[ Python 常見問題 ] How to sample a random number from a probability distribution in Python

 Source From Here

Question
A number selected randomly using a probability distribution will return a number according to the relative weights. For example, a number selected from [1, 2] with relative weights [.9, .1] will have a 90% chance of being 1 and a 10% chance of being 2.

HowTo
USE random.choices() TO SELECT A RANDOM NUMBER

Call random.choices(population, weights) to sample a random number from population based on the probability distribution weights. The weights are relative, meaning the percentage of each number being picked depends on what the weights sum to. e.g.:

view plaincopy to clipboardprint?

1. import random  
2. from collections import defaultdict  
3.   
4. a_list = [1, 2]  
5. distribution = [.9, .1] # weights add up to 1  
6. fdict = defaultdict(int)  
7.   
8. loop_num = 1000  
9. for i in range(loop_num):  
10.     random_number = random.choices(a_list, distribution)[0]  
11.     fdict[random_number] += 1  
12.       
13. for n, f in fdict.items():  
14.     print(f"number {n} appears with probability as {f*100/loop_num}%")  

Output:

number 1 appears with probability as 91.4%
number 2 appears with probability as 8.6%