[R-sig-eco] mrpp between pairs of factor levels

2010-09-24 Thread cristabel.duran
Dear list, I'm performing a mrpp analysis for my data (spp), the grouping factor is mhab. mhab has five levels: b br c nd nongap t So, for a mrpp with factor mhab with all levels I'm doing: mrpp(data=spp, grouping=mhab, dist=bray, permutations=1000) BUT now I want is to

[R-sig-eco] Suggestion for Paired Population Trend Analysis

2010-09-24 Thread Peter Jørgensen
Dear R-eco'ers, I am about to analyze a dataset with population trends from two decades (1990's and 2000's) for about 500 populations. I would like to incorporate both trends in the same model (as the response variable) to get an idea of how the populations respond to different environmental

Re: [R-sig-eco] mrpp between pairs of factor levels

2010-09-24 Thread Luciano Selzer
Cristabel, I don't know if this what you are after. Why don't you just use the subset of data you are interested in. mrpp(data=subset(spp, mhab == b mhab == nd), grouping=mhab, dist=bray, permutations=1000) or spp1 - subset(spp, mhab == b mhab == nd) spp1$mhab - factor(spp1$mhab) #to drop

Re: [R-sig-eco] Suggestion for Paired Population Trend Analysis

2010-09-24 Thread Ben Bolker
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 10-09-24 07:26 AM, Peter Jørgensen wrote: Dear R-eco'ers, I am about to analyze a dataset with population trends from two decades (1990's and 2000's) for about 500 populations. I would like to incorporate both trends in the same model (as the