Quick Start

Installation

Let’s first install node-graph via pip:

$ pip install node-graph

First workflow

Suppose we want to calculate (x + y) * z in two steps.

• step 1: add x and y

• step 2: then multiply the result with z.

Create task

Task is the basic building block of a workflow. One can create a task from a Python function using the decorator:

from node_graph.decorator import task


@task()
def add(x, y):
    return x + y


@task()
def multiply(x, y):
    return x * y


@task.graph()
def AddMultiply(x, y, z):
    the_sum = add(x=x, y=y).result
    return multiply(x=the_sum, y=z).result


ng = AddMultiply.build(x=1, y=2, z=3)
ng

Engines and provenance Run graphs directly in Python:

from node_graph.engine.local import LocalEngine

graph = AddMultiply.build(x=1, y=2, z=3)

engine = LocalEngine()
results = engine.run(graph)
print("results:", results)

results: {'result': 9}

Provenance for visualization

In interactive notebooks you can display the provenance graph inline

engine.recorder

Node graph programming

You can also create a graph programmatically. Three steps:

• create a empty Graph

• add nodes: add and multiply.

• link the output of the add task to one of the x input of the multiply task.

from node_graph import Graph

ng = Graph("first_workflow")
ng.add_task(add, name="add", x=2, y=3)
ng.add_task(multiply, name="multiply", y=4)
ng.add_link(ng.tasks.add.outputs.result, ng.tasks.multiply.inputs.x)
ng

What’s Next