library(MASS)
 library(lattice)
 library(cluster)
 Sigma<- matrix(c(0.04,0,0,0.04),c(2,2))
 x<- mvrnorm(15,c(1.5,2), Sigma)
 y<- mvrnorm(15,c(2,1.5), Sigma)
 z<- mvrnorm(15,c(1.5,1), Sigma)
 w<- mvrnorm(15,c(1.5,0), Sigma)

 x<- cbind(x,1)
 y<- cbind(y,2)
 z<- cbind(z,3)
 w<- cbind(w,4)
 data1 <- data.frame(rbind(x,y,z,w))
 dataset<- cbind(data1$X1,data1$X2)
 distmatrix<- dist(dataset)
 mycol <- c("red", "black", "green", "blue")
 plot(dataset,col=mycol[data1$X3],main="Real clusters")


 hrs<-hclust(distmatrix,method="single")
 hrc<-hclust(distmatrix,method="complete")
 hra<-hclust(distmatrix,method="average")
 hrw<-hclust(distmatrix,method="ward")

 membs<- cutree(hrs,k=4)
 membc<- cutree(hrc,k=4)
 memba<- cutree(hra,k=4)
 membw<- cutree(hrw,k=4)

 c(sum(membs==1),sum(membs==2),sum(membs==3),sum(membs==4))

 c(sum(membc==1),sum(membc==2),sum(membc==3),sum(membc==4))

 c(sum(memba==1),sum(memba==2),sum(memba==3),sum(memba==4))

 c(sum(membw==1),sum(membw==2),sum(membw==3),sum(membw==4))

 hrs$order
 hrc$order
 hra$order
 hrw$order
 plot(hrs,hang=-1); plot(hra,hang=-1); plot(hrc,hang=-1); plot(hrw,hang=-1)



 dv<-diana(distmatrix)
 plot(dv)
 dv1 <- cutree(as.hclust(dv), k = 4)
 dv1
 c(sum(dv1==1),sum(dv1==2),sum(dv1==3),sum(dv1==4))



(cl <- kmeans(dataset, 4, algorithm="MacQueen"))

 plot(dataset, col = mycol[cl$cluster], main="K-means")
 points(cl$centers, col = mycol, pch = 8)



 pamx<-pam(distmatrix,4)
 pamx
 plot(pamx)


 library("kohonen")
 data("wines")
 wines.sc <- scale(wines)
 set.seed(7)
 wine.som <- som(data = wines.sc,grid=somgrid(5,4,"hexagonal"))
 plot(wine.som, main= "Wine data")


 fuzzyc <- fanny(distmatrix,4)
 fuzzyc
 plot(fuzzyc)


 library(cluster)
 data(ruspini)
 plot(ruspini,main="Ruspini data")
 distmatrix <- dist(ruspini)



 mycol <- c("red", "black", "green", "blue")
 hca<-hclust(distmatrix,method="average")
 memba<- cutree(hca,k=4)
 c(sum(memba==1),sum(memba==2),sum(memba==3),sum(memba==4))
 dataset<- cbind(ruspini$x,ruspini$y,memba)
 plot(dataset,col=mycol[memba],main="Hierarchical Clustering - Average Linkage")


 dv<-diana(distmatrix)
 dc <- cutree(as.hclust(dv), k = 4)
 dataset<- cbind(ruspini$x,ruspini$y,dc)
 plot(dataset,col=mycol[dc],main="DIANA clustering")


 cl <- kmeans(ruspini, 4, algorithm="MacQueen")
 plot(ruspini, col = mycol[cl$cluster], main="K-means")
 points(cl$centers, col = mycol, pch = 8)


 pamx<-pam(distmatrix,4)
 clusters<-pamx$clustering
 dataset<- cbind(ruspini$x,ruspini$y,clusters)
 plot(dataset,col=mycol[clusters],main="Partitioning Around Medoids")


 library("kohonen")
 ruspini.sc <- scale(ruspini)
 set.seed(7)
 ruspini.som <- som(data = ruspini.sc,grid=somgrid(2,2,"rectangular"))
 plot(ruspini.som, main= "Ruspini data")



 fuzzyc <- fanny(distmatrix,4)
 clusters<-fuzzyc$clustering
 dataset<- cbind(ruspini$x,ruspini$y,clusters)
 plot(dataset,col=mycol[clusters],main="Fuzzy Analysis Clustering")