R help - Apr 2006 - Sign of loadings and scores from PCA in cross validation

Hi,

The help file on princomp says that 'The signs of the columns of the 
loadings and scores are arbitrary, and so may differ between differnt 
programs for PCA, and even between different builds of R'. Does anyone 
know 

1. Whether this applys to all functions like svd, eigen ... etc? 

2. During a leave-one-out cross step, I notice the loadings also change 
sige when a different observation is held out. This causes the sign change 
of principal component score of the held-out observation. Since I use the 
same version of R, same procedure, and same program, to make the 
leave-one-out PC scores consistent, does that mean I have the take the 
absolute value of these scores before analyzing them?

Thanks in advance for any help!

Anne
	[[alternative HTML version deleted]]