Learn about excluded attributes in analyses

Beginning with PI AF 2018, excluded attributes perform differently in these types of analyses: Expression, event frame generation, and statistical quality control (SQC). An excluded attribute is an attribute that does not apply to a particular element, and is excluded from an element's attribute list. For more information on excluded attributes, see Excluded attribute property.

Prior to 2018, analyses configured with excluded attributes would return Calc Failed when evaluated and go into error. Beginning with PI AF 2018, such analyses will be suspended and marked by an icon () in the Analyses Viewer. If you hover over the icon, you will see a tooltip with more information.

If you do not want the analyses with excluded attributes to go into a suspended state, you can validate the excluded attribute with the BadVal expression function.

To prevent analyses with excluded attributes from being suspended, use the BadVal expression function to directly validate the excluded attribute. For example, the following expression syntax validates the att1 attribute, allowing the analysis to continue to run:

If BadVal('att1') Then ('att2') Else ('att1' + 'att2')

In the example above, validation is only successful if the exact syntax is used; there cannot be a nested expression within BadVal.

Note: Any change in the configuration of the excluded attribute in the analysis is automatically picked up by PI Analysis Service. For example, if you decide to change the property of 'att1' to be no longer excluded and check it in, the service will detect it and run the analysis with 'att1.'

When you select Event-Triggered as a scheduling option, you are required to select a triggering input. Note that if the excluded attribute is the only available triggering input, the analysis will be suspended whether or not you validate it with BadVal. It is because analyses do not trigger on excluded attributes. Make sure that you have other inputs to trigger on in your analyses. For more information, see Understand analysis scheduling.