Planning Analytics

Hi everyone!

I'm currently exploring MDX calculations that behave differently based on whether a member is consolidated or leaf (just like we're able to do in Cube rules with the C: and N: syntax). The intent there is to create a complex calculation that would be used in only one report. As a MDX calculation, it would live in the report query and I don't have to "pollute" the core model (and its global performance & maintainability) with yet another cube rule that might be a temporary/limited need. 

The only way I've managed to create such rules is with the ISLEAF() function, either trapped in a IIF() function or in a CASE statement. The calculation works well but it corrupts the view when applied: interacting with view stops working (can't change selections, can't apply suppress 0, drill, swap dimensions etc...). However the result is correct, which tells me that the issue might be rather on PAW side than on TM1server side.

Alternatively, I've seen some kind of system attribute in PAW (Type) which seems to give the N/C info, however I've never been able to pull it out with CURRENTMEMBER.PROPERTIES("Type"). I'm guessing that perhaps the displayed name in PAW is different from the real technical name...

Have any of you guys been able to make such calculation work?

HI Paul,

This has come up before and I would say is a bug in PAW.

As soon as you have something like in IIF() or CASE and other functions with your calculated member, synchronisation seems to stop working.

Given that the selectors are synchronised to items in the context/slicers, this does not make sense hence I am saying Bug.

@STUART KING - are you aware of this?

Hi everyone!

I'm currently exploring MDX calculations that behave differently based on whether a member is consolidated or leaf (just like we're able to do in Cube rules with the C: and N: syntax). The intent there is to create a complex calculation that would be used in only one report. As a MDX calculation, it would live in the report query and I don't have to "pollute" the core model (and its global performance & maintainability) with yet another cube rule that might be a temporary/limited need. 

The only way I've managed to create such rules is with the ISLEAF() function, either trapped in a IIF() function or in a CASE statement. The calculation works well but it corrupts the view when applied: interacting with view stops working (can't change selections, can't apply suppress 0, drill, swap dimensions etc...). However the result is correct, which tells me that the issue might be rather on PAW side than on TM1server side.

Alternatively, I've seen some kind of system attribute in PAW (Type) which seems to give the N/C info, however I've never been able to pull it out with CURRENTMEMBER.PROPERTIES("Type"). I'm guessing that perhaps the displayed name in PAW is different from the real technical name...

Have any of you guys been able to make such calculation work?