Sum by multiple Categories in a Measure in a visual with no said Category

[Anonymous]

Hello All,

Im looking for a meausure very similar to this https://community.fabric.microsoft.com/t5/Desktop/Sum-by-Category-Measure-in-a-visual-with-no-Catego... but with multiple categories.

I have a table with 4 columns, Shop Name, City, product and quantity.

Shop Name	City Name	Product	QTY
Fruit Galore	Antwerp	carrots	137
Fruit Fest	London	carrots	245
Fruit Fest	Antwerp	carrots	453
Fruit Fest	Antwerp	Apple	20

I want to show the total quantity in the network for every city/product combo related to the shop but not showing the city and product in visual. The resulting table would look like this.

Shop Name	Qty
Fruit Galore	590
Fruit Fest	855

Fruit Galore is 590 because total  Antwerp/Carrot-453+137

Fruit Fest is the sum of everything since it was every combination. 

Is there a dax measure that can do this?

[Jihwan_Kim · Accepted solution]

Hi,

I am not sure if I understood your question correctly, but I tried to create a sample pbix file like below.

I assume the semantic model looks like the below, and please check the below picture and the attached pbix file.

TREATAS function - DAX | Microsoft Learn

Expected result measure: =
VAR _t =
    SUMMARIZE ( Data, City[City Name], 'Product'[Product] )
VAR _filterfact =
    CALCULATETABLE (
        ALLSELECTED ( Data ),
        TREATAS ( _t, Data[City Name], Data[Product] )
    )
RETURN
    IF ( HASONEVALUE ( Shop[Shop Name] ), SUMX ( _filterfact, Data[QTY] ) )

[Anonymous]

Thank you, the output is what im looking for but im using a different data set with a similar semantic model. I have never really used the virtual table before, can you explain the logic and purpose of using the variables _t  and _filterfact? I want to recreate the measure with tables I'm using. 

_t is creating every city product combinantion in a table correct? 

_filterfact is recreating the 'data' table virtually to be used in the calculate function?

[Jihwan_Kim]

Hi,

Thank you for your message, and you are correct.

Please check the comments down below.

Expected result measure: =
VAR _t =       //  it shows all the combination of city and product of each row (shop name).
    SUMMARIZE ( Data, City[City Name], 'Product'[Product] )

VAR _filterfact =          // from all fact table ( remove filter context), apply the combination of city and product that are generated from above.
    CALCULATETABLE (
        ALLSELECTED ( Data ),
        TREATAS ( _t, Data[City Name], Data[Product] )
    )

RETURN
    IF ( HASONEVALUE ( Shop[Shop Name] ), SUMX ( _filterfact, Data[QTY] ) )

[Anonymous]

follow-up question as some of my requirements have changed. How would I do this as a calculated column?