[R-sig-ME] Help GLMM Random intercept and slope

[ONKELINX, Thierry]

Enrique,

Your model is probably to complex for the data you have.

You'll need to provide more info about your data (number of groups of fFemale and fEmbrion and the number of levels per factor)

Best regards,

Thierry
________________________________________
Van: r-sig-mixed-models-bounces at r-project.org [r-sig-mixed-models-bounces at r-project.org] namens Enrique González Ortegón [quique.ortegon at gmail.com]
Verzonden: dinsdag 16 augustus 2011 1:01
Aan: r-sig-mixed-models at r-project.org
Onderwerp: [R-sig-ME] Help GLMM Random intercept and slope

I am using the function lmer of the library(lme4) to do a GLMM



All models with random intercept works successfully with all my factors.

M4 <- lmer (Stages2 ~ factor(Temperatura) * factor (Sal) *
factor(Food)*factor(Embrion) + (1 | fFemale/fEmbrion), data = PS, family =
binomial)



A model with a random intercept and slope using only one fixed factor in
the random part works successfully Example:
M4 <- lmer (Stages2 ~ factor(Temperatura) * factor (Sal) *
factor(Food)*factor(Embrion) + (1 + factor(Temperatura) |
fFemale/fEmbrion), data = PS, family = binomial)

but when I include in the random part several factors M4 ........... +(1 +
factor(Temperatura) +factor(Food) + factor (Sal)|
fFemale/fEmbrion)........., it
doesn't work.

#Warning message:In mer_finalize(ans, verbose) :
# function evaluation limit reached without convergence (9)



if I use the function lme of the library nlme

the message of R is:

# Error in lme.formula(Stages2 ~ fTemperatura * fSal * fFood * fEmbrion, :
nlminb problem, convergence error code = 1 message = iteration limit reached
without convergence (9)
I want to test if the relationship between the variable Y and fixed
factors is different in each Female. For this reason I need to include
in the random part all the fixed factors.

I think my main problem is that I don't know how can I include 3 factors
in a random intercept and slope model.

All my explanatory variables are factors I mean categoric variables, the
matrix has only 0 and
1. I am using a binomial.

If you know how can I write in R for one of this library the code, it
would be OK for me
thanks in advance.
Enrique
--
Gonzalez,Enrique

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models