Networkx and Matplotlib: how to access node attributes and show them as annotation

Question

I'm trying to get the on hover to work for this networkx matplotlib graph, but I'm getting a key error issue. I want to create a node with the key name as a string and have that as the label. However, it looks like the on hover function that gets called breaks at the xy=pos[node] section. Here's some output of what some of that looks like:

obj in this output is me printing the node keys.
ind is me hovering over the node.
node is me trying to figure out the link between the node key and what hover is looking for.

pos: {'A': array([-0.88308537, -0.21952651]), 'B': array([ 0.70105714, -0.78047349]), 'C': array([0.18202824, 1.        ])}

obj: A
obj: B
obj: C
ind: {'ind': array([], dtype=int32)}
ind: {'ind': array([2], dtype=int32)}
node: 2
ind: {'ind': array([2], dtype=int32)}
node: 2
ind: {'ind': array([2], dtype=int32)}

Code below:

import networkx as nx
import matplotlib.pyplot as plt
from networkx.readwrite import edgelist
import numpy as np

G = nx.DiGraph()

G.add_node("A", attr1=20)
G.add_node("B", attr1=25)
G.add_node("C", attr1=30)

fig, ax = plt.subplots()
pos = nx.spring_layout(G)
print('pos:', pos)

nodes = nx.draw_networkx_nodes(G, pos, node_size=500)

for obj in G.nodes:
    print('obj:', obj)

nx.draw_networkx_edges(G, pos, edgelist=G.edges(), edge_color='black')

annot = ax.annotate("", xy=(0,0), xytext=(20,20),textcoords="offset points",
                    bbox=dict(boxstyle="round", fc="w"),
                    arrowprops=dict(arrowstyle="->"))

annot.set_visible(False)

def update_annot(ind):
    node = ind["ind"][0]
    print('node:',node)
    xy = pos[node]
    print('here:', xy)
    annot.xy = xy
    node_attr = {'node': node}
    node_attr.update(G.nodes[node])
    text = '
'.join(f'{k}: {v}' for k, v in node_attr.items())
    annot.set_text(text)

def hover(event):
    vis = annot.get_visible()
    if event.inaxes == ax:
        cont, ind = nodes.contains(event)
        print('ind:',ind)
        if cont:
            update_annot(ind)
            annot.set_visible(True)
            fig.canvas.draw_idle()
        else:
            if vis:
                annot.set_visible(False)
                fig.canvas.draw_idle()

fig.canvas.mpl_connect("motion_notify_event", hover)
plt.show()

Error:

xy = pos[node]
KeyError: 2

It looks like it's trying to find a key with a number instead of a letter. That's why I'm getting a key error. But I'm not sure as to why I can't get it to accept a string instead.

user_12046409 · Answer

node_index = ind["ind"][0] is an integer index in the list of nodes. Networkx stores its nodes in a dictionary, using the node name. G.nodes[2] gives an error, because G.nodes is a dictionary which is indexed by node names.

You can convert G.nodes to a list, and find the name at node_index: node_name = list(G.nodes)[node_index]. With that name, you can find the attributes: node_attr = G.nodes[node_name].

To work with hover annotations, I'd suggest the mplcursors library, which hides many of the internal bookkeeping.

Here is the updated example:

import matplotlib.pyplot as plt
import networkx as nx
import mplcursors

G = nx.DiGraph()
G.add_node("A", attr1=20)
G.add_node("B", attr1=25)
G.add_node("C", attr1=30)

fig, ax = plt.subplots()
pos = nx.spring_layout(G)
nodes = nx.draw_networkx_nodes(G, pos, node_size=500, node_color='dodgerblue')

nx.draw_networkx_labels(G, pos, font_color='yellow')
nx.draw_networkx_edges(G, pos, edgelist=G.edges(), edge_color='black')

def update_annot(sel):
    node_index = sel.target.index
    node_name = list(G.nodes)[node_index]
    node_attr = G.nodes[node_name]
    text = node_name + ' ' + '
'.join(f'{k}: {v}' for k, v in node_attr.items())
    sel.annotation.set_text(text)

cursor = mplcursors.cursor(nodes, hover=True)
cursor.connect('add', update_annot)

plt.show()

Networkx and Matplotlib: how to access node attributes and show them as annotation

1 Answers