[R-sig-Geo] Principal Component Analysis - Selecting components? + right choice?

[Corrado]

Dear R-sig-geo friends,

I have been testing some of the suggested approaches, and they seem too 
confirm each others conclusion.

I think one of the nice things about PC is that it does actually generate 
variables that are a mixture of the original variables (in this case a 
combination of precipitation and temperature), and, in a sense, it seems to 
account for interaction between variables. Do you think this statement is 
true?

Best,

-- 
Corrado Topi

Global Climate Change & Biodiversity Indicators
Area 18,Department of Biology
University of York, York, YO10 5YW, UK
Phone: + 44 (0) 1904 328645, E-mail: ct529 at york.ac.uk