User guide¶
The user guide follows the lifecycle of a statistical analysis. Method-specific likelihood details live in the LMM, GLMM, and MMRM guides; every Python symbol is also indexed in the API reference.
Supported inputs, missingness contracts, source-row audits, formula grammar, factor handling, and design-matrix inspection.
Probability calculations, default links, wrappers, censoring/survival objects, and the boundary between the family catalog and fit support.
Random blocks, structured residual covariance, visit ordering, parameterization, validation, singularity checks, and simulation.
Engine and method choice, compilation, numerical controls, structured convergence, warning codes, and recovery workflow.
Coefficient covariance, linear inference, bootstrap, result comparison, scale, and degrees-of-freedom labels.
Auditable residual tables, random effects, singularity, reproducible simulation, and visual predictive checks.
FitResult, manifests, hashes, portable archives, reports, traceability, and
validation bundles.