Skip to content
Browse

/errors/python/zerodivisionerror-division-by-zero

ZeroDivisionError: division by zero

verified onenvironmentlocal · windows 10.0.26200 arm64 · GNU bash 5.3.15 (Git Bash) · Python 3.13.15updated
tool
Python
exit code
1
versions
Python 3.x
os
any

ZeroDivisionError: division by zero means the right-hand side of / was 0. Python raises it for integers and floats. //, % and divmod() raise it too, with their own text. Check the divisor first, and ask why it is zero: it is often the length of an empty list.

Reproduce it

bash
python -c "print(1 / 0)"

Output · exit 1

Traceback (most recent call last):
  File "<string>", line 1, in <module>
    print(1 / 0)
          ~~^~~
ZeroDivisionError: division by zero

The other operators use other messages

bash
cat > ops.py <<'EOF'
for expr in ["1 / 0", "1 // 0", "1 % 0", "divmod(1, 0)", "1.0 / 0", "0 ** -1"]:
    try:
        eval(expr)
    except ZeroDivisionError as e:
        print(f"{expr:14} -> ZeroDivisionError: {e}")
EOF
python ops.py

Output

1 / 0          -> ZeroDivisionError: division by zero
1 // 0         -> ZeroDivisionError: integer division or modulo by zero
1 % 0          -> ZeroDivisionError: integer modulo by zero
divmod(1, 0)   -> ZeroDivisionError: integer division or modulo by zero
1.0 / 0        -> ZeroDivisionError: float division by zero
0 ** -1        -> ZeroDivisionError: 0.0 cannot be raised to a negative power

Fix: guard the divisor

bash
cat > avg.py <<'EOF'
def average(values):
    return sum(values) / len(values) if values else 0.0

print(average([2, 4]))
print(average([]))
EOF
python avg.py

Output

3.0
0.0

Other number types

Fraction and Decimal raise a ZeroDivisionError too, while some math functions report a zero divisor as a different exception:

bash
cat > types.py <<'EOF2'
import math
from decimal import Decimal, DivisionByZero
from fractions import Fraction

print(issubclass(DivisionByZero, ZeroDivisionError))
for label, f in [("Fraction(1, 0)", lambda: Fraction(1, 0)),
                 ("Decimal(1) / Decimal(0)", lambda: Decimal(1) / Decimal(0)),
                 ("math.fmod(1, 0)", lambda: math.fmod(1, 0))]:
    try:
        f()
    except ZeroDivisionError as e:
        print(f"{label:24} -> {type(e).__name__}")
    except ValueError as e:
        print(f"{label:24} -> ValueError: {e}")
EOF2
python types.py

Output

True
Fraction(1, 0)           -> ZeroDivisionError
Decimal(1) / Decimal(0)  -> DivisionByZero
math.fmod(1, 0)          -> ValueError: math domain error

How to recognise the cause

The traceback shows the line; the question is why the divisor is zero there. Print it, or the expression that produced it, just before the division:

Divisor expressionTypical reason it is zero
len(items)The list, query result or file was empty
count, total, nA counter was never incremented because the loop did not run
end - startTwo timestamps or positions are equal, such as a very fast operation timed with a coarse clock
int(value) or float(value)A missing field was read as 0, or a default of 0 was used
width, height, scaleAn uninitialised size, such as a window or image that has not been laid out yet

Choosing a fix

Guard the divisor when zero is a legitimate case with an obvious answer: an average of nothing can be 0.0 or None, a percentage of zero items can be shown as "n/a". Catch ZeroDivisionError when the division is deep inside a calculation and the caller decides what to show. If zero should never happen, do not hide it: raise a clear error earlier, for example ValueError("no samples"), so the real bug is found where it starts.

Prevention

Validate inputs where they enter the program, and make functions that divide by a length state what they return for an empty input. Tests with an empty list and with a single zero value catch most of these errors before production.

FAQ

Why does Python raise an error instead of returning infinity?

Python's int and float division raise ZeroDivisionError for a zero divisor rather than returning inf or nan. If you want IEEE-style results, use NumPy or check the divisor and return float('inf') yourself.

Can I catch a division by zero from the decimal module the same way?

Yes. With the default context, decimal raises decimal.DivisionByZero, which is a subclass of ZeroDivisionError, so except ZeroDivisionError catches it too.

Does math.fmod(x, 0) raise ZeroDivisionError?

No. math.fmod(1, 0) raises ValueError: math domain error, so code that uses math functions may need to catch ValueError as well.

Sources

Related pages

errors · pythonValueError: invalid literal for int() with base 10Python raises ValueError: invalid literal for int() with base 10 when text is not a whole number: '3.5', '' or '1,000'. Clean the string, or use float() first.errors · pythonKeyError: 'key' in PythonPython raises KeyError when you read a dict key that does not exist. Check the key with in, read it with dict.get(key, default), or fix the spelling or type.errors · pythonTypeError: 'NoneType' object is not subscriptablePython raises this TypeError when you index a value that is None, often the result of list.sort(), dict.get() or re.match(). Find where None comes from.compare · pythonPython is vs ==: identity vs equalityIn Python == compares values and is compares identity (the same object). Use is only for None and sentinel objects. Why 256 is 256 but 257 may not be.