[R] Multivariate response methods question

[Gavin Simpson]

On Thu, 2008-01-03 at 10:10 -0600, markleeds at verizon.net wrote:
> >From: Max <mnevill at exitcheck.net>
> >Date: 2008/01/03 Thu AM 09:31:40 CST
> >To: r-help at stat.math.ethz.ch
> >Subject: Re: [R] Multivariate response methods question
> 
> Hi Max:  multinom allows ( and probably polr also ) allows a
> categorical response that can take on any number of values so I am
> unsure what you mean by not
> allowing multiple responses ?
> 
>                                 Mark

Mark,

For example, say we have abundance data on 10 species from a sample of
acid grassland quadrats, expressed on one of the cover scales ecologists
often use (Van der Maarel, Braun-Blanquet, etc) or as presence/absence.
We have associated soil chemistry data and other morphological variables
that we feel are important for predicting the abundances of the 10
species. The responses are the species and the predictors the soil
chemistry and morphological variables. In this example the responses are
multivariate *and* ordinal (or nominal in the case of presence/absence).

The OP might also want to look at Canonical Correspondence Analysis,
such as that implemented in cca in package vegan (?predict.caa and type
= "response"). It can handle multivariate responses of this nature. Or
look at capscale() (also in vegan) which fits constrained analysis of
principle coordinates and an appropriate dissimilarity coefficient for
ordinal data (e.g. Gower's coefficient for mixed data, such as that
implemented in function distance in package analogue).

These methods are more familiar to me as an ecologist and from fields
such as quantitative geology and chemometrics (of old), as multivariate
methods, than the VGAM approach I recommended in my other reply.

G

> 
> >Dimitris,
> >
> >Thankyou for pointing me at those functions. It seems like they're 
> >limited to only one response variable, unless there's a way of 
> >inserting multiple responses that I'm unaware of (which could easily be 
> >true) into the formulas.
> >
> >Thanks,
> >
> >-Max
> >
> >
> >
> >After serious thinking Dimitris Rizopoulos wrote :
> >> you can have a look at the following functions:
> >>
> >> help("polr", package = "MASS")
> >> help("lrm", package = "Design")
> >> help("multinom", package = "nnet")
> >>
> >>
> >> I hope it helps.
> >>
> >> Best,
> >> Dimitris
> >>
> >> ----
> >> Dimitris Rizopoulos
> >> Ph.D. Student
> >> Biostatistical Centre
> >> School of Public Health
> >> Catholic University of Leuven
> >>
> >> Address: Kapucijnenvoer 35, Leuven, Belgium
> >> Tel: +32/(0)16/336899
> >> Fax: +32/(0)16/337015
> >> Web: http://med.kuleuven.be/biostat/
> >>      http://www.student.kuleuven.be/~m0390867/dimitris.htm
> >>
> >>
> >> ----- Original Message ----- 
> >> From: "Max" <mnevill at exitcheck.net>
> >> To: <r-help at stat.math.ethz.ch>
> >> Sent: Wednesday, January 02, 2008 9:36 PM
> >> Subject: [R] Multivariate response methods question
> >>
> >>
> >>> Hi Everyone,
> >>> 
> >>> I have some data that predicts both a nominal and ordinal response
> >>> variable.  I was wondering what packages in R would help me analyze 
> >>> the
> >>> data?
> >>> 
> >>> I was also curious if anyone could recomend me some textbooks that
> >>> would help with the analysis of such data? I have the 5th edition of
> >>> "Applied Multivariate Statistical Analysis" by Richard A. Johnson 
> >>> and
> >>> Dean W. Wichern already, but that doesn't seem to cover the material
> >>> I'm looking for.
> >>> 
> >>> Thanks,
> >>> 
> >>> -Max
> >>> 
> >>> ______________________________________________
> >>> R-help at 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.
> >>> 
> >>
> >>
> >> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
> >>
> >> ______________________________________________
> >> R-help at 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 at 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 at 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.
-- 
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
 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
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%