[Gate-users] problem with gjm

Jan De Beenhouwer jan.debeenhouwer at ugent.be
Thu Mar 4 21:50:33 CET 2010


Hi,

The job merger shouldn't depend on the random engine you used. Maybe
something else is wrong though. If you can provide me with a '.split' file
and a few small root files to merge, then I will take a look at it.

All the best,

Jan

--
dr. ir. Jan De Beenhouwer
Ghent University - IBBT
IBITECH - MEDISIP
De Pintelaan 185
Block B - 502a
B-9000 Ghent
Tel. +32-9-332.02.58
Fax: +32-9-332.33.90
E-mail: <Jan.DeBeenhouwer at UGent.be>



> -----Original Message-----
> From: gate-users-bounces at lists.opengatecollaboration.org [mailto:gate-
> users-bounces at lists.opengatecollaboration.org] On Behalf Of Peter
> Herrick
> Sent: donderdag 4 maart 2010 19:58
> To: 'Simon Stute'
> Cc: gate-users at lists.opengatecollaboration.org
> Subject: Re: [Gate-users] problem with gjm
> 
> Hi all,
> 
> 
> 
> Sadly I meant with respect to a block of crystals in my detector – the
> #counts in the blockID  histogram (in root) seem to fluctuate by much
> greater than the anticipated amount à (#counts) ^ 0.5. Changing from
> the
> MersenneTwister Engine seemed to fix this problem  - examining an
> individual
> rootfile before merging them all seemed to indicate a fluctuation much
> closer to the anticipated amount, but the gjm gave the error.
> 
> Any other ideas?
> 
> 
> 
> Thanks,
> 
> Peter.
> 
> 
> 
> From: Simon Stute [mailto:gate.stute at gmail.com]
> Sent: 04 March 2010 15:29
> To: Peter Herrick
> Cc: gate-users at lists.opengatecollaboration.org
> Subject: Re: [Gate-users] problem with gjm
> 
> 
> 
> Hi Peter,
> 
> When you talk about fluctuations in the number of singles for each
> "block",
> do you mean split macros (in the meaning of a job) or do you physically
> mean
> a block in your detector ?
> In the first case, did you use the following command in your main macro
> 
> /gate/cluster/setTimeSplitHalflife
> 
> Because if you did not use it, each split is equally divided along the
> simulation time, so depending on the half-life of your isotope each
> split
> will contain a different number of singles.
> 
> Hope this helps !
> Simon
> 
> On Thu, Mar 4, 2010 at 3:40 PM, Peter Herrick <pdeh2 at cam.ac.uk> wrote:
> 
> Dear GATE users,
> 
> I have been using version 4.0  with linux in a cluster arrangement to
> simulate a PET scanner. I have found that the fluctuations in the
> number of
> singles events for each block are MUCH greater than those expected from
> poisson statistics - After looking at the seeds created when I employ
> gjs,
> the same numbers are appearing in each 'seed.rndm' file in different
> orders.
> To solve I have tried switching back to the HepJamesRandom engine, but
> as a
> result the Job merger is no longer working - I wondered if anyone has
> any
> ideas?
> 
> The change that I implemented was the one Simon suggests in the thread
> 'same
> running on Gate' which can be found below.
> 
> http://lists.opengatecollaboration.org/mailman/private/gate-users/2009-
> April
> /subject.html
> 
> but there is no mention of gjm not working. The error I recieve is:
> 
> 
>  *** Break *** segmentation violation
> (no debugging symbols found)
> Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1".
> Attaching to program: /proc/17367/exe, process 17367
> done.
> Failed to read a valid object file image from memory.
> 0xb7f95410 in ?? ()
> 
> which I have never seen before. Any help or suggestions would be much
> appreciated.
> 
> Thanks,
> 
> Peter Herrick,
> Wolfson Brain Imaging Centre
> University of Cambridge
> 
> _______________________________________________
> Gate-users mailing list
> Gate-users at lists.opengatecollaboration.org
> http://lists.opengatecollaboration.org/mailman/listinfo/gate-users
> 
> 
> 
> _______________________________________________
> Gate-users mailing list
> Gate-users at lists.opengatecollaboration.org
> http://lists.opengatecollaboration.org/mailman/listinfo/gate-users



More information about the Gate-users mailing list