loglikMaltipooCollapsed: Calculations for the Collapsed Maltipoo Model

In jsilve24/mongrel: Bayesian Multinomial Logistic Normal Regression

Description

Functions providing access to the Log Likelihood, Gradient, and Hessian of the collapsed maltipoo model. Note: These are convenience functions but are not as optimized as direct coding of the MaltipooCollapsed C++ class due to a lack of Memoization. By contrast function optimMaltipooCollapsed is much more optimized and massively cuts down on repeated calculations. A more efficient Rcpp module based implementation of these functions may following if the future. For model details see optimMaltipooCollapsed documentation

Usage

loglikMaltipooCollapsed(Y, upsilon, Theta, X, KInv, U, eta, ell,
  sylv = FALSE)

gradMaltipooCollapsed(Y, upsilon, Theta, X, KInv, U, eta, ell,
  sylv = FALSE)

hessMaltipooCollapsed(Y, upsilon, Theta, X, KInv, U, eta, ell,
  sylv = FALSE)

Arguments

Y	D x N matrix of counts
upsilon	(must be > D)
Theta	D-1 x Q matrix the prior mean for regression coefficients
X	Q x N matrix of covariates
KInv	D-1 x D-1 symmetric positive-definite matrix
U	a PQxQ matrix of stacked variance components
eta	matrix (D-1)xN of parameter values at which to calculate quantities
ell	P-vector of scale factors for each variance component (aka VCScale)
sylv	(default:false) if true and if N < D-1 will use sylvester determinant identity to speed computation

jsilve24/mongrel documentation built on Jan. 27, 2022, 9:54 p.m.