library(cluster)
library(rattle)
library(NbClust)
data(wine,package="rattle")
head(wine)
##Exercise1###
Column_1<-as.numeric(wine[,1])
scale(Column_1)
wine<-wine[,-1]

wssplot<-function(data=wine,nc=15,seed=1234){
  wss<-(nrow(data-1)*sum(apply(data,2,var))
  for (i in 2:nc){
    set.seed(seed)
    wss[i]<-sum(kmeans(data,centers=i)$withinss) }
  plot(1:nc,wss,type="b",xlab="Number of Clusters",
       ylab="Within groups sum of squares")
  }
wssplot(df)


##Exercise 2####
library(NbClust)
set.seed(1234)
nc <- NbClust(data=wine, min.nc=2, max.nc=15, method="kmeans")
barplot(table(nc$Best.n[1,]),
        xlab="Numer of Clusters", ylab="Number of Criteria",
        main="Number of Clusters Chosen by 26 Criteria")

####Exercise 3#####
##Clusters via method 2 = 3 clusters

###Exercise 4####
fit.km<-kmeans(wine, centers= 3, nstart= 2)

###Exercise 5#####
table(fit.km$cluster)
table(wine$Type)
##It does not appear that accurate

##Exercise 6###
clusplot(pam(fit.km$cluster,3))
###given the clusters cover 100% of the data points, yes it appears accurate