[Numpy-discussion] reshape 2D array into 3D

[Sebastian Berg]

On Mon, 2017-07-10 at 16:16 +0300, eat wrote:
> Hi,
> 
> On Mon, Jul 10, 2017 at 3:20 PM, <paul.carrico at free.fr> wrote:
> > Dear All
> > I'm looking in a way to reshape a 2D matrix into a 3D one ; in my
> > example I want to move the columns from the 4th to the 8th in the
> > 2nd plane  (3rd dimension i guess)
> > a =  np.random.rand(5,8); print(a)
> > I tried
> > a = p.reshape(d, (2,5,4), ) but it is not what I'm expecting
> > 
> > Nota : it looks like the following task (while I want to split it
> > in 2 levels and not in 4), but I've not understood at all
> > https://stackoverflow.com/questions/31686989/numpy-reshape-and-part
> > ition-2d-array-to-3d
> > 
> 
> Is this what you are looking for: 
> import numpy as np
> 
> a= np.arange(40).reshape(5, 8)
> 
> a
> Out[]: 
> array([[ 0,  1,  2,  3,  4,  5,  6,  7],
>        [ 8,  9, 10, 11, 12, 13, 14, 15],
>        [16, 17, 18, 19, 20, 21, 22, 23],
>        [24, 25, 26, 27, 28, 29, 30, 31],
>        [32, 33, 34, 35, 36, 37, 38, 39]])
> 
> np.lib.stride_tricks.as_strided(a, (2, 5, 4), (16, 32, 4))
> Out[]: 
> array([[[ 0,  1,  2,  3],
>         [ 8,  9, 10, 11],
>         [16, 17, 18, 19],
>         [24, 25, 26, 27],
>         [32, 33, 34, 35]],
> 
>        [[ 4,  5,  6,  7],
>         [12, 13, 14, 15],
>         [20, 21, 22, 23],
>         [28, 29, 30, 31],
>         [36, 37, 38, 39]]])
> 

While maybe what he wants, I would avoid stride tricks if you can
achieve the same thing with a reshape + transpose. Far more safe if you
hardcode the strides, and much shorter if you don't, plus easier to
read usually.

One thing some people might get confused about with reshape is the
order, numpy reshape defaults to C-order, while other packages may use
fortran order for reshaping, you can actually change the order you want
to use (though it is in general a good idea to prefer C-order in numpy
probably).

- Sebastian


> Regards,
> -eat
> > Thanks for your support
> > 
> > Paul
> > 
> > _______________________________________________
> > NumPy-Discussion mailing list
> > NumPy-Discussion at python.org
> > https://mail.python.org/mailman/listinfo/numpy-discussion
> > 
> 
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion at python.org
> https://mail.python.org/mailman/listinfo/numpy-discussion
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: This is a digitally signed message part
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20170710/41bff162/attachment.sig>