Difference between elements of subsequent pairs in numpy arrays (not numpy.diff)

Question

Is there quick way to calculate the difference between two elements in subsequent pairs in python arrays? For example, consider x:

x = np.array([1,5,3,8])

How can I calculate from x the differences between subsequent pairs? My desired output is:

np.array([4,5])

cs95 · Accepted Answer · 2017-12-12 14:59:09Z

4

You can slice in strides of 2 and subtract:

>>> x[1::2] - x[::2]
array([4, 5])

Another solution is to reshape and call np.diff:

>>> np.diff(x.reshape(-1, 2), axis=1).ravel()     
array([4, 5])

A generalised version of this that works for any N * M array would look something like this -

r = np.diff(x.reshape(-1, 2), axis=1).reshape(-1, x.shape[1] // 2)

edited Dec 12, 2017 at 14:59

answered Dec 12, 2017 at 14:00

cs95

406k106 gold badges744 silver badges794 bronze badges

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

4 Comments

user7740248 Over a year ago

Hey thanks for editing my comment to make it more legible. Your answer does provide the solution so thanks! However it does work if you scale it to any arbitrary 2 dimensional array with an even number of rows. For example if I want to to this for a N by M array where M is even. For example: x = np.array([[1,5,3,8],[1,2,3,4]]) with desired output: array([[4,5],[1,1]])

cs95 Over a year ago

@user7740248 Are you asking or are you telling?

user7740248 Over a year ago

Sorry I posted my response on accident I just edited for a follow up question

cs95 Over a year ago

@user7740248 Okay. For any N * M array where M is even, this will work: np.diff(x.reshape(-1, 2), axis=1).reshape(-1, x.shape[1] // 2).

Collectives™ on Stack Overflow

Difference between elements of subsequent pairs in numpy arrays (not numpy.diff)

1 Answer 1

4 Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Related