fit_bayes_cjs: Fit Bayesian CJS Model

In mackerman44/telemetyr: Tools for processing, analysis, and visualization of telemetry data

Fit Bayesian CJS Model

Description

This wrapper function writes a text file describing a Bayesian Cormack Jolly-Seber model in the JAGS language, prepares the data, runs the MCMC and summarises the results.

Usage

fit_bayes_cjs(
  file_path = NULL,
  cap_hist_wide = NULL,
  tag_meta = NULL,
  drop_col_nm = "duty_cycle",
  drop_values = c("batch_2", "batch_3"),
  n_chains = 4,
  n_adapt = 1000,
  n_burnin = 2500,
  n_iter = 2500,
  n_thin = 5,
  params_to_save = c("phi", "p", "survship"),
  rng_seed = 4,
  ...
)

Arguments

file_path	name (with file path) to save the model as. Probably a .txt file
cap_hist_wide	wide capture history, one row per tag, one column per detection site. Includes columns named tag_id and cap_hist. Part of the output from prep_capture_history.
tag_meta	metadata for each tag, including a column named tag_id.
drop_col_nm	name of column in tag_meta to be used to filter out any particular tags.
drop_values	character vector of values in drop_col_nm column of tag_meta that should be excluded from the CJS model.
n_chains	the number of parallel chains for the model
n_adapt	the number of iterations for adaptation. If n_adapt = 0 then no adaptation takes place.
n_burnin	the number of iterations of the Markov chain to run during the burn-in phase
n_iter	the number of iterations to monitor
n_thin	the thinning interval for monitors
params_to_save	a character vector giving the names of variables to be monitored
rng_seed	random number generator seed, to make results reproducible
...	other parameters to be fed to rjags::jags.model, update.jags or rjags::coda.samples

Value

tibble

Author(s)

Kevin See and Mike Ackerman

mackerman44/telemetyr documentation built on Feb. 15, 2025, 1:08 a.m.