Reputation: 7951
B.add_nodes_from(a, bipartite=1)
B.add_nodes_from(b, bipartite=0)
nx.draw(B, with_labels = True)
plt.savefig("graph.png")
I am getting the following figure. How can I make it look like a proper bipartite graph?
Upvotes: 23
Views: 26134
Reputation: 1
X = {n for n, d in B.nodes(data=True) if d["bipartite"] == 0}
nx.draw_networkx(B, pos=nx.drawing.layout.bipartite_layout(B, X))
Upvotes: -1
Reputation: 1270
Inspired in mdml's answer, another way to plot a complete bipartite graph in a "classical" way:
import networkx as nx
import matplotlib.pyplot as plt
m, n = 5, 10
K = nx.complete_bipartite_graph(m, n)
pos = {}
pos.update((i, (i - m/2, 1)) for i in range(m))
pos.update((i, (i - m - n/2, 0)) for i in range(m, m + n))
fig, ax = plt.subplots()
fig.set_size_inches(15, 4)
nx.draw(K, with_labels=True, pos=pos, node_size=300, width=0.4)
plt.show()
Upvotes: 4
Reputation: 1883
NetworkX already has a function to do exactly this.
Its called networkx.drawing.layout.bipartite_layout
You use it to generate the dictionary that is fed to the drawing functions like nx.draw
via the pos
argument like so:
nx.draw_networkx(
B,
pos = nx.drawing.layout.bipartite_layout(B, B_first_partition_nodes),
width = edge_widths*5) # Or whatever other display options you like
Where B
is the full bipartite graph (represented as a regular networkx graph), and B_first_partition_nodes
are the nodes you wish to place in the first partition.
This generates a dictionary of numeric positions that is passed to the pos
argument of the drawing function. You can specify layout options as well, see the main page.
Upvotes: 18
Reputation: 31
Another example, combining graph with bipartite graph:
G = nx.read_edgelist('file.txt', delimiter="\t")
aux = G.edges(data=True)
B = nx.Graph()
B.add_nodes_from(list(employees), bipartite=0)
B.add_nodes_from(list(movies), bipartite=1)
B.add_edges_from(aux)
%matplotlib notebook
import [matplotlib][1].pyplot as plt
plt.figure()
edges = B.edges()
print(edges)
X, Y = bipartite.sets(B)
pos = dict()
pos.update( (n, (1, i)) for i, n in enumerate(X) ) # put nodes from X at x=1
pos.update( (n, (2, i)) for i, n in enumerate(Y) ) # put nodes from Y at x=2
nx.draw_networkx(B, pos=pos, edges=edges)
plt.show()
Upvotes: 3
Reputation: 22872
You could do something like this, to draw nodes from each partition at a particular x
coordinate:
X, Y = bipartite.sets(B)
pos = dict()
pos.update( (n, (1, i)) for i, n in enumerate(X) ) # put nodes from X at x=1
pos.update( (n, (2, i)) for i, n in enumerate(Y) ) # put nodes from Y at x=2
nx.draw(B, pos=pos)
plt.show()
The key is creating the dict
for the the nx.draw
pos
parameter, which is:
A dictionary with nodes as keys and positions as values.
See the docs.
Upvotes: 30