R Notebooks, featuring data analysis and code from select chapters of Ottar N. Bjornstad's Epidemics, for the purpose of learning time series analysis of infectious disease data. This includes using discrete-time and continuous-time SIR simulation models as a latent process which mechanistically constrains the trajectories of infectious disease dynamics. We use the time series data analysis tools to infer what the parameters in the SIR simulation models are. From this, for example, we can discover the seasonal variation in interaction strength, for example due to the schedule of the school year.
TSIR belongs to a class of techniques for fitting mechanistic models to data. POMP also belongs to a similar class, but is more sophisticated.
- Chapter 3: R_0 - stochastic simulation of SIR
- Chapter 6: Time Series Analysis - Background, review of techniques, time & frequency domain analysis methods, wavelet analysis
- Chapter 7: TSIR - Estimating parameters in time series data
- Chapter 8: Trajectory Matching