CODEWITHSUNDEEP
pythonnumpysumslicenumpy-ndarray
tony selcuk
tony selcuk

Reputation: 687

Sum function using slices with a numpy array Python

Is there a way I could index through a numpy list just like how I would be able to within a normal list function. I want to go through 3 elements in the list moving up by one point every single time and summing all the slices. So it would go through 1,2,3 for the first sum and then it would go through 2,3,4 for the second sum etc. The code down below gives me a scalar error, is there a way I could perform the function below without using a for loop.

import numpy as np
n = 3 
arr = np.array([1, 2, 3, 4, 5, 6, 7, 8,9, 10, 11,12, 13, 14, 15, 16, 17, 18, 19, 20, 21 ,22, 23, 24, 25])
start = np.arange(0, len(arr)-n, 1)
stop = np.arange(n-1, len(arr), 1)
sum_arr = np.sum(arr[start:stop])

Upvotes: 0

Views: 460

Answers (2)

Stef
Stef

Reputation: 30579

To sum a sliding range of n elements you can use convolve1d with all weights set to 1. Use 'constant' boundary mode with the default fill value of 0. As the filter window is centered by default you need to adjust the length of the result at both ends.

import numpy as np 
from scipy.ndimage import convolve1d

arr = np.arange(1,26)

for n in range(2,6):
   k,r = divmod(n, 2)
   print(n, convolve1d(arr, np.ones(n), mode='constant')[k+r-1:-k])

Result:

2 [ 3  5  7  9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49]
3 [ 6  9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72]
4 [ 10 14 18 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 86 90 94]
5 [ 15  20  25  30  35  40  45  50  55  60  65  70  75  80  85  90  95 100 105 110 115]

Upvotes: 0

davidmpeletz
davidmpeletz

Reputation: 46

I think this should work:

sum_arr = arr[1:-1] + arr[2:] + arr[:-2]

This creates an array that's two values shorter than arr because the last element in arr doesn't have two extra elements to create a sum with.

If you wanted the array to be of the same length as the original arr, you could append two extra zeros to the arr array like so:

arr = np.append(arr, [0, 0])
sum_arr = arr[1:-1] + arr[2:] + arr[:-2]

Upvotes: 1

Related Questions