A Data Splitter solution contains one or more PGraphs. The first PGraph is referred to as the "main graph". Subsequent PGraphs are referred to as "node groups".
A node group is defined as a PGraph and used as a Pattern. A node group is created by adding a new entry to the node group list (Definitions | Node groups). When the new entry is selected a clear drawing surface appears, at which point a new PGraph can be defined. It has the same properties as the initial (main) PGraph, including its own start node. When an existing entry is selected that node group's graphical definition appears on the drawing surface.
If the pattern associated with a node is a node group, the node group is transitioned into if its start node is recognized in the input stream. PGraph traversal continues within the node group until a "Return" action is performed.
A node within a node group can refer to another node group, including its own node group. This allows recursion and processing of "nested" patterns.
Node groups are selected for viewing and modification via menu option Definitions | Node groups. The main graph is returned to via menu option View | Main graph.