[Numpy-discussion] Release of NumPy

[Alan G Isaac]

> On 17/04/2008, Alan G Isaac <aisaac at american.edu> wrote:
>>         - it retains more ndarray behavior


On Thu, 17 Apr 2008, Stéfan van der Walt apparently wrote:
> That is simply not true.  The patch I sent provides everything you 
> asked for without breaking any ndarray behaviour, whereas your 
> suggestion breaks x[0] == x[0,:]. 


The behavior I was referring to is:
iteration through the matrix would
yield 1d arrays (and all their
associated behavior).

Sorry that was not clear.

Again, please note that I have said
multiple times that if iteration
yields instead 1 by N matrices that
can be indexed with with a single index
("vectors") I will be happy.

What is more, Travis's question seems
to kill my proposal, since it seems to
imply that

        - the behavior of matrices and sparse matrices 
          should match
        - that x[0] should yield a sparse matrix if x is 
          a sparse matrix

I do not use sparse matrices, so I have no opinion on these 
point.

Cheers,
Alan

PS I think that it would be good behavior
for x.A[0][0] to match x[0].A[0]