Apply a Bayesian [zero-inflated] gamma /
Weibull / lognormal / independant / simple Poisson model to count data
to return possible values for mean count, coefficient of variation, and
zero-inflation, as either summary statistics or mcmc objects.
Convergence is assessed for each dataset by calculating the Gelman-Rubin
statistic for each parameter, see autorun.jags.
Optionally, the log likelihood for the model fit is also calculated.
The time taken to complete each analysis (not including calculation of
the likelihood) is also recorded. The lower level functions in the
runjags package are used for calling JAGS.
Note: The GUI interface for R in Windows may not continually refresh the
output window, making it difficult to track the progress of the
simulation (if silent.jags is FALSE). To avoid this, you can run the
function from the terminal version of R (located in the Program
Files/R/bin/ folder).
Usage
count.analysis(data = stop("Data must be specified"), model="ZILP",
alt.prior = FALSE, adjust.zi.mean = FALSE, raw.output = FALSE,
likelihood=FALSE, ...)
Arguments
Value
Either a vector similar to that obtained from
bayescount containing an indication of the
error/crash/convergence status, the number of sampled updates used, and
a lower/upper 95% highest posterior density interval (see
HPDinterval), and median estimate for each relevant
parameter (optionally including the likelihood), or an MCMC object
representing the estimates at each iteration for both chains (optionally
including the likelihood).
# use a zero-inflated lognormal Poisson model to analyse some count# data, and suppressing JAGS output:# results <- bayescount.single(data=c(0,5,3,7,0,4,3,8,0),
model="ZILP", silent.jags=TRUE)