In this talk, I will introduce the R package, Dynamic Modeling in R (dynr), which provides a suite of fast and accessible functions for estimating and visualizing the results from fitting (possibly) regime-switching linear and nonlinear dynamic systems models in discrete as well as continuous time. Software usage examples will be demonstrated with longitudinal modeling examples involving individuals’ and couples’ self-report emotions, automated measures of physical proximity between mothers and infants during lab-based interactions, and infant-mother nighttime actigraphy data. I will highlight some graphical functions in dynr for estimating and exploring relations among derivatives, and utility functions such as dynr.mi, a routine for handling possibly non-ignorable missingness in the dependent variables and/or covariates in a user-specified discrete-time dynamic systems model via multiple imputation (MI); and dynr.taste, a routine for identifying and removing the impact of outlying data at the observed and latent levels – in other words, a procedure to help users get a taste of a particular model and make informed model adjustments.

Register