Numpy中无法存储多次模拟结果，仅保留最后一次数据的问题

核心问题定位 #

你的代码里，每次模拟循环内执行pandemiclength2 = numpy.append(pandemiclength1, day)，但pandemiclength1在循环过程中始终是初始的空列表，而且pandemiclength1 = pandemiclength2是放在最外层循环外面的，这就导致每次循环都只是把当前的day追加到空列表里，最后只剩最后一次模拟的结果。

修复方案 #

方法一：用列表收集结果（高效且适合新手） #

numpy数组每次追加会重新分配内存，小数据量没问题，但先通过普通列表收集所有结果，最后转成numpy数组更高效：

1. 初始化空列表pandemic_lengths = []
2. 每次模拟结束后，把当前的day追加到列表：pandemic_lengths.append(day)
3. 所有模拟完成后，转成numpy数组：pandemic_lengths_np = numpy.array(pandemic_lengths)

方法二：修复numpy数组追加逻辑 #

如果坚持用numpy数组，要确保每次追加是基于上一次的结果：

1. 初始化pandemiclength = numpy.array([])
2. 每次模拟结束后执行：pandemiclength = numpy.append(pandemiclength, day)，这样每次都是在已有数组后追加新值。

修改后的完整代码示例（方法一） #

import numpy

# 模拟次数
simulation_times = 3
pandemic_lengths = []

# 循环执行模拟
for _ in range(simulation_times):
    # 变量初始化
    day = 0
    p = 0.01
    infected = 0
    daysinfected = numpy.zeros(60)
    z = 0

    # 第0天模拟
    print(' ')
    print('==========================')
    print("DAY: 0")
    print('==========================')
    result = numpy.random.binomial(1, p, 60)
    print(result)
    for ele in result:
        if ele == 1:
            infected += 1
    print("Number Infected:", infected)

    daysinfected = numpy.where(result == 1, 3, 0)
    print("Days Infected")
    print(daysinfected)

    # 后续天数循环，直到无感染者
    while infected > 0:
        day += 1
        print(' ')
        print('==========================')
        print("DAY:", day)
        print('==========================')

        # 更新每个个体的剩余感染天数
        daysinfected = numpy.where(daysinfected > 0, daysinfected - 1, 0)

        # 为康复者设置免疫标记
        for i in range(len(daysinfected)):
            if result[i] == 1 and daysinfected[i] == 0:
                result[i] = -1

        infected -= z
        print("Children Infected at Start of Day:", infected)

        # 模拟当天的感染传播
        for i in range(infected):
            print("---trial", i, "---")
            new_values = numpy.random.binomial(1, p, 60)
            result = numpy.where(result == 0, new_values, result)
            infected = 0
            for ele in result:
                if ele == 1:
                    infected += 1

            # 为新感染者设置感染天数
            for i in range(len(daysinfected)):
                if result[i] == 1 and daysinfected[i] == 0:
                    daysinfected[i] = 3

            print(result)
            print("Number Infected:", infected)
            print("Days Infected")
            print(daysinfected)

        z = numpy.count_nonzero(daysinfected == 1)
        print("Totalabouttoflip", z)

        print("Total Infected at End of", day, "days is", infected)
        print("Days Infected - less 1")
        print(daysinfected)

    # 收集当前模拟的疫情持续天数
    pandemic_lengths.append(day)

# 转成numpy数组，方便后续导出Excel
pandemic_lengths_np = numpy.array(pandemic_lengths)
print("The Results of", simulation_times, "simulations is", pandemic_lengths_np)

额外优化建议 #

• 用for _ in range(simulation_times)代替while循环，避免手动管理计数变量，逻辑更清晰
• 变量名改用下划线分隔（如pandemic_lengths），符合Python命名规范，可读性更强
• 可以把单轮模拟的核心逻辑封装成函数，比如def run_simulation(p, num_children=60):，然后循环调用函数收集结果，代码结构更整洁

内容的提问来源于stack exchange，提问作者lwadz88