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.
--
http://mail.python.org/mailman/listinfo/python-list
