Custom Operations

This example shows how to fuse a small arithmetic expression into one CPU worker using cpu.custom_operation.

Expression

(input - dark) / flat

Files

• examples/custom_operations/pipeline.yaml

• examples/custom_operations/run_example.py

Run it

python examples/custom_operations/run_example.py

What it demonstrates

• custom_operation kernels for small fused workflows

• named auxiliary streams for expression bindings

• a safe restricted expression language

• a compact alternative to chaining several simple arithmetic kernels

Supported expression building blocks

• elementwise +, -, *, /

• unary + and -

• matrix multiplication with @

• abs, minimum, maximum, min, max, and clip

• numeric constants

When to use it

Use a custom-operation kernel when the computation is simple, local, and easier to reason about as one expression than as a longer multi-stage graph.