[R-sig-eco] MANCOVA in R

[Jari Oksanen]

On Fri, 2011-11-25 at 10:47 +0100, marcelino.delacruz at upm.es wrote:
> Hi,
> 
> maybe adonis? e.g.,
> 
> library(vegan)
> adonis(Y~Lat+Lat2+Year+Lat:Year+Lat2:Year)
> 
> 
Michel,

This (adonis) does not do MANCOVA in its basic setting. Normal lm() does
multivariate linear models when the dependent variable is a matrix. If
it is a data.frame, you must cast it as a matrix:

lm(as.matrix(Y) ~  Year*(Lat + Lat2))

Old-fashioned people use name MANCOVA when a multiple linear model has
factors and interaction terms. 

The tests can be found with anova() of its results and details of the
fit with summary(). Although I said that adonis() does not do MANCOVA,
you can get the same F-values but with permutation statistics if you use
Euclidean distances:

adonis(dist(Y) ~  Year*(Lat + Lat2))

However, that should not be called as MANCOVA (because it is not
MANCOVA).

Cheers, Jari Oksanen

> Con fecha 25/11/2011, "Michel Rapinski" <mrapi090 at uottawa.ca> escribió:
> 
> >Hi,
> >
> >I am trying to perform an MANCOVA in R with permutation tests of
> >significance.
> >
> >The model I want to test is:
> >
> >Y~Lat+Lat2+Year+Lat:Year+Lat2:Year
> >
> >Where Year is the grouping variable, while Lat and Lat2 are continuous
> >variables representing Latitude and its square.
> >
> >Does anyone know how to do this in R?
> >
> >Thank you,
> >
> >Michel
> >
> >
> >Michel Rapinski, MSc. Student
> >Inst. of Plant Biology Research, Montreal Botanical Garden
> >Université de Montréal
> >Montréal, QC H1X 2B2
> >Tel: 514.772-1710
> >Fax: 514.872.9406
> >michael.rapinski at umontreal.ca
> >
> >University of Ottawa
> >mrapi090 at uottawa.ca
> >
> >_______________________________________________
> >R-sig-ecology mailing list
> >R-sig-ecology at r-project.org
> >https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
> 
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology