What is meant with "Summarizing" here?

[Anonymous]

Hi!

Suppose I am given the following table named Orders:

What is meant by

"Summarizing Orders by the CustomerID, OrderID and OrderDate columns"?

(Found this  here in Question#35)

Does this simply mean to apply the DAX-forumula SUMMARIZE, i.e., 

SUMMARIZE(Orders, Orders[CustomerID], Orders[OrderID], Orders[OrderDate])?

Best

[Greg_Deckler]

@Anonymous That's how I read it but you probably want something like this:

SUMMARIZE(Orders, Orders[CustomerID], Orders[OrderID], Orders[OrderDate], "UnitPrice",AVERAGE(Orders[UnitPrice]),"Quantity",SUM(Orders[Quantity]),"Discount",SUM(Orders[Discount]),"SalesTotal",SUM(Orders[SalesTotal]))

[Anonymous]

@Greg_Deckler Thank you for your answer.

Why do I have to aggregate all the other original columns?

[Greg_Deckler]

@Anonymous Well, you don't have to but then you just wind up with a table with three columns and I'm not sure that tells you much. In actuality, you would only need to aggregate the Sales Total column when summarizing (or using Group By in Power Query). I would answer the questions, Yes, No and Yes btw. You could also ditch the Quantity column if all you care about is Total Sales.

[Anonymous]

@Greg_Deckler Regarding the linked question, I am really not sure about the correct answers.

As to the first, I do not know what they mean (and this is why I asked this question). If it is meant as you read and suggest, why does this make the answer yes? Why does this reduce the model size for instance?

[Greg_Deckler]

@Anonymous Well, with the provided example data, it wouldn't reduce the overall dataset size. But, one would have to imagine that there are orders from the same customer on the same date that would include multiple products and thus grouping or summarizing the table would result in a smaller dataset. But, if every product ordered results in a new OrderID then that wouldn't be the case. The question sucks pretty much IMHO. Vague.