CODEWITHSUNDEEP
rmatrixnetwork-programmingsna
A Lee
A Lee

Reputation: 13

Converting edgelist (occurance) to incidence matrix for social network

I'm new to R/network analysis, and apologize if this is a stupid question. I'm currently trying to convert an edgelist into a network matrix, where the shared category of a group indicates the presence of an edge. (ie. 1 if they are in the same group, 0 if they are not... it might also be possible to indicate different numbers depending on the group... 1 for the first, 2 for the second, etc.)

If this helps, here's a sample of my data (I'm working with two data sets, much larger than this):

edgelist:

netherlands, business and human rights
burkina faso, business and human rights
japan, business and human rights
china, civil and political rights
myanmar, civil and political rights
fiji, civil and political rights
netherlands, civil and political rights
japan, civil and political rights
ethiopia, civil society
egypt, civil society
myanmar, civil society

matrix:

burkina faso, china, japan, netherlands, ethiopia, egypt, myanmar, fiji
burkina faso, 0, 1, 1, 0, 0, 0, 0, 0
china 0, 0, 1, 1, 0, 0, 1, 1
japan 1, 1, 0, 1, 0, 0, 1, 1
netherlands 1, 1, 1, 0, 0, 0, 1, 1
ethiopia 0, 0, 0, 0, 0, 1, 1, 0
egypt 0, 0, 0, 0, 1, 0, 1, 0
myanmar 0, 0, 0, 0, 1, 1, 0, 0
fiji 0, 1, 1, 1, 0, 0, 0, 0

etc.

I've tried reshape and as.integer but am struggling to know how that might work in my case...

Thanks in advance, everyone.

Upvotes: 1

Views: 292

Answers (1)

Onyambu
Onyambu

Reputation: 79238

you could do:

library(igraph)
a <- aggregate(V1~V2,dat,function(x)combn(as.character(x),2,paste0,collapse=","))[2]

as.matrix(as_adj(graph_from_data_frame(read.csv(text=unique(unlist(a)),h=F),F)))


             netherlands burkina faso china myanmar fiji ethiopia egypt japan
netherlands            0            1     1       1    1        0     0     1
burkina faso           1            0     0       0    0        0     0     1
china                  1            0     0       1    1        0     0     1
myanmar                1            0     1       0    1        1     1     1
fiji                   1            0     1       1    0        0     0     1
ethiopia               0            0     0       1    0        0     1     0
egypt                  0            0     0       1    0        1     0     0
japan                  1            1     1       1    1        0     0     0

If that seems difficult, you could break it down to:

my_f <-function(x)  do.call(rbind,combn(as.character(x),2,simplify = F))

a <-unique(do.call(rbind,aggregate(V1~V2,dat,my_f)[,2]))

as.matrix(as_adj(graph_from_edgelist(a)))

DATA:

             V1                          V2
1   netherlands   business and human rights
2  burkina faso   business and human rights
3         japan   business and human rights
4         china  civil and political rights
5       myanmar  civil and political rights
6          fiji  civil and political rights
7   netherlands  civil and political rights
8         japan  civil and political rights
9      ethiopia               civil society
10        egypt               civil society
11      myanmar               civil society

Upvotes: 1

Related Questions