El dl 11 de 12 del 2006 a les 14:16 +0100, en/na Alexandre Fayolle va
escriure:
>  > > I can work around this by using an intermediate string representation:
>  > > 
>  > > temp = Numeric.fromstring(atest.tostring(), atest.typecode())
>  > > temp.shape = atest.shape
>  > 
>  > Another (faster) workaround would be:
>  > 
>  > temp2 = Numeric.fromstring(atest._data, typecode=atest.typecode())
>  
>  Nice! 

Well, I've to say that this approach only work for contiguous,
non-offseted arrays, as can be seen in:

In [59]:atest = numarray.arange(10)
In [60]:Numeric.fromstring(atest[5:]._data, typecode=atest.typecode())
Out[60]:array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9],'i')  # wrong!
In [61]:Numeric.fromstring(atest[5:].tostring(), atest.typecode())
Out[61]:array([5, 6, 7, 8, 9],'i')                 # good
In [62]:Numeric.fromstring(atest[::2]._data, typecode=atest.typecode())
Out[62]:array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9],'i')  # wrong!
In [63]:Numeric.fromstring(atest[::2].tostring(), atest.typecode())
Out[63]:array([0, 2, 4, 6, 8],'i')                 # good

So, be careful when using it. I'd rather keep using your approach, which
is the faster one that is completely general.

-- 
Francesc Altet    |  Be careful about using the following code --
Carabos Coop. V.  |  I've only proven that it works, 
www.carabos.com   |  I haven't tested it. -- Donald Knuth

_______________________________________________
Numpy-discussion mailing list
Numpy-discussion@[...].org
http://projects.scipy.org/mailman/listinfo/numpy-discussion