dbt-expectations for data quality - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of a dbt-expectations test on null values

Given the following dbt-expectations test configuration, what will be the result when run on a table where 5% of the rows have null values in the email column?

tests:
  - dbt_expectations.expect_column_values_to_not_be_null:
      column: email
      mostly: 0.95

A Test passes because 95% of <code>email</code> values are not null

B Test fails because any null value causes failure

C Test passes only if no null values exist

D Test fails because <code>mostly</code> parameter is ignored

💡 Hint

The mostly parameter allows some exceptions up to the given threshold.

The expect_column_values_to_not_be_null test with mostly: 0.95 means the test passes if at least 95% of the values are not null. Since 5% are null, the test passes.

❓ data_output

intermediate

Resulting data from a uniqueness test

You run the following dbt-expectations test on the user_id column:

tests:
  - dbt_expectations.expect_column_values_to_be_unique:
      column: user_id

What is the expected output if the table contains duplicate user_id values?

A Test passes and returns all rows

B Test passes but returns duplicate rows

C Test fails but returns no rows

D Test fails and returns rows with duplicate <code>user_id</code> values

💡 Hint

Uniqueness tests identify duplicates and fail if any exist.

The uniqueness test fails if duplicates exist and returns the rows causing failure.

❓ visualization

advanced

Visualizing distribution with dbt-expectations

You want to check if the age column in your dataset follows a normal distribution using dbt-expectations. Which visualization would best help you confirm this after running the test?

A Pie chart showing percentage of null vs non-null <code>age</code> values

B Histogram of <code>age</code> values with a normal curve overlay

C Scatter plot of <code>age</code> vs <code>user_id</code>

D Bar chart of counts of unique <code>age</code> values

💡 Hint

Normal distribution is best visualized with histograms and density curves.

A histogram with a normal curve overlay shows how closely the data matches a normal distribution.

🧠 Conceptual

advanced

Understanding the use of mostly in dbt-expectations

What is the main purpose of the mostly parameter in dbt-expectations tests?

A To specify the minimum proportion of rows that must meet the expectation for the test to pass

B To set the maximum number of rows to test

C To define the exact number of rows that must fail the test

D To ignore null values in the tested column

💡 Hint

Think about tolerance for exceptions in data quality tests.

The mostly parameter sets a threshold for how many rows must satisfy the condition for the test to pass.

🔧 Debug

expert

Identifying error in a dbt-expectations test configuration

You wrote this test in your dbt model:

tests:
  - dbt_expectations.expect_column_values_to_be_in_set:
      column: status
      value_set: ['active', 'inactive', 'pending']

When running dbt, you get an error. What is the cause?

A The parameter name should be <code>value_set</code> not <code>value_set</code> (typo in option name)

B The correct parameter name is <code>value_set</code> but it should be <code>value_set</code> (case sensitive)

C The correct parameter name is <code>value_set</code> but it should be <code>value_set</code> (wrong parameter name)

D The parameter name should be <code>value_set</code> but it must be <code>value_set</code> as a string, not a list

💡 Hint

Check the exact parameter name required by dbt-expectations for this test.

The correct parameter name is value_set, but the test uses value_set which is invalid and causes an error.