Usage
fecr_mcmc(fec.pre, fec.post, rawCounts = FALSE, f.pre = 50,
f.post = f.pre, model = "paired", priors.mu = list(hyperpars = c(1,
0.001), proposalDist = "kl"), priors.phi = list(hyperpars = c(1, 0.1),
proposalDist = "unif", v = 0.5), priors.delta = list(priorDist = c("gamma",
"beta")[1], hyperpars = c(1, 1), proposalDist = NULL),
priors.psiB = list(hyperpars = c(1, 1)), priors.psiA = list(hyperpars =
c(1, 1)), priors.deltaPsi = list(hyperpars = c(1, 1), proposalDist =
"beta"), maxiter.pilot = 15, nburnin = 5000, nsamples = 10000,
thin = 1, initials = NULL, verbose = TRUE, .verboselevel = 0, ...)Arguments
fec.pre
vector with faecal egg counts before treatment
fec.post
vector with faecal egg counts after treatment
rawCounts
logical indicating whether fec.pre and fec.post corresponds to raw counts
(as counted on the McMaster slide), or to calculated EpGs (raw counts times correction factor).
Defaults to FALSE.
f.pre
correction factor(s) before treatment
f.post
correction factor(s) after treatment
model
string with model formulation ("paired" or "unpaired")
priors.mu
list with hyper-prior/proposal information for $\mu$
priors.phi
list with hyper-prior/proposal information for $\phi$
priors.delta
list with hyper-prior/proposal information for $\delta$
priors.psiA
list with hyper-prior/proposal information for $\psi_A$
priors.psiB
list with hyper-prior/proposal information for $\psi_B$
priors.deltaPsi
list with hyper-prior/proposal information for $\delta_\psi$
maxiter.pilot
maximal number of tries to determine a good tuning value
for the proposal distribution for $\phi$
nburnin
number of burn-in iterations
nsamples
number of desired samples
initials
named list with starting values for the parameters
mu, phi, muiPre, muiPost, yPre, yPost,
delta, psiB and psiA
verbose
print progress information
.verboselevel
print additional information, mainly for
debugging information, larger values print more details
...
extra arguments (not used atm)