Data tests

Data tests improve the reliability of your data. In dbt, data tests are placed in the tests/ folder and return the rows that break a rule. Tests succeed when no rows are returned. As with the models, dbt-polars lets you write a test in SQL or in Python.

Writing a test

The test below fails when a daily average temperature is below -20°C or above 40°C:

Python
SQL

# quickstart/tests/meteo_daily_temperature_range.py
import polars as pl

def test(dbt, session):
    daily: pl.LazyFrame = dbt.ref("meteo_daily")
    return daily.filter((pl.col("temperature") < -20) | (pl.col("temperature") > 40))

A Python test can also return a boolean, where True means the test passed.

-- quickstart/tests/meteo_daily_temperature_range.sql
select *
from {{ ref('meteo_daily') }}
where temperature < -20 or temperature > 40

Run all tests with:

dbt test

Writing a test​

Further reading​

Writing a test

Further reading