Hi all
I was wondering if i might have overseen a feature in Tabular/PowerBI which i found would be valuable.
I'm looking for a way to filter out/release unused dimension attributes in the model when exposed to end users (maybe also from a model performance/size point of view)
Considder the following:
Sales Fact table holds all sold amount of products to all known customers. Attached to the fact table I have a product table and a customer table. Now since the model holds data for the last 5 years ex. i might only have data for 60% of the customers in the customer table and 80% of the products in the product table.
So what i'm looking for is a way to hide/disable customers and/or products that are not used. Either by releasing them from the model/memory or simply just by hiding them towards end users.
Now I know that I can enable bidirectional crossfiltering between the fact and product/customer table, but this is not what i'm after. Also this will be in a scenario with multiple fact tables and rather large dimensional tables.
I also know that I could go ahead and handle this in the relational layer by checking all ex. customer keys in all fact tables on load, but as the model when processed, kind of already knows what is used or not, this could be a nice feature to be able to enable/disable on all dimension tables.
What are your thoughts ?
