Group by the multiple values in one column

[Anonymous]

Hello. I have a some columns in my file that correspond to customer's orders: order date, payment date, Number of order, product typology, etc.  The behavior of the file is: when some customer order more than 1 item, it creates two lines for the same order, for example:

Num_Order       Orderdate             SKU              typology         

1234                   10/09/2020           123456        23

1234                   10/09/2020           765432         22

I want to do a group by by Num_Order so i can identify how many orders are mono orders (1 item) and how many are multi orders (more than 1 item). The problem is that I need to add to this group by the item's typology and customers can order itens from the same typology and different as well. 

I need this information for doing an analysis of how many customers did a multi order for all itens typologies except "23", by excluing the order lines that have typology "23" and maintain the rest. 

[amitchandak · Accepted solution]

@Anonymous , You can count orders like this

Single Item orders = countx(filter(Summarize(Table,Table[Num_Order], "_1", count(Table[SKU])), [_1]=1),[Num_Order])
multiple Item orders = countx(filter(Summarize(Table,Table[Num_Order], "_1", count(Table[SKU])), [_1]>1),[Num_Order])

Two have these as bucket values or dimension value you need have brining or segmentation

https://www.daxpatterns.com/dynamic-segmentation/
https://www.daxpatterns.com/static-segmentation/
https://www.poweredsolutions.co/2020/01/11/dax-vs-power-query-static-segmentation-in-power-bi-dax-power-query/
https://radacad.com/grouping-and-binning-step-towards-better-data-visualization

[Greg_Deckler]

@Anonymous Seems like a matrix visualization with multiple columns in the Rows field area?

[Anonymous]

It is a matrix with multiple column fields that correspond to order's fields like number of order, payment date, shipment status, Sku number, and SKU typology. 

Each row correspond to each order's entry. That means that if a client orders only one item, it corresponds to his order but if a client orders 4 itens, the 4 rows correspond to client's order. 

I want to group by by number of order to understand if a customer orders one item or more than one. The problem is that I want to understand which sku typologies clients order and I can't group by different values (a client can order different itens' typologies like socks and a tshirt, that corresponds to different numbers (23 and 45 for example).