[SciPy-user] ndimage.convolve behaviour?

David Warde-Farley dwf@cs.toronto....
Mon Jun 29 04:36:16 CDT 2009


On 23-Jun-09, at 3:38 PM, Robert Kern wrote:

>> Ah, thanks. Is there some reason/convention for this that I'm unaware
>> of?
>
> Yes. Convolution flips the filter around and correlation doesn't.
> That's the convention. :-)

So the only difference is which direction the axes are read? Odd. Is  
this a convention in other packages too?

>> Should this be documented somewhere?
>
> Oh, probably.

I've added reciprocal 'See Also' listings but I'm unsure how to word  
it to imply that "this is the same function with the stride/axes  
direction reversed", without it sounding a bit hokey.

David


More information about the SciPy-user mailing list