Creating data models

[BellPower]

Hi all, I have been using Power BI in a medium sized company for about 2 years now. I have created many good dashboards and have mostly taught myself. I used to work with Transact SQL and create visuals in an ERP through creating views in SQL and using the views as my data source, so grouping, merging etc always took place in SQL. The company I work for now works with a PSA system using an API. I often work with multiple fact tables. I have read a lot that it's best to create a star schema, but I still find it easiest to merge the tables in Power Query and expand only the needed columns to have all my data in 1 table. Is this bad modeling? The data isn't huge, but I want to become better so would like to have some comments.

[amitchandak · Accepted solution]

@BellPower , When you need a measure that needs to go across rows, you should use the ALL function or time intelligence functions. This approach works best with a star schema. While it may work for smaller datasets, for complex problem solving, always use a star schema with a proper date/calendar table.

refer video from Guyinacube- https://www.youtube.com/watch?v=vZndrBBPiQc