Add optional support to fit to count data with CFR/IFR correction

epiforecasts / inc2prev

Estimate epidemiological quantities from repeated cross-sectional prevalence measurements

http://epiforecasts.io/inc2prev/

BSD 3-Clause "New" or "Revised" License

16 stars 7 forks source link

Add optional support to fit to count data with CFR/IFR correction #14

Open seabbs opened 2 years ago

seabbs commented 2 years ago

Allows tracking of for example severity over time.

sbfnk commented 2 years ago

If what you're thinking to have a time-varying (with some smoothing) scaling factor (and possibly delay, though may run into identifiability issues again)? Very much agree that this would be a great addition.

seabbs commented 2 years ago

yes and I don't think it will as long as the delay distribution is not time-varying or at least not time-varying on the same length scale as the scaling factor (as in the CFR work).