Description Usage Arguments Examples
Function glmm
estimates GLMM using methods basec on state space modelling.
1 2 3 4 5 |
response.var |
Name of the response variable in data. |
grouping.var |
Name of the grouping variable in data. Only one grouping variable is allowed. |
data |
Optional data frame environment containing the variables in the model. |
distribution |
Character vector defining the distributions used for each group. Either length of one (same distribution for all groups) or length of p, number of groups. Default is "gaussian". |
init.random |
Initial values for random effect covariances. |
init.dispersion |
Initial values for dispersion paremeters. |
init.fixed |
Initial values for fixed effects. |
correlating.effects |
Logical. Default is TRUE. |
estimate.dispersion |
Logical. Is the dispersion parameter estimated or fixed. |
return.model |
Logical. Is the estimated model returned in output. |
nsim |
Integer. Number of independent samples used in importance sampling. Default is 0, which corresponds to Laplace approximation. |
maxiter |
Integer. Number of iterations for in iterative weighted least squares. |
method |
"REML", "REML-ML" or "ML". "REML-ML" uses estimates obtained from REML for ML estimation. Note that ML is much slower. |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 | data(butterfly)
require(lme4)
system.time(fit.SSglmm.reml<-glmm(response.var="Colias", grouping.var="site", fixed=~habitat + building + urbanveg,random=~1,
REML=TRUE,data=butterfly,distribution="poisson",return.model=TRUE))
system.time(fit.SSglmm.ml<-glmm(response.var="Colias", grouping.var="site", fixed=~habitat + building + urbanveg,random=~1,
REML=FALSE,data=butterfly,distribution="poisson",return.model=TRUE))
fit.glmer<-glmer(Colias ~ habitat + building + urbanveg + (1|site),
family=poisson, data=butterfly,control=glmerControl(optimizer = "bobyqa"))
sqrt(fit.SSglmm.reml$random$P)
sqrt(fit.SSglmm.ml$random$P)
VarCorr(fit.glmer)
fixef(fit.glmer)
fit.SSglmm.ml$fixed$coef
fit.SSglmm.reml$fixed$coef
require(mvabund)
data(spider)
x<-apply(spider$x,2,rep,times=12)
dataspider<-data.frame(species=rep(names(spider$abund),each=28),abund=unlist(spider$abund),x)
out<-glmm(response.var="abund", grouping.var="species", fixed=~soil.dry+bare.sand, random=~1,
data=dataspider,distribution="poisson")
out2<-glmm(response.var="abund", grouping.var="species", fixed=~soil.dry+bare.sand,
data=dataspider,distribution="poisson")
fit<-manyglm(mvabund(spider$abund)~soil.dry+bare.sand,data=data.frame(spider$x),family="poisson")
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.