predictive: Predictive distributions

In alessandromagrini/Medbetareg: Generalized Beta regression to elicit conditional distributions of medical variables

Description

Plot of a predictive distribution chosen by the user.

Usage

predictive(x, cnfg, nrep = 500, title = NULL)

Arguments

x	An object of class mbr.
cnfg	A numerical vector representing a configuration of explanatory variables (original scale).
nrep	An integer positive number representing the number of Monte Carlo replications. Default is 500.
title	The title of the plot. If NULL (the default), the configuration of explanatory variables is shown as a title.

See Also

newPrior

Examples

assess.test <- 'RESP RespRate (0,15,25,40);
  CEV intraShunt (0,2,5,100) 0.5 hp 0.5 hp 5;
  CEV deadSpace (0,0,30,100) 0.5 hp 0.5 hp 5;
  CEV extraShunt (0,0,5,100) 0.5 hp 0.5 hp 5;
  CEV redAlvSpace (0,0,5,100) 0.5 hp 0.5 hp 5;
  BEV Panic 0.25 hp 25;
  BEV Neuromusc 0.6 lp 100;
  INTER intraShunt deadSpace 0.9 hp 5;
  TAU 0.3 n'
###  NOT RUN: replicate the results in Magrini et al. (2018)
# set.seed(10)
# prior.test <- newPrior(assess.test, nrep=5000)
# set.seed(10)
# predictive(prior.test, xcfg1, nrep=50000, title="Configuration 1")
# set.seed(10)
# predictive(prior.test, xcfg2, nrep=50000, title="Configuration 2")
# set.seed(10)
# predictive(prior.test, xcfg3, nrep=50000, title="Configuration 3")
# set.seed(10)
# predictive(prior.test, xcfg4, nrep=50000, title="Configuration 4")
#############
prior.test <- newPrior(assess.test, nrep=100)
predictive(prior.test, c(5,30,5,5,0,0), nrep=500)

alessandromagrini/Medbetareg documentation built on March 7, 2021, 4:54 p.m.