For loops for dynamic SQL in dbt - Mini Project: Build & Apply

For loops for dynamic SQL

📖 Scenario: You work as a data analyst using dbt to build SQL models. You want to create a dynamic SQL query that selects sales data for multiple regions without repeating code.

🎯 Goal: Build a dbt model using a for loop in Jinja to dynamically generate SQL code for multiple regions.

📋 What You'll Learn

Create a list variable called regions with the values 'north', 'south', and 'east'.

Create a variable called sales_threshold and set it to 1000.

Use a for loop with the variable region to iterate over regions and generate a SQL SELECT statement for each region.

Print the final SQL query string that combines all region queries with UNION ALL.

💡 Why This Matters

🌍 Real World

Data analysts often need to write SQL queries that repeat similar logic for different categories or regions. Using loops helps avoid repetitive code and makes maintenance easier.

💼 Career

Knowing how to generate dynamic SQL with loops is useful for dbt developers and data engineers who build scalable and maintainable data transformation pipelines.

1

Create the list of regions

Create a list variable called regions with the exact values 'north', 'south', and 'east'.

dbt

# Create the list variable regions with 'north', 'south', and 'east'
# Your code here

Hint

Use square brackets [] to create a list and separate values with commas.

2

Set the sales threshold

Create a variable called sales_threshold and set it to the integer 1000.

dbt

regions = ['north', 'south', 'east']
# Create the variable sales_threshold and set it to 1000
# Your code here

Hint

Assign the number 1000 directly to the variable sales_threshold.

3

Generate SQL queries with a for loop

Use a for loop with the variable region to iterate over regions. Inside the loop, create a list called queries that contains a SQL SELECT statement for each region selecting all columns from sales_data where region equals the current region and sales is greater than sales_threshold. Use string formatting with f-strings.

dbt

regions = ['north', 'south', 'east']
sales_threshold = 1000
# Use a for loop to create a list called queries with SQL SELECT statements for each region
# Your code here

Hint

Start with an empty list queries = []. Use a for loop to add formatted SQL strings to this list.

4

Print the combined SQL query

Use print to display the final SQL query string that joins all queries in queries with UNION ALL between them.

dbt

regions = ['north', 'south', 'east']
sales_threshold = 1000
queries = []
for region in regions:
    query = f"SELECT * FROM sales_data WHERE region = '{region}' AND sales > {sales_threshold}" 
    queries.append(query)
# Print the combined SQL query string joined by ' UNION ALL '
# Your code here

Hint

Use the join method on the string ' UNION ALL ' to combine the list queries.

Practice

1. What is the main purpose of using {% for %} loops in dbt SQL models?

easy

A. To create permanent tables in the database

B. To stop SQL queries from running

C. To write Python code inside SQL models

D. To repeat SQL code patterns dynamically and reduce manual repetition

Solution

1. Step 1: Understand the role of {% for %} in dbt

   The {% for %} loop is used to repeat code blocks dynamically in dbt SQL models.

2. Step 2: Identify the benefit of repeating code

   Repeating code patterns saves time and reduces errors by avoiding manual copy-pasting.

3. Final Answer:

   To repeat SQL code patterns dynamically and reduce manual repetition -> Option D

4. Quick Check:

   For loops = dynamic repetition [OK]

Hint: For loops repeat code blocks automatically in dbt [OK]

Common Mistakes:

• Thinking for loops create tables
• Confusing dbt loops with Python loops
• Believing loops stop queries

2. Which of the following is the correct syntax to start a for loop in a dbt SQL model?

easy

A. {% loop item in list %}

B. {{ for item in list }}

C. {% for item in list %}

D. <% for item in list %>

Solution

1. Step 1: Recall dbt Jinja syntax for loops

   dbt uses Jinja templating where loops start with {% for variable in sequence %}.

2. Step 2: Compare options to correct syntax

   Only {% for item in list %} matches the correct Jinja for loop syntax.

3. Final Answer:

   {% for item in list %} -> Option C

4. Quick Check:

   dbt for loop syntax = {% for %} [OK]

Hint: dbt loops start with {% for variable in sequence %} [OK]

Common Mistakes:

• Using {{ }} instead of {% %} for loops
• Using wrong keywords like 'loop'
• Using angle brackets instead of curly braces

3. Given the following dbt code snippet, what will be the output SQL?

{% for col in ['id', 'name', 'age'] %}SELECT {{ col }} FROM users{% if not loop.last %} UNION ALL {% endif %}{% endfor %}

medium

A. SELECT id FROM users UNION ALL SELECT name FROM users UNION ALL SELECT age FROM users

B. SELECT id, name, age FROM users

C. SELECT id FROM users SELECT name FROM users SELECT age FROM users

D. SyntaxError due to missing commas

Solution

1. Step 1: Understand the loop iteration

   The loop runs over ['id', 'name', 'age'], producing three SELECT statements.

2. Step 2: Check the UNION ALL insertion

   Between each SELECT except the last, 'UNION ALL' is added to combine results.

3. Final Answer:

   SELECT id FROM users UNION ALL SELECT name FROM users UNION ALL SELECT age FROM users -> Option A

4. Quick Check:

   Loop + UNION ALL = combined selects [OK]

Hint: Look for loop.last to add UNION ALL except last [OK]

Common Mistakes:

• Ignoring the UNION ALL between selects
• Assuming all columns selected in one query
• Thinking commas separate selects

4. Identify the error in this dbt for loop snippet:

{% for col in columns %}SELECT {{ col }} FROM users{% endfor %}

Assuming columns = ['id', 'name'].

medium

A. Wrong loop syntax with {{ }} instead of {% %}

B. Missing UNION ALL or separator between SELECT statements

C. Incorrect variable name inside the loop

D. Loop variable should be capitalized

Solution

1. Step 1: Analyze the loop output

   The loop outputs two SELECT statements back to back without any separator.

2. Step 2: Identify missing SQL syntax

   Without UNION ALL or a separator, the SQL is invalid and will cause errors.

3. Final Answer:

   Missing UNION ALL or separator between SELECT statements -> Option B

4. Quick Check:

   Multiple SELECTs need UNION ALL [OK]

Hint: Add UNION ALL between selects in loops [OK]

Common Mistakes:

• Forgetting to add UNION ALL
• Using wrong loop syntax
• Misnaming variables inside loop

5. You want to create a dynamic SQL query in dbt that selects multiple columns from a table, but only if the column name is not empty or None. Which for loop code snippet correctly achieves this?

hard

A. {% for col in columns if col %}SELECT {{ col }} FROM users{% if not loop.last %} UNION ALL {% endif %}{% endfor %}

B. {% for col in columns %}SELECT {{ col }} FROM users{% if col != '' or col != None %} UNION ALL {% endif %}{% endfor %}

C. {% for col in columns %}{% if col %}SELECT {{ col }} FROM users UNION ALL {% endif %}{% endfor %}

D. {% for col in columns %}SELECT {{ col }} FROM users UNION ALL {% endfor %}

Solution

1. Step 1: Filter columns inside the for loop

   {% for col in columns if col %}SELECT {{ col }} FROM users{% if not loop.last %} UNION ALL {% endif %}{% endfor %} uses for col in columns if col to skip empty or None values correctly.

2. Step 2: Add UNION ALL only between valid selects

   {% for col in columns if col %}SELECT {{ col }} FROM users{% if not loop.last %} UNION ALL {% endif %}{% endfor %} adds UNION ALL only if not the last valid column, avoiding trailing UNION ALL.

3. Final Answer:

   {% for col in columns if col %}SELECT {{ col }} FROM users{% if not loop.last %} UNION ALL {% endif %}{% endfor %} -> Option A

4. Quick Check:

   Filter in loop + conditional UNION ALL = correct [OK]

Hint: Use 'if col' in loop to skip empty values [OK]

Common Mistakes:

• Using incorrect if condition syntax
• Adding UNION ALL after last select
• Not filtering out empty or None columns