Window functions are T-SQL functions used to perform a calculation over a set of rows. A window function is similar to an aggregate function. The difference is that window functions do not group the results into a single output row.

Most people know how to use aggregate functions to calculate aggregate values, such as sum, average, maximum, minimum, and count. The results of these aggregate functions are grouped by specific data using GROUP BY clause.

Then how if we want to display the detail data and aggregated data side-by-side? For example, to display the EmployeeID, Gender, Age, and AVG(Age) by Gender in one row. It’s possible doing this using sub-query, but we can do that in a single query using window functions.

Let’s take a look at the following example to see the difference between aggregate function vs. window function.

Table: Students

We use the following table for our example case. Let’s calculate the average GPA of each Major using aggregate function and a window function.

Note: you can find the script to create the table and insert the data at the bottom of this post.

Basic aggregate function

Calculate average GPA by Major.

SELECT	Major, AVG(GPA) AvgGpa
FROM	Students
GROUP BY Major

Result:

As you can see in the above result, the average values are grouped by the three rows of Major.

Window function

The window function is used if we want to show the aggregate value without grouping.

For example, we want to compare the students’ GPA with the average GPA of their Major side by side. The columns we want to display are the students’ name, major, GPA, and the average GPA of each Major.

We can solve the above problem by using a window function as below.

 1 2 3 4 5 SELECT Fullname , Major , GPA , AVG(GPA) OVER(PARTITION BY Major) AvgGpa FROM Students

Result:

The average GPA of each Major is displayed in each row without grouping.

Script to create table and insert data

 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 CREATE TABLE Students ( ID int IDENTITY(1,1) NOT NULL, Fullname nvarchar(50) NULL, Age int NOT NULL, Major nvarchar(50) NULL, GPA decimal(3, 2) NULL, CONSTRAINT PK_Students PRIMARY KEY(ID) ) GO INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Olivia Mardhiyah', 20, 'Graphic Design', 3.40) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Cahyanto Saefullah', 19, 'Computer Science', 3.20) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Novi Nuraini', 22, 'Computer Science', 2.94) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Irfan Zulkarnaen', 21, 'Electrical Engineering', 2.47) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Hilda Zaenab Kuswandari', 19, 'Computer Science', 3.67) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Gunawan Raden Pradipta', 20, 'Electrical Engineering', 3.10) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Yessi Bauti Farida', 22, 'Computer Science', 3.86) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Umar Hakim', 24, 'Electrical Engineering', 2.20) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Elvina Handayani', 20, 'Graphic Design', 3.62) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Rani Purwanti', 25, 'Computer Science', 3.51) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Harjasa Wibowo', 22, 'Electrical Engineering', 3.20) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Agnes Sabrina Pertiwi', 21, 'Graphic Design', 2.82) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Luwes Kuswoyo', 20, 'Electrical Engineering', 2.80) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Bagio Semar Prasetyo', 18, 'Computer Science', 2.68) INSERT INTO Students (Fullname, Age, Major, GPA) VALUES('Koko Hidayah', 19, 'Electrical Engineering', 2.95)