Demo Vignettes
Welcome to epymorph! This collection of vignettes serve as practical, example-focused exercises which complement the material in the User Guide and the API documentation. We encourage you to wander freely between the three resources, as you find helpful.
Single-node usage
1. Introduction to epymorph
Using a single-node SIRS model and showcasing the basic simulation and visualization capabilities of epymorph. We use epymorph to explore a few fundamentals of epidemiology.
2. Birth and Death
Creating an SIR model with host demography, to visualize damped oscillations.
3. Hospitalization
Creating an SIRH model (with hospitalization), showcasing forking in the IPM specification.
4. Time-varying Transmission Rate
Creating an SIRS model with seasonal forcing, demonstrating epymorph’s method of specifying function-based parameter values.
5. Parameter Estimation with Particle Filtering
Using simulated and real hospitalization data to estimate time-varying transmission rate with epymorph’s built-in particle filter methods.
Multi-node usage
1. Intro to Movement Models
Creating multi-node simulations, demonstrating movement models of various complexity, and introduction to the choropleth mapping functionality.
2. Age-stratified, multi-node SIRH model with multiple movement models
Creating a multi-strate RUME, exploring how to add different movement models per strata.
Parameter Fitting and Model Forecasting
1. Location- and Time-varying Parameter Estimation with Particle Filtering *2.0 beta!
Using real hospitalization data to estimate State-specific and time-varying transmission rates with epymorph’s particle filter with localization.
2. Comparing the Particle Filter and Ensemble Kalman Filter *2.0 beta!
Comparing inference performed using the ensemble Kalman filter and particle filter on synthetic data.
3. Forecasting using the Particle Filter *2.0 beta!
Forecasting influenza hospitalization numbers using real hospitalization data from the CDC and the particle filter.
Feel free to explore each vignette in order or jump to the one that interests you the most.