Guides

Welcome to the yieldgraph guides. These pages walk you through building pipelines — from a simple two-step chain all the way to multi-branch, threaded production pipelines.

• Getting Started

  ---

  Install yieldgraph and build your first pipeline in minutes. Covers sources, transforms, output, and the tuple data model.

  Getting Started

• Patterns & Recipes

  ---

  Fan-out branches, attach_to, error handling, cooperative cancellation, progress monitoring, and threaded execution.

  Patterns & Recipes

• Configuration

  ---

  Environment variables, log levels, the LoggingBehavior mixin, and the ENV / LOG constants objects.

  Configuration

---

Quick reference

from yieldgraph import Graph

# Define steps
def source(graph):
    for x in range(1, 6):
        yield x

def square(x):
    yield x ** 2

# Build
g = Graph()
g.add_chain(source, square)

# Run
g.run()

# Consume
print(g.output)
# [(1,), (4,), (9,), (16,), (25,)]

Task	How
Build a linear pipeline	g.add_chain(fn1, fn2, fn3)
Branch from a node	g.add_chain(fn, attach_to="node_name")
Run sequentially	g.run()
Run with threads	YIELDGRAPH_THREADED=1 python ...
Check success	g.succeeded
Get errors	node.errors for each node
Cancel early	graph.cancelled = True inside a node

Or from Python:

import os
os.environ['YIELDGRAPH_THREADED'] = '1'
g.run()

Cancellation

Set graph.cancelled = True at any point (e.g. from a signal handler or UI callback) and all nodes will stop after their current yield.

Accessing results

After run() completes:

g.succeeded   # True if no errors
g.output      # list of items from the terminal node(s)
g.error       # error message string, or None