You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am trying to understand how to use the scoring system, and to keep things simple I am trying to run the system to evaluate the ground-truth .trk files.
The steps I follow are:
export the path, i.e.: export PYTHONPATH=~/Desktop/ismrm_2015_tractography_challenge_scoring
Run the scoring script on the CST_Left.trk ground-truth file downloaded from http://www.tractometer.org/ismrm_2015_challenge/data ./scripts/score_tractogram.py ~/Downloads/ISMRM_2015_Tracto_challenge_ground_truth_bundles_TRK_v2/CST_left.trk scoring_data/ results/
However, I get the following error:
WARNING:root:The number of streamlines specified in header ([0]) does not match the actual number of streamlines contained in this file (7217). The latter will be used.
WARNING:root:The number of streamlines specified in header ([0]) does not match the actual number of streamlines contained in this file (7217). The latter will be used.
WARNING:root:The number of streamlines specified in header ([0]) does not match the actual number of streamlines contained in this file (7217). The latter will be used.
Need "vox_to_ras" field to get affine with which to convert points; this is present for headers >= version 2
Traceback (most recent call last):
File "./scripts/score_tractogram.py", line 175, in <module>
main()
File "./scripts/score_tractogram.py", line 168, in main
segments_dir, base_name, args.verbose)
File "/home/aritche/Desktop/ismrm_2015_tractography_challenge_scoring/challenge_scoring/metrics/scoring.py", line 154, in score_submission
full_strl = [s for s in streamlines_gen]
File "/home/aritche/Desktop/ismrm_2015_tractography_challenge_scoring/challenge_scoring/io/streamlines.py", line 94, in _get_tracts_over_grid
str(er))
ValueError:
------ ERROR ------
TrackVis header is malformed or incomplete.
Please make sure all fields are correctly set.
If I instead use the .tck files downloaded from the same website, I don't have any issues and the expected results are generated (f1 for everything is 0, except for CST_left which is 0.8294, valid bundles = 1, etc.).
What do I change in the .trk header files to fix this problem?
The text was updated successfully, but these errors were encountered:
Just to make sure: how did you setup your environment to run the scorer? I feel like this might be a discrepancy in behavior between python 2 and 3. @AntoineTheb is currently finalizing PR #6 to migrate and fully support Python 3. If that's the issue, the solution should come soon.
Hi,
I am trying to understand how to use the scoring system, and to keep things simple I am trying to run the system to evaluate the ground-truth .trk files.
The steps I follow are:
export PYTHONPATH=~/Desktop/ismrm_2015_tractography_challenge_scoring
./scripts/score_tractogram.py ~/Downloads/ISMRM_2015_Tracto_challenge_ground_truth_bundles_TRK_v2/CST_left.trk scoring_data/ results/
However, I get the following error:
If I instead use the .tck files downloaded from the same website, I don't have any issues and the expected results are generated (f1 for everything is 0, except for CST_left which is 0.8294, valid bundles = 1, etc.).
What do I change in the .trk header files to fix this problem?
The text was updated successfully, but these errors were encountered: