Resampling Transformation for the Markov Chain Monte Carlo Estimation Simulation of the Bayesian Hierarchical Model for Identifying the Most Harmful Medication Errors

Share:

Description

This function implements the transformation needed to apply the importance link function resampling methodology based on the Markov Chain Monte Carlo simulations obtained with the bhm.mcmc command (see the References).

Usage

1
bhm.resample(model, dat, p.resample = 0.1, k, eta)

Arguments

model

an object of class "mederrFit".

dat

an object of class "mederrData".

p.resample

proportion of simulations resampled from the model argument.

k

required vector of k values to be used in the resampling process.

eta

required vector of η values to be used in the resampling process.

Value

bhm.resample returns an object of the class "mederrResample".

Author(s)

Sergio Venturini sergio.venturini@unibocconi.it,

Jessica A. Myers jmyers6@partners.org

References

MacEachern, S. and Peruggia, M. (2000), "Importance Link Function Estimation for Markov Chain Monte Carlo Methods", Journal of Computational and Graphical Statistics, 9, 99-121.

Myers, J. A., Venturini, S., Dominici, F. and Morlock, L. (2011), "Random Effects Models for Identifying the Most Harmful Medication Errors in a Large, Voluntary Reporting Database". Technical Report.

See Also

mederrData, mederrFit, bhm.mcmc.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
## Not run: 
data("simdata", package = "mederrRank")
summary(simdata)

fit <- bhm.mcmc(simdata, nsim = 1000, burnin = 500, scale.factor = 1.1)
resamp <- bhm.resample(fit, simdata, p.resample = .1,
	k = c(3, 6, 10, 30, 60, Inf), eta = c(.5, .8, 1, 1.25, 2))
fit2 <- bhm.constr.resamp(fit, resamp, k = 3, eta = .8)
plot(fit, fit2, simdata)

theta0 <- c(10, 6, 100, 100, .1)
ans <- mixnegbinom.em(simdata, theta0, 50000, 0.01, se = TRUE,
	stratified = TRUE)

summary(fit2, ans, simdata)

## End(Not run)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.