Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Nick Cox <n.j.cox@durham.ac.uk> |
| To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
| Subject | st: RE: Principal Component Analysis accounting for clustering |
| Date | Tue, 19 Jun 2012 18:41:04 +0100 |
Yes; that's wrong. First, -pca- has no -cluster()- option. It's not clear why you thought there was one. Second, what would "accounting for clustering" look like? In PCA, at least as handled in Stata, there is no question of estimation or error structure as the PCA is just a mathematical transformation of the data. You may have in mind some flavour of factor analysis, and it may be that -sem- offers more of what you want. Nick n.j.cox@durham.ac.uk Armen Martirosyan i am doing Principal Component Analysis (PCA) but would like to also account for clustering. So when i write the following syntax pca var 1 var 2 var 3 var n, cluster (cluster variable) i am getting "option cluster not allowed". am i doing something wrong or it is just impossible to account for clustering in PCA? * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/