pg_mvgp_univariate_mra: Bayesian Polya-gamma regression
In jtipton25/pgR: Bayesian Polya-Gamma Regression

pg_mvgp_univariate_mra

R Documentation

Bayesian Polya-gamma regression

Description

this function runs the Bayesian multinomial regression using Polya-gamma data augmentation

Usage

pg_mvgp_univariate_mra(
  Y,
  X,
  Z0,
  locs,
  params,
  priors,
  M = 4,
  n_coarse_grid = 10,
  model = "iid error",
  n_cores = 1L,
  inits = NULL,
  config = NULL,
  verbose = FALSE,
  use_spam = TRUE,
  n_chain = 1
)

Arguments

`Y`	is a n x J matrix of compositional count data.
`X`	is a n x p matrix of fixed effects (like latitude, elevation, etc)
`Z0`	is a n x Q matrix of observed climate variables
`locs`	is a n x 2 matrix of observation locations.
`params`	is a list of parameter settings. The list `params` must contain the following values: `n_adapt`: A positive integer number of adaptive MCMC iterations. `n_mcmc`: A positive integer number of total MCMC iterations post adaptation. `n_thin`: A positive integer number of MCMC iterations per saved sample. `n_message`: A positive integer number of frequency of iterations to output a progress message. For example, `n_message = 50` outputs progress messages every 50 iterations.
`priors`	is the list of prior settings.
`M`	The number of resolutions.
`n_coarse_grid`	The number of basis functions in one direction (e.g. `n_coarse_grid = 10` results in a 10x10 course grid which is further extended by the number of additional padding basis functions given by `n_padding`.
`model`	is the form of the polya-gamma model. Currently, this option is not active the only model is the "iid error" model. This option allows for independent species-specific overdispersion variance terms.
`n_cores`	is the number of cores for parallel computation using openMP.
`inits`	is the list of initial values if the user wishes to specify initial values. If these values are not specified, then the initial values will be randomly sampled from the prior.
`config`	is the list of configuration values if the user wishes to specify initial values. If these values are not specified, then default a configuration will be used.
`verbose`	is a logicial input that determines whether to print more detailed messages.
`use_spam`	is a boolean flag to determine whether the output is a list of spam matrix objects (`use_spam = TRUE`) or a an n x n sparse Matrix of class "dgCMatrix" `use_spam = FALSE` (see spam and Matrix packages for details).
`n_chain`	is the MCMC chain id. The default is 1.