negLASSO: Normal-exponential-gamma Bayesian LASSO

In abnormally-distributed/Bayezilla: Bayesian Models With JAGS and Several Utilities for Data Exploration and Model Evaluation.

Description

This implements the normal-exponential-gamma "hyperlasso" of Griffin & Brown (2011). This model has independent normal priors on each coefficient, whose precision is modeled by independent, predictor specific, exponential distributions. The exponential distributions in turn have their respective rate parameters modeled through independent gamma(.5, 1 / lambda^2) distributions. Lambda is a single top-level hyperparameter here given a gamma(0.50 , 0.20) prior.

The model specification is given below:



Model Specification:


The normal-exponential-gamma (NEG) lasso is very similar to the adaptive Bayesian Lasso (adaLASSO), which also makes use of a normal-exponential-gamma hierarchy, except that it is parameterized slightly differently.

Usage

negLASSO(formula, data, family = "gaussian", lambda.prior = "dmouch",
  dof = 0.5, log_lik = FALSE, iter = 10000, warmup = 1000,
  adapt = 2000, chains = 4, thin = 1, method = "rjparallel",
  cl = makeCluster(3), ...)

Arguments

formula	the model formula
data	a data frame.
family	one of "gaussian", "binomial", or "poisson".
lambda.prior	either "dmouch" (the default) or "gamma"
dof	the degrees of freedom for the normal-exponential-gamma prior. Defaults to 0.50.
log_lik	Should the log likelihood be monitored? The default is FALSE.
iter	How many post-warmup samples? Defaults to 10000.
warmup	How many warmup samples? Defaults to 1000.
adapt	How many adaptation steps? Defaults to 2000.
chains	How many chains? Defaults to 4.
thin	Thinning interval. Defaults to 1.
method	Defaults to "rjparallel". For another parallel option, choose "parallel" or "rjags" for a single core run.
cl	Use parallel::makeCluster(# clusters) to specify clusters for the parallel methods. Defaults to 3.
...	Other arguments to run.jags.

Value

A run.jags object

References

Griffin, J. E. and Brown, P. J. (2011), Bayesian Hyper‐LASSOs With Non-Convex Penalization. Australian & New Zealand Journal of Statistics, 53: 423-442. doi:10.1111/j.1467-842X.2011.00641.x

Examples

negLASSO()

abnormally-distributed/Bayezilla documentation built on Oct. 31, 2019, 1:57 a.m.