Manualseg

[akhanf]

fixes to get manual seg import working again

[jordandekraker]

does this require a fresh install?

[akhanf]

No

…

On Thu., Feb. 11, 2021, 12:17 p.m. jordandekraker, ***@***.***> wrote: does this require a fresh install? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#34 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACXV2XILGVTFWJDVHEOMSU3S6QGLDANCNFSM4XPFEARQ> .

[akhanf]

@jordandekraker I reverted your two recent commits since they're not related to manual seg input..
It is def get_autotop_inputseg_cmd and rule run_autotop_inputseg that are invoked when manual segs are used, which does pass an image too..

Can you check to make sure your data is named as in the test dataset and you can do a dry run like below but on your data?

hippunfold test_data/bids_segT2w/ /tmp/test_out participant -np --modality segT2w

[jordandekraker]

I think this problem is solved now with 3668df2

[akhanf]

Sounds good I'll merge this in

[jordandekraker]

Unfortunately this seems to have broken again :/

Building DAG of jobs...
MissingInputException in line 38 of /scratch/jdekrake/HippUnfold/hippunfold/hippunfold/workflow/rules/autotop.smk:
Missing input files for rule run_autotop:

work/sub-001/seg_segT2w/sub-001_hemi-Lflip_space-corobl_desc-nnunet_dseg.nii.gz

Traceback (most recent call last):
  File "/scratch/jdekrake/HippUnfold/venv/bin/hippunfold", line 33, in <module>
    sys.exit(load_entry_point('hippunfold', 'console_scripts', 'hippunfold')())
  File "/scratch/jdekrake/HippUnfold/hippunfold/hippunfold/run.py", line 14, in main
    app.run_snakemake()
  File "/scratch/jdekrake/HippUnfold/venv/lib/python3.7/site-packages/snakebids/app.py", line 209, in run_snakemake
    run(snakemake_cmd)
  File "/scratch/jdekrake/HippUnfold/venv/lib/python3.7/site-packages/snakebids/app.py", line 41, in run
    raise Exception("Non zero return code: %d"%process.returncode)
Exception: Non zero return code: 1

When I use this with actual data, a /results directory is created but no /work directory.

If i remember correctly, manual segs would be put into the /work directory with the name sub-001_hemi-Lflip_space-corobl_desc-nnunet_dseg.nii.gz. Could perhaps be that because nnunet.smk is no longer being called, this file is not being recognized as a dependency for autotop.smk and therefore the preprocessing is also not recognized as a dependency?

[akhanf]

hmm let me take a look -- I set aside some time now for hippunfold stuff, so good timing!
Are you using manual segs naming like the bids_segT2w test dataset?

[akhanf]

just tried it is failing for that bids test dataset too -- looks like I never added that test case which is why it slipped it -- anyways will have a look to fix it now!

[jordandekraker]

sorry just realize that I didn't copy the command for the error above - it was the test case code. But this seems to be working now, thanks for taking a look and the quick fix!

[akhanf]

No worries! Was a good lesson for me to always make sure I add new workflow variants to the test case, as it is easy to break things when we add new features.. Also learned that you can easily compare things in github by just adding compare to the url.. Going to have a v0.4.0 ready soon as well, just putting some final changes in --

…

On Wed, Apr 21, 2021 at 9:59 AM jordandekraker ***@***.***> wrote: sorry just realize that I didn't copy the command for the error above - it was the test case code. But this seems to be working now, thanks for taking a look and the quick fix! — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub <#34 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACXV2XLQB67MH2IYO3XFL5DTJ3K2LANCNFSM4XPFEARQ> .

-- *Ali R. Khan, PhD* Canada Research Chair in Computational Neuroimaging Assistant Professor, Medical Biophysics Scientist, Robarts Research Institute Schulich School of Medicine & Dentistry, Western University e. ***@***.*** t. 519.931.5777 (x24280) w. www.khanlab.ca