bits: Bayesian Iterated Screening (ultra-high, high or low...

Description Usage Arguments Value References

View source: R/bits.r

Description

Perform Bayesian iterated screening in Gaussian regression models

Usage

1
bits(X, y, lam = 1, w = NULL, pp = FALSE, max.var = nrow(X))

Arguments

X

An n\times p matrix. Sparse matrices are supported and every care is taken not to make copies of this (typically) giant matrix. No need to center or scale.

y

The response vector of length n.

lam

The slab precision parameter. Default: n/p^2.

w

The prior inclusion probability of each variable. Default: sqrt(n)/p as suggested by the theory of Wang et al. (2019).

pp

Booliean: If FALSE (default) the algorithm stops after including max.var many variables. If true, the posterior probability stopping rule is used.

max.var

The maximum number of variables to be included.

Value

A list with components

model.pp

An integer vector of screened model under posterior probability stopping rule. This will be null if only "eBIC" stopping criterion was used.

mdoel.ebic

An integer vector of screened model under eBIC criterion. This will be NULL if only "PP" stopping criterion was used.

postprobs

The sequence of posterior probabilities until the last included variable. This will be null if only "eBIC" stopping criterion was used. Here the last included variable is the last one included by either "PP" or "eBIC" if criteria="both" was selected

ebics

The sequence of eBIC values until the last included variable. This will be null if only "PP" stopping criterion was used. Here the last included variable is the last one included by either "PP" or "eBIC" if criteria="both" was selected

References

Wang, R., Dutta, S., Roy, V.(2021) Bayesian iterative screening in ultra-high dimensional settings https://arxiv.org/abs/2107.10175


dongjli/bravo documentation built on Sept. 20, 2021, 3:33 a.m.