R: MCMC Probability Update Function for Hurdle Model Count Data...

update_probs {hurdlr}

R Documentation

MCMC Probability Update Function for Hurdle Model Count Data Regression

Description

MCMC algorithm for updating the likelihood probabilities in hurdle model regression using hurdle.

Usage

update_probs(y, x, hurd, p, q, beta.prior.mean, beta.prior.sd, pZ, pT, pE, beta,
  XB2, XB3, beta.acc, beta.tune)

Arguments

`y`	numeric response vector.
`x`	optional numeric predictor matrix.
`hurd`	numeric threshold for 'extreme' observations of two-hurdle models.
`p`	numeric vector of current 'p' probability parameter values for zero-value observations.
`q`	numeric vector of current 'q' probability parameter values for 'extreme' observations.
`beta.prior.mean`	mu parameter for normal prior distributions.
`beta.prior.sd`	standard deviation for normal prior distributions.
`pZ`	numeric vector of current 'zero probability' likelihood values.
`pT`	numeric vector of current 'typical probability' likelihood values.
`pE`	numeric vector of current 'extreme probability' likelihood values.
`beta`	numeric matrix of current regression coefficient parameter values.
`XB2`	`x*beta[,2]` product matrix.
`XB3`	`x*beta[,3]` product matrix.
`beta.acc`	numeric matrix of current MCMC acceptance rates for regression coefficient parameters.
`beta.tune`	numeric matrix of current MCMC tuning values for regression coefficient estimation.

Value

A list of MCMC-updated regression coefficients for the estimation of the parameters 'p' (the probability of a zero-value observation) and 'q' (the probability of an 'extreme' observation) as well as each coefficient's MCMC acceptance ratio.

Author(s)