Question

I’m trying to solve the problem of having a co-occurence matrix. I have a datafile of transactions and items, and I want to see a matrix of the number of transactions where items appear together.

I’m a newbie in R programming and I’m having some fun finding out all the shortcuts that R has, rather than creating specific loops (I used to use C years ago and only sticking to Excel macros and SPSS now). I have checked the solutions here, but haven’t found one that works (the closest is the solution given here: Co-occurrence matrix using SAC? – but it produced an error message when I used projecting_tm, I suspected that the cbind wasn’t successful in my case.

Essentially I have a table containing the following:

TrxID Items Quant
Trx1 A 3
Trx1 B 1
Trx1 C 1
Trx2 E 3
Trx2 B 1
Trx3 B 1
Trx3 C 4
Trx4 D 1
Trx4 E 1
Trx4 A 1
Trx5 F 5
Trx5 B 3
Trx5 C 2
Trx5 D 1, etc.

I want to create something like:

   A B C D E F
A  0 1 1 0 1 1
B  1 0 3 1 1 0
C  1 3 0 1 0 0
D  1 1 1 0 1 1
E  1 1 0 1 0 0
F  0 1 1 1 0 0

What I did was (and you’d probably laugh at my rookie R approach):

library(igraph)
library(tnet)

trx <- read.table("FileName.txt", header=TRUE) 
transID <- t(trx[1])
items <- t(trx[2])

id_item <- cbind(items,transID)
item_item <- projecting_tm(id_item, method="sum")
item_item <- tnet_igraph(item_item,type="weighted one-mode tnet")
item_matrix <-get.adjacency(item_item,attr="weight")
item_matrix

As mentioned above the cbind was probably unsuccessful, so the projecting_tm couldn’t give me any result.

Any alternative approach or a correction to my method?

Your help would be much appreciated!

Answer 1

I’d use a combination of the reshape2 package and matrix algebra:

#read in your data
dat <- read.table(text="TrxID Items Quant
Trx1 A 3
Trx1 B 1
Trx1 C 1
Trx2 E 3
Trx2 B 1
Trx3 B 1
Trx3 C 4
Trx4 D 1
Trx4 E 1
Trx4 A 1
Trx5 F 5
Trx5 B 3
Trx5 C 2
Trx5 D 1", header=T)

#making the boolean matrix   
library(reshape2)
dat2 <- melt(dat)
w <- dcast(dat2, Items~TrxID)
x <- as.matrix(w[,-1])
x[is.na(x)] <- 0
x <- apply(x, 2,  function(x) as.numeric(x > 0))  #recode as 0/1
v <- x %*% t(x)                                   #the magic matrix 
diag(v) <- 0                                      #repalce diagonal
dimnames(v) <- list(w[, 1], w[,1])                #name the dimensions
v

For the graphing maybe…

g <- graph.adjacency(v, weighted=TRUE, mode ='undirected')
g <- simplify(g)
# set labels and degrees of vertices
V(g)$label <- V(g)$name
V(g)$degree <- degree(g)
plot(g)