Closed peterN908 closed 3 years ago
Is it that the number of gradient directions on the test data is insufficient to support N_shm > 2?
Hi @peterN908 ,
Is it that the number of gradient directions on the test data is insufficient
Yes.
I included a table in our documentation describing the numbers. Tests are for quick checking only, so I released a data with only 13 diffusion-weighted gradients, which forces N_shm
to be 2.
When I run the ./pipeline_test.sh on the branch 'origin/spm-bspline' with N_shm = 6
Curious, how do you do that? I do not provide a way to set up N_shm
manually as part of ./pipeline_test.sh
.
Thanks for your help I hard coded in the n_shm because the config file could not be found (similar issue to https://github.com/pnlbwh/dMRIharmonization/issues/65)
the config file could not be found
That I can try to solve. So you tried ./pipeline_test.sh
as it is and it complained about the inexistence?
Yes exactly
I checked the /TMP/ folder and the config files being created certainly do exist
Thanks for reporting, I shall try to reproduce at my end.
With ./pipeline_test.sh
, the issue could not be reproduced. Regardless, feel free to switch to the master branch. That employs tempfile.gettempdir
method to write the configuration to a tempfile.
Hi there :)
When I run the ./pipeline_test.sh on the branch 'origin/spm-bspline' with N_shm = 6 I receive the following error:
file xxx/dMRIharmonization/lib/tests/connectom_prisma/template/Scale_L4.nii.gz does not exist .
Would you have any tips in debugging this?
FYI; if i re-run the pipeline_test.sh with N_shm = 2 I do not get this issue