Hello!
I have a pretty complex hierarchy and I want to eliminate some attributes in the hierarchy which reads bad values but maintain the shape of the AF hierarchy. Is it possible with AF transformer? the examples I see are all transforming the existing hierarchy or creating a new template which I do not want to do.
Yes. You can place filters at the Task, Element and Attribute level to screen items out of the SearchShape which you don't want carried forward into the OutputShape, including screening out bad/garbage attribute values.
My AF hierarchy is template based and I am having a hard time duplicating the template into the AF transformer. Each element has a template and I need to drill down and eliminate certain attributes mapped to a particular pi point.
Is discardfilter the only way to eliminate garbage attributes?
thank you. I am trying to discard attributes whose value is a "NOData" PI Point. I am unable to delete using PI Builder as my AF hierarchy is template based and I can't seem to delete anything.
By design you cannot delete attributes that were instantiated from template. Afterall, that the whole idea of a template. We do give you a way to exclude the attribute that you don't need. Have you tried that?
--
Steve Kwan
Hi Stephen,
My AF hierarchy is completely template based. i tried discard filters and it didnt work as the AF hierarchy is template based.
You can solve this by running AF Transform twice. Use the output DB from the first run as the input for your second run (I think you can also just add a new <Task> and define additional <SearchShapes> and <OutputElementShapes> tied into that task.
Essentially you can just screen out the null elements in the second pass with a Task-level data filter, like this...
<Task Id="1000">
<SearchShapes>
<SearchShape Id="1003" ReaderId="1004" FilterMatchType="Any" >
<DataFilters>
<Filter Expression="[10.value]==null" />
</DataFilters>
</SearchShape>
</SearchShapes>
<OutputElementShapes>
<OutputElement Id="1005" />
</OutputElementShapes>
<TaskWriters>
<TaskWriter Id="1006" />
</TaskWriters>
</Task>
You can solve this by running AF Transform twice. Use the output DB from the first run as the input for your second run (I think you can also just add a new <Task> and define additional <SearchShapes> and <OutputElementShapes> tied into that task.
Essentially you can just screen out the null elements in the second pass with a Task-level data filter, like this...
<Task Id="1000">
<SearchShapes>
<SearchShape Id="1003" ReaderId="1004" FilterMatchType="Any" >
<DataFilters>
<Filter Expression="[10.value]==null" />
</DataFilters>
</SearchShape>
</SearchShapes>
<OutputElementShapes>
<OutputElement Id="1005" />
</OutputElementShapes>
<TaskWriters>
<TaskWriter Id="1006" />
</TaskWriters>
</Task>