compose_ants_transforms change

[ntustison]

With the recent ANTsPy changes, this random transform generator example used to work but now throws an error when trying to compose and linear and deformable ants transform as illustrated as follows:

>>> import ants
>>> import antspynet
>>> 
>>> image = ants.image_read(ants.get_ants_data("r16"))
>>> 
>>> linear_transform = ants.create_ants_transform(transform_type=
>>>     "AffineTransform", precision='float', dimension=image.dimension)
>>> 
>>> displacement_field = ants.simulate_displacement_field(image,
>>>     field_type="bspline", number_of_random_points=1000,
>>>     sd_noise=10.0, enforce_stationary_boundary=True,
>>>     number_of_fitting_levels=4, mesh_size=1,
>>>     sd_smoothing=4.0)
>>> displacement_field_xfrm = ants.transform_from_displacement_field(displacement_field)
>>> 
>>> xfrm = ants.compose_ants_transforms([linear_transform, displacement_field_xfrm])

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/ntustison/miniconda3/envs/antsx/lib/python3.11/site-packages/ants/core/ants_transform.py", line 415, in compose_ants_transforms
    itk_composed_tx = libfn(tx_ptr_list, precision, dimension)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: composeTransformsF2(): incompatible function arguments. The following argument types are supported:
    1. composeTransformsF2(arg0: list[ants.lib.AntsTransformF22], arg1: str, arg2: int, /) -> ants.lib.AntsTransformF22

Invoked with types: list, str, int

Is this a change in functionality, a bug, or is my recollection incorrect? Note that it would seem that currently ants.compose_ants_transforms(...) only works with linear transforms.

[ncullen93]

Yes I see that the input argument type on the composeTransforms c++ function is more strictly defined now and it's causing that issue. I will try to fix it.

Although I'm not sure it worked before because everything gets casted to the same itk::Transform type in the old function also.

[ncullen93]

Ok I've got a working version in a PR right now.

[ntustison]

Awesome. Thanks.