I would like to exclude some anomalous values in my trend. Any suggestions other than exporting to excel and delete the values I don't want in my graph?
Can you give some examples of what you want to exclude? Why do you want to exclude these events? Is your data noisy because of sensor inaccuracy or something else? A solution could be to filter your data using an AF analysis or something similar and then plotting the cleaned data but some more context into your issue can help us give you a better answer.
The device I am collecting data on is a circuit regulator. I'm doing an analysis on the Primary Load Voltage on each phase. I want to factor out any power outage condition (value = 0) from my analysis. I do not have access rights to add or change the AF. Our IT department has everything locked down for read only. In order to make a change, I would need to have a change promoted up through the Development and QA environments. I was hoping for a quick method to do this analysis.
I see. You can specify a filter expression in PI System Explorer for a Time Series Data call. See my example below:
This is just a standard call for all recorded data from yesterday at midnight until today at midnight.
I can then add a filter expression. In this case I used the TagVal function and specified to only return data if the value is above 50.
The analogous functionality also exists in ProcessBook and Datalink.
Yes, I agree and that is what I ended up doing but I really wanted to be able to compare the three phase against each other in a single graph. I am unfamiliar with the Analogous Functionality in Process Book. Can you show me this?
I think you may want to take a look at some documentation on Calculation Data Sets in ProcessBook
Create Data Sets to Visualize Client-Side Calculations
Retrieving data ...