Can indexed variables have probability distributions associated with them?
For example, I have an index called CITIES. And I want to define a variable called RAINFALL, indexed over CITIES, such that the variable instance has a different probability distribution for each city in CITIES. For Example, Monterey may have a uniform distribution between 7 and 15 inches of rain, and Philadelphia may have some triangular distribution. Given an index variable called CITIES, you can create an uncertain variable, RAINFALL, as a Table indexed by CITIES. (Select the Table option from the ‘Expr’ popup menu for the definition of RAINFALL. In the index chooser, select CITIES as the index.) Then you can type in any distribution, e.g. Normal(10, 2) or Triangular(0, 4, 6), into each cell of the table. The only limitation is that the Distribution object finder is not available to help you do this. If you want to specify a discrete probability mass function for each city, you can select Probability Table for RAINFALL, and create a table indexed by CITIES and RAINFALL. You create a
