Re: [igraph] injecting a network and simulating it.

[suvirbhargav]

thanks.That worked!

Increasing vertices and edges to higher values makes it slow.(v=10000,e=20000) though.

Any optimization technique i can use?

Regards

Suvir

On Fri, Apr 26, 2013 at 4:22 AM, Gábor Csárdi <address@hidden> wrote:

On Thu, Apr 25, 2013 at 6:08 PM, suvirbhargav <address@hidden> wrote:

i changed the code a bit by running neighbour opinion function in loop and comparing the opinion "1" of net from past five time step .(assumption is network will repeat itself in these five time step in steady state)

code: http://scsys.co.uk:8002/244444

How do i change the opinion based not only on neighbour but also node itself .

(as adopted opinion is majority opinion including opinion from both neigbour and itself )

Then use the Opinion attribute of the 'x' vertex as well. 

 

Also,cluster size is there in igrpah already there, but i want to calculate largest cluster with opinion "1".

Delete all the vertices that are '2' with delete.vertices(net, V(g)[opinion==2]) and then call clusters(). 

G.

 

Thanks & Regards

Suvir

On Thu, Apr 25, 2013 at 7:02 PM, Gábor Csárdi <address@hidden> wrote:

See ?sample and the prob argument. G.

On Thu, Apr 25, 2013 at 12:54 PM, suvirbhargav <address@hidden> wrote:

thanks.

In below ,How can i control the percentage of both 1 and 2,say 20% of 1 and rest 2?

V(net)$Opinion <- sample(1:2, vcount(net), replace=TRUE)

Sorry,this is R question though,may be 

Regards

Suvir

On Thu, Apr 25, 2013 at 4:37 PM, Gábor Csárdi <address@hidden> wrote:

On Thu, Apr 25, 2013 at 10:34 AM, Gábor Csárdi <address@hidden> wrote:

I see. Then what you could do is adding a vertex attribute, named (say) "Opinion", and then update that based on the network structure, e.g. (in R):

Sorry, sent it prematurely, this is the correct code:

library(igraph)

set.seed(42)

net <- erdos.renyi.game(40, 2/40)

V(net)$Opinion <- sample(1:2, vcount(net), replace=TRUE)

coords <- layout.auto(net)

colbar <- c("orange", "cyan")

layout(rbind(1:2))

par(mar=c(0,0,0,0))

plot(net, vertex.color=colbar[V(net)$Opinion], layout=coords)

## Set the opinion based on the neighbors

V(net)$Opinion <- sapply(V(net), function(x) {

  if (degree(net, x)==0) {

    V(net)$Opinion[x]

  } else { 

    round(mean(V(net)$Opinion[neighbors(net,x)]))

  }

})

plot(net, vertex.color=colbar[V(net)$Opinion], layout=coords)

G.

[...]

--
Gabor Csardi <address@hidden>     MTA KFKI RMKI

_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help

_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help

--
Gabor Csardi <address@hidden>     MTA KFKI RMKI

_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help

_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help

--
Gabor Csardi <address@hidden>     MTA KFKI RMKI

_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help