Process Flow Diagram Logic |
Here are some general rules that govern the placement of nodes in a process flow diagram:
The Input Data Source node cannot be preceded by any other nodes.
All nodes except the Input Data Source and SAS Code nodes must be preceded by a node that exports a data set.
The SAS Code node can be defined in any stage of the process flow diagram. It does not require an input data set that is defined in the Input Data Source node. If you create a SAS data set with code you write in the SAS Code node (for example, a data set that contains target and predict variables), you can use a successor Data Set Attributes node to assign model roles to the variables in the SAS data set.
The Assessment node must be preceded by one or more modeling nodes.
The Score node must be preceded by a node that produces score code. For example, the modeling nodes produce score code.
The Data Set Attributes can have only one node preceding it.
The Ensemble node must be preceded by a modeling node.
The Reporter node generates HTML reports for all of the predecessor nodes. It does not generate reports for the successor nodes. When you run the flow from the Reporter node, Enterprise Miner makes a first pass through the flow to analyze the data. After the analysis is complete, Enterprise Miner makes a second pass to generate the HTML reports. The node icons are colored green during the analysis pass and yellow during the reporting pass.
You can have one Group Processing node per process flow. Group processing occurs when you run the process flow path. The Group Processing node can be connected to any successor node, but the only nodes that accumulate results from each pass are the modeling nodes and the Score node.
Copyright © 2006 by SAS Institute Inc., Cary, NC, USA. All rights reserved.