numpy-discussion

[Numpy-discussion] slicing suggestion

[JJ <jos...@ya...>]

Hello.
I have a suggestion that might make slicing using
matrices more user-friendly.  I often have a matrix of
row or column numbers that I wish to use as a slice. 
If K was a matrix of row numbers (nx1) and M was a nxm
matrix, then I would use ans = M[K.A.ravel(),:] to
obtain the matrix I want.  It turns out that I use
.A.ravel() quite a lot in my code, as I usually work
with matrices rather than arrays.  My suggestion is to
create a new attribute, such as .AR, so that the
following could be used: M[K.AR,:].  I believe this
would be more concise, easier to read, and well used. 
If slices are made in both directions of the matrix,
then the .A.ravel() becomes even more unwieldy.  Does
anyone else like this idea?

John     

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [Numpy-discussion] slicing suggestion

[Keith G. <kwg...@gm...>]

On 10/20/06, JJ <jos...@ya...> wrote:
> My suggestion is to
> create a new attribute, such as .AR, so that the
> following could be used: M[K.AR,:]

It would be even better if M[K,:] worked. Would such a patch be
accepted? (Not that I know how to make it.)

Re: [Numpy-discussion] slicing suggestion

[Travis O. <oli...@ie...>]

JJ wrote:
> Hello.
> I have a suggestion that might make slicing using
> matrices more user-friendly.  I often have a matrix of
> row or column numbers that I wish to use as a slice. 
> If K was a matrix of row numbers (nx1) and M was a nxm
> matrix, then I would use ans = M[K.A.ravel(),:] to
>   
I had thought that something like that would be useful.  It's called

.A1

-Travis

Re: [Numpy-discussion] slicing suggestion

[Travis O. <oli...@ie...>]

Keith Goodman wrote:
> On 10/20/06, JJ <jos...@ya...> wrote:
>   
>> My suggestion is to
>> create a new attribute, such as .AR, so that the
>> following could be used: M[K.AR,:]
>>     
>
> It would be even better if M[K,:] worked. Would such a patch be
> accepted? (Not that I know how to make it.)
>   
What exactly do you want to work?

-Travis

Re: [Numpy-discussion] slicing suggestion

[Keith G. <kwg...@gm...>]

On 10/23/06, Travis Oliphant <oli...@ie...> wrote:
> Keith Goodman wrote:
> > On 10/20/06, JJ <jos...@ya...> wrote:
> >
> >> My suggestion is to
> >> create a new attribute, such as .AR, so that the
> >> following could be used: M[K.AR,:]
> >>
> >
> > It would be even better if M[K,:] worked. Would such a patch be
> > accepted? (Not that I know how to make it.)
> >
> What exactly do you want to work?

x
matrix([[ 0,  1,  2,  3],
        [ 4,  5,  6,  7],
        [ 8,  9, 10, 11],
        [12, 13, 14, 15]])

idx
matrix([[1],
        [3]])

I'd like (if it doesn't break the consistency of numpy):

--------------------------------------

x[idx, :] to give

matrix([[ 4,  5, 6 , 7],
              12, 13, 14 ,15]])

instead of

matrix([[[ 4,  5,  6,  7]],

        [[12, 13, 14, 15]]])

--------------------------------------

x[:, idx] top give a 4x2 matrix instead of an error

--------------------------------------

x[x[:,0] > 4, :] to give a 2x4 matrix instead of a 1x2 matrix

--------------------------------------

I'd also like a pony.