search

neherlab / covid19_scenarios

Models of COVID-19 outbreak trajectories and hospital demand
https://covid19-scenarios.org
MIT License
1.36k stars 352 forks source link

Default time range of simulation should be ETA of vaccine #44

Closed MarcelloPerathoner closed 4 years ago

MarcelloPerathoner commented 4 years ago

~18 month is the ETA of the vaccine. That seems like the most natural time range to simulate.

ivan-aksamentov commented 4 years ago

@MarcelloPerathoner Hi Marcello,

Yes it might be a reasonable default if you want to predict the grand totals towards the end of the pandemic.

However the most "interesting" part of the outbreak will be unrolling in the coming weeks to months and we want to put emphasis on how mitigation/containment measures influence the numbers around the peak. Also, showing longer intervals make plots quite crammed (which is a user interface/design problem).

We may change the defaults as situation evolves and as we introduce user interface improvements.

Meanwhile, you can always adjust the time range in "Simulation time range" parameter.

MarcelloPerathoner commented 4 years ago

This paper https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf suggests that a too strong suppression in the first 5 months will cause the peak to appear in December. Your default timescale will completely miss this point. See Figure 3.

ivan-aksamentov commented 4 years ago

@MarcelloPerathoner Thanks for the suggestions! I'll let our science forks to consider this. CC @nnoll @rneher

nnoll commented 4 years ago

For now, we want the model to focus on understanding the immediate hospital demand in the next 3-6 months, and the various control we have on it. Increasing the default time-range blurs this focus, as well as introduces worries of uncertainty in predictions looking that far forward in time, which our model has no way of calculating and/or displaying right now.

For now, I think we'll opt to leave it at 6 months, however the simulation time range can be edited for you to explore longer scenarios. Thank you for your suggestion -- it's certainly a good point and something we can revisit later down the road as the model improves. So I'll close this for now but feel free to re-open later!