When should you call it PCoA vs. PCA

Asked Aug 21 '13 at 20:15

Active Oct 03 '20 at 14:09

Viewed 661 times

I'm reviewing a paper. In it the authors use a genetic distance metric to created a distance matrix between subjects, then they run classic MDS on this distance matrix. Throughout the MS they call this a principal components analysis (PCA). I was trained to call this principal coordinates analysis (PCoA) or classic Torgerson's multidimensional scaling. Can anyone clarify which terms are correct, and if more than one, which is preferred?

edited Oct 03 '20 at 14:09

ttnphns

57,480
49
284
501

asked Aug 21 '13 at 20:15

lef

3

See http://stats.stackexchange.com/questions/14002/whats-the-difference-between-principal-components-analysis-and-multidimensional – Ellis Valentiner Aug 21 '13 at 20:34
In PCA, you usually have a objects by variables dataset as the input. In PCoA, the input is a square distance matrix.This is their difference. Main math operations and theory behind both are essentially the same. – ttnphns Oct 03 '20 at 14:13

When should you call it PCoA vs. PCA

0 Answers0