Page 1 of 1
Minimum in consolidation
Posted: Fri Mar 30, 2012 8:47 am
by Toto
Hello,
I would like to write a rule which gives me:
a) the minimum for the consolidation of a dimension
b) the value of the last child for the dimension consolidation
So I need something like: ['All Elements'] = MIN('All Children');
As stated in the examples, the average is calculated with one measures has a count value and then I would divide the sum trough the counter.
At the moment I do not have an idea how to convert this approach to MINIMUM and LASTCHILD.
Is there a standard-way for doing this?
Thanks a lot for your help!
Toto
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 9:22 am
by Elessar
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 10:43 am
by lotsaram
Last Child can be achieved with a combination of ELCOMP(dim, par, ELCOMPN(dim, par))
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 11:38 am
by Michel Zijlema
I noticed that the Consolidated... functions are listed as TI only in the
10.1 (online) documentation.
Anyone here who tested whether this really is the case?
Michel
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 12:33 pm
by qml
Michel, these functions were added in 9.5.2, but only documented in a
technote.
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 12:46 pm
by Michel Zijlema
qml wrote:Michel, these functions were added in 9.5.2, but only documented in a
technote.
Hi Kamil,
I'm aware of that - what I noticed though is that in the technote the functions are listed as both rules and TI functions, while in the (on-line) 10.1 documentation (ref guide) the functions are listed as TI only.
I'm wondering whether this is an error in the documentation or whether something changed here...[spooky smiley required]
Michel
Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 12:58 pm
by qml
Ahh, gotcha. Sorry for misreading your post.
It's quite possible this is IBM's way of getting rid of the implementation and performance problems people were having with these functions when used in rules. They might have made them TI only instead of actually attempting to address these issues.

Re: Minimum in consolidation
Posted: Fri Mar 30, 2012 1:01 pm
by Duncan P
I have tried them in a test model just to check functionality and performance.
They do work in rules but you will need to check them yourself as some of the flag parameters do not work quite as documented. I can't remember the exact details.
They are very different from ConsolidateChildren. Whereas that works with the direct children the ConsolidatedMin/Max/Avg/Count functions work directly off the leaf data in the same way as normal consolidation. As a result they use the feeder information and perform similarly to normal consolidation as well.
Re: Minimum in consolidation
Posted: Sat Mar 31, 2012 10:18 am
by Toto
Hello,
thanks for all your replies.
They are very different from ConsolidateChildren. Whereas that works with the direct children the ConsolidatedMin/Max/Avg/Count functions work directly off the leaf data in the same way as normal consolidation. As a result they use the feeder information and perform similarly to normal consolidation as well.
Is this the way the normal consolidation works/should work? So the sum of something should ALWAYS the sum of the leaf elements, and not the sum of the aggregated levels between leaf elements and the total elements?
I have the following dimension structure:
All Units
--- Total of units per country
---- single unit
When I use the ConsolidateMin function for the level 'Total of units per country' the standard consolidation seems to aggregate the level 'Total of units per country' to the 'All Units'. So it would not be the aggregation of the leaf levels.
Another question: If the functions ConsolidatedMin/Max etc. are only documented/supported for TI-process, what other ways I could use to achieve this in rules?
Thanks again!
Toto
Re: Minimum in consolidation
Posted: Sat Mar 31, 2012 10:43 am
by Duncan P
This is interesting. Is there any chance you could post the rule you are using and the relevant parts of the dimension structure?
Re: Minimum in consolidation
Posted: Sun Apr 01, 2012 3:32 pm
by Toto
Hi Duncan,
sorry, my mistake. It is as you said, the consolidation sums up the values from the leaf level.
So, only one question is open for me: If functions like ConsolidateMin, ConsolidateMax are only officially supported for TI-processes, which other options do I have in the rules editor?
Thank a lot to all of you!
Toto
Re: Minimum in consolidation
Posted: Sun Apr 01, 2012 6:17 pm
by Duncan P
Hi Toto,
From
http://www-01.ibm.com/support/docview.w ... wg21472749 which applies to 9.5.2
we have
These functions can be used in TurboIntegrator processes or Rules.
and from
http://www-01.ibm.com/support/docview.w ... wg27024004 which applies to 10.1.0 (albeit only mentioning ConsolidatedMin in passing)
we also have
This function can be used in TurboIntegrator processes or Rules.
so it is at least documented somewhere as working in both versions.
Cheers,
Duncan.