#Chay Bagan CIS 430 HW3
Universities = read.csv("Universities.csv", header = TRUE)

Universities$State = NULL
Universities$Public..1...Private..2. = NULL

head(Universities)

Universities = na.omit(Universities)

Universities = princomp(Universities[2:18], cor=TRUE, score=TRUE)
summary(Universities)
plot(Universities)
#biplot(Universities)
#We should normalize the data because a PCA without normalization will always perform worse than one with
#With that being said luckily in this dataset we can still see a great deal of information from PCA without normalization
#The key components are clearly 1 and 2 as they account for atleast 4-5x the amount of variance compared to any other component