R语言中K-Means聚类算法

• K-Means聚类算法

# 设置工作空间
# 把“数据及程序”文件夹拷贝到F盘下，再用setwd设置工作空间
setwd("E:\\R_workspace\\R语言数据分析与挖掘实战\\chp5")

# 读入数据
Data <- read.csv("./data/consumption_data.csv", header = TRUE)[, 2:4]

View(Data)
fix(Data)

# center=3 为3个聚类中心，默认为  欧式距离
km <- kmeans(Data, center = 3)
print(km)

# 每组分类所占的比例
km$size / sum(km$size)

# 数据分组
aaa <- data.frame(Data, km$cluster)
Data1 <- Data[which(aaa$km.cluster == 1), ]
Data2 <- Data[which(aaa$km.cluster == 2), ]
Data3 <- Data[which(aaa$km.cluster == 3), ]

# 类别1的情况
summary(Data1)
View(Data1)

# 客户分群“1”的概率密度函数图
par(mfrow = c(1,3))
plot(density(Data1[, 1]), col = "red", main = "R")
plot(density(Data1[, 2]), col = "red", main = "F")
plot(density(Data1[, 3]), col = "red", main = "M")

# 客户分群“2”的概率密度函数图
par(mfrow = c(1, 3))
plot(density(Data2[, 1]), col="red", main = "R")
plot(density(Data2[, 2]), col="red", main = "F")
plot(density(Data2[, 3]), col="red", main = "M")

# 客户分群“3”的概率密度函数图
par(mfrow = c(1, 3))
plot(density(Data3[, 1]), col="red", main = "R")
plot(density(Data3[, 2]), col="red", main = "F")
plot(density(Data3[, 3]), col="red", main = "M")