vbglmss.mixedSS: Variational approximation of Spike and Slab regression

In antiphon/vbss: Spike and Slab generalized linear regression (GLM) using variational approximation

Description

The function fits a regression model where observations can be from logistic and/or gaussian likelihood. The coefficients are shared, and can have Spike and Slab (SS) and/or Gaussian (G) priors.

Usage

  vbglmss.mixedSS(y, Xy, Zy, h, Xh, Zh,
    prior = list(theta = list(), beta = list(), binary = list(), gaussian = list()),
    verbose = FALSE, damp = 0, eps = 0.01, max.iter = 100)

Arguments

y	Binary observations
Xy	SS covariates for y
Zy	G covariates for y
h	Gaussian observations
Xh	SS covariates for h
Zh	G covariates for h
prior	A list of priors. See details.
verbose	print additional runtime messages
damp	Damping coefficient in SS mean update. damp * old + (1-damp)*new.
eps	the approximation loop convergence threshold
max.iter	– TODO

Value

A list with posteriors.

antiphon/vbss documentation built on May 10, 2019, 12:22 p.m.