blakeelias
commented
3 years ago Population 1000:
Population 5000:
Population 10000:
Population 50000
No change in number of tolerated cases. 0 is green (re-open: R = 2.5) 1 is white (R=0.9) 2 is red (lock-down; R = 0.7). 3: R = 0.6 4: R = 0.6 5: R = 0.6 6: R = 0.6
Number of tolerable cases doesn't vary with population. This makes sense.
Wrong intuition: being a big country somehow makes that single case more tolerable. Right intuition: having 1 case is almost as bad as having 100 cases -- you're always just a few days away from that if you don't act.
It does vary with number of imported cases: with stronger border control (low importation), you can get to very small numbers of cases, so you'll choose to do it. With weak border control (high importation), getting to small case counts will be too expensive, so you'll "give up" and tolerate some higher number of cases.
Wrong intuition: Big countries will have to tolerate some background number of cases. Right intuition: Countries with weak borders have to tolerate some background number of cases. (Fair, the bigger the country, the harder to keep the border secure. But the point is that the level of background cases depends on things you can actively control [eg border policy], rather than things you can't [eg size].)
python main.py --num_population 1000 5000 10000 50000 100000 500000 1000000 --imported_cases_per_step_range 0.1 --powers 1.0 --action_frequency 1 --horizon 28 --distr_family poisson --R_0 2.5 --dynamics SIS --policy-optimization --no-policy-comparison --vaccine_schedule none