#Supone que todos los par?mentros son aletorios
tmp_fit.model.rpl<-function (model, design, design.nsim=1) {
#browser()
nalt<-length(unique(design$alt))
param<-summary(model)$CoefTable[,1]
nparam<-length(param)
medias<-param[1:(nparam/2)]
desvios<-param[(nparam/2+1):nparam]
chids<-unique(design$chid)
nchid<-length(unique(design$chid))
simul.betas<-matrix(ncol=nparam/2,nrow=nchid)
for (i in 1:nchid) {
simul.betas[i,]<-rnorm(nparam/2)*desvios+medias
}
# P es una matriz donde se guardan las distintas estimaciones luego de i iteraciones para analizar convergencia
p<-matrix(0,nrow=design.nsim,ncol=nalt)
p[1,]<- fit.rpl(simul.betas, design)
if (design.nsim>1){
for (i in 2:design.nsim) {
p[i,]<- (p[i-1,]*(i-1)+fit.rpl(simul.betas, design))/i
}
}
#names(p)<-rownames(design)
p[design.nsim,]
}
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.