inst/exampleZIHR.R

In UHM: Unified Zero-Inflated Hurdle Regression Models

# Example 1
data(dataD)
index <- 1:(dim(dataD)[1])
IND_new <- sample(index, .5 * length(index))
datat <- dataD[IND_new, ]
datav <- dataD[-IND_new, ]
modelY <- y~x1 + x2
modelZ <- z~x1
D1 <- ZIHR(modelY, modelZ,
           data = datat, n.chains = 2, n.iter = 1000,
           n.burnin = 500, n.thin = 1, family = "Poisson"
)

\donttest{
  SummaryZIHR(D1)
  Prediction(D1, data = datav)


  D2 <- ZIHR(modelY, modelZ,
             data = datat, n.chains = 2, n.iter = 1000,
             n.burnin = 500, n.thin = 1, family = "Bell"
  )
  SummaryZIHR(D2)



  # Example 2
  data(dataC)
  modelY <- y~x1 + x2
  modelZ <- z~x1
  C <- ZIHR(modelY, modelZ,
            data = dataC, n.chains = 2, n.iter = 1000,
            n.burnin = 500, n.thin = 1, family = "Gaussian"
  )
  SummaryZIHR(C)

  Prediction(C, data = datav)



  # Example 3
  data(dataP)
  modelY <- y~x1 + x2
  modelZ <- z~x1
  P1 <- ZIHR(modelY, modelZ,
             data = dataP, n.chains = 2, n.iter = 1000,
             n.burnin = 500, n.thin = 1, family = "Exponential"
  )
  SummaryZIHR(P1)

  P2 <- ZIHR(modelY, modelZ,
             data = dataP, n.chains = 2, n.iter = 1000,
             n.burnin = 500, n.thin = 1, family = "Gamma"
  )
  SummaryZIHR(P2)

  P3 <- ZIHR(modelY, modelZ,
             data = dataP, n.chains = 2, n.iter = 1000,
             n.burnin = 500, n.thin = 1, family = "Weibull"
  )
  SummaryZIHR(P3)


  # Example B
  data(dataB)
  modelY <- y~x1 + x2
  modelZ <- z~x1
  P <- ZIHR(modelY, modelZ,
            data = dataB, n.chains = 2, n.iter = 1000,
            n.burnin = 500, n.thin = 1, family = "Beta"
  )
  SummaryZIHR(P)

  # Example C
  data(dataI)
  modelY <- y~x1 + x2
  modelZ <- z~x1
  P4 <- ZIHR(modelY, modelZ,
             data = dataI, n.chains = 2, n.iter = 1000,
             n.burnin = 500, n.thin = 1, family = "inverse.gaussian"
  )
  SummaryZIHR(P4)
}