Data Modeling & Transformation Flashcards

Which macro filters rows in an incremental model so only new data is processed on subsequent runs?

Answer: {{ is_incremental() }}

is_incremental() returns true when the model already exists, is configured incremental, and no full refresh is requested.

In an incremental model's SQL, which relation does {{ this }} refer to?

Answer: The currently building model's existing table in the warehouse

{{ this }} resolves to the already-built relation of the model, used to find the max timestamp.

What does running `dbt run --full-refresh` do to an incremental model?

Answer: Drops and rebuilds it from scratch

Full refresh rebuilds incremental models as if they did not exist, ignoring is_incremental() logic.

Which config tells an incremental model how to match existing rows for updating?

Answer: unique_key

unique_key identifies rows to update or replace, depending on the incremental strategy.

Which materialization creates no object in the warehouse and inlines the model as a CTE in downstream models?

Answer: ephemeral

Ephemeral models are interpolated as CTEs into dependent models and are not built.

What does the on_schema_change config control for an incremental model?

Answer: How new or removed columns in the source query are handled

Options such as ignore, fail, append_new_columns, and sync_all_columns govern schema drift.

Which is a typical drawback of using the view materialization for a heavily queried, complex model?

Answer: It reruns the underlying query every time it is queried

Views store only the query, so each downstream query pays the compute cost.