This repository contains the code for the paper Xiaoming Fu et al. "Simultaneous Identification of Changepoints and Model Parameters in Switching Dynamical Systems" (2024). To run the code repository locally, please check the Project.toml
file for the required packages and install them in your local Julia environment.
scripts/Bayesian_ChangePointStudy
main scripts to run the Bayesian inference for the models in the paper.scripts/plot_scripts
scripts to generate the figures in the paper.src
main Julia code used to build the ode models, adjoint sensitivity analysis, and Bayesian inference. Some of the code that calculates gradient information of the parameters are adapted from the SciMLSensitivity.jl with copyrightCopyright (c) 2016-2020: ChrisRackauckas, Julia Computing.
data
directory to save the generated simulation data and the raw data from covid-19-data.test
directory to run the tests for the correctness of reverse-mode AD and the gradient calculation of the parameters.