Sum a List with Different Operator in Python

Question

In python we can sum a list as: sum(list_of_integers).

Now sum is just an operation among two elements with the operator +.

What if I want to sum a list with different operator like or, and, xor etc?

I could do it manually one by one using a for loop, but there must be a better way.

I think he means if you have a list [x1, x2, x3] then he wants to do `x1 or x2 or x3` without using a manual loop over all elements — FlyingTeller, Feb 15 '18 at 07:54
Use `functools.reduce`, but some operations have other builtins. Example, `all` -> `and` and `any` -> `or`. — cs95, Feb 15 '18 at 07:56
An example for XOR: `from functools import reduce reduce(lambda x,y:x^y, l)`. Is this what you want? — Sohaib Farooqi, Feb 15 '18 at 07:56
@bro-grammer reduce is not built in in python NameError: name 'reduce' is not defined — mourinho, Feb 15 '18 at 08:00
@mourinho It definitely is, i tested python2.6, 2.7 and 3 and they all work — jamylak, Feb 15 '18 at 08:42

score 9 · Accepted Answer · answered Feb 15 '18 at 08:00

9

functools.reduce is perfect for this use-case. It takes a function to apply to the accumulated value and the next value, the iterable you want to reduce, and optionally an initialiser.

For example, bitwise-or-ing every value in a list:

import functools

functools.reduce(lambda a, b: a ^ b, [1, 2, 3])

This is equivalent to 1 ^ 2 ^ 3.

answered Feb 15 '18 at 08:00

Christian Scott

893
7
19

but reduce is not built-in in python anyway without importing any library..thanks – mourinho Feb 15 '18 at 08:02
3

Not using the standard library seems like an extremely arbitrary restriction. Is this for homework? – Christian Scott Feb 15 '18 at 08:06
5

@mourinho `reduce` *used* to be part of the `builtins` module in Python 2, but it was relegated to `functools` because, well, [Guido just doesn't like it](https://stackoverflow.com/questions/181543/what-is-the-problem-with-reduce). Not wanting to use core libraries seems like an artificial restriction. Otherwise, note, Guido recommends a for-loop. And that seems like the most straightforward way if you don't want `reduce` – juanpa.arrivillaga Feb 15 '18 at 08:09
@mourinho Then you should've mentioned that in the question. There's no reason not to use the standard library – jamylak Feb 15 '18 at 08:40

Graipher · Answer 2 · 2018-03-03T20:27:48.027

The alternative to functools.reduce is to write an explicit for loop:

def xor_reduce(args):
    result = 0
    for x in args:
        result ^= x
    return result

xor_reduce([1, 2, 3])

If you are going for the reduce way (not so unreasonable for this, IMO), I would make use of the operator module:

from functools import reduce
from operator import xor

reduce(xor, [1, 2, 3])

The operator module (which is in the standard library and should therefore always be available) also defines all other standard operations as functions, but for or and and a trailing _ is added because they are reserved keywords:

from operator import or_, and_
reduce(or_, [1, 2, 3])
reduce(and_, [1, 2, 3])

Although for these two you could use the built-in functions any and all:

any([1, 2, 3])
all([1, 2, 3])

This is much better than the accepted answer, showing a manual and the `reduce` way to do it. Absolutely no need for a `lambda`. Faster and more elegant — jamylak, Feb 15 '18 at 08:41

Sum a List with Different Operator in Python

2 Answers2