Networkx: colour nodes differently only when certain attribute determining the colour is present-CodePudding

While I am building a networkx graph, I algorithmically sometimes add a custcol attribute only to a few of them such as:

import networkx as nx
import matplotlib.pyplot as plt
G = nx.DiGraph()
G.add_edges_from([
    ('A','B'),
    ('B','C'),
    ('C','D'),
    ('D','E'),
    ('F','B'),
    ('B','G'),
    ('B','D'),
])

# in real life the following would be an algorithm deciding if the node
# should be custom coloured, and which colour it should get
G.nodes['C']['custcol'] = 'red' # simple setting for the example

# now let's explore the created example nodes
for node in G.nodes(data=True):
    print(node)

which would print out:

('A', {})
('B', {})
('C', {'custcol': 'red'})
('D', {})
('E', {})
('F', {})
('G', {})

I am now displaying the network with a single draw such as:

NXDOPTS = {
    "node_color": "orange",
    "edge_color": "powderblue",
    "node_size": 400,
    "width": 2,
}
nx.draw(G, with_labels=True, **NXDOPTS)

which would generate the following picture: What would be the best/pythonic/more efficient way of drawing node 'C' (in this example) with the colour of its custcol attribute? Of course this in reality will need to be applied to a few dozen nodes with a few different colours that are decided case by base when creating them.

CodePudding user response：

You can loop through the node data, and create a list of corresponding colors:

NXDOPTS = {
     "node_color": [data["custcol"] if "custcol" in data else "orange" for _, data in G.nodes(data=True)],
     "edge_color": "powderblue",
     "node_size": 400,
     "width": 2,
}
nx.draw(G, with_labels=True, **NXDOPTS)