Convert long table of linked observations to wide adjacency matrix [duplicate] - r

Question

This question already has an answer here:
How to convert two factors to adjacency matrix in R?
(1 answer)
Closed 4 years ago.
I am facing a challenge that I cannot manage to solve. I have a list of observations x_i (dimension is large, something around 30k) and a list of observations y_j (also large). x_i and y_i are id of the same units (say firms).
I have a dataframe of two columns that links x_i and y_j: if they appear on the same line, it means that they are connected. What I would like is to convert this network into a large matrix M of size (unique(union(x, y))) and which takes the value 1 if the two firms are connected.
Here is an example in small dimensions:
x1 x2
x3 x6
x4 x5
x1 x5
What I would like is a matrix:
0 1 0 0 1 0
0 0 0 0 0 0
0 0 0 0 0 1
0 0 0 1 0 0
0 0 0 0 0 0
0 0 0 0 0 0
Right now, the only solution I could think of is a double loop combined with a search in the initial dataframe:
list_firm = union(as.vector(df[1]), as.vector(df[2]))
list_firm <- sort(list_firm[[1]])
list_firm <- unique(list_firm)
M <- Matrix(nrow = length(list_firm), ncol = length(list_firm))
for (i in list_firm) {
for (j in list_firm) {
M[i, j] = !is.null(which(df$col1 == i & df$col2 == j))
}
}
Where df is the two columns data frame. This is obviously much too long to run.
Any suggestion? This would be very welcome

Answer 1

We convert the columns to factor with levels specified as the unique elements of both columns and get the frequency with table
lvls <- sort(unique(unlist(df)))
df[] <- lapply(df, factor, levels = lvls)
table(df)
# col2
#col1 x1 x2 x3 x4 x5 x6
# x1 0 1 0 0 1 0
# x2 0 0 0 0 0 0
# x3 0 0 0 0 0 1
# x4 0 0 0 0 1 0
# x5 0 0 0 0 0 0
# x6 0 0 0 0 0 0
data
df <- structure(list(col1 = c("x1", "x3", "x4", "x1"), col2 = c("x2",
"x6", "x5", "x5")), class = "data.frame", row.names = c(NA, -4L
))

Answer 2

The answer provided by #akrun in the comments is a good one. However, this is a good scenario to take advantage of a different data structure than data frames. Basically, what you're looking for is an adjacency matrix, which is a data structure in social network analysis. To achieve this, we can use the igraph package in R.
library(igraph)
library(dplyr)
df = data_frame(source=c('x1', 'x3', 'x4', 'x1'), target=c('x2', 'x6', 'x5', 'x5'))
g = graph_from_data_frame(df, directed=FALSE)
output = as.matrix(get.adjacency(g))
x1 x3 x4 x2 x6 x5
x1 0 0 0 1 0 1
x3 0 0 0 0 1 0
x4 0 0 0 0 0 1
x2 1 0 0 0 0 0
x6 0 1 0 0 0 0
x5 1 0 1 0 0 0
The output columns aren't in the exact order as your example, but this is a trivial problem to solve if needed.