ex cluster analysis.R

examples of cluster analysis in R

# *-----------------------------------------------------------------
# | PROGRAM NAME: ex cluster analysis.R
# | DATE: 3/6/20 
# | CREATED BY: MATT BOGARD 
# | PROJECT FILE:        
# *----------------------------------------------------------------
# | PURPOSE: basic mechanics of cluster analysis
# *----------------------------------------------------------------


### motivated by quick R https://www.statmethods.net/advstats/cluster.html

# rm(list=ls()) # get rid of any existing data 

options("scipen" =100, "digits" = 4) # override R's tendency to use scientific notation

library(psych) # use data from psych library

#Loading the dataset
df = bfi

# Prepare Data
df <- bfi[,1:24]                                    #select variables to include in fa
df<- na.omit(df) # listwise deletion of missing

#---------------------------------
# kmeans clustering
#--------------------------------

df <- scale(df) # standardize variables

# Determine number of clusters
wss <- (nrow(df)-1)*sum(apply(df,2,var))
for (i in 2:15) wss[i] <- sum(kmeans(df,
                                     centers=i,iter.max = 50)$withinss)
plot(1:15, wss, type="b", xlab="Number of Clusters",
     ylab="Within groups sum of squares")

# based on plot determine # of clusters (~ 6)

# K-Means Cluster Analysis
fit <- kmeans(df, 5) # 5 cluster solution

# get cluster means
aggregate(df,by=list(fit$cluster),FUN=mean)

# append cluster assignment
df <- data.frame(df, fit$cluster)


# Cluster Plot against 1st 2 principal components

# vary parameters for most readable graph
library(cluster)
clusplot(df, fit$cluster, color=TRUE, shade=TRUE,
         labels=2, lines=0)

#----------------------------------------------
# hierarchical clustering
#----------------------------------------------

# Ward Hierarchical Clustering
d <- dist(df, method = "euclidean") # distance matrix
fit <- hclust(d, method="ward")
plot(fit) # display dendogram
groups <- cutree(fit, k=5) # cut tree into 5 clusters
# draw dendogram with red borders around the 5 clusters
rect.hclust(fit, k=5, border="red")

#--------------------------------------
# Model Based Clustering
#-------------------------------------

library(mclust)
fit <- Mclust(df)
plot(fit) # plot results
summary(fit) # display the best model