Hi Elaine, It's perfectly OK to develop regression models that predict wing data based on body mass. One problem that you should address is the degree of correlation among the data due to shared evolutionary history. 5 species is too small a sample to check for phylogenetic "signal" in your data (rule of thumb is about 20 spp. are necessary). One way would be to continue to ignore phylogenetic correlations, and use lm(). A better way would be to include estimates of the phylogenetic correlations in the analysis, using a Brownian motion model, and a known phylogeny. You can use the gls() function in package nlme for this. However, if you use the predict() function on a gls object, you will still only get predictions assuming the new species are independent of all the others. If your new species are nested within the phylogeny of the original 5 species, you need to make some adjustments to your predictions. See
Garland, T., Jr., and A. R. Ives. 2000. Using the past to predict the present: Confidence intervals for regression equations in phylogenetic comparative methods. American Naturalist 155:346-364. Another problem that you might encounter is multiple measurements (on different individuals) per species. See: Ives, A. R., P. E. Midford, and T. Garland, Jr. 2007. Within-species variation and measurement error in phylogenetic comparative methods. Systematic Biology 56:252-270 You should also check out the book by Emmanuel Paradis: http://ape.mpl.ird.fr/APER.html There are numerous other references in this field, but these will get you most of the way. Also, the r-sig-phylo email list can be of assistance. Cheers, Simon. On 27/08/12 12:46, R. Michael Weylandt wrote: > Hi Elaine, > > As posed, your question is rather difficult to respond to properly. > I'd imagine the answer is the predict() function, but you'll probably > want to rephrase if that's not enough to get what you want. What > worries me is that I would be inclined to model your bird species as > categorical variables (or perhaps use mixed models) so I'm not sure > it's valid to extrapolate to different species. > > As far as asking questions in a technical forum goes, here are two > helpful general links: > > www.catb.org/esr/faqs/smart-questions.html > http://mattgemmell.com/2008/12/08/what-have-you-tried/ > > and some R specific advice here: > http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example > > Cheers, > Michael > > On Sun, Aug 26, 2012 at 9:33 PM, Elaine Kuo <elaine.kuo...@gmail.com> wrote: >> Hello, >> >> I have weight and wing length data of 5 kinds of birds (B1 to B5). >> Weight data (predictor) and wing data (response) were used to generated a >> simple regression. >> (using lm) >> >> Now some weight data are found but without wing data (B6, B7). >> I want to use the generated simple regression to produce wing data of B6 >> and B7 by their weight data. >> Please kindly advise R code to carry it out. >> Thank you. >> >> Elaine >> >> [[alternative HTML version deleted]] >> >> ______________________________________________ >> 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. > ______________________________________________ > 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. -- Simon Blomberg, BSc (Hons), PhD, MAppStat, AStat. Lecturer and Consultant Statistician School of Biological Sciences The University of Queensland St. Lucia Queensland 4072 Australia T: +61 7 3365 2506 email: S.Blomberg1_at_uq.edu.au http://www.uq.edu.au/~uqsblomb/ Policies: 1. I will NOT analyse your data for you. 2. Your deadline is your problem. Statistics is the grammar of science - Karl Pearson. [[alternative HTML version deleted]] ______________________________________________ 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.
