MultiDiGraph edges from networkx draw with connectionStyle

Question

Is it possible to somehow draw different edges at the same nodes with different curvatures using connectionstyle?

I wrote the following code, but I got all three edges overlapped:

import networkx as nx
import matplotlib.pyplot as plt

G = nx.MultiDiGraph()
G.add_node('n1')
G.add_node('n2')
G.add_edge('n1', 'n2', 0)
G.add_edge('n1', 'n2', 1)
G.add_edge('n1', 'n2', 2)

pos = nx.spring_layout(G)
nx.draw(G, pos, with_labels=True, connectionstyle='arc3, rad = 0.3')

plt.show()

Answer 1

This can be done by plotting each edge with a different rad argument - as shown. Note that my approach here uses f-strings which require Python 3.6 - below that you will have to build the string using a different method.

Code:

import networkx as nx
import matplotlib.pyplot as plt

G = nx.MultiDiGraph()
G.add_node('n1')
G.add_node('n2')
G.add_edge('n1', 'n2', rad=0.1)
G.add_edge('n1', 'n2', rad=0.2)
G.add_edge('n1', 'n2', rad=0.3)

plt.figure(figsize=(6,6))

pos = nx.spring_layout(G)
nx.draw_networkx_nodes(G, pos)
nx.draw_networkx_labels(G, pos)

for edge in G.edges(data=True):
    nx.draw_networkx_edges(G, pos, edgelist=[(edge[0],edge[1])], connectionstyle=f'arc3, rad = {edge[2]["rad"]}')

plt.show()

Output:

We could even create a new function to do this for us:

import networkx as nx
import matplotlib.pyplot as plt

def new_add_edge(G, a, b):
    if (a, b) in G.edges:
        max_rad = max(x[2]['rad'] for x in G.edges(data=True) if sorted(x[:2]) == sorted([a,b]))
    else:
        max_rad = 0
    G.add_edge(a, b, rad=max_rad+0.1)

G = nx.MultiDiGraph()
G.add_node('n1')
G.add_node('n2')

for i in range(5):
    new_add_edge(G, 'n1', 'n2')

for i in range(5):
    new_add_edge(G, 'n2', 'n1')

plt.figure(figsize=(6,6))

pos = nx.spring_layout(G)
nx.draw_networkx_nodes(G, pos)
nx.draw_networkx_labels(G, pos)

for edge in G.edges(data=True):
    nx.draw_networkx_edges(G, pos, edgelist=[(edge[0],edge[1])], connectionstyle=f'arc3, rad = {edge[2]["rad"]}')

plt.show()

Output: