DPpackage (version 1.1-7.4)

PTrandom: Extracts Random Effects

Description

Extracts random effects from PTglmm objects: PTlmm, PTolmm, and PTglmm.

Usage

PTrandom(object,centered=FALSE,predictive=FALSE,ngrid=1000,gridl=NULL)

Arguments

object

PT fitted model object from which random effects estimates can be extracted.

centered

logical variable indicating whether the random effects should be extracted centered, bi, or uncentered thetai. This option cannot be only used to get the density estimates.

predictive

logical variable indicating whether actual or predictive information of the random effects should be extracted.

ngrid

number of grid points where the density estimate is evaluated. This is only used if dimension of the random effects is lower or equal than 2 and if predictive=TRUE. The default value is 1000.

gridl

The limits of the interval or rectangle covered by the grid as c(xl,xu) or c(xl, xu, yl, yu), respectively. If not specified the grid is defined automatically. This is only used if dimension of the random effects is lower or equal than 2 and if predictive=TRUE.

Examples

# NOT RUN {
    # School Girls Data Example
      data(schoolgirls)
      attach(schoolgirls)
 
    # Prior information
      prior<-list(alpha=1,
                  M=4,
                  tau1=0.01,tau2=0.01,
                  nu0=4.01,
                  tinv=diag(10,2),
                  mub=rep(0,2),
                  Sb=diag(1000,2))
 
    # Initial state
      state <- NULL
 
    # MCMC parameters
 
      nburn<-5000
      nsave<-5000
      nskip<-0
      ndisplay<-100
      mcmc <- list(nburn=nburn,nsave=nsave,nskip=nskip,ndisplay=ndisplay,
                    tune1=1.5,tune2=1.1)
 
    # Fitting the model
     
      fit1<-PTlmm(fixed=height~1,random=~age|child,prior=prior,mcmc=mcmc,
                  state=state,status=TRUE)
      fit1
 
 
    # Extract random effects
    
      PTrandom(fit1)
      
      plot(PTrandom(fit1))

    # Extract predictive information of random effects
    
      PTrandom(fit1,predictive=TRUE)
      plot(PTrandom(fit1,predictive=TRUE,gridl=c(75,89,3.8,7.5)))
# }