PS I ran bedpostx in all cases with "bedpostx subjectfolder" (I stripped down my scripts so they only contained the bedpost command). And when submitting the command as tcsh in a *.sh file, the log files get huge both when running on my local computer (Mac) and cluster (Linux). Bedpostx finished in all cases and the output looked good in all cases.
>> Hi Jon and Stam,
>> I have experienced the same, I get huge log files running bedpostx. The files that are huge are bedpostX/logs/log* (for each
>> slice I get a log of about 6GB, with 62 slices leading to ~372GB per subject). I have 4x256 directions for each subject and I
>> preprocessed the data with topup and eddy (don't know if that matters).
>> This happens only when using fsl 5.0.2, not fsl 4. Also, it only happens when I submit the bedpostx command as tcsh within
>> a *.sh file, but not when I submit the command as tcsh within a *.txt file. It happens also when I run bedpostx directly from
>> the command line but only if I run it on my local computer (Mac OS 10.8) but not on our department's computational cluster
>> (Linux).
>> Thanks,
>> Tanja
|