How to create a numpy array filled with differences of elements of another array

Question

I have a numpy array X. I need to create another array (say Y) of the same size, which has elements

Y[i] = X[i+1]-X[i-1]

Can I do that without looping over array elements?

What's Y[0] (i-1 = -1)? Can you add some small sample input and desired output? — pault
– pault, Commented Feb 21, 2018 at 18:33
docs.scipy.org/doc/numpy/reference/generated/numpy.diff.html — Osman Mamun
– Osman Mamun, Commented Feb 21, 2018 at 18:44
@pault, Y[0] and Y[-1] will be handled separately (in fact I just assign fixed values to those two elements). — Tajimura
– Tajimura, Commented Feb 21, 2018 at 19:06
@mamun, diff is nice, but it gives differences between neighbouring elements, while what I need is "leapfrog" difference: not between 2nd and 1st elements (or, say 6th and 5th), but rather between 3rd and 1st ones (7th and 5th, so on). — Tajimura
– Tajimura, Commented Feb 21, 2018 at 19:09

Jello · Accepted Answer · 2018-02-21 18:46:26Z

2

You could make new arrays with shifted values and then subtract them from one another. Something like this:

import numpy as np

X  = np.arange(10)
X1 = np.roll(X,-1) #right shift
X2 = np.roll(X,1)  #left shift
Y  = X1 - X2

answered Feb 21, 2018 at 18:46

Jello

4206 silver badges13 bronze badges

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

1 Comment

this one is perfect. Going to use it if there's no builtin method for that thing.