python复杂网络库networkx：绘图draw

柔光的暖阳◎ 2022-07-13 12:27 365阅读 0赞

http://[blog.csdn.net/pipisorry/article/details/54291831][blog.csdn.net_pipisorry_article_details_54291831]

## networkx使用matplotlib绘制函数 ##

<table> 
 <tbody> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw.html#networkx.drawing.nx_pylab.draw" title="draw" rel="nofollow">draw</a>(G[,&nbsp;pos,&nbsp;ax,&nbsp;hold])</td> 
   <td>Draw the graph G with Matplotlib.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_networkx.html#networkx.drawing.nx_pylab.draw_networkx" title="draw_networkx" rel="nofollow">draw_networkx</a>(G[,&nbsp;pos,&nbsp;arrows,&nbsp;with_labels])</td> 
   <td>Draw the graph G using Matplotlib.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_networkx_nodes.html#networkx.drawing.nx_pylab.draw_networkx_nodes" title="draw_networkx_nodes" rel="nofollow">draw_networkx_nodes</a>(G,&nbsp;pos[,&nbsp;nodelist,&nbsp;...])</td> 
   <td>Draw the nodes of the graph G.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_networkx_edges.html#networkx.drawing.nx_pylab.draw_networkx_edges" title="draw_networkx_edges" rel="nofollow">draw_networkx_edges</a>(G,&nbsp;pos[,&nbsp;edgelist,&nbsp;...])</td> 
   <td>Draw the edges of the graph G.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_networkx_labels.html#networkx.drawing.nx_pylab.draw_networkx_labels" title="draw_networkx_labels" rel="nofollow">draw_networkx_labels</a>(G,&nbsp;pos[,&nbsp;labels,&nbsp;...])</td> 
   <td>Draw node labels on the graph G.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_networkx_edge_labels.html#networkx.drawing.nx_pylab.draw_networkx_edge_labels" title="draw_networkx_edge_labels" rel="nofollow">draw_networkx_edge_labels</a>(G,&nbsp;pos[,&nbsp;...])</td> 
   <td>Draw edge labels.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_circular.html#networkx.drawing.nx_pylab.draw_circular" title="draw_circular" rel="nofollow">draw_circular</a>(G,&nbsp;**kwargs)</td> 
   <td>Draw the graph G with a circular layout.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_random.html#networkx.drawing.nx_pylab.draw_random" title="draw_random" rel="nofollow">draw_random</a>(G,&nbsp;**kwargs)</td> 
   <td>Draw the graph G with a random layout.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_spectral.html#networkx.drawing.nx_pylab.draw_spectral" title="draw_spectral" rel="nofollow">draw_spectral</a>(G,&nbsp;**kwargs)</td> 
   <td>Draw the graph G with a spectral layout.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_spring.html#networkx.drawing.nx_pylab.draw_spring" title="draw_spring" rel="nofollow">draw_spring</a>(G,&nbsp;**kwargs)</td> 
   <td>Draw the graph G with a spring layout.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_shell.html#networkx.drawing.nx_pylab.draw_shell" title="draw_shell" rel="nofollow">draw_shell</a>(G,&nbsp;**kwargs)</td> 
   <td>Draw networkx graph with shell layout.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pylab.draw_graphviz.html#networkx.drawing.nx_pylab.draw_graphviz" title="draw_graphviz" rel="nofollow">draw_graphviz</a>(G[,&nbsp;prog])</td> 
   <td>Draw networkx graph with graphviz layout.</td> 
  </tr> 
 </tbody> 
</table>

## networkx使用Graphviz AGraph (dot)绘制函数 ##

<table> 
 <tbody> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.from_agraph.html#networkx.drawing.nx_agraph.from_agraph" title="from_agraph" rel="nofollow">from_agraph</a>(A[,&nbsp;create_using])</td> 
   <td>Return a NetworkX Graph or DiGraph from a PyGraphviz graph.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.to_agraph.html#networkx.drawing.nx_agraph.to_agraph" title="to_agraph" rel="nofollow">to_agraph</a>(N)</td> 
   <td>Return a pygraphviz graph from a NetworkX graph N.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.write_dot.html#networkx.drawing.nx_agraph.write_dot" title="write_dot" rel="nofollow">write_dot</a>(G,&nbsp;path)</td> 
   <td>Write NetworkX graph G to Graphviz dot format on path.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.read_dot.html#networkx.drawing.nx_agraph.read_dot" title="read_dot" rel="nofollow">read_dot</a>(path)</td> 
   <td>Return a NetworkX graph from a dot file on path.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.graphviz_layout.html#networkx.drawing.nx_agraph.graphviz_layout" title="graphviz_layout" rel="nofollow">graphviz_layout</a>(G[,&nbsp;prog,&nbsp;root,&nbsp;args])</td> 
   <td>Create node positions for G using Graphviz.</td> 
  </tr> 
  <tr> 
   <td> <p><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_agraph.pygraphviz_layout.html#networkx.drawing.nx_agraph.pygraphviz_layout" title="pygraphviz_layout" rel="nofollow">pygraphviz_layout</a>(G[,&nbsp;prog,&nbsp;root,&nbsp;args])</p> </td> 
   <td>Create node positions for G using Graphviz.</td> 
  </tr> 
 </tbody> 
</table>

## networkx使用Graphviz with pydot绘制函数 ##

<table> 
 <tbody> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.from_pydot.html#networkx.drawing.nx_pydot.from_pydot" title="from_pydot" rel="nofollow">from_pydot</a>(P)</td> 
   <td>Return a NetworkX graph from a Pydot graph.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.to_pydot.html#networkx.drawing.nx_pydot.to_pydot" title="to_pydot" rel="nofollow">to_pydot</a>(N[,&nbsp;strict])</td> 
   <td>Return a pydot graph from a NetworkX graph N.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.write_dot.html#networkx.drawing.nx_pydot.write_dot" title="write_dot" rel="nofollow">write_dot</a>(G,&nbsp;path)</td> 
   <td>Write NetworkX graph G to Graphviz dot format on path.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.read_dot.html#networkx.drawing.nx_pydot.read_dot" title="read_dot" rel="nofollow">read_dot</a>(path)</td> 
   <td>Return a NetworkX MultiGraph or MultiDiGraph from a dot file on path.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.graphviz_layout.html#networkx.drawing.nx_pydot.graphviz_layout" title="graphviz_layout" rel="nofollow">graphviz_layout</a>(G[,&nbsp;prog,&nbsp;root])</td> 
   <td>Create node positions using Pydot and Graphviz.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.nx_pydot.pydot_layout.html#networkx.drawing.nx_pydot.pydot_layout" title="pydot_layout" rel="nofollow">pydot_layout</a>(G[,&nbsp;prog,&nbsp;root])</td> 
   <td>Create node positions using Pydot and Graphviz.</td> 
  </tr> 
 </tbody> 
</table>

## 图布局Graph Layout ##

<table> 
 <tbody> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.circular_layout.html#networkx.drawing.layout.circular_layout" title="circular_layout" rel="nofollow">circular_layout</a>(G[,&nbsp;dim,&nbsp;scale,&nbsp;center])</td> 
   <td>Position nodes on a circle.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.fruchterman_reingold_layout.html#networkx.drawing.layout.fruchterman_reingold_layout" title="fruchterman_reingold_layout" rel="nofollow">fruchterman_reingold_layout</a>(G[,&nbsp;dim,&nbsp;k,&nbsp;...])</td> 
   <td>Position nodes using Fruchterman-Reingold force-directed algorithm.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.random_layout.html#networkx.drawing.layout.random_layout" title="random_layout" rel="nofollow">random_layout</a>(G[,&nbsp;dim,&nbsp;scale,&nbsp;center])</td> 
   <td>Position nodes uniformly at random.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.shell_layout.html#networkx.drawing.layout.shell_layout" title="shell_layout" rel="nofollow">shell_layout</a>(G[,&nbsp;nlist,&nbsp;dim,&nbsp;scale,&nbsp;center])</td> 
   <td>Position nodes in concentric circles.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.spring_layout.html#networkx.drawing.layout.spring_layout" title="spring_layout" rel="nofollow">spring_layout</a>(G[,&nbsp;dim,&nbsp;k,&nbsp;pos,&nbsp;fixed,&nbsp;...])</td> 
   <td>Position nodes using Fruchterman-Reingold force-directed algorithm.</td> 
  </tr> 
  <tr> 
   <td><a href="http://networkx.readthedocs.io/en/networkx-1.11/reference/generated/networkx.drawing.layout.spectral_layout.html#networkx.drawing.layout.spectral_layout" title="spectral_layout" rel="nofollow">spectral_layout</a>(G[,&nbsp;dim,&nbsp;weight,&nbsp;scale,&nbsp;center])</td> 
   <td>Position nodes using the eigenvectors of the graph Laplacian.</td> 
  </tr> 
 </tbody> 
</table>

[皮皮blog][blog]

# networkx绘图示例 #

### 示例1：绘制二分图 ###

def draw\_bipartite(edges, bi1\_node\_w\_dict, bi2\_node\_w\_dict):  
    '''  
    :param edges: edges及权重  
    :param bi1\_node\_w\_dict: bipartite1中node的大小  
    '''  
    import networkx as nx  
    import numpy as np  
    from networkx.algorithms import bipartite  
    import matplotlib as mpl  
    import matplotlib.pyplot as plt

''' node显示label和显示大小设置；edge设置'''  
    \# node labels的显示名转换（显示名转换；两边node名不能冲突）  
    bi1\_node\_key\_dict = \{k: 'c' + str(k) + '-' + str(v) for k, v in bi1\_node\_w\_dict.items()\}  
    bi1\_nodes = \[bi1\_node\_key\_dict.get(k) for k in bi1\_node\_w\_dict.keys()\]  
    bi1\_node\_w = bi1\_node\_w\_dict.values()  \# node 显示大小；要和bi1\_nodes一一对应不能乱了

bi2\_node\_key\_dict = \{k: 'a' + str(k) + '-' + str(v) for k, v in bi2\_node\_w\_dict.items()\}  
    bi2\_nodes = \[bi2\_node\_key\_dict.get(k) for k in bi2\_node\_w\_dict.keys()\]  
    bi2\_node\_w = bi2\_node\_w\_dict.values()

edges = \[(bi1\_node\_key\_dict.get(k1), bi2\_node\_key\_dict.get(k2), e\_w)  
             for k1, k2, e\_w in edges\]

print("bi1\_nodes:\{\}".format(bi1\_nodes))  
    print("bi2\_nodes:\{\}".format(bi2\_nodes))  
    print("edges:\{\}".format(edges))

B = nx.Graph()  \# 创建空图  
    B.add\_nodes\_from(bi1\_nodes, bipartite=0)  
    B.add\_nodes\_from(bi2\_nodes, bipartite=1)  
    B.add\_weighted\_edges\_from(edges)  
    \# print(nx.is\_bipartite(B))  \# 判断是否是二分图  
    \# X, Y = bipartite.sets(B) \# 不需要这种方式获取

''' node weights normalization '''  
    \# print("bi1\_node\_w:\\n\{\}".format(bi1\_node\_w))  
    min1, max1 = min(bi1\_node\_w), max(bi1\_node\_w)  
    bi1\_node\_w\_norm = \[(i - min1) / (max1 - min1) for i in bi1\_node\_w\]  
    \# print("bi1\_node\_w\_norm:\\n\{\}".format(bi1\_node\_w\_norm))

\# print("bi2\_node\_w:\\n\{\}".format(bi2\_node\_w))  
    min2, max2 = min(bi2\_node\_w), max(bi2\_node\_w)  
    bi2\_node\_w\_norm = \[(i - min2) / (max2 - min2) for i in bi2\_node\_w\]  
    \# print("bi2\_node\_w\_norm:\\n\{\}".format(bi2\_node\_w\_norm))

'''绘制node'''  
    pos = dict()  
    pos.update((n, (1, i)) for i, n in enumerate(bi1\_nodes))  \# put nodes from X at x=1  
    pos.update((n, (2, i)) for i, n in enumerate(bi2\_nodes))  \# put nodes from X at x=2  
    \# nx.draw pos parameter: A dictionary with nodes as keys and positions as values  
    nx.draw\_networkx\_nodes(B, pos, nodelist=bi1\_nodes, alpha=bi1\_node\_w\_norm,  
                           node\_size=\[5 \* i for i in bi1\_node\_w\])  \# node\_color='slategray',  
    nx.draw\_networkx\_nodes(B, pos, nodelist=bi2\_nodes, alpha=bi2\_node\_w\_norm,  
                           node\_size=\[5 \* i for i in bi2\_node\_w\])  \# node\_color='steelblue',  
    \# 设置nodes中label大小等属性  
    nx.draw\_networkx\_labels(B, pos, font\_color='black', font\_family='arial',  
                            font\_size=8)  \# label\_size\_by\_freq is not allowed

'''绘制edges'''  
    colors = \[B.edges\[u, i\]\['weight'\] for u, i in B.edges\]  
    edges = nx.draw\_networkx\_edges(B, pos=pos, edge\_color=colors,  \# edge\_cmap=plt.cm.Blues,  
                                   width=np.asarray(colors) / 10, edge\_vmin=0, alpha=0.9)  
    \# nx.draw(B, pos=pos, alpha=0.5) \#这个会同时绘制node和edges

'''设置颜色'''  
    pc = mpl.collections.PatchCollection(\[\])  \# , cmap=plt.cm.Blues)  
    pc.set\_array(colors)  
    plt.colorbar(pc)

ax = plt.gca()  
    ax.set\_axis\_off()  
    \# plt.legend()  
    plt.show()

actions\_cnt = \{2: 130, 4: 36, 3: 16, 0: 9, 5: 6\}  
clusters\_cnt = \{4: 44, 9: 37, 6: 36, 8: 29, 2: 28, 5: 22, 1: 1\}  
rudt\_cnt\_edges = \[(1, 2, 1), (2, 2, 22), (2, 4, 3), (2, 3, 3), (4, 2, 22), (4, 4, 10), (4, 0, 7),  
                  (4, 3, 5), (5, 2, 22), (6, 2, 22), (6, 4, 7), (6, 5, 4), (6, 3, 3), (8, 2, 21),  
                  (8, 4, 6), (8, 3, 2), (9, 2, 20), (9, 4, 10), (9, 3, 3), (9, 0, 2), (9, 5, 2)\]

draw\_bipartite(rudt\_cnt\_edges, clusters\_cnt, actions\_cnt)

![156e456af20e4503bf2e8fb9da87e6ea.png][]

\[[Python-NetworkX绘制网络图][Python-NetworkX]\]

\[[Bipartite graph in NetworkX][]\]

### 示例2： ###

def drawGraph(fg, title='weighted_graph'):
        import matplotlib.pyplot as plt
        import networkx as nx
    
        pos = nx.spring_layout(fg)
        nx.draw_networkx_nodes(fg, pos, node_shape='.', node_size=20)
        nx.draw_networkx_edges(fg, pos)
        nx.draw_networkx_labels(fg, pos)
        nx.draw_networkx_edge_labels(fg, pos, edge_labels=nx.get_edge_attributes(fg, 'weight'))
    
        plt.savefig(os.path.join(CWD, '../PGT/middlewares/' + title + '.png'))
        plt.show()

###  ###

### 其它示例 ###

\[[Python-NetworkX绘制网络图][Python-NetworkX]\]

from: [python复杂网络库networkx：绘图draw\_皮皮blog-CSDN博客][blog.csdn.net_pipisorry_article_details_54291831]

ref:官方文档\[Drawing graphs[¶][Link 1]\]

\[[networkx reference][]\]

\[[Drawing][]\]

[blog.csdn.net_pipisorry_article_details_54291831]: http://blog.csdn.net/pipisorry/article/details/54291831
[blog]: http://blog.csdn.net/pipisorry
[156e456af20e4503bf2e8fb9da87e6ea.png]: /images/20220713/3933708062df4bafb1f6b19ba312994a.png
[Python-NetworkX]: https://zhuanlan.zhihu.com/p/51886917
[Bipartite graph in NetworkX]: https://stackoverflow.com/questions/27084004/bipartite-graph-in-networkx
[Link 1]: https://networkx.org/documentation/stable/tutorial.html#drawing-graphs
[networkx reference]: https://networkx.github.io/documentation/stable/reference/index.html
[Drawing]: https://networkx.github.io/documentation/stable/reference/drawing.html