Earn a 50% discount on the DP-600 certification exam by completing the Fabric 30 Days to Learn It challenge.
As is well established, best practice is to have a Fact table with columns that are your 'facts' for each row. What is the suggested approach to when those columns may have dimensions that are unique to them? For example certain groups of columns may always 'go together' and in visualizations it may be desirable to filter or organize the data based on these dimensions.
Solved! Go to Solution.
Regarding your specific question about columns that may have dimensions that are unique to them, it is recommended to split the single large fact table into multiple fact tables based on these dimensional relationships.
Patrick from Guy in a Cube in his classic 2019 video addressed
best practices in Power BI Data Modeling:
Data modeling best practices - Part 1 - in Power BI and Analysis Services
As well EnterpriseDNA from their Blog (they have learning paths and courses, and many Forum Posts on Data Modeling best practices):
Data Modeling In Power BI: Tips & Best Practices (enterprisedna.co)
Thanks,
Love GiC and have also used eDNA. I'll take a look, good chance I've watched the GiC video. 🙂
Sounds like my intution on splitting the Fact table is the 'right' path to go down.
Regarding your specific question about columns that may have dimensions that are unique to them, it is recommended to split the single large fact table into multiple fact tables based on these dimensional relationships.
Patrick from Guy in a Cube in his classic 2019 video addressed
best practices in Power BI Data Modeling:
Data modeling best practices - Part 1 - in Power BI and Analysis Services
As well EnterpriseDNA from their Blog (they have learning paths and courses, and many Forum Posts on Data Modeling best practices):
Data Modeling In Power BI: Tips & Best Practices (enterprisedna.co)
User | Count |
---|---|
100 | |
87 | |
81 | |
76 | |
71 |
User | Count |
---|---|
113 | |
104 | |
100 | |
73 | |
65 |