NetworkX Cheat Sheet

A Programming › Python cheat sheet in English by gonz95alo

NetworkX library useful functions

12,644 views 1,860 downloads 7 Sep 2022 History

PRs

The basics

Import library	`import networx as nx`
Import matplotlib for graph drawing	`import matplotlib.pyplot as plt`
Initialize (Di)graphs	`G = nx.Graph(), G = nx.DiGraph()`
Create a (Di)graph from another graph	`G = nx.Graph(H), G = nx.DiGraph(H)`
Copy existing graphs	`G2 = G1.copy()`

Unweighted node and edge creation/deletion

Add nodes	`G.add_node(1) G.add_nodes_from(['a', 'b'])`
Add (directed) edges	`G.add_edge(1, 2) G.add_edges_from([(1, 'a'), (1, 'c')])`
Delete nodes	`G.remove_node(1) G.remove_nodes_from(['a', 'b'])`
Delete edges	`G.remove_edge(1, 2) G.remove_edges_from([(1, 'a'), (1, 'c')])`

If G is directed, then these take into account edge direction.

Weights and attributes

Add nodes with attribute	`G.add_nodes_from([1, 2], color = 'blue')`
Change node attributes	`G.nodes[1]['color'] = 'red'`
Add edges with attribute	`G.add_edges_from([(1,2), (2,4)], weight=3)`
Change edge attributes	`G.edges[1,2]['weight'] = 2`
Set attributes from dictionary	`nx.set_node_attributes(G, {1: {'city': 'Madrid'}})`

Attributes can have any name (color, timestamp, weight, etc). However for edge weights they should have 'weight', as many functions for weighted graphs assume it.

Basic graph properties

Number of nodes	`N = len(G.nodes())`
Number of edges	`E = len(G.edges())`
Node-degree dictionary	`degreedict = G.degrees`
Directedness	`G.is_directed(), G.is_undirected()`
Planarity	`G.is_planar()`
Diameter, radius	`d = nx.diameter(G), r = nx.radius(G)`
Average shortest path length	`aspl = nx.average_shortest_path_length(G)`

Iterate over a graph

Iterate over nodes/edges	`nodelist = [n for n in G.nodes()] edgelist = [n for n in G.edges()]`
Iterate over node/edge attributes	`node_attrs = [n for n in G.nodes(data=True)] edge_attrs = [n for n in G.edges(data=True)]`
Iterate over in/out edges	`in_list = [e for e in G.in_edges()] out_list = [e for e in G.out_edges()]`
Iterate over node's neighbors	`neighs = [n for n in G.neighbors(node)]`
Iterate over successors/predecessors	`succ = [n for n in G.successors(node)] pred = [n for n in G.predecessors(node)]`

Connectivity properties

Connectedness	`G.is_connected() G.is_weakly_connected(), G.is_strongly_connected()`
Contains node or edge?	`G.has_node(node), G.has_edge(*edge)`
List of connected components	`cc = nx.connected_components(G) wcc = nx.weakly_connected_components(G) scc = nx.strongly_connected_components(G)`
Largest (simply) connected component	`largest_cc = max(nx.connected_components(G), key=len) LCC = G.subgraph(largest_cc).copy()`

Some matricial representations

Adjacency matrix (standard)	`A = nx.to_numpy_array(G, dtype=int)`
Adjacency matrix (sparse)	`A = nx.to_scipy_sparse_array(G)`
Adjacency matrix eigenvalues	`eigs = nx.adjacency_spectrum(G)`
Graph from adjacency matrix	`G = nx.from_numpy_array(A) G = nx.from_scipy_sparse_array(A)`
Incidence matrix (sparse)	`I = nx.incidence_matrix(G)`
Laplacian matrix (sparse)	`L = nx.laplacian_matrix(G)`
Laplacian matrix eigenvalues	`eigs = nx.laplacian_spectrum(G)`
Google matrix (standard)	`Goog = nx.google_matrix(G)`

Sparse matrices/arrays are memory efficient. See the Scipy page on them for their methods, or convert them to numpy arrays with sparse_array.todense()

Graph plotting

Plotting command, standard options	`nx.draw(G, pos=pos, with_labels=True, ax=axis )`
Node positioning options (I)	`pos = [nx.circular_layout(G), nx.planar_layout(G), nx.random_layout(G)]`
Node positioning options (II)	`pos = [nx.shell_layout(G), nx.spring_layout(G), nx.spiral_layout(G)]`
Node colors and sizes in nx.draw()	`node_color = colorlist node_size = sizelist`
Edge colors and widths in nx.draw()	`edge_color = colorlist width = widthlist`
Node fine-tuning	`nx.draw_networkx_nodes(G, other options)`
Edge fine-tuning	`nx.draw_networkx_edges(G, other options)`

Centrality measures

Degree	`C = nx.degree_centrality(G)`
Betweenness	`C = nx.betweeness_centrality(G)`
Closeness	`C = nx.closeness_centrality(G)`
Eigenvector	`C = nx.eigenvector_centrality_numpy(G)`
Katz	`C = nx.katz_centrality_numpy(G, alpha=0.1, beta=1.0)`
PageRank	`C = nx.pagerank(G, alpha=0.1, personalization=None)`
HITS	`C = nx.hits(G)`

All measures return a dictionary {node:value}. They do not take into account weights, one needs to provide an extra argument weight='weight' in order to consider them.

Undirected graph generators

Path graph	`G = nx.path_graph(N)`
Cycle graph	`G = nx.cycle_graph(N)`
Star graph	`G = nx.star_graph(N)`
Complete graph	`G = nx.complete_graph(N)`
Erdös-Renyi (random)	`G = nx.erdos_renyi_graph(N, p)`
Barabasi-Albert (scale-free)	`G = nx.barabasi_albert_graph(N, m)`
Watts-Strogatz (small world)	`G = nx.watts_strogatz_graph(N, p)`

Badges

Downloaded from Unusual Locations × 1

Comments

No comments yet.