DOC: BUG: typo in ndimage LowLevelCallable tutorial example

[mikofski]

Is there a typo in the ndimage LowLevelCallable tutorial example?

AFAICT the C-extension exports the Python function example.get_transform which is imported in the next section, but then it's passed as transform() to LowLevelCallable, shouldn't it be passed as get_transform()?

I'm working on implementing LowLevelCallable for cython_optimize, and I was looking at the tutorial examples for ndimage, but I'm having a hard time interpreting when the C methods are passed to the LowLevelCallable. If this is indeed a typo, I'm happy to send a PR to fix it, but just wanted to check that I wasn't misunderstanding the usage. Thanks!

Reproducing code example:

import ctypes
import numpy as np
from scipy import ndimage, LowLevelCallable

from example import get_transform  # <- this line never used?

shift = 0.5

user_data = ctypes.c_double(shift)
ptr = ctypes.cast(ctypes.pointer(user_data), ctypes.c_void_p)
callback = LowLevelCallable(transform(), ptr)  # <- shouldn't this be get_transform() ?
im = np.arange(12).reshape(4, 3).astype(np.float64)
print(ndimage.geometric_transform(im, callback))

Error message:

>>> callback = LowLevelCallable(transform(), ptr)
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-2-88e997e74745> in <module>()
----> 1 callback = LowLevelCallable(transform(), ptr)

NameError: name 'transform' is not defined

Scipy/Numpy/Python version information:

import sys, scipy, numpy; print(scipy.__version__, numpy.__version__, sys.version_info)
1.1.0 1.15.0 sys.version_info(major=3, minor=6, micro=6, releaselevel='final', serial=0)

[mikofski]

I confirmed that it is a typo, and that changing it to get_transform() returns the expected output:

>>> from scipy import ndimage

>>> def transform(output_coordinates, shift):
...     input_coordinates = output_coordinates[0] - shift, output_coordinates[1] - shift
...     return input_coordinates

>>> im = np.arange(12).reshape(4, 3).astype(np.float64)
>>> shift = 0.5
>>> print(ndimage.geometric_transform(im, transform, extra_arguments=(shift,)))
[[0.     0.     0.    ]
 [0.     1.3625 2.7375]
 [0.     4.8125 6.1875]
 [0.     8.2625 9.6375]]

while the original typo raises an exception above