The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
I have the auto featured disabled as I do not want 15 date tables created in my report (I have a lot of different data values in this report).
I have one actual data table, and everything else are just flat "dates". Not all of these dates on the fact tables are connected to main date table, which is fine. I know I can use USERELATIONSHIP for calculations, but that is not what I am trying to do. I just need to see the results of a measure on a chart that has one of the flat date values. I need see the values in Year, Qtr Month but this is not a date that is connected to my date table.
There are a couple of these flat dates that I would actually like to see a date hierarchy (Year, Qtr, Month). Is there a DAX way to create a date hierarchy, or another way to go about this?
Solved! Go to Solution.
@Anonymous That's correct. In the event of auto time intelligence this can lead to data model size bloat as essentially a separate table is created for each date that spans the range of dates in that column. Since these are all unique values, they don't compress particularly well. There are some videos from Guy in a Cube that demonstrated dramatic size and performance improvements by turning auto time intelligence (automatic date hiearchies) off. I would tend to refer to it as a recommended approach/best practice.
@Anonymous You can create columns in your table using YEAR([Date]), QUARTER([Date]), MONTH([Date]) and then just build an ad-hoc hierarchy in your visual using these columns.
@Greg_Deckler Thanks Greg. Is this the recommended approach versus having the auto time intelligence turned on? In either approach, additional columns get created. Is there a performance benefit to one or the other? I guess the other point is that by doing it with DAX you can add hierarchies to only the dates that you want, rather than have hiearchies created for all of them automatically.
@Anonymous That's correct. In the event of auto time intelligence this can lead to data model size bloat as essentially a separate table is created for each date that spans the range of dates in that column. Since these are all unique values, they don't compress particularly well. There are some videos from Guy in a Cube that demonstrated dramatic size and performance improvements by turning auto time intelligence (automatic date hiearchies) off. I would tend to refer to it as a recommended approach/best practice.