Write readable and high-performance queries with Window Functions

In SQL Server 2008, we gained a new and powerful feature in our bag of T-SQL tricks: Window Functions. The actual Window Function is the OVER clause, that allows you to determine partitions or “groups” in your query, before applying another function. In practice, this means you can create groups in your dataset, that can be molded further by applying functions to that groups.

In the past, I’ve written a number of posts about Window Functions in SQL Server:

Row_Number: Unique ID in select statement
Calculating Running Totals
Removing duplicates in a dataset

But there’s more!

 
Finding missing numbers in sequence
How many times did you look at a table, and you noticed that one of the values in a sequence is missing? Or a date range in a table, where someone deleted a few records. So all of a sudden, your year doesn’t have 365 days (or 366 in a leap year, or 365.242199 days which is the years average), but 315 days. There go your management reports that are calculated on a per day average. So how do you find the missing dates, without having to write a cursor or create a full blown date table like your BI colleagues do? You guessed right: a query with Window Function!

Let’s start with declaring a table, and insert a sequence with missing values:

DECLARE @Sequence TABLE
  (Number INT)


INSERT INTO @Sequence
  (Number)
VALUES
  (1),
  (10),
  (7),
  (4),
  (2),
  (8),
  (5)

 
So how can we get the “nearest numbers” from that table with just a single select statement?

SELECT
  LAG(Number, 1, 0) OVER (ORDER BY Number) AS LAG_Value,
  Number,
  LEAD(Number, 1, 0) OVER (ORDER BY Number) AS LEAD_Value
FROM @Sequence AS S
ORDER BY Number ASC

 
The LAG and LEAD are standard t-sql functions from sql server 2012 on. These functions give you the opportunity to access the previous or next row, without the need for a so called “self-join”. So what you see is the number, the record preceding that value (LAG) and the following value. So in this case, number 2 is preceded by Number 1, and followed by Number 4.

 
The lemonade stand
Now let’s look at another example. How about you? When you grew up, you wanted to save money for a new mobile phone right? In my case it was either a disc man, a Walkman, or a stereo set. But let’s stick with the modern equivalent of the Walkman for now: the MP3 player. So to earn money for the MP3 player, our fictitious friend who is called Joe, decides to start a lemonade stand. He needs to save up at least $150 to buy a new MP3 player. So every glass of lemonade he sells is accounted for, and at the end of the day he sums up all his
earnings, and puts it into a table:

DECLARE @Profit TABLE
  (DayNumber INT,
   Sales DECIMAL(10,2))


INSERT INTO @Profit
  (DayNumber, Sales)
VALUES
  (1,  6.90),
  (2,  4.17),
  (3,  2.69),
  (4,  7.26),
  (5,  2.93),
  (6,  8.98),
  (7,  7.25),
  (8,  5.88),
  (9,  1.51),
  (10, 7.97),
  (11, 3.44),
  (12, 3.76),
  (13, 9.96),
  (14, 0.92),
  (15, 8.28),
  (16, 6.05),
  (17, 9.40),
  (18, 4.03),
  (19, 9.14),
  (20, 7.25),
  (21, 0.06),
  (22, 9.12),
  (23, 7.39),
  (24, 6.57),
  (25, 4.54),
  (26, 0.09),
  (27, 4.42),
  (28, 9.53),
  (29, 5.09),
  (30, 0.89)

 
So as you can see, he earns quite a lot of money this way! But because he’s eager to buy his new MP3 player, he wants to see his day totals, and the amount he needs to buy his new toy. But because Joe is a really smart guy, he doesn’t want to do this with a lot of self-joins, and he wants his results fast. So looking at performance, what is the easiest way to query this data? How about this:

DECLARE @Goal DECIMAL(10,2) = 150.00


SELECT
  DayNumber,
  Sales,
  @Goal - SUM(Sales) OVER(ORDER BY DayNumber
                  ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS MoneyNeeded
FROM @Profit

 
He declares a “goal” variable, that is set to the amount he needs for his new gadget. So for every row, we calculate the sum of that day, plus all the sales he made from the first day (UNBOUNDED PRECEDING) and today (CURRENT ROW). After day 28 he has earned enough to buy his MP3 player. But now he wants to know what his average sales were. So he calculates the average of his sales, based on every sale he’s made so far:

SELECT
  DayNumber,
  Sales,
  AVG(Sales) OVER(ORDER BY DayNumber
                  ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS MoneyNeeded
FROM @Profit

 
So where does it stop?
So now that we’ve seen the SUM and AVERAGE option, what do we have left? How far can we take this? Thinking about it, how about a daily checkup if we hit a lowest or highest Sales amount? We can do this with the MIN and MAX option on the same query:

SELECT
  DayNumber,
  Sales,
  MIN(Sales) OVER(ORDER BY DayNumber
                  ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS LowestSale
FROM @Profit


SELECT
  DayNumber,
  Sales,
  MAX(Sales) OVER(ORDER BY DayNumber
                  ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS HighestSale
FROM @Profit

 
Now Joe can save his daily sales in the evening, and check if he had a very good, or a very bad day.

 
Not only for the average Joe…
So how can the business profit from all this? In some cases, Window Functions make it easier to output a certain resultset. In some cases it even gives you a whole new way to output data with a well performing (single) query, that was impossible until now. So if you’re running SQL Server 2008 or higher, start using (or at least start exploring) Window Functions right away!

 

To generate the random floats, I’ve used the generator of FYIcenter.com

If you want to read more about this topic, don’t forget to check out these blog posts:

Julie Koesmarno: ABC Classification With SQL Server Window Function
Mickey Stuewe: A Date At The End of The Month
Chris Yates: Windows functions who knew

Calculating Running Totals

When you work with data and database systems, eventually you need to calculate running totals on (for example) product sales or financial data. There are several methods to calculate these amounts. In this post I’ll try to show the pros and cons to the different solutions.

Let’s start with creating the resources for the examples. The most basic example I could think of, is one with only the necessary information: Date and Turnover.

CREATE TABLE Dough
	(Date DATE,
	 Turnover FLOAT)

And then insert some data:

INSERT INTO Dough
	(Date, Turnover)
VALUES
	('2011-01-01', 1000),
	('2011-02-01', 1250),
	('2011-03-01', 1500),
	('2011-04-01', 1750),
	('2011-05-01', 2000),
	('2011-06-01', 2250),
	('2011-07-01', 2250),
	('2011-08-01', 2000),
	('2011-09-01', 1750),
	('2011-10-01', 1500),
	('2011-11-01', 1250),
	('2011-12-01', 1000)

INSERT INTO Dough
	(Date, Turnover)
VALUES
	('2012-01-01', 100),
	('2012-02-01', 125),
	('2012-03-01', 150),
	('2012-04-01', 175),
	('2012-05-01', 200),
	('2012-06-01', 225),
	('2012-07-01', 225),
	('2012-08-01', 200),
	('2012-09-01', 175),
	('2012-10-01', 150),
	('2012-11-01', 125),
	('2012-12-01', 100)

With this resource, we can start on the examples.

Different solutions


When looking at this question, you’ll notice that there are more solutions to return the correct result. The following queries return the same result, but all the solutions are written for a specific version of SQL Server.

SQL 2000
If you’re using SQL Server 2000 (and I certainly hope you don’t have to anymore ;)), you can use the query with the INNER JOIN. This can be used on all SQL Server versions:

SELECT
	A.Date,
	A.Turnover,
	SUM(B.Turnover)		AS RunningTotal
FROM Dough A
INNER JOIN Dough B
	ON YEAR(B.Date) = YEAR(A.Date)
	AND B.Date <= A.Date
GROUP BY A.Date, A.Turnover
ORDER BY A.Date ASC

SQL 2005
In SQL Server 2005 they entered a new join type, called CROSS JOIN:

SELECT
	A.Date,
	A.Turnover,
	SUM(B.Turnover)		AS RunningTotal
FROM Dough A
CROSS JOIN Dough B
WHERE YEAR(B.Date) = YEAR(A.Date)
AND B.Date <= A.Date
GROUP BY A.Date, A.Turnover
ORDER BY A.Date ASC

The example with the INNER JOIN and the CROSS JOIN generate the same execution plan.

SQL 2012
With the release of SQL Server 2012 they handed us (SQL developers) a whole new “bag of tricks”. One of these “tricks” is the window function.

The first time I saw the window function, was at a Techdays NL 2012 session. This session was hosted by Bob Beauchemin (Blog | @bobbeauch). The sessions (T-SQL improvements in SQL Server 2012) is worth watching. Even if you’re using SQL Server 2012 already!

With the window function you can compute and group data, and this is done with the rows you specify.

SELECT
	Date,
	TurnOver,
	SUM(TurnOver) OVER (PARTITION BY YEAR(Date)
						ORDER BY Date ASC
						ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)		AS RunningTotals
FROM Dough

Performance


Seeing all these different solutions for the same question, I (and you probably will too) wonder about the performance of these queries. One very quick conclusion: they all return the same records ;).

When using SET STATISTICS IO, you can see the amount of disk activity generated by your statement. If you run this for the queries above, you will get the following results:

INNER JOIN:

Table ‘Dough’. Scan count 2, logical reads 25, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.

CROSS JOIN:

Table ‘Dough’. Scan count 2, logical reads 25, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.

OVER:

Table ‘Worktable’. Scan count 0, logical reads 0, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.
Table ‘Dough’. Scan count 1, logical reads 1, physical reads 0, read-ahead reads 0, lob logical reads 0, lob physical reads 0, lob read-ahead reads 0.

In the OVER query, you see a table called “Worktable”. This is an “extra object” that is generated by SQL Server because you use the OVER statement.

Conclusion


As shown above, there are several different ways to get to the same result. In this example I didn’t show you the cursor solution. This because it’s a bad practice, a bad performer, and a little bit to preserve my good name ;). If you do want to see this, please leave me a comment, and I’ll add it to this post.

But with every solution you’ll see as much discussion about reasons to use is, as discussions on why NOT to use it. And in this case, you might be bound to a specific SQL Server version, so you can’t use a specific approach.

But if you ask me for my opinion, I’ll go for the last option. Not only because I’ve got the privilege to work with SQL Server 2012 in my daily work, but also because it’s the best performer and you’ll end up with the most readable code.

I’m guessing you have a totally different opinion, so please leave a comment with your ideas and/or approaches to this challenge! Also, comments and questions are also greatly appreciated!