[Numpy-discussion] Tools / data structures for statistical analysis and related applications

Keith Goodman kwgoodman at gmail.com
Fri Jun 11 15:02:41 EDT 2010


On Fri, Jun 11, 2010 at 11:51 AM, Gael Varoquaux
<gael.varoquaux at normalesup.org> wrote:
> On Fri, Jun 11, 2010 at 12:57:37PM -0500, Bruce Southey wrote:
>>  2. Do we really need to build custom data structures (larry, pandas,
>>  tabular, etc.) or are structured ndarrays enough? (My conclusion is
>>  that we do need to, but others might disagree). If so, how much
>>  performance are we willing to trade for functionality?
>
> For this purpose, I like Fernando's data array:
> http://github.com/fperez/datarray
> A very simple subclass of ndarrays that answers my most-wanted feature in
> terms of richer data structures.
>
> But, on the other hand, I have never really used R's advanced features,
> so I may be missing something.
>
> My 2 cents,
>
> Gaël

I looks like datarray labels the axes. The data object in pandas and
la, on the other hand, label the elements along each axis.



More information about the NumPy-Discussion mailing list