When to create dimensions?

[wilson_smyth]

I have been given an excel file of data, nothing out of the ordinarly. Filtering will be done based on some of the columns in that file.

Its easy to just import the file and with little or no work start creating reports.

My questions is, when is it more prudent to move those filtering columns to a seperate dimension table?

IMO, its when multiple fact tables need filtering by the same column but i want to hear if there are other unforseen benefits to creating dimensoin tables of these filter columns at an earlier stage in dev?

[OwenAuger]

@wilson_smyth

There is a good discussion on this topic here:

https://community.powerbi.com/t5/Desktop/Performance-considerations-when-building-a-Power-BI-datamod...

When writing DAX expressions, it is certainly easier to manipulate the filter context of a dimension if it is contained in its own table. And time intelligence requires a standalone date table (assuming you don't use the auto-generated date table).

[v-chuncz-msft]

@wilson_smyth,

To help improve Power BI, you may just Give Feedback.