forecast_infections

[seabbs]

To bring estimate_infections into line with forecast_secondary and friends there should the the option of forecasting infections in forecast_infections.

This would work in the same way as simualte_infections but instead of taking a modified Rt time series it should simulate the generative process for the Rt model when not contained in the posterior predictions. This would decoupled forecasting from estimation.

See simulate_infections here: https://epiforecasts.io/EpiNow2/reference/simulate_infections.html

[sbfnk]

I agree with this, but to be in line with the secondary models shouldn't it be the other way round? simulate_infections from parameters and forecast_infections from an existing estimate? Or we swap them for the secondary model.