LAG function for previous row access in SQL - Step-by-Step Execution

Concept Flow - LAG function for previous row access

Start with ordered rows

↓

For each row

↓

Look at previous row in order

↓

Return previous row's value

↓

If no previous row, return NULL or default

↓

Output current row with previous value

↓

Repeat for all rows

↓

End

The LAG function looks at the previous row's value in a defined order and returns it for the current row, or NULL if there is no previous row.

Execution Sample

SQL

SELECT id, sales, LAG(sales) OVER (ORDER BY id) AS prev_sales
FROM sales_data;

This query returns each row's sales and the sales value from the previous row ordered by id.

Execution Table

Step	Current Row (id)	Current sales	Previous Row sales (LAG)	Output Row (id, sales, prev_sales)
1	1	100	NULL	(1, 100, NULL)
2	2	150	100	(2, 150, 100)
3	3	200	150	(3, 200, 150)
4	4	130	200	(4, 130, 200)
5	5	170	130	(5, 170, 130)
6	End	-	-	No more rows, stop

💡 Reached end of rows; no more previous rows to access.

Variable Tracker

Variable	Start	After Row 1	After Row 2	After Row 3	After Row 4	After Row 5	Final
Current Row id	N/A	1	2	3	4	5	End
Current sales	N/A	100	150	200	130	170	N/A
LAG(sales)	N/A	NULL	100	150	200	130	N/A

Key Moments - 3 Insights

Why does the first row's LAG value return NULL?

Because there is no previous row before the first row, the LAG function returns NULL as shown in execution_table step 1.

How does the LAG function know which row is previous?

It uses the ORDER BY clause inside the OVER() clause to define the order, so the previous row is the one with the immediately smaller id, as seen in the execution_table.

What happens if the order is changed?

The previous row changes according to the new order, so the LAG values will reflect the previous row in that new order, changing the output rows accordingly.

Visual Quiz - 3 Questions

Look at the execution_table, what is the LAG(sales) value at step 3?

A 200

B 150

C 100

D NULL

💡 Hint

Check the 'Previous Row sales (LAG)' column at step 3 in the execution_table.

At which step does the LAG function first return a non-NULL value?

A Step 3

B Step 1

C Step 2

D Step 4

💡 Hint

Look at the 'Previous Row sales (LAG)' column in execution_table for the first non-NULL value.

If the ORDER BY was changed to ORDER BY sales DESC, how would the LAG value for id=2 change?

A It would be 170

B It would be NULL

C It would be 150

D It would be 100

💡 Hint

Think about the new order by sales descending and which row comes before id=2 in that order.

Concept Snapshot

LAG(value) OVER (ORDER BY column) returns the previous row's value in the specified order.
If no previous row exists, it returns NULL by default.
Useful for comparing current row with previous row in ordered data.
ORDER BY inside OVER() defines the row sequence.
Can specify default value as third argument to LAG.
Helps analyze trends or differences between rows.

Full Transcript

The LAG function in SQL lets you access the value from the previous row in a result set ordered by a column. It works by looking at each row in order and returning the value from the row before it. If there is no previous row, it returns NULL. For example, if you have sales data ordered by id, LAG(sales) will give you the sales from the previous id. This is useful to compare current and previous values easily. The order is important because it defines which row is considered previous. Changing the order changes the LAG results. Beginners often wonder why the first row returns NULL and how the previous row is determined. The execution table shows step by step how each row's previous value is found. This helps understand the flow and output clearly.