filter all nodes with only one edge

[Hugo Koopmans]

Hi all,

I am new to networkX and have the following question:
I have created a network graph and now I would like to filter out all
nodes with only one edge.
So I keep the network of nodes(people) who have more then one friend
in the network...

thank you for a reply

hugo

[Aric Hagberg]

Here is the "one-liner" way ([17])

In [15]: G=nx.fast_gnp_random_graph(10,0.1)

In [16]: G.degree()
Out[16]: {0: 4, 1: 3, 2: 3, 3: 2, 4: 1, 5: 2, 6: 4, 7: 2, 8: 1, 9: 2}

In [17]: G.remove_nodes_from((n for n,d in G.degree_iter() if d==1))

In [18]: G.degree()
Out[18]: {0: 4, 1: 2, 2: 2, 3: 2, 5: 2, 6: 4, 7: 2, 9: 2}

Aric

[Amitabh Sharma]

Hi Aric,
In the similar line, I would like to ask you how to retain the edges having link with two friends i.e. n1 is connected to friend1 and also with friend2. How to do this in networkx.

Amitabh

> --
> 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.
>

[Christopher Ellison]

Amitabh Sharma wrote the following on 02/16/2011 01:09 PM:
> Hi Aric, In the similar line, I would like to ask you how to retain
> the edges having link with two friends i.e. n1 is connected to
> friend1 and also with friend2. How to do this in networkx.
>

Hi Amitabh,

It's nearly identical to what Aric showed you. Previously you wanted to
remove all nodes which had degree equal to 1. Now you want to remove
all nodes which have degree not equal to 2. So replace,

d==1

with

d!=2

Chris

[Dan Schult]

A one liner version could be:

G.remove_nodes_from((n for n in G if n not in G[friend1] or n not in G[friend2]))

But its probably better to do something like:

bothnbrs=set([friend1,friend2]) ^ (set(G[friend1]) & set(G[friend2]))
H=G.subgraph(bothnbrs)

Dan

[Amitabh Sharma]

Thanks Dan!
It worked, I have to just check it in the graph..soo easy :)

cheers

[Hugo Koopmans]

Thank you guys for this quick and excellent response!

much appriciated

hugo

Vriendelijke groet / Kind regards,

Hugo Koopmans

managing partner

DIKW Consulting
___________________________________
data > information > knowledge > wisdom

[w] www.dikw.com
[m] +31 6 4310 6780