Topup output not found

[arnaudbore]

Dear nipype experts,

When I run topup it looks for a ".nii.gz" file as output 'out_fieldcoef'.
Because all my FSL settings are in ".nii" file it doesn't find anything and then crashes.

Thank you for your help.
Arnaud

[satra]

this is being fixed here:

#794

[arnaudbore]

Thank you Satra. Can you tell me when is going to be in the new version of nipype ? Maybe 0.9.3 ?

[satra]

definitely in 0.9.3 - it should be in master soon. we'll see if we can release 0.9.3 by end of february. there are a few other things that need to happen as well.

[oesteban]

The output 'out_fieldcoef' is assigned here: https://github.com/nipy/nipype/blob/master/nipype/interfaces/fsl/epi.py#L222 using the fsl.base method to generate filenames, as the argument change_ext is not overridden, it should observe the environment settings.

Is this happening to you with other interfaces of FSL?

[oesteban]

@arnaudbore Did you find out the problem?

[satra]

@oesteban - i think this was fixed in #794

@arnaudbore - did you check?

[rfdougherty]

I don't think it's been fixed. I just pulled master and still get a FileNotFoundError because it's looking for the fieldcoef file in the current directory rather than the directory specified in out_base (which is where the fieldcoef file actually is).

[oesteban]

That's a bright hypothesis, I did not think of out_base. I'm currently looking at this. Thanks @rfdougherty .

EDIT: ok, yes, a fast doctest writing out_base shows that the outputs are not updated.

[oesteban]

@satra this is the diagnostic:

we have two input traits that should name_source from the same trait (out_base) that should be sourced itself from in_file. This dependency is broken by the sorted return of traits within _parse_inputs.

I've thought of two solutions:

1. Review this sorted iteration of traits, so that dependencies are identified and the sources set in front of the list.

2. Modify _filename_from_souce to be recursive. This is probably a faster fix, but also produces risky redundancies in parsing.

[oesteban]

Submitted a fix, using option 2). If inputs are not changed during parsing, it should work out.

[satra]

@rfdougherty and @oesteban - could you take a look at this notebook before we make any changes?

http://nbviewer.ipython.org/urls/dl.dropbox.com/s/l84fgl333fperqb/Eddy.ipynb?dl=0

this workflow ran fine for me. i'll run it again to make sure things haven't changed.

[rfdougherty]

Hi Satra,

Everything does work fine if cwd == out_base. Reading through your workflow, it seems that's how you run it. E.g., " merged_file = os . path . join ( os . getcwd (), 'merged.nii.gz' )". But it will fail if out_base != cwd.

cheers,
bob

Robert F. Dougherty, PhD
Research Director
Stanford Center for Cognitive and Neurobiological Imaging
70 Jordan Hall * Stanford CA 94305 * 650-725-0051
http://www.stanford.edu/~bobd

----- Original Message -----

From: "Satrajit Ghosh" notifications@github.com
To: "nipy/nipype" nipype@noreply.github.com
Cc: "Bob Dougherty" bobd@stanford.edu
Sent: Sunday, September 21, 2014 6:20:24 AM
Subject: Re: [nipype] Topup output not found (#796)

@rfdougherty and @oesteban - could you take a look at this notebook before we make any changes?

http://nbviewer.ipython.org/urls/dl.dropbox.com/s/l84fgl333fperqb/Eddy.ipynb?dl=0

this workflow ran fine for me. i'll run it again to make sure things haven't changed.

—
Reply to this email directly or view it on GitHub .