Estimation of Extended Mixed Models Using Latent Classes and Latent Processes: The R Package lcmm

The R package lcmm provides a series of functions to estimate statistical\nmodels based on linear mixed model theory. It includes the estimation of mixed\nmodels and latent class mixed models for Gaussian longitudinal outcomes (hlme),\ncurvilinear and ordinal univariate longitudinal outcomes (lcmm) and curvilinear\nmultivariate outcomes (multlcmm), as well as joint latent class mixed models\n(Jointlcmm) for a (Gaussian or curvilinear) longitudinal outcome and a\ntime-to-event that can be possibly left-truncated right-censored and defined in\na competing setting. Maximum likelihood esimators are obtained using a modified\nMarquardt algorithm with strict convergence criteria based on the parameters\nand likelihood stability, and on the negativity of the second derivatives. The\npackage also provides various post-fit functions including goodness-of-fit\nanalyses, classification, plots, predicted trajectories, individual dynamic\nprediction of the event and predictive accuracy assessment. This paper\nconstitutes a companion paper to the package by introducing each family of\nmodels, the estimation technique, some implementation details and giving\nexamples through a dataset on cognitive aging.\n