Skip to main content

dbt feature compatibility

Overview of features dbt-polars already supports and plans to support in the future.

StatusMeaning
✅Supported
🟡Partially supported
🚧Not supported yet, but planned
❌️Not supported, unplanned

dbt core features​

FeatureStatusRemarks
table materialization✅
incremental materialization🟡✅ append
✅ merge
✅ delete+insert
🚧 microbatch
ephemeral materialization✅
view materialization❌️dbt-polars has no database to hold views. Use table or ephemeral instead.
Snapshots✅
Seeds✅
Sources✅
Python models✅
Data tests (generic and singular SQL)✅
Python singular tests✅
Unit tests (defined in YAML)🚧
persist_docs✅
dbt docs generate✅
dbt show✅
dbt clone❌️
Hooks❌️
Grants❌️
Model contracts🚧

dbt core extensions​

dbt-polars adds two features that are not supported in dbt-core:

  • Ephemeral Python models
  • Python data tests

Macros​

dbt-core publishes a list of cross-database macros that each warehouse should implement.

Macro groupStatusRemarks
Date/time✅
Hash❌️
Type conversion✅
Strings✅
Arrays🟡Missing:
- array_append
- array_concat
Aggregates🟡Missing: listagg with an order or limit
Set operations✅
Other✅

Unsupported macros raise a compilation error. Use a Python model for the same logic.