New in the Community? Get started here

Schneider Electric Exchange Community

Discuss and solve problems in energy management and automation. Join conversations and share insights on products and solutions. Co-innovate and collaborate with a global network of peers.

Register Now
Resource Advisor - Performance Analytics Forum
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

How to use RA metrics in calculated measurements right?


Hi, I got this problem with calculated measurements using metrics from RA:

  •      Calculation is very slow
  •      An error occurs rendering charts if I choose anything except "Last x month" not including current month in the timeline options.

Hence my questions:

1. Is there a best practice using advanced directives in the calculation engine in order to improve calculation speed?

2. Should I use some disaggregation method in order to make it availabe in any option in the timeline?

3. If I use SUM PER MONTH on interval data, does it crashes the calculation on another aggregation option than monthly?

Thanks!

4 REPLIES 4
Commander

Re: How to use RA metrics in calculated measurements right?

Hi Etienne, can you share some details of what you're trying to do that's crashing or slowing the system?

Some RA metrics can be disaggregated to work at the hourly or interval level, but in many cases this is "possible but not recommended".  For example, if you have monthly units of production, then disaggregating to show energy/unit at the hourly level isn't very helpful, because it assumes that the factory is producing at the same rate, 24 hours a day, 7 days a week.  This is very unlikely.

One tool we use a lot to share short software demos is called Jing.  It's free, and allows you to take short videos that you can then upload to the cloud and share links.  That would be a great way to show us what you're trying to do, if screenshots are too hard to explain.

Michael Schmitz
Highlighted
Lieutenant

Re: How to use RA metrics in calculated measurements right?

Hi Etienne,

As Mike said, some further details would be helpful.  However, I tend to use disaggregation command for everything I pull into PAM. It just makes things look better as you move around different timeframes.  The trick is to use the disaggregate value appropriately (either constant, or linearinterpolate).

If you use SUM PER MONTH, it shouldnt crash, but rather just not show anything - unless you allow disaggregation.

For an error to occur (as opposed to data just not showing), that sounds to me like a bug.  I see things like that on a reasonably regular basis. If you ticket it (ideally using JING as Mike suggested), it usually gets resolved quite quickly.

Lastly, in terms of calculation best practice for speed, thats a good question! My advice would be to try to simplify any equations as much as possible, but outside of that, I too would be interested to know any general advice. eg, one I've always wondered is if its any slower having an intermediate calculated measurement and we should be building everything into the one expression.

Re: How to use RA metrics in calculated measurements right?

Hello,

It seems to have been corrected, maybe in the March release. I cannot replicate the error today.

The calculation I'm speaking about is the one detailed in the normalized consumption walkthrough #PAM Calculating Weather Normalized Consumption in Resource Advisor

There is an IF statement and several RA metrics in it. I just simplified it removing all SUM PER MONTH statements and adding a disaggregation with linear interpolation. It doesn't look to be faster but now the daily values will mean something.

Lieutenant

Re: How to use RA metrics in calculated measurements right?

Hi Ettienne, I know there was an IF statement bug where it wasnt working, which was fixed in the release, so maybe that had something to do with it?

I do think the SUM PER MONTHS were superfluous as the RA metric were coming in monthly anyway, but I doubt it would have slowed the visualisation down much anyway.  If your baseline was on monthly data, but us were receiving PAM interval data, then they would definitely be needed.

Otherwise, I dont think there is much you can do to optimise your expression really.  I was trying to think if there is any way to remove the IF statement, but came up with nothing.