Hi again,
Reproducing part of my data (which indeed are here well standardized), I
obtain similar results (with regard to the R output) than previously.
These are "part" of an experimentation where people with different athletic
(ath) training exercised during 120 minutes. Plasma lactate was one of the
metabolic variables examined during this study.
Thanks for your help and the time spent for me
Best regards
Virgile
id=sort(c(rep(seq(1:7),15)))
time=factor(c(rep(c(0,20,40,60,80,100,120),15)))
ath=factor(c(rep(1,56),rep(2,49)))
lact=c(rnorm(8,1.27,0.32),rnorm(8,1.48,0.58),rnorm(8,1.39,0.48),rnorm(8,1.71,0.44),rnorm(8,1.78,0.56),rnorm(8,1.79,0.57),rnorm(8,1.71,0.4),rnorm(7,1.35,0.18),rnorm(7,2.15,1.4),rnorm(7,1.82,1.08),rnorm(7,2.13,1.2),rnorm(7,2.3,1.23),rnorm(7,2.38,1.31),rnorm(7,2.7,1.57))
data=data.frame(id,time,ath,lact)
lmelac.ex=lmer(lact~time*ath+(1|id),na.action=na.omit,data=data)
summary(lmelac.ex)
mult2 <- glht(lmelac.ex, linfct = mcp("ath:time" = "Tukey"))
Erreur dans mcp2matrix(model, linfct = linfct) :
Variable(s) ‘ath:time’ have been specified in ‘linfct’ but cannot be found
in ‘model’!
mult2 <- glht(lmelac.ex, linfct = mcp(ath:time = "Tukey"))
Erreur : syntax error
Dieter Menne wrote:
>
>
>
> Billeke wrote:
>>
>> I'm a bit lost while performing multiple comparisons on a lme model of
>> that type:
>>
>> lmeglu=lme(glucose~Ath*tim,random=~1|Vol,na.action=na.omit,data=data)
>> multc = glht(lmeglu, linfct = mcp(Ath = "Tukey", tim = "Tukey"))
>>
>> This works fine for identifying the effect of each factor. However, when
>> I look for their interactions, l only obtain error messages.
>> For example this leads to a syntax error...
>> multc2 <- glht(lmeglu, linfct = mcp(Ath:tim = "Tukey", tim = "Tukey"))
>>
>> I try to identify/locate differences in glucose concentration (glucose in
>> the model) between groups (Ath) at each time point (tim).
>>
>>
>
> Following the documentation, glht should not work with lme at all, but I
> know it does and I have used it several times without interactions. But I
> remember that the author, Torsten Hothorn, mumbled something like "keeping
> track with lme/lme4 variants is like baby-sitting fleas" when I contacted
> him on the subject, so there might be reasons the lme variant is
> undocumented.
>
> If you could prepare a small, reproducible example (use random data) that
> shows the effect and the detailed error message, it could help tracking
> the problem down.
>
> Dieter
>
>
>
>
--
View this message in context:
http://www.nabble.com/Multiple-comparison-on-lme-model-with-2-fixed-factors-tp25002051p25006669.html
Sent from the R help mailing list archive at Nabble.com.
______________________________________________
R-help@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.
