[Numpy-discussion] Style for pad implementation in 'pad' namespace or functions under np.lib
Charles R Harris
Wed Mar 28 17:08:48 CDT 2012
On Wed, Mar 28, 2012 at 1:57 PM, Ralf Gommers
> On Wed, Mar 28, 2012 at 9:31 PM, Tim Cera <firstname.lastname@example.org> wrote:
>> I have been developing a set of pad functions to pad arrays in different
>> ways. Really close to having it accepted into numpy, but I want to revisit
>> an implementation issue that I have become worried about. Should these
>> functions be collected into a 'pad' namespace or put raw into np.lib?
>> Do I really care about this? Not really since it isn't a utility issue,
>> but I contend that it would look better to pull these functions into their
>> own namespace.
>> +1 one for a separate 'pad' or 'padding' namespace. Adding 10 functions
> to the main numpy namespace feels like too much.
I think there is also a question of using a prefix pad_xxx for the function
names as opposed to pad.xxx.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the NumPy-Discussion