kmeans

data <- a0926[, -1]

# 去除第一欄編號

# 建立一個分群模型

##分4群，nstart=10 defaut執行10次 收斂資料區

km <- kmeans(data, centers = 4, nstart = 10)

kmeans.cluster <- kmeans(data, centers=4)

# Elbow Method 應用在 K-Means #資料太多...

fviz_nbclust(data,

FUNcluster = kmeans,# K-Means

method = "wss", # total within sum of square

k.max = 6 # max number of clusters to consider) +

labs(title="Elbow Method for K-Means")

# Determine number of clusters # it's work!!!!!!!!!

wss <- (nrow(data)-1)*sum(apply(data,2,var))

for (i in 2:15) wss[i] <- sum(kmeans(data,

centers=i)$withinss)

plot(1:15, wss, type="b", xlab="Number of Clusters",

ylab="Within groups sum of squares")

# 視覺化 k-means 分群結果(基於ggplot2的語法)

require(factoextra)

fviz_cluster(kmeans.cluster, # 分群結果

data = data, # 資料

geom = c("point"), # 點和標籤(point & label)

frame.type = "norm") # 框架型態

xlab = NULL,

ylab = NULL,

#跑分群之後畫分佈 #失敗

plot(formula = belonging ~ achieve, data = data, col = kmeans$cluster, main = "將歸屬感做分群", xlab = "學業成就", ylab = "歸屬感")

# 視覺化 k-means 分群結果(基於ggplot2的語法)

require(factoextra)

fviz_cluster(kmeans.cluster, # 分群結果

data = data, # 資料

geom = c("point"), # 點和標籤(point & label)

frame.type = "norm") # 框架型態

# Elbow Method 應用在 K-Means #資料太多...

fviz_nbclust(data,

FUNcluster = kmeans,# K-Means

method = "wss", # total within sum of square

k.max = 6 # max number of clusters to consider

) +

labs(title="Elbow Method for K-Means") +

geom_vline(xintercept = 3, # 在 X=3的地方

linetype = 2) # 畫一條垂直虛線

fit <- kmeans(data, 4)

library(cluster)

clusplot(data, fit$cluster, color=TRUE, shade=TRUE,

labels=2, lines=0)

fviz_cluster(kmeans.cluster, # 分群結果

data = data, # 資料

xlab = "data$belonging", ylab = "data$achieve",

geom = c("point"), # 點和標籤

frame.type = "norm") # 框架型態