What's a good strategy to get a decent overview of big correlation matrices or pairs?

Question

Often some data.frame contains 20+ variables and you want to get a first overview (of the correlation structure). Even on a 30\" screen you run quickly out of s

Answer 1

What about doing a PCA on the correlation matrix? Then the angle between variables show their correlation.

library(HSAUR)
heptathlon
round(cor(heptathlon[,-8]),2)   # correlations [without score]

require(vegan)
PCA <-  rda(heptathlon[,-8], scale=TRUE)   # run a PCA
biplot(PCA, display = "species")   # correlation biplot
#  The angles between descriptors in the biplot reflect their correlations