#Read Preference Data
htv <- read.csv("R/BendigoHtV.csv", header=T)
htv
## Party NAT GRN RUA LAB LIB ANJ IND FAM
## 1 NAT 1 8 5 6 2 7 4 3
## 2 GRN 5 1 8 2 4 3 7 6
## 3 RUA 4 8 1 7 5 6 2 3
## 4 LAB 4 2 7 1 5 3 8 6
## 5 LIB 2 7 8 6 1 4 5 3
## 6 ANJ 7 2 8 3 4 1 5 6
## 7 IND 4 7 2 8 5 6 1 3
## 8 FAM 3 8 4 7 2 6 5 1
#Convert preferences to points matrix
#Second preference = 7 points
row.names(htv) <- htv[,1]
htv <- as.matrix(9-htv[,-1])
#Remove self-preference (first preference)
diag(htv) <- 0
htv
## NAT GRN RUA LAB LIB ANJ IND FAM
## NAT 0 1 4 3 7 2 5 6
## GRN 4 0 1 7 5 6 2 3
## RUA 5 1 0 2 4 3 7 6
## LAB 5 7 2 0 4 6 1 3
## LIB 7 2 1 3 0 5 4 6
## ANJ 2 7 1 6 5 0 4 3
## IND 5 2 7 1 4 3 0 6
## FAM 6 1 5 2 7 3 4 0
#Number of points received by parties
#Each partie can distribute 28 points (1-7)
#Max=7*7, min=7*1
barplot(sort(colSums(htv)), ylim=c(7,49))
abline(h=7, col="red")
abline(h=49, col="red")
#Calculate net scores (strength of ties) between parties
htv[lower.tri(htv)] <- htv[lower.tri(htv)]+htv[upper.tri(htv)]
htv[upper.tri(htv)] <- 0
htv
## NAT GRN RUA LAB LIB ANJ IND FAM
## NAT 0 0 0 0 0 0 0 0
## GRN 5 0 0 0 0 0 0 0
## RUA 9 6 0 0 0 0 0 0
## LAB 6 11 8 0 0 0 0 0
## LIB 10 6 6 4 0 0 0 0
## ANJ 9 9 6 10 8 0 0 0
## IND 7 8 9 5 10 9 0 0
## FAM 13 4 12 8 10 6 10 0
#Network analysis
#Spinglass community detection based on the strength of ties
library(igraph)
##
## Attaching package: 'igraph'
## The following objects are masked from 'package:stats':
##
## decompose, spectrum
## The following object is masked from 'package:base':
##
## union
g <- graph.adjacency(htv, weighted=T, mode = "undirected")
c <- spinglass.community(g)
plot(c,g)
