numpy dtype duplicate the array

Question

I try to add a dtype in a numpy array (for the creation after of a ply file), but when I write this:

list_example = np.array([[1,2,3]], dtype=[('example', 'i4', (3,))]

The result is :

array([[([1, 1, 1],), ([2, 2, 2],), ([3, 3, 3],)]],
      dtype=[('example', '<i4', (3,))])

And i don't understand why it's duplicate every components of my array. Because if I do this:

np.array([([1,2,3], 255)], dtype = [('example', 'i4', (3,)), ('color', 'u1')])

I obtain:

array([([1, 2, 3], 255)], dtype=[('example', '<i4', (3,)), ('color', 'u1')])

where my array dont duplicate [1,2,3]

Thanks for your help!

hpaulj · Accepted Answer · 2021-08-26 15:30:21Z

2

When creating structured arrays, the distinction between list and tuple is important. The inputs should match the print display, as your working case shows.

For the first dtype, try:

In [525]: np.zeros(1, dtype=[('example', 'i4', (3,))])
Out[525]: array([([0, 0, 0],)], dtype=[('example', '<i4', (3,))])

imitating that:

In [526]: np.array([([1,2,3],)], dtype=[('example', 'i4', (3,))])
Out[526]: array([([1, 2, 3],)], dtype=[('example', '<i4', (3,))])

np.array([([1,2,3])], dtype=[('example', 'i4', (3,))]) doesn't help because the () include a single list, without a comma, so they do nothing. Your dtype has one field, so the input must have a single element tuple per record.

answered Aug 26, 2021 at 15:30

hpaulj

233k14 gold badges260 silver badges392 bronze badges

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

Comments

Mohammad · Accepted Answer · 2021-08-26 10:57:32Z

1

You can use rec.fromarrays

>>> np.rec.fromarrays([([1,2,3])], dtype=[('example', 'i4', (3,))])

rec.array(([1, 2, 3],),
          dtype=[('example', '<i4', (3,))])

answered Aug 26, 2021 at 10:57

Mohammad

3,4162 gold badges22 silver badges35 bronze badges

Collectives™ on Stack Overflow

numpy dtype duplicate the array

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related