[Numpy-discussion] First impressions on migrating to NumPy

[Francesc A. <fa...@ca...>]

Hi,

I'm starting to (slowly) replace numarray by NumPy at the core of PyTables,=
=20
specially at those places where the speed of NumPy is *much* better, that i=
s,=20
in the creation of arrays (there are places in PyTables where this is=20
critical, most specially in indexation) and in copying arrays. In both case=
s,=20
NumPy performs between 8x to 40x than numarray and this is, well...,=20
excellent :-)

Also, the big unification between numerical homogeneous  arrays, string=20
homogeneous arrays (with unicode support added) and heterogeneous arrays=20
(recarrays, with nested records support there also!) is simplyfying very mu=
ch=20
the code in PyTables where there are many places where one have to=20
distinguish between those different objects in numarray. Fortunately, this=
=20
distinction is not necessary anymore in many of this places.

=46urthermore, I'm seeing that most of the corner cases where numarray do w=
ell=20
(this was the main reason I was conservative about migrating anyway), are=20
also very well resolved in NumPy (in some cases better, as for one, NumPy h=
as=20
chosen NULL terminated strings for internal representation, instead of spac=
e=20
padding in numarray that gave me lots of headaches). Of course, there are=20
some glitches that I'll report appropriately, but overall, NumPy is behavin=
g=20
better than expected (and I already had *great* expectations).

Well, I just wanted to report these experiences just in case other people i=
s=20
pondering about migrating as well to NumPy. But also wanted to thanks (once=
=20
more), the excellent work of the NumPy crew, and specially Travis for their=
=20
first-class work.

Thanks!

=2D-=20
>0,0<   Francesc Altet =A0 =A0 http://www.carabos.com/
V   V   C=E1rabos Coop. V. =A0=A0Enjoy Data
 "-"