When carrying out probabilistic simulations, you may often run hundreds or thousands of realizations. In order to analyze the results, it is often quite useful to classify the realizations into categories. A category is simply defined by a condition relating one or more outputs in the model (e.g., those realizations in which the discount rate was above 3.5%; those realizations in which the profit exceeded $1,000,000; those realizations in which the peak concentration was between 1 mg/l and 10 mg/l).
The power of categories is that once you have defined them, you can use them in two ways:
•Within certain kinds of result charts (e.g., scatter plots, time histories, distributions), realizations from each category can be displayed in a different color (and/or symbol).
•Within all result displays, you can choose to screen out one or more categories, so that the results that are shown (in charts and tables) only include those realizations in the categories which you have chosen to include.
Classification categories are defined at the bottom of the Monte Carlo Result Display Properties dialog:
This dialog can be accessed from the Monte Carlo tab of the Simulation Settings dialog (via the Result Options… button), and is also accessible from result properties dialogs.
By default, all results are placed in a single category (defined as “All realizations”). You can add categories by pressing the Add button. When you do so, it adds a row above the selected row. For each category that you add, you must define a Label and a Condition.
In this example, three categories have been defined:
It is critical to understand how the categories of realizations are created by GoldSim. In particular, the order of the Classification Conditions is important. If multiple Classification Conditions are true for a realization, the realization is assigned to the category with the first True Condition in the list. Hence, in the example shown above:
•The “Low” category consists of all realizations in which X < 0.2. As indicated by the Net % column, 11% of the realizations fall into this category.
•The “Medium” category consists of all realizations in which X >= 0.2 (and hence they do not fall into the “Low” category above it) AND X < 0.6. As indicated by the Net % column, 46% of the realizations fall into this category.
•The “High” category consists of all remaining realizations (all realizations in which the conditions in the first two categories are false). In this case, it implies that the category consists of all realizations in which X >= 0.6. As indicated by the Net % column, 43% of the realizations fall into this category.
As pointed out above, the Net % columns indicates the percentage of realizations falling into each category. The Gross % column shows the percentage of realizations falling into the selected category and all the categories above it.
Note: If you place your
cursor over the Net % field, a tool-tip will display the actual
realizations that fall into that category; if you place your cursor over the
Gross % field, a tool-tip will display the actual realizations that fall
into that category and all the categories above it.
Note: The Gross % and
Net % are only shown when the model is in Result Mode (or when the model is
paused).
You can use the Move Up and Move Down buttons to change the order of the Conditions (and hence the definition of the categories).
Several points should be noted regarding defining the categories:
•You cannot edit the Condition for (or move) the final category. It always represents “everything else”. That is, it represents all realizations that do not fall into one of the previous categories.
•The Condition can be a complex expression, but it must be a scalar condition.
•You can edit the Conditions in Edit Mode, Result Mode or Run Mode (but not Scenario Mode). However, in Result Mode or Run Mode, you cannot reference an output for which Final Values have not been saved.
Categories are powerful because in Result Mode, within certain kinds of result charts (e.g., scatter plots, time histories, distributions), realizations from each category can be displayed in a different color (and/or symbol). For example, this Multi-Variate Result (a 2D scatter plot) breaks down the results by category:
Note that for each category, you can select a Style. This is used to control things such as colore when plotting the results when you have created multiple categories.
The manner in which these Styles are used is a function of the type of result being plotted.
In addition, you can choose to screen out one or more categories of realizations by clearing the Include checkbox for a category:
When you do so, when showing results (in both charts and tables), only those realizations in the categories which you have chosen to include are displayed.
Note: You can edit the
Include box in Edit Mode, Result Mode Scenario Mode or Run Mode.
Obviously, if you have screened out results in such a way, it is critical that this is made apparent that the results being shown are screened. Hence, when results are screened, this is indicated in the right-hand corner of the status bar. A hashtag (and a comment) are added after “Results”:
Learn more about:
Using Result Classification and Screening in Time History Results
Using Result Classification and Screening in Distribution Results
Using Result Classification and Screening in Final Value Results
Using Result Classification and Screening in Multi-Variate Results