During testing for #290 I noted that forecasts that were reasonable for underlying latent process, the forecasts could be unreasonable for cases (e.g. explosive growth). A simple fix would be including the population constraints into the modelling as per new EpiAware functionality.
This isn't in the analysis plan and also doesn't really represent a "standard" so I would vote to hold on this for now. Potentially a prior tweak would help in the first case.
During testing for #290 I noted that forecasts that were reasonable for underlying latent process, the forecasts could be unreasonable for cases (e.g. explosive growth). A simple fix would be including the population constraints into the modelling as per new
EpiAware
functionality.