bmrarm: MCMC sampler to implement a bmrarm model

In nickseedorff/bmrarm: Bayesian Mixed Response Autoregressive Models

MCMC sampler to implement a bmrarm model

Description

MCMC sampler to implement a bmrarm model

Usage

bmrarm(
  formula,
  data,
  ordinal_outcome,
  time_var,
  patient_var,
  random_slope = T,
  ar_cov = TRUE,
  nsim = 1000,
  burn_in = 100,
  thin = 10,
  seed = 14,
  sig_prior = 1e+05,
  sd_vec = c(0.15, 0.3, rep(0.2, 4)),
  N_burn_trunc = 10,
  prior_siw_uni = c(0.2, 5),
  prior_siw_df = NULL,
  prior_siw_scale_mat = NULL
)

Arguments

formula	an object of class "formula"; a symbolic description of the model to be fitted
data	a dataframe containing outcome variables, covariates, and a patient or subject identifier
ordinal_outcome	a character string, which contains the variable name for the ordinal outcome
time_var	a character string, which contains the variable name for time indexing; this must be integer valued and should indicate the observation number.
patient_var	a character string, which contains the variable name for patient indexing
ar_cov	logical, indicating use of an autoregressive error term. The default is TRUE
nsim	positive integer, number of iterations with default of 1000
burn_in	positive integer, number of iterations to remove with default of 100
thin	positive integer, specifies the period of saving samples. Default of 10
seed	positive integer, seed for random number generation
sig_prior	scalar, variance term for prior on the beta coefficients
sd_vec	numeric vector, containing standard deviations for Metroplis Hastings proposals, either of length 4 or 6. The values are associated with the MH-within-gibbs step, the ar term, and the random effects
N_burn_trunc	integer, number of burn-in draws from the truncated multivariate normal Gibbs sampler
prior_siw_uni	prior bounds for the uniform distribution associated with the SIW prior
prior_siw_df	degrees of freedom for the SIW prior. The default is 1 + number of random effects for a single person (2 or 4).
prior_siw_scale_mat	scale matrix for the SIW prior. The default is an identity matrix
random_Slope	logical, indicating use of random slopes. The default is TRUE

Value

bmrarm

nickseedorff/bmrarm documentation built on April 17, 2025, 9:43 p.m.