Randomness and simulation
Python for AP CS Principles Lesson 10 2:02 English narration · English + 中文 subtitles burned in
Chapters
Transcript
Some programs need chance — games, experiments, models of the real world.
有些程序需要"随机"——游戏、实验、对真实世界的模型。
The random tools live in a module, so you write import random at the top.
随机工具放在一个模块里，所以你在开头写 import random。
Then randint of one and six rolls a die, and every time you run it you may get a different number.
然后 randint(1, 6) 就是掷一个骰子， 每跑一次你都可能得到不同的数字。
One detail matters for the exam: both ends are included, so a one and a six are just as possible as anything between.
有一个细节对考试很要紧：两端都是包含在内的， 所以 1 和 6 跟中间任何一个数一样有可能出现。
Which raises a problem: if the numbers change every run, how do you test the program?
这就带出一个问题： 如果每次跑出来的数字都不一样，你怎么测试这个程序？
The answer is a seed.
答案是"种子"。
A seed sets the random starting point, and with the same seed you get the same numbers every time.
种子设定随机的起点， 用同一个种子，你每次都会得到同样的数字。
Look at the third column — it is identical on every run.
看第三列——它每一次运行都一模一样。
That is not random being broken; it is random being controlled, so results can be repeated.
这不是随机坏掉了；这是随机被"控制"住了， 好让结果可以复现。
Now a simulation: a model of a real process, built out of those random values.
现在来做一个模拟： 用那些随机值搭出来的、对真实过程的模型。
Flip a coin many times, counting heads.
抛很多次硬币，数正面的次数。
At ten flips the answer looks nothing like a fair coin — thirty percent.
抛 10 次的时候，结果看上去完全不像一枚公平的硬币——30%。
Keep going.
继续抛。
At a hundred it is closer, and at a thousand it settles right by the halfway line.
100 次的时候接近了些，1000 次的时候就稳稳落在一半那条线旁边。
That is why a simulation runs many trials: only then does the long-run pattern appear.
这就是模拟要跑很多次的原因： 只有这样，长期的规律才会显现出来。
On the exam the same thing is written with RANDOM in capitals.
在考卷上，同样的东西写作大写的 RANDOM。
RANDOM of a and b gives a whole number from a to b, which is exactly randint — both ends included.
RANDOM(a, b) 给出一个从 a 到 b 的整数， 跟 randint 完全一样——两端都包含在内。
Four things to take with you.
带走四点。
One: randint gives a whole number, both ends included.
第一：randint 给出一个整数，两端都包含。
Two: a seed makes a random run repeat exactly.
第二：种子让一次随机运行可以精确复现。
Three: a simulation models a real process with chance.
第三：模拟用随机来给真实过程建模。
Four: run many trials to see the long-run pattern.
第四：多跑几次，才看得见长期的规律。
Now build a simulation in the tasks below.
现在去下面的题里做一个模拟。