R | NbClust聚類分析

library("factoextra")
library("NbClust")
data("USArrests")

### 數據標準化
data =  scale(USArrests)
head(data, n=5)


#確認分類數目
set.seed(1234) #因爲method選擇的是kmeans，所以如果不設定種子，每次跑得結果可能不同
nb_clust <- NbClust(data,  distance = "euclidean",
                    min.nc=2, max.nc=15, method = "kmeans",
                    index = "alllong", alphaBeale = 0.1)

#確認分類數目
fviz_nbclust(data, kmeans, method = "wss") + geom_vline(xintercept = 4, linetype = 2)

#利用k-mean是進行聚類
km_fit = kmeans(data, 4, nstart = 25) ## nstart 重複迭代kmeans的次數
print(km_fit)

fviz_cluster(km_fit, data, palette = c("#2E9FDF", "#00AFBB", "#E7B800", "#FC4E07"),  geom = "point" ,ellipse.type = "convex", star.plot = TRUE, repel = TRUE, ggtheme = theme_grey() )

#先求樣本之間兩兩相似性 
result <- dist(data, method = "euclidean")
#產生層次結構
result_hc <- hclust(d = result, method = "ward.D2")
#進行初步展示
fviz_dend(result_hc, cex = 0.6)

fviz_dend(result_hc, k = 4, cex = 0.5, k_colors = c("#2E9FDF", "#00AFBB", "#E7B800", "#FC4E07") )