Reduce function that is used for reducing the iterable (must take two arguments). (positional-only). Learn More about it here.

Parameter Details

function function that is used for reducing the iterable (must take two arguments). (positional-only)

iterable iterable that’s going to be reduced. (positional-only)

initializer start-value of the reduction. (optional, positional-only)


No import needed
No import required…
from functools import reduce # … but it can be loaded from the functools module

from functools import reduce # mandatory

reduce reduces an iterable by applying a function repeatedly on the next element of an iterable and the cumulative result so far.

def add(s1, s2):
return s1 + s2
asequence = [1, 2, 3]
reduce(add, asequence) # equivalent to: add(add(1,2),3)

Out: 6

In this example, we defined our own add function. However, Python comes with a standard equivalent function in the operator module:

import operator
reduce(operator.add, asequence)

Out: 6

reduce can also be passed a starting value:

reduce(add, asequence, 10)

Out: 16

Using reduce

def multiply(s1, s2):
print('{arg1} * {arg2} = {res}'.format(arg1=s1,
return s1 * s2
asequence = [1, 2, 3]

Given an initializer the function is started by applying it to the initializer and the first iterable element:

cumprod = reduce(multiply, asequence, 5)

Out: 5 * 1 = 5

52=10 103=30
Out: 30

Without initializer parameter the reduce starts by applying the function to the first two list elements:

cumprod = reduce(multiply, asequence)
Out: 1 * 2 = 2
Out: 6

Cumulative product

import operator
reduce(operator.mul, [10, 5, -3])

Out: -150

Non short-circuit variant of any/all

it will not terminate the iteration before the iterable has been completely iterated over so it can be used to create a non short-circuit any() or all() function:

import operator

non short-circuit “all”

reduce(operator.and_, [False, True, True, True]) # = False

non short-circuit “any”

reduce(operator.or_, [True, False, False, False]) # = True

Learn More


Please enter your comment!
Please enter your name here