Avoiding zerodivisionerror in between the equation

Question

In an optimization problem,if function zero(x ,y) in the code given below is objective function with initial values of x,y as 0,0.Then, to solve the problem i need to compute the obj. function as well as its derivitive. So, when i am doing so,i am getting ZeroDivisionError: division by zero.

[x,y] = [0,0]
def zero(x ,y):
    return 3 + x/y + 2*xy + 1/x

Output is:

 Traceback (most recent call last):
  File "<ipython-input-5-4e3e5e87c5f1>", line 1, in <module>
    runfile('C:/Users/HP/Documents/Python Scripts/Python Scripts_majrpjct/working/practice_cvxopt.py', wdir='C:/Users/HP/Documents/Python Scripts/Python Scripts_majrpjct/working')
  File "C:\ProgramData\Anaconda2\lib\site-packages\spyder\utils\site\sitecustomize.py", line 866, in runfile
    execfile(filename, namespace)
  File "C:\ProgramData\Anaconda2\lib\site-packages\spyder\utils\site\sitecustomize.py", line 87, in execfile
    exec(compile(scripttext, filename, 'exec'), glob, loc)
  File "C:/Users/HP/Documents/Python Scripts/Python Scripts_majrpjct/working/practice_cvxopt.py", line 394, in <module>
    print   zero(x,y)
  File "C:/Users/HP/Documents/Python Scripts/Python Scripts_majrpjct/working/practice_cvxopt.py", line 392, in zero
    return 3 + x/y
ZeroDivisionError: division by zero

Can anyone suggest me a solution so as to get correct answer by ignoring ZeroDivisionError values.

Ok, so what _should_ the correct answer be? What would you prefer `zero(0,0)` to return? — Kevin, Jun 30 '17 at 17:47
[`try/except`](https://docs.python.org/3/reference/compound_stmts.html#try)? — Christian Dean, Jun 30 '17 at 17:49
"if we consider P = ( 3+ x/y) , if (x,y) = (0,0), then the value of P = 3" - you need to get your math straight first, because 3+0/0 is not 3. The division operation is not defined for second argument 0. — user2357112, Jun 30 '17 at 17:49
0/0 is undefined, what is your argument for you to say is 0? — eyllanesc, Jun 30 '17 at 17:50
He said what he wanted to do when y = 0. He wants it to be zero. That's a valid thing to want it to be, depending on the situation. He just doesn't know how to make that happen. Forgive my limited pronouns if you're not a he. — Alan Leuthard, Jun 30 '17 at 17:51
See [this answer](https://stackoverflow.com/a/3931746/355230) to another question that describes how/whether to use `try/except` to handle cases like this. — martineau, Jun 30 '17 at 18:10

Alan Leuthard · Answer 1 · 2017-06-30T17:59:19.277

This is what try blocks are for! If you don't want to sanitize your inputs and are coding for a situation where a division by zero IS zero (the other option is infinity which makes for some much more interesting math), then a simple try block is the way to go. If you need precision, sanitize your inputs (make sure y isn't zero before sending it to the function).

Code:

def zero(x ,y):
    try:
        return 3 + x/y
    except ZeroDivisionError:
        return 3
        #  or do whatever else you want to do when y is zero.

TMueller83 · Answer 2 · 2017-06-30T17:59:14.277

0

How about this:

def zero(x, y):
    if abs(x) < 1e-10 and abs(y) < 1e-10:
        return 3.0
    elif abs(y) < 1e-10:
        return 0.0
    else:
        return 3 + x/y

edited Jun 30 '17 at 17:59

answered Jun 30 '17 at 17:50

TMueller83

412
3
11

But the thing is,i need this part of code to be used in my optimization problem where the calculations are more complex involving variouscomponents and calculated through iterations. – Manasaveena Jun 30 '17 at 18:03
Then you have to make the `if`, `elif`, `else` tests more complex, and you have to specify the result for each limiting case. If `x` and `y` are for example arrays, then you could use – TMueller83 Jun 30 '17 at 18:07
... `y=np.array(y)` and test if there are zeros in `y` by `np.any(abs(y) < 1e-10)`, Where `np` refers to `numpy` imported as `np`. – TMueller83 Jun 30 '17 at 18:17

Avoiding zerodivisionerror in between the equation

2 Answers2