About Drawing Weighted Multi-Directional Graphs with Python

I'd like to draw a weighted multi-direction graph of the edge with the coordinates of the node specified using python.
Is there any good way?

I tried drawing using networkx and pyplot, but it didn't work because the arrows overlapped.

nodes = pd.DataFrame([[[1, 3]], [[0, 2]], [[3, 3]], [[5, 2]], [[2, 4]]], columns=['coords'])
nodes.index = ['A', 'B', 'C', 'D', 'E']
edges = pd.DataFrame([['A','B', 4], ['B','A', 20], ['C','B', 3], ['D','D' ,10], ['C','A', 7], ['E','A' ,2], ['D','C' ,15], ['C','E', 8], ['D','A', 2]], columns=['start', 'end', 'weight'])

G = nx.MultiDiGraph()
G.add_nodes_from(nodes.index)
for index, row in edges.iterrows():
    G.add_edge(row.start, row.end, weight=row.weight)

nx.draw_networkx_labels(G, nodes.coords, font_size=16)
nx.draw_networkx_edge_labels(G, nodes.coords, weight=edges.weight)
nx.draw(G, nodes.coords)

plt.show()

python jupyter-notebook

2022-09-29 22:28

1 Answers

If you don't want to overlap the edges (arrows), you should use graphviz.
You can shorten the position of the edge by weight (*), but you can also express the weight by label of the edge.
I haven't checked on jupyter-notebook, but I have written it on Anaconda's Spyder.
g With graphviz, it is difficult to place nodes and edges in the desired position, and the goodness of graphviz cannot be utilized.

Below is a sample code.

import graphviz
g=graphviz.Digraph(format='svg')
g.attr('node',shape='circle')
g.attr('node', style='filed')
g.attr('node',filcolor='orange')
g.attr('node', color='orange')

for node in nodes.index:
    g.node (node, node)

for index, row in edges.iterrows():
    g.attr('edge', weight=str(row.weight))
    g.edge(row.start, row.end, label='{weight:'+str(row.weight)+'}')

print(g)
g.view()

When executing, please add it to the code of the person who asked the question.
*Please add the following to the beginning of the code of the person who asked the question to execute it.

import networkx as nx
import pandas aspd

Original image

Drawing with graphviz

2022-09-29 22:28

If you have any answers or tips

Popular Tags

python x 4647

android x 1593

java x 1494

javascript x 1427

c x 927

c++ x 878

ruby-on-rails x 696

php x 692

python3 x 685

html x 656