I am a little out of the loop. Just to clarify, Marie is trying to create an average subject out of 400+ subjects that have been run through the freesurfer pipeline? I don't see a problem with that (except possibly memory issues). If the subjects are from a homogeneous population (whatever homogeneous means), I would even say the more subjects the better.
--Thomas
On Jun 7, 2010, at 2:43 PM, Nick Schmansky wrote:
Marie,
I fixed the memory leak that was causing your out-of-memory error when running mris_make_average_surface. A new mris_make_average_surface for Mac Intel Leopard is posted here:
ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/misc/macos-leopard-intel/
I'm still looking into the mri_average failure. I dont know what to say about the license error, as the code that handles the license hasnt changed since v2.0.
As an aside, I'm skeptical of the utility of a volume or surface composed of 400+ subjects. The volume will be too blurry to be of use, and the surface will have lost any group-specific patterns that might have existed. If retaining the unique morphometric properties of your subject set is important, you might want to consider using just 20 or so randomly chosen subjects. Perhaps Bruce or others can comment.
Nick
On Wed, 2010-06-02 at 18:53 +0200, Marie Schaer wrote:
Hi Nick,
Thank you for your response. I added the flag --debug, please find in attachment the log. I am working on a Mac Intel Quad-Core Intel Xeon, Processor 3 GHz, RAM 5GB. I also tried with another similar machine which 9GB, but it exited with the same error exactly at the same point. Even using the -no-aseg option it stops exactly at the same point. I also provide you a screenshot with the top command running, a few minutes before the error, and just after the error.
As for the license error, I just had this problem using the Freesurfer version 4.5, and you are right, it may be a connection problem because our data are on a server (mounted with afp). It is however strange that it exits always at exactly the same point... Anyway, no license problem with version 4.4 or 4.3.
Many thanks for your help,
Marie
On 1 juin 10, at 16:54, Nick Schmansky wrote:
can you try adding the flag --debug to make_average_volume? it will output a lot of stuff, but will pinpoint at which stage it is running out of memory. also, try running 'top' in another terminal to monitor the amount of memory it's consuming while running.
as for the license error, that is bizarre, i can't think of a reason for that, unless your subjects are located on an NFS drive (a network). we've seen lots of problems with NFS and the Mac OS, where the connection drops for a couple seconds.
n.
On Mon, 2010-05-31 at 22:39 +0200, Marie Schaer wrote:
Hi everyone,
I did not find a solution to my problem yet: how to launch make_average_subject / make_average_volume / make_average_surface on a large number of subjects on a Mac Intel? Did someone encounter this error before?
Many thanks in advance for your help,
Marie
On 29 mai 10, at 20:07, Marie Schaer wrote:
Hi Nick,
Sorry for disturbing again. As advised, I tried to launch separately make_average_volume and make_average_surface. Even with the --no- aseg option with make_average_volume I got the "cannot allocate memory" error. And with the make_average_surface I got the "license error" in the middle of one subject processing... Any idea?
Thank you very much,
Marie
On 28 mai 10, at 21:37, Nick Schmansky wrote:
try adding --no-aseg to the command line. that out-of-memory error seems to originate with creation of the average aseg.mgz file (hopefully you wont need it for your subject set, typically you dont need aseg.mgz since aseg stats are usually gathered from a set of aseg.stats files, not from an average aseg.mgz).
doug: maybe we should make --no-aseg the default?
n.
On Fri, 2010-05-28 at 21:31 +0200, Marie Schaer wrote: > Hello everybody, > > I have a problem running make_average_subject for a large number > of > subjects (>400) on a Mac Intel. I tried with different > Freesurfer > versions (all the log outputs are attached). Using the > freesurfer > version 4.3 and 4.4 the make_average_subject script always > stop at > the > 297th subject with the error "Cannot allocate memory". I already > tried > to uncomment the line "limit descriptors unlimited" or to set it > to > various values, as proposed in the mailing list, but without > success > (it always stop at the 297th subject). Then I tried with the > Freesurfer version 4.5 (therefore I had to recompute the > lh.inflated.H > and lh.inflated.K which are not automatically produced in > previous > versions), but with the 4.5 version I got a strange "license > error" > message in the middle of the script (and my licence is certainly > fine > as I can run any other script without problems). > > Does anyone have an idea? Many thanks in advance for your > precious > help, > > Marie > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer