CoBrALab / antsRegistration-MAGeT

A reimplementation of MAGeTbrain using only ANTs tools.
Other
16 stars 6 forks source link

output directory not created #90

Closed steelec closed 6 years ago

steelec commented 7 years ago

When running mb.sh -- multiatlas-resample multiatlas-vote (to perform atlas-template QC)

after the completion of the atlas-template registrations (135 of 135 atlas-template registrations completed) but before the completion of the run (i.e., jobs still in the que!)

the following error occurs: ls: cannot access output/multiatlas/labels/majorityvote: No such file or directory

full output from mb.sh -- multiatlas-resample multiatlas-vote 2017-09-17 13:55:31 UTC [ info] Found: 2017-09-17 13:55:31 UTC [ info] 5 atlases in input/atlas 2017-09-17 13:55:31 UTC [ info] 3 labels per atlas in input/atlas 2017-09-17 13:55:31 UTC [ info] 27 templates in input/template 2017-09-17 13:55:31 UTC [ info] 1428 subjects in input/subject 2017-09-17 13:55:31 UTC [ info] 0 models in input/models 2017-09-17 13:55:31 UTC [ info] Progress: 2017-09-17 13:55:31 UTC [ info] 135 of 135 atlas-template registrations completed 2017-09-17 13:55:40 UTC [ info] 159 of 38529 template-subject registrations completed 2017-09-17 13:55:50 UTC [ info] 0 of 578340 resample labels completed 2017-09-17 13:55:50 UTC [ info] 0 of 4284 voted labels completed 2017-09-17 13:55:51 UTC [ info] 0 of 405 multiatlas resample labels completed ls: cannot access output/multiatlas/labels/majorityvote: No such file or directory 2017-09-17 13:55:51 UTC [ info] 0 of 81 multiatlas voted labels completed 2017-09-17 13:55:51 UTC [ info] Computing Multiatlas/Template Label Resamples 45241991.gpc-sched-ib0 2017-09-17 13:55:55 UTC [ info] Computing Multiatlas/Template Votes 45241992.gpc-sched-ib0 2017-09-17 13:55:56 UTC [ info] Cleaning up. Done

gdevenyi commented 7 years ago

Hrm, so, this is weird.

If this was the first time multitlas commands were run, then output/multiatlas didn't exist and https://github.com/CobraLab/antsRegistration-MAGeT/blob/master/bin/mb.sh#L247-L251 shouldn't have run.

The error indicates that that code was run.

So, my question is, were any multiatlas commands run previously?

steelec commented 7 years ago

It is possible. The command may have been mistakenly run once outside of the scinet gpc (i.e, still on the login node). Chris

On Sep 19, 2017 9:52 AM, "Gabriel A. Devenyi" notifications@github.com wrote:

Hrm, so, this is weird.

If this was the first time multitlas commands were run, then output/multiatlas didn't exist and https://github.com/CobraLab/antsRegistration-MAGeT/blob/mast er/bin/mb.sh#L247-L251 shouldn't have run.

The error indicates that that code was run.

So, my question is, were any multiatlas commands run previously?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/CobraLab/antsRegistration-MAGeT/issues/90#issuecomment-330545819, or mute the thread https://github.com/notifications/unsubscribe-auth/ACm4wYC5tzD7tbIciVRkZDpVfIrTpqPTks5sj8cUgaJpZM4PaKBN .