Random effects
in
AD Model Builder

Do you find standard statistical packages too restrictive?

ADMB-RE provides great flexibility for use of random effects in nonlinear models

News
List of publications where ADMB-RE has been used.
glmmADMB: R-package for fitting mixed models to overdispersed and zero inflated count data. glmmADMB is implemented using ADMB-RE, but it is free!


Examples
Nonlinear mixed models (nlme type)
Frailty models in survival analysis
Semiparametric regression
Stochastic volatility models

Fields of application
Statistical modelling
Financial time series
Fisheries assessments
Pharmacokinetics

More information
Example collection
User manual
User forum
A note for ADMB users

Buy ADMB-RE
Contact us at orders@otter-rsch.com

HOME
AD Model Builder home page

Free evaluation version of ADMB-RE can be downloaded here

Features

  • Nested and crossed random effects
  • Exact marginal likelihood by importance sampling
  • Seamless switch between maximum likelihood and MCMC based inference

Technical details

  • Model specification in C++ like language
  • Hyper-parameters (variance components etc.) estimated by maximum likelihood
  • Marginal likelihood evaluated by the Laplace approximation or importance sampling
  • ADMB-RE calculates exact derivatives using Automatic Differentiation
  • All the useful features of ordinary AD Model Builder are available

Why choose ADMB-RE?

  • Flexibility: In principle you can implement any random effect you can think of
  • Convenience: Computational details are transparent. Your only responsibility is to formulate the loglikelihood
  • Computational efficiency: ADMB-RE is up to 50 times faster than winBUGS
  • Robustness: With exact derivatives you can fit highly nonlinear models
  • Convergence diagnostic: The gradient of the likelihood function provides a clear convergence diagnostic, while with MCMC judging convergence is difficult.

Updated August 2006

© Otter Research Ltd.