Hi Steve,
obviously I'm not the first with RAM/SwapSpace problems in tbss_skeleton.
My actual "all_FA" has t=15, about 57MB, and I guess the large required
memory (>>1GB) is due to converting INT16->FLOAT/DOUBLE for calculation
precision.
1) FSL(v3.3) installation on Linux (2.4.20-SMP, SUSE 8.2) - 1GB RAM, 1GB Swap
tbss_skeleton is aborted (segmentation fault) when RAM limit is reached; the
Swap is not touched at all (but it is e.g. with avwmaths++, so I don't think
there's a problem with my system)
2) FSL(v3.3) installed on Cygwin (WinXP) - 1GB RAM, 2GB virt. memory
tbss_skeleton is NOT aborted - ;-)) - but it's running using all memory
space just below 3GB, about 30 min.
However, I was thinking if it might be possible to split all data sets in
'all_FA' and run 'tbss_skeleton' on the splitted images separately. At least
to my understanding there should be no dependence or reference among the
single data sets in the calculation process
"tbss_skeleton -i mean_FA -p 2000 mean_FA_skeleton_mask_dst lower-cingulum
all_FA all_FA_skeletonised"
Am I right? Or did I miss an important fact in your TBSS paper?
Thanks a lot for a short comment!
Kind regards,
Jochen
--------------------------------------
Dept. of Neuroradiology
University Hospital Basel, Switzerland
On Sun, 28 Jan 2007 05:00:32 +0000, Steve Smith <[log in to unmask]> wrote:
>It may be, or it may be that you're just running our of RAM or swap,
>in which case this would also depend on what other programs were
>running at the time.
>Steve
>
>
>On 27 Jan 2007, at 16:37, Tal Herbsman wrote:
>
>> On Wed, 24 Jan 2007 15:35:21 +0000, Stephen Smith
>> <[log in to unmask]> wrote:
>>
>>> Hi - yes most likely you are running out of RAM or swap space, or the
>> number of subjects is just too
>>> big for a 32-bit machine (if that's what you're on).
>>>
>>
>> Is 30 subjects too big? 29 seems to work just fine.
>>
>> Tal.
>>
>>
>>
>>> Cheers, Steve.
>>>
>>>
>>>
>>> On Wed, 24 Jan 2007 02:57:39 +0000, Tal Herbsman
>>> <[log in to unmask]> wrote:
>>>
>>>> Hi.
>>>> My tbss_4_prestats pass fails when tbss_skeleton aborts with no
>>>> explanation
>>>> after a minute or so. It worked on another data set only last
>>>> week so I am
>>>> at a loss as to why it's failing now.
>>>>
>>>> Anybody have any idea what's happpening or how I can debug it?
>>>>
>>>> Tal.
>>>> =============================================================
>>> ===========
>>> =====================================================================
>>> ===
>
>
>------------------------------------------------------------------------
>---
>Stephen M. Smith, Professor of Biomedical Engineering
>Associate Director, Oxford University FMRIB Centre
>
>FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
>+44 (0) 1865 222726 (fax 222717)
>[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>------------------------------------------------------------------------
>---
>=========================================================================
|