使用 pos 值显示来自 pandas 数据帧 networkx/错误的边缘标签时出错.不知道是哪个 [英] Error displaying Edge labels from pandas dataframe networkx/ error with pos values. dont know which is it

查看:99
本文介绍了使用 pos 值显示来自 pandas 数据帧 networkx/错误的边缘标签时出错.不知道是哪个的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我的数据框有 3 列,源、目标和值.看起来像这样

My dataframe has 3 columns, source , target and value. it looks like this

源目标值波巴费特 C-3PO 4波巴费特 CHEWBACCA 3波巴·费特达斯·维达 8波巴费特汉 7

source target value BOBA FETT C-3PO 4 BOBA FETT CHEWBACCA 3 BOBA FETT DARTH VADER 8 BOBA FETT HAN 7

G = nx.from_pandas_edgelist(links,source='source',target='target', edge_attr='value')

G = nx.from_pandas_edgelist(links,source='source',target='target', edge_attr='value')

我用它来添加我的边缘列表

I use this to add my edgelist

nx.draw_networkx_edge_labels(G,pos=nx.Graph(G),edge_labels={(u,v):w for u,v,w inG.edges(data='value')})

nx.draw_networkx_edge_labels(G,pos=nx.Graph(G),edge_labels={(u,v):w for u,v,w in G.edges(data='value')})


我试过这个来显示边缘标签.我希望将值"显示为我的边缘标签它给出了这个错误:


I tried this to show edge labels. I want the 'value' to be dispplayed as my edge label it gives this error:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-23-d49aadc4a5f6> in <module>
      1 ##try edge list here
----> 2 nx.draw_networkx_edge_labels(G,pos=nx.Graph(G),edge_labels={(u,v):w for u,v,w in G.edges(data='value')})
      3 

~\Anaconda3\lib\site-packages\networkx\drawing\nx_pylab.py in draw_networkx_edge_labels(G, pos, edge_labels, label_pos, font_size, font_color, font_family, font_weight, alpha, bbox, ax, rotate, **kwds)
    939     text_items = {}
    940     for (n1, n2), label in labels.items():
--> 941         (x1, y1) = pos[n1]
    942         (x2, y2) = pos[n2]
    943         (x, y) = (x1 * label_pos + x2 * (1.0 - label_pos),

ValueError: too many values to unpack (expected 2)

推荐答案

你提供给 pos 的值应该是节点位置的字典,而不是实际的图形(里面有一堆布局函数networkx 来构建它.像这样的东西应该可以修复它.

The value you provide to pos should be a dictionary of node positions not the actual graph (there are a bunch of layout functions in networkx to build this. Something like this should fix it.

gpos = nx.circular_layout(G) # make a variable containing not positions
nx.draw_networkx(G,pos=gpos) # draw the graph
nx.draw_networkx_edge_labels(G,pos=gpos,edge_labels={(u,v):w for u,v,w in G.edges(data='value')}) # add edge labels

如果您多次调用以绘制图形的其余部分,最好创建一个包含节点位置的变量,并将该变量传递给上述函数的 pos 参数.

If you make multiple calls to plot the rest of the graph it's best to make a variable that contains your node positions and pass that same variable to the pos argument of the functions as above.

这篇关于使用 pos 值显示来自 pandas 数据帧 networkx/错误的边缘标签时出错.不知道是哪个的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆