Before, I wrote a diary "I'm worried about whether or not to simulate the infection of coronavirus (COVID-19)"
In fact, I've been working on creating a simple simulator (C/C ++) using a virtual city traffic simulation (railway and bus only) as a hobby for the past year or so.
Currently, as many (virtual) humans as the ordinance-designated cities can be moved on my personal computer.
However, I thought that it would take a little more than 2 hours for a simulation of 20 hours a day, so it couldn't be used easily.(For an one hour future prediction, 10 minutes is enough)
And the bottleneck of this simulator is "moving".
In the U.S. presidential election, I don't have to move people to polling places. In this simulator, 100 units of trains and buses, and tens of thousands to hundreds of thousands of people move at a stretch and must track them in seconds.
To be honest, when I start calculating this simulation, I feel like I can hear a scream from my computer.
------
So, when I thought, "It is difficult to use it as a simulator as it is" "I wonder if a day's simulation can be done in about 20 minutes" and about remodeling, this new corona infection disaster occurred.
I've been playing with a variety of predictive techniques so far, I could not predict the spread of the new corona infection disaster.
The current reality is far from reality.
Also, with this simulator, I can't do simulations of school closures, or simulations of midnight Shinjuku and Shibuya drunks.
All I can do is a city traffic simulation in a virtual city on the premise of commuting to work.
In this simulation, throwing in a certain number of corona infected people and roughly determining the infection rate according to the riding rate, I can watch the number of infected people increasing just keeping turning the computer.
By the way, because of me, deaths according to age will also be counted.
If I sort out the results of this simulation, I may be able to watch an insane and nasty (virtual) reality.
(It's very unusual, however) I thought, "I don't want to see the simulation results"
(To be continued)