NetworkX Introduction

Introduction

The primary package for analyzing network graphs in Python is NetworkX. In this lesson you'll get a brief introduction to the package, recreating the basic graphs from the previous lesson by adding nodes and edges and then creating a visual.

Objectives

You will be able to:

• Create basic network graphs using NetworkX
• Add nodes to network graphs with NetworkX
• Add edges to network graphs with NetworkX
• Visualize network graphs with NetworkX

Creating a Graph

Creating the initial graph is incredible simple. Observe:

import networkx as nx
G = nx.Graph()

Adding Nodes

From there, adding nodes is just as easy. Simply call the add_node method from you graph instance.

G.add_node("Bob")

Of course, you can also combine this with some of your previous Python prowess!

people = ["Sally", "Kate", "Jen", "Jake", "Doug"]
for person in people:
    G.add_node(person)

Adding Edges

Similarly, adding edges is also quite straightforward.

G.add_edge("Bob", "Sally")

Once again, you can also take advantage of your knowledge of python data structures to create a nested data structure and then feed these pairs into the add_edge method.

relations = {"Bob": ["Jen", "Kate"],
            "Jen": ["Bob", "Sally", "Jake", "Doug", "Kate"],
            "Doug": ["Bob"]
            }
for p1 in relations.keys():
    p2s = relations[p1]
    for p2 in p2s:
        G.add_edge(p1, p2)

Visualizing the Graph

Finally, let's take a look at visualizing this graph! The only required parameter to nx.draw() is specifying the graph itself. In addition, demonstrated below are a number of optional parameters:

• with_labels (boolean) - would you like labels for your nodes?
• node_color (color) - what color do you want your nodes?
• node_size (real) - how big do you want your nodes? (300 is default)
• alpha (real) - node transparency, must be between 0 and 1, 1 being the default
• font_weight (string) - additional formatting for the label text

%matplotlib inline
nx.draw(G, with_labels=True, node_color="#1cf0c7", node_size=1500, alpha=.7, font_weight="bold")

Additional Resources

• NetworkX General Documentation
• NetworkX Drawing Options Documentation

Summary

Well done! In this lab, you got a brief introduction to using NetworkX to create and visualize graph networks. In the upcoming lab, you'll get a chance to further practice these skills before moving on to common algorithms and metrics for processing and interpreting network graphs.