[R] Results from mixed linear models

Julien Leblud julien.leblud at umons.ac.be
Thu Jul 12 09:37:00 CEST 2012


Dear list,

I'm quite confused when interpreting results from a mixed linear model.

For example, working on Iris data frame, I want to know the effect of 
species on slope of the model  "Petal.Length~Sepal.Length"

I write this :

data(iris)
reg01 <- lm(Petal.Length~Sepal.Length + Sepal.Length:Species, data=iris)
summary(reg01)

It gives me a summary table with lm for the first factor, i.e setosa, 
and the effects of each levels of Species on interactions, which mean on 
the slope.

Now, i would like a model with intercept =0. I should write this:

data(iris)
reg02 <- lm(Petal.Length~Sepal.Length + Sepal.Length:Species - 1 , 
data=iris)
summary(reg02)

...But results is meaningless for me...

Any idea to fit a model like this?
Thanks a lot !

Julien

-- 
*********************************************
Julien Leblud

Numerical Ecology of Aquatic Systems
Mons University, Belgium

http://econum.umons.ac.be/labo/

..............................<°}))><........

Ecologie Numérique des Milieux Aquatiques
Université de Mons

tel. +32 (0) 65 37 36 65

Pentagone (aile 3D)
6, Avenue du Champ de Mars
7000 Mons, Belgium



More information about the R-help mailing list