累加列表中具有整数的所有元素的总和

Question

I'm trying to get the sum of all numbers for each list through itertool accumulate function: This is the code:

from itertools import *
import operator

list_numbers = [[45,43,56,67],[41,423,526,627,897],[545,453,656,667]]

Iterable_list = iter(list_numbers)

while(True):
    try:
        print(list(accumulate(next(Iterable_list),operator.add))[-1])
    except StopIteration as ex:
        print(ex)
        break

The above works well.

if I try to take out the max of each list:

while(True):
    try:
        print(list(accumulate(next(Iterable_list),func=max))[-1])
    except StopIteration as ex:
        print(ex)
        break

this also works well.

My interpretation was, if I use func = sum , it will do the sum of each list like I tried below but it prompts error message:

'int' object is not iterable


while(True):
    try:
        print(list(accumulate(next(Iterable_list),func=sum))[-1])
    except StopIteration as ex:
        print(ex)
        break

Why it doesn't it work and in which scenario can we use func = sum ?

Answer 1

You could understand better the behavior of itertools.accumulate if you use a lambda function to explore the inputs that func will receive.

Let's use lambda x: print(x) instead of sum first, to print the values:

TypeError: <lambda>() takes 1 positional argument but 2 were given

Ok, so func will receive two values. Let's try lambda x, y: print(x, y) :

45 43
None 56
None 67
None
41 423
None 526
None 627
None 897
None
545 453
None 656
None 667
None

Why are we seeing None ? Because print() does not return any value, so it returns a None .

This means that func should receive two values from the iterator (L1 and L2), apply an operation on both values and return a single value C. Then, apply the same operation on the previous result (C) with the next value in the iterator (L3).

---

Why sum() fails?

Well, because it expects a single argument, an iterable, and optionally accepts a second ( start ). Now, itertools.accumulate is passing two values and Python assumes you're trying to use sum(L1, start=L2) , but L1 is not iterable since its just a number... TypeError: 'int' object is not iterable .

In which scenario can we use func=sum ?

sum() could help us if it could receive two values and return a single one, which is possible ( sum(iterable, start=start) ). Now, consider that the output of sum will be used as the first input of sum with the next element in the iterable, something like: sum(sum(L1, start=L2), start=L3) and so on. This means that sum(L1, start=L2) should return an iterable for us to be able to use sum . It's hard to find an example for this, meaning that sum might not be the best option for any particular case when using itertools.accumulate , but this doesn't mean that sum is not useful by itself.

Answer 2

This can be done using Zip() or map() method or using numpy.

Using zip():

def nested_sum(lst):
      
 return [sum(i) for i in zip(*lst)]
      
lst = [[2, 4, 6], [2, 2, 1], [1, 2, 3]] # A Sample List
print(nested_sum(lst))

#Output
[12,5,6]