TI Subset insert Consolidated Elementes

[maps]

I have problems creating a subset.

Lets asume I have a hierarchy like the following in the dimension cars.

total cars
- Big cars
-x1
-x2
-x3

- Small cars
-x4
-x5
-x6
The subset I like to create can vary. The lists of the elements of the subset I get out of a database. So my TI-Process always
destroys the subset and creates a new one.

DB Table with Subset elements:
x1; Big cars; total cars;
x3; Big cars; total cars;
x6; Small cars; total cars;

Hierarchy
total cars
- Big cars
-x1
-x3

- Small cars
-x6

Now I have Problems to insert the consolidated elements. When I add an C-Elements it automatically adds all of the childelements to the subset. So for example when I add Big cars, it automatically adds x1,x2,x3... Did I miss something? Like function to add childs to a consolidated element?

Any ideas how to implement that?

[qml]

When I add an C-Elements it automatically adds all of the childelements to the subset. So for example when I add Big cars, it automatically adds x1,x2,x3...

Are you absolutely sure this is what happens? Are you sure these elements don't show up only when you click on the plus symbol next to the C element in the subset to see its children? Because if you're sure, then this behaviour is not an expected one, and would be a bug, already most likely noticed by someone. So, are you absolutely sure that what you're saying is true?

[lotsaram]

When I add an C-Elements it automatically adds all of the childelements to the subset. So for example when I add Big cars, it automatically adds x1,x2,x3...

Are you absolutely sure this is what happens? Are you sure these elements don't show up only when you click on the plus symbol next to the C element in the subset to see its children? Because if you're sure, then this behaviour is not an expected one, and would be a bug, already most likely noticed by someone. So, are you absolutely sure that what you're saying is true?

I second that. SubsetElementInsert adds only the specifically named member. I have never seen the function do anything else (except maybe throw an error if an invalid index value is passed to the function.)

Of course once the subset is opened in the subset editor and the user actually DOES SOMETHING like clicking drill down all or drilling down on a member then you are not actually looking at the named subset anymore but at a list of members derived from the original subset. In the cube viewer there is no way to make a subset "read only".

[maps]

Okay I have 2 Problems.

The first Problem is that my hierarchy is pretty deep and I have a lot of bottom elements. I wanted to use the subset to show the user only the active elements of the dimension. The other Elements I can not delete because of historical data. So when I create the view with all nodes expanded the view is not readable for the user in addition the view will show the rejected elements when the user closes pushes the + again.

The next problem is creating the hierarchy with the TI-Process .To get the view with all nodes expanded I first thought I just need to add all bottom elements first and walk up the hierarchy inserting the c element. Order to insert: x1,x2,x3,samll cars, big cars, total cars. For that i would need for every level one ti-process but who cares... But it's even worse
The result is:
x1
x2
x3
small cars [+]
big cars [+]
total cars [+]


To get what i want i need to insert the elements in the following order x1,x2,small cars,x3, big cars, total cars. I have no idea how to create the subset out of the db-table.

[rmackenzie]

I wanted to use the subset to show the user only the active elements of the dimension

I just need to add all bottom elements first and walk up the hierarchy inserting the c element

If I understood correctly, you do want the subset to show a hierarchized view of the dimension, but excluding the inactive elements. If that is the case then you could use this MDX expression:

Code: Select all

{EXCEPT(
  {TM1DRILLDOWNMEMBER( {[Car].[Total Cars]}, ALL, RECURSIVE )},
  {FILTER( {TM1SUBSETALL([Car])}, [Car].[Active Car] = 'NO')}
)}

I think this is the 'top-down' alternative to the method you are having problems with. You should have the active flag attribute as a string so you can apply it to consolidated elements that only contain inactive elements. As long as you maintain that attribute then the expression should generate the correct subset.

However, this doesn't address the problem you mention here:

in addition the view will show the rejected elements when the user closes pushes the + again

To solve this, you could think about having an alternate hierarchy in your dimension that is made up of only the active elements. If the dimension is fast-moving in respect of the 'active status' of its elements then this could be awkward. But, if the active status only changes e.g. once per financial year then you could have several hierarchies - FY09 Active Cars, FY10 Active Cars, FY11 Active Cars, etc etc.

HTH,
Robin