[Numpy-tickets] [NumPy] #655: Library for padding arrays

[NumPy]

#655: Library for padding arrays
-------------------------+--------------------------------------------------
 Reporter:  timcera      |        Owner:  somebody
     Type:  enhancement  |       Status:  new     
 Priority:  low          |    Milestone:  1.2.0   
Component:  numpy.lib    |      Version:  none    
 Severity:  minor        |   Resolution:          
 Keywords:               |  
-------------------------+--------------------------------------------------
Comment (by timcera):

 Replying to [comment:2 peridot]:
 > The patch is actually a more-or-less self-contained python package. The
 best thing to do is probably to drop it in as numpy.pad (particularly as
 it has functions like numpy.pad.maximum). It includes tests and
 documentation (albeit not exaclty in numpy standard format) so it's really
 a packaging issue.

 Thank you for looking at 'pad'.  I followed the scipy/numpy standards (as
 near as I could make them out) from the time that I first started
 programming 'pad' with the intent of it being a drop in module.  I can
 update to the latest standards if that would ease addition to numpy.

 > The package makes it easy to pad vectors on one end or the other with an
 arbitrary number of values, with wrapped or reflected values, or
 mean/median/maximum/minimum values of the whole array or part of it.

 Actually, more than vectors, it can pad N-dimensional arrays.  It is
 useful for any FFT (1 or 2D), plus any convolution or digital filter.

-- 
Ticket URL: <http://scipy.org/scipy/numpy/ticket/655#comment:3>
NumPy <http://projects.scipy.org/scipy/numpy>
The fundamental package needed for scientific computing with Python.