进化生物学中的模拟--太多的循环？

Question

当时，我成功地为我需要做的工作编写了模拟代码。然而，我对python还是个新手。因此，我现在正在使模拟更有效率。例如：

if random.random() < mm:
    z = numpy.random.choice(pat)

    if random.random() < 0.5:
        if random.random() < mut:
            if maleadult[z][0] == 0:
                malejuv[y][x][0] = 1
            elif maleadult[z][0] == 1:
                malejuv[y][x][0] = 0
        else:
            malejuv[y][x][0] = maleadult[z][0]
    else:
        if random.random() < mut:
            if femaleadult[z][0] == 0:
                malejuv[y][x][0] = 1
            elif femaleadult[z][0] == 1:
                malejuv[y][x][0] = 0
        else:
            malejuv[y][x][0] = femaleadult[z][0]

    if random.random() < 0.5:
        if random.random() < mut:
            if maleadult[z][1] == 0:
                malejuv[y][x][1] = 1
            elif maleadult[z][1] == 1:
                malejuv[y][x][1] = 0
        else:
            malejuv[y][x][1] = maleadult[z][1]
    else:
        if random.random() < mut:
            if femaleadult[z][1] == 0:
                malejuv[y][x][1] = 1
            elif femaleadult[z][1] == 1:
                malejuv[y][x][1] = 0
        else:
            malejuv[y][x][1] = femaleadult[z][0]

其中：

 mm - male dispersal, 
 mf - female dispersal, 
 mut - mutations, 
 pat - patch, 
 maleadult - adult male, 
 femaleadult - adult female, 
 malejuv - juvenile male, 
 femalejuv - juvenile female. 

如您所见，代码很大。这只适用于雄性，当它们散开的时候。代码的其余部分非常相似。这些都是标准的遗传和人口统计过程--但我觉得这是可以改进的。我觉得这些过程很简单，所以可能没有必要编写这么大的代码。

有没有人有任何想法来缩短这段时间，从而使其更有效率？

Answer 1

您的示例没有任何循环，但看起来可以简化一个：

if random.random() < mm:
    z = numpy.random.choice(pat)

    for i in range(2):

        if random.random() < 0.5:
            if random.random() < mut:
                if maleadult[z][i] == 0:
                    malejuv[y][x][i] = 1
                elif maleadult[z][i] == 1:
                    malejuv[y][x][i] = 0
            else:
                malejuv[y][x][i] = maleadult[z][i]
        else:
            if random.random() < mut:
                if femaleadult[z][i] == 0:
                    malejuv[y][x][i] = 1
                elif femaleadult[z][i] == 1:
                    malejuv[y][x][i] = 0
            else:
                malejuv[y][x][i] = femaleadult[z][i]

​

还可以对其进行修改的pass a mutable object as reference to a function，这允许进一步减少几乎冗余的代码。我添加了一些数据来测试它：

​

#!python3
#coding=utf-8

import random

maleadult = [[["male adult"], ["another male adult"], ]]
femaleadult = [[["female adult"], ["another female adult"], ]]
malejuv =   [[[["male juv"],["another male juv"]]]]

mut = 0.5
mm = 1.0
x = 0
y = 0
z = 0


def some_logic(a, j):
    """ does something """
    if random.random() < mut:
        if a[z][i] == 0:
            j[y][x][i] = 1
        elif a[z][i] == 1:
            j[y][x][i] = 0
        # added!
        else:
            j[y][x][i] = 0
    else:
        j[y][x][i] = a[z][i]


if random.random() < mm:
    z = 0 #numpy.random.choice(pat)
    for i in range(2):
        print(i)
        if random.random() < 0.5:
            some_logic(maleadult, malejuv)
        else:
            some_logic(femaleadult, malejuv)

print(maleadult)
print(malejuv)
print(femaleadult)