Learn R Programming
SDGLM (version 0.4.0)

sca_mcmc1: Alternative Sca-MCMC Implementation for Variable Selection

Description

An alternative implementation of Sca-MCMC for variable selection with binary inclusion indicators.

Usage

sca_mcmc1(
  y,
  X,
  family = c("poisson", "pareto", "gamma"),
  method = c("ScaI", "ScaII", "ScaIII"),
  N_chain = 8,
  n_iter = 5000,
  beta_star = NULL,
  alpha_gamma = 2
)

Value

List containing:

beta_chain

Array [n_iter x p x N_chain] of sampled inclusion vectors.

family

Model family used.

method

Mutation strategy used.

Arguments

y

Numeric response vector.

X

Design matrix (n x p).

family

Character, one of "poisson", "pareto", "gamma".

method

Mutation strategy: "ScaI", "ScaII", or "ScaIII".

N_chain

Number of parallel tempered chains (>1).

n_iter

Total MCMC iterations.

beta_star

Target (true) inclusion vector (for adaptive Q0; optional).

alpha_gamma

Shape parameter if family = "gamma" (default 2).