how to filter an numpy.ndarray by a spevific number?

Question

roi_pixel_img = crop_img[indices_list]
print (roi_pixel_img)

when i add (I use only to use the entire array (meaning only a part):

np.set_printoptions(threshold=sys.maxsize)

th output is:

the whole part happens in a while loop because I'm extracting pixels in this section, which is irrelevant to the question.

My goal is not to include the lines with [0 255 255] in this array, how can I do that?

the type of roi_pixel_img is numpy.ndarray.

is it even possible to answer this question without an example code for you ?

its variable, for example (6730, 3). i mean the only variable part that is the first, fixed is 3 — mika
– mika, Commented Apr 27, 2022 at 17:31

Professor Pantsless · Accepted Answer · 2022-04-28 13:04:11Z

1

You can do this by creating an indexing array:

r = (roi_pixel_img == [0,255,255]).all(axis = -1)

roi_pixel_img[~r]

The roi_pixel_img == [0,255,255] statement will result in an array with the same shape as roi_pixel_img (say (N, 3)) and will compare element-wise, eg [0,255,0] will result in [True, True, False]. Using .all(axis = -1) Will reduce along the last axis (in this case axis = 1 would produce the same result) and will result in True if all the element match. So r will have shape (N, ).

Using ~r to index will exclude the matching pixels and due to the shape will be broadcast appropriately by numpy.

edited Apr 28, 2022 at 13:04

answered Apr 27, 2022 at 19:40

Professor Pantsless

4844 silver badges6 bronze badges

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

1 Comment

mika Over a year ago

works! could you explain why do you use .all on axis ?

Collectives™ on Stack Overflow

how to filter an numpy.ndarray by a spevific number?

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related