Hi, I'm working to understand the new "aggregations" feature in PBI desktop (and now the service). I completely understand how I can "omit" dimensions to summarize a fact table...but I'm a bit confused about how to still use a dimension but summarize to a non-leaf, parent hierarchy level.
For instance, in several of the videos, Christian Wade states that they are summarizing the fact table at the product sub-category level, which is a level above the detailed product level. From the screen shots, it's clear that they've snow-flaked the dimension - i.e. the product, product sub-categeory, and product category tables are all separate.
Is this a requirement for doing aggregations at non-leaf levels? I understand why it's desireable from a "there aren't repeating keys because the grain of the dimension table is too low"...but on the flip side, it also would prevent anyone from building a simple hierarchy on the dimension (can't build a product hierarchy if the different levels are in their own tables).
In my case, I want my aggregate table to store data at the month level instead of the day level. But I hate to split out separate tables to handle day level and month/qtr/year level, and then not be able to have a simple hierarchy to drill through to them.
Does anyone know a way to work around this?
Thanks,
Scott
