[Numpy-discussion] Fancy indexing question:

[Christopher Barker]

HI all,

I'm having a bit of trouble getting fancy indexing to do what I want.

Say I have a 2-d array:

 >>> a
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]])

I want to extract a sub-array:

The 1st, 3rd, and 4th rows:
 >>> i
[1, 3, 4]

and the 1st and 3rd columns:
 >>> j
[1, 3]

so I should get a 3x2 array:

[[ 5,  7],
  [13, 15],
  [17, 19]]

The obvious (to me!) way to do this:

 >>> a[i,j]
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
ValueError: shape mismatch: objects cannot be broadcast to a single shape

fails.

I can do:

extract the rows:

 >>> rows = a[i]
 >>> rows
array([[ 4,  5,  6,  7],
        [12, 13, 14, 15],
        [16, 17, 18, 19]])

then the columns:

 >>> sub_array = rows[:,j]
 >>> sub_array
array([[ 5,  7],
        [13, 15],
        [17, 19]])

or, of course:
 >>> sub_array = a[i][:,j]
 >>> sub_array
array([[ 5,  7],
        [13, 15],
        [17, 19]])


however, it seems I am doing one more data copy and array creation than 
I want -- can this be done as a single indexing operation?

thanks,

Chris



-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker@noaa.gov