Choose which calculationgroup to use?

Hi,  Pbiuserr 

If you want to know more about compute groups, you can refer to the Compute groups article in SQLBI and then evaluate whether your current choice can be realized, you can click on the following link:

Using calculation groups to selectively replace measures in DAX expressions - SQLBI

Official Documentation:

Create calculation groups in Power BI (preview) - Power BI | Microsoft Learn

As you mentioned, you want to control the appearance of your set of columns in the slicer and you can use a custom matrix header. I used the following sample data:

I've created a matrix header table:

 

Use the following DAX expression to create a measure to switch columns:

 

Measure = SUMX('Fact','Fact'[sales])
Measure A = VAR _seleted = SELECTEDVALUE(Header[name]) RETURN CALCULATE(SUM('Fact'[sales]),FORMAT('Fact'[Date],"MMM")=_seleted)
Measure B = DIVIDE(SUM('Fact'[sales]),10)

Data = 
SWITCH(TRUE(),
    SELECTEDVALUE(Header[Indicator])="Month",
        SWITCH(TRUE(),
            SELECTEDVALUE(Header[name])="Jan",[Measure A],
            SELECTEDVALUE(Header[name])="Feb",[Measure A],
            SELECTEDVALUE(Header[name])="Mar",[Measure A]
        ),
    SELECTEDVALUE(Header[Indicator])="Other calculation",
        SWITCH(TRUE(),
            SELECTEDVALUE(Header[name])="Total",[Measure]
        ),
    SELECTEDVALUE(Header[Indicator])="Other calculation1",
        SWITCH(TRUE(),
            SELECTEDVALUE(Header[name])="Zther calculation",[Measure B]
        )
)

 

Here's what it looks like:

 

This will display the groups selected by the user. You can refer to the related cases below:

Solved: How to show a Row Total as Row element in a matrix... - Microsoft Fabric Community

Solved: Matrix Overlapping Groups - Microsoft Fabric Community
Solved: How to add a custom measure at the end of matrix v... - Microsoft Fabric Community

 

 

 

Best Regards

Jianpeng Li

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.