Are we over-reacting to the corona virus? Really, what would happen if life went about as usual? In this simulation we model the spread of an epidemic such as COVID-19 through the population. We expose variables the control a global random interaction versus a localized neighborhood interaction, number of such interactions, incubation period and disease period and fraction of people who quarantine. We try to understand what would happen if life went about as usual versus if we practiced very strict social distancing policies.
We use the latest 4 days data of the case count to project the next week for the US. This starts off with the last 3 data points and fits an adoption curve for those data points by solving for the coefficients algebraically. It then adds the last data point and performs a regression. The user can update this model for any country, state or city. Save yourself the mental math or needing to whip out your calculator.
We use the latest 4 days data of the case count to project the next week for China. This starts off with the last 3 data points and fits an adoption curve for those data points by solving for the coefficients algebraically. It then adds the last data point and performs a regression. The user can update this model for any country, state or city. Save yourself the mental math or needing to whip out your calculator.
In this simulation we model the impact of R0 on the total case count. When the virus broke out in Wuhan, the approximate R0 was greater than 2. Under strict quarantine practices and social distancing countries have brought it down to 1.5 quickly. Once we get R0 under 1, then we can contain the disease. Plug in your best estimates for R0 for 1, 2, 4 and 8 weeks from now to see the impact on the spread of the virus.
In this simulation, we attempt to reproduce resource usage and resource shortage caused due to COVID-19. We attempt to reproduce the approach used by IHME COVID-19 Heath Service Utilization Forecasting Team. we can overall reproduce with default assumptions the conclusions they obtained. An overage shortage of 49K beds and and 15K ICU beds and a requirement of 19K invasive ventilators. All of these parameters match our numbers here up to the hundreds. Most importantly, you can play around with the parameters to see what would happen if you were to change them to be higher or lower.
Would a dominant genetic trait automatically increase in frequency? Would the rare variant get wiped out quickly? Do we expect the dominant trait to systematically increase in frequency generation after generation until it wipes out the rare variant?
This is a really simple simulation but helps you visualize statistical significance of a single binomial variable. Let's say you have n trails and you observed k 'successes'. You often want to know, what is the success rate and the bounds around that success rate. As an example, let's say we toss a coin 444 times and we saw 386 heads. You may want to know if the bias of 386 / 444 ~ 0.87 a good estimate or do you need more tosses? This occurs A LOT in the real-world. For example, you see 444 patients and you saw 386 of a certain case. You have 444 users and 386 of them retained after a certain period of time.
How much work do you spend by moving the toilet seat up and down? Does it change based on the ratio of men to women? Does it change based on the fraction of times you perform business 1 or 2? How much more work is it if you always left the toilet seat down as opposed to being lazy and just leaving it in the position you finished your business in? Click to find out!
Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice?
How long does it take to board a flight in random order? How does boarding by zones help speed up? How bad can things be in worst case and how does this compare with other cases? What happens if you board from window to aisle instead?
Let's say that you need to interview a series of candidates and at the end of each interview you need to decide if you wanted to hire or pass on the candidate. How should you strategize to maximize the probability of finding the best candidate? Such problems occurs all the time in real life situations. While not every aspect of the real-world can be reduced to simulations, maybe you can gain some intuition by realizing how to trade-off between waiting to find the best person and being decisive and hiring the person in front of you.
You are trying to hire the best (actually top-10%) person among N candidates. . You want to examine k of them and pick the best. You are successful if you got to the top 10% (90th percentile by skill) candidate. How many candidates do you need to interview to achieve this? Again, said another way, how many candidates do you need to interview and choose the best among them so that you have a 90% chance of finding the 90th percentile candidate.
How many coin tosses do you need to perform on an average to get a single head? Seems straight-forward and the answer is indeed 2 as intuition would suggest. How about the number of coin tosses until we see hh (two consecutive heads)? Is that the same as the average number of tosses before we see ht (heads followed by tails)? What about hth (heads-tails-heads) versus hhh (heads-heads-heads) versus htt (heads-tails-tails)? Are they all equally likely? Find out in this simulation to see if your intuition is guiding you or misleading you!