你的代码有一个小问题，那就是它相信它每次都有输出，但实际上没有。
为了更精确的概率得到0.8或更多的2个变量是严重的偏差与非常小的概率。所以你可以反复调用猜测随机获得输出。我相信它需要平均70 - 92猜测，在最坏的情况下，它可以上升到130+尝试。
希望我解释清楚。
验证码：

import numpy as np
ar = np.array([0,1,1,0])
number_of_portfolios = 5

tickers = ['A','B','C','D'] 

portfolio_weights = []
found=0
count=0
while count<5:
    weights = np.random.random_sample(len(tickers))
    weights = np.round(weights / np.sum(weights),3)
    if sum(ar * weights) > 0.8:
        portfolio_weights.append(weights)
        count+=1
print("obtained output is :",portfolio_weights)

如果你有兴趣知道平均需要多少次尝试，运行这个来查看尝试的次数。

import numpy as np
ar = np.array([0,1,1,0])
number_of_portfolios = 5

tickers = ['A','B','C','D'] 

portfolio_weights = []
found=0
count=0
tries=0

while count<5:
    weights = np.random.random_sample(len(tickers))
    weights = np.round(weights / np.sum(weights),3)
    tries+=1
    if sum(ar * weights) > 0.8:
        portfolio_weights.append(weights)
        count+=1
print("obtained output is :",portfolio_weights)
print("number of tries it took to yeild result :",tries)

如果这有助于留下一个赞成票是赞赏的。