theta_var_est_grid: Basic Grid Optimizer for Likelihood

In agfh: Agnostic Fay-Herriot Model for Small Area Statistics

Basic Grid Optimizer for Likelihood

Description

A basic grid search optimizer. Here, used to estimate the variance in the latent model by maximum likelihood.

Usage

  theta_var_est_grid(likelihood_theta_var)

Arguments

likelihood_theta_var

some flavor of likelihood function in terms of latent variance

Details

likelihood_theta_var may be created using adj_resid_likelihood_theta_var_maker or similar.

We recommended implementing a more robust optimizer.

Value

The scalar value that optimizes likelihood_theta_var, or an error if this value is on the search boundary [10^{-6}, 10^2].

Source

Marten Thompson [e-mail]

Examples

  X <- matrix(1:10, ncol=1)
  Y <- 2*X + rnorm(10, sd=1.1)
  D <- rep(1, 10)
  adj.lik <- adj_resid_likelihood_theta_var_maker(X, Y, D)
  theta_var_est_grid(adj.lik)

agfh documentation built on July 9, 2023, 6:44 p.m.