Python networkx graph labels

Question

I have two data frames that I am using to create a graph with networkx in Python. The dataframe df1(node coordinates) and df2(edge info), look as such:

    location     x      y
0   The Wall     145    570
2   White Harbor 140    480

and

    location    x             y 
56  The Wall    Winterfell    259 
57  Winterfell  White Harbor  247

This is the code I implemented to try and graph it:

plt.figure()
G=nx.Graph()

for i, x in enumerate(df1['location']):
  G.add_node(x, pos=(df1['x'][i], df1['y'][i]))

for x, x2, w in zip(df2['location'], df2['x'], df2['y']):
  G.add_edge(x, x2, weight=w)

plt.figure(figsize=(15,15)) 

pos = nx.get_node_attributes(G, 'pos')
weights = nx.get_edge_attributes(G, 'weight') 
nx.draw(G, pos=pos, node_size=40, with_labels=True, fontsize=9)
nx.draw_networkx_edge_labels(G, pos=pos, edge_labels=weights)

plt.show()

I ran this a couple times before and it seemed to work, now however after re-opening jupyter notebook and running it again, it will not work. I mainly have two major issues.

If I try just running the line nx.draw(G, pos=pos, node_size=40, with_labels=True, fontsize=9), my graph will show up, but no labels will be shown, even with with_labels set to true.
Secondly this line nx.draw_networkx_edge_labels(G, pos=pos, edge_labels=weights) is now showing me the error can't multiply sequence by non-int of type 'float'

I have been looking at this for a couple of hours and I can't seem to fix it, any thoughts?

Edit: I can get the labels to show up if a exclude pos=pos from nx.draw, but it will not work if I include it

yatu yatu · Accepted Answer · 2019-02-20T09:12:31

The problem is that you are not specifying any pos attribute for the node Winterfell, and then when you try to access it in draw_networkx_edge_labels it doesn't find it.

If you try giving it a position attribute, say:

      location    x    y
0      TheWall  145  570
1   Winterfell  142  520
2  WhiteHarbor  140  480

Then the attributes of all nodes can be correctly accessed and the network is corectly drawn:

plt.figure()
G=nx.Graph()

df1 = df1.reset_index(drop=True)
df2 = df2.reset_index(drop=True)

for i, x in enumerate(df1['location']):
    G.add_node(x, pos=(df1.loc[i,'x'], df1.loc[i,'y']))

for x, x2, w in zip(df2['location'], df2['x'], df2['y']):
    G.add_edge(x, x2, weight=w)

plt.figure(figsize=(15,15)) 

pos = nx.get_node_attributes(G, 'pos')
weights = nx.get_edge_attributes(G, 'weight') 
nx.draw(G, pos=pos, node_size=40, with_labels=True, fontsize=9)
nx.draw_networkx_edge_labels(G, pos=pos, edge_labels=weights)

plt.show()

Python networkx graph labels

1 Answers