Series Transform (Niagara Only)

Periscope has the ability to use trends created by the Series Transform feature of Niagara 4.  This is a powerful feature that allows a user to perform calculations on some input histories.  This includes things such as combining or scaling histories as well as data cleansing.  

Series Transforms are an advanced feature and require configuration on the Niagara back-end.

Using a Series Transform Graph

In order to to use a Series Transform graph within Periscope a special "seriesTransform" record must be created.  This record tells Periscope which histories are being input into the Series Transform graph and where to find them.  The following table shows the tags required:

Haystack TagNiagara TypeDescription
hs:hisMarkerThe his tag is used to define a historized point
hs:seriesTransformBOrdThe BOrd to the Transform Graph component
hs:sourceName#StringThe name of the History Source within the Transform Graph. The '#' allows configuration of multiple history sources. This number should start at "1" and continue up to the n number of sources.
hs:sourceOrd#BOrd

The BOrd to the history that will be plugged into the transform graph.

hs:unitStringThe unit facet (the value facet used in the transform graph is not used with Periscope).

Example

Here we have a simple TransformGraph located at station:|slot:/PeriscopeTechDemo/TransformGraph. This graph contains one History Source Node simply called "HistorySource" which goes through a Scale Node and finishes at the Terminal.  In order to apply this transformation and make the result usable in Periscope we need to setup a record with the appropriate tags.

This screenshot shows the correctly applied tags so that we can use our Series Transform graph within Periscope.  Note that we only have one History Source input in our Transform Graph so our "sourceName" and "sourceOrd" end in 1.  If we had a 2nd history source then we'd have a sourceName1, sourceOrd1, sourceName2, sourceOrd2.

 In this case we're using a BNumericWritable as our component type but any BComponent can be used.  The benefit of using a BNumericWritable is that we can easily apply a "unit" facet which is then passed to Periscope.


Series Transform histories are calculated on the fly and should be used cautiously to avoid overloading the server.