Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.
Hi there,
I have the following calculated columns in the 'Product' table of the Contoso data model (available here). 'Product'[ProductKey] has 2517 unique values.
A) ‘Product’[Test 1]=CALCULATE(ALL(’Product’[ProductKey])) yields the ProductKey per row.
B) ‘Product’[Test 2]=CALCULATE(COUNTROWS(ALL(’Product’[ProductKey]))) however yields 2517 in all rows (size of the 'Product' table)
I find it puzzling that the filter resulting from context transition does affect the result in ‘Product’[Test 1] but it doesn't in ‘Product’[Test 2]. Could someone explain what is going on?
Thanks very much
Solved! Go to Solution.
Hi @AlB and @MattAllington
I did email Jeffrey Wang about this issue recently and he confirmed it is a bug and that it will be fixed in a future release. I don't have a timeframe, unfortunately.
Hi @AlB,
The sample file you have offered could not be opened due to the license problems coudl you please just share some sample file to have a test?
And the formula you have offered in ‘Product’[Test 1]=CALCULATE(ALL(’Product’[ProductKey])) seemed wrong, if you want to use the CALCULATE function, you should have an Aggregate function in it, could you please modify your formula and test it again?
Regards,
Daniel He
There is no error. Both 'Test 1' and 'Test 2' work as you can see in the attached file. I am just trying to understand why the interaction between context transition and ALL() seems to be different in 'Test 1' than in 'Test 2'
Thanks
Hi @AlB,
Based on my research, it is due to the simple useage of calculate funtion and the dax engine-The VertiPaq Engine in DAX, when you are using one parameter in calculate function, it takes the existing row contexts (if any) and transforms them into an equivalent filter context.
You could refer to below link:
https://powerpivotpro.com/2014/03/becoming-one-with-calculate/
https://www.microsoftpressstore.com/articles/article.aspx?p=2449192&seqNum=2
And you could refer to the Chapter 5 Understanding CALCULATE and CALCULATETABLE in the book
"The Definitive Guide to DAX: Business intelligence with Microsoft Excel, SQL Server Analysis Services, and Power BI"
Regards,
Daniel He
Thanks @v-danhe-msft
That's however what I explained in my previous post and it does not justify the behaviour
My recollection was that @Phil_Seamark contacted Jeff Wang about this, and Jeff says it is a bug. Phil, can you confirm?
Hi @AlB and @MattAllington
I did email Jeffrey Wang about this issue recently and he confirmed it is a bug and that it will be fixed in a future release. I don't have a timeframe, unfortunately.
When a Table expression returns a single row and a single column..It can be used as a scalar.
So for example.You have a simple table (say Table1)
Column1
1 |
1 |
1 |
Lets define a Calculated Tableas ALL(Table1[Column1])
Since ALL returns unique values in a column/Table, above results in a Table of singlecolumn and single row containing 1
So now if you define even a MEASURE as following ..it will work
1 |
1 |
2 |
Thanks @Zubair_Muhammad
I understand that. My question has more to do with why the filter resulting from context transition has an effect on 'Test 1' but not on 'Test 2'.
Judging from the result, in 'Test 1' the filter resulting from context transition contains the value of ProductKey in the current row. That seems to override the ALL() to select only one value, a scalar, and it works. But following that reasoning, 'Test 2' should yield a 1 instead of 2517. as the ALL() is overridden by that one-row filter from context transition. It is that apparently different behaviour what I do not understand. What is your take on that?
Thanks
Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City
Check out the April 2024 Power BI update to learn about new features.
User | Count |
---|---|
109 | |
98 | |
77 | |
66 | |
54 |
User | Count |
---|---|
144 | |
104 | |
100 | |
86 | |
64 |