[Numpy-discussion] Cutting 1.0.2 release

[Gary Ruben]

Thanks Alan & Chris,

My apologies. I was trying ones(), zeros() and empty() in ipython 0.7.2 
with the -pylab option and getting the wrong functions. On my system, 
ipython -pylab imports the namespace of the oldnumeric wrapper versions 
of ones(), zeros() and empty() and I had assumed these were the core 
numpy versions. Maybe this has been changed in ipython 0.7.3.

Gary R.


Alan G Isaac wrote:
> On Tue, 30 Jan 2007, Gary Ruben apparently wrote: 
>> One question, which may be an issue: Should ones, zeros 
>> and empty be generating arrays of floats by default now? 
> 
>>>> import numpy as n
>>>> z = n.zeros((2,2))
>>>> z.dtype
> dtype('float64')
>>>> o = n.ones((2,2))
>>>> o.dtype
> dtype('float64')
>>>> e = n.empty((2,2))
>>>> e.dtype
> dtype('float64')
>>>> n.__version__
> '1.0'
> 
> hth,
> Alan Isaac