Back to modelling time. The other day I had written a blogpost exploring a few scenarios of spread of Covid-19, and kinda what regulation should look like. https://noenthuda.com/blog/2020/03/30/simulating-covid-19-scenarios/

This morning I was "presenting the blogpost" to a bunch of people from @TakshashilaInst , and thought it might add value to change parameters in the middle of the discussion.

So I used @rstudio Shiny to build up some web dashboards for simulations

These are public, so I thought I'll share them here so you can play around.

Play around with the parameters on the left and hit "recompute" when done.

First up, we have the one-person shop / delivery model.

http://206.189.17.47:3838/CovidDelivery/

This simulates how a one-customer-at-a-time shop or delivery can spread the disease. We played around with multiple scenarios. The broad conclusion is that it doesn't spread much.

It's pretty safe. Deliveries and small shops can be opened up any time.

Next we came to "supermarkets". Here, multiple people come in and go out as per their own pleasure. This was the unclear one.

Once incidence in a geographical area has crossed a threshold (maybe 5%) these become dangerous. Else they're fine

http://206.189.17.47:3838/CovidShop/

While I call it "shop" it can be used to describe any other establishments where people come and go as they please, and where they are in close proximity.

Can include temples, gyms, etc. as well.

Then we have the "office" simulation.

http://206.189.17.47:3838/CovidOffice/

Here, a bunch of people in close proximity who spend a period of time together. Close enough proximity that anyone can infect anyone else.

What simulations show is that even if we start with one infected person, thing can blow up quickly and everyone can be infected in the matter of a small number of hours.

However, if there is a high likelihood that nobody is infected, then such establishments can be open as well

Again I said "office" here, but can include any situation where a bunch of people get together in one place and stay together for an extended period of time.

Includes schools, gym classes, mosque prayers, etc.

Finally there's the simulation for "conference". Large number of people in one place, but they congregate in small groups at a time.

You can think of the Nizamuddin Mosque or the Boston Biomed (?) conference as examples of this

http://206.189.17.47:3838/CovidConference/

Even sporting events and large parties (think Kannika Kapoor) can be described using this "conference" model. Again, there is a chance of quick blowup if we start with a high incidence of infection.

Play around with the links above (some of them may not be directly clickable, so you might need to copy paste them in the bar).

And share any other insights you might get on how the disease spread may be stopped.

Finished.