Programmers make their own epidemic simulation program: The spread of the epidemic is more terrible than you think

btty038

Programmers make their own epidemic simulation program: The spread of the epidemic is more terrible than you think [Copy link]

• The spread of the epidemic is more terrible than you think

yang_alex

In reality, there are more variables and the models are more complex.

bigbat

Mathematical models are very complex. The usual simple probability model is the "normal distribution" model. This program does not even do a local diffusion well. It is far from the actual model. The real model requires many parameters. Therefore, some models are very expensive. It costs hundreds of thousands.

bigbat

Normal distribution model: in five steps

1. Define a range, such as 0-2000. Generate a random number Rx, and it can be assumed that the probability of Rx in the range of 0-2000 is balanced.

2. Construct a function, such as the normal function (Gaussian probability function), substitute x=Rx into the function y=f(x) to find the function value Ys, and use cP=Ys/Ymax (the maximum value of the Ymax function).

3. Generate a random number Ry between 0 and Ymax. Compare Ry with cP, if Ry>cP, discard it, if Ry<=cP, Ry is retained and displayed.

4. Repeat this process several times, and the ratio of the total number of cycles Nz to the effective number Ne is the probability that it roughly conforms to the normal distribution.

5. If you want to display it more vividly, you can use a circular pattern. The effective value Ry is the distance from the center of the circle. Rx is the angle of the circumference. Polar coordinates.

btty038

So much development means too many variables