ndimage.shift produces incorrect output (Trac #1520)

[scipy-gitbot]

Original ticket http://projects.scipy.org/scipy/ticket/1520 on 2011-09-15 by @stefanv, assigned to unknown.

def shift(input, shift):
   """
   Shift an array like  scipy.ndimage.interpolation.shift(input, shift, mode="wrap", order=0) but faster
   @param in: 2d numpy array
   @param d: 2-tuple of integers=20
   @return: shifted image
   """
   re =  numpy.zeros_like(input)
   s0, s1 =  input.shape
   d0 =  shift[0] % s0
   d1 =  shift[1] % s1
   r0 =  (-d0) % s0
   r1 =  (-d1) % s1
   re[d0:, d1:] =  input[:r0, :r1]
   re[:d0, d1:] =  input[r0:, :r1]
   re[d0:, :d1] =  input[:r0, r1:]
   re[:d0, :d1] =  input[r0:, r1:]
   return re

a=np.arange(12).reshape((3,4))
print "Input:\n", a, "\n"
print "Ndimage:\n", interpolation.shift(a, (1, 1), mode="wrap", order=0), "\n"
print "Custom:\n", shift(a, (1, 1)), "\n"

Output:

Input:
[[ 0  1  2  3]
 [ 4  5  6  7]
 [ 8  9 10 11]] 

Ndimage:
[[6 4 5 6]
 [2 0 1 2]
 [6 4 5 6]] 

Custom:
[[11  8  9 10]
 [ 3  0  1  2]
 [ 7  4  5  6]] 

[scipy-gitbot]

trac user jjhelmus wrote on 2012-05-08

This ticket identies the same problem as Ticket gh-1323.

[rgommers]

Added example to gh-1323, closing this as duplicate.