LogLikeC: Calculate the conditional likelihood for the univariate and...

In schildjs/ods4lda: Outcome Dependent Sampling (ODS) Designs for Linear Mixed Models Data Analysis

Description

Calculate the conditional likelihood for the univariate and bivariate sampling cases across all subjects (Keep.liC=FALSE) or the subject specific contributions to the conditional likelihood along with the log-transformed ascertainment correction for multiple imputation (Keep.liC=TRUE).

Usage

LogLikeC(y, x, z, w.function, id, beta, sigma0, sigma1, rho, sigmae,
  cutpoints, SampProb, SampProbi, Keep.liC = FALSE)

Arguments

y	response vector
x	sum(n_i) by p design matrix for fixed effects
z	sum(n_i) by 2 design matric for random effects (intercept and slope)
w.function	options include "mean" "intercept" "slope" and "bivar"
id	sum(n_i) vector of subject ids
beta	mean model parameter p-vector
sigma0	std dev of the random intercept distribution
sigma1	std dev of the random slope distribution
rho	correlation between the random intercept and slope
sigmae	std dev of the measurement error distribution
cutpoints	cutpoints defining the sampling regions. (a vector of length 4 c(xlow, xhigh, ylow, yhigh))
SampProb	Sampling probabilities from within each region (vector of length 2 c(central region, outlying region)).
SampProbi	Subject specific sampling probabilities. A vector of length sum(n_i). Not used unless using weighted Likelihood
Keep.liC	If FALSE, the function returns the conditional log likelihood across all subjects. If TRUE, subject specific contributions and exponentiated subject specific ascertainment corrections are returned in a list.

Value

If Keep.liC=FALSE, conditional log likelihood. If Keep.liC=TRUE, a two-element list that contains subject specific likelihood contributions and exponentiated ascertainment corrections.

schildjs/ods4lda documentation built on March 16, 2020, 8:16 a.m.