在Networkx中导出最小生成树的节点坐标

Question

我在Python中使用Networkx在数据中生成相关性的最小生成树，我想找到节点的（x，y）坐标。 我读了这个问题： 使用NetworkX导出图形的布局位置并尝试应用它，但是我真的不明白为什么我没有得到预期的结果。

这是我使用的代码：

import networkx as nx
import random as rn
import pylab as pl
def create_tree(data):
    corr_matrix = define_correlation_matrix(data)
    G = nx.Graph(corr_matrix)
    pos = nx.minimum_spanning_tree(G)
    rn.seed = 5
    colors = 'bcgmry'
    components = nx.connected_components(pos)
    for i in components:
        component = pos.subgraph(i)
        nx.draw_graphviz(component,
                         node_color = colors[rn.randint(0,   len(colors)-1)], 
                         node_size = 15,
                         edge_color = [corr_matrix[i][j]*0.5 for (i,j) in component.edges()],
                         with_labels = True,
                         labels = dict([(x,data[x]) for x in component.nodes()]))
    nx.set_node_attributes(G,'pos',pos)
    print G.node
    pl.show()

但是我没有得到坐标：

{0: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}, 1: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}, 2: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}, 3: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}, 4: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}, 5: {'pos': <networkx.classes.graph.Graph object at 0x10513ea50>}}

您知道如何更改代码以具有坐标吗？

谢谢。