CODEWITHSUNDEEP
r
1478963
1478963

Reputation: 1216

Row Uniqueness in R

I have a list of undirected graph edges like follows:

A B

B A

C D

I would like to extract unique edges.

So the end result would return only

A B  (or)  B A
C D

I know how to do this in python, and how to select unique rows in R. But how would I do this additional constraint in R.

Upvotes: 1

Views: 224

Answers (2)

MrFlick
MrFlick

Reputation: 206283

If you are working with graph data, you might also consider using the igraph package in R. If this is your input data

dd<-matrix(c("A", "B", "C", "B", "A", "D"), ncol=2)
#     [,1] [,2]
# [1,] "A"  "B" 
# [2,] "B"  "A" 
# [3,] "C"  "D"

Then you could do

library(igraph)
gg<-simplify(graph.edgelist(dd, directed=F))
get.edgelist(gg)

#     [,1] [,2]
# [1,] "A"  "B" 
# [2,] "C"  "D"

Here the simplify remove the redundant bidirectional edge from A-B. Then you can also do fun stuff like plot it

plot(gg)

enter image description here

Upvotes: 1

ping
ping

Reputation: 1336

Maybe there's a better way, but one option would be to sort the data so that each row is in alphabetical order, so that rows containing the same items will become identical, and using unique on the result:

edges <- data.frame(a=c("A", "B", "C"), b=c("B","A","D")) 
unique.edges <- unique(t(apply(edges,1, sort)) )

output here is a character matrix like:

#     [,1] [,2]
#[1,] "A"  "B" 
#[2,] "C"  "D"

Upvotes: 2

Related Questions