While testing for PR #1119 , I found a small bug with the ADNI to BIDS converter. Trying to run successively different modalities and save them in the same folder raises an error when creating the 'scans.tsv' file.
To Reproduce
Run successively the ADNI-2-BIDS converter on the same dataset for two different modalities and direct it to the same output folder. For example, running :
...
File "/Users/alice.joubert/clinica/clinica/iotools/converters/adni_to_bids/adni_to_bids_cli.py", line 79, in cli
adni_to_bids.convert_clinical_data(
File "/Users/alice.joubert/clinica/clinica/iotools/converters/adni_to_bids/adni_to_bids.py", line 155, in convert_clinical_data
adni_utils.create_adni_scans_files(conversion_path, bids_subjs_paths)
File "/Users/alice.joubert/clinica/clinica/iotools/converters/adni_to_bids/adni_utils.py", line 1124, in create_adni_scans_files
conversion_df = converted_dict[converted_mod]
~~~~~~~~~~~~~~^^^^^^^^^^^^^^^
KeyError: 't1'
And the 'scans.tsv' file is saved with only FLAIR - related lines in it.
Expected behavior
While it may not be the primary way the adni-to-bids converter is used, it would be great to take into account what was already done with the dataset.
While testing for PR #1119 , I found a small bug with the ADNI to BIDS converter. Trying to run successively different modalities and save them in the same folder raises an error when creating the 'scans.tsv' file.
To Reproduce Run successively the ADNI-2-BIDS converter on the same dataset for two different modalities and direct it to the same output folder. For example, running :
clinica convert adni-to-bids dir_raw_data dir_clinical_data dir_output_bids -m T1
clinica convert adni-to-bids dir_raw_data dir_clinical_data dir_output_bids -m FLAIR
Gives an error :
And the 'scans.tsv' file is saved with only FLAIR - related lines in it.
Expected behavior While it may not be the primary way the adni-to-bids converter is used, it would be great to take into account what was already done with the dataset.