Home → Troubleshooting → @RISK with Projects → Continuous Output Data Graphed as Discrete
Applies to: @RISK 6.x/7.x, Professional and Industrial Editions
In @RISK with a project, the histograms for some outputs are displayed as Discrete Probability, and others as Probability Density. (In Application Settings I have preferred distribution format set to Automatic.)
The Discrete probability graph looks very strange at first sight, and whereas it can easily be changed, I was wondering why this behavior occurs —I have not seen this happening with @RISK in pure Excel models.
Project data have many date values, which means it's likely that many iterations will come up with duplicate values. But when the preferred distribution format is automatic, @RISK looks at the number of duplicates in each distribution to decide whether to graph it as continuous or discrete. If there are no duplicate values in the distribution, or not too many, @RISK graphs it as continuous. If there are too many duplicate values, @RISK graphs the distribution as discrete.
If @RISK doesn't do what you want for a particular distribution, you can tell @RISK how to treat it by using the RiskIsDiscrete( ) property function. RiskIsDiscrete(TRUE) tells @RISK to treat the distribution as discrete, and RiskIsDiscrete(FALSE) tells @RISK to treat it as continuous.
For output distributions, click on the desired cell. Then click Add Output, enter a name if you wish, and click the fx icon. You can then set data type to Automatic, Discrete, or Continuous. You can set the data type in this way when you first define an output or if you have already defined the cell as an output.
If you prefer, you can add a RiskIsDiscrete( ) function manually as the fourth argument:
=RiskOutput("MyOutput",,,RiskIsDiscrete(TRUE)) + ...
or
=RiskOutput("MyOutput",,,RiskIsDiscrete(TRUE)) + ...
You can also apply the RiskIsDiscrete( ) property function to inputs, as in this example:
=RiskResample(1,B1:B100, RiskIsDiscrete(FALSE))
@RISK will graph this function as continuous, regardless of duplicate values or the lack of them.
Last edited: 2015-10-15