Hello,
I'm trying to figure out the right approach to build and share data models in Power BI and was hoping for your collective expert advise. I have a need to define relationships between tables and save the models to allow other users to use them as a static base to build reports. I think that sharing datasets may be the way to go as I can then assign security by user or group. However, some of the joins required are by more than one field. I've read that you can create concantenated fields to build unique ID's in both tables (fact & dimension) and then just join on the newly created fields. With that being said, in some models I could end up creating multiple new calculated columns just to facilitate the joins and I'm not sure if that is the way to go.
Most of the tables I need to model are Oracle DB tables. Occassionally I need to mix in other data sources to include .csv files. How are Power BI developers defining these relationships and sharing them? Are shared datasets the way to go? Does anyone use modeling software as an intermediary to define the model and just connect Power BI to it? For those of you who may have used Hyperion Interactive Reporting, I'm esentially trying to build and share data models similarly to that.
Thanks for any pointers you can help me with. Much appreciated.
