Warning This project has been archived and is no longer actively maintained or supported. We highly recommend migrating to the official Databricks SDK for Python available at https://github.com/databricks/databricks-sdk-py.

azure-databricks-sdk-python is a Python SDK for the Azure Databricks REST API 2.0.

Easily, perform all the operations as if on the Databricks UI:

from azure_databricks_sdk_python import Client
from azure_databricks_sdk_python.types.clusters import AutoScale, ClusterAttributes

client = Client(databricks_instance="<instance>", personal_access_token="<token>")
spark_conf = {'spark.speculation': True}
autoscale = AutoScale(min_workers=0, max_workers=1)
attributes = ClusterAttributes(cluster_name="my-cluster",
                            spark_version="7.2.x-scala2.12",
                            node_type_id="Standard_D3_v2",
                            spark_conf=spark_conf,
                            autoscale=autoscale)
created = client.clusters.create(attributes)
print(created.cluster_id)

azure-databricks-sdk-python is ready for your use-case:

• Clear standard to access to APIs.
• Contains custom types for the API results and requests.
• Support for Personal Access token authentification.
• Support for Azure AD authentification.
• Support for the use of Azure AD service principals.
• Allows free-style API calls with a force mode -(bypass types validation).
• Error handeling and proxy support.

Officially supports 3.6+, and runs great on PyPy.

Please refer to the progress below:

As for specific API wrappers:

Check the documentation on readthedocs.org.