SUM, AVG, COUNT as window functions in PostgreSQL - Step-by-Step Execution

Concept Flow - SUM, AVG, COUNT as window functions

Start with table rows

↓

Apply window function

↓

Partition rows if specified

↓

Order rows if specified

↓

Calculate running SUM, AVG, COUNT

↓

Add results as new columns

↓

Return result set with window function values

Window functions compute values across sets of rows related to the current row, adding results as new columns without collapsing rows.

Execution Sample

PostgreSQL

SELECT dept, emp, salary,
       SUM(salary) OVER (PARTITION BY dept) AS dept_sum,
       AVG(salary) OVER (PARTITION BY dept) AS dept_avg,
       COUNT(*) OVER (PARTITION BY dept) AS dept_count
FROM employees;

Calculates sum, average, and count of salaries per department, showing these as new columns for each employee.

Execution Table

Step	Row (dept, emp, salary)	SUM(salary) OVER (PARTITION BY dept)	AVG(salary) OVER (PARTITION BY dept)	COUNT(*) OVER (PARTITION BY dept)
1	(Sales, Alice, 5000)	15000	5000	3
2	(Sales, Bob, 7000)	15000	5000	3
3	(Sales, Carol, 3000)	15000	5000	3
4	(HR, Dave, 4000)	9000	4500	2
5	(HR, Eve, 5000)	9000	4500	2
6	(IT, Frank, 6000)	6000	6000	1
Exit	All rows processed	-	-	-

💡 All rows processed, window functions computed per partition

Variable Tracker

Variable	Start	After Row 1	After Row 2	After Row 3	After Row 4	After Row 5	After Row 6	Final
dept_sum (Sales)	0	15000	15000	15000	15000	15000	15000	15000
dept_avg (Sales)	0	5000	5000	5000	5000	5000	5000	5000
dept_count (Sales)	0	3	3	3	3	3	3	3
dept_sum (HR)	0	0	0	0	9000	9000	9000	9000
dept_avg (HR)	0	0	0	0	4500	4500	4500	4500
dept_count (HR)	0	0	0	0	2	2	2	2
dept_sum (IT)	0	0	0	0	0	0	6000	6000
dept_avg (IT)	0	0	0	0	0	0	6000	6000
dept_count (IT)	0	0	0	0	0	0	1	1

Key Moments - 3 Insights

Why does the SUM value repeat for each row in the same department?

Because window functions calculate over the partition but do not collapse rows, so each row in the partition shows the same SUM value (see execution_table rows 1-3).

How is AVG calculated without grouping rows?

AVG is computed as SUM divided by COUNT over the partition for each row, shown as a new column without changing the number of rows (see execution_table AVG column).

Why does COUNT(*) show the total number of rows in the partition for every row?

COUNT(*) counts all rows in the partition and repeats this count for each row, unlike aggregate functions that collapse rows (see execution_table COUNT column).

Visual Quiz - 3 Questions

Look at the execution_table, what is the SUM(salary) for Bob in Sales at step 2?

A 15000

B 7000

C 5000

D 3000

💡 Hint

Check the SUM(salary) column for step 2 in the execution_table.

At which step does the COUNT(*) for HR become 2?

A Step 3

B Step 4

C Step 5

D Step 6

💡 Hint

Look at the COUNT(*) column and find when HR rows start showing 2.

If a new employee with salary 4000 is added to IT, how would the AVG for IT change in variable_tracker?

A It would stay 6000

B It would become 4000

C It would become 5000

D It would become 10000

💡 Hint

AVG is total salary divided by count; adding 4000 to 6000 total over 2 rows changes average.

Concept Snapshot

SUM, AVG, COUNT as window functions:
- Use OVER() clause to apply without grouping
- PARTITION BY divides rows into groups
- Functions compute values per partition, repeated per row
- Rows stay same count, new columns added
- Useful for running totals, averages, counts per group

Full Transcript

This lesson shows how SUM, AVG, and COUNT work as window functions in PostgreSQL. Window functions calculate values across a set of rows related to the current row, defined by PARTITION BY. Unlike aggregate functions, they do not reduce the number of rows but add new columns with the computed results. For example, SUM(salary) OVER (PARTITION BY dept) calculates the total salary per department and shows it on every row in that department. AVG and COUNT work similarly, computing average salary and number of employees per department. The execution table traces each row's values step-by-step, showing how the window functions produce repeated results per partition. Key moments clarify why values repeat and how averages are computed without grouping. The quiz tests understanding of these repeated values and how changes affect averages. This helps beginners see how window functions add powerful analytics without losing row detail.