A Kiefer-Wolfowitz NPMLE procedure for estimation of a Gaussian model with independent mean and variance prior components with weighted longitudinal data. This version iterates back and forth from Gamma and Gaussian forms of the likelihood.
WLVmix(y, id, w, u = 300, v = 300, eps = 1e-04, maxit = 2, ...)A list consisting of the following components:
midpoints of the mean bin boundaries
the function values of the mixing density of the means
midpoints of the variance bin boundaries
the function values of the mixing density of the variances.
vector of log likelihood values for each iteration
Bayes rule estimate of the mixing density means.
Bayes rule estimate of the mixing density variances.
Mosek convergence status for each iteration
A vector of observations
A strata indicator vector indicating grouping of y
A vector of weights corresponding to y
A vector of bin boundaries for the mean effects
A vector of bin boundaries for the variance effects
Convergence tolerance for iterations
A limit on the number of allowed iterations
optional parameters to be passed to KWDual to control optimization
J. Gu and R. Koenker
Gu, J. and R. Koenker (2015) Empirical Bayesball Remixed: Empirical Bayes Methods for Longitudinal Data, J. Applied Econometrics, 32, 575-599.
Koenker, R. and J. Gu, (2017) REBayes: An R Package for Empirical Bayes Mixture Methods, Journal of Statistical Software, 82, 1--26.
WGLVmix for a more general bivariate mixing distribution version and WTLVmix for an alternative estimator exploiting a Student/Gamma decomposition