Connect to SEP using Python clients#

Use Starburst Enterprise platform (SEP) in conjunction with the following Python tools and libraries to create powerful data pipelines and analytics workflows:

  • PyStarburst: a library that implements the Python DataFrame API used to query and transform data in SEP clusters in a data pipeline.

  • Trino Python client: a Trino client library for the Python programming language.

  • Ibis: a Python library with a DataFrame interface for working with data across a variety of backends, including a Trino interface that supports SEP.

See the list of Python clients for further Python client support in Starburst products.

Enable PyStarburst in SEP#

PyStarburst uses the DataFrame API server-side on the coordinator and must be enabled manually. To enable PyStarburst in SEP, set the following configuration property to true in your SEP coordinator in config.properties:

dataframe-api-enabled=true

When using role-based access control such as built-in access control or Apache Ranger, the user role must have the EXECUTE privilege on the system.builtin.analyze_logical_plan entity.