2) In the basic model it was possible to generate a directory having a lot of balancing levels (See Screenshot):
3) This was our goal, but for the convenience of users it would be desirable to hide the balancing levels, that is, to get such a view of the table:
A solution is needed that would be able to hide these levels automatically - a filter (in the example above it is hidden in the manual). However, when configuring a filter, I can only use modules that have the same change and at the same level as the original module (in my case it's level 8). But I can not select the L8 layer in the module and set the values on the nodes that could identify fictitious nodes.
Different formulas and their combinations that can determine them can not work with the summation type "formula" (because they use ITEM or LOOKUP). To use properties of other levels too it is impossible (an error: different directories / levels).
There was another "manual" version using the "Select Item To Show" and searching for balancing elements by code. But this function has a limit on the number of values and does not allow me to save my choice:
What other approaches can be used to hide the balancing levels of a balanced hierarchy?
I managed to solve my problem using the approach of this article.
the solution is based on the use of the summation method "Ratio":
1) Define the date hub fictitious elements through text formulas. When loading lists into the main model, set the property for the balanced element to 1;
2) Create a module for filtering. Lists we use the lower level of the composite list. In the module we add a Technical line item (always equal to 1, the summation method is a formula), and a line item for each level of the list that can be fictitious. Lists and name of these line items correspond to each level.
3) As values for these line item, enter any identical number (I chose 5) and multiply it by our property "Balancing" (1 or 0). Summation method: Ratio with settings previous level / technical position.
4) Use the filter on the lower level: hide the values that are not equal to the number indicated by us (5 in my example).