library(cluster)
#get csv file
#survey<-survey[,2:11]
items <- c("id","informed","info_trust","certification_trust","region","fair_price","small_diverse","agbio_conserv","price","reliability","taste")
colnames(survey) <- items
# store labels in separate variable and exclude id from dataset to do clustering
id<-survey$id
survey$id<-NULL
#gower's distance, ward method, 5 clusters
data(survey)
d.sur <- daisy(survey, metric = "gower", stand = FALSE)
d.sur
surveyhclust<-hclust(d.sur, method="ward")
plot(surveyhclust, main="Default from hclust")
rect.hclust(surveyhclust,5)
groups.5<-cutree(surveyhclust,5)
sapply(unique(groups.5),function(g)id[groups.5==g])
a5=aggregate(survey,list(groups.5),FUN=mean)
x<-data.frame(Cluster=a5[,1], Freq=as.vector(table(groups.5)),a5[,-1])
x
write.table(x,file="gower ward five clus.csv", sep=",")
#then export cluster and id info to match up with survey responses
survey_2 <- data.frame(survey,groups.5,id)
write.table(survey_2, file="survey with clusters.csv",sep=";")
clusplot(survey, groups.5, color=TRUE, shade=TRUE, labels=2, lines=0, main= 'Customer groups')
#then relationship between clusters and socioeconomic data
library(psych)
data<-read.csv(file location and name here)
cluster<-data[,2]
informed<-data[,3]
info_trust<-data[,4]
cert<-data[,5]
region<-data[,6]
fair<-data[,7]
smalldiverse<-data[,8]
conserv<-data[,9]
price<-data[,10]
convenience<-data[,11]
taste<-data[,12]
freq_eggs<-data[,13]
where_eggs<-data[,14]
freq_meat<-data[,15]
where_meat<-data[,16]
gender<-data[,17]
age<-data[,18]
household<-data[,19]
mytable_freqegg <- table(cluster,freq_eggs)
mytable_freqegg
margin.table(mytable_freqegg, 1) # A frequencies (summed over B)
margin.table(mytable_freqegg, 2) # B frequencies (summed over A)
prop.table(mytable_freqegg, 1) # row percentages
chisq.test(mytable_freqegg)
mytable_whereegg <- table(cluster,where_eggs)
mytable_whereegg
margin.table(mytable_whereegg, 1) # A frequencies (summed over B)
margin.table(mytable_whereegg, 2) # B frequencies (summed over A)
prop.table(mytable_whereegg, 1) # row percentages
chisq.test(mytable_whereegg)
mytable_freqmeat <- table(cluster,freq_meat)
mytable_freqmeat
margin.table(mytable_freqmeat, 1) # A frequencies (summed over B)
margin.table(mytable_freqmeat, 2) # B frequencies (summed over A)
prop.table(mytable_freqmeat, 1) # row percentages
chisq.test(mytable_freqmeat)
mytable_wheremeat <- table(cluster,where_meat)
mytable_wheremeat
margin.table(mytable_wheremeat, 1) # A frequencies (summed over B)
margin.table(mytable_wheremeat, 2) # B frequencies (summed over A)
prop.table(mytable_wheremeat, 1) # row percentages
chisq.test(mytable_wheremeat)
mytable_gender <- table(cluster,gender)
mytable_gender
margin.table(mytable_gender, 1) # A frequencies (summed over B)
margin.table(mytable_gender, 2) # B frequencies (summed over A)
prop.table(mytable_gender, 1) # row percentages
chisq.test(mytable_gender)
mytable_age <- table(cluster,age)
mytable_age
margin.table(mytable_age, 1) # A frequencies (summed over B)
margin.table(mytable_age, 2) # B frequencies (summed over A)
prop.table(mytable_age, 1) # row percentages
chisq.test(mytable_age)

mytable_household <- table(cluster,household)
mytable_household
margin.table(mytable_household, 1) # A frequencies (summed over B)
margin.table(mytable_household, 2) # B frequencies (summed over A)
prop.table(mytable_household, 1) # row percentages
chisq.test(mytable_household)
mytable_edu <- table(cluster,edu)
mytable_edu
margin.table(mytable_edu, 1) # A frequencies (summed over B)
margin.table(mytable_edu, 2) # B frequencies (summed over A)
prop.table(mytable_edu, 1) # row percentages
chisq.test(mytable_edu)
mytable_income <- table(cluster,income)
mytable_income
margin.table(mytable_income, 1) # A frequencies (summed over B)
margin.table(mytable_income, 2) # B frequencies (summed over A)
prop.table(mytable_income, 1) # row percentages
chisq.test(mytable_income)