I am trying to do an operation relative to the location of the element during an elementwise numpy operation. I cannot figure out how access the current location in order to sum it with the next three elements. I have included a sample array as well as a sample solution I am trying to achieve. Since the real array is over 1MM rows a for loop would be too slow. Thank you!
import numpy as np
k = np.random.rand(10,1)*10
k.astype(int)
Try np.convolve
:
N = 3
x = np.convolve(k.ravel(), np.ones(N, dtype=int), "full")[N - 1 :].reshape(-1, 1)
print(x)
Prints:
[[17]
[14]
[ 5]
[14]
[14]
[15]
[11]
[14]
[13]
[ 8]]
The input array was:
[[8]
[9]
[0]
[5]
[0]
[9]
[5]
[1]
[5]
[8]]