Skip to main content

How to migrate your after scripts to Python dbt models

Python dbt models (or Python Data Models) are a way to include a Python transformation of data inside of your dbt DAG.

We will explore how what we were able to achieve before with write_to_source and write_to_model in after scripts is now possible more clearly with Python dbt models.

When to use a Python Data Model vs an after script

The rule-of-thumb is that if you are writing to the data warehouse, you should be using a Python Data Model.

We are deprecating the use of write_to_source and write_to_model outside of Python Data Models.


If you are already using write_to_model to enrich an existing table, you can remove said table and replace with the model.

Example commit: