R/Bayesiangammareg.R

In Bayesiangammareg: Bayesian Gamma Regression: Joint Mean and Shape Modeling

Bayesiangammareg <-
function(Y,X,Z,nsim,bpri,Bpri,gpri,Gpri,burn,jump,graph1,graph2,meanlink="log"){
  
  if(meanlink=="log"){est<-GammaLog(Y,X,Z,nsim,bpri,Bpri,gpri,Gpri,burn,jump,graph1,graph2)} 
  else{
  if(meanlink=="ide"){est<-GammaIdentity(Y,X,Z,nsim,bpri,Bpri,gpri,Gpri,burn,jump,graph1,graph2)}
  else{stop("No link avali")}
  }
  
  
  est$coefficients <- matrix(c(est$Bestimado,est$Gammaest))
  names <- c(colnames(est$X),colnames(est$Z))
  par <-rep(c("beta.","gamma."),c(ncol(est$X),ncol(est$Z)))
  rownames(est$coefficients) <-paste(par,names,sep="")
  
  est$desv <-  matrix(c(est$DesvBeta, est$DesvGamma))
  est$interv<- rbind(est$B,est$G)
  est$fitted.values <- est$yestimado
  est$residuals <- est$residuals
  est$precision <- est$phi
  est$variance <- est$variance
  est$beta.mcmc<-est$beta.mcmc
  est$gamma.mcmc<-est$gamma.mcmc
  est$beta.mcmc.short<-est$beta.mcmc.auto
  est$gamma.mcmc.short<-est$gamma.mcmc.auto
  est$Y<-est$Y
  est$X<-est$X  
  est$Z<-est$Z 
  
  est$call <- match.call()
  
  class(est) <- "Bayesiangammareg"
  
  return(est) 
}