[Numpy-discussion] array indexing in numarray

Joost van Evert joostvanevert at gmail.com
Fri Sep 23 01:33:43 CDT 2005


Hi list,

I have two questions on numarray indexing:
1) does anyone know an elegant way to assign values to a non-contiguous
slice. Like:

        In [1]:from numarray import *
        
        In [2]:a = arange(25).resize((5,5))
        
        In [3]:a
        Out[3]:
        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]])
        
        In [4]:i = [1,3]
        
        In [5]:transpose(a[i])[i]=1

	----------------------------------------------------------------
        exceptions.ValueError                                Traceback
        (most recent call last)
        
        ...
        
        ValueError: Invalid destination array: partial indices require
        contiguous non-byteswapped destination
        
2) And second: why didn't you choose to return all combinations of
indices:

        In [8]:a[i,i]
        Out[8]:array([ 6, 18])

Where I, in my humble opinion, would prefer to see all possible
combinations. e.g. a[[1,3],[1,3]] = a[[1,1,3,3],[1,3,1,3]]

Anyone has an idea how to easyly obtain this behaviour?

Greets+thanks,
Joost




More information about the Numpy-discussion mailing list