clm: Conjugate Linear model

In jsilve24/driver: Utility Functions from Justin Silverman

Description

Conjugate Linear model

Usage

clm(y, X, mu0, Lambda0, a0, b0, calc_ml = TRUE)

Arguments

y	k x 1 vector
X	n x k matrix
mu0	k x 1
Lambda0	k x k
a0	scalar
b0	scalar
calc_lml	if TRUE also returns model evidence (log marginal likelihood)

Details

Model:

y_i ~ N(x_i^T β, σ^2)

β ~ N(μ0, σ^2 Λ0)

σ^2 ~ IG(a0, b0)

from https://en.wikipedia.org/wiki/Bayesian_linear_regression

Value

list with elements muN, LambdaN, aN, bN, lml

jsilve24/driver documentation built on Jan. 30, 2022, 12:07 p.m.