Forum Discussion

Thor2022's avatar
Thor2022
Frequent Visitor
3 years ago
Solved

Filter the contents inside MATRIX visual

Dear All  I have a matric that contains the following information:  - Rows: Warehouse names: A, B and C  - Columns: Years  - Values: Profits    and I have a multiselection slicer that contains ...
Tips and Tricks
  • PaulDBrown's avatar
    PaulDBrown
    3 years ago

    Thor2022 

    Actually, thinking about it, the average method could potentially not achieve what you are looking for.

    If you have a set of values for a year like {30, 29, 30, 31}, the average method will hide both 30 values.

    The following  method however will work: compare the minimum value in a set to the maximum value. If they are the same, all values are the same...

    So...

     

    No duplicate profits =
VAR _Min =
    MINX ( ALL ( 'Warehouse table'[Warehouse] ), [Your measure] )
VAR _MAX =
    MAXX ( ALL ( 'Warehouse table'[Warehouse] ), [Your measure] )
RETURN
    IF (
        AND (
            ISINSCOPE ( 'Warehouse table'[Warehouse] ),
            ISINSCOPE ( 'Year Table'[Year] )
        ),
        IF ( _Min = _Max, BLANK (), [Your measure] )
    )

     

     

    PS. I tried deleting the previous message but the forum won't let me for that particular post for some reason....

     

     

PaulDBrown's avatar
PaulDBrown
Community Champion
3 years ago

Here is one way. Basically you compare the sum (profit) of each warehouse with the average for all warehouses in a year.

First the model:

 The calculation (without totals)

 

No duplicate profits =
VAR _Profit =
    SUM ( fTable[Profit] )
VAR _Average =
    CALCULATE ( AVERAGE ( fTable[Profit] ), ALL ( 'Warehouse table'[Warehouse] ) )
RETURN
    IF (
        AND (
            ISINSCOPE ( 'Warehouse table'[Warehouse] ),
            ISINSCOPE ( 'Year Table'[Year] )
        ),
        IF ( _Profit = _Average, BLANK (), _Profit )
    )

 

and if you need totals:

 

No Dups with totals = 
SUMX(SUMMARIZE(fTable, 'Warehouse table'[Warehouse], 'Year Table'[Year]), [No duplicate profits])

 

 

Sample PBIX attached

pi_eye's avatar
pi_eye
Resolver IV
3 years ago

Hi Thor2022 

 

This is interesting, because for some reason it's not as simple to aggregate a distinct count on another measure! However it is always possible in DAX 😉

 

I created a new measure rather than chain it in called "New count" which evaluates to the total number of distinct sums. I am use group by to create an aggregated table which I can pass the summed column through to a  distinct count.

New Count = calculate(
        countx(
            distinct(SELECTCOLUMNS( GROUPBY(WareHouse,WareHouse[WareHouse],"SumVal",sumx(CURRENTGROUP(),WareHouse[Value])),[SumVal]) ) ,
            [SumVal] ),
        ALLSELECTED(WareHouse[WareHouse])
)
 
On it's own, it looks like this: (sorry I can't seem to cut and paste images right now)
 12345678 
WH142123333 
WH242123333 
WH342123333 
WH442123333 
          
 

Added into the final expression gives you: 

Show diff Values = If([New Count]=1,"", Sum(WareHouse[Value]))

 

Let me know if this works,

Pi

  • Thor2022's avatar
    Thor2022
    Frequent Visitor
    3 years ago

    Hi Pi 

    I am sure its a great solution , but give me some time to digest it , since some of the used functions are new for me 🙂 

    Regards