In my graph-package (as in graph theory, nodes connected by edges) I have a vector indicating for each edge the node of origin from, a vector indicating for each edge the node of destination to and a vector indicating the curve of each edge curve.
By default I want edges to have a curve of 0 if there is only one edge between two nodes and curve of 0.2 if there are two edges between two nodes. The code that I use now is a for-loop, and it is kinda slow:
curve <- rep(0,5)
from<-c(1,2,3,3,2)
to<-c(2,3,4,2,1)
for (i in 1:length(from))
{
if (any(from==to[i] & to==from[i]))
{
curve[i]=0.2
}
}
So basically I look for each edge (one index in from and one in to) if there is any other pair in from and to that use the same nodes (numbers).
What I am looking for are two things:
A way to identify if there is any pair of nodes that have two edges between them (so I can omit the loop if not)
A way to speed up this loop
#
EDIT:
To make this abit clearer, another example:
from <- c(4L, 6L, 7L, 8L, 1L, 9L, 5L, 1L, 2L, 1L, 10L, 2L, 6L, 7L, 10L, 4L, 9L)
to <- c(1L, 1L, 1L, 2L, 3L, 3L, 4L, 5L, 6L, 7L, 7L, 8L, 8L, 8L, 8L, 10L, 10L)
cbind(from,to)
from to
[1,] 4 1
[2,] 6 1
[3,] 7 1
[4,] 8 2
[5,] 1 3
[6,] 9 3
[7,] 5 4
[8,] 1 5
[9,] 2 6
[10,] 1 7
[11,] 10 7
[12,] 2 8
[13,] 6 8
[14,] 7 8
[15,] 10 8
[16,] 4 10
[17,] 9 10
In these two vectors, pair 3 is identical to pair 10 (both 1 and 7 in different orders) and pairs 4 and 12 are identical (both 2 and 8). So I would want curve to become:
[1,] 0.0
[2,] 0.0
[3,] 0.2
[4,] 0.2
[5,] 0.0
[6,] 0.0
[7,] 0.0
[8,] 0.0
[9,] 0.0
[10,] 0.2
[11,] 0.0
[12,] 0.2
[13,] 0.0
[14,] 0.0
[15,] 0.0
[16,] 0.0
[17,] 0.0
(as I vector, I transposed twice to get row numbers).