[Numpy-discussion] when and where to use numpy arrays vs nested lists

Francesc Altet faltet@carabos....
Thu Mar 1 13:53:35 CST 2007


El dj 01 de 03 del 2007 a les 12:47 -0700, en/na Mark P. Miller va
escriure:
>  > try with
>  >
>  >>>> array1.itemset((5,5),9)
> 
> 
> Yep...tried that. But I don't get it!
> 
>  >>> import numpy
>  >>> array1 = numpy.zeros((10,10),int)
>  >>> array1
> array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])
>  >>> array1.itemset((5,5),9)
>  >>> array1
> array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 9],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
>         [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])
>  >>> array1.item(5,5)
> 0
>  >>> array1.item(5,9)
> 9
> 

Ops, this seems a bug with your numpy version:

In [46]:array1 = numpy.zeros((10,10),int)
In [47]:array1.itemset((5,5),9)
In [48]:array1
Out[48]:
array([[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 9, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])
In [49]:array1.item(5,5)
Out[49]:9
In [50]:numpy.__version__
Out[50]:'1.0.2.dev3546'

So, this seems solved in SVN.

-- 
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



More information about the Numpy-discussion mailing list