Thanks!

Wow, what a thorough explaination!  I was tring to explain these concepts to some colleagues over lunch the other day, and your article beats the snot out of my back-of-napkin diagrams!

Thanks!

Very nice.  The graphics you took the time to do the "Alice's Restaurant" thing to really make this a very clear article on what Rollups and Cubes actually do.  The graphics clearly demonstrate why you might want to use the Grouping function of a Rollup or Cube.

There's lot's of folks who have no idea what a "subtree cost" is, how to determine it, or what unit of measurement it is.  Also, although very obvious from the graphics, simple example Selects for the isolation of subtotals and grandtotals would be very helpful for newbies.  A very brief explanation of these two things would have made this a "six star" article.  

Outstanding article overall!  Thanks for sharing it with us.

I liked it very much. I would like to see more articles like this on the near future. My only "negative" comment is I was reading the article and "copy and paste" the source code and it has a lot typos, here is the code will run:

SELECT
  Cust.country ,
  Cat.CategoryName ,
  Count(*) AS ProductsBought
 FROM dbo.Customers AS Cust
 INNER JOIN dbo.Orders AS Ord
  ON Cust.CustomerId = Ord.CustomerId
 INNER JOIN dbo.[Order Details] AS OrdDet
  ON Ord.OrderId = OrdDet.OrderId
 INNER JOIN dbo.Products AS Prod
  ON OrdDet.ProductId = Prod.ProductId
 INNER JOIN dbo.Categories AS Cat
  ON Prod.CategoryId = Cat.CategoryId
 GROUP BY
  Cust.Country ,
  Cat.CategoryName
 WITH ROLLUP;

SELECT Cust.country ,
       Cat.CategoryName ,
       Count(*) AS ProductsBought
 FROM dbo.Customers AS Cust
   INNER JOIN dbo.Orders AS Ord
      ON Cust.CustomerId = Ord.CustomerId
   INNER JOIN dbo.[Order Details] AS OrdDet
      ON Ord.OrderId = OrdDet.OrderId
   INNER JOIN dbo.Products AS Prod
      ON OrdDet.ProductId = Prod.ProductId
   INNER JOIN dbo.Categories AS Cat
      ON Prod.CategoryId = Cat.CategoryId
 GROUP BY
   Cust.Country ,
 Cat.CategoryName
order by Cust.country;

GROUP BY clause.
SELECT
  Cust.country ,
  Cat.CategoryName ,
  Count(*) AS ProductsBought,
  GROUPING(Cust.country) AS SubtotalForCountries,
  GROUPING(Cat.CategoryName) AS SubtotalForCategories
 FROM dbo.Customers AS Cust
 INNER JOIN dbo.Orders AS Ord
  ON Cust.CustomerId = Ord.CustomerId
 INNER JOIN dbo.[Order Details] AS OrdDet
  ON Ord.OrderId = OrdDet.OrderId
 INNER JOIN dbo.Products AS Prod
  ON OrdDet.ProductId = Prod.ProductId
 INNER JOIN dbo.Categories AS Cat
  ON Prod.CategoryId = Cat.CategoryId
 GROUP BY
  Cust.Country ,
  Cat.CategoryName
 WITH CUBE ;

SELECT
  Prod.ProductId ,
  Prod.ProductName ,
  Cat.CategoryName ,
  Prod.UnitsInStock * Prod.UnitPrice AS StockValue
 FROM dbo.Products AS Prod
 INNER JOIN dbo.Categories AS Cat
  ON Prod.CategoryId = Cat.CategoryId
 ORDER BY
  Cat.CategoryName
 COMPUTE SUM(Prod.UnitsInStock * Prod.UnitPrice);

SELECT
  Prod.ProductId ,
  Prod.ProductName ,
  Cat.CategoryName ,
  Prod.UnitsInStock * Prod.UnitPrice AS StockValue
 FROM dbo.Products AS Prod
 INNER JOIN dbo.Categories AS Cat
  ON Prod.CategoryId = Cat.CategoryId
 ORDER BY
  Cat.CategoryName
 COMPUTE SUM(Prod.UnitsInStock * Prod.UnitPrice) 
   BY Cat.CategoryName;

SELECT
  Prod.ProductId ,
  Prod.ProductName ,
  Cat.CategoryName ,
  Prod.UnitsInStock * Prod.UnitPrice AS StockValue
 FROM dbo.Products AS Prod
 INNER JOIN dbo.Categories AS Cat
  ON Prod.CategoryId = Cat.CategoryId
 ORDER BY Cat.CategoryName
 COMPUTE SUM(Prod.UnitsInStock * Prod.UnitPrice)
   BY Cat.CategoryName
 COMPUTE SUM(Prod.UnitsInStock * Prod.UnitPrice);

One more time.... Great article.

A very good explaination!