Analysis of Iris Dataset in R Programming

First of all you need to install ‘multigraph’ library if already not installed

• load ‘multigraph’ library

library(multigraph)

• Load Iris dataset as matrix in matrix variable

matrix <- data.matrix(iris)

• Matrix transpose because ‘cor’ method calculate column wise correlation

trn <- t(matrix)

• Calculate pairwise correlation (150*150) in pwcor variable

pwcor <- cor(trn)

• Save calculated (150*150) correlation in csv file

write.csv(pwcor, file = "correlation.csv")

• Define scope of node / edge / graph characteristics as list object

scp3 <- list(cex = 1, fsize = 3, pch = c(19, 15), lwd = 1.5, vcol = 2:3)

• Map pairwise correlation (150*150) to ‘circ2’ layout base graph

bmgraph(pwcor, layout = "circ2", scope = scp3, main = "Iris correlation using bmgraph")

• Calculate mean, median and summary

mean(pwcor)
median(pwcor)
summary(pwcor)

DONE :)

Feel free to extend this post and don’t forget to share with others … :)