DAT224x: Developing a SQL Server Analysis Services Multidimensional DataModel

I like SSAS Dimensional. It's a thorough, powerful and robust tool.

I like the write-back capability, which brings enterprise planning and budgeting into the Microsoft BI landscape, very much in the guise of IBM Cognos TM1 or older Cognos Planning. I like the possibility to embed KPIs and drill-through actions into the cube, ready to be consumed into Excel, without any need for further developments on the reporting side. I like the powerful Business Intelligence wizards, which allow to give life to a chart of account for instance, through unary operators, which relate each account line to the other.

The equivalent tool in the Cognos landscape would be 20+ years old Powerplay transformer, which does the job in terms of creating cubes, but certainly doesn't provide all the possibilities above. No write-back, no KPIs, no unary operator. Powerplay transformer is just a brick in Cognos BI, whereas SSAS Dimensional is an enterprise-level hierarchical BI semantic layer, that can be queried by Excel, SSRS, or Power BI, and offers a planning and budgeting solution.

Subjects covered:

• Data Sources, Data Source Views, Dimensions, Attribute relationships, Parent-Child Hierarchies
• Cubes, Perspectives, Translations, Business Intelligence functions
• MDX,
• Calculated Members, Named Sets, Scoped Assignments, KPIs, Actions
• Aggregations, Processing, Write-back, Security, Deployment
• Architectural Decisions, in particular the choice between SSAS Tabular and SSAS Dimensional...