dbt-expectations for data quality - Mini Project: Build & Apply

Data Quality Checks with dbt-expectations

📖 Scenario: You work as a data analyst in a retail company. You want to make sure the sales data you use is clean and reliable before making reports. Using dbt-expectations, you will add simple data quality tests to your sales data model.

🎯 Goal: Build a dbt model for sales data and add dbt-expectations tests to check for missing values and valid ranges. Finally, run the tests and see the results.

📋 What You'll Learn

Create a dbt model file with sample sales data

Configure the model to materialize as a table

Write dbt-expectations tests to check for nulls and valid sales amounts

Run the tests and print the test results

💡 Why This Matters

🌍 Real World

Data analysts and engineers use dbt-expectations to automate data quality checks and catch errors early in data pipelines.

💼 Career

Knowing how to write and run data quality tests with dbt-expectations is a valuable skill for data engineers and analytics engineers.

1

Create a dbt model with sample sales data

Create a dbt model file called sales.sql with a select statement that returns these exact rows: order_id 1, 2, 3; customer_id 101, 102, 103; and sales_amount 100, NULL, 250.

dbt

-- Create a dbt model called sales.sql with the specified rows
-- Your code here

Hint

Use a select * from (values ...) statement with the exact rows and column names.

2

Enable dbt-expectations in the model config

Add a config block at the top of sales.sql to set materialized to table.

dbt

{% raw %}{% config(
  materialized='table'
) %}{% endraw %}

select * from (values
  (1, 101, 100),
  (2, 102, null),
  (3, 103, 250)
) as sales(order_id, customer_id, sales_amount)

-- Add the config block above the select statement
-- Your code here

Hint

Use the {% config() %} Jinja block at the top of the model file.

3

Add dbt-expectations tests for data quality

Create a test file called test_sales.yml with dbt-expectations tests: one to check sales_amount has no nulls using expect_column_values_to_not_be_null, and one to check sales_amount values are between 0 and 1000 using expect_column_values_to_be_between.

dbt

# Create test_sales.yml with dbt-expectations tests
# Your code here

Hint

Use YAML format with models and tests keys. Add the two dbt-expectations tests exactly as shown.

4

Run dbt tests and print the results

Run dbt test --select sales in your terminal and print the summary line that shows how many tests passed and failed.

dbt

# Run 'dbt test --select sales' in terminal
# Print the summary line below
# Your code here

Hint

Run the dbt test command and print the exact summary line showing test results.

Practice

1. What is the main purpose of dbt-expectations in data projects?

easy

A. To create visual dashboards for data

B. To load data into the database

C. To write tests that check data quality in your models

D. To write SQL queries faster

Solution

1. Step 1: Understand dbt-expectations role

   dbt-expectations is a package that helps add tests to check data quality.

2. Step 2: Identify main function

   Its main function is to write simple tests like checking for nulls or uniqueness in data models.

3. Final Answer:

   To write tests that check data quality in your models -> Option C

4. Quick Check:

   dbt-expectations = data quality tests [OK]

Hint: Remember: dbt-expectations is for testing data quality [OK]

Common Mistakes:

• Confusing dbt-expectations with visualization tools
• Thinking it loads or transforms data
• Assuming it speeds up SQL writing

2. Which of the following is the correct way to add a null check test using dbt-expectations in a schema.yml file?

easy

A. tests: - expect_column_values_to_be_null

B. tests: - dbt_expectations.expect_column_values_to_not_be_null

C. tests: - dbt-expectations.expect_column_values_to_not_be_null

D. tests: - expect_column_values_to_not_be_null

Solution

1. Step 1: Recall dbt-expectations test syntax

   In schema.yml, tests are listed by their function name without the package prefix.

2. Step 2: Identify correct test name

   The correct test for checking no nulls is expect_column_values_to_not_be_null.

3. Final Answer:

   tests: - expect_column_values_to_not_be_null -> Option D

4. Quick Check:

   Test names exclude package prefix [OK]

Hint: Use test function names without package prefix in schema.yml [OK]

Common Mistakes:

• Including package name in test list
• Using wrong test name like 'expect_column_values_to_be_null'
• Using hyphens instead of underscores

3. Given this snippet in a schema.yml file:

tests:
  - expect_column_values_to_be_between:
      min_value: 10
      max_value: 100

What does this test check for the column?

medium

A. Values are between 10 and 100 inclusive

B. Values are unique

C. Values are not null

D. All values are exactly 10 or 100

Solution

1. Step 1: Understand the test name

   The test expect_column_values_to_be_between checks if values fall within a range.

2. Step 2: Interpret parameters

   Parameters min_value: 10 and max_value: 100 mean values must be between 10 and 100, including these limits.

3. Final Answer:

   Values are between 10 and 100 inclusive -> Option A

4. Quick Check:

   Between test checks range limits [OK]

Hint: Between test checks inclusive range by default [OK]

Common Mistakes:

• Thinking values must be exactly min or max
• Confusing with null or uniqueness tests
• Assuming exclusive range instead of inclusive

4. You wrote this test in your schema.yml:

tests:
  - expect_column_values_to_be_unique_or_nulls

But dbt throws an error when running tests. What is the likely cause?

medium

A. The test name is incorrect; it should be expect_column_values_to_be_unique_or_nulls

B. The test name is incorrect; it should be expect_column_values_to_be_unique_or_null

C. The test requires parameters min_value and max_value

D. The test cannot be used in schema.yml files

Solution

1. Step 1: Check test name spelling

   The correct test name is expect_column_values_to_be_unique_or_null without the trailing 's'.

2. Step 2: Understand test usage

   This test does not require parameters and is valid in schema.yml files.

3. Final Answer:

   The test name is incorrect; it should be expect_column_values_to_be_unique_or_null -> Option B

4. Quick Check:

   Exact test names matter for dbt-expectations [OK]

Hint: Check exact test function names for typos [OK]

Common Mistakes:

• Adding extra characters to test names
• Assuming parameters are needed for all tests
• Thinking some tests can't be in schema.yml

5. You want to ensure a column email in your users model has no nulls and all values match a simple email pattern using dbt-expectations. Which test setup in schema.yml is correct?

hard

A. tests:
- expect_column_values_to_not_be_null
- expect_column_values_to_match_regex:
regex: '^[\w.-]+@[\w.-]+\.\w+$'

B. tests:
- expect_column_values_to_be_null
- expect_column_values_to_match_regex:
regex: '^[\w.-]+@[\w.-]+\.\w+$'

C. tests:
- expect_column_values_to_not_be_null
- expect_column_values_to_be_unique

D. tests:
- expect_column_values_to_not_be_null
- expect_column_values_to_match_regex:
regex: '[e-mail]'

Solution

1. Step 1: Identify null check test

   Use expect_column_values_to_not_be_null to ensure no null emails.

2. Step 2: Identify regex pattern test

   Use expect_column_values_to_match_regex with a regex matching email format.

3. Step 3: Evaluate options

   tests: - expect_column_values_to_not_be_null - expect_column_values_to_match_regex: regex: '^[\w.-]+@[\w.-]+\.\w+$' correctly combines both tests with a valid regex pattern.

4. Final Answer:

   tests: - expect_column_values_to_not_be_null - expect_column_values_to_match_regex: regex: '^[\w.-]+@[\w.-]+\.\w+$' -> Option A

5. Quick Check:

   Null check + regex pattern = correct test setup [OK]

Hint: Combine null check and regex for pattern validation [OK]

Common Mistakes:

• Using expect_column_values_to_be_null instead of not null
• Using a fixed string instead of regex pattern
• Omitting the regex test for pattern check