[R-sig-eco] pairwise distance metrics

[Gavin Simpson]

On Sun, 2010-01-24 at 17:14 -0800, Dan Rabosky wrote:
> Hi -
> 
> This isn't really an R question (but the analysis is performed in  
> R!). Anyway, I'm trying to compute pairwise distances between a set  
> of sites, where the site data (habitat) consist of:
> 
> (1) Percent cover of various plant groups (eg "shrub", "tussock grass")
> 
> and
> 
> (2) several additional quantitative (non percentage) metrics, such as  
> the volume of dead wood, the mean diameter of shrubs, and the  
> "hardness" of the soil at each site.
> 
> I'm having a hard time deciding which distance metric is best for  
> computing an overall pairwise distance between sites, given this  
> mixture of datatypes. I have never seen proportional similarity  
> metrics used for such data and am unsure whether they would be  
> acceptable - although Bray-Curtis distances seem to accord well with  
> my intuition about how I think the data are structured. Actually, I'm  
> not familiar with any studies that have computed pairwise distances  
> for such mixed data, so don't have much to go on.

One such mixed mode dissimilarity coefficient is Gower's general
(dis)similarity coefficient. The Environmetrics Task View at CRAN
contains details of which packages can calculate this coefficient:

http://cran.r-project.org/web/views/Environmetrics.html

Package 'cluster' using function daisy() would be the most direct way as
this package is distributed as part of R.

HTH

G

> 
> Thanks for your help,
> ~Dan Rabosky
> 
> 
> 
> 
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
-- 
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
 Dr. Gavin Simpson             [t] +44 (0)20 7679 0522
 ECRC, UCL Geography,          [f] +44 (0)20 7679 0565
 Pearson Building,             [e] gavin.simpsonATNOSPAMucl.ac.uk
 Gower Street, London          [w] http://www.ucl.ac.uk/~ucfagls/
 UK. WC1E 6BT.                 [w] http://www.freshwaters.org.uk
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%