TSDataFlowEvent produced by a transform

Hi,
I need to implement a dfe produced by a transform but I don’t find documentation about it.

Can someone explain ho to configure the dfe (for example, what to include in the annotation) to use it?

Thanks,
Laura

What do you mean “produced by a transform”, do you mean that when data is being loaded (e.g. through a Canonical) and then transformed and finally persisted to DB you want to trigger a DFE?

Normally you would think about this a little differently. At some point an Analytic is going to be executed, and your process() function would be called. That process function will require some data (normally metrics). You should configure your dfe with those metrics, and then any time data is persisted which affects any of those metrics (whether through a transform or otherwise) the DFE will triggern and your Analytic will run.

If you can describe a bit your use case i could try to help you design your DFE a bit.

I think she’s talking about the TSDataFlowEvent type documentation, which states:

A DFE may be constructed in various ways, but the most common type is loaded from a system metric SimpleMetic/CompoundMetric or produced by a transform.

(emphasis mine)

Yes @matt , I’m talking about it. If possible to define in the annotation @DFE , instead of for example
@DFE(period=‘QUARTER_HOUR’, grain=‘QUARTER_HOUR’, metric=‘METRIC’)

a transform ? In this case I suppose that the dfe is triggered when the transform is “called” and the process function receives as input the raw data that will be stored in the DB instead of the normalized value returned by the metric, correct?

I would like to verify the real data that the system receives. A problem that I have using metric is that : if I upload a measure, the system can re-generate the timeSeries and I can have more than one “dfe trigger”, that is not what I want. I want for one measure / upload --> one trigger. If the canonical file uploaded contains more than one rows, I want verify each rows.

tranform in doc is referring to An analytic process output that would have transformed the input dfes into output. It is not canonical data transform.