I have analyzed some datasets using prcomp and some of my data is nice and amenable to PCA. But the summary of one set is showing that at least 6 components are needed to cover 80% of the variance. I've checked and the correlation matrix should work better than the covariance matrix.
When I plot the results using biplot, is that plotting all components, taking all components into consideration or only the first two? If that is not the best way to look at my data with so much variance, then what is?
Also, does anyone know of a tutorial on how to plot a biplot of only the top values and not all - this would be useful as I have a few datasets with >1,000 entries.
Any help greatly appreciated.