The longevity R package provides provide maximum likelihood estimation routine for modelling of survival data that are subject to non-informative censoring and truncation mechanisms. It includes a selection of 12 parametric models of varying complexity, with a focus on tools for extreme value analysis and more specifically univariate peaks over threshold modelling. The package comes with visual diagnostics that account for the sampling scheme for lifetime data, utilities for univariate threshold selection, non-parametric maximum likelihood estimation, goodness-of-fit diagnostics and model comparisons tools. The different methods therein are illustrated using aggregated tabular data of longevity from Japan, and truncated lifelengths Dutch records.
翻译:暂无翻译