Python : Array group by Sum value

Question

Many topics exist to sum an array based on another array indexes but i couldn't find what i am looking for.

I have an array [1,1,1,1,2,3,3,0,0,2,4] and i would like to get indexes by which the sum until now is greater or equal to 4. So I would get:

[3, 5, 9, 10] because:

Sum([1,1,1,1]) >= 4

Sum([2,3]) >= 4

Sum([3,0,0,2]) >= 4

I need the indexes of the element at which the sum value is fulfilled.

I could do it with a loop but i am looking for doing it efficiently with numpy or the like.

What if the sum for the current group never reaches exactly 4, and the next item takes it over 4? Are the items guaranteed to be integers, or could there be floats? — PM 2Ring
– PM 2Ring, Commented Nov 8, 2017 at 10:52
Right, i updated the question, i was not clear enough, i need the slicing based on the first index where the cumsum passed the sum value i am looking for. — isaac.hazan
– isaac.hazan, Commented Nov 8, 2017 at 11:08
Think you should update with a more relevant sample. Maybe like : [1, 1, 3, 1, 2, 2, 3, 0, 0, 1, 4], so that the answers won't fall prey to the simplistic sample. — Divakar
– Divakar, Commented Nov 8, 2017 at 11:11

DeepSpace · Accepted Answer · 2017-11-08 11:06:22Z

2

You can use np.where in conjunction with np.cumsum:

import numpy as np

li = [1,1,1,1,2,2,3,0,0,1,4]

print(np.where(np.cumsum(li) % 4 == 0))
# (array([ 3,  5,  9, 10], dtype=int32),)

The gotcha here of course is that it will also find indexes where the sum is 8, 12, 16, 20 etc.

Unfortunately I'm not sure there is way to "reset" cumsum, if there was you could do something like np.where(np.cumsum(li) - 4 == 0) (but currently it will only return 3).

However, you could slice the array every time cumsum reaches 4, but you didn't want to use loops.

edited Nov 8, 2017 at 11:06

answered Nov 8, 2017 at 10:53

DeepSpace

82.1k12 gold badges119 silver badges166 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Divakar Over a year ago

What if it's [1, 1, 3, 1, 2, 2, 3, 0, 0, 1, 4]? As an example of what PM 2Ring commented about.

Pulkit · Accepted Answer · 2017-11-08 11:03:55Z

1

array =[1,1,1,1,2,2,3,0,0,1,4]
print(np.where(np.cumsum(array)%4==0))

now, if you change your array then you can check whenever sum goes beyond 4 then u can take the last index and proceed further.

answered Nov 8, 2017 at 11:03

Pulkit

7161 gold badge9 silver badges19 bronze badges

Collectives™ on Stack Overflow

Python : Array group by Sum value

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related