mm.post.var: Compute the posterior variance across Q difference proposed...

In annaheath/EVSI: EVSI: Calculating the Expected Value of Sample Information

Compute the posterior variance across Q difference proposed data collection exercises This can be used to estimate the EVSI using the Heath et al. method.

Description

Compute the posterior variance across Q difference proposed data collection exercises This can be used to estimate the EVSI using the Heath et al. method.

Usage

mm.post.var(model.stats, data, N.name = NULL, N.size = NULL,
  effects = NULL, costs = NULL, he = NULL, evi = NULL,
  parameters = NULL, Q = 30, data.stats = NULL, update = c("bugs",
  "jags"), n.burnin = 1000, n.thin = 1, n.iter = 5000)

Arguments

model.stats	A .txt file containing the model file of a Bayesian model. This should be a BUGS or JAGS model.
data	A string or vector of strings that defines the name of the future data in the model file. If the future data has already been generated then the data arguement can be given as a list. The elements of the data list should be data lists for JAGS or BUGS.
N.name	A string that defines the name of the variable that defines the sample size of the future data in the model file. If NULL then the sample size of the data must be defined in data.stats.
N.size	A scalar or vector that defines the sample sizes for the future study. A scalar N.size will compute the EVSI for a single sample size. If a vector is passed for N.size then the EVSI can be estimated for all sample sizes within the range of values given. If NULL then the sample size of the data must be defined in data.stats.
effects	This can either be given as a string which defines the name of the effectivness measure in the BUGS/JAGS model. Or it can be given as a function that takes the output of the model and calculates the effectiveness measure. The inputs of this function should be parameter names found in the model file.
costs	This has a similar format to e but gives the value of the costs in the model file or a function defining the costs.
he	A bcea object containing the base case analysis for the model.
evi	An evppi object that contains the EVPPI analysis for the parameters that are being informed by the sample information.
parameters	A list of the names of the parameters that are being targetted by the study. This can be given OR an evppi object needs to be given.
Q	The number of quadrature points used the estimate the EVSI.
data.stats	A data file for the BUGS/JAGS model. This is the data used to inform the base case analysis. If empty then it is assumed that the models are
update	Defines the Bayesian engine that should be used to update the the Bayesian model file given in model.stats.
n.burnin	The burnin for the JAGS/BUGS model
n.thin	The thinning for the JAGS/BUGS model
n.iter	The number of interations for the JAGS/BUGS model

Value

An var.mm object. 1. evsi An array containing the EVSI by wtp, N and across different uncertaincies 2. attrib A list of wtp, N and prob describing the attributes of the evsi matrix. 3. evppi An evppi object containing all the information about the calculation of EVPPI. 4. he A bcea object containing all the information about the underlying health economic model

Examples

NULL

annaheath/EVSI documentation built on June 25, 2022, 6:26 a.m.