Structured numpy array not modified in place

Question

I have a structured numpy array which I'm trying to modify in-place, but the new values are not reflected.

import numpy as np

dt = {'names':['A', 'B', 'C'],
        'formats': [np.int64, np.int64, np.dtype('U8')]}
arr = np.empty(0, dtype=dt)

arr = np.append(arr, np.array([(1, 100, 'ab')], dtype = dt))
arr = np.append(arr, np.array([(2, 800, 'ax')], dtype = dt))
arr = np.append(arr, np.array([(3, 700, 'asb')], dtype = dt))
arr = np.append(arr, np.array([(4, 600, 'gdf')], dtype = dt))
arr = np.append(arr, np.array([(5, 500, 'hfg')], dtype = dt))

print(arr)

arr[arr['A'] == 1]['B'] = 555

print(arr)

Is it even possible to change values in structured array? What could be the workaround?

Please don't suggest Pandas or other library based solution since I'm only allowed to use numpy at work.

As explenation you're creating a copy which you then modify. — Daraan
– Daraan, Commented Oct 12, 2022 at 13:07

Florent Monin · Accepted Answer · 2022-10-12 13:43:56Z

5

You can reverse the indexing you're doing: first select the field you're interested in, which returns a view on the original array (as described in the documentation), as a normal numpy array, that you can modify:

import numpy as np

dt = {'names':['A', 'B', 'C'],
        'formats': [np.int64, np.int64, np.dtype('U8')]}
arr = np.empty(0, dtype=dt)

arr = np.append(arr, np.array([(1, 100, 'ab')], dtype = dt))
arr = np.append(arr, np.array([(2, 800, 'ax')], dtype = dt))
arr = np.append(arr, np.array([(3, 700, 'asb')], dtype = dt))
arr = np.append(arr, np.array([(4, 600, 'gdf')], dtype = dt))
arr = np.append(arr, np.array([(5, 500, 'hfg')], dtype = dt))

print(arr)

arr['B'][arr['A'] == 1] = 555

print(arr)

edited Oct 12, 2022 at 13:43

answered Oct 12, 2022 at 13:03

Florent Monin

1,3421 gold badge7 silver badges17 bronze badges

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

Collectives™ on Stack Overflow

Structured numpy array not modified in place

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related