G'day all,

 

Although PCA is only a rotation of axes through a point cloud I've always thought the algorithim assumed normalty, ot at least symmetry. I thought this came from the fact that the mean is usually subtracted from all the variables first (or am I getting PCA mixed up with factor analysis?).

 

However, recently I was taken to task over this by someone pointing out that a rotation of axes through a point cloud doesn't need to assume normality, and I think I agree with them.

 

Could someone please tell me if PCA assumes normality, and why? Or is it just some of the eigenvalue tests for significance that require normality.

 

Thanx

 

 

 

Christopher G Howden
Statistical Ecologist
(Work) 02 9895 7130
(Mob) 0410 689 945