 |
 |
Hi, one last e-mail...I guess I found the difference. Turns out that the original fsf file also states "set fmri(ncopeinputs) 2", so it is the 'GUI-version' which does something different by correcting this to the amount of COPES it actually finds (and this is of course what I saw and why I thought the command-line version did something odd). So all I need to do is adjust the template.fsf and everything is fine. My apologies for all the confusion! Best, Johannes 2014-04-14 12:13 GMT+02:00 Johannes Klene <[log in to unmask]>: > Hi Stephen, > When I look at the design.fsf files in the output folder they are indeed > different on line 274 (thought you meant the 'input'.fsf file earlier). In > the design.fsf from the GUI version it says correctly "set > fmri(ncopeinputs) 8", whereas the command-line version says "set > fmri(ncopeinputs) 2". So despite the fact that I'm loading and subsequently > running the same input .fsf file, the GUI and the command line run produce > different design.fsf and consequently produce different outputs. For some > reason, the command line version changes the ncopeinputs argument. I've > attached the original 'input'.fsf file for a random subject if that helps > in solving this dilemma. > Thanks a lot for your suggestions. Best, Johannes > > > > 2014-04-14 11:50 GMT+02:00 Stephen Smith <[log in to unmask]>: > > Hi - as I said recently on the list - there is no difference between >> running from the GUI and running feat from the command line - in the sense >> that in both cases feat is run from the command line in the end. So the >> place to start is to compare the design.fsf files inside the output FEAT >> folder. >> >> Cheers. >> >> >> >> On 14 Apr 2014, at 10:47, Johannes Klene <[log in to unmask]> wrote: >> >> Hi, >> sorry little update. Forgot to check whether it also happens when I just >> run a single feat from the command line (i.e. feat design.fsf). Then the >> same thing happens (again without any error)! So it doesn't have anything >> to do with batching, but rather feat command line vs feat GUI. Any >> suggestions? >> Johannes >> >> >> 2014-04-14 11:02 GMT+02:00 Johannes Klene <[log in to unmask]>: >> >>> Hi Stephen, >>> thanks for the suggestions. However, the error logs are blank as if >>> everything ran correctly and the design.fsf's are exactly the same; when I >>> load the .fsf file manually into the GUI and press 'go' everything works >>> fine, but when I run the exact same fsf file as part of a for-loop (nothing >>> special, literally 'for each subject, do feat subject.fsf, done') only the >>> first two copes are outputted... >>> Best, Johannes >>> >>> >>> >>> 2014-04-14 10:54 GMT+02:00 Stephen Smith <[log in to unmask]>: >>> >>> Hi - you should check in all error logs, and also compare in detail the >>>> design.fsf files in the various cases, to see what is not being setup >>>> correctly in your batch-script runs. >>>> Cheers. >>>> >>>> >>>> >>>> On 14 Apr 2014, at 09:51, Johannes Klene <[log in to unmask]> wrote: >>>> >>>> Dear FSL experts, >>>> I'm having an odd problem with batching my second-level FEAT analyses, >>>> where I average the four sessions per subject for eight contrasts. When I >>>> run feat through the GUI everything goes fine, and I get eight copes >>>> outputted every time (tried a bunch of different subjects, all fine). >>>> However, when I use either a for-loop or submit the feat *.fsf to our >>>> computing cluster (necessary, as N>400), for everyone only the first two >>>> copes are outputted. Do you have any idea what might be going on here? >>>> Best, Johannes. >>>> >>>> >>>> >>>> >>>> --------------------------------------------------------------------------- >>>> 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 >>>> >>>> --------------------------------------------------------------------------- >>>> >>>> Stop the cultural destruction of Tibet <http://smithinks.net/> >>>> >>>> >>>> >>>> >>>> >>> >> >> >> >> --------------------------------------------------------------------------- >> 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 >> >> --------------------------------------------------------------------------- >> >> Stop the cultural destruction of Tibet <http://smithinks.net> >> >> >> >> >> >