How to rectify code from zero division error:

Question

我正在编写一个程序来求一个数的立方根。我得到除以零的错误。我已将所有值更改为浮动值，但仍然出现错误。

这是代码。

def cube_root(num):
    import random
    g = random.randint(1,num+1)
    new_g = float(num/(g*g))
    check_ifclose(g,new_g,num)

def check_ifclose(g,new_g,num):
    from math import isclose
    if isclose(g, new_g, abs_tol=0.0000001):
        print("The cube root of ",num,"is ",float(new_g))
    else:
        g = float(new_g)
        new_g = float(num/(g*g))
        check_ifclose(g,new_g,num)

这是我遇到的错误

ZeroDivisionError                         Traceback (most recent call last)
<ipython-input-53-7ad3ad630bd7> in <module>
----> 1 cube_root(16)

<ipython-input-52-46121e93a109> in cube_root(num)
      3     g = random.randint(1,num+1)
      4     new_g = float(num/(g*g))
----> 5     check_ifclose(g,new_g,num)
      6 
      7 def check_ifclose(g,new_g,num):

<ipython-input-52-46121e93a109> in check_ifclose(g, new_g, num)
     12         g = float(new_g)
     13         new_g = float(num/(g*g))
---> 14         check_ifclose(g,new_g,num)

ZeroDivisionError: float division by zero

Answer 1

问题是，如果您的数字非常小，如 g（小于 1e-200），则 g * g python "treats" 为 0。在这里查看 https://docs.python.org/3.7/tutorial/floatingpoint.html。

Answer 2

在new_g = float(num/(g*g))中，g*g是浮点数。小数点多于 308，导致值变为 0.0.

求立方根可以试试num**1/3

Answer 3

为什么不在每次迭代时打印值？你会看到发生了什么，然后你可以尝试调试它并重新考虑你的解决方案

就这样：

def cube_root(num):
    import random
    g = random.randint(1,num+1)
    new_g = float(num/(g*g))
    print("g", g, "new_g", new_g, "num", num)
    check_ifclose(g,new_g,num)

def check_ifclose(g,new_g,num):
    from math import isclose
    print("g", g, "new_g", new_g, "num", num)
    if isclose(g, new_g, abs_tol=0.0000001):
        print("The cube root of ",num,"is ",float(new_g))
    else:
        g = float(new_g)
        new_g = float(num/(g*g))
        check_ifclose(g,new_g,num)

然后 cube_root(16) 你会得到输出：

>>> cube_root(16)
g 2 new_g 4.0 num 16
g 2 new_g 4.0 num 16
g 4.0 new_g 1.0 num 16
g 1.0 new_g 16.0 num 16
g 16.0 new_g 0.0625 num 16
g 0.0625 new_g 4096.0 num 16
g 4096.0 new_g 9.5367431640625e-07 num 16
g 9.5367431640625e-07 new_g 17592186044416.0 num 16
g 17592186044416.0 new_g 5.169878828456423e-26 num 16
g 5.169878828456423e-26 new_g 5.986310706507379e+51 num 16
g 5.986310706507379e+51 new_g 4.464794497196387e-103 num 16
g 4.464794497196387e-103 new_g 8.02633041618099e+205 num 16
g 8.02633041618099e+205 new_g 0.0 num 16
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 6, in cube_root
  File "<stdin>", line 9, in check_ifclose
  File "<stdin>", line 9, in check_ifclose
  File "<stdin>", line 9, in check_ifclose
  [Previous line repeated 8 more times]
  File "<stdin>", line 8, in check_ifclose
ZeroDivisionError: float division by zero

如你所见，这就像 Razmik 所说的小数字。

但是为什么呢？

因为您的解决方案存在缺陷。你确定你应该这样计算 new_g 吗？如果你看这里，它会变得混乱 - 非常小，然后很大，然后又很小...你为什么不试试，例如将 new_g 设置为旧的和当前的新的平均值以平衡它？

编辑：决定检查我最后的建议，它现在工作正常。

def cube_root(num):
    import random
    g = random.randint(1,num+1)
    new_g = float(num/(g*g))
    print("g", g, "new_g", new_g, "num", num)
    check_ifclose(g,new_g,num)

def check_ifclose(g,new_g,num):
    from math import isclose
    print("g", g, "new_g", new_g, "num", num)
    if isclose(g, new_g, abs_tol=0.0000001):
        print("The cube root of ",num,"is ",float(new_g))
    else:
        g = float(new_g)
        new_g = (g+float(num/(g*g)))/2
        check_ifclose(g,new_g,num)

>>> cube_root(16)
g 3 new_g 1.7777777777777777 num 16
g 3 new_g 1.7777777777777777 num 16
g 1.7777777777777777 new_g 3.420138888888889 num 16
g 3.420138888888889 new_g 2.3939850310351836 num 16
g 2.3939850310351836 new_g 2.592869433541053 num 16
g 2.592869433541053 new_g 2.4863846555355664 num 16
g 2.4863846555355664 new_g 2.537249169224923 num 16
g 2.537249169224923 new_g 2.51131729000868 num 16
g 2.51131729000868 new_g 2.524147960725603 num 16
g 2.524147960725603 new_g 2.517700180896523 num 16
g 2.517700180896523 new_g 2.52091579334918 num 16
g 2.52091579334918 new_g 2.519305938864469 num 16
g 2.519305938864469 new_g 2.520110351423893 num 16
g 2.520110351423893 new_g 2.51970801680198 num 16
g 2.51970801680198 new_g 2.5199091519863974 num 16
g 2.5199091519863974 new_g 2.5198085763676823 num 16
g 2.5198085763676823 new_g 2.519858862169772 num 16
g 2.519858862169772 new_g 2.519833718766991 num 16
g 2.519833718766991 new_g 2.5198462903429375 num 16
g 2.5198462903429375 new_g 2.519840004523604 num 16
g 2.519840004523604 new_g 2.519843147425431 num 16
g 2.519843147425431 new_g 2.519841575972557 num 16
g 2.519841575972557 new_g 2.519842361698504 num 16
g 2.519842361698504 new_g 2.519841968835408 num 16
g 2.519841968835408 new_g 2.5198421652669256 num 16
g 2.5198421652669256 new_g 2.519842067051159 num 16
The cube root of  16 is  2.519842067051159

How to rectify code from zero division error:

How to rectify code from zero division error:

python

divide-by-zero

data-structures

编辑：决定检查我最后的建议，它现在工作正常。