Subgraph Drawing

subhabangalore at gmail.com subhabangalore at gmail.com
Sun Jan 13 23:01:21 EST 2013


On Monday, January 14, 2013 6:05:49 AM UTC+5:30, Steven D'Aprano wrote:
> On Sun, 13 Jan 2013 12:05:54 -0800, subhabangalore wrote:
> 
> 
> 
> > Dear Group,
> 
> > 
> 
> > I have two questions, if I take a subseries of the matrix as in
> 
> > eigenvalue here, provided I have one graph of the full form in G, how
> 
> > may I show it, as if I do the nx.draw(G) it takes only the original
> 
> > graph.
> 
> 
> 
> Is this what you mean? If not, you will have to explain your question 
> 
> better.
> 
> 
> 
> 
> 
> L = = nx.laplacian(G)
> 
> E = numpy.linalg.eigvals(L)
> 
> nx.draw(E)
> 
> 
> 
> 
> 
> > >>> print numpy.linalg.eigvals(L)
> 
> > [  8.00000000e+00   2.22044605e-16   1.00000000e+00   1.00000000e+00
> 
> >    1.00000000e+00   1.00000000e+00   1.00000000e+00   1.00000000e+00]
> 
> > for more than 1000 nodes it is coming too slow on Windows 7 machine with
> 
> > 3GB RAM.
> 
> 
> 
> Get a faster machine. Or use fewer nodes. Or be patient and wait.
> 
> 
> 
> Solving a graph problem with 1000 nodes is a fairly big problem for a 
> 
> desktop PC. It will take time. Calculations don't just happen instantly, 
> 
> the more work you have to do the longer they take.
> 
> 
> 
> The last alternative is to ask on a specialist numpy list. But I expect 
> 
> they will probably tell you the same thing.
> 
> 
> 
> 
> 
> -- 
> 
> Steven

Dear Steven,

Thank you for your kind effort. You got the problem right. But it is giving following error,
Traceback (most recent call last):
  File "<pyshell#5>", line 1, in <module>
    nx.draw(E)
  File "C:\Python27\lib\site-packages\networkx\drawing\nx_pylab.py", line 138, in draw
    draw_networkx(G,pos=pos,ax=ax,**kwds)
  File "C:\Python27\lib\site-packages\networkx\drawing\nx_pylab.py", line 267, in draw_networkx
    pos=nx.drawing.spring_layout(G) # default to spring layout
  File "C:\Python27\lib\site-packages\networkx\drawing\layout.py", line 241, in fruchterman_reingold_layout
    A=nx.to_numpy_matrix(G,weight=weight)
  File "C:\Python27\lib\site-packages\networkx\convert.py", line 492, in to_numpy_matrix
    nodelist = G.nodes()
AttributeError: 'numpy.ndarray' object has no attribute 'nodes'
>>> 

there are other solution of converting back the matrix to graph should I try that?

Regards,
Subhabrata. 



More information about the Python-list mailing list