Modify values of a numpy array based on a condition

Question

Hello I have a numpy array as follows:

list_of_values = [-0.5, -0.8, -0.6, -0.2, 0.2, 0.5, 0.7, 0.9, -0.3, 0.2, 0.6, 0.4, 0, 0.2, 0, -0.3, -0.2, 0, -0.6, 0, 0.3]
array= np.array(list_of_values)

I need an array of the same size, having a 1 when the value goes from negative to positive, and a -1 when the value goes from positive to negative, and a 0 for the rest. If there is a 0 in list_of_values, and it bounces back, it should not be considered as the 0 line has been crossed.

This is the exact output that I am looking for

output = [0, 0, 0, 0, 1, 0, 0, 0, -1, 1, 0, 0, 0, 0, 0, -1, 0, 0, 0, 0, 1]

The real list_of_values is large, with 3000+ numbers, so the solution should use Numpy (some for loop is not valid), and it should be quick.

Thank you

Sayandip Dutta · Accepted Answer · 2021-06-08 07:41:21Z

1

You can use numpy.ediff1d:

>>> import numpy as np
>>> arr = np.array([-1, -0.8, -0.6, -0.1, 0.1, 0.5, 0.7, 1, -0.3, 0.2, 0.6])
>>> np.ediff1d(np.sign(arr), to_begin=0) // 2
array([ 0.,  0.,  0.,  0.,  1.,  0.,  0.,  0., -1.,  1.,  0.])

If you'd like to consider 0 as positive:

>>> arr = np.array([-1, -0.8, -0.6, -0.1, 0.0, 0.5, 0.7, 1, -0.3, 0.2, 0.6])
>>> np.ediff1d(np.where(arr==0, 1, np.sign(arr)), to_begin=0)//2
array([ 0.,  0.,  0.,  0.,  1.,  0.,  0.,  0., -1.,  1.,  0.])

EDIT

Based on your edit, you can do the following:

>>> arr = [-0.5, -0.8, -0.6, -0.2, 0.2, 0.5, 0.7, 0.9, -0.3, 0.2, 0.6, 0.4, 0, 0.2, 0, -0.3, -0.2, 0, -0.6, 0, 0.3]
>>> output = np.array(arr)
>>> nonzero_idx = output != 0
>>> nonzero_arr = output[nonzero_idx]
>>> res = np.ediff1d(np.sign(nonzero_arr), to_begin=0) // 2
>>> output[nonzero_idx] = res
>>> output
array([ 0.,  0.,  0.,  0.,  1.,  0.,  0.,  0., -1.,  1.,  0.,  0.,  0.,
        0.,  0., -1.,  0.,  0.,  0.,  0.,  1.])

edited Jun 8, 2021 at 7:41

answered Jun 6, 2021 at 12:41

Sayandip Dutta

15.9k4 gold badges27 silver badges57 bronze badges

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

2 Comments

Arman Mojaver Over a year ago

Thank you for your answer I really appreciate it. What if I wanted to treat 0 as indifferent, meaning that if [..., +, 0, +, ...], the third value would not be -1, and the other way around, if [..., -, 0, -, ....], the third value would not be +1

Sayandip Dutta Over a year ago

Well actually I didn't know what you wanted in case of 0. Could you add an input with 0 and expected output in your question? I can update the answer accordingly.

Terra · Accepted Answer · 2021-06-06 12:55:50Z

0

The answer provided by Cyttorak is correct and a lot prettier than mine, however, if you don't mind a little bit uglier code and want speed, this is faster.

a2 = np.array([-1, 1])
array = np.array([-1, -0.8, -0.6, -0.1, 0.1, 0.5, 0.7, 1, -0.3, 0.2, 0.6])
array = np.where(array >= 0, -1, 1)
output = np.convolve(array, a2, mode='same')
output = np.where((output == 2) | (output == -2), output, 0) // 2

Edit: sorry, forgot to copy-paste a2 :D

answered Jun 6, 2021 at 12:55

Terra

1871 silver badge10 bronze badges

Collectives™ on Stack Overflow

Modify values of a numpy array based on a condition

2 Answers 2

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related