API Reference

This is the API reference for the fused Python package, which allows you to build and run serverless data workflows.

The reference is intended to be limited to low-level descriptions of various programmatic functionality. If you're just getting started with Fused, we recommend looking at the Guide first or exploring an Example.

Decorators

Function	Description
@fused.udf	Decorator to define a User Defined Function
@fused.cache	Decorator to cache function results

UDF Execution

Class/Function	Description
Udf	UDF class — call directly or use .map() for parallel execution
JobPool	Manage parallel job execution (returned by udf.map())
fused.run	[Legacy] Execute a UDF
fused.run_async	[Legacy] Execute a UDF asynchronously
fused.submit	[Legacy] Submit a UDF for parallel execution

Loading UDFs

Function	Description
fused.load	Load a UDF from file, URL, or Workbench
fused.load_async	Load a UDF asynchronously

Data Access

Function	Description
fused.get_chunk_from_table	Read a chunk from a partitioned table
fused.get_chunks_metadata	Get metadata for table chunks
fused.find_dataset	Search for datasets in the catalog
fused.register_dataset	Register a dataset in the catalog

File Operations

Function	Description
fused.download	Download files from URLs
fused.file_path	Get local file path for cached files
fused.ingest	Ingest geospatial data into optimized format
fused.ingest_nongeospatial	Ingest non-geospatial data

Configuration & Utilities

Function	Description
fused.context	Access UDF execution context
fused.options	Configure Fused runtime options
fused.secrets	Access secrets and credentials
fused.types	Type definitions for UDF parameters
fused.h3	H3 hexagonal grid utilities
fused.api	Low-level API access

Classes

Class	Description
Udf	UDF object with metadata and execution methods
JobPool	Pool for managing parallel job execution