I have a table of PRODUCTs. These PRODUCTs are sold by PARTNERs and are part of an ORDER_ITEMS, which (as expected) compose an ORDER.
On occasions, these PRODUCTs are involved in EVENTs (say, "maintenance").
Customer wants to create an "event index" which is defined by #EVENTS per PRODUCT/ sum( sold PRODUCTs).
One interesting analysis is to filter by PARTNER. In other words, I would like to calculate which PARTNER generate less EVENTs.
The problem is that this is creating a circular reference in my data model (as seen in the picture below), given the fact that I need to filter by PRODUCTs "which were sold" and PRODUCTs "which were involved in EVENTs".
I have created two PRODUCTs tables ("PRODUTOS"), and initially I have thought of applying set analysis (filter by product involved in events and calculate sum({<PRODUCT_ID={"$(=getFieldSelections(PRODUCT_EVENT_ID))"}>} SOLD_PRODUCTS) (something along these lines...)
However, this approach seems TOO convoluted.
How can I break this (potential) circular reference and apply this kind of filter in these two different contexts?
