Hi
it seems that melodic is being killed by the OS and I suspect it to be
related to the amount of memory it requests - after calling smoothest
melodic needs to allocate new memory in order to calculate the adjusted
eigenspectrum ...
If you can easily provide a download link for me to grab the data I'm
happy to have a look on a machine with more memory. Also it'd be good to
know what happens when you monitor top while executing melodic
cheers
Christian
Rombouts, S.A.R.B. wrote:
> Hi Christian,
>
> I have tried your command /usr/local/bin/fsl/bin/smoothest -d 700 -r
> filtered_func_data
>
> >-m /fsl_analysis/700vol.ica/mask
>
> This is what I get:
>
> DLH 7.23278
> VOLUME 6064
> RESELS 0.638767
>
> Is this useful for you?
>
> Thanks, Serge.
>
>
>
>
> >-----Original Message-----
> >From: Christian Beckmann [mailto:[log in to unmask]]
> >Sent: Wednesday, October 16, 2002 2:30 PM
> >To: [log in to unmask]
> >Subject: Re: [FSL] An upper limit to the number of time points
> >using MELODIC?
> >
> >
> >Hi
> >
> >looks like smoothest isn't happy, can you please try
> >
> >/usr/local/bin/fsl/bin/smoothest -d 700 -r filtered_func_data
> >-m /fsl_analysis/700vol.ica/mask
> >
> >In cases like yours where the number of voxels is relatively
> >small compared to the number of time points you might even
> >want to switch off the automatic dim. estimation - I wouldn't
> >necessarily assume that it gives reliable estimates - you can
> >do this from the GUI or usingthe -d option from the command line
> >
> >ta
> >Christian
> >
> >
> >Rombouts, S.A.R.B. wrote:
> >
> >>Dear FSL experts,
> >>
> >>I am trying to analyze my fMRI data using melodic:
> >>Number of slices = 3
> >>Matrix = 64x64
> >>Number of volumes = 1024
> >>TR = 0.3 sec
> >>
> >>The system crashes after some calculations. The strange
> >
> >thing is that
> >
> >>I have no problems when I only use the first 100, 200, or
> >
> >500 volumes.
> >
> >>However, when I try to use melodic with 700 volumes or more,
> >
> >I get the
> >
> >>error:
> >>
> >>######
> >>Reading data file filtered_func_data ... done
> >>Create mask ... done
> >>Excluding voxels with constant value
> >>Data size : 700 x 6064
> >>Removing mean image ... done
> >>Estimating data covariance ... done
> >>Perform variance-normalisation ... done
> >>Removing mean time course ... done
> >>Starting PCA ... done
> >>Estimating the number of sources from the data (PPCA) ...
> >> Calling Smoothest: /usr/local/bin/fsl/bin/smoothest -d 700 -r
> >>filtered_func_data -m /fsl_analysis/700vol.ica/mask >
> >>/fsl_analysis/700vol.ica/smoothest
> >>
> >>child killed: SIGABRT
> >>#####
> >>
> >>I am sure that the amount of memory is not a problem on the system I
> >>use (RAM = 1 GB, swap = 2.5 GB) and all my volumes are ok.
> >>
> >>Is there possibly an upper limit to the number of volumes
> >
> >melodic can
> >
> >>handle?
> >>
> >>Thanks for any help,
> >>Serge.
> >>
> >>
> >>Serge Rombouts, PhD
> >>Functional MRI Laboratory
> >>Vrije Universiteit Medical Center
> >>Amsterdam, The Netherlands
> >>Tel. +31 20 4440316
> >>Fax +31 20 4444147
> >>_www.fmri.nl_
> >>
> >
> >
> >--
> > Christian F. Beckmann
> > Address: Oxford University Centre for Functional
> > Magnetic Resonance Imaging of the Brain,
> > John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> > Email: [log in to unmask]
> >http://www.fmrib.ox.ac.uk/~beckmann/
> > Phone: +44(0)1865
> >222782 Fax: +44(0)1865 222717 Mob: +44(0)7811 189123
> >
>
>
>
--
Christian F. Beckmann
Address: Oxford University Centre for Functional
Magnetic Resonance Imaging of the Brain,
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
Email: [log in to unmask] http://www.fmrib.ox.ac.uk/~beckmann/
Phone: +44(0)1865 222782 Fax: +44(0)1865 222717 Mob: +44(0)7811 189123
|