Rated R (A Teaser)

| 1 Comment

How many lines of TSQL codes and Window Functions you need to come up with this?


With R, it takes about 2 commands:

> order > summary(order)

To give you an idea, I prepared the dataset into a view like this:

CREATE VIEW SalesDetails
 soh.OrderDate AS [Date],
 ppc.Name AS Category,
 pps.Name AS Subcat,
 pp.Name as Product,
 SUM(sd.OrderQty) AS Qty,
 SUM(sd.LineTotal) AS LineTotal
FROM Sales.SalesPerson sp
 INNER JOIN Sales.SalesOrderHeader AS soh
 ON sp.BusinessEntityID = soh.SalesPersonID
 INNER JOIN Sales.SalesOrderDetail AS sd
 ON sd.SalesOrderID = soh.SalesOrderID
 INNER JOIN Production.Product AS pp
 ON sd.ProductID = pp.ProductID
 INNER JOIN Production.ProductSubcategory AS pps
 ON pp.ProductSubcategoryID = pps.ProductSubcategoryID
 INNER JOIN Production.ProductCategory AS ppc
 ON ppc.ProductCategoryID = pps.ProductCategoryID
GROUP BY ppc.Name, soh.OrderDate, soh.SalesOrderID, ppc.name, pps.Name, pp.Name,

And, that’s it. I just fed that view to the R engine and it summarized the dataset just by using those two commands.

This is not the perfect dataset to test with R.

I’d love to dig deeper into R because of its powerful analytical features.

What is R?

“R is a language and environment for statistical computing and graphics.”

If you’re into statistics, linear/non-linear modelling, or simply want to try another tool to analyze your data warehouse, give R a shot.

If you want to start digging with R, Ted Malone ( b | t ) has a nice introduction to using R with SQL Server 2012. Get more information about R on the R Project website.

Author: Marlon Ribunal

I'm here to learn and share things about data and the technologies around data.