Hi,
The difference between normal flirt and redhat9-flirt is
in the code optimisation of the cost function calculations.
They should give the same result, but we've noticed that
in some cases they do not, so the redhat9 version is a
simplified version with more checks in to make sure that
the compiler can't make any incorrect shortcuts.
If you can't get a particular subject to work repeatably,
then I suggest that recompiling with the native compiler is
the best solution. Once again, you may need some
modified sources to let you do that, and if so, please let
us know.
All the best,
Mark
On 11 Jul 2005, at 18:39, Helmar Waiczies wrote:
> both computer are brand new and have plenty of swap-space, the OS is
> working fine
> just that they give different reproducible results (s.b.)
> i repeated it a couple of times, always the same.
>
> > can u tell me what is the difference between the "normal" flirt and
> the
> > redhat9-flirt version ?
>
> Best Regards
> Helmar
>
> At 18:10 11.07.2005, you wrote:
>> Hi - if you are sometimes getting crashes and sometimes not, that
>> probably
>> either means that you are running out of swap (you can try increasing
>> your
>> swap) or that you have dodgy RAM or that your OS is corrupted - bugs
>> in
>> the programs themselves are almost always repeatable.
>>
>> Cheers, Steve.
>>
>>
>> On Thu, 23 Jun 2005, Dr. Helmar Waiczies wrote:
>>
>> > Hi Steve,
>> >
>> > thnx for your help!
>> >
>> > i installed FSL 3.2 a couple of weeks ago, so i had already the
>> newest
>> > flirt-version
>> > but i didnt know there is a special patched version for
>> Redhat9+SuSE9
>> > (works on Fedora Core3 as well)
>> >
>> > can u tell me what is the difference between the "normal" flirt and
>> the
>> > redhat9-flirt version ?
>> >
>> > one thing which i noticed is, when i do the same calculation on a
>> > Redhat9-System and on a FC3-System
>> > im getting different results (sometimes quite a big difference),
>> e.g.
>> > RH9: -1.67855; 0.1096; 0.2614; -0.1497; -0.4177; 0.52205; 0.4052
>> > FC3: -1.6527; 0.1084; 0.44285; -0.1376; -0.41855; 0.10985; -0.32775
>> >
>> > strange thing is, i tested three patients and only one of them shows
>> > different results
>> >
>> > I'll test this further and recalculate the whole dataset and let u
>> know
>> >
>> > thnx
>> > Helmar
>> >
>> >
>> >
>> > Am Mittwoch, den 22.06.2005, 11:35 +0100 schrieb Stephen Smith:
>> >
>> > > Hi - I think this is an old FLIRT bug - if you take the current
>> FSL
>> > > download or just the relevant patch file in
>> > > http://www.fmrib.ox.ac.uk/fsldownloads/patches/3.2/
>> > > then all should be fine.
>> > >
>> > > Cheers, Steve.
>> > >
>> > >
>> > >
>> > > On Sun, 19 Jun 2005, Helmar Waiczies wrote:
>> > >
>> > > > Hi,
>> > > >
>> > > > ive a little problem when using Siena
>> > > > somtimes im getting Segmentation Faults like the one shown below
>> > > >
>> > > > ---------- register brains and skulls
>> -------------------------------
>> > > > (do not worry about histogram warnings)
>> > > > /usr/local/fsl/bin/pairreg: line 92: 16476 Segmentation Fault
>> $flirt -in
>> > > > $bin -ref $bref -init $matf.3.$$ -schedule
>> > > > ${FSLDIR}/etc/flirtsch/pairreg3.sch -omat $matf $args
>> > > > or
>> > > > /usr/local/fsl/bin/pairreg: line 90: 1256 Segmentation Fault
>> $flirt -in
>> > > > $sin -ref $sref -init $matf.1.$$ -schedule
>> > > > ${FSLDIR}/etc/flirtsch/pairreg2.sch -omat $matf.2.$$ $args
>> > > > or
>> > > > /usr/local/fsl/bin/pairreg: line 89: 7710
>> Speicherzugriffsfehler $flirt
>> > > > -in $bin -ref $bref -schedule
>> ${FSLDIR}/etc/flirtsch/pairreg1.sch -omat
>> > > > $matf.1.$$ $args
>> > > >
>> > > > usually followed by
>> > > > Cannot read input-matrix
>> > > > Could not open matrix ....
>> > > >
>> > > > strange thing is, if i repeat the same calculation a couple of
>> times,
>> > > > sometimes it'll get
>> > > > through without error, but sometimes not ...
>> > > >
>> > > > im using two different computers (both 3GHz P4, 120GB harddisk)
>> > > > 1) Redhat9 and 2GB RAM
>> > > > 2) Fedora Core3 1GB RAM
>> > > >
>> > > > its the same on both computers
>> > > > :/
>> > > >
>> > > >
>> > > > Memory usage is usually around 60-70% (with 1GB RAM) so i dont
>> think im running
>> > > > out of memory.
>> > > >
>> > > > looking at the pairreg-file didnt help me either, since i dont
>> know what
>> > > > the problem might be
>> > > >
>> > > > any suggestions how to sort out this problem ?!
>> > > >
>> > > > thnx
>> > > > Helmar
>> > > >
>> > >
>> >
>>
>> --
>> Stephen M. Smith DPhil
>> Associate Director, FMRIB and Analysis Research Coordinator
>>
>> Oxford University Centre for Functional MRI of the Brain
>> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
>> +44 (0) 1865 222726 (fax 222717)
>>
>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>
> Der Horizont vieler Menschen ist ein Kreis mit dem Radius Null - und
> das nennen sie dann ihren Standpunkt.
|