Blows up with `KeyError: 'ses-02'` whenever that subject does not have T1w in any other session but one

Not yet sure if it is related to

since no details on error was provided there. Invocation based on a container from

for a sample participant, with mount points only under current PWD (for proper isolation etc)

yoh@typhon:/mnt/DATA/data/yoh/1076_spacetop$ mkdir -p derivatives scratch
yoh@typhon:/mnt/DATA/data/yoh/1076_spacetop$ singularity run -B $PWD:$PWD:ro -B $PWD/derivatives:$PWD/derivatives:rw -B $PWD/scratch:$PWD/scratch:rw --scratch $PWD/scratch --pwd $PWD --net --network none ../dsst-defacing-pipeline/dsst-defacing-pipeline-test.sing -p 0001 -- . derivatives/dss-defacing-1

and leading to

====================
Dataset Summary
====================
Total number of sessions with 'anat' directory in the dataset: 115
Sessions with 'anat' directory with at least one T1w scan: 115

Please find the mapping file in JSON format at /mnt/DATA/data/yoh/1076_spacetop/derivatives/dss-defacing-1/primary_to_others_mapping.json
and other helpful logs at /mnt/DATA/data/yoh/1076_spacetop/derivatives/dss-defacing-1/logs

Defacing in Serial, one at a time
Running @afni_refacer_run on sub-0001_ses-01_acq-MPRAGEXp3X08mm_T1w.nii.gz
Find command logs at /mnt/DATA/data/yoh/1076_spacetop/derivatives/dss-defacing-1/bids_defaced/sub-0001/ses-01/anat/defacing_pipeline.log
@afni_refacer_run command completed on sub-0001_ses-01_acq-MPRAGEXp3X08mm_T1w.nii.gz

Reorganizing ses-01 with defaced images into BIDS tree...

[]
Preparing for QC by visual inspection...

Traceback (most recent call last):
  File "/opt/dsst-defacing-pipeline/run.py", line 192, in <module>
    main()
  File "/opt/dsst-defacing-pipeline/run.py", line 132, in main
    missing_refacer_out = deface.deface_primary_scan(
                          ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/dsst-defacing-pipeline/deface.py", line 218, in deface_primary_scan
    primary_t1 = mapping_dict[subj_id][sess_id]['primary_t1']
                 ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^
KeyError: 'ses-02'

and the reason is that anatomical T1 is present only in session 1:

yoh@typhon:/mnt/DATA/data/yoh/1076_spacetop$ ls -l sub-0001/ses-*/anat
total 60
-rw-r--r-- 1 yoh yoh 56954 May 21 10:35 sub-0001_ses-01_acq-MPRAGEXp3X08mm_T1w.json
lrwxrwxrwx 1 yoh yoh   145 May 21 10:35 sub-0001_ses-01_acq-MPRAGEXp3X08mm_T1w.nii.gz -> ../../../.git/annex/objects/QZ/PK/MD5E-s19195514--f8c60f8bdbba3d715a0e23accb7255cd.nii.gz/MD5E-s19195514--f8c60f8bdbba3d715a0e23accb7255cd.nii.gz

yoh@typhon:/mnt/DATA/data/yoh/1076_spacetop$ ls -ld sub-0001/ses-*/anat
drwxr-xr-x 1 yoh yoh 176 May 21 10:35 sub-0001/ses-01/anat

how to mitigate?

nimh-dsst / dsst-defacing-pipeline

Blows up with `KeyError: 'ses-02'` whenever that subject does not have T1w in any other session but one #42

38

41