Where is the warp api of python but not the 'fast_warp' of cython?

[Joevaen]

Description

The 'fast_warps' api is here. It is based on Cython.I want to know where is the old python 'warp' api?

Is it here?

Way to reproduce

# Place the full code we need to recreate your issue here
# upload all necessary images to github too!

Version information

# Paste the output of the following python commands
from __future__ import print_function
import sys; print(sys.version)
import platform; print(platform.platform())
import skimage; print("scikit-image version: {}".format(skimage.__version__))
import numpy; print("numpy version: {}".format(numpy.__version__))

# your output here
3.7.6 (default, Jan  8 2020, 19:59:22) 
[GCC 7.3.0]
Linux-5.4.0-65-generic-x86_64-with-debian-buster-sid
scikit-image version: 0.17.2
numpy version: 1.19.2

[jni]

@Joevaen I don't really understand your question? Here is our documentation for skimage.transform.warp, which is what I think you're after:

https://scikit-image.org/docs/stable/api/skimage.transform.html#skimage.transform.warp

It hasn't changed places in about a decade so I don't know where the confusion is! 😅

I'm going to close this issue as I don't think it's a code issue, but please feel free to continue the discussion if you need further clarification!

[Joevaen]

Sorry about not describe my question clearly.
I notice there is two file about warp in scikit-image/skimage/transform/. Usually, some projects like to using the _warps_cy.pyx because it is based on Cython and it is faster. Am I right?
But it is not enough quickly for me, so I want to try to modify the python 'warp' which doesn't use Cython. Maybe Cupy taking place of Numpy will be better option to speed it up. I just wanna try.
In the _warps_cy.pyx, there is api named _warp_fast like below:

Does it correspond to the api named warpin _warps.py like below:

if they have same input image, do they have same output?
Thank for your kindness!
Best,
Joevaen.

[jni]

Yes, under certain circumstances, they are the same. You can see that the _warp_fast Cython function is imported into Python in this line:

scikit-image/skimage/transform/_warps.py

Line 8 in be6e3dc

from ._warps_cy import _warp_fast

And then warp calls _warp_fast here:

scikit-image/skimage/transform/_warps.py

Lines 871 to 884 in be6e3dc

	if matrix is not None:
	matrix = matrix.astype(image.dtype)
	ctype = 'float32_t' if image.dtype == np.float32 else 'float64_t'
	if image.ndim == 2:
	warped = _warp_fast[ctype](image, matrix,
	output_shape=output_shape,
	order=order, mode=mode, cval=cval)
	elif image.ndim == 3:
	dims = []
	for dim in range(image.shape[2]):
	dims.append(_warp_fast[ctype](image[..., dim], matrix,
	output_shape=output_shape,
	order=order, mode=mode,
	cval=cval))

This is used when the transform is linear and the interpolation is of order 0, 1, or 3. If those conditions aren't met, then a slower path is taken.