No CACULATE challenge

[AlexisOlson]

For those haters of CALCULATE like @Greg_Deckler, I challenge you to create a measure without CALCULATE that performs anywhere close to the same speed as this very simple measure that switches to an inactive relationship. It must also not break if a simple filter from any of the dimension tables is applied.

Sales Amount (Delivered) = 
CALCULATE (
    SUM ( Sales[SalesAmount] ),
    USERELATIONSHIP ( Sales[DeliveredDateKey], 'Calendar'[DateKey] )
)

Please use the attached file.

Report preview:

As expected, we can see that sales during periods without a discount promotion have products delivered in promotion months.

[tamerj1 · Accepted solution]

Hi @AlexisOlson  & @Greg_Deckler 
This one uses neither CALCULATE nor CALCULATETABLE and performs faster than the original one

Sales Amount (Delivered) 2 = 
VAR SelectedSales = ALLSELECTED ( Sales )
VAR SummarySales = 
    SUMMARIZE ( 
        SelectedSales, 
        Sales[DeliveredDateKey],
        "@SalesAnount", SUM ( Sales[SalesAmount] ) 
    )
RETURN
    SUMX ( 
        VALUES ( 'Calendar'[DateKey] ),
        VAR FilteredSales = FILTER ( SummarySales, [DeliveredDateKey] = 'Calendar'[DateKey] ) 
        RETURN
            SUMX ( FilteredSales, [@SalesAnount] )
    )

[tamerj1]

@Greg_Deckler 

Yes I actually placed the solution of the first problem in a separate reply right before the the solution of the 2nd problem. 

[Greg_Deckler]

@tamerj1 Somehow I missed that, the way these forums do threading is mildly infuriating at times.

[Greg_Deckler]

@AlexisOlson It's a good more or less single use case where CALCULATE is beneficial but really just because USERELATIONSHIP was coded to work with CALCULATE. USERELATIONSHIP could have been coded to support FILTER as well. It's not really CALCULATE, it's the additional functions that were coded to work with CALCULATE. But, there are plenty of examples of the reverse. Try writing a version of these with only using CALCULATE and no X aggregator that A. Actually works and B. Performs significantly faster.

1. Write a running total measure that uses a single table data model only using only CALCULATE and no X aggregator
2. Write a working measure total for a semi-additive measure using only CALCULATE and no X aggregator. Measure Totals, The Final Word - Microsoft Power BI Community
3. Write a version of Open Tickets measure using only CALCULATE and no X aggregator
4. Write a Deseasonalized Correlation Coefficient measure using only CALCULATE and no X aggregator
5. Write a Net Work Days measure using only CALCULATE and no X aggregator
6. Write a Simple Linear Regression measure formula using only CALULATE and n X aggregator
7. Write Cthulhu using only CALCULATE and no X aggregator
8. Write an MTBF measure using CALCULATE and no X aggregator
9. Write a While loop measure using CALCULATE
10. Write a For loop measure using CALCULATE
11. Write a Days of Supply measure using only CALCULATE and no X aggregator
12. Write Hour Breakdown measure using only CALCULATE and no X aggregator
13. Write Lookup Value measure using only CALCULATE and no X aggregator
14. Write Overworked measure using only CALCULATE and no X aggregator

I could go on, I only got through 3 pages of 11 of the Quick Measure Gallery.

The point here is the the No CALCULATE approach is a far better and more flexible approach to writing DAX that allows you to solve real-world problems, problems that CALCULATE could never hope to touch. Yes, there are times when CALCULATE is a good idea for one reason or another. But, this fixation the DAX community has on CALCULATE is unhealthy. It makes DAX harder to learn and breeds this crazy culture where people want to use it everywhere when it is simply making their lives harder and is absolutely unnecessary the vast majority of the time when the simple approach of:

1. Create some VAR's
2. Create a Table VAR
3. X Aggregator

Will solve the VAST majority of problems in DAX without ever needing CALCULATE. Simple. No reason to worry about the internal workings of CALCULATE or context transition or pretty much any of the stuff that people find "hard" about DAX.

[AlexisOlson]

@Greg_Deckler X aggregators are absolutely essential. I have no interest in comparing No CALCULATE vs No X aggregator since I'd rather not handicap myself at all.

[tamerj1]

@Greg_Deckler 
Totally agree 

[tamerj1]

@Greg_Deckler 

🤣🤣🤣