Page 1 of 1
Excluding elements with wildcard search
Posted: Sat Jul 20, 2024 7:33 pm
by raghun1995
Hello Team..I am new to planning analytics and this is my first project.
One of the hurdles I am facing is that I am trying to hide elements from a dimension with a specific pattern. There are 100's of elements which end with '_XYZ' and I need to hide them from users.
I tried several options like MDX subset, a while loop etc. Nothing seem to be working. The subset approach I tried works for the first time but it immediately starts showing the elements again once I drill up and drill down through in the hierarchy.
The data I am working is of Employee data.
Appreciate any kind of pointers.
Re: Excluding elements with wildcard search
Posted: Sat Jul 20, 2024 8:26 pm
by declanr
Hi,
As you have found; subsets are just a list of members - but it is just a pre-applied "selection" of members within the dimension. The user is still interacting with the dimension; so the second they make any sort of change (e.g. expanding or collapsing a consolidation) they are no longer looking at the subset and are instead just looking at a different "ad-hoc" list of members.
So the solution you are looking for really depends on the requirement:
1/ If the users should NEVER be able to see certain members in ANY cube - you would use ELEMENT SECURITY set to NONE to stop those members ever appearing for certain user groups
2/ If the users should NEVER be able to see data within a certain cube (or cubes) for some members; but they should be able to see data for those members in OTHER cubes then you would have to look at CELL SECURITY. This way the "member" could still exist but you wouldn't be able to see their salaries etc.
3/ If its not a security problem to see those members; but they make certain consolidations "messy" to view (e.g. employee _XYZ is in my department and I'm allowed to see them; BUT they are a contractor and I only want to see Permanent Employees) then you would look at either having additional consolidations OR alternative hierarchies for the dimension (e.g. the base Hierarchy has ALL employees; but then you have another Hierarchy that excludes all Contractors)
If you let us know what exactly is driving your requirement I am sure someone will be able to give a more targeted and suitable solution.
Thanks,
Declan
Re: Excluding elements with wildcard search
Posted: Sun Jul 21, 2024 6:02 am
by raghun1995
Thanks a lot Declan! Element security seems to be an option.
So for your question on what exactly is my requirement, I will have a to open a can of worms.
Dealing with employee data seems to be little tricky. The way I need to show a global leader hierarchy is that I need to show managers under a leader and one below and one below and it goes on till 6 levels. So the way I get a source data file is that, they left a blank next to each manager and the backend team did so because the manager at any level needs to be counted himself. Thats how the whole problem started and I am dealing with ragged hierarchy.
To address the blank cell problem next to each manager, I had to insert a dummy record using TI process, wherever I find a blank record next to a manager using the condition on field 'manager flag'='y'. That dummy record, I am naming it as 'ManagerName_Self'. As a result, there are 100's of such dummy elements that got generated.
Now I need to hide these dummy records to all the end users/leaders/managers (not specific to any group/person) because they wouldn't want to see their name under their own hierarchy.
I tried numerous ways to address this requirement starting from rules to inserting dummy records. I could not write a rule where I could say, whenever there is a consolidation, then please add 1. If I was successful in this, I would not need to deal with dummy records.
Hope I explained to some extent.