Hi - yes, the main look over "sessions" (ie all first-level analyses)
starts in fsl/tcl/feat.tcl around line 5153:
for { set session 1 } { $session <= $fmri(multiple) } { incr session 1 } {
however, it's not obvious to me how you would go about parallelising this
tcl loop. I think you would find it a _lot_ easier to write a script that
takes the input design.fsf file and uses that to acheive what you want.
1. find all the feat_files() lines
2. create n new design.fsf files
3. foreach design.fsf file, at the bottom;
set fmri(multiple) 1
set feat_files(1) [to one of the feat_files() values from the list]
then call this with "feat designN.fsf" on the appropriate processor
The above should be trivial, hopefully will do what you want.
Cheers, Steve.
On Tue, 12 Aug 2003, Alex Korb wrote:
> So I am trying to get FEAT analysis to run faster with
> multiple processors (I read the posts of a few months
> ago), and I figured an easy first step would be to
> first split up the analyses of different subjects into
> different processess. That is, if I want to do a
> first level analysis on ten subjects, then I want to
> run all 10 subjects at the same time on different
> processors. Granted I could just set up 10 individual
> first level analyses but that would take more time. I
> presumed the key to this would be in Feat.tcl, and
> that there would be some FOR loop that would serially
> go through all the subjects and run sequentially
> McFLIRT, BET etc, for each subject. However, I
> couldn't find such a FOR loop, though it must exist
> somewhere, and I was hoping someone could point me to
> it. Thanks, Alex Korb
>
Stephen M. Smith MA DPhil CEng MIEE
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
|