pagerank of weighted network

[satyam]

Hi
How do I compute the pagerank of a weighted network?
I have a file which reads like this

1 2 14
1 3 15
2 5 16

and so on...Here the first column is node i and 2nd column is node j.
The third column is weight of the edge (from 1-->2).
thanks
Satyam

[Nicholas Dronen]

I don't have python where I am right now, so I can't write an example,
but I recommend using read_edgelist to create a graph, then calling
the pagerank function. Both functions are well documented. Looking
at the pagerank code in networkx 1.5, it appears to do the right thing
with weighted directed networks, so I think you should be OK there.

Regards,

Nick

[Aric Hagberg]

Yes, that should do it. e.g.

In [1]: import networkx as nx

In [2]: cat graph.edgelist

1 2 14
1 3 15
2 5 16

In [3]: G=nx.read_weighted_edgelist('graph.edgelist',nodetype=int)

In [4]: nx.pagerank(G)
Out[4]:
{1: 0.31576945228063863,
2: 0.32367032787416605,
3: 0.17632967212583389,
5: 0.18423054771936137}

[satyam mukherjee]

Thanks. Should work with my dataset :)

--
You received this message because you are subscribed to the Google Groups "networkx-discuss" group.
To post to this group, send email to networkx...@googlegroups.com.
To unsubscribe from this group, send email to networkx-discu...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/networkx-discuss?hl=en.

--
-------------------------------------------
WHEN LIFE GIVES U HUNDRED REASONS TO CRY,SHOW LIFE THAT U HAVE THOUSAND REASONS TO SMILE :-)

satyam mukherjee
224-436-3672 (Mob)
847-491-7238 (Off)

[satyam mukherjee]

Well it doesn't work correctly I guess. Just consider this edgelist

1 2 14

2 3 11

3 1 21

3 2 31

>>> G=nx.read_weighted_edgelist('graph.edgelist',nodetype=int)

>>> print G.edges(data=True)

[(1, 2, {'weight': 14.0}), (1, 3, {'weight': 21.0}), (2, 3, {'weight': 31.0})]

Just see everything gets interchanged. Any suggestion how to overcome this?

Thanks

Satyam

[Moritz Beber]

Hi,

On 08/18/2011 03:26 AM, satyam mukherjee wrote:

Well it doesn't work correctly I guess. Just consider this edgelist

1 2 14

2 3 11

3 1 21

3 2 31

>>> G=nx.read_weighted_edgelist('graph.edgelist',nodetype=int)

>>> print G.edges(data=True)

[(1, 2, {'weight': 14.0}), (1, 3, {'weight': 21.0}), (2, 3, {'weight': 31.0})]

Just see everything gets interchanged. Any suggestion how to overcome this?

You probably want a directed graph, i.e., pass create_using=nx.DiGraph() to read_weighted_edgelist.

Moritz

[satyam mukherjee]

Thanks.

Also I am trying to use the weighted pagerank in the following way :

P(i) = (1-d)(\Sum_{j}P(j) w_{ji}/s_{j}^{out} ) + d/N

where s_{j}^{out} is the strength of the outgoing link 

s_{j}^{out} = \sum_{i}w_{ji}

currently in the weighted pagerank algorithm, we have the first term like 

w_{ji}/k_{j}^{out}. Basically one can take a look at this paper.

http://www.plosone.org/article/info:doi%2F10.1371%2Fjournal.pone.0017249

--

You received this message because you are subscribed to the Google Groups "networkx-discuss" group.
To post to this group, send email to networkx...@googlegroups.com.
To unsubscribe from this group, send email to networkx-discu...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/networkx-discuss?hl=en.

[satyam mukherjee]

Ok. I found the solution to do this :

In the file stochastic.py pass

degree = W.out_degree(weighted=True). 

Thanks

Satyam