comp_logml: Compute Maximum log-Likelihood

In PatrickPfeifferDSc/bite: Bayesian Inference On Treatment Effects For Panel Data

Description

Internal. comp_logml will return a numeric value, the maximum log-likelihood y = X*alpha+epsilon, epsilon N(0,S), alpha ~ N(0,inv(iA0))

Usage

comp_logml(ySy, XSy, invBN, indic, iB0)

Arguments

ySy	matrix product of y vector standardized by standard deviation
XSy	matrix product of data, standardizing matrix and y outcome
invBN	inverse beta matrix
indic	indices for regression effects, their likelihood to be accessed
iB0	k x k prior precision matrix

Details

ySy ... y'S^-1y XSy ... X'S^-1y invBN=X'S^-1X+ iB0 iB0 k x k prior precision matrix indic k indicators for regression effects

Value

Returns an object containing the numeric value of the marginal Likelihood and beta parameters

See Also

lcondx

PatrickPfeifferDSc/bite documentation built on Aug. 22, 2019, 9:57 a.m.