Data model

[katybr]

Hello! 

I have a question to know which is the best practice. 

I have a project where I have 4 tables:

-Program (IdProgram,Name,Responsable...)
-Parts of the program (IdProgram,IdSection,SectionName...)

-Parts (IdSection,IdParts,Description...)

-Conformities(IdParts,IdConformities,Answer..)

I can work with all those 4 tables related in the data model, each table have one columns that can be related with another table.

My question is, is it best to work with the 4 table separately or merge all the 4 tables into one?

Thank you 🙂

[PaulDBrown · Accepted solution]

Best practices suggest creating dimension tables with unique values for fields common to different tables, creating one-to-many (single direction) relationships between these dimension tables and the corresponding fields in the fact tables. 
you then use the diemnsion table fields in measures, slicers, visuals, filters...
The dimension tables provide the filter context for correct filtering and calculations

[Anonymous]

@katybr Hi, I don't have much details about your model, but it's depend on each table's complexity and weight. Also I believe if you merge 4 queries in one, it wouldn't be good practice in term of query performance.