Antony Unwin wrote: > > On 27 Sep 2007, at 5:11 pm, Peter Dalgaard wrote: > >> There was a competition in 2004, and this is the display that won. > > Thanks for clearing that up. > >> It was deliberately designed as a "show-off" for the home page, and as >> such, I don't think it can be the same sort of graphic that you'd use >> for real analysis. It does have the nice feature of displaying results >> of simple, yet non-trivial statistical analyses (PCA, clustering) >> without requiring a lengthy explanation. > > Do you really think of PCA and clustering as "statistical" analyses? > And anyway, surely it's a factor analysis not PCA? Perhaps lengthy > explanations are needed after all. I would need an explanation for > the curious plots lower right. It's probably due to my sheltered > upbringing, as I was always taught to keep factor analyses at a > respectable distance. > > I liked Hadley's comment. It struck me as fortun(at)e. > Hmm, well, ... Actually it _is_ PCA (it says so: princomp), but it quite possibly ought to have been a factor analysis. The former is often used as a substitute for the latter (after rescaling) -- SPSS does this, causing a good deal of confusion to users expecting factanal () to have an option for PCA -- but that's not exactly a good reason to perpetuate it.
Whether it counts as statistical analysis? I tend to think that it does, although it isn't among the tools that I reach for most frequently. -- O__ ---- Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
