It is somewhat similar. The procedure I describe is normalizing (using z-scores instead of raw data) the data. The difference, as far as I can tell, is that nor

It is somewhat similar. The procedure I describe is normalizing (using z-scores instead of raw data) the data. The difference, as far as I can tell, is that normalizing retains the correlation structure of the data, while whitening uncorrelates the observations. R is the lingua franca of the academic statisticians, so you might not derive a huge amount of value from this, but this question was asked a couple years ago on Stats.SE: http://stats.stackexchange.com/questions/53/pca-on-correlation-or-covariance http://stats.stackexchange.com/questions/53/pca-on-correlati...