SQL

Windows
Function in SQL

https://www.linkedin.com/in/nivetha-anand/

1
01

What is a window function

and what does it do?

A Windows function performs calculations across a set

of rows related to the current row, unlike the aggregate
function which group rows into a single result. Windows
functions allows you to calculate values like running
totals, moving averages and ranking without grouping
rows.

https://www.linkedin.com/in/nivetha-anand/

2
 02

Key features of Windows

function:
Calculations across rows: Window functions allow you to
perform calculations that consider the context of
multiple rows, such as calculating a running total or a
moving average.

Maintain individual rows: Unlike aggregate functions,

window functions do not group rows into a single
output row; instead, they produce a result for each row
in the result set.

Use of the OVER clause: Window functions are identified

by the OVER clause, which defines the window frame
(the set of rows to which the calculation applies).

3
03

Common use cases:

Running totals: Calculate the sum of values up to the
current row.

Moving averages: Calculate the average of values within

a specific window of rows.

Rankings: Assign ranks to rows based on a specific

criteria.

Partitioning: Divide the data into partitions and perform

calculations within each partition.

https://www.linkedin.com/in/nivetha-anand/

4
04

Syntax

SELECT column_name,
FUNCTION_NAME(expression) OVER (
PARTITION BY column_name
ORDER BY column_name
ROWS/RANGE frame_specification)

https://www.linkedin.com/in/nivetha-anand/

5
 05

Syntax
🛠 Breakdown of Syntax Components:
FUNCTION_NAME(expression) → The aggregation or ranking
function applied to the selected rows (e.g., SUM(), AVG(),
RANK(), LAG(), etc.).
OVER() → Specifies the window (or subset of rows) over
which the function operates.
PARTITION BY → (Optional) Divides the result set into
smaller groups before applying the function.
ORDER BY → (Optional) Defines the order of rows within
each partition.
ROWS/RANGE frame specification → (Optional) Defines a
subset of rows within the partition for calculations like
running totals or moving averages.

https://www.linkedin.com/in/nivetha-anand/

6
 06

When and where to use some of the

below commonly used functions:
1️⃣ LEAD()
Use case: Fetch the next row’s value without using a self-join.
Example: Finding the next month’s revenue for each month in a
sales dataset.
2️⃣ LAG()
Use case: Fetch the previous row’s value, useful for comparing
trends.
Example: Finding the previous month’s revenue to calculate
month-over-month change.
3️⃣ RANK()
Use case: Assigns a rank with gaps if there are duplicate values.
Example: Ranking students by scores; if two students have the
same score, they get the same rank, but the next rank is skipped
(1,1,3…).

https://www.linkedin.com/in/nivetha-anand/

7
 07

When and where to use some of the

below commonly used functions:

4️⃣ DENSE_RANK()
Use case: Assigns a continuous rank without gaps.
Example: Similar to RANK(), but if two students tie, the next rank
follows sequentially (1,1,2…).

5️⃣ ROW_NUMBER()
Use case: Assigns a unique row number within each partition.
Example: When selecting the latest transaction per customer.

https://www.linkedin.com/in/nivetha-anand/

8
 08

Window Function Problems:

Row Number()
📌 Problem:
Assign a unique row number to employees within
each department based on salary in descending
order.

https://www.linkedin.com/in/nivetha-anand/

9
 09

Window Function Problems:

📊 DATA SET (EMPLOYEES TABLE)

https://www.linkedin.com/in/nivetha-anand/

10
 10

SELECT
employee_id,
department,
salary,
ROW_NUMBER() OVER (PARTITION BY department
ORDER BY salary DESC) AS row_num
FROM employees;

https://www.linkedin.com/in/nivetha-anand/

11
 11

Window Function Problems:

✅ EXPECTED OUTPUT

https://www.linkedin.com/in/nivetha-anand/

12
 08

Window Function Problems:

DENSE_RANK()
📌 Problem
Rank employees within each department based
on salary in descending order. If two employees
have the same salary, they should receive the
same rank, and the next rank should NOT be
skipped.

https://www.linkedin.com/in/nivetha-anand/

13
 09

Window Function Problems:

📊 DATA SET (EMPLOYEES TABLE)

https://www.linkedin.com/in/nivetha-anand/

14
06

SELECT
employee_id,
department,
salary,
DENSE_RANK() OVER (PARTITION BY
department ORDER BY salary DESC) AS
dense_rank
FROM employees;

https://www.linkedin.com/in/nivetha-anand/

15
 09

Window Function Problems:

✅ EXPECTED OUTPUT

https://www.linkedin.com/in/nivetha-anand/

16
07

Explanation:
In Sales:
The highest salary ($95,000) gets rank 1.
Two employees have $85,000, so they share rank 2.
The next salary ($72,000) gets rank 3 (no rank is
skipped).
In IT:
The highest salary ($120,000) gets rank 1.
Two employees have $110,000, so they share rank 2.
The next salary ($100,000) gets rank 3 (no rank is
skipped).
In HR:
The highest salary ($70,000) gets rank 1.
The next salary ($68,000) gets rank 2.

https://www.linkedin.com/in/nivetha-anand/

17
 08

Window Function Problems:

RANK()
📌 Problem
Rank employees within each department based
on salary in descending order. If two employees
have the same salary, they should receive the
same rank, but the next rank should be skipped.

https://www.linkedin.com/in/nivetha-anand/

18
 09

Window Function Problems:

📊 DATA SET (EMPLOYEES TABLE)

https://www.linkedin.com/in/nivetha-anand/

19
10

SELECT
employee_id,
department,
salary,
RANK() OVER (PARTITION BY department ORDER
BY salary DESC) AS rank
FROM employees;

https://www.linkedin.com/in/nivetha-anand/

20
 11

Window Function Problems:

✅ EXPECTED OUTPUT

https://www.linkedin.com/in/nivetha-anand/

21
 12

Explanation:
In Sales:
The highest salary ($95,000) gets rank 1.
Two employees have $85,000, so they share rank 2.
The next rank is 4 (rank 3 is skipped).
In IT:
The highest salary ($120,000) gets rank 1.
Two employees have $110,000, so they share rank 2.
The next salary ($100,000) gets rank 4 (rank 3 is
skipped).
In HR:
The highest salary ($70,000) gets rank 1.
The next salary ($68,000) gets rank 2.

https://www.linkedin.com/in/nivetha-anand/

22
 13

Rank() vs Dense_Rank() - Key Difference

https://www.linkedin.com/in/nivetha-anand/

23
 14

When to use which?

✅ Use RANK() if you need true ranking

positions (e.g., competition scoring, where
ties affect ranking).

✅ Use DENSE_RANK() if you want

consecutive ranking without gaps (e.g.,
salary bands, grading systems).

https://www.linkedin.com/in/nivetha-anand/

24
 Was this post
helpful to you?
Please like, share, save, comment and repost!

♻️

https://www.linkedin.com/in/nivetha-anand/
25