morphologist 2023 won't open: AttributeError: 'NoneType' object has no attribute '_parameterized'

[HdeV999]

Dear BrainVISA team,
I was happily running batch processing scripts when something unadvisable happened: I exceeded the quotas of my home directory during the processing, and it affected my installation. The most obvious result was that my .brainvisa/options.minf was overwritten with a blank file, but I believe I solved that by reconfiguring my options.
A side effect is that I can't open the morphologist 2023 process anymore: I get a "AttributeError: 'NoneType' object has no attribute '_parameterized'" error when I try to open the morphologist 2023 window.
Incidentally, when I try to run the magic command bv_maker, I fail the source step: master (1e06678e) x merge failed U-49 axon/master
This should be unrelated but may still be relevant to try and solve the issue...
This is my environment specification:
opensource-master-5.3
type: dev
distro: opensource
branch: master
system: ubuntu-22.04
image_version: 5.3
container_type: singularity
image: /home/hd488/casa_distro/casa-dev-5.3-22.sif
directory: /home/hd488/casa_distro/brainvisa-opensource-master
Thanks for any help on the matter...

[HdeV999]

Update: I tried running a bv_maker configure build doc to restore my previous build even without being able to run the source step, it worked, but I'm still faced with the same morphologist 2023 issue, i.e. when I double click on the process I get an error window with
AttributeError: 'NoneType' object has no attribute '_parameterized'

[denisri]

Perhaps the disk full issue happened during a git operation.
Try going to src/axon/master
and there, do:

git reset --hard origin/master
git pull

then re-run bv_maker

[denisri]

So much longer, it's weird... The error or warning messages are "new" since we have replaced the way Anatomist renders offscreen, I have noticed similar messages but it seems to do the job anyway, and in my experiments this part has more or less the same performance as it used to, so I doubt this is the bottleneck. But yes an additional step has been added for the report. It's not that long (maybe 10 seconds per subject). We can try to unselect it just to see if it makes a difference. I'll send a modified process for Freesurfer importation when I have a minute, so that you can compare.

[HdeV999]

Thanks!

[denisri]

Hi,
Can you try this modified process ?
Save it in your $HOME/.brainvisa/processes/ directory (if it doesn't exist, just create it) and restart brainvisa. It should replace the original process.
It has an additional parameter, perform_morpho_report (the last one) that you can un-check.
https://brainvisa.info/download/tmp/Import_FROM_FreeSurfer_TO_Morpho.py

[HdeV999]

Thanks a lot! Trying it with my current pipeline, will let you know as soon as it runs (my cluster queues can be a bit long).

[HdeV999]

It works! Here is an extract from my log files:
progress: 91.66666666666666 % ...

Skip unselected node: morpho_report

progress: 100.0 % ...

Process Morphologist 2023 finished on 2024/08/20 23:16 (14 minutes 59
seconds)

OK

Process Import FreeSurfer grey/white segmentation to Morphologist
finished on 2024/08/20 23:16 (17 minutes 11 seconds)

It didn't bring my processing time back to 2 hours though, but this may be related to other changes in my configuration. Thanks a lot!