Hybrid Source Scenarios - Additional Documentation / Details

[ajslack]

One of the new features in the 6.6 release is the introduction of Hybrid Source Scenarios.  I've watched the 6.6 release and read through the abbreviated examples in the admin guide. Both have good high level content, but they seem to be missing details around the Options section within the Hybrid Source Data of Scenario properties.

Does anyone have any additional details around what options are available and a description of each? Or more content around Hybrid Source scenarios in general?

Thanks,

Added screenshot to provide context. 

[EricOsmanski · Accepted solution]

Hello - I am going to ask that more documentation around the Options are added to the documentation. Below is more detail on the Options:

• ExecuteCopyDuringCalc - this is a True/False setting and by default it is set to True. This would only be used with the 2 Copy binding types and not the Share binding type. Having this setting as True indicates that the copy will execute during the Data Unit Calculation Sequence (DUCS). Or in other words, anytime you calculate a Data Unit the copy will execute. Making this setting False, you would need to use a Data Management Job to execute the Hybrid Source Data Copy. On a DM Step, there is a new setting called "Execute Scenario Hybrid Source Data Copy". You would make this True and then the Calculation defined on this step would use the settings from the Hybrid Source Data properties to run the Copy. This would NOT run during the DUCS and be a "point in time" calculation. Use cases for this may be versioning or seeding where constant copying of the same numbers would be unnecessary overhead on every calculation.
• CopyAsDurableData - this is a True/False setting and by default it is set to False. This would only be used with the 2 Copy binding types and not the Share binding type. By default, copied data is stored as calculated. Set this to true to flag this data as durable. Calculated data is cleared by OS at the beginning of the DUCS (if the scenario setting is set to True) and will re-calculate. Durable data is NOT cleared by OS at the beginning of the DUCS and must be cleared via a rule. These 2 settings go hand in hand. Likely, if you have ExecuteCopyDuringCalc as True, CopyAsDurableData would be False. And vice versa.
• RemoveZeros - this will remove zeros for the copied data. This is only if all periods of the intersection being copied are 0.
• SourceTimePeriodForYearlyResult - this is used when the Input Frequency of the target scenario is Yearly and the source scenario is an Input Frequency that is more granular. It is used for performance reasons to only share or copy one time period to the target scenario which limits the Data Unit size and will improve calculation and/or query performance. 
• SourceViewForYearlyResult - use in conjunction with the setting above, this can either be Period/YTD. For the time period being copied/shared you can choose whether you want to copy/share the Periodic or YTD value.

[ChuckUsher]

Hey Eric - thanks for the above, and additional questions: 

I was experimenting with Hybrid Copies, and I'd like to only pull over some data from upper level Entities from some Actual months into a Forecast. 

In regard to the Entity dimension, it only seemed to pull over data when I had Base Entities as part of my calculation filter.  When I tried to pull upper levels alone it would not bring over the data (this is certainly consistent with the setting name of 'Copy Input Data...').  

Also I tried to limit the T# Time Periods with a Member Filter (only some for a Year) and it seemed like this was ignored.

1. Are there any ways to do the non-Base Entity and Time limitations with Hybrid copies?
2. Do the Member Filters only apply to the non-Data Unit dimensions?

Thank you!

[EricOsmanski]

That is correct - the member filters on Hybrid Source data are only Account-level dimensions. Data Unit dimensions are invalid. DU dimensions would be supplied when a calc/trans/cons is ran.

[ChuckUsher]

(thank you very much Eric!)

[camagruder]

Hi Eric, thanks for the additional information. Very helpful!

Do you know if it's possible to copy the last month's data from the source scenario into future years of a target scenario? Our source scenario is monthly and forecast through 2024. The target scenario is for long range planning and we want to have the data from 2024M12 populate the target scenario from 2025M1 - 2032M12.

[EricOsmanski]

That cannot be done with Hybrid Scenarios. There is no Time (or any Data Unit) filtering/conditional logic unless you did "Using Business Rule" as the Data Binding Type. But I would argue you should just write a Custom Calc BR at that point. One of the main  benefits of Hybrid is the performance and that is mainly achieved by copying all periods in a year simultaneous (and why there is no time element at this moment).

[Tom_Abbate]

Does the Copy Input Data from Source Scenario option copy all data cell storage types or just input?

[EricOsmanski]

It will copy all storage types.

[miked]

Hi Eric, when you say that it will copy all storage types, you mean that it will copy them all but when it pastes or writes to the target scenario as calculation or durable calculation right?  There is not a way to have it copy the input data as input data, the calculated data as calculated and the durable calculated data as durable calculated to the target scenario right?

Thx,

Mike

[EricOsmanski]

That's correct Mike.

[miked]

Thx Eric.  One more question while I have your ear.  If I want to do a very simple 'Pointer' scenario like 'Current Forecast' that just points at a regular Forecast scenario like FC3+9, can I just use the 'Share Data From Source Scenario' binding type and then provide the Source Scenario Name (FC3+9) and just leave all of the other settings blank?  It looks like it worked for me on base and parent entities.  Am I missing anything or is this valid use case?  This is very handy and saves having to write some logic.  Feels very Essbase-y (that's a compliment in this case).

Mike

[EricOsmanski]

That is a valid use case and would work great.

Another use case is to share the data to a new scenario where you can filter out dimensionality not required for certain reporting so you can reduce the data unit and improve reporting performance.

[miked]

ok got it thx so much.  much appreciated