Application Programmers Manual: Variables and Outcomes

From DSL

Every node in a graphical probabilistic model represents a variable. Variables can assume values, also referred to as outcomes. Discrete variables have a finite number of outcomes, continuous variables can assume any of an infinite number of outcomes from a continuous interval.

An Example:

A discrete node Disease may take two values (and have effectively two possible outcomes): present and absent. A discrete node Test can have three outcomes: positive, negative, and neutral.

Node type, its outcomes, and their names can be defined by the programmer using several SMILE functions. Check the description of DSL_nodeDefinition class for more information.