Transform graphs and nodes

Transform graphs comprise a collection of graph nodes contained as properties within the parent transform graph component. The default view of the transform graph component is the Wire Sheet view, where you work with the various nodes to create your report.

Figure 10. Series transform graph Wire Sheet view and source histories

Typically, you assemble transform graphs by interconnecting graph nodes in the Wire Sheet view of the parent transform graph node, with source nodes and other nodes integrating into a final terminal node. On the Wire Sheet, each graph node connects into the graph by linking the output property of the component to the input property of the target component. Niagara histories provide the most likely source of tabular data. Each graph produces a single table result, as represented by a single terminal node.

As a third-party developer, you may create additional nodes using the series transform API. However, Niagara4 provides seven individual objects on the seriesTransform palette. These components provide four general categories of series transform objects:

The Container (TransformGraph node) holds all the other graph nodes and has a ResolveGraph action that you use to generate the data transformation after you configure all the nodes.
The Source (HistorySource node) is the single object that identifies the source data for your transformation.
The Terminal (Terminal node) is the final node of any transform graph. Each transform graph must have one and only one terminal node.
The Transforms (Transform nodes) provided on the palette include the Aggregate, Composite, Rollup, Scale, TimeShift, Filter and Cleanser nodes.
Figure 11. Series Transform Graph palette

By adding components from the seriesTransform palette to a Transform Graph Wire Sheet view, you can create a static definition of how to transform run-time data into a new single-output set of information. You create the transform graph using HistorySource nodes, one or more transform nodes, and a single terminal node. By configuring the data schema of each transformation you design how the data flow from one transformation node to the next.

NOTE: A transform graph container must contain all nodes and all nodes within the transform graph container must be uniquely named.

Figure 12. Minimum set of nodes on the seriesTransformGraph Wire Sheet view

Each of the graph nodes described in the following sections includes a special popup editor that opens when you click the graph node component in the Wire Sheet view. To access the Property Sheet view, select it from the Workbench View Selector.