What is the most pythonic way to check if an object is a number?

user_15055 0
Viewed 39110

Given an arbitrary python object, what's the best way to determine whether it is a number? Here is is defined as acts like a number in certain circumstances.

For example, say you are writing a vector class. If given another vector, you want to find the dot product. If given a scalar, you want to scale the whole vector.

Checking if something is int, float, long, bool is annoying and doesn't cover user-defined objects that might act like numbers. But, checking for __mul__, for example, isn't good enough because the vector class I just described would define __mul__, but it wouldn't be the kind of number I want.

numbers python types
15 Answers

To summarize / evaluate existing methods:

Candidate    | type                      | delnan | mat | shrewmouse | ant6n
-------------------------------------------------------------------------
0            | <type 'int'>              |      1 |   1 |          1 |     1
0.0          | <type 'float'>            |      1 |   1 |          1 |     1
0j           | <type 'complex'>          |      1 |   1 |          1 |     0
Decimal('0') | <class 'decimal.Decimal'> |      1 |   0 |          1 |     1
True         | <type 'bool'>             |      1 |   1 |          1 |     1
False        | <type 'bool'>             |      1 |   1 |          1 |     1
''           | <type 'str'>              |      0 |   0 |          0 |     0
None         | <type 'NoneType'>         |      0 |   0 |          0 |     0
'0'          | <type 'str'>              |      0 |   0 |          0 |     1
'1'          | <type 'str'>              |      0 |   0 |          0 |     1
[]           | <type 'list'>             |      0 |   0 |          0 |     0
[1]          | <type 'list'>             |      0 |   0 |          0 |     0
[1, 2]       | <type 'list'>             |      0 |   0 |          0 |     0
(1,)         | <type 'tuple'>            |      0 |   0 |          0 |     0
(1, 2)       | <type 'tuple'>            |      0 |   0 |          0 |     0

(I came here by this question)

Code

#!/usr/bin/env python

"""Check if a variable is a number."""

import decimal


def delnan_is_number(candidate):
    import numbers
    return isinstance(candidate, numbers.Number)


def mat_is_number(candidate):
    return isinstance(candidate, (int, long, float, complex))


def shrewmouse_is_number(candidate):
    try:
        return 0 == candidate * 0
    except:
        return False


def ant6n_is_number(candidate):
    try:
        float(candidate)
        return True
    except:
        return False

# Test
candidates = (0, 0.0, 0j, decimal.Decimal(0),
              True, False, '', None, '0', '1', [], [1], [1, 2], (1, ), (1, 2))

methods = [delnan_is_number, mat_is_number, shrewmouse_is_number, ant6n_is_number]

print("Candidate    | type                      | delnan | mat | shrewmouse | ant6n")
print("-------------------------------------------------------------------------")
for candidate in candidates:
    results = [m(candidate) for m in methods]
    print("{:<12} | {:<25} | {:>6} | {:>3} | {:>10} | {:>5}"
          .format(repr(candidate), type(candidate), *results))
user_5627690

can be implemented in a simple try exception block

def check_if_number(str1):
    try:
        int(float(str1))
        return 'number'
    except:
        return 'not a number'

a = check_if_number('32322')
print (a)
# number
user_34649710

If you want to call different methods depending on the argument type(s), look into multipledispatch.

For example, say you are writing a vector class. If given another vector, you want to find the dot product. If given a scalar, you want to scale the whole vector.

from multipledispatch import dispatch

class Vector(list):

    @dispatch(object)
    def __mul__(self, scalar):
        return Vector( x*scalar for x in self)

    @dispatch(list)
    def __mul__(self, other):
        return sum(x*y for x,y in zip(self, other))


>>> Vector([1,2,3]) * Vector([2,4,5])   # Vector time Vector is dot product
25
>>> Vector([1,2,3]) * 2                 # Vector times scalar is scaling
[2, 4, 6]

Unfortunately, (to my knowledge) we can't write @dispatch(Vector) since we are still defining the type Vector, so that type name is not yet defined. Instead, I'm using the base type list, which allows you to even find the dot product of a Vector and a list.

user_36900240

Short and simple way :

obj = 12345
print(isinstance(obj,int))

Output :

True

If the object is a string, 'False' will be returned :

obj = 'some string'
print(isinstance(obj,int))

Output :

False
user_83801730

You have a data item, say rec_day that when written to a file will be a float. But during program processing it can be either float, int or str type (the str is used when initializing a new record and contains a dummy flag value).

You can then check to see if you have a number with this

                type(rec_day) != str

I've structured a python program this way and just put in 'maintenance patch' using this as a numeric check. Is it the Pythonic way? Most likely not since I used to program in COBOL.

user_89298140
Related