Append Python 3D Numpy Array

Question

I have some problem about 3D python numpy

import numpy as np
n = 5
m = 4

Sc = np.random.rand(m,n,n)
S1 = np.zeros((1,n+2))
S2 = np.zeros((n,1))

A0 = np.r_[S1, np.c_[S2, Sc[0], S2], S1]
A1 = np.r_[S1, np.c_[S2, Sc[1], S2], S1]
#print(A)
#print(B)
A = np.array([A0,A1])
A.shape
Atmp = np.r_[S1, np.c_[S2, Sc[2], S2], S1]

Dimension of A = (2, 7, 7)

and dimension of Atmp = (7,7).

How to append Atmp to A ?

Atmp + A doesn't give any errors for me. Maybe A += Atmp? You need to clarify exactly what you want. — Daniel F
– Daniel F, Commented Aug 7, 2017 at 8:22
A = np.array((A0, A1, Atmp)) or A = np.vstack((A, Atmp[None,...])) — gboffi
– gboffi, Commented Aug 7, 2017 at 9:05
np.append is another way of usingnp.concatenate, and often a confusing one. np.r_ and np.c_ are also concatenate frontends. — hpaulj
– hpaulj, Commented Aug 7, 2017 at 9:07

Daniel F · Accepted Answer · 2017-08-07 08:38:40Z

4

Don't concatenate/append/stack arrays if you can help it, especially big ones. It's very wasteful of memory and slow.

Assign A = np.empty((m, n+2, n+2)) and then fill it with A[i] = np.r_[S1, np.c_[S2, Sc[i], S2], S1]. Or do it vectorized and get rid of the for loops:

A = np.zeros((m, n+2, n+2))
A[:,1:-1,1:-1] = Sc

or even do it in one line:

A = np.pad(Sc, ((0,0),(1,1),(1,1)), 'constant', constant_values = 0)

edited Aug 7, 2017 at 8:38

answered Aug 7, 2017 at 8:33

Daniel F

14.5k2 gold badges34 silver badges59 bronze badges

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

3 Comments

Rio Harapan Pangihutan Over a year ago

Thank you for your answer, It's very helpful for my problem. And this is a big dimension array problem, can you suggest some reading for this method?

Daniel F Over a year ago

indexing docs and np.pad

Rio Harapan Pangihutan Over a year ago

I think this answer not only give me the answer of the problem but also new way to use big problem array in python. Thakn you very much and I vote this for the acceptance answer

AGN Gazer · Accepted Answer · 2017-08-07 08:32:28Z

2

You can try this:

A = np.concatenate([A, [Atmp]])

answered Aug 7, 2017 at 8:32

AGN Gazer

8,4272 gold badges31 silver badges49 bronze badges

1 Comment

Rio Harapan Pangihutan Over a year ago

Thank you for your answer, I think, this answer work for my problem.

Landmaster · Accepted Answer · 2017-08-07 08:32:24Z

1

You can use reshape to put the array in the right form:

np.reshape(Atmp,(1, Atmp.shape[0], Atmp.shape[1]))

and then append as

np.vstack([A, np.reshape(Atmp,(1, 7, 7))])

answered Aug 7, 2017 at 8:32

Landmaster

1,0532 gold badges13 silver badges22 bronze badges

Collectives™ on Stack Overflow

Append Python 3D Numpy Array

3 Answers 3

3 Comments

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related