runSTADEMmodel: Run JAGS model
In KevinSee/STADEM: Estimate escapement past a dam with a window and trap facility

runSTADEMmodel

R Documentation

Run JAGS model

Description

Run the JAGS model

Usage

runSTADEMmodel(
  file_name = NULL,
  mcmc_chainLength = 100,
  mcmc_burn = 10,
  mcmc_thin = 2,
  mcmc_chains = 1,
  jags_data = NULL,
  weekly_params = FALSE,
  seed = NULL,
  verbose = FALSE,
  parallel = TRUE,
  DIC = FALSE,
  win_model = c("neg_bin", "neg_bin2", "pois", "quasi_pois", "log_space"),
  trap_est = TRUE,
  use_jagsUI = FALSE
)

Arguments

`file_name`	name (with file path) to save the model as
`mcmc_burn`	number of burn-in samples in each MCMC chain
`mcmc_thin`	thinning interval for MCMC samples to save
`mcmc_chains`	number of MCMC chains
`jags_data`	list of data being passed to `JAGS` model, created by `prepJAGS`
`weekly_params`	should weekly estimates be saved? Default is `FALSE`
`seed`	starting seed value for RNG, to make results reproducible
`verbose`	passed to the `jags` function
`parallel`	passed to the `jags` function
`DIC`	passed to the `jags` function
`win_model`	what type of distribution should be used when modeling the window counts. `neg_bin` is a standard negative binomial distribution. `neg_bin2` is a more flexible version of a negative binomial, allowing the mean-variance relationship to take different forms. `pois` is a Poisson distribution. `quasi_pois` is the quasi-Poisson distribution. `log_space` assumes that the window counts have normal errors in log-space.
`trap_est`	should an estimate of escapement based on the adult fish trap rate be used as a second observation of true total escapement, together with the window counts? Default is `TRUE`. If `FALSE`, `win_model` is automatically set to `pois`.
`use_jagsUI`	should the 'jagsUI' package be used to fit the model? Default is 'FALSE', meaning the 'rjags' package is used instead.
`mcmc_chainlength`	number of total samples in each MCMC chain