Python: NetworkX

NetworkX: Examining and Removing Graph Elements

Examining graph elements

NetworkX provides four basic properties to facilitate examining graphs.

1. G.nodes - to view nodes

2. G.edges - to view edges

3. G.adj - to view neighbor nodes of each node

4. G.degree - to view degree of each node

These provide a continually updated read-only view into the graph structure and help us understand the underlying network and its properties.

Now, we will create a sample graph using the below code.

import networkx as nx graph_edges = [(0, 2), (0, 3), (0, 4), (1, 2), (1, 3), (1, 4), (2, 4), (3, 4)] G = nx.Graph() G.add_edges_from(graph_edges)

The graph we just created looks like the one below.

Now we will apply the four properties we discuss earlier to the above graph.

import networkx as nx graph_edges = [(0, 2), (0, 3), (0, 4), (1, 2), (1, 3), (1, 4), (2, 4), (3, 4)] G = nx.Graph() G.add_edges_from(graph_edges) print('Nodes:', end=' ') print(G.nodes) print('Edges:', end=' ') print(G.edges) print('Neighbors of 0:', end=' ') print(list(G.adj[0])) print('Degrees of nodes:', end=' ') print(G.degree)

The output is as below.

Nodes: [0, 2, 3, 4, 1]

Edges: [(0, 2), (0, 3), (0, 4), (2, 1), (2, 4), (3, 1), (3, 4), (4, 1)]

Neighbors of 0: [2, 3, 4]

Degrees of nodes: [(0, 3), (2, 3), (3, 3), (4, 4), (1, 3)]

Removing graph elements

Removing nodes and edges of a graph are done in the same way as adding.

We can use remove_node(), remove_nodes_from(), remove_edge() and remove_edges_from() methods to remove nodes and/or edges from a graph.

We can remove the edge (2,4) from the earlier graph using the below code.

import networkx as nx graph_edges = [(0, 2), (0, 3), (0, 4), (1, 2), (1, 3), (1, 4), (2, 4), (3, 4)] G = nx.Graph() G.add_edges_from(graph_edges) G.remove_edge(2, 4)

The graph looks like this

Similarly, we can remove the node 4 using the below code.

import networkx as nx graph_edges = [(0, 2), (0, 3), (0, 4), (1, 2), (1, 3), (1, 4), (2, 4), (3, 4)] G = nx.Graph() G.add_edges_from(graph_edges) G.remove_node(4)

The graph then looks like this