Python isdigit vs isnumeric: Which One Checks Numbers?

Python isdigit vs isnumeric: isdigit() returns True for digit characters, including superscripts like ², while isnumeric() also accepts fractions like ⅕, Roman numerals like Ⅻ and CJK numerals like 三. Neither one accepts '-5', '3.14' or ' 7'. If you want strings that int() can parse, isdecimal() is the closer match.

What is the difference between isdigit() and isnumeric()?

The Python docs define both methods through Unicode properties. str.isdigit() is True when every character has Numeric_Type Digit or Decimal and there is at least one character. str.isnumeric() is True when every character has Numeric_Type Digit, Decimal or Numeric. That extra Numeric group is where fractions, Roman numerals and many CJK numerals live. str.isdecimal() is the narrowest: it only accepts characters in the Unicode Nd category, which are the characters that can form base-10 numbers.

So every string that passes isdecimal() also passes isdigit(), and every string that passes isdigit() also passes isnumeric().

What did a real test show?

We ran this on Python 3.13.5:

tests = ['123', '٠١٢', '²', '⅕', 'Ⅻ', '三', '-5', '3.14', ' 7', '', '１２']
for s in tests:
    print(repr(s), s.isdecimal(), s.isdigit(), s.isnumeric())

The minus sign, the decimal point and the space are not digits, so all three methods reject those strings. An empty string is always False.

Why can int() fail when isdigit() is True?

int() parses base-10 digits from the Unicode Nd category, plus an optional sign and surrounding whitespace. A superscript two passes isdigit() but is not an Nd character, so int() rejects it. In our run, int('²') raised ValueError: invalid literal for int() with base 10: '²'. Meanwhile int('٠١٢') and int('１２') both returned 12, because those are decimal digits in other scripts.

That is the main trap with isdigit(): it can say yes to a string that int() then refuses.

How do you check for negative numbers or decimals?

None of the three methods handles a sign, a decimal point or spaces. For real input, try the conversion and catch the error:

def to_int(s):
    try:
        return int(s)
    except ValueError:
        return None

In our run, to_int('-5') returned -5, to_int(' 7') returned 7, and to_int('3.14') and to_int('²') returned None. Use float() the same way when decimals are allowed.

How do you accept only ASCII 0 to 9?

Combine two checks: s.isascii() and s.isdigit(). In our run that returned True for '12' and False for '١٢', which is useful when a form field or ID should only contain plain ASCII digits.

What does unicodedata say about these characters?

The unicodedata module exposes the values behind these methods. unicodedata.numeric('⅕') returned 0.2 and unicodedata.digit('²') returned 2 in our run. That is a good way to see why a character passes one method and fails another.

FAQ

Which should I use to validate a number typed by a user?Usually neither. Call int() or float() inside try/except. If you only need to know whether int() will accept a string of plain digits, isdecimal() is the better match than isdigit().

Does bytes have isnumeric()?No. bytes has isdigit(), which checks ASCII 0 to 9, but it has no isnumeric() or isdecimal(). In our run b'123'.isdigit() returned True.

Is isnumeric() True for '½'?Yes. Vulgar fractions have Numeric_Type=Numeric, the same as '⅕' in the table.

Related

• removeprefix vs lstrip in Python
• str.translate vs replace
• Python re vs str methods

Sources

• 01docs.python.org/3/library/stdtypes.html
• 02docs.python.org/3/library/functions.html
• 03docs.python.org/3/library/unicodedata.html