The new valueat() function seems to be quite an elegant function, but I can’t get it to behave the way I want when timelines are involved. I can understand why valueat() lists would be limited to a single level of categories, but Quantrix documentation suggests that timelines are a “special” set of categories that can be easily recursed / traversed by functions. I would think (HOPE??!!) that valueat should be designed with timeline support in mind.
My goal is to be able to reference a list of assumptions from one matrix (with a a “generic” Year + Quarter timeline with no start date) in another matrix with a different Year + Quarter timeline with an explicit start date (say 2014 Q1), and configure an offset (say, configure an offset of -3 such at the assumptions for Y1 Q1 appears in the other matrix’s 2014 Q3, Y1 Q2 assumptions appear in 2014 Q4, etc.
Please see the attached model. The above scenario is found in the matrices Calendar and Assumptions, but the behavior is weird – I can only get the 1st data item to appear, but not the whole list.
The desired results are found in Calendar 2.
If experts out there can offer alternative formulas to achieve what I want, I would appreciate it!