How to efficiently shuffle numpy array in chunks

Question

I have a numpy array that looks like the following [-1,0,1,0,1,2,1,2,3,...,n-1,n,n+1,n,n+1,n+2..] I would like to shuffle the array in chunks of 3, is there an efficient way to do it in numpy?

I know you can shuffle a numpy array using the following shuffle method, but this gives me a fully shuffled array. Is there a way to shuffle it in chunks in numpy?

import numpy.random as rng

ind = numpy.arange(100)
rng = numpy.random.RandomState(123)
rng.shuffle(ind)

ErikR · Accepted Answer · 2015-09-10 02:54:34Z

8

Reshape into 3 columns. shuffle doc says it just shuffles the 1st dimension:

ind=np.arange(99)  # multple of 3
ind=ind.reshape(-1,3)
rng.shuffle(ind)
ind.flatten()

edited Sep 10, 2015 at 2:54

ErikR

52.2k9 gold badges78 silver badges136 bronze badges

answered Sep 10, 2015 at 2:33

hpaulj

233k14 gold badges260 silver badges392 bronze badges

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

2 Comments

Guillem Cucurull Over a year ago

To keep the array flattened after the process, shouldn't the last operation be ind = ind.flatten()?

hpaulj Over a year ago

Right. flatten isn't an inplace operation. I adapted the code snippet from an interactive session, where I got an automatic print of the last line.

Collectives™ on Stack Overflow

How to efficiently shuffle numpy array in chunks

1 Answer 1

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related