Re: [igraph] Enforcing directions of edges

[Benika H]

Also, just got an error

Error in which(c1$csize) : argument to 'which' is not logical

Calls: <Anonymous> ... withCallingHandlers -> withVisible -> eval -> eval -> which

Execution halted

On Wed, Feb 4, 2015 at 5:45 PM, Benika H <address@hidden> wrote:

Oh ok. I'm new to this. I will send data.

hgnc_mapped2.txt is the 1st edgelist (y in my code)

newdata.txt is the 2nd edgelist (z in my code)

I got an error saying Warning in layout[, 1] + label.dist * cos(-label.degree) * (vertex.size +## : longer object length is not a multiple of shorter object length

I'm not getting any errors now, but I'm still getting all communities except thrones I requested.

On Wed, Feb 4, 2015 at 5:33 PM, Chris Watson <address@hidden> wrote:

We can't reproduce your code because we don't have your data. You also don't show any error messages you receive, or how the output/result is different than what you expect to see.

On Wed, Feb 4, 2015 at 5:30 PM, Benika H <address@hidden> wrote:

Hi Tamas, Could you help me with my question.

I have an edgelist (1st file) which I created the graph object. I want to use walktrap community detection on that network. After getting the communities, I need to plot only the communities with at most 8 vertices and at least 2 vertices (no isolates). Using those communities, I want to join the edgelist from the 2nd file to the matching vertices in a column. Here, I think I can use graph.union. However, I can't seem to get the communities with a sizes 2-8..

My data are in files: 

hgnc_mapped looks like  

ABCB7 MARS 0.054839

ABCB7 MAX 0.0638109

ABCB7 MRPS34 0.112394

ABCB7 NDUFA8 0.123633

ABCB7 NUP133 0.0810968

combined looks like

hsa-mir-1180 MAPK7

hsa-mir-1228 NACA

hsa-mir-1248 IRF9

hsa-mir-1248 PSME2

hsa-mir-1254 KIAA1279

hsa-mir-125b-1PCDHGB3

               

My code so far is:

edgeList <- read.table("hgnc_mapped.txt", header=TRUE, sep="\t")

eQTLList <- read.table("combined.txt", header=TRUE, sep="\t")

t <- as.data.frame(edgeList)

u <- as.data.frame(eQTLList)

y <- graph.data.frame(t, directed=FALSE, vertices=NULL)

V(y)

z <- graph.data.frame(u, directed=TRUE)

# new_g <- delete.vertices(y,which(degree(y) < 1) - 1)

# V(new_g)

```

```{r cache=TRUE, dependson='network-data'}

#Run the community detection algorithm

wc <- walktrap.community(y, weights = edgeList$Weight, steps=6, merges =TRUE, modularity = TRUE, membership = TRUE)

```

c1 <- clusters(y)

c1$membership

# small.clusters <- which(c1$size > 7)

# vertices.to.delete <- which((c1$membership)==small.clusters)-1

# g <- delete.vertices(y, vertices.to.delete)

layout <- layout.fruchterman.reingold(y)

x <- (which(c1$csize) < 8 & (c1$csize) > 1)

vertices <- which(c1$membership==x)

g1 <- induced.subgraph(y, vertices)

plot(g1,layout=layout[vertices,])

Thanks in advance.....

_______________________________________________
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