A node is a graphical representation of a decision, value, uncertainty, strategy table, or controlled event. There are six types of nodes available in DPL: Value Nodes, Discrete Chance Nodes, Continuous Chance Nodes, Decision Nodes, Strategy Nodes, and Controlled Nodes.
A value node can be a scalar, one dimensional array, series or two dimensional array. If a value node is a scalar, it represents a single number or an expression. Arrays and series represent multiple number or expressions or a combination of both. Value nodes appear as blue rounded rectangles in the influence diagram. Value nodes do not appear directly in the decision tree though they may be used in get/pay expressions in the tree.
Discrete Chance Nodes
A discrete chance node represents an event whose outcomes are uncertain. The probability distribution for a discrete chance node is specified by specifying probabilities or a named distribution for a discrete set of outcomes. A discrete chance node will always be in one of these states. Discrete chance nodes appear as green ovals in the Influence Diagram or green circles with its number of outcomes as branches in the Decision Tree .
Continuous Chance Nodes
A continuous chance node also represents an event whose outcomes are uncertain. The probability distribution for a continuous chance node is specified by specifying the parameters for a named distribution. Continuous chance nodes are used in Monte Carlo simulation. The values that a continuous chance node takes in an analysis depends on the number of Monte Carlo samples. Continuous chance nodes appear as dark green ovals in the Influence Diagram or dark green circles with a semicircle and gray "fan" in the Decision Tree.
A decision node represents an event with alternatives. During a DPL decision analysis, one alternative is determined to be optimal. Decision nodes appear as yellow rectangles in the Influence Diagram or yellow squares with its number of alternatives as branches in the Decision Tree.
A Strategy Node contains a Strategy Table, which is a collection of decision nodes and a set of defined strategies. During evaluation, one strategy is selected as optimal. Strategy nodes appear as large divided yellow rectangles in the Influence Diagram or yellow squares in the Decision Tree.
A controlled node represents an event whose state is set as an action on a branch in a decision tree Controlled nodes appear as gray rectangles in the Influence Diagram or gray squares in the Decision Tree with the branch (state) it's controlled to bolded.
Versions: DPL Professional, DPL Enterprise, DPL Portfolio