skipTrack.MCMC {skipTrack}

Perform one chain of MCMC sampling for the skipTrack model.

Description

This function runs a single Markov Chain Monte Carlo (MCMC) chain to update parameters in the skipTrack hierarchical model.

Usage

skipTrack.MCMC(
  Y,
  cluster,
  X = matrix(1, nrow = length(cluster)),
  Z = matrix(1, nrow = length(cluster)),
  numSkips = 10,
  reps = 1000,
  fixedSkips = FALSE,
  initialParams = list(pi = rep(1/(numSkips + 1), numSkips + 1), muis = rep(log(30),
    length(unique(cluster))), tauis = rep(5, length(unique(cluster))), rho = 1, cijs =
    sample(1:3, length(Y), replace = TRUE), alphas = rep(1, numSkips + 1), Beta =
    matrix(rep(0, ncol(as.matrix(X))), 1), Gamma = matrix(rep(0, ncol(as.matrix(Z))), 1),
    rhoBeta = 0.01, rhoGamma = 1000, phi = 0.01, rhoPhi = 1000),
  verbose = FALSE
)

Arguments

Y

A vector of observed cycle lengths.

cluster

A vector indicating the individual cluster/group membership for each observation Y.

X

A matrix (length(Y) x length(Beta)) of covariates for cycle length mean. Default is a vector of 1's.

Z

A matrix (length(Y) x length(Gamma)) of covariates for cycle length precision. Default is a vector of 1's.

numSkips

The maximum number of skips to allow. Default is 10.

reps

The number of MCMC iterations (steps) to perform. Default is 1000.

fixedSkips

If TRUE cycle skip information (cijs) is not updated in sample steps and the inputs are instead assumed to be true.

initialParams

A list of initial parameter values for the MCMC algorithm. Default values are provided for pi, muis, tauis, rho, cijs, alphas, Beta, Gamma, phi, rhoBeta, rhoGamma, and rhoPhi.

verbose

logical. If true progress bars and additional info are printed to the console.

Value

A list containing the MCMC draws for each parameter at each iteration. Each element in the list is itself a list containing:

ijDat

A data.frame with updated parameters at the individual-observation level: Individual, ys, cijs, muis, tauis.

iDat

A data.frame with updated parameters at the individual level: Individual, mus, taus, thetas.

rho

Updated value of the global parameter rho.

pi

Updated value of the global parameter pi.

Xi

Matrix of covariates for cycle length mean.

Zi

Matrix of covariates for cycle length precision.

Beta

Updated matrix of coefficients for cycle length mean.

Gamma

Updated matrix of coefficients for cycle length precision.

priorAlphas

Vector of prior alpha values for updating pi.

indFirst

A logical vector indicating the first occurrence of each individual.

rhoBeta

Hyperprior parameter rhoBeta, used to update Beta.

rhoGamma

Value of the proposal parameter rhoGamma.

phi

Updated value of the parameter phi.

rhoPhi

Value of the proposal parameter rhoPhi.

fixedSkips

Logical. Indicates if skips were fixed.

See Also

sampleStep