How to transform weighted edgelist into adjacency matrix in R

1
votes

I would like to transform a weighted, directed edge list into an adjacency matrix with the weights for the sender and receiver in different cells. How can I do this most efficiently?

here is an example:

 el <- rbind(c("acotr1", "actor2", "actor1sendsActor2", "actor2sendsActor1"), c(1,2,5.5,6.5), c(1,3, 3.5, 1),  c(4,1,1.5,0))
 colnames(el) <- el[1,]
 el <- el[-1,]

el looks as follows

    acotr1 actor2 actor1sendsActor2 actor2sendsActor1
[1,] "1"    "2"    "5.5"             "6.5"            
[2,] "1"    "3"    "3.5"             "1"              
[3,] "4"    "1"    "1.5"             "0"

Creating a binary edge list can be easily achieved by using

 as.matrix(table(el[,1], el[,2]))

where el[,1], el[,2] are the names of the nodes in the network.

but I would like to have 

  1    2    3    4
1 .    5.5  3.5  0
2 6.5  .    .    .
3 1    .    .    .
4 1.5  .    .    .
1
rigraph

1 Answers

1
votes

First let's convert the matrix to a numeric one:

mode(el) <- "numeric"
el
#      acotr1 actor2 actor1sendsActor2 actor2sendsActor1
# [1,]      1      2               5.5               6.5
# [2,]      1      3               3.5               1.0
# [3,]      4      1               1.5               0.0

I don't think there's a magic shortcut for weighted (especially with two columns) cases, but the following is also succinct:

# Creating an adjacency matrix of zeros for 1, ..., max(agents)
M <- matrix(0, max(el[, 1:2]), max(el[, 1:2]))
# Using the 3rd column
M[el[, 1:2]] <- el[, 3]
# Using the 4th column
M[el[, 2:1]] <- el[, 4]
M
#      [,1] [,2] [,3] [,4]
# [1,]  0.0  5.5  3.5    0
# [2,]  6.5  0.0  0.0    0
# [3,]  1.0  0.0  0.0    0
# [4,]  1.5  0.0  0.0    0