Help in creating a PAW report

[balaji1225]

Hey,

I need inputs on how I can create a report in PAW for the following scenario where

I have a cube with 2 dimensions with one dimension is having elements by concatenation of both Account and Location like Account1-Location1, Account1-Location2 etc and other dimension being the measure dimension with all the values.

How I can create a view with Account as X - axis and location as filter for the report and the measure dimension being the Y axis. We can easily create this in Perspectives/PAX by using DBRW but I am blank on how we can do it in PAW. Appreciate if anyone can answer this naive question.
Thanks

[burnstripe]

You could create two alternate hierarchies
Account
Location
In your dim1 which has the concatenation of account and location

So you had
Acc1-loc1
Acc1-loc2
Acc2-loc1
Acc2-loc2

In the account hierarchy you would have
Parent: acc1 child: acc1-loc1
Parent: acc1 child: acc1-loc2

Parent: acc2 child: acc2-loc1
Parent: acc2 child: acc2-loc2

In the location hierarchy you'd have
Parent: loc1 child: acc1-loc1
Parent: loc1 child: acc2-loc1

Parent: loc2 child: acc1-loc2
Parent: loc2 child: acc2-loc2

Then in paw you can add these alternate (related) hierarchies. They can then be split onto separate axis or filtered separately. You can do this by right clicking on the acc-loc dim and selecting add related hierarchy...

[MarenC]

Hi,

Alternatively, you could create a new user attribute called something like Location and then in the MDX for the concatenation dimension do a filterbypattern for the value of the client attribute.

On the PAW screen you could include the client attribute cube with the Location attribute and the logged on user (doesn't have to be a client attribute could be a new cube that includes the client dim).

The user would select the location and the concatenated dimension would return the required elements.

I have never tried this myself with filterbypattern or with a graph but cannot think why it wouldn't work.

Maren