Learning SQL: UNION and UNION ALL

What are UNION and UNION ALL?

The UNION and UNION ALL operators in SQL are used to combine the results of two or more SELECT queries into a single result set. They allow you to merge data from different tables or queries, making them essential for reporting and data analysis.

• UNION combines results and removes duplicate rows.
• UNION ALL combines results but keeps all rows, including duplicates.

Think of them as tools for stacking datasets vertically—like appending one list to another.

How Do UNION and UNION ALL Work?

Both operators require that the queries being combined have the same number of columns and compatible data types. The key difference is in how they handle duplicates:

• UNION performs a distinct operation, eliminating duplicate rows.
• UNION ALL does not remove duplicates, making it faster since it skips the deduplication step.

Basic Syntax:

-- Using UNION (removes duplicates)
SELECT column1, column2 FROM table1
UNION
SELECT column1, column2 FROM table2;

-- Using UNION ALL (keeps duplicates)
SELECT column1, column2 FROM table1
UNION ALL
SELECT column1, column2 FROM table2;

Example: Combining Customer Data

Imagine you have two tables:

• customers_east (stores customers from the East region)
• customers_west (stores customers from the West region)

You want to create a single list of all customers.

Using UNION (No Duplicates)

SELECT customer_id, customer_name FROM customers_east
UNION
SELECT customer_id, customer_name FROM customers_west;

What Happens? If a customer exists in both tables, only one copy appears in the result.

Using UNION ALL (All Rows, Including Duplicates)

SELECT customer_id, customer_name FROM customers_east
UNION ALL
SELECT customer_id, customer_name FROM customers_west;

What Happens? If a customer exists in both tables, both copies appear in the result.

When to Use UNION vs. UNION ALL?

Scenario	Use	Reason
Need unique records only	UNION	Removes duplicates
Want all records (faster)	UNION ALL	No duplicate check, better performance
Combining similar datasets	UNION ALL	If duplicates are acceptable or unlikely

Key Points to Remember

1. Column Matching: Queries must have the same number of columns with compatible data types.

2. Performance: UNION ALL is faster than UNION because it doesn’t remove duplicates.

3. Ordering: If you need sorted results, add ORDER BY at the end of the last query.

   SELECT name FROM employees
   UNION ALL
   SELECT name FROM contractors
   ORDER BY name;

4. Use in Complex Queries: You can combine UNION with WHERE, GROUP BY, and other clauses.

Recap

• UNION merges results and removes duplicates.
• UNION ALL merges results and keeps duplicates.
• Both require matching columns and data types.
• Use UNION ALL when duplicates don’t matter for better performance.

Resources

PostgreSQL UNION

Thank you!

Thank you for your time and for reading this!