Cognos Analytics

Hi Veljko,

for those purposes there's the "for" clause in calculations. It determines the aggregation level for the calculation.

I created a small sample based on IBM sample data Great Outdoors. count( distinct City) and count (distinct City for Region). Then a third calculation for the percentage.

 That's the result:

You would nedd a count( distinct Stores for City) in your data.

Hope this helps.

Hi Robert,

Thank you for your replay.

First problem is that I can't get this editor like you for calculation. Mine looks like this 

Hi Veljko,

you can open the calculation editor by using the context menu (right click) on folders or tables:

Or you click the link on the bottom left of the screen here:

Yes, I have done like that at first, but since I always was getting 1 as a result for a % calculation, I thought that maybe the problem is with editor :)

I am using DMR, since at your example is a relation model, maybe that is a problem? Also, I am trying to show this calculation at visualization where city is changed dynamically.  

Ah! DMR - way more easy than thought. What you need is a completeTuple + currentMember(<Hierarchy>). 

I created a sample based on the IBM GO Dynamic Cubes sample cube. Just create the complete tuple and then use it to calculate the percentage of your count distinct measure to this tuple:

In this example I would like to have the value of the count distinct measure in the combination of the dynamically assigned time and region. All other hierarchies the "All" or default member is used.

Works:

The only down side is that count distinct measures of DCs are not supported in summaries of tables. Don't know if that also applies to DMR but if you use charts - all is good.

It works perfect. Thank you very much Robert :)