Normalize12 does not accept an existing file as "apply_to_files" input

[fednem]

Summary

When I try to provide the path towards the file to whom apply the deformation field it throws an error message

Actual behavior

When I provide the path towards the file to whom apply the deformation field as in

nrm = Normalize12()
nrm.inputs.apply_to_files = "/path/to/file.nii"

it throws the following error

TraitError: Each element of the 'apply_to_files' trait of a Normalize12InputSpec instance must be a pathlike object or string representing an existing file or a list of items which are a pathlike object or string representing an existing file, but a value of '/media/zipat/DATA/temp/AD_75_T1.nii' <class 'str'> was specified.

Expected behavior

The expected behavior would be to accept the input, especially since it takes as deformation_field a file in the same directory and I have double checked that the files I am providing exist.

Script/Workflow details

https://gist.github.com/fednem/e4725d8812af75a51d681214da3b401f

Platform details:

/home/zipat/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
{'commit_hash': '4b69cfcfa',
 'commit_source': 'installation',
 'networkx_version': '2.1',
 'nibabel_version': '2.4.0',
 'nipype_version': '1.2.2',
 'numpy_version': '1.14.3',
 'pkg_path': '/home/zipat/anaconda3/lib/python3.6/site-packages/nipype',
 'scipy_version': '1.1.0',
 'sys_executable': '/home/zipat/anaconda3/bin/python',
 'sys_platform': 'linux',
 'sys_version': '3.6.5 |Anaconda, Inc.| (default, Apr 29 2018, 16:14:56) \n'
                '[GCC 7.2.0]',
 'traits_version': '5.1.2'}

[fednem]

After having restarted the python console and reinstalled nipype, it does accept the file as input, but it does not run, giving the following error

Traceback (most recent call last):

  File "<ipython-input-28-56bd4d3919aa>", line 10, in <module>
    nrm.run()

  File "/home/zipat/anaconda3/lib/python3.6/site-packages/nipype/interfaces/base/core.py", line 378, in run
    outputs = self.aggregate_outputs(runtime)

  File "/home/zipat/anaconda3/lib/python3.6/site-packages/nipype/interfaces/base/core.py", line 481, in aggregate_outputs
    raise FileNotFoundError(val, message=msg)

TypeError: __init__() got an unexpected keyword argument 'message'

Which seems to me to be a similar problem, since it is complaining about not finding a file that it is absolutely there.

[effigies]

For the FileNotFoundError, I've proposed a fix in #3035.

As to why the file isn't being found, what kind of environment are you running in? Is the process running on the same machine and with the same permissions as your user?

[effigies]

@fednem Can you confirm whether this persists with the current master?

[fednem]

I have just tried again and the issue seems to be resolved. Il giorno gio 19 set 2019 alle ore 22:22 Chris Markiewicz < notifications@github.com> ha scritto:

…

@fednem <https://github.com/fednem> Can you confirm whether this persists with the current master? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3030?email_source=notifications&email_token=AFH4JCH2DGKE62XR46Y5EW3QKPNSFA5CNFSM4IXCCCLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD7EW3WQ#issuecomment-533294554>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFH4JCCBTL2XGXFU4OYWRL3QKPNSFANCNFSM4IXCCCLA> .

-- Federico Nemmi, PhD

[effigies]

Great, thanks for the confirmation!

[axiezai]

Hi, I'm running into the same problem:

traits.trait_errors.TraitError: Each element of the 'T1_files' trait of a ReconAllInputSpec instance must be a pathlike object or string representing an existing file, but a value of '/dwi_preproc/derivatives/pipetography/_subject_id_01/apply_warp/mapflow/_apply_warp1/acpc_t1.nii.gz' <class 'str'> was specified.

I'm running nipype with it's fsl and freesurfer interfaces in singularity/docker image:

root@e1498729b2f7:/# nipypecli version
1.5.0-rc1

The crazy thing is, I built the workflow on my Mac, which does not spit out this error, but when I run this on a remote linux machine in the containers i build with the same version of nipype, I get this error. The file does exist. Any suggestions? Thank you in advance.

[fednem]

Hey ! Maybe a stupid suggestion: have you checked that a gunzipped nifti is a valid input to the function interface you're calling? Il giorno mer 27 mag 2020 alle ore 04:47 Xi He <notifications@github.com> ha scritto:

…

Hi, I'm running into the same problem: traits.trait_errors.TraitError: Each element of the 'T1_files' trait of a ReconAllInputSpec instance must be a pathlike object or string representing an existing file, but a value of '/dwi_preproc/derivatives/pipetography/_subject_id_01/apply_warp/mapflow/_apply_warp1/acpc_t1.nii.gz' <class 'str'> was specified. I'm running nipype with it's fsl and freesurfer interfaces in singularity/docker image: ***@***.***:/# nipypecli version 1.5.0-rc1 The crazy thing is, I built the workflow on my Mac, which does not spit out this error, but when I run this on a remote linux machine in the containers i build with the same version of nipype, I get this error. The file does exist. Any suggestions? Thank you in advance. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#3030 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFH4JCGABUK6LKYMQVUFLYDRTR5NDANCNFSM4IXCCCLA> .

-- Federico Nemmi, PhD

[axiezai]

That is indeed a good suggestion, I changed the previous step's output to be .nii instead of .nii.gz so recon-all get's a .nii for input, now the workflow is complaining that the input to the previous step is a str that points to a file that doesn't exist, but it does:

traits.trait_errors.TraitError: The 'in_file' trait of an ApplyWarpInputSpec instance must be a pathlike object or string representing an existing file, but a value of '/BIDS_output/sub-01/ses-002/anat/sub-01_ses-002_T1w.nii.gz' <class 'str'> was specified.

Just to be sure:

$ ls /BIDS_output/sub-01/ses-002/anat/
sub-01_ses-002_T1w.json  sub-01_ses-002_T1w.nii.gz

Additionally, another node in my workflow for mrtrix3's dwipreproc is also failing, i'm feeding in a .mif image as suggested by mrtrix, and getting the same error that the .mif cannot be found but it does exist.

[effigies]

Can you open a new issue?

[axiezai]

hi @effigies,

After reading through more issues and documentations, I was able to solve the issue by configuring my workflow with {'use_relative_paths':'True', 'hash_method': 'content'}.

SO I won't be opening a new issue. Thank you everyone for the help.