optimal_grad: Log-likelihood Gradient

In PeppeSaccardi/acmvup: Additive Covariance Modeling via Unconstrained Parametrization

Description

This function compute the gradient of the log-likelihood function given the data and the covariates

Usage

optimal_grad(par, data, lista_phi, lista_d, fnscale = 1)

Arguments

par	The point in which the gradient has to be computed
data	The matrix of the observed data
lista_phi	The list containing all the matrices of covariates to model each phi element
lista_d	The list containing all the matrices of covariates to model each d element
fnscale	Scale coefficient: default value equal to 1

Value

Gradient vector of the log-likelihood function

Examples

data <- matrix(rnorm(300), ncol=3)
lista_d <- list()
lista_phi <- list()
lista_d[[1]] <- matrix(c(rep(1,100),rnorm(100)), byrow = FALSE, ncol=2)
lista_d[[2]] <- matrix(c(rep(1,100),rnorm(200)), byrow = FALSE, ncol=3)
lista_d[[3]] <- matrix(rep(1,100), byrow = FALSE, ncol=1)
lista_phi[[1]] <- matrix(c(rep(1,100),rnorm(200)),byrow = FALSE, ncol=3)
lista_phi[[2]] <- matrix(rep(1,100),ncol=1)
lista_phi[[3]] <- matrix(c(rep(1,100),rnorm(100)),byrow = FALSE, ncol=2)
par <- rnorm(12)
optimal_grad(par,data,lista_phi,lista_d)

PeppeSaccardi/acmvup documentation built on Dec. 31, 2020, 3:28 p.m.