How to calculate sum using list comprehension

Question

How do I get the cumulative sum of this list using list comprehension:

list_comp=[1,4,9,16] 

Here's what I tried but it prints the double of each item in the list

print([x+x for x in list_comp])

I expect the results to be: list_comp=[1,5,14,30] But I'm getting this:

>> [2, 8, 18, 32]

Solution 1

List comprehensions are for mapping or filtering of lists. They cannot have an internal state, which is what you would need to do this efficiently.

However in Python 3 there is itertools.accumulate for that:

import itertools

print(list(itertools.accumulate([1, 4, 9, 16])))  # [1,5,14,30]

Solution 2

You can do it like this, by combining slicing and a list comprehension, but it is ungainly to do a cumulative sum efficiently.

comp=[1,4,9,16] 
[sum(comp[:idx+1]) for idx in range(len(comp))]

I would not recommend using this, it recalculates the sum n times!

---

A proper way to cumsum can be done like this:

def cumsum(seq):
    cumulative = [seq[0]]
    for elt in seq[1:]:
        cumulative.append(cumulative[-1] + elt)
    return cumulative

---

itertools.accumulate is another way that @NielsWerner demonstrated.

---

Further efficiency can be found with the numpy library, and the cumsum function of this library.

Solution 3

the walrus operator can help to solve this task:

lst = [1,4,9,16] # expected result: [1,5,14,30] 
buf = 0
res = [buf := buf + num for num in lst]
print(res)

[Note : walrus operator is introduced in python 3.8]

Comments

[Chris_Rands]

Possible duplicate of How to find the cumulative sum of numbers in a list?

[Manjunath]

I think accumulate is introduced in Python 3?

[Nils Werner]

Yes, I will add that. But the question targets Python 3 anyways.

[Pratik Charwad]

walrus operator is introduced in python 3.8