Wolfram Language Paclet Repository
Community-contributed installable additions to the Wolfram Language
MMA MixedEffects
Native Gaussian, generalized, and nonlinear mixed-effects modeling with ordinary Wolfram expressions, structured random and residual effects, small-sample inference, progress reporting, and property-driven fitted-model objects.
Fit models
— fit Gaussian ML or REML models with crossed or nested random effects, residual correlation, and heteroscedasticity.
— fit GLMMs using Laplace or adaptive Gauss-Hermite quadrature and a broad response-family library.
— fit nonlinear mixed models from symbolically differentiated Wolfram expressions.
Model structure
— construct explicit composite grouping levels for multilevel designs.
— construct validated core, negative-binomial, beta, lognormal, Student-t, zero-inflated, and hurdle families.
Fitted results
— extract estimates, covariance components, predictions, residuals, likelihood criteria, inference, and convergence diagnostics.
Workflow
Specify fixed bases as ordinary expressions; add one or more random structures; fit; inspect convergence and singularity separately; select population, conditional, or marginal prediction; then validate residual and covariance assumptions.
Long fits can expose ProgressReporting, EvaluationMonitor, and StepMonitor without starting parallel kernels. R and Python are optional external validation tools, never runtime dependencies.
""