TraitError with 'transform0DerivedInitialMovingTranslation.mat' not found

[sgiavasis]

Hi guys,

I'm running antsRegistration on the latest Nipype (Master branch, commit 4d855cb) using the parameter initial-moving-transform-com, set to 0:

antsRegistration --collapse-linear-transforms-to-fixed-image-header 0 --collapse-output-transforms 0 --dimensionality 3 --initial-moving-transform [ {reference}, {input}, 0 ] .... etc.

In the node's folder in the working directory, I'm getting every warp file I would get if I ran the same command from terminal:
transform0DerivedInitialMovingTranslation.mat
transform1Rigid.mat
transform2Affine.mat
transform3InverseWarp.nii.gz
transform3Warp.nii.gz

However, upon completion of the command, Nipype throws this error:

TraitError: Each element of the 'forward_transforms' trait of a RegistrationOutputSpec instance must be an existing file name, but a value of 'transform0DerivedInitialMovingTranslation.mat' <type 'str'> was specified.

Any thoughts? Thanks!

[chrisgorgo]

@oesteban could you have a look?

[oesteban]

@sgiavasis could you handle over me the 'command.txt' output or the output of the cmdline of the interface?

from nipype.interfaces.ants import Registration
reg = Registration()
reg.inputs.fixed_image = ['fixed1.nii', 'fixed2.nii']
reg.inputs.moving_image = ['moving1.nii', 'moving2.nii']
reg.inputs.output_transform_prefix = "output_"
...
reg.cmdline

[oesteban]

This is potentially related to #883

[sgiavasis]

Hi oesteban,

Thanks for getting back to me. Here's the command.txt:

antsRegistration --collapse-linear-transforms-to-fixed-image-header 0 --collapse-output-transforms 0 --dimensionality 3 --initial-moving-transform [ {reference}, {input}, 0 ] --interpolation Linear --output [ transform, transform_Warped.nii.gz ] --transform Rigid[ 0.1 ] --metric MI[ {reference}, {input}, 1, 32, Regular, 0.25 ] --convergence [ 1000x500x250x100, 1e-08, 10 ] --smoothing-sigmas 3.0x2.0x1.0x0.0 --shrink-factors 8x4x2x1 --use-histogram-matching 1 --transform Affine[ 0.1 ] --metric MI[ {reference}, {input}, 1, 32, Regular, 0.25 ] --convergence [ 1000x500x250x100, 1e-08, 10 ] --smoothing-sigmas 3.0x2.0x1.0x0.0 --shrink-factors 8x4x2x1 --use-histogram-matching 1 --transform SyN[ 0.1, 3.0, 0.0 ] --metric CC[ {reference}, {input}, 1, 4 ] --convergence [ 100x100x70x20, 1e-09, 15 ] --smoothing-sigmas 3.0x2.0x1.0x0.0 --shrink-factors 6x4x2x1 --use-histogram-matching 1 --winsorize-image-intensities [ 0.01, 0.99 ]

Is that what you were asking for?

[oesteban]

Thanks @sgiavasis, that's great.

[chrisgorgo]

I unfortunately don't have access to ANTS at the moment, but we need to get this fix before releasing. @sgiavasis is there anything you guys would like to merge before?

[sgiavasis]

Let me check, but I think Nipype now encompasses any interfaces I hacked in months ago - I'm testing the newer dev of CPAC with the current Nipype commit and I haven't had any troubles yet. When are you releasing?

[chrisgorgo]

Last Thursday ;)

On Mon, Oct 13, 2014 at 2:39 PM, sgiavasis notifications@github.com wrote:

Let me check, but I think Nipype now encompasses any interfaces I hacked
in months ago - I'm testing the newer dev of CPAC with the current Nipype
commit and I haven't had any troubles yet. When are you releasing?

—
Reply to this email directly or view it on GitHub
#944 (comment).

[sgiavasis]

That will be here in no time!

[chrisgorgo]

Not unless we use DHL...

On Mon, Oct 13, 2014 at 2:41 PM, sgiavasis notifications@github.com wrote:

That will be here in no time!

—
Reply to this email directly or view it on GitHub
#944 (comment).

[oesteban]

@sgiavasis @chrisfilo I'm currently looking at this

[oesteban]

Ok, seems like I forgot to add an os.path.abspath for that output, I'm currently submitting the fix.

@sgiavasis could you please give it a try?