What's for dynr: A Package for Linear and Nonlinear Dynamic Modeling in R

Area of Science:

• Behavioral Sciences
• Computational Statistics
• Psychometrics

Background:

• Intensive longitudinal data in behavioral sciences are often noisy and multivariate.
• Models with regime switches are increasingly used but lack accessible software.
• Existing software is often slow and not freely available.

Purpose of the Study:

• Introduce the dynr R package for analyzing complex longitudinal data.
• Provide the mathematical and computational foundations of the dynr package.
• Demonstrate the utility of dynr through illustrative examples.

Main Methods:

• Developed an R package (dynr) using C for computational efficiency.
• Implemented a broad class of linear and nonlinear discrete- and continuous-time models.
• Incorporated regime-switching properties and linear Gaussian measurement functions.

Main Results:

• The dynr package offers a fast and freely accessible solution for regime-switching models.
• Handles both discrete- and continuous-time dynamics with discontinuities.
• Successfully demonstrated unique features through two case studies.

Conclusions:

• dynr provides a powerful and user-friendly tool for advanced longitudinal data analysis.
• Facilitates the study of complex dynamics and regime shifts in behavioral research.
• Addresses the scarcity of accessible and efficient software for such models.