pandemic-statespace

a Bayesian statistical support to the PoPS global pest spread model

currently using jags and the data inputs obtained for a given case study with the PoPS Global Data Acquisition notebook

Objectives

• Expand on statistical tests of case study viability with temporal data
• Move towards Bayesian callibration of PoPS Global
• Benchmark model to compare performance
• Flexible option to test new modules, incorporate additional data sources, and more!

Status

Working from a static binomial regression of model predictors to a dynamic state-space model

Current (working) models

1. Static binomial
2. Dynamic (temporal) binomial
3. Dynamic binomial with interaction term
4. Dynamic state-space (basic)

Workflow

1. workspace: import packages, define case study-relevant fields
2. data: import and format static and temporal data
3. models: some work, some don't (listed)!
4. runModel: pick a model, run it, and review intermediate outputs (convergence diagnostics and summary statistics)
5. simResults: simulate and visualize results

Example outputs