Description
Summary
ANTSInputSpec has an incorrect trait type traits.Float, rather than traits.Int. I've tested this with ants v2.2.0, I believe the last for which ANTS was the expected registration call.
Actual behavior
My input parameters:
transformation_model = "SyN",
output_transform_prefix = "ANTS_OUT_",
dimension = 3,
metric = ["CC"],
metric_weight = [1.0],
number_of_iterations = [50, 50, 50, 50],
regularization = "Gauss",
regularization_gradient_field_sigma = 0,
regularization_deformation_field_sigma = 3,
gradient_step_length = 25,
number_of_time_steps = 3,
delta_time = 0.05,
radius = [5]
The command line generated was:
ANTS 3 --image-metric CC[ $fixed, $moving, 1, 5 ] --number-of-iterations 50x50x50x50 --output-naming ANTS_OUT_ --regularization Gauss[0.0,3.0] --transformation-model SyN[25.,3.0,0.050] --use-Histogram-Matching 1
After getting to the SyN transformation step, ANTS crashed with the following message:
Requested Transformation Model: SyN : Using
SyN diffeomorphic model for transformation.
Standard error:
Exception caught in ANTS:
/build/ants-x9F8a4/ants-2.2.0/Utilities/antsCommandLineParser.h:126:
itk::ERROR: CommandLineParser(0x5616707e9f80): ERROR: Parse error occured during command line argument processing
ERROR: Unable to convert '3.0' to type 'j' as unsigned int
Return code: 1
Expected behavior
ANTS conducts the SyN transformation with the following modified command:
ANTS 3 --image-metric CC[ $fixed, $moving, 1, 5 ] --number-of-iterations 50x50x50x50 --output-naming ANTS_OUT_ --regularization Gauss[0.0,3.0] --transformation-model SyN[25.,3,0.050] --use-Histogram-Matching 1
Platform details:
{'commit_hash': 'a0f756d1c',
'commit_source': 'installation',
'networkx_version': '1.11',
'nibabel_version': '2.5.1',
'nipype_version': '1.2.0',
'numpy_version': '1.16.5',
'pkg_path': '/usr/local/lib/python2.7/dist-packages/nipype',
'scipy_version': '1.2.2',
'sys_executable': '/usr/bin/python',
'sys_platform': 'linux2',
'sys_version': '2.7.17 (default, Nov 7 2019, 10:07:09) \n[GCC 7.4.0]',
'traits_version': '4.6.0'}
Execution environment
Run inside a docker container with Ubuntu 18.04 with the versions above.