Closed valosekj closed 1 year ago
CSV files in the r20230221 release have a different number of lines, e.g.:
$ cd data_normalize_metric_pam50 $ for file in sub-*;do echo $file;cat $file | wc -l;done sub-amu01_T2w_PAM50.csv 364 sub-amu02_T2w_PAM50.csv 413 ...
This is caused by omitting the -vert 1:20 flag for the sct_process_segmentation command:
-vert 1:20
sct_process_segmentation
sct_process_segmentation -i ${file_t2_seg}.nii.gz -vertfile ${file_t2_seg}_labeled.nii.gz -perslice 1 -normalize-PAM50 1 -o ${PATH_RESULTS}/${file_t2/_RPI_r/}_PAM50.csv
(source)
The different number of lines consequently causes an error in sct_compute_compression.py script during averaging of all CSV files (line 298).
sct_compute_compression.py
Usage of the -v 2 flag preserves the same number of lines across subjects; see here.
-v 2
Implemented in https://github.com/sct-pipeline/dcm-metric-normalization/commit/cfce6bb845f6551fc6ee214726d80bbc067d0571.
CSV files in the r20230221 release have a different number of lines, e.g.:
This is caused by omitting the
-vert 1:20
flag for thesct_process_segmentation
command:(source)
The different number of lines consequently causes an error in
sct_compute_compression.py
script during averaging of all CSV files (line 298).