed values should be an integer between 0 and 4294967295  (2^32 -1)

Multiple simulation seeds has three options:

All use same seed: all simulations will use the first seed value specified in the Seed Value(s) field. If you use the same number of random variables in each simulation version of your model, this ensures that the difference is not due to random sampling

Use different seeds – First must be defined: the first simulation uses the first seed value specified in the Seed Value(s) field, and the remaining simulations will use different values. This helps evaluate whether the model results are materially affected by the random values being generated

Use different seeds – All must be defined: seed values for each simulation will use the list specified in the Seed Value(s) field. This can be useful if you wish to check the effect of using different seed values on your model, but is not commonly used.

Go To Sample:

Switching ON the Go To Sample feature will allow the user to load any simulation sample into the spreadsheet after the simulation has been performed. Switching this feature on might increase the simulation start time, and if this feature is not required, the user can switch it off to speed up the simulation.
If the feature if turned on, the Go To Sample functionality becomes available in the Data list view of the Results Viewer window.

Warning: The Go To Sample feature will not work properly if the spreadsheet model has any non-ModelRisk volatile functions (like Excel's RAND() function for example)

Save and Load group

You can save simulation settings, and load previously saved settings. The simulation settings file takes a .vmro extension.

Macro Settings Tab

The Macro settings tab controls VBA macros that are run before and after simulation samples and simulation runs. The tab allows specifying 4 places at which the VBA macros can be run:

1. Before each simulation

2. Before each sample

3. After each sample

4. After each simulation

The VBA macros can be chosen from the list of available macros saved in the current workbook:

 

 Application Settings tab

Application settings control some basic ModelRisk default actions.

General group

Remove Vose Software details from reports will produce reports from the Results Viewer without any acknowledgment of the software used.

Save Results determines whether, when one saves a spreadsheet model for which a simulation has been performed, ModelRisk will prompt the user to save a simulation results file or do it automatically.

Simulation Settings group

Samples is the default number of samples that would be used for any newly created model.

The default number of samples is easily over-ridden by typing a different number in the ModelRisk ribbon, or in the Model Settings window.

Stop on output error determines the default setting, which can be over-ridden for an individual model run in the Model Settings window.

Show results window at end of simulation determines whether up the Simulation Results window will automatically pop once ModelRisk has completed a simulation. The setting can be over-ridden for an individual model run in the Model Settings window.

The Global timeout sets the maximum number of seconds for any ModelRisk function to calculate its value. It's useful when one uses calculation-intensive functions and enters parameters that result in a very long calculation loop.

Simulation Results group

Histogram Bars determines how many bars there will be in a histogram plot of simulation results. The Auto option will select the number of bars according to Sturges' rule to reflect the level of detail implied by the number of samples taken in a simulation run. The number of bars can still be changed for individual histogram plots in the Simulation Results window.

Histogram View determines whether the default plotting of histograms should be bars or lines. This can still be changed for individual histogram plots in the Simulation Results window. It is more common to use bars, but lines are better if you habitually plot several variables in the same histogram plot.

Histogram sliders determines the default location for sliders in terms of cumulative probabilities. This can still be changed for individual histogram plots in the Simulation Results window.

Box-plot percentiles determines the default percentiles plotted around the 50th percentile for box-plots. This can still be changed for individual histogram plots in the Simulation Results window.

Trend-plot percentiles determines the default percentiles plotted around the 50th percentile or mean for trend plots. This can still be changed for individual trend plots in the Simulation Results window.

Library group

Allow the user to set the location of the files created by the ModelRisk Library (Industrial edition only).