Figure 6
From: Dynamic calibration with approximate Bayesian computation for a microsimulation of disease spread
Model predictions generated by taking 30 samples from the posterior distribution at the end of each calibration window and running the DyME model for an additional 63 days.
