flows <- read.csv("gn2-flows-10ge.data", col.names=c("source","dest","proto","sip","dip","sport","dport","packets","bytes"))
cor(flows[flows$proto=="TCP",4:9])
tcp <- flows[flows$proto=="TCP",4:9]
summary(tcp)
tcp.pca <- princomp(tcp, cor=T)
tcp.pca
plot(tcp.pca)
loadings(tcp.pca)
scores(tcp.pca)
tcp.pc <- predict(tcp.pca)
plot(tcp.pc[,1:2])
pary <- read.table("adragr.data", col.names=c("source", "dest", "sip", "dip", "bytes"))
akamai <- pary[pary$source=="195.113.232.88", c(2,4,5)]
akamai.dist <- dist(akamai[,2:3])
akamai.dend <-hclust(akamai.dist, method="average")
plot(akamai.dend)