similar to: glmmPQL and spatial correlation

Hi all, I am analyzing data on habitat utilization of seals in the Southern Ocean. My data show spatial autocorrelation, which I'm interested in incorporating into my model. I am trying to model the presence of dives (versus simulated pseudo-absences) using a binomial generalized binomial model (glmmPQL), since I can incorporate the autocorrelation structure to the model using that package.

Hi, I have a spatial data set with many observations (~50,000) and would like to keep as much data as possible. There is spatial dependence, so I am attempting a mixed model in R with a spherical variogram defining the correlation as a function of distance between points. I have tried nlme, lme, glmmML, and glmmPQL. In all case the matrix needed (seems to be (N^2)/2 - N) is too large for my

Greetings all The help file for GAMM in mgcv indicates that the log likelihood for a GAMM reported using summary(my.gamm$lme) (as an example) is not correct. However, in a past R-help post (included below), there is some indication that the likelihood ratio test in anova.lme(mygamm$lme, mygamm1$lme) is valid. How can I tell if anova.lme results are meaningful (are AIC, BIC, and logLik

Hi. I'm trying to perform what should be a reasonably basic analysis of some spatial presence/absence data but am somewhat overwhelmed by the options available and could do with a helpful pointer. My researches so far indicate that if my data were normal, I would simply use gls() (in nlme) and one of the various corSpatial functions (eg. corSpher() to be analagous to similar analysis in SAS)

I have spatial data on a sphere (the Earth) for which I would like to run an gls model assuming that the errors are autcorrelated, i.e. including a corSpatial correlation in the model specification. In this case the distance metric should be calculated on the sphere, therefore metric = "euclidean" in (for example) corSpher would be incorrect. I would be grateful for help on how to

Hello. As advised by Mick Crawley in his book on S+, I'm trying to use the lme function to examine a linear relationship between two variables measured at 60 locations in 12 sites, while taking account of any spatial autocorrelation (i.e. similarity in variation between the two variables that is due to site). I am using the function as follows:

I need to use the glmmPQL function for an assignment, but when I call for the summary of the function, it gives the AIC a value of NA. How do I get R to give me the AIC value? -- View this message in context: http://www.nabble.com/glmmPQL-help-tf1955675.html#a5363876 Sent from the R help forum at Nabble.com.

Dear all, I try to fit a glmmPQL on a huge data with 384189 individuals id=1:384189: working in 1520 establishments est:1:1516. The minimum number of individuals in every establishment is 30. This works for a subsample excluding establishemnet cells smaller than 100, but fail when we include smaller cells: R> summary(glmmPQL(count ~ + I( age-ave(age,est) )* ave(age,est) + + I(

> I'm trying to use "predict" with a linear mixed-effects logistic > regression model fitted with nlmmPQL from the MASS library. > Unfortunately, I'm getting an error "non-conformable arguments" in > predict.lme, and I would like to understand why. I'd like to briefly describe how I ended up working around this problem. The issue is that predict.lme

Hi, I want to do a logistic regression model with random effects but I need to sepcify my own design matrix for the random effects. Can I do it in glmmPQL or anything similar? If so, how? Thanks, Minya

Hi R-users,   I intend to apply a mixed model on a set of longitudinal data, with a negative binomial distributed dependent variable, and after following the discussions on R help list I saw that more experienced people recommended using lmer (from lme4 pack), glmmPQL (from MASS) or glmm.admb (from glmmADMB pack)     My first problem: yesterday this syntax was ok, now I get this weird message (I

Dear R users, Is there a way to compare glmmPQL models differing in their fixed-effects structure (similar to the ANOVA approach in lme) ? Thank you very much for your help! Chris. ---------------------------------------------------------------- This mail was sent through http://webmail.uni-jena.de

I'm a bit puzzled by how to write out additive random effects in glmmPQL. In my situation, I have a factorial design on two (categorical) random factors, A and B. At each combination, I have a binary response, y, and two binary fixed covariates, C and D. If everything were fixed, I would use glm(y ~ A + B + C + D, family = binomial) My first thought was to use glmmPQL(y ~ A + B, random

Hi all, I've been attempting to fit a logistic glmm using glmmPQL in order to estimate variance components for a score test, where the model is of the form logit(mu) = X*a+ Z1*b1 + Z2*b2. Z1 and Z2 are actually reduced rank square root matrices of the assumed covariance structure (up to a constant) of random effects c1 and c2, respectively, such that b1 ~ N(0,sig.1^2*I) and c1 ~

Hi ! I try to compare some models obtained from glmmPQL. model1 <- glmmPQL(y~red*yellow+I(red^2)+I(yellow^2)+densite8+I(densite8^2)+freq8_4 +I(freq8_4^2), random=~1|num, binomial); model2 <- glmmPQL(y~red*yellow+I(red^2)+I(yellow^2)+densite8+I(densite8^2)+freq8_4 , random=~1|num, binomial); anova(model1, model2) here is the answer : Erreur dans anova.lme(model1, model2) : Objects must

Hi, I'm trying to perform a power simulation for a simple multilevel model, using the function glmmPQL in R version 2.8.1. I want to extract the p-value for the fixed-effects portion of the regression, but I'm having trouble doing that. I can extract the coefficients (summary(fit)$coeff), and the covariance matrix (summary(fit)$varFix), but I can't grab the p-value (or the

Dear R-listers I wonder if someone can help me with the syntax for the random effect in glmmPQL()? I have a data set with a response variable "y" (counts), two dependent variables: "treat" (4 levels) and "site" (2 levels). The latter, I want to use as a random variable. How do I specify this in the function? Is it like this:

I asked a few days ago about the difference in results I saw between the MASS function glmmPQL (due to Venables and Ripley) and the lme function from the package nlme (due to Pinheiro and Bates). When the two tools apply to the same model (gaussian, link=identity, correlation=AR1), I was getting different results and wondered if there was an argument in favor of one or the other. Several

Greetings- Using R 1.5.0 under linux and the latest MASS and nlme, I am trying to develop a three-level (two levels of nesting) model with a dichotomous oucome variable. The unconditional model is thus: > doubt1.pql<-glmmPQL(fixed = r.info.doubt ~ 1, random = ~1 | groupid/participantid, + family = binomial, data = fgdata.10statements.df) iteration 1 iteration 2 iteration 3 iteration 4

Hello, Might anyone know how to conduct Wald-type F-tests of the fixed effects estimated by glmmPQL? I see this implemented in SAS (GLIMMIX), and have seen it recommended in user group discussions, but haven't come across any code to accomplish it. I understand the anova function treats a glmmPQL fit as an lme fit, with the test assumptions based on maximum likelihood, which is inappropriate