Truncated Power Prior Distribution for Models

Share:

Description

Creates an object representing the prior distribution on models for BAS using a truncated Distribution on the Model Size where the probability of gamma = p^-kappa |gamma| where gamma is the vecotr of model indicators

Usage

1

Arguments

kappa

parameter in the prior distribution that controls sparsity

trunc

parameter that determines truncation in the distribution i.e. P(gamma; alpha, beta, trunc) = 0 if |gamma| > trunc.

Details

The beta-binomial distribution on model size is obtained by assigning each variable inclusion indicator independent Bernoulli distributions with probability w, and then giving w a beta(alpha,beta) distribution. Marginalizing over w leads to the distribution on the number of included predictos having a beta-binomial distribution. The default hyperparameters lead to a uniform distribution over model size. The Truncated version assigns zero probability to all models of size > trunc.

Value

returns an object of class "prior", with the family and hyerparameters.

Author(s)

Merlise Clyde

See Also

bas.lm, Bernoulli,uniform

Examples

1
2
3
4
5
6
7
tr.power.prior(2, 8)
library(MASS)
data(UScrime)
UScrime[,-2] = log(UScrime[,-2])
crime.bic =  bas.lm(y ~ ., data=UScrime, n.models=2^15, prior="BIC",
                    modelprior=tr.power.prior(2,8),
                    initprobs= "eplogp")

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.