dfpHorShoeSte: Performs one iteration of DFP for the Horseshoe Lasso

In Rene-Gutierrez/DynParRegReg:

Description

This function performs one step of the DFP with the Horseshoe Lasso prior. The inputs are described in Algorithm 1 and the Supplementary Materials of Guhaniyogi & Gutierrez.

Usage

dfpHorShoeSte(P, XX, Xy, yy, sN, hb, hs, hl, ht, hv, he, nmcmc)

Arguments

P	Current partition to be used.
XX	Sufficient statistic X'X.
Xy	Sufficient statistic X'y.
yy	Sufficient statistic y'y.
sN	Number of Observations.
hb	Point estimate of the coefficients.
hs	Point estimates of the variance of the error.
hl	Point estimate for the local shrinking parameter.
ht	Point estimate for the global shrinking parameters.
hv	Point estimate for the auxilary local shrinking hyper-parameter.
he	Point estimate for the auxilary global shrinking hyper-parameters.
nmcmc	Number of samples of the parameters.

Value

A list containing samples for the model parameters.

sb

A matrix with samples for coefficients, 1 sample per row.

sl

A matrix with samples for local shrinkage parameters, 1 sample per row.

sv

A matrix with samples for auxilary local shrinkage hyper-parameters, 1 sample per row.

ss

Samples for sigma.

st

Samples for the global shrinkage parameter.

se

Samples for the auxilary global shrinkage hyper-parameter.

Author(s)

Rene Gutierrez Marquez

Rene-Gutierrez/DynParRegReg documentation built on Dec. 18, 2021, 9:57 a.m.