Building a DAG of models in dbt - Mini Project: Build & Apply

Building a DAG of models

📖 Scenario: You work in a company that uses dbt to organize data transformations. You want to build a simple Directed Acyclic Graph (DAG) of models where one model depends on another. This helps keep your data organized and updated in the right order.

🎯 Goal: Create two dbt models where the second model depends on the first. This will build a DAG of models showing how data flows from one step to the next.

📋 What You'll Learn

Create a base model with a simple SQL SELECT statement

Create a second model that selects from the first model using the ref() function

Use the ref() function to define dependencies between models

Run the models to see the DAG in action

💡 Why This Matters

🌍 Real World

In real companies, dbt helps organize complex data transformations by building DAGs of models. This keeps data pipelines clear and maintainable.

💼 Career

Data engineers and analysts use dbt to build reliable data workflows. Knowing how to build DAGs is essential for managing dependencies and ensuring data quality.

1

Create the base model

Create a dbt model file called base_customers.sql with a SQL query that selects id and name from a table called raw_customers. Write the exact SQL: select id, name from raw_customers.

dbt

-- Create the base model SQL query
-- Your code here

Need a hint?

This model is the starting point. It just selects data from the raw source table.

2

Create a dependent model

Create a second dbt model file called active_customers.sql. Use the ref() function to select all columns from the base_customers model. Write the exact SQL: select * from {{ ref('base_customers') }}.

dbt

-- Create the dependent model SQL query
-- Your code here

Need a hint?

This model depends on the base model. The ref() function tells dbt about this dependency.

3

Add a filter to the dependent model

Modify the active_customers.sql model to select only customers with id greater than 100. Use the exact SQL: select * from {{ ref('base_customers') }} where id > 100.

dbt

-- Modify the dependent model to filter customers with id > 100
-- Your code here

Need a hint?

Adding a filter shows how you can transform data in dependent models.

4

Print the DAG structure

Run the dbt command to show the DAG graph. Write the exact command: dbt ls --select active_customers --output graph.

dbt

# Run the command to show the DAG graph
# Your code here

Need a hint?

This command lists the models and shows their dependencies as a graph.