How to draw a simple relation graph in Python
Intro
The process of drawing a simple relation graph in python can be broken down into 2 steps.
- Define a graph.
- Draw a graph.
Step 1. Define a graph
In this step, we will use the networkx package.
Install tutorial
If you are using conda, you can just type conda install networkx
If you are using pip, you can just type pip install networkx
Nodes
First of all, you need to create a graph.
import networkx as nx
G = nx.Graph()You can use different ways to add nodes.
- Add one node at a time.
- Add nodes from any iterable container
- Add nodes along with node attributes. In this way, you can define many attributes of a node, such as color, size, etc.
- Add nodes from another graph directly
G.add_node(1) # method 1
G.add_nodes_from([2, 3, 4, 5]) # method 2
G.add_nodes_from([ # method 3
(6, {"color": "red"}),
(7, {"color": "blue"})
])
G2 = nx.Graph() # method 4
G2.add_nodes_from([8, 9, 10])
G.add_nodes_from(G2)
# you can verify nodes by
print(G.nodes)
# NodeView((1, 2, 3, 4, 5, 6, 7, 8, 9, 10))Edges
Also, networkx has many ways to add edges, which is quite similar to add nodes. Let’s just jump to the code🤗
G.add_edge(1, 2) # method 1
G.add_edges_from([(1, 3), (1, 4)]) # method 2
G2.add_edge(8, 9) # method 3
G2.add_edge(8, 10)
# WARNING: you can't just use G2 instead of G2.edges
G.add_edges_from(G2.edges) # method 4 Get some information about a graph
Usually, we want to know some information about a graph. For example, we may want to know the number of nodes and edges, the degree of a node, etc. Of course, networkx implements this for us.⬇️
# nodes information
print(list(G.nodes)) # output: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
print(G.number_of_nodes()) # output: 10
# edges information
print(list(G.edges)) # output: [(1, 2), (1, 3), (1, 4), (8, 9), (8, 10)]
print(G.number_of_edges()) # output: 5
# neighbors of a node
print(list(G.neighbors(1))) # output: [2, 3, 4]
# degree information
print(G.degree(1)) # output: 3, because we have edges: (1, 2), (1, 3), (1, 4)manipulate nodes && edges
The operation of removing is the inverse process of the operation of adding.
G.remove_node(5)
G.remove_nodes_from(G2)
# similar to remove_node && remove_nodes_from
# you can add parameters you like
# G.remove_edge()
# G.remove_edges_from()๐ค
๐ The node type doesn’t have to be an integer. It can also be a string.
new_G = nx.Graph() # Let's define a new graph
new_G.add_node("abcd") # Add a string node
new_G.add_nodes_from("abcd") # this line is different, we treat the string as an iterable cotainer
print(list(new_G.nodes)) # output: ['abcd', 'a', 'n', 'c', 'd']Step 2. Draw a graph
Although we can draw a graph defined by networkx in matplotlib.pyplot, the generated image is static and not pretty in my opinion. So I will just skip the tutorial of drawing in matplotlib.pyplot. I will use pyvis package instead.
Install tutorial
I didn’t find how to use conda to install pyvis, so you may just use pip install pyvis. I remember that we can use pip in conda environment, but we need to run conda install pip at first(Not quite sure๐จ)
How to use pyvis
from pyvis.network import Network
net = Network('500px', '500px')
net.from_nx(G) # The G is defined in Step 1.
net.show('net.html')Then you can open the net.html in your local machine to interact with it. It looks like this⬇️
I don’t know how to integrate my html file into this blog. You may check the result in this (If you follow my tutorial)
๐ค
๐ In
pyvis, the valid node attributes are⬇️. So when we usenetworkxto add nodes, we can directly use these attributes. In this way, thepyviswill translate them for us.
['size', 'value', 'title', 'x', 'y', 'label', 'color']Similarly, if you want to customize your edges’ attributes, you may check the docs
Then you can use
G.add_edge(a, b, attribute1=, attribute2, ...)